[
    {
        "paper_doi": "10.1186/s12936-021-03611-7",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design:  2-arm, open-label, parallel-arm, cRCTUnit of allocation: village. Villages not reaching the minimum population for inclusion were combined with the nearest neighbouring village to form a single cluster.Number of units: 168 clusters (84 IRS, 84 no IRS) were involved in the passive surveillance component of the study. 86 clusters (43 IRS, 43 no-IRS) participated in the active cohort component of the study.Outcome assessment/surveillance type: active surveillance of a cohort of children aged < 5 year (18 per cluster), passive surveillance of people of all ages through the national health system, and annual cross-sectional surveys near the peak of the transmission season (April-May).Length of follow-up: Not reportedAdjustment: primary analysis was done on intention-to-treat basis, assuming that all individuals living in an IRS cluster received IRS in their household. The effect of IRS was estimated using negative binomial regression models with the GEE approach. Sensitivity analyses and additional per-protocol analysis adjustments were done considering ITN ownership and usage, household socioeconomic status, and cluster size (as defined by number of households).\n\n\nParticipants: Number of participants: 1536 (active cohort), 139,286 across 194 villages under passive surveillanceInclusion criteria for participants: children aged < 5 years (active cohort); all ages (passive surveillance and cross-sectional surveys)\n\n\nInterventions: IRS active ingredient and dosage:  pirimiphos-methyl - 1 g/m2Formulation: Actellic 300 CSFrequency of spraying: annually for 2 yearsTime of spraying: October-November 2016 and 2017Spraying conducted by: President's Malaria Initiative Africa Indoor Residual Spraying (PMI AIRS)Coverage: all eligible structures (ones in which people slept and that had sprayable surfaces)Compliance: 83% of target buildings sprayed in 2016, 85% of targets sprayed in 2017  LLINActive ingredient and dosage: alphacypermethrin (in 2017 distribution campaign)Time of implementation: mass distribution campaigns in 2013 and 2017Coverage measure: ownership among all ages: 54% in 2017 and 95% in 2018Coverage in IRS arm: Not reportedCoverage in control arm: Not reportedCompliance measure: usage in households owning >= 1 ITN: 89% in 2018.No differences between study arms in proportion of children aged < 5  years who were reported to have slept under a net the night before monthly study; implemented household surveys, with estimates ranging from 59% to 67% before the mass distribution campaign and from 92% to 94% after the campaign.Cointerventions: none described\n\n\nOutcomes: Malaria infection incidence in an active cohort of children aged < 5 yearsMalaria case incidence in all ages through passive surveillance of national health system data (confirmed case defined as fever, either reported or measured plus a positive RDT)Malaria prevalence in all agesAdult mosquito densityEIR\n\n\nLocation profile: Study location: Mopeia district, Zambezia province, MozambiqueMalaria endemicity: highly endemic (> 60% parasite prevalence)EIR: < 1 Infectious bites per household per monthPlasmodium species: P falciparum\n\n\nVector profile: Primary (and secondary) vector species:An funestus and An gambiae s.l.Phenotypic resistance profile: resistant to pyrethroids (34-52% mortality after exposure to deltamethrin, 33-40% mortality after exposure to lambda-cyhalothrin)Method of mosquito collection:  vector densities were monitored monthly in a subset of 10 sentinel study villages: 5 IRS and 5 no-IRS villages, using overnight CDC light trap collections in 8 houses per village and paired indoor-outdoor human landing collections at 1 house per village, for 3 nights each month (note: no analysis of HLC results. Rate ratios were calculated using CDC light trap data only).\n\n\nNotes: For inclusion in the review meta-analyses, we calculated adjusted risk ratios for prevalence from the reported adjusted odds ratios following the methodology stated in Section 12.5.4.4 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a)\n\n",
        "objective": "To summarize the effect on malaria of additionally implementing IRS, using non\u2010pyrethroid\u2010like or pyrethroid\u2010like insecticides, in communities currently using ITNs.",
        "full_paper": "Background\nAttaining the goal of reducing the global malaria burden is threatened by recent setbacks in maintaining the effectiveness of vector control interventions partly due to the emergence of pyrethroid resistant vectors.\nOne potential strategy to address these setbacks could be combining indoor residual spraying (IRS) with non-pyrethroids and standard insecticide-treated nets (ITNs).\nThis study aimed to provide evidence on the incremental epidemiological benefit of using third-generation IRS product in a highly endemic area with high ITN ownership.\nMethods\nA cluster-randomized, open-label, parallel-arms, superiority trial was conducted in the Mopeia district in Zambezia, Mozambique from 2016 to 2018.\nThe district had received mass distribution of alphacypermethrin ITNs two years before the trial and again mid-way.\n86 clusters were defined, stratified and randomized to receive or not receive IRS with pirimiphos-methyl (Actellic\u00ae300 CS).\nEfficacy of adding IRS was assessed through malaria incidence in a cohort of children under five followed prospectively for two years, enhanced passive surveillance at health facilities and by community health workers, and yearly cross-sectional surveys at the peak of the transmission season.\nFindings\nA total of 1536 children were enrolled in the cohort.\nChildren in the IRS arm experienced 4,801 cases (incidence rate of 3,532 per 10,000 children-month at risk) versus 5,758 cases in the no-IRS arm (incidence rate of 4,297 per 10,000 children-month at risk), resulting in a crude risk reduction of 18% and an incidence risk ratio of 0.82 (95%\u00a0CI 0.79\u20130.86, p-value\u2009<\u20090.001).\nFacility and community passive surveillance showed a malaria incidence of 278 per 10,000 person-month in the IRS group (43,974 cases over 22 months) versus 358 (95%\u00a0CI 355\u2013360) per 10,000 person-month at risk in the no-IRS group (58,030 cases over 22 months), resulting in an incidence rate ratio of 0.65 (95%\u00a0CI 0.60\u20130.71, p\u2009<\u20090.001).\nIn the 2018 survey, prevalence in children under five in the IRS arm was significantly lower than in the no-IRS arm (OR 0.54, 95%\u00a0CI, 0.31\u20130.92, p\u2009=\u20090.0241).\nConclusion\nIn a highly endemic area with high ITN access and emerging pyrethroid resistance, adding IRS with pirimiphos-methyl resulted in significant additional protection for children under five years of age.\nTrial registration: ClinicalTrials.gov identifier NCT02910934, registered 22 September 2016, https://clinicaltrials.gov/ct2/show/NCT02910934?term=NCT02910934&draw=2&rank=1.\nBackground\nThere has been remarkable success in the global fight against malaria since 2000.\nDuring the period 2000\u20132015, coordinated malaria control efforts helped reduce worldwide malaria mortality rates in all ages by 47%, averting an estimated 4.3 million malaria deaths.\nThis progress was particularly impressive in Africa, where infection prevalence was halved and clinical cases reduced by 40%, averting an estimated 663 million cases, during the same time period.\nMost of this progress (81% of the cases averted) in Africa can be attributed to the successful scale-up of malaria vector control with conventional pyrethroid insecticide-treated nets (ITNs) and indoor residual spraying (IRS).\nDespite this overall success, recent trends indicate that maintaining high intervention coverage is challenging and that the number of malaria cases has increased slightly, but consistently, every year since 2016, this is mainly driven by a few high-burden countries.\nThis interrupted progress has put the malaria fight at a crossroads and threatens attaining the disease burden reduction targets set forth by the World Health Organization (WHO) in the Global Technical Strategy for Malaria 2016\u20132030 (GTS).\nFurther complicating the picture for vector control is knowledge that the continued effectiveness of currently available tools is threatened by the spread of insecticide resistance (especially pyrethroid resistance) in key vector populations.\nIndeed, extensive modelling conducted in preparation of the GTS suggests that innovative approaches are needed to get back on track to achieving the proposed goals.\nThese needs include better access to prevention and treatment interventions, better distribution systems, better tools with non-pyrethroid insecticides, and optimized combinations of available tools.\nThe efficacy of ITNs to reduce malaria incidence, prevalence, and even all-cause child mortality has been well established, and so this intervention has become the main malaria vector control method worldwide with an estimated 72% of households at risk in sub-Saharan Africa owning at least one ITN in 2017, as compared with 47% in 2010.\nUsage has also steadily increased; it is estimated that 50% of the population at risk in sub-Saharan Africa, including 61% of children under 5\u00a0years and 61% of pregnant women, slept under an ITN in 2017.\nThe impact and cost-effectiveness of IRS as a malaria control intervention has also been clearly established by historical and programme documentation.\nOne major challenge has been the increased cost of IRS with new insecticides or third generation IRS products (3GIRS).\nThis increased cost of IRS products was associated with a reduction in IRS coverage throughout sub-Saharan Africa.\nGlobally, the proportion of the population at risk protected by IRS was 5% in 2010 but declined to 3% in 2017 as countries identified insecticide resistance and new effective 3GIRS insecticides were more expensive.\nIn sub-Saharan Africa, IRS coverage experienced a marked decline from 10.1% (80 million people protected) in 2010 to 5.4% (51 million people protected) in 2016 before rising again to 6.6% (64 million people protected) in 2017.\nThis and other intervention coverage gaps, as well as a funding plateau, have been identified as important contributors to the stall in progress seen in 2017 and 2018.\nThe use of 3GIRS, with longer residual activity, in addition to high ITN coverage is one approach that could improve vector control and enhance disease burden reduction in some situations.\nThe current evidence for this potential benefit is mixed.\nAlthough modelling suggests additional incremental impact and observational studies suggest added value for IRS in addition to ITNs, experimental hut studies, non-randomized, and cluster-randomized trials show variable impact that is highly dependent on transmission intensity, vector bionomics, ITN coverage, insecticide resistance profiles, implementation strategies, and other factors.\nA recent metanalysis on the combined used of IRS and long-lasting insecticidal nets (LLINs) concluded that care is needed when using the limited available evidence for policy decisions.\nRegarding cost-effectiveness, there are important logistical costs associated with IRS, and while, a 2011 systematic review of IRS found it was cost-effective in low income setting, only one trial has explicitly evaluated the cost-effectiveness of the combined approach, and did so in the unique context of a low-burden region of Ethiopia.\nMoreover, the added value and cost-effectiveness of IRS in addition to ITNs in the context of intense transmission areas are critical questions for the new \u201cHigh burden to high impact\u201d strategy developed by the WHO and RBM.\nAlthough implementation is often sub-national, at least 35 countries in Africa already recommend combining ITNs and IRS, and the latter is often deployed in areas targeted by mass ITN distribution campaigns.\nCombining IRS and ITNs can result in different insecticides in the same area.\nIndeed, the WHO guidelines for vector control suggests that combined deployment can be used as part of an insecticide resistance management strategy, but specifically cautions against introducing a second intervention to compensate for deficiencies in the implementation of the first.\nRobust data are needed to guide decisions about prioritizing and combining vector control strategies in the context of different transmission dynamics, changing insecticide resistance patterns, and limited funds.\nTo help address this information gap in a high-intensity transmission setting with evidence of emerging pyrethroid resistance, a cluster-randomized trial was conducted, assessing the impact of IRS with a microencapsulated formulation of the organophosphate insecticide pirimiphos-methyl (PM) on malaria transmission, compared to no IRS.\nBoth arms received ITNs in accordance with the national distribution campaigns, which resulted in high ITN access in the IRS and no-IRS arms.\nThis study explores the epidemiological outcomes of malaria incidence and prevalence over two years.\nMethods\nThe overall study concept, setting, and methods of this open-label, controlled, parallel-arm, superiority trial have been previously published.\nStudy setting\nThe study occurred in rural Mopeia District (population 162,000) in the Zambezia Province of Mozambique during 2016\u20132018.\nZambezia is highly endemic for malaria, with parasite prevalence exceeding 60% and significant direct and indirect costs associated with the disease in some recent assessments.\nThe main vectors were Anopheles funestus and Anopheles gambiae sensu lato (s.l.) and data from neighbouring districts showed pyrethroid resistance in Anopheles gambiae s.l..\nAccess to ITNs was relatively high at baseline in 2016, as Mopeia District received 175,000 pyrethroid ITNs in a mass distribution campaign in 2013 (which represented more than one ITN per habitant) and benefits from routine distribution in antenatal clinics.\nMopeia received IRS (with DDT and then pyrethroids) from 2007\u20132011 and in 2014.\nIn Mozambique, ITN coverage is sustained through routine distribution at antenatal care clinics.\nIn 2017, the NMCP conducted a mass ITN distribution campaign with alphacypermethrin-treated ITNs in Mopeia.\nOwnership among all ages in Mopeia was 54% during the 2017 cross-sectional study and 95% in 2018.\nNet use in Zambezia among households with at least one ITN was 89% in the 2018 Malaria Indicator Survey.\nThe standard of care at public health facilities (testing of all fevers with a rapid diagnostic test (RDT) or microscopy and provision of treatment with artemisinin-based combination therapy to all positive cases) and from community health workers remained unaltered beyond study efforts to prevent stock outs of malaria commodities.\nThere were 30 community health workers providing passive testing, treatment and reporting in Mopeia throughout the study period.\nIntervention\nGiven expected impact at community level, IRS with PM was implemented only in the IRS-assigned clusters by the President\u2019s Malaria Initiative Africa Indoor Residual Spraying (PMI AIRS) project from October\u2013November in both 2016 and 2017 (Fig.\u00a01).\nSpraying was conducted according to PMI AIRS standard operating procedures, including community and household consent.\nStudy design\nThe study employed a two-arm, cluster randomized, controlled study design.\nA household and population enumeration was conducted from June\u2013July 2016, which identified 139,286 total residents (26,320 under five years old) living in 21,328 households distributed across 194 villages.\nCluster limits were delineated using expanded village borders through Voronoi polygons, with villages not reaching the minimum population for inclusion combined with the nearest neighbouring village to form a single cluster.\nThe primary research question was: In an area with high malaria endemicity and high ITN access, what is the incremental benefit of IRS with PM on reducing malaria transmission in a cohort of children under five years of age?\nThe primary outcome was malaria infection incidence in an active cohort of children under five years of age at community level.\nSecondary outcomes included: (1) passively reported confirmed case incidence in all ages through the national health system, including health facilities and community health workers and, (2) malaria prevalence in all ages from annual cross-sectional surveys near the peak of the transmission season (April\u2013May).\nRandomization and masking\nThe 168 clusters were stratified into three groups according to the number of households (<\u200969\u2009=\u2009small; 69\u2013125\u2009=\u2009medium;\u2009>\u2009125\u2009=\u2009large), and randomized 1:1 into one of the two arms, IRS and no-IRS, by drawing lots during a public community-engagement ceremony.\nEntomological surveillance\nThe standard PMI AIRS Mozambique vector surveillance methods and study-specific sampling strategies have been previously described.\nIn short, vector densities were monitored monthly in a subset of ten sentinel study villages (selected based on preliminary mosquito density surveys, ease of access and safety) five IRS and five no-IRS villages.\nIn each sentinel village, overnight CDC light trap collections were conducted at eight houses for three consecutive nights every month.\nAt one additional house per village, paired indoor-outdoor human landing collections were conducted overnight on the same three consecutive nights.\nSubsequent molecular analyses of the collected specimens (species confirmation, Plasmodium spp. infection rates, and appropriate pyrethroid resistance marker frequencies) have been reported in a separate publication.\nStandard WHO cone wall bioassay tests were performed at a subset of five randomly selected households in each of three villages in Mopeia to assess initial spray quality and estimate the residual efficacy of PM.\nLarval collections and subsequent insecticide resistance profiling of An. gambiae s.l. (2017 and 2018) and An. funestus s.l. (2018) using the WHO tube test bioassay also followed standard PMI AIRS Mozambique methods.\nPrimary outcome measures\nActive cohort\n86 total clusters (43 IRS, 43 no-IRS) were selected for participation in the active cohort component of the study.\nEligible households were selected from the core zones of each cluster using a fried-egg design with a 1-km buffer zone at the margins of each cluster, effectively leaving a buffer of at least 2\u00a0km between spray-discordant core zones.\nNo buffers were included between clusters that had been randomized to the same study arm.\nMalaria infection incidence at community level was determined by enrolling a cohort of children under five years of age under parental informed consent (18 children per cluster, 774 per study arm).\nThese children were visited monthly by a trained field worker that administered a short questionnaire to the caregiver and performed an HRP2-based rapid diagnostic test (RDT).\nEvery child with a positive RDT received treatment with artemether-lumefantrine (AL) according to Mozambique National Malaria Control Programme guidelines at baseline and in every subsequent visit.\nPerson-time at risk was reduced by ten days after each treatment to account for the prophylactic effect of lumefantrine.\nPassive case detection\nThe incidence of confirmed malaria cases (defined as fever, either reported or measured plus a positive RDT) that sought care in the public health system in Mopeia was measured using an enhanced passive surveillance approach: a study worker was placed in each of the 13 health facilities in the district to assure the quality of malaria case recording and to register the village origin of every case by village study-code.\nCross-sectional surveys\nA cross-sectional survey was conducted in April\u2013May in 2017 and again in 2018 to assess malaria prevalence in all-ages and to gather behavioural information as well data on costing, and health care expenditure.\nStatistical considerations\nFor the active cohort, 42 clusters of 12 children per arm had 80% power at a 5% significance level to detect a reduction in baseline incidence of 30% (from estimated 700/1,000 children-years to 490/1,000 children-years), using a robust K of 0.5.\nThe number of clusters per arm was 43 and the number of children per cluster was 18 at enrolment to account for potential sample loss.\nPower and sample size calculations were conducted using the Hayes and Bennett formula.\nThe sampling strategy for each cross-sectional survey (770 individuals, half under five years of age) aimed for 5% precision to measure an estimated prevalence of 50% in a population of 128,000.\nPrimary analysis was done on intention-to-treat, assuming that all individuals living in an IRS cluster received IRS in their household.\nThe effect of IRS was estimated using negative binomial regression models with the generalized estimating equations (GEE) approach.\nThis effect was adjusted for the variables identified as potential confounders in univariate models; the interaction term between IRS and ITNs was included in the multivariate analysis.\nSensitivity analyses and additional per protocol analysis adjustments were done considering ITN ownership and usage, household socioeconomic status, and cluster size (as defined by number of households).\nThe analysis was performed using Stata Statistical Software (StataCorp 2017).\nEthical reviews and registration\nAll procedures were reviewed and approved by PATH\u2019s Research Ethics Committee, CISM\u2019s IRB, and the National Ethics Committee of Mozambique as well as the PMI Operational Research Committee.\nThis study was reviewed by the Centers for Disease Control and Prevention (CDC) and determined to be human subjects research with non-engagement by CDC staff.\nThe trial was registered at clinicaltrials.gov with the identifier NCT02910934.\nResults\nTotal population and study flow\nThe study enumeration and enrolment process are depicted in Fig.\n2.\nIRS quality, acceptance and ITN distribution\nThe IRS campaigns were well accepted in Mopeia, with 16,500 structures (83%) sprayed of the 19,992 target in 2016 and 16,936 structures (85%) of the 19,950 target sprayed in 2017.\nThese targets were based on PMI AIRS-led structure enumeration which was conducted pre-spray each year.\nStandard WHO cone bioassays using susceptible An. gambiae sensu stricto (s.s.) in houses from Mopeia district indicated that both IRS campaigns were of high quality, with all houses tested in both years showing 100% mortality within 48\u00a0h of spraying.\nAdditionally, PM was efficacious for at least three months in 2017.\nIn 2018, results again showed residual efficacy for a minimum of three months on all wall surface types tested, though on mud walls efficacy was of longer duration and lasted for at least four months.\nIn June\u2013July 2017, all villages in Mopeia received 120,765 ITNs in the context of the mass distribution campaign.\nThere were no reported stock-outs of RDTs or anti-malarials reported at health facilities in Mopeia during the study period.\nFollowing the 2017 campaign, the four-month time point measurement showed 82% mosquito mortality on mud walls of Cero village and a five-month measurement in neighbouring Mocuba and Morrumbala districts showed 95% mortality.\nActive cohort detection\nBaseline characteristics were calculated using the active cohort first measurement to ensure comparability between clusters as the passive case recording did not delineate village of origin prior to the study period.\nA total of 1,536 children under five years of age (765 the no-IRS arm and 771 in the IRS arm), were enrolled in the active cohort from the 86 clusters (43 ITNs-only and 43 ITNs\u2009+\u2009IRS).\nThe distribution of cluster size was equal in both groups, with 14 small, 14 medium and 15 large clusters per arm.\nThe baseline characteristics of the cohort are shown in Table 1.\nThere were no major differences in terms of distance to the nearest health facility from the cluster\u00b4s centroid, ITN ownership, basic socioeconomic characteristics, age, or gender of the children enrolled.\nMore than 60% of the children had a positive RDT at enrolment.\nThe comparison of factors potentially associated with a positive RDT between both groups at baseline is presented in the Supplementary Materials.\nSpecifically, there were slight associations between malaria test positivity and living in a medium or large cluster, having a household sibling also testing positive, younger age, longer distances to the nearest health facility, and a history of fever in the last 48\u00a0h (Additional file 1: Table S1).\nITN ownership data collected at baseline and after the mass distribution campaign showed consistency across both arms and a large increase from 61\u201363% in January 2017 to 90% ownership of at least one ITN by the end of the trial (Additional file 1: Fig.\u00a0S1).\nThe children in the IRS arm experienced a significantly lower malaria infection incidence throughout the study.\nThere were 4,801 cases in the IRS arm (incidence rate of 3,532 per 10,000 children-month at risk) versus 5,758 cases in the no-IRS arm (incidence rate of 4,297 per 10,000 children-month at risk).\nThe crude risk reduction was 18% and the incidence risk ratio (IRR) was 0.82 (95%\u00a0CI: 0.79, 0.86, p-value\u2009<\u20090.001) (Table 2 and Fig.\u00a03).\nUsing these data, the IRS campaign in Mopeia averted between 15,697 and 21,651 malaria infections in the 12,670 children under five years of age living in the IRS clusters from January 2017 to October 2018.\nA sensitivity analysis was conducted, adjusting to account for residual HRP2 RDT positivity for up to 30\u00a0days after an infection, but did not show significant changes, IRR ranging from 0.79 to 0.86 (Supplementary Fig.\u00a02).\nThe coefficient of variation (k) between clusters for RDT positive test results was calculated to be 0.336 using GEE.\nGiven the crude incidence reduction of 18%, from 4,297 infections per 10,000 children-month (5.1 cases per child-year) in the no-IRS arm to 3,532 per 10,000 children-month (4.2 cases per child-year) in the IRS arm, the study had 74% power for its primary outcome with 43 clusters of 18 children per arm.\nUnivariate GEE with negative binomial models were used to explore which potentially confounding factors should be included in the multivariate model.\nThese results are shown in Table 3.\nVariables identified as having a significant influence on the IRR in the univariate analysis were included in a multivariate model using GEE.\nTable 4 shows the corrected IRR associated with IRS alone (i.e. no ITN owned), ITN use the night before, combined IRS\u2009+\u2009ITN use, having a sibling who tested positive, cluster size or distance to the nearest HF.\nThe combined effect of sleeping under an ITN the night before in a cluster that received the IRS intervention was significantly greater than the effect of either intervention used alone: the adjusted IRR for the interaction term was 0.62 (95% CI 0.57 \u2013 0.67; p\u2009<\u20090.001) corresponding to an incidence reduction of 38% (95%\u00a0CI 33%-43%).\nThe incidence reduction associated with IRS alone was 19% (95%\u00a0CI 13\u201326%) and 23% (95%\u00a0CI 18\u201328%) with ITN use alone (Table 4).\nThere was a 21% risk increase in children with at least one other sibling in the cohort that tested positive.\nThere was small but statistically significant reduction in the IRR in larger clusters and a higher risk of malaria in clusters with longer distances to health facilities (Table 4).\nSensitivity analyses were performed including only data after the ITN distribution campaign or adjusting the reference category for the IRR and no major changes in these results were noted (Additional file 1: Table S2).\nPassive case detection at HF\nThere was a total of 188 distinct villages coded during the district-wide pre-study enumeration, 81 that received IRS and 107 that did not, as per randomization which excluded a few villages that were not accessible for logistical or instability reasons.\nThe total enumerated population was 138,685, of which 18.8% (26,097) were under the age of 5\u00a0years.\nSlightly less than half the total enumerated population lived in IRS villages.\n(68,725\u2009=\u200949.6%).\nThere were 380,727 total visits to health facilities and to community health workers in Mopeia recorded during the study period; of these, 365,741 (96%) had a village code corresponding to a spray status with the rest corresponding to patients from outside the district boundaries, patients unwilling to disclose their home address, or visits with no village code recorded.\nOf visits with a corresponding village code, 174,126 (49%) included suspected malaria cases (patients presenting with, or reporting a history of, fever) that had an RDT performed: 102,004 (59%) of these RDTs were positive.\nFrom no-IRS villages, a total of 58,030 RDT-confirmed malaria cases were recorded over 22\u00a0months, resulting in a crude all-ages case incidence rate of 361 per 10,000 person-months at risk.\nAlmost half (48.7%) of these confirmed cases were in children under five years of age, an age-specific case incidence rate of 916 per 10,000 child-months at risk in this population.\nThere were significantly fewer confirmed cases of malaria recorded from IRS villages: 43,973 total cases, resulting in a crude all-ages case incidence rate of 278 per 10,000 person-months at risk.\nOf the cases from IRS villages, 45.5% were in children under five years of age, an age-specific case incidence rate of 687 per 10,000 child-months at risk.\nThe malaria incidence at health facility in the overall population was 358 (95%\u00a0CI: 355\u2013360) per 10,000 person-month at risk in the no-IRS group (58,030 cases over 22\u00a0months) and 278 per 10,000 person-month in the IRS group (43,974 cases over 22\u00a0months), resulting in an incidence rate ratio of 0.65 (95%\u00a0CI 0.60\u20130.71, p\u2009<\u20090.001).\nThe number of averted cases was estimated to be between 15,697 and 21,651.\nMonthly case incidence in both arms for the overall population and children under five years of age are shown in Table 5 and Fig.\u00a04.\nThe crude incidence was adjusted using a negative binomial regression model with variables identified via univariate regression.\nThese results confirmed the lower IRR in larger clusters (IRR: 0.98 per every 100 population increase 95%\u00a0CI 0.98\u20130.99 p\u2009<\u20090.0001) and also revealed an increased risk of malaria detection in clusters with shorter linear distance to a health facility; people living closer to a health facility received an RDT with higher frequency (IRR: 0.68 per every 5-km increase in distance 95%\u00a0CI 0.65\u20130.71, p\u2009<\u20090.0001).\nData are presented in Supplementary Tables 3 and 4.\nCross sectionals\nA total of 822 participants were surveyed in 2017 and 805 in 2018.\nBoth samples were balanced in terms of age, gender, ITN ownership, and other relevant factors (Table 6).\nIn the 2017 survey, conducted before the mass distribution of ITNs, there was no significant difference in prevalence at the peak of the transmission season between both study arms, even when correcting by age of the participant or ITN ownership (Tables 7 and 8).\nIn the 2018 survey, conducted ten months after ITN distribution, prevalence in children under five years of age in the IRS arm was significantly lower than in the no-IRS arm (OR 0.54, 95%\u00a0CI 0.31\u20130.92, p\u2009=\u20090.0241).\nThe incremental protective effect was particularly marked among ITN owners compared to those with no ITNs (Table 8).\nEntomological characterization\nA full analysis of the entomological impact of the IRS campaigns will be presented in a complementary manuscript.\nIn terms of characterizing the underlying vector bionomics at the sentinel sites, more than 90% of all anophelines collected (23,974/25,735) were An. funestus s.l. and 97% of those tested to date by PCR have been confirmed as An. funestus s.s. (2,234 / 2,309).\nSamples of An. gambiae s.l. were also present, though in substantially lower densities (1,320/25,735; 5% of all anophelines collected, with 82% of those tested [336/411] being Anopheles arabiensis).\nBaseline, pre-intervention, dry season CDC light trap collections from September and October 2016 indicated slightly higher An. funestus s.l. densities at the IRS sentinel sites compared to the no-IRS sentinel sites (geometric mean 4.5 [3.5\u20135.8] mosquitoes per trap-night vs. 2.5 [1.8\u20133.4] mosquitoes per trap-night).\nThe WHO tube test results from 2015 showed that pyrethroid resistance was evident in An. gambiae s.l. populations from the nearby districts of Mocuba (52% mortality against deltamethrin/40% against lambda cyhalothrin) and Morrumbala (34% mortality against deltamethrin/33% against lambda cyhalothrin), although data from Mopeia district in 2017 showed that An. gambiae s.l. was 100% susceptible to both alphacypermethrin and to PM.\nAnopheles funestus s.l. from Mopeia were tested in 2018 and were 100% susceptible to PM and DDT, but showed signs of emerging resistance to alphacypermethrin (85% mortality), deltamethrin (88% mortality), and bendiocarb (89% mortality).\nDiscussion\nThe IRS campaigns of 2016 and 2017 made positive contributions to malaria control in this high transmission district of Mozambique, as evident by: (1) reduced infection incidence in IRS clusters relative to no-IRS clusters, even in the presence of high ITN ownership; (2) reduced confirmed clinical case incidence at public health clinics; and among those detected by community health workers, and (3) reduced odds of malaria infection in the population under five years of age during the 2018 prevalence survey.\nMalaria policy makers and implementers face difficult decisions regarding the best available tools and their optimal deployment.\nPrior cluster-randomized trials have provided differing results, suggesting that the added value of the combination of ITNs and IRS is variable and likely.\ndependent on local factors like transmission intensity, vector bionomics, insecticide resistance profiles, and implementation strategy.\nThis study generated robust evidence to help support those making policy and implementation decisions about the use of IRS with a non-pyrethroid insecticide in communities with high rates of malaria transmission, high ITN ownership of standard pyrethroid-only ITNs, and evidence of emerging pyrethroid resistance in the local vector populations.\nThis study found significant added malaria protection by adding IRS with PM to a policy of universal coverage with a pyrethroid-only ITN in Mopeia.\nThis was quantified as 18% protective efficacy when considering new P. falciparum infections detected in the incidence cohort, and around 28% protective efficacy when considering confirmed cases reporting to the public health system for treatment.\nThese suggest that when resources are available, combining these two interventions will reduce malaria incidence.\nSome of the reasons contributing to this incremental impact include the indoor insecticide-mosaic created by combining pyrethroid ITNs and organophosphate IRS and the benefit obtained at the household level from at least one insecticide present independently of compliance with ITN use.\nThe adjusted analysis performed with the active cohort data confirmed that the interaction of IRS and ITNs leads to the greater incidence reduction, namely 38%.\nWhile this study provides valuable evidence on the combination of interventions, it also highlights the need to generate evidence on the value of IRS in combination with ITNs with piperonyl butoxide and non-pyrethroid ITNs to inform programmatic decision-making.\nThe reduced odds ratio of malaria infection observed in the population under five years of age during the 2018 prevalence survey is particularly interesting as it aligns with the protective effect of IRS also observed in the active and passive surveillance components of the study.\nIn IRS clusters, under-five prevalence was held relatively stable from 2017 (50%) to 2018 (47%), while in no-IRS clusters under-five prevalence increased from 47 to 62%.\nThis apparent increase in under-five prevalence in no-IRS clusters occurred in conjunction with (1) the mass ITN distribution campaign of 2017 that improved ITN access to more than 90% and (2) even as prevalence in the over-five population fell from 40 to 25% during the same time.\nAdditionally, the combination of IRS and ITNs appeared to significantly reduce the odds of malaria infection in the under-five population by almost 50% during the five months after the second spray campaign and following the mass distribution campaign.\nThese interesting trends highlight how complex the relationship between malaria infection incidence, malaria clinical case incidence, and malaria infection prevalence can be, particularly in very highly endemic areas with year-round transmission.\nThese results are in contrast from those of previous cluster-randomized trials conducted in lower transmission settings in which no benefit with a combined IRS and ITN approach showed no added benefit, but in concordance with the positive results seen in higher transmission settings.\nThis raises the question of whether the incremental impact maybe dependent on the transmission level.\nThis study employed a robust cluster-randomized design, with a multiplicity of outcome measures, a large sample size, and close community engagement, but there are some important limitations.\nChildren in the active cohort were subject to screen and treatment every visit, resulting in early detection of infections followed by prompt treatment, which also provided temporary prophylaxis.\nGiven that malaria infections were more common in children in the no-IRS arm, they received proportionally more treatments (and potential prophylactic benefit) which may have reduced the difference between arms, resulting in an underestimation of the true added benefit of the combined approach as adjustments done at analysis cannot fully correct for this prophylaxis.\nAdditionally, the net reduction in malaria incidence was lower than originally expected for sample size calculations; this was, however, partly compensated by the coefficient of variation, which, once retrospectively calculated from empirical data, was lower than assumed.\nAnother potential source of bias could be false-positive RDT results.\nThe RDTs used are based on the HRP2 antigen, which can persist for several weeks after treatment potentially inflating estimates of incidence in the active cohort and adding uncertainty around the number of true infections.\nThis effect would, however, occur equally in both groups.\nA sensitivity analysis adjusting incidence rates by censoring positive RDT results from a second consecutive household visit showed no major difference with the main findings presented here (Supplementary Fig.\u00a02).\nBoth study arms also benefitted from the study team efforts to avoid stock-outs which could have contributed to lower the incidence at cohort and health facility level in both arms.\nDespite these potential biases, it is reassuring to note the consistency among all outcome measures with active cohort, passive surveillance and cross sectionals.\nMalaria remains a challenge that will require multiple preventive as well as therapeutic intervention strategies, and many of these tools will need to be used in combination to maximize impact and reach elimination goals.\nUnderstanding when and where to combine vector control strategies requires locally relevant data to ensure that resources are invested wisely, particularly in the context of the \u201cHigh burden for high impact\u201d strategy.\nThis study demonstrated added value for IRS with a non-pyrethroid active ingredient in the context of high coverage with standard (pyrethroid-only) ITNs, as well as good access to malaria case management commodities.\nThis supports consideration of co-investment strategy (IRS and ITNs) in areas such as Zambezia, where transmission is high and the local primary vector species, An. funestus s.s., shows moderate levels of pyrethroid resistance.\nThe cost-effectiveness of this combined approach has been analysed in the context of this trial resulting in a separate manuscript.\nConclusion\nIn 2017, Mozambique had 5% of the global share of malaria cases.\nThe results of this trial suggest consideration for the combined deployment of non-pyrethroid IRS with ITNs in areas of high transmission and emerging pyrethroid resistance.\nThis strategy could prove especially valuable in the context of an overall increase in malaria burden and strategy put in place in an attempt to get back on track to achieving the 2030 goals as outlined in the WHO Global Technical Strategy.\nData availability and materials\nThe datasets generated and/or analysed during the current study are available in the Dip\u00f2sit Digital de la Universitat de Barcelona repository, http://diposit.ub.edu/dspace/handle/2445/101776.\nEthics approval and consent to participate\nAll procedures were reviewed and approved by PATH\u2019s Research Ethics Committee, CISM\u2019s IRB, and the National Ethics Committee of Mozambique as well as the PMI Operational Research Committee.\nThis study was reviewed by the Centers for Disease Control and Prevention (CDC) and determined to be human subjects research with non-engagement by CDC staff.\nStudy timeline, interventions and assessments. Considering at least nine months of efficacious indoor residual spraying (IRS) with pirimiphos-methyl (Actellic\u00ae300 CS), there was an overlap of IRS with older nets throughout 2017 and newer nets throughout 2018\nStudy flow chart. ACD: active case detection, PCD: passive case detection. *of the enrolled 1,536 children, three were under six months at enrolment and 54 were between 5 and 5.5\u00a0years\nCohort incidence by spray status (a); cohort cumulative incidence by spray status (b); and spray IRR (with 95% confidence interval) at cohort level (c). IRS campaigns highlighted in blue and mass ITN distribution highlighted in grey, ACT treatment correction of time at risk: 10\u00a0days\nMonthly population incidence at health facilities by spray status (a); cumulative population incidence at health facilities by spray status (b); and monthly incidence rate ratio (c). IRS campaigns highlighted in blue and mass ITN distribution highlighted in grey\n\nBaseline characteristics of the children in the active cohort, Mopeia, Mozambique\n | Spray Status | p-value\nNo-IRS | IRS\nCluster Characteristics (N\u2009=\u200986)\n\u00a0Km to nearest health facilitya | 6.1 (4.7)  | 6.8 (4.3)  | 0.4828\u00a0b\nHousehold Characteristics (N\u2009=\u20091536) e\n\u00a0ITN ownership\u00a0c | 470 / 765 (61.4%) | 487 / 771 (63.2%) | 0.4850\u00a0d\n\u00a0Number of ITNs in the household\u00a0a | 1.4 (0.7) [469] | 1.3 (0.6) [486] | 0.0868\u00a0b\n\u00a0Electricity in the household\u00a0c | 13 / 765 (1.7%) | 8 / 771 (1.0%) | 0.2641\u00a0d\n\u00a0Head of household with any formal education\u00a0c | 321 / 765 (42.0%) | 308 / 771 (39.9%) | 0.4225\u00a0d\n\u00a0Head of household farmer\u00a0c | 615 / 765 (80.4%) | 653 / 771 (84.7%) | 0.0263\u00a0d\nHouseholds enrolled (N\u2009=\u20091305) e\n\u00a0Siblings enrolled\u00a0c | 106 / 645 (16.4%) | 106 / 660 (16.1%) | 0.8549\u00a0d\nChildren enrolled (N\u2009=\u20091536) e\n\u00a0Gender: female\u00a0c | 362 / 765 (47.3%) | 389 / 771 (50.5%) | 0.2193\u00a0d\n Age (months) at enrolment\u00a0a | 32.4 (16.0) [765] | 31.2 (16.2) [771] | 0.1396\u00a0b\nCluster size\u00a0c |  |  | \n\u00a0Small | 243 (31\u00b7 8%) | 247 (32.0%) | 0.9917\u00a0d\n\u00a0Medium | 252 (32.9%) | 252 (32.7%) | \n\u00a0Large | 270 (35.3%) | 272 (35.3%) | \n\u00a0RDT positive\u00a0c | 474 / 765 (62.0%) | 499 / 771 (64.7%) | 0.2616\u00a0d\n\naArithmetic Mean (SD) [n], bt-test, cn (Column percentage), dChi-squared test, eNote that more than one child per household could be recruited, resulting in different denominators for children in the cohort and households in the cohort\n\nIncidence per 10,000 children-months. Time at risk corrected by ten days after each ACT treatment\nStudy | No IRS | IRS | Crude IRR | (95% Conf. Interval)\nMonth | RDT\u2009+\u2009 | Cohort months at risk | Cumulative cases | RDT\u2009+\u2009 | Cohort months at risk | Cumulative cases\n1 | 398 |  | 398 | 380 |  | 380 |  | \n2 | 374 | 600 | 772 | 422 | 705 | 802 | 0.96 | (0.83, 1.11)\n3 | 406 | 812 | 1178 | 353 | 767 | 1155 | 0.92 | (0.80, 1.06)\n4 | 345 | 644 | 1523 | 234 | 602 | 1389 | 0.73 | (0.61, 0.86)\n5 | 393 | 759 | 1916 | 331 | 762 | 1720 | 0.84 | (0.72, 0.97)\n6 | 355 | 678 | 2271 | 282 | 657 | 2002 | 0.82 | (0.70, 0.96)\n7 | 301 | 653 | 2572 | 282 | 692 | 2284 | 0.88 | (0.75, 1.04)\n8 | 227 | 663 | 2799 | 199 | 700 | 2483 | 0.83 | (0.68, 1.01)\n9 | 207 | 590 | 3006 | 163 | 628 | 2646 | 0.74 | (0.60, 0.91)\n10 | 209 | 724 | 3215 | 167 | 721 | 2813 | 0.80 | (0.65, 0.99)\n11 | 146 | 725 | 3361 | 137 | 747 | 2950 | 0.91 | (0.72, 1.16)\n12 | 154 | 698 | 3515 | 88 | 681 | 3038 | 0.59 | (0.45, 0.77)\n13 | 242 | 678 | 3757 | 172 | 693 | 3210 | 0.70 | (0.57, 0.85)\n14 | 233 | 619 | 3990 | 162 | 644 | 3372 | 0.67 | (0.54, 0.82)\n15 | 239 | 612 | 4229 | 204 | 633 | 3576 | 0.83 | (0.68, 1.00)\n16 | 326 | 688 | 4555 | 258 | 666 | 3834 | 0.82 | (0.69, 0.96)\n17 | 319 | 672 | 4874 | 265 | 675 | 4099 | 0.83 | (0.70, 0.98)\n18 | 260 | 632 | 5134 | 242 | 640 | 4341 | 0.92 | (0.77, 1.10)\n19 | 225 | 654 | 5359 | 160 | 653 | 4501 | 0.71 | (0.58, 0.88)\n20 | 187 | 655 | 5546 | 118 | 651 | 4619 | 0.63 | (0.50, 0.80)\n21 | 197 | 610 | 5743 | 170 | 574 | 4789 | 0.92 | (0.74, 1.13)\n22 | 15 | 37 | 5758 | 12 | 101 | 4801 | 0.30 | (0.13, 0.68)\n\nIRSindoor residual spraying, RDT rapid diagnostic test, IRR incidence rate ratio\n\nUnivariate analysis of covariables and their association with RDT positive status at monthly follow-up in active cohort of children under five\nVariable | Crude | (95% Conf. Interval) | p-value\nIRR\nSpray Status\u00a0a | 0.82 | (0.79; 0.89) | \u2009<\u20090.0001\nCluster size\nSmall | 1.00 |  | \u2009<\u20090.0001\nMedium | 0.95 | (0.89; 1.02)\nLarge | 0.8 | (0.75; 0.86)\nChild gender\u00a0b | 0.95 | (0.90; 1.01) | 0.1077\nSibling tested positive\u00a0c\u00a0(n\u2009=\u200928,998, m\u2009=\u20091,534) | 1.26 | (1.18; 1.33) | \u2009<\u20090.0001\nHead of household with any formal education\u00a0c\u00a0(n\u2009=\u200928,998, m\u2009=\u20091534) | 1.04 | (0.98; 1.10) | 0.1752\nHead of household farmer\u00a0c\u00a0(n\u2009=\u200928,998, m\u2009=\u20091,534) | 0.98 | (0.91; 1.06) | 0.6856\nElectricity in the household\u00a0c\u00a0(n\u2009=\u200928,998, m\u2009=\u20091,534) | 1.05 | (0.76; 1.44) | 0.7696\nChild with history of fever in the last 48\u00a0h\u00a0c\u00a0(n\u2009=\u200929,005, m\u2009=\u20091,536) | 1.90 | (1.83; 1.98) | \u2009<\u20090.0001\nParticipant slept under an ITN last night\u00a0c\u00a0(n\u2009=\u200927,479, m\u2009=\u20091,521) | 0.78 | (0.75; 0.81) | \u2009<\u20090.0001\nNumber of ITNs in household\u00a0d\u00a0(n\u2009=\u200923,175, m\u2009=\u20091,535) | 0.91 | (0.90; 0.92) | \u2009<\u20090.0001\nChild age (in months)\u00a0d | 0.99 | (0.99; 0.99) | \u2009<\u20090.0001\nKm to nearest health facility\u00a0d | 1.01 | (1.01; 1.02) | \u2009<\u20090.0001\n\nn\u2009=\u2009number of observations, m\u2009=\u2009number of subjects. (n\u2009=\u200929,020, m\u2009=\u20091,536), otherwise, specified. aCrude IRR for IRS vs. no-IRS cluster. bCrude IRR for Female vs. Male. cCrude IRR for Yes vs. No. dCrude IRR per unit increase. IRR incidence rate ratio\n\nAdjusted incidence using a multi-variable generalized estimating equation model\nVariable | Adjusted | (95% Conf. Interval) | p-value\nIRR\nIRS only\u00a0a | 0.81 | (0.74; 0.87) | \u2009<\u20090.0001\nITN use only\u00a0a | 0.77 | (0.72; 0.82) | \u2009<\u20090.0001\nIRS\u2009+\u2009ITN use\u00a0a | 0.62 | (0.57; 0.67) | \u2009<\u20090.0001\nSibling tested positive\u00a0a | 1.21 | (1.13; 1.29) | \u2009<\u20090.0001\nCluster size\n\u00a0Small | 1 |  | 0.0001\n\u00a0Medium | 0.95 | (0.89; 1.02) | \n\u00a0Large | 0.85 | (0.79; 0.92) | \nKm to nearest health facility\u00a0b | 1.01 | (1.01; 1.02) | 0.0001\n\naAdjusted IRR using children without ITN or IRS as referent group; bAdjusted IRR per 1-km increase. Number of observations\u2009=\u200927,479, number of subjects\u2009=\u20091,521. IRS: indoor residual spraying, ITN insecticide treated net, RDT rapid diagnostic test, IRR incidence rate ratio\n\nMalaria case incidence at health facilities in the overall and under 5\u00a0years of age population\nOverall population | Under five years\nStudy | No IRS | IRS | IRR | 95%CI | Study | No IRS | IRS | IRR | 95%CI\nMonth | RDT- | RDT\u2009+\u2009 | RDT- | RDT\u2009+\u2009 | Month | RDT- | RDT\u2009+\u2009 | RDT- | RDT\u2009+\u2009\n0 | 740 | 807 | 677 | 685 | 0.87 | (0.79, 0.97) | 0 | 340 | 424 | 315 | 328 | 0.83 | (0.72, 0.96)\n1 | 1289 | 2379 | 1238 | 1811 | 0.78 | (0.73, 0.83) | 1 | 559 | 1233 | 523 | 889 | 0.78 | (0.71, 0.85)\n2 | 1354 | 2881 | 1323 | 2539 | 0.9 | (0.86, 0.95) | 2 | 501 | 1390 | 540 | 1157 | 0.9 | (0.83, 0.97)\n3 | 1306 | 2419 | 1381 | 2004 | 0.85 | (0.80, 0.90) | 3 | 461 | 1138 | 468 | 877 | 0.83 | (0.76, 0.91)\n4 | 1228 | 2286 | 1084 | 1762 | 0.79 | (0.74, 0.84) | 4 | 396 | 1098 | 404 | 747 | 0.73 | (0.67, 0.81)\n5 | 1353 | 2577 | 1252 | 2195 | 0.87 | (0.83, 0.93) | 5 | 442 | 1324 | 450 | 1023 | 0.83 | (0.77, 0.90)\n6 | 1332 | 2606 | 1109 | 2135 | 0.84 | (0.79, 0.89) | 6 | 450 | 1306 | 379 | 966 | 0.8 | (0.73, 0.87)\n7 | 1367 | 2246 | 1230 | 1785 | 0.82 | (0.77, 0.87) | 7 | 537 | 1159 | 487 | 854 | 0.79 | (0.73, 0.87)\n8 | 1618 | 2336 | 1408 | 1726 | 0.76 | (0.71, 0.81) | 8 | 573 | 1127 | 567 | 790 | 0.76 | (0.69, 0.83)\n9 | 1410 | 1999 | 1240 | 1573 | 0.81 | (0.76, 0.86) | 9 | 568 | 1019 | 519 | 786 | 0.83 | (0.76, 0.91)\n10 | 1449 | 1898 | 1342 | 1562 | 0.84 | (0.79, 0.90) | 10 | 555 | 987 | 506 | 745 | 0.81 | (0.74, 0.90)\n11 | 1384 | 1441 | 1196 | 1043 | 0.74 | (0.69, 0.81) | 11 | 553 | 673 | 468 | 458 | 0.73 | (0.65, 0.83)\n12 | 1306 | 1644 | 1130 | 1115 | 0.7 | (0.64, 0.75) | 12 | 508 | 744 | 471 | 468 | 0.68 | (0.60, 0.76)\n13 | 2485 | 3455 | 2014 | 2365 | 0.7 | (0.67, 0.74) | 13 | 861 | 1580 | 705 | 975 | 0.66 | (0.61, 0.72)\n14 | 2228 | 3063 | 1831 | 2030 | 0.68 | (0.64, 0.72) | 14 | 752 | 1353 | 615 | 784 | 0.62 | (0.57, 0.68)\n15 | 2471 | 3653 | 2378 | 2646 | 0.74 | (0.71, 0.78) | 15 | 808 | 1764 | 812 | 1165 | 0.71 | (0.66, 0.77)\n16 | 2189 | 3708 | 2144 | 2770 | 0.77 | (0.73, 0.81) | 16 | 823 | 1877 | 873 | 1313 | 0.75 | (0.70, 0.81)\n17 | 2111 | 3986 | 1840 | 2999 | 0.77 | (0.74, 0.81) | 17 | 702 | 1910 | 610 | 1289 | 0.73 | (0.68, 0.78)\n18 | 1961 | 3207 | 1556 | 2212 | 0.71 | (0.67, 0.75) | 18 | 668 | 1597 | 569 | 1078 | 0.73 | (0.67, 0.79)\n19 | 1971 | 2741 | 1456 | 1915 | 0.72 | (0.68, 0.76) | 19 | 660 | 1355 | 547 | 921 | 0.73 | (0.67, 0.80)\n20 | 2051 | 2460 | 1837 | 1775 | 0.74 | (0.70, 0.79) | 20 | 697 | 1206 | 725 | 900 | 0.8 | (0.74, 0.88)\n21 | 1891 | 2487 | 1696 | 1961 | 0.81 | (0.76, 0.86) | 21 | 658 | 1213 | 663 | 944 | 0.84 | (0.77, 0.91)\n22 | 1723 | 1751 | 1543 | 1366 | 0.8 | (0.75, 0.86) | 22 | 605 | 810 | 554 | 560 | 0.74 | (0.67, 0.83)\nTotal | 38,217 | 58,030 | 33,905 | 43,974 | 0.78 | (0.77, 0.79) | Total | 13,677 | 28,287 | 12,770 | 20,017 | 0.76 | (0.75, 0.78)\n\nIRS: indoor residual spraying, RDT: rapid diagnostic test, IRR: incidence rate ratio. The overall (under 5) population considered in the table was 68,725 (12,670) for IRS and 169,960 (13,427) for non-IRS clusters.\n\nCharacteristics of the cross-sectional samples by spray status in 2017 and 2018, Mopeia, Mozambique\n | 2017 |  |  | 2018 |  | \n | Spray Status |  | p-value | Spray Status |  | p-value\n | No IRS | IRS |  | No IRS | IRS | \nGender Female | 169 / 420 (40.2%) | 174 / 397 (43.8%) | 0.2986 | 210 / 407 (51.6%) | 186 / 398 (46.7%) | 0.1676\nAge under 5 | 232 / 420 (55.2%) | 202 / 397 (50.9%) | 0.2123 | 195 / 407 (47.9%) | 205 / 398 (51.5%) | 0.3076\nDistance to nearest HFa | 7.1 | 6.8 | 0.7702 | 6.9 | 7.1 | 0.8821\nITN ownership | 235 / 419 (56.1%) | 204 / 397 (51.4%) | 0.1783 | 384 / 407 (94.3%) | 379 / 398 (95.2%) | 0.5758\nElectricity in the household | 2 / 420 (0.5%) | 19 / 397 (4.8%) | 0.0001 | 2 / 407 (0.5%) | 4 / 398 (1.0%) | 0.4465\nHead of household with any formal education | 206 / 419 (49.2%) | 203 / 397 (51.1%) | 0.574 | 299 / 407 (73.5%) | 292 / 398 (73.4%) | 0.975\nHead of household farmer | 350 / 419 (83.5%) | 344 / 397 (86.6%) | 0.2119 | 368 / 407 (90.4%) | 348 / 398 (87.4%) | 0.1776\n\na Mean linear distance from village centroid in km. IRS: indoor residual spraying\n\nPrevalence and odds ratio of malaria in the overall and under-five populations by spray status and age category in 2017 and 2018, Mopeia, Mozambique\n | 2017 | 2018\n | Spray Status | OR(95% CI) | p-value | Spray Status | OR(95% CI) | p-value\n | No-IRS | IRS |  | No-IRS | IRS | \nUnder 5 | 109 / 231 (47%) | 100 / 202 (50%) | 1.10 (0.62,1.93) | 0.7473 | 121 / 195 (62%) | 96 / 205 (47%) | 0.54 (0.31,0.92) | 0.0241\nOver 5 | 74 / 187 (40%) | 71 / 195 (36%) | 0.87 (0.55,1.38) | 0.567 | 52 / 212 (25%) | 40 / 193 (21%) | 0.80 (0.52,1.24) | 0.3241\nOverall | 183 / 418 (44%) | 171 / 397 (43%) | 0.97 (0.65,1.46) | 0.8894 | 173 / 407 (43%) | 136 / 398 (34%) | 0.70 (0.49,1.00) | 0.051\n\n\nPrevalence and odds ratio in the overall and under-five populations according to spray status and ITN ownership\n | 2017 | 2018\n | Spray Status | OR (95% CI) | p-value | Spray Status | OR (95% CI) | p-value\n | No IRS | IRS | No IRS | IRS\nITNowned\u00a01 | 96 / 235 (41%) | 94 / 204 (46%) | 1.24 (0.76,2.00) | 0.3869 | 166 / 384 (43%) | 125 / 379 (33%) | 0.65 (0.46,0.92) | 0.0139\nNo ITN owned\u00a01 | 87 / 183 (48%) | 77 / 193 (40%) | 0.73 (0.43,1.24) | 0.2459 | 7 / 23 (30%) | 11 / 19 (58%) | 3.14 (0.80,12.32) | 0.1004\nOverall\u00a01 | 183 / 418 (44%) | 171 / 397 (43%) | 0.97 (0.65,1.46) | 0.8894 | 173 / 407 (43%) | 136 / 398 (34%) | 0.70 (0.49,1.00) | 0.0514\n",
        "label": "low",
        "id": "task4_RLD_test_212"
    },
    {
        "paper_doi": "10.1136/bmjgast-2016-000124",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Randomized, double-blinded, placebo-controlled trialLength of follow-up: not reported\n\n\nParticipants: Number: 120Inclusion criteria:Children aged 3 to 60 months with acute diarrhoea, as evidenced by a Vesikari score of > 11Exclusion criteria:Children who had severe vomiting; those with a clinical diagnosis of dysentery or a known diagnosis of liver or renal failure; children who had prescriptions of probiotics or any other antidiarrhoeal medication.Missing data: 9 participants\n\n\nInterventions: Group 1: 60 children received racecadotril additional to oral rehydration or intravenous rehydration and zincGroup 2: 60 children received oral rehydration or venous rehydration and zincThey received either intravenous fluids as per WHO plan C (30 mL/kg followed by 70 mL/kg over 1 and 5 hours, respectively, in infants and over 30 min and 2.5 hours for those over 12 months of age) or low osmolality oral rehydration solution as per WHO plan B (75 mL/kg over 4 hours). Zinc was prescribed at 10 to 20 mg/day.The test arm received racecadotril at a dose recommended by the manufacturer: 10 mg per dose for children below 12 months of age and 30 mg for those over 12 months of age.\n\n\nOutcomes: Number of stools in the first 48 hours after introduction of the drug. The difference in the median number of stools on intention-to-treat analysis revealed no statistically significant difference.Duration of inpatient stay. This was measured as the number of days from the start of the medication to the day of discharge as determined by the attending physician.Duration of illness. This was defined as the duration from the time of introduction of the drug to the appearance of <= 3 formed stools in 24 hours.Adverse events associated with racecadotril.\n\n\nNotes: Location: KenyaThe trial received part of its funding from pharmaceutical industry\n\n",
        "objective": "To assess the efficacy and safety of racecadotril for treating acute diarrhoea in children under five years of age.",
        "full_paper": "Background\nDiarrhoea is the second most common cause of death in children under 5\u2005years of age in Kenya.\nIt is usually treated with oral rehydration, zinc and continued feeding.\nRacecadotril has been in use for over 2 decades; however, there is a paucity of data regarding its efficacy from Africa.\nObjectives\nThe objectives of this study were: to compare the number of stools in the first 48\u2005hours in children with severe gastroenteritis requiring admission and treated with either racecadotril or placebo, to study the impact of racecadotril on duration of inpatient stay as well as duration of diarrhoea and to describe the side effect profile of racecadotril.\nMethods\nThis was a randomised, double-blinded, placebo-controlled trial.\nIt enrolled children between the age of 3 and 60\u2005months who were admitted with severe acute gastroenteritis.\nThey received either racecadotril or placebo in addition to oral rehydration solution (ORS) and zinc and were followed up daily.\nResults\n120 children were enrolled into the study.\nThere were no differences in the demographics or outcomes between the 2 groups.\nStools at 48\u2005hours: median (IQR) of 5 (3\u20137) and 5 (2.5\u20137.5), respectively; p=0.63.\nThe duration of inpatient stay: median (IQR): 4\u2005days (1.5\u20136.5) and 4.5 (1.8\u20136.3); p=0.71.\nThe duration of illness: 3\u2005days (2\u20134) and 2\u2005days (1\u20133); p=0.77.\nThe relative risk of a severe adverse event was 3-fold higher in the drug group but was not statistically significant (95% CI 0.63 to 14.7); p=0.16.\nConclusions\nRacecadotril has no impact on the number of stools at 48\u2005hours, the duration of hospital stay or the duration of diarrhoea in children admitted with severe gastroenteritis and managed with ORS and zinc.\nTrial registration number\nPACTR201403000694398; Pre-results.\nSummary box\nWhat is already known about this subject?\nCurrent treatment of acute gastroenteritis as recommended by the WHO includes the use of oral rehydration solution and zinc.\nStudies carried out on racecadotril in children have proved its efficacy in reducing duration and severity of diarrhoea.\nThe European Society of Paediatric Gastroenterology, Hepatology and Nutrition has stated that racecadotril may be used in the management of acute gastroenteritis in children.\nWhat are the new findings?\nRacecadotril did not significantly impact the duration or severity of diarrhoeal disease in children with severe gastroenteritis.\nOne of the first studies using racecadotril to treat children in East Africa.\nHow might it impact on clinical practice in the foreseeable future?\nDefining which populations of children with acute gastroenteritis would have a maximum benefit from racecadotril.\nThis could allow recommendations on its use to be refined.\nIntroduction\nDiarrhoea is one of the most common causes of morbidity and mortality in children under 5\u2005years in the developing world.\nKenya is ranked in the top 10 countries bearing the burden of high mortality rates in children suffering from diarrhoeal disease with a prevalence of 16.7%.\nIn the 1970s\u20131980s, trends in mortality showed a significant decrease with the advent of oral rehydration solution (ORS).\nSince then this decline has slowed with current data at 530\u2005000 deaths per year.\nRacecadotril is a diesterified derivative of thiorphan.\nOnce absorbed after oral administration, it is converted to its parent compound (thiorphan) which acts to increase the half-life of enterocyte methionine-enkephalin (a potent antisecretory agent).\nRacecadotril has been in use for the past 20\u2005years in Europe and has been consistently useful in the management of diarrhoeal disease with an acceptable adverse effect profile.\nAlthough zinc has been shown to significantly affect the duration and severity of diarrhoea and forms part of the core treatment recommendations by the WHO, none of the trials studying efficacy of racecadotril have included it in their protocols.\nDespite a number of studies performed in different parts of the world, there has been no published data involving racecadotril from the African continent.\nOf the few studies carried out in the developing world, majority have been inpatient studies with methodological flaws noted in many.\nThis study was carried out to bridge gaps in knowledge about the use of racecadotril and its usefulness when used with zinc in African children with severe gastroenteritis.\nMethodology\nStudy design\nThis was a prospective randomised, double-blind, placebo-controlled, parallel-group study conducted at the Kenyatta National Hospital.\nThis is a tertiary level, national referral hospital and has four dedicated paediatric wards.\nParticipants were assigned to the control arm or treatment arm with a 1:1 ratio.\nEligibility criteria\nThe inclusion criterion was: children aged 3\u201360\u2005months requiring admission for severe acute gastroenteritis, as evidenced by a Vesikari score of >11 (figures 1 and 2) with written parental consent.\nExclusions were: children who had severe vomiting (scoring 3 on the Vesikari score for maximum number of vomiting episodes per day); those with a clinical diagnosis of dysentery or a known diagnosis of liver or renal failure; children who had prescriptions of probiotics or any other antidiarrhoeal medication (loperamide, attapulgite, activated charcoal, diphenoxylate and kaolin).\nInterventions\nAll children were admitted and started on WHO recommended treatment by the attending physician.\nThey received either intravenous fluids as per WHO plan C (30\u2005mL/kg followed by 70\u2005mL/kg over 1 and 5\u2005hours, respectively, in infants and over 30\u2005min and 2.5\u2005hours for those over 12\u2005months of age) or low osmolality ORS as per WHO plan B (75\u2005mL/kg over 4\u2005hours).\nZinc was prescribed at 10\u201320\u2005mg/day.\nParticipants were recruited within 24\u2005hours of admission after initial hydration had been completed.\nEligibility was determined by calculating the Vesikari score based on the information recorded by the primary physician in the patients' chart.\nThis included the patients' history up to and the examination at the time of admission.\nA computer program was used to generate random numbers in blocks with varying sizes.\nThese were used to assign patients randomly to either the test or control arm of the study.\nBoth the drug and placebo were supplied by the manufacturer, Racedot\u2014Macleods (India) and distributed by Sai Pharmaceuticals (Kenya).\nThese were packed in tamper proof brown bags, sealed and labelled by a study pharmacist based off site.\nThe test arm received racecadotril at a dose recommended by the manufacturer: 10\u2005mg per dose for children below 12\u2005months of age and 30\u2005mg for those over 12\u2005months of age.\nThis was administered as granules dissolved in 10\u2005mL of water and taken three times a day for a maximum duration of 3\u2005days.\nThe control arm had a placebo preparation administered in the same way.\nAn initial dose of the drug or placebo was given at the time of enrolment.\nThe parents were taught how to administer the drug at this time and this was confirmed by return demonstration.\nThe participants were followed up using daily interviews asking about: the number of stools, the consistency using a Bristol stool chart, the presence of blood in stool, any new symptoms and the introduction of an antidiarrhoeal or other medication.\nTreatment was given either until the stools were formed or for a total of 3\u2005days, whichever came first.\nChildren who were discharged before complete resolution of symptoms were followed up by phone interviews until recovery.\nOutcome measures\nThe primary outcome measure was the number of stools in the first 48\u2005hours after introduction of the drug.\nThe secondary outcomes included the duration of inpatient stay, the duration of illness and the number of adverse events associated with racecadotril.\nSample size calculation\nEstimation of sample size was carried out using the formula for comparison of two sample means.\nThe study was powered at 90% with a significance level of 95%.\nAssumptions for the formula were based on the study by Cojocaru et al using his outcome of total number of stools at 48\u2005hours.\nA provision of 10% was made for attrition and a final sample size of 60 participants in each arm was derived.\nThe attached online supplementary file contains the complete sample size calculation and references.\nStatistical methods\nBaseline data for the two groups were collected and analysed by comparing proportions to determine the presence of any significant differences.\nOnce the data were collected and analysed, it was noted to be positively skewed.\nMedians with IQRs were calculated for the outcome measures and the two groups were compared using the Mann-Whitney U test according to intention-to-treat analysis.\nSTATA V.11.0 software was used for the above.\nEthical consideration\nThe study was conducted after the approval of the Ethics and Research Committees of the Aga Khan University Hospital and the Kenyatta National Hospital/University of Nairobi.\nThe development of known side effects, suspected adverse effects or new symptoms were communicated to a Drug Safety Monitoring Board (DSMB).\nAdverse events were graded based on the Division of Allergy and Infectious Diseases (DAIDS) table for grading the severity of adult and paediatric adverse events summary sheet.\nResults\nA total of 154 children with severe gastroenteritis were screened for the study during the data collection period from 3 February to 26 March 2014.\nThe details of recruitment and allocation are shown in figure 3.\nAnalysis included all 60 participants in each group.\nThree participants from the drug group and two from the placebo group that did not complete the protocol were included in the analysis.\nAll results for the outcome measures (except adverse events) are given as medians with IQRs.\nBaseline characteristics\nThere was a statistically significant predominance of girls in the drug group (p=0.04) despite an overall male prevalence in the study.\nThis difference was believed to arise due to chance as randomisation was carried out according to protocol.\nThe difference in dehydration levels between the groups was not statistically significant (p=0.25).\nIt was noted that most children in the study were mild to moderately dehydrated with severe dehydration present in only 15% of the drug group and 27% of the placebo group.\nThis finding was attributed to use of the Vesikari score as an inclusion criteria.\nSimilar results were found in the study carried out by Schnadower et al.\nBaseline data are summarised in table 1.\nNumber of stools at 48\u2005hours\nThe difference in the median number of stools on intention-to-treat analysis revealed no statistically significant difference.\nMedian (IQR) 5 stools for the drug and 5 (2.5\u20137.5) for the placebo (p=0.63).\nDuration of inpatient stay\nThis was measured as the number of days from the start of the medication to the day of discharge as determined by the attending physician.\nThe duration was shorter in children in the drug group: median (IQR) of 4 (1.5\u20136) days as compared with placebo: 4.5 (1.8\u20136.3) days.\nThe results were not statistically significant (p=0.71).\nDuration of illness\nThis was defined as the duration from the time of introduction of the drug to the appearance of \u22643 formed stools in 24\u2005hours.\nThis was noted to be lower in the placebo group: median (IQR) of 2 days and 3 (2.5\u20133.5) days for the drug.\nThe results were not statistically significant (p=0.77).\nAdverse events\nThe proportion of patients experiencing any adverse event was similar for the two groups (23%).\nThere were four mortalities overall, two in each group.\nConvulsions were noted in three children who were on the drug (5%) but none in children on the placebo.\nAll three children were investigated for concurrent meningitis and this diagnosis was confirmed in one.\nOverall, the proportions of children who had serious adverse reactions were higher in the racecadotril group, although this was not statistically significant: relative risk 3.0 (CI 0.63 to 14.27); p=0.16.\nNon-serious adverse events: relative risk 0.81 (CI 0.36 to 1.83); p=0.62.\nThe adverse events are listed in table 2.\nAnalysis of the data using means and SD, with comparison using the Student\u2019s t-test, was also carried out.\nThe results showed no significant difference.\nLinear regression analysis revealed malnutrition was a significant factor for duration of in-hospital stay.\nThis analysis is presented in the online supplementary file.\nDiscussion\nThis study, carried out in children with severe acute gastroenteritis who were admitted to hospital and on treatment with zinc and ORS, compared the impact of racecadotril to a placebo.\nThere was no difference noted between the group on racecadotril and that on placebo for the outcomes measured.\nWhile this result is consistent with that seen by Santos et al in their study carried out in 2009, it differs from most other studies carried out on the subject.\nThe Vesikari score made up part of the inclusion criteria for our study, thus ensuring all children recruited had severe gastroenteritis.\nThis is a composite score that was originally developed by Ruuska and Vesikari.\nIt incorporates the patient's temperature, level of dehydration, need for hospitalisation as well as the duration and frequency of vomiting and diarrhoea.\nAlthough the Vesikari score was originally used to determine severity of gastroenteritis at the conclusion of illness, it was shown by Schnadower et al to be useful for grading episodes at the point of first contact and during follow-up.\nThey recommended its use in clinical trials involving acute gastroenteritis disease.\nThis recommendation was later endorsed by the European Society of Pediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) in their 2014 update on the guidelines for management of acute gastroenteritis in children.\nTo the best of our knowledge, the incorporation of the Vesikari score in the inclusion criteria has never been performed before in any studies on racecadotril and this may have led to the lack of effect seen with the medication in this study.\nIn their review carried out in 2011, Lehert et al alluded to the same noting that baseline level of dehydration was a significant negative predictor for duration of diarrhoea after introduction of racecadotril.\nPrevious studies on racecadotril have included children with diarrhoea lasting <5\u2005days.\nThe WHO defines persistent diarrhoea as that which lasts 14\u2005days or longer and children with disease duration less than this are treated for acute gastroenteritis.\nAlmost half of the participants included in this study had diarrhoea of 5\u2005days or more (46% in both groups).\nContinued damage to the intestinal mucosa results in an osmotic effect which contributes to the mechanism of diarrhoeal causation.\nRacecadotril has a minimal effect on this mechanism of diarrhoea and this may have contributed to the results in this study.\nAnother difference in the population studied was the inclusion of children with malnutrition.\nIn a trial carried out by Jean et al, children with weight for age <80% were excluded.\nIn this study, children with severe undernutrition made up 15% of the drug group and 26% of the placebo group.\nHowever, the mean weight of participants in this study was 7.4\u2005kg which was similar to a study carried out by Savitha (as quoted in the Lehert et al's review).\nIn both the Jean et al and Savitha studies, there was a significant improvement in the stool output for children on racecadotril.\nThe difference in the length of inpatient stay between the two groups constituted about half a day.\nThis was not statistically significant but was comparable with other studies carried out at the same site.\nOsano et al studied admissions of children with acute gastroenteritis at the Kenyatta National Hospital and noted that the average duration of stay was 4.2\u2005days.\nNone of the other inpatient studies carried out in children on racecadotril have described differences in duration of stays for their participants.\nCojocaru et al stated that the effect of racecadotril on duration of hospitalisation could not be commented on due to the short patient stay and follow-up in his study.\nThe use of zinc in this study may have contributed to similar durations of illness between the two groups.\nPrevious studies have shown a link between the use of zinc and the duration of illness in developing countries.\nIn most studies performed, racecadotril has been used alongside ORS although none have incorporated zinc into the treatment protocol.\nClinically significant adverse events were equally distributed between the two groups.\nThe results are similar to those in other studies including the review carried out by Lehert et al.\nAdverse events in this study involved mainly the skin and upper respiratory tract (13%), which are comparable to those found in the literature.\nConvulsions were noted in the drug group of this study, a finding that has not been reported in studies with racecadotril thus far.\nThe significance of this finding cannot be commented on due to the sample size.\nLimitations\nAlthough the number of stools was recorded on a daily basis, measurement of stool weights may have been a better measure to determine stool output as has been done in similar studies on the subject.\nThe recording of number of stools as reported by the parents/guardians may have introduced a recall bias that affected the results.\nThis may have been avoided by investigators observing the number of stools rather than parents providing information.\nThe presence of comorbid conditions was not considered when the participants were recruited to the trial.\nSome of these conditions (especially HIV) may have had an impact on the results.\nConclusion\nThis study does not demonstrate a significant reduction in the number of stools at 48\u2005hours for children admitted with severe gastroenteritis after introduction of racecadotril.\nFurthermore, it does not reduce the duration of inpatient stay or the duration of diarrhoea when compared with placebo.\nThe Vesikari clinical severity scoring system parameters and scores. Table adapted from Vesikari Clinical Severity Scoring System Manual.\nGrading of severity of diarrhoea according to the Vesikari score. Table adapted from Vesikari Clinical Severity Scoring System Manual.\nFlow diagram showing progress through phases of the trial. The final participants analysed are on ITT and PPA basis. ITT, intention to treat; PPA, per protocol analysis.\n\nTable showing baseline characteristics of the participants enrolled in the trial and the p value between the drug and placebo groups to show presence of statistical significance\nVariable | Racecadotriln (%) | Placebon (%) | p Value*\nSex: male | 27 (45) | 38 (63) | \nFemale | 33 (55) | 22 (37) | 0.04\u2020\nMalnutrition\n\u2003None | 34 (57) | 28 (47) | 0.27\n\u2003Mild to moderate | 17 (28) | 16 (27)\n\u2003Severe | 09 (15) | 16 (27)\nDuration of diarrhoea\u2021 (days)\n\u2003\u22644 | 31 (51) | 32 (53) | 0.97\n\u20035 | 15 (23) | 14 (23)\n\u2003\u22656 | 14 (23) | 14 (23)\nDehydration\n\u2003No | 3 (5) | 4(6) | 0.43\n\u2003Some | 26 (43) | 32 (53)\n\u2003Severe | 31 (51) | 24 (40)\n\n*Pearson constant on \u03c72 test.\n\u2020Significant difference in sex ratio between the placebo and drug groups, suspected to be due to chance.\n\u2021Duration of diarrhoea given as the number of patients (and %) who fell into each of the three groups as given by the Vesikari score.\n\nTable showing the number of adverse events in the drug and placebo groups, respectively\nEvent | Drug group (frequency) | Placebo group (frequency)\nSkin rashes | 6 | 8\nPruritus | 0 | 0\nAngiooedema | 2 | 1\nTonsilitis | 1 | 3\nMortality* | 2 | 2\nOthers (convulsions*) | 3 | 0\nTotal | 14 | 14\n\n*Severe adverse events as given by the DAIDS table. No statistical difference between the drug and placebo group: relative risk 3.0 (CI 0.63 to 14.27); p=0.16.\nDAIDS, Division of Allergy and Infectious Diseases.",
        "label": "low",
        "id": "task4_RLD_test_590"
    },
    {
        "paper_doi": "10.1186/1479-5868-5-63",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Setting: community-conference rooms at local hotels and the basement of a church (settings where the experiment took place, but not part of the experiment per se), USADesign: randomised controlled trialRecruitment: advertisements in local newspapers and flyers in community locations and in person at high schoolsAllocation to groups: no information on sequence generation was reported\n\n\nParticipants: Adolescents and adults recruited from suburban and urban populations in the local community including a high school whose students regularly ate fast food. Total number of participants recruited is not given although 605 participants completed the study procedures, of whom 301 were in the two study groups included in this reviewAge: 16-25 years: 25% (n = 147); 26-40 years: 19% (n = 115), 41-60 years: 42% (n = 248); > 60 years: 14% (n = 84)Gender: male 41% (n = 241), female 59% (n = 353)Ethnicity: Hispanic/Latino 3% (n = 20); Non-Hispanic/Latino, 97% (n = 567)\n\n\nInterventions: Intervention 1: menu with energy (kcal) information on a bright yellow background plus recommendation of daily energy intake for men and women in a box on the right-hand bottom corner of the menu (n = 151)Intervention 2: menu with no energy label, but with value pricing (the unit cost decreases as portion size increases) (n = 143)Intervention 3: menu with energy (kcal) information plus recommendation of daily energy intake (as above) and value pricing (n = 150)Control: menu with no energy labelling and no value pricing (n = 150)\n\n\nOutcomes: Nutrient composition of meal purchased and consumed: absolute measure of the energy and nutritional content (fat, saturated fat, carbohydrate, protein, fibre, vitamin C and calcium) were calculated using a food composition table available from the McDonald's corporation in combination with the gram weight information for the amount of each food item selected and consumed\n\n\nNotes: Participants chose items available from a McDonalds lunch/dinner menu, and research staff drove to nearby McDonald's restaurant to purchase meals ordered by the participants. Study sessions were held on weekday and weekend evenings (4:50 pm to 7:30 pm) between October 2005 and April 2006. Subgroup analyses were conducted by the study authors for: men and women; those who reported seeing the intervention menus and those who did not; those who reported that nutrition was important to them and those who did not; and those who reported that price was important to them and those who did not. Information about randomisation and raw data for subgroup analyses requested, but no response from author. Data were extracted on intervention 1 vs control. Interventions 2 and 3, which involved a price component, were not eligible for inclusion in this review. The research was supported by a NIDDK gran\n\n",
        "objective": "To assess the impact of nutritional labelling for food and non\u2010alcoholic drinks on purchasing and consumption of healthier items. Our secondary objective was to explore possible effect moderators of nutritional labelling on purchasing and consumption.",
        "full_paper": "Background\nAlthough point-of-purchase calorie labeling at restaurants has been proposed as a strategy for improving consumer food choices, a limited number of studies have evaluated this approach.\nLikewise, little research has been conducted to evaluate the influence of value size pricing on restaurant meal choices.\nMethods\nTo examine the effect of point-of-purchase calorie information and value size pricing on fast food meal choices a randomized 2 \u00d7 2 factorial experiment was conducted in which participants ordered a fast food meal from one of four menus that varied with respect to whether calorie information was provided and whether value size pricing was used.\nStudy participants included 594 adolescents and adults who regularly ate at fast food restaurants.\nStudy staff recorded the foods ordered and consumed by each participant.\nParticipants also completed surveys to assess attitudes, beliefs and practices related to fast food and nutrition.\nResults\nNo significant differences in the energy composition of meals ordered or eaten were found between menu conditions.\nThe average energy content of meals ordered by those randomized to a menu that included calorie information and did not include value size pricing was 842 kcals compared with 827 kcals for those who ordered their meal from a menu that did not include calorie information but had value size pricing (control menu).\nResults were similar in most analyses conducted stratified by factors such as age, race and education level.\nConclusion\nAdditional research is needed to better evaluate the effects of calorie labeling and value size pricing on fast food meal choices.\nStudies in which participants are repeatedly exposed to these factors are needed since long term exposure may be required for behavior change.\nBackground\nThe prevalence of overweight and obesity in the United States has increased dramatically.\nOne factor that many believe to be an important contributor to this increase is the number of meals and snacks eaten away from home.\nOver the past several decades the proportion of total food expenditures spent on food away from home has increased from 34% in 1974 to about half in 2004.\nFoods available at restaurants and other away from home eating locations tend to be higher in calories and fat and often larger in portion size compared to foods eaten from home, which may contribute to energy intake in excess of energy expenditure.\nIn recognition of the potential role of meals eaten away from home on excess energy intake, a number of strategies to promote more healthful food choices when eating out have been proposed.\nOne recommended approach is to increase the availability of nutrition information for foods eaten and prepared away from home.\nMoreover, it has been suggested that fast food chain restaurants be required to provide calorie information on their menu boards or on product packaging.\nIn theory, the provision of calorie information at the point-of-purchase for restaurant products may help consumers limit excess calorie intake.\nAside from providing point-of-purchase nutrition information, it has also been suggested that the food industry reduce its use of value size pricing as a marketing technique.\nValue size pricing involves structuring product prices such that the per unit cost (e.g., price per ounce) decreases as portion size increases.\nIt has been speculated that this product pricing structure leads to higher energy intake when eating out because value conscious consumers may be prone to purchase larger-sized food items.\nAlthough a number of studies have been conducted to evaluate the effect of point-of-purchase nutrition promotions on foods purchased away from home, the relevance of most of these studies to calorie labeling in a restaurant context is limited for a number of reasons.\nFirst, many of the studies evaluated point-of-purchase promotional activities that focused solely on identifying and promoting more healthful food choices (e.g., low-fat foods) versus providing nutrient content information, such as calorie information, for the full range of available foods.\nMany of the studies were conducted in a workplace or university setting instead of a restaurant.\nAmong those studies that evaluated comprehensive calorie labeling (e.g., providing calorie information for most food items), a number evaluated self-reported behavioral intentions rather than actual food purchases.\nIn consideration of the limitations of previous studies, we conducted an experimental trial to examine the effects on food purchases of providing calorie information at the point-of-purchase for food items on a fast food restaurant menu.\nThe influence of value size pricing on fast food meal choices and consumption was also examined.\nMethods\nOverview\nA randomized controlled 2 \u00d7 2 factorial experiment was conducted in which adolescents and adults who reported eating regularly at fast food restaurants were asked to purchase and consume a fast food restaurant meal from one of four randomly assigned menus.\nThe menus varied as to whether calorie information was provided and value size pricing was used (see Figure 1).\nMenu items ordered and consumed by each study participant were recorded by trained study staff.\nStudy Menus\nFour paper menus designed to be similar in format to menu boards at fast food restaurants were developed.\nFood items on all four menus were lunch and dinner items available at McDonald's at the start of the experiment (October 2005).\nThus, the menu included a variety of foods and beverages including hamburgers (n = 9), fish and chicken entrees (n = 6), salads (n = 4), french fries (n = 1), beverages (n = 9), and desserts (n = 4).\nAll of the portion size options available at McDonalds at the time of the study were included on the menu as well.\nTo blind participants to the meal source some food descriptions were modified.\nFor example, the Big Mac\u2122 hamburger was given the generic description 'Double Cheese Burger Deluxe'.\nEach of the four study menus is briefly described as follows:\nCalorie Menu\nThe calorie menu included calorie information for each menu item.\nThis information was provided in a column between the food description and product price columns (see Figure 2 for menu excerpt).\nTo draw attention to it, the background color of the calorie column was bright yellow.\nTo put the calorie information in context the average daily calorie needs of adult men and women were provided in a 'Calories Count' information box in the bottom right hand bottom corner of the menu (see Figure 3).\nThe calorie content information for each item was obtained from the McDonald's Corporation web-site.\nFood items were priced in accord with McDonald's usual food pricing for the area.\nPrice Menu\nThe price menu was modified for items with more than one portion size option so that the value size pricing structure was eliminated from the menu.\nThe prices listed for food items that had more than one portion were calculated so that the price per ounce was standardized across portion size options (see Figure 4 for menu excerpt).\nTo calculate the standardized prices, a standard price per ounce for each product type (e.g., soft drinks) was determined and applied to calculate the total price for each product portion size.\nThe standard price per ounce used in this calculation was based on the price per ounce of the medium-sized portion available (price of medium serving divided by the total number of ounces in a medium serving).\nUsing this price per ounce estimate, the price for each product portion size option was calculated by multiplying the standard price per ounce by the number of ounces in each product portion.\nThe price menu did not include calorie information.\nCalorie plus Price Menu\nThe calorie plus price menu included calorie information and price modification.\nThe calorie information provided was identical to that on the calorie menu.\nThe price modification was identical to that of the price menu (see Figure 5 for menu excerpt).\nControl Menu\nThe control menu did not include calorie information and all menu items were priced in accord with usual McDonald's pricing, and thus included value size pricing (see Figure 6 for menu excerpt).\nParticipant Recruitment\nParticipants were recruited from suburban and urban communities in the Minneapolis St. Paul, Minnesota metropolitan area via advertisements placed in community newspapers and flyers posted in community locations.\nAlso, recruitment was conducted in-person at selected area high school.\nA $25 gift certificate to a discount store was offered as an incentive to participate.\nThose who called the recruitment phone number provided in advertisements and flyers were screened for the following eligibility criteria: 1) \u2265 16 years of age; 2) eat at fast food restaurants \u2265 1 time/week; and 3) able to read and speak English.\nThose who met eligibility criteria were told that participation would involve completing a two-hour evening study session at which they would be required to purchase a fast food restaurant meal for their dinner and complete several questionnaires.\nTo minimize subject reactivity participants were blinded to the menu manipulation aspect of the study.\nThe study purpose was described as \"learning more about fast food meal choices\".\nParticipants were also blinded to the source of the fast food.\nThose who indicated an interest in participating were scheduled for a study session, and randomly assigned to one of the four menu conditions.\nStudy Session Procedures\nStudy sessions were held on weekday and weekend evenings (4:50 pm\u20137:30 pm) between October 2005 and April 2006.\nThe sessions were held in three sites.\nTwo of the sites were conference rooms in suburban hotels.\nOne site was the basement of an urban church.\nEach location was within six blocks of a McDonald's restaurant.\nUpon arrival at the study session each participant met individually with a study staff member who provided him/her with the study menu to which they were randomly assigned.\nThe participant was asked to order their dinner from the menu, and the study staff member recorded the order.\nAfter ordering participants were told that payment for their meal would be collected from them at the end of the study session, and then they were escorted to a room set up as a dining area.\nParticipants were asked to complete a survey while waiting for their food.\nThe survey included questions about fast food consumption frequency, opinions about fast food restaurants, and food shopping and preparation practices.\nImmediately after participants placed their dinner order study staff drove to the nearby McDonald's restaurant to purchase the meals ordered by participants.\nParticipants were then provided their food by study staff.\nWhen participants were finished eating, leftover food was collected and covertly measured using a digital food scale.\nAfter meals were consumed participants were escorted to an exit interview area where a final interview was completed by a research staff person.\nThis interview included questions about nutrition knowledge and beliefs and self-reported height and weight.\nAfter participants completed the interview they were informed of the true intent of the study (debriefed) and told that they would not have to pay for their meal.\nThey were then asked several questions to determine whether they had noticed the menu manipulations, and whether they were aware of the true purpose of the study or the source of the food prior to the study session.\nThe University of Minnesota Institutional Review Board approved all study procedures.\nDetermining Nutrient Composition of Meals Selected and Consumed\nThe nutrient composition of the meals selected and consumed by participants were calculated using a food composition table available from the McDonald's corporation in combination with the gram weight information for the amount of each food item selected and consumed.\nAlthough energy (kcal) was the primary focus, estimates for total fat, total carbohydrate, total protein, saturated fat, dietary fiber, vitamin C, and calcium were also generated so that possible experimental effects on selection and consumption of these nutrients could be examined.\nData Analyses\nA total of 605 individuals completed the study procedures.\nEleven of these individuals were excluded from the analyses because they disclosed during the debriefing that they knew before participating in the study that calories might be listed or price would be modified on the menu (n = 2) or knew that they would not have to pay for their meal (n = 9).\nMeans and frequencies were calculated to describe the characteristics of the study sample.\nTo evaluate whether randomization was effective in equally distributing potential confounders across experimental conditions, characteristics of participants such as age, sex, and education level were compared across experimental groups.\nGeneral linear modeling (GLM) was conducted to test for differences in the food and nutrient composition of meals selected and consumed by those in each experimental condition.\nIn each model, the food or nutrient under consideration was included as the dependent variable and experimental condition was included as the independent variable.\nPossible differences in experimental condition effects by age (16\u201328 years, 29\u201349 years, 50+ years), sex (male, female), highest education level (college graduate or higher, less than college graduate), body weight (normal weight, overweight or obese), and perceived importance of nutrition and price when eating fast food were examined by conducting analyses stratified by these factors, and qualitatively comparing results.\nTests for interactions were not conducted because power for two-way interactions was weak.\nAll analyses were conducted using SAS (Version 9.1.2, 2004; SAS Institute, Cary, NC).\nStatistical Power\nThe study was designed to have strong power (85%) to detect small main effects (delta < 100 kcal; effect size = 0.2) under ideal conditions (e.g., low variance, perfect randomization, Gaussian errors).\nModest to strong effect sizes may be detected in analyses conducted stratified by factors such as sex and education level.\nFor example, in analyses involving one-half of the sample, modest effects (delta < 140 kcal) are detectable.\nResults\nThe demographic characteristics of participants are shown in Table 1.\nMore females (59.4%) than males (40.6%) participated.\nApproximately three-fourths of participants were white, with blacks comprising the second largest racial/ethnic group (10.9%).\nThe demographic characteristics of participants in each experimental group were similar.\nWhen asked to rate the importance of price, taste, nutrition and convenience when purchasing food from a fast food restaurant and when buying groceries, taste was the most highly rated factor for both; 97.6% and 98.5% reported taste as very important or somewhat important when buying fast food and groceries, respectively (Table 2).\nIn contrast, nutrition was the least likely to be rated as very important or somewhat important, with 58.2% and 83.5% of participants rating nutrition as very important or somewhat important when buying fast food and groceries respectively.\nThe average energy and nutrient composition of meals ordered and consumed by those in each experimental group (Table 3) were similar.\nFor example, there were no significant differences (p = 0.25) in the average number of calories consumed by those in the calorie, price, calorie plus price, and control menu conditions were 805, 813, 761, and 739 respectively.\nSelection and consumption of major food categories (e.g., sugar-sweetened soft drinks, diet soft drinks, French fries, salads, etc.) and portion sizes were also examined, with no significant differences found (data not shown).\nThe average energy content of meals selected and consumed were similar across experimental conditions among those in each age, education level, and body weight strata.\nHowever, significant differences in energy intake across experimental condition were observed among males (p = 0.01), those who reported that nutrition was important when buying food from a fast food restaurant (p < 0.01), and those who reported price was not important when buying food from a fast food restaurant (p = 0.01).\nAverage energy intake was higher among males in the calorie, price, and calorie plus price experimental conditions compared to those who selected their meal from the control menu (Figure 7).\nAmong those who reported that nutrition was important when buying fast food, average energy intake was significantly lower among those who received the control and calorie plus price menus relative to those in the other two experimental conditions (Figure 8).\nAmong those who reported that price was not important when buying food from a fast food restaurant, average energy intake was lowest among those in the control condition (598 kcal) and highest among those in the calorie plus price condition (948 kcal) (Figure 9).\nAs part of the post-meal interview, participants in the calorie, price, and calorie plus price menu conditions were asked questions to assess whether they noticed the menu manipulations.\nThose who received a menu with calories listed were asked, \"Did you notice that calorie counts were listed on your menu this evening?\".\nAbout half (54%) of those in the calorie condition and 59% of those in the calorie plus price menu conditions reported noticing the calorie information (Table 4).\nThose with a higher level of education, whites, and those 15\u201325 years of age were more likely to report noticing the calorie information (data not shown).\nParticipants who received the price or calorie plus price menu were asked, \"Did you notice on the menu from which you ordered tonight that prices were set such that for a given unit of food (like a chicken nugget or ounce of soda), you paid the same price for that unit no matter what size you ordered?\"\nDue to the complexity of the question, an illustration of this pricing structure was provided and explained as part of the question.\nLess than one-fifth reported noticing this pricing structure (Table 4).\nThe demographic characteristics of those who reported noticing the price modification were similar to those who did not (data not shown).\nTo evaluate whether calorie information may have influenced food choices among those who reported noticing this information relative to those who did not, a linear regression analysis was conducted.\nThis analysis was restricted to participants in the calories and calories plus price experimental conditions.\nCovariates included in the analysis were factors found to differ between those who reported noticing the calorie information and those who did not notice this information, specifically age, race, education level, and site.\nResults from the multivariate analysis indicated that average energy intake was comparable between those who reported noticing the calorie information and those who did not (690 kcal versus 671 kcal; p = 0.65).\nResults were similarly null in a comparable analysis conducted to compare energy intake of those who noticed and did not notice the pricing structure modification (p= 0.90).\nDiscussion\nResults of the present study showed that providing calorie information at the point-of-purchase on a fast food restaurant menu had little effect on food selection and consumption among a sample of adolescents and adults who eat regularly at fast food restaurants.\nThese results contribute to a limited literature on point-of-purchase calorie labeling.\nTo date, seven studies have examined the influence of providing calorie composition information at the point-of-purchase for most food items available in a cafeteria or restaurant setting.\nAmong these studies, one found no evidence of an effect of calorie labeling on food choices.\nIn contrast, six of the seven studies found some evidence in support of the hypothesis that calorie information may positively influence food choices, however, results from most of these studies were weak or inconsistent.\nFor example, Conklin et al. found that only 18% of college freshman living on a campus where point-of-purchase nutrition information was available in the dining commons agreed that the available information affected their choice of food.\nA host of factors may explain the weak and inconsistent results in the literature.\nFirstly, the calorie labeling formats utilized varied across studies.\nFor example, in one study 5 cm by 5 cm cards with calories in red ink were placed as close as possible to food items in a hospital cafeteria.\nIn contrast, in another cafeteria study calorie information for all menu items was presented on two large posters at the cafeteria entrances, with leaflets distributed to patrons to encourage use of this information.\nAlso, in three studies calorie information was provided along with other nutrient composition information such as saturated fat and fiber.\nIn the present study, calorie information alone was provided on a restaurant menu in a column between the food item name and price.\nTo draw attention to it, the column was highlighted in bright yellow.\nNonetheless, only slightly more than one-half of those who ordered from a menu with calories listed reported noticing this information.\nThe calorie content of meals selected by those who noticed the information compared to those who did not were similar, suggesting our null results are not solely due to the failure of some to notice the calorie information.\nThe designs of studies conducted to date have varied greatly, with all having limitations.\nMost notably, four studies measured behavioral intentions rather than actual food choices.\nConsequently, social desirably bias in reporting is a significant concern in these studies.\nOther major weaknesses include use of quasi-experimental designs where factors other than the experimental conditions being tested may have differed across test periods due to lack of randomization.\nThe present study is the first to measure actual food choices within an experimental design where participants were randomly assigned to experimental condition.\nHowever, it has methodological weaknesses.\nParticipants were exposed to the calorie information on only one occasion.\nThis is a critical shortcoming if repeated exposure to calorie information is required before awareness or behavior change may be expected.\nA final issue is that the weak and inconsistent results across studies may reflect heterogeneity in response to calorie labeling, with certain population subgroups more apt to utilize calorie information when it is provided.\nFor example, a number of studies have found that females are more likely than males to use nutrition information on packaged food products.\nConsequently, it is perhaps not surprising that Milich et al found calorie labeling to effect cafeteria food choices in a sample of female hospital employees, whereas Mayer et al.\nfound no significant effect of calorie labeling on cafeteria food choices in a study involving male and females employees of a Fortune 500 company.\nThe results of the present study are consistent with the notion that certain population subgroups may be more likely to use calorie information when it is provided.\nIn the present study males appeared to use the calorie information to choose a higher calorie meal.\nThis finding could be an artifact of multiple comparisons, as a significant number of subgroup analyses were conducted.\nConversely, this result could reflect a desire among males for an energy dense meal.\nTo our knowledge, only one other study has reported findings suggesting an unintended consequence of calorie labeling.\nYamamoto et al. conducted a study in which adolescents were asked to order meals from three different restaurant menus that did not contain nutrition information, and then reorder their meals after being shown a version of the menus that included calorie and fat content information for menu items.\nApproximately 17% of meal orders were changed in response to the calorie and fat information.\nAmong the meals that were modified, 20.4% were modified in a way that resulted in a higher calorie content meal.\nIn the present study the elimination of value size pricing was found to have little influence on food selection or consumption.\nThis finding is somewhat surprising given that a number of studies have documented that price changes may influence food choices.\nThe price shifts we evaluated tended to be smaller in magnitude compared to those evaluated in previous studies, which could explain why our results conflict with previous findings.\nIt is also possible that the null results are due to the study design which provided only one exposure to the price modification.\nWhen queried regarding whether they had noticed the modified pricing structure, less than one-fifth responded affirmatively.\nSince most fast food restaurant chains utilize a value size price structure, it is possible that study participants generally assumed the larger sized items were the better value without considering the prices listed on the menu.\nIn consideration of this potential methodological issue, future studies designed to evaluate value size pricing should ensure repeated exposure to price modification.\nThe present study has a number of strengths.\nThe study measured actual food choices rather than behavioral intent.\nConsequently, social desirability bias in reporting is likely less of a concern and internal validity is probably better than studies that only measured behavioral intent.\nA randomized design was employed ensuring that potential confounding factors, such as age and sex were equally distributed across experimental conditions.\nAnother strength of the present study is that participants were a community sample of adolescents and adults who ate regularly at fast food restaurants.\nConsequently, the external validity of results may be stronger than many previous studies.\nLimitations of the present study include that participants were exposed to the experimental condition only once.\nAs mentioned earlier, this is problematic as it is possible that repeated exposure to calorie information and standardized pricing may be required before behavior is impacted.\nAlthough participants were blinded to the true intent of the study, and the ordering and dining procedure was set-up to be as naturalistic as possible, subject reactivity remains a concern.\nOf particular concern is the possibility that the participation incentive undermined price sensitivity.\nIt is important to note that most of the study limitations just described could have been avoided if the study had been conducted in fast food restaurants where menu boards and prices could be manipulated for prolonged periods of time.\nUnfortunately we were not able to find any fast food restaurants willing to collaborate with us, and thus we were not able to implement a more rigorously designed study.\nConclusion\nIn conclusion, results from this study indicate that providing calorie information for food items on fast food restaurant menus may have little effect on the food choices made by adolescents and adults who regularly eat at these establishments.\nIt is possible that skills for using point-of-purchase nutrition information must be built before the information provided may be effectively used.\nFor example, Hawthorne at el. found that knowledge of the basic use of the nutrition facts label on packaged food products was low among a sample of adolescents.\nAfter a brief training on use of the label, understanding of the label was significantly improved.\nMore likely though is the need to increase concern about nutrition when eating at fast food restaurants, as factors such as taste and convenience appear to be far more important consideration for most consumers.\nAlthough the design of the present study is more methodologically rigorous than most previous research, it has significant shortcomings.\nConsequently additional research is warranted to more rigorously evaluate calorie labeling and value size pricing in the context of fast food restaurants.\nCollaborating with fast food restaurant establishments will be a critical but challenging need.\nDiagram of 2 \u00d7 2 experimental design menus.\nExcerpt from calorie menu.\nCalorie reference information provided in the bottom right hand corner of the calorie and calorie plus price menus.\nExcerpt from price menu.\nExcerpt from calories plus price menu.\nExcerpt from control menu.\nAverage energy intake by experimental condition among females (n = 353) and males (n = 241).\nAverage energy intake by experimental condition among those who reported nutrition was very important or somewhat important (n = 341) or not very important or not at all important (n = 245) when buying foods from a fast food restaurant.\nAverage energy intake by experimental condition among those who reported price was very important or somewhat important (n = 496) or not very important or not at all important (n = 96) when buying foods from a fast food restaurant.\n\nDemographic characteristics of participants by experimental group\n | Total(n = 594)% (n) | Caloriea(n = 151)% (n) | Priceb(n = 143)% (n) | Calorie + pricec(n = 150)% (n) | Controld(n = 150)% (n) | p-valuee\nAge (years) |  |  |  |  |  | \n\u200316\u201325 | 24.8 (147) | 19.9 (30) | 31.5 (45) | 21.3 (32) | 26.7 (40) | 0.08\n\u200326\u201340 | 19.4 (115) | 14.6 (22) | 20.3 (29) | 22.0 (33) | 20.7 (31) | \n\u200341\u201360 | 41.8 (248) | 46.4 (70) | 35.7 (51) | 41.3 (62) | 43.3 (65) | \n\u2003\u2265 61 | 14.1 (84) | 19.2 (29) | 12.6 (18) | 15.3 (23) | 9.3 (14) | \nSex |  |  |  |  |  | \n\u2003Male | 40.6 (241) | 37.7 (57) | 37.8 (54) | 46.0 (69) | 40.7 (61) | 0.42\n\u2003Female | 59.4 (353) | 62.3 (94) | 62.2 (89) | 54.0 (81) | 59.3 (89) | \nEthnicity |  |  |  |  |  | \n\u2003Hispanic/Latino | 3.4 (20) | 1.3 (2) | 5.7 (8) | 4.0 (6) | 2.7 (4) | 0.21\n\u2003Not Hispanic/Latino | 96.6 (567) | 98.7 (148) | 94.3 (133) | 96.0 (144) | 97.3 (142) | \n\nEducation levelf |  |  |  |  |  | \n\u2003High school | 25.3 (150) | 29.1 (44) | 23.1 (33) | 20.7 (31) | 28.2 (42) | 0.54\n\u2003graduate or less | 38.8 (230) | 39.1 (59) | 41.3 (59) | 40.0 (60) | 34.9 (52) | \n\u2003Some college | 35.9 (213) | 31.8 (48) | 35.7 (51) | 39.3 (59) | 36.9 (55) | \n\u2003College graduate or higher |  |  |  |  |  | \nBody weightg |  |  |  |  |  | \n\u2003Normal weight | 42.6 (249) | 43.2 (64) | 45.1 (64) | 40.1 (59) | 41.9 (62) | 0.83\n\u2003Overweight | 27.9 (163) | 27.0 (40) | 26.1 (37) | 32.7 (48) | 25.7 (38) | \n\u2003Obese | 29.6 (173) | 29.7 (44) | 28.9 (41) | 27.2 (40) | 32.4 (48) | \n\na calorie menu included calorie information for each menu item\nb price menu had price modification (standardized pricing) for food items with more than one portion size option\nc calorie + price menu included calorie information and price modification (standardized pricing) for food items with more than one portion size option.\nd control menu did not include calorie information and had usual food pricing\ne p-value calculated from chi-square test.\nf For participants 16\u201319 years of age, the reported education level is that of their parent with the highest degree or level of education.\ng For those 16\u201319 years of age: CDC growth charts were used to calculate percentiles for sex and age. In this table, those < 85th percentile were classified as normal weight; 85\u201394th percentile were classified as overweight; and \u2265 95th percentile were classified as obese; For those \u2265 20 years of age: body mass index < 25 was classified as normal weight; 25\u201329.9 was classified as overweight; and \u2265 30 was classified as obese.\n\nImportance of taste, price, nutrition, and convenience when purchasing food from a fast food restaurant and the grocery store\n | Very important % (n) | Somewhat important % (n) | Not very important% (n) | Not at all important % (n)\nFast food |  |  |  | \nTaste | 76.9 (456) | 20.7 (123) | 1.7 (10) | 0.7 (4)\nConvenience | 56.4 (333) | 35.4 (209) | 6.8 (40) | 1.4 (8)\nPrice | 40.4 (239) | 43.4 (257) | 13.2 (78) | 3.0 (18)\nNutrition | 20.8 (122) | 37.4 (219) | 29.4 (172) | 12.5 (73)\nGroceries |  |  |  | \nTaste | 78.3 (461) | 20.2 (119) | 1.5 (9) | 0 (0)\nConvenience | 34.5 (202) | 47.7 (279) | 14.5 (85) | 3.3 (19)\nPrice | 59.4 (350) | 34.6 (204) | 5.3 (31) | 0.7 (4)\nNutrition | 39.7 (233) | 43.8 (257) | 11.4 (67) | 5.1 (30)\n\n\nMean nutrient contents of meals ordered and consumed by participants in each experimental group\n | Caloriea | Priceb | Calorie + Pricec | Controld | p-valuee\nOrdered |  |  |  |  | \nEnergy, kcal | 873.6 (439.1) | 881.7 (353.6) | 842.3 (425.3) | 827.5 (400.6) | 0.62\nTotal fat, g | 34.3 (19.3) | 35.1 (15.1) | 32.7 (17.0) | 32.5 (18.6) | 0.55\nTotal carbohydrate, g | 110.3 (63.3) | 112.7 (55.8) | 108.0 (65.2) | 105.7 (39.2) | 0.77\nTotal protein, g | 32.4 (14.5) | 30.5 (11.3) | 30.4 (13.) | 29.9 (12.0) | 0.37\nSaturated fat, g | 10.7 (7.6) | 10.4 (5.7) | 10.3 (6.7) | 9.7 (6.7) | 0.61\nDietary fiber, g | 5.0 (2.8) | 5.2 (2.8) | 4.8 (2.7) | 4.6 (2.9) | 0.33\nVitamin C, mg | 27.1 (47.1) | 24.1 (39.0) | 19.9 (32.4) | 27.0 (45.2) | 0.39\nCalcium, mg | 314.2 (233.8) | 270.7 (190.8) | 303.3 (234.7) | 272.7 (213.7) | 0.22\nConsumed |  |  |  |  | \nEnergy, kcal | 804.7 (423.9) | 813.3 (331.6) | 761.0 (356.8) | 739.0 (358.2) | 0.25\nTotal fat, g | 32.1 (19.1) | 32.8 (15.1) | 30.1 (15.3) | 29.6 (16.3) | 0.29\nTotal carbohydrate, g | 100.0 (58.6) | 102.7 (49.8) | 96.0 (53.3) | 92.0 (52.4) | 0.34\nTotal protein, g | 30.4 (14.4) | 28.5 (10.2) | 28.0 (11.4) | 27.8 (11.1) | 0.20\nSaturated fat, g | 9.9 (7.5) | 9.8 (5.5) | 9.4 (6.0) | 8.9 (6.1) | 0.50\nDietary fiber, g | 4.7 (2.8) | 4.8 (2.8) | 4.5 (2.6) | 4.2 (2.7) | 0.22\nVitamin C, mg | 26.2 (46.2) | 22.5 (36.3) | 20.2 (34.3) | 24.5 (39.4) | 0.60\nCalcium, mg | 285.1 (215.0) | 248.1 (163.7) | 265.3 (191.4) | 246.1 (191.0) | 0.26\n\na calorie menu included calorie information for each menu item\nb price menu had price modification (standardized pricing) for food items with more than one portion size option\nc calorie + price menu included calorie information and price modification (standardized pricing) for food items with more than one portion size option.\nd control menu did not include calorie information and had usual food pricing\ne p-value calculated from an analysis of variance analysis\n\nPercent of those in the calorie, price, and calories + price experimental conditions who reported noticing the menu modifications\n | Caloriea% (n) | Priceb% (n) | Calories + pricec% (n)\nNoticed calories | 54.3 (82) | NAd | 58.7 (88)\nNoticed price modification | NAe | 16.1 (23) | 16.7 (25)\n\na calorie menu included calorie information for each menu item\nb price menu had price modification (standardized pricing) for food items with more than one portion size option\nc calorie + price menu included calorie information and price modification (standardized pricing) for food items with more than one portion size option.\nd Not applicable because price was not modified on the menu\ne Not applicable because calories were not included on the menu",
        "label": "low",
        "id": "task4_RLD_test_121"
    },
    {
        "paper_doi": "10.1186/1471-2431-14-187",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Cluster-RCT in 6 districts of Southern Tanzania. INSIST (Improving Newborn Survival in Southern Tanzania) community intervention trial.\n\n\nParticipants: Sample size: 512 women delivering 521 babies completed the final survey (9 pairs of twins).Clusters: 65 intervention and 67 control clusters were randomised. 1 control cluster was lost to follow-up.Individuals: women were eligible for the survey if they had given birth in the previous year and were aged 13-49.\n\n\nInterventions: Target: health systemArm 1: the intervention consisted of prenatal (3) and postnatal (2) counselling home visits to support pregnant women and educate women about birth preparation, delivery and recommended newborn care practices. Counselling tools included picture cards and a doll.Arm 2: pregnant women received standard care.\n\n\nOutcomes: Trial primary outcomes: breastfeeding within an hour of delivery, birth attendants for home deliveries washing hands before childbirth or wearing gloves, and babies fed only breast milk in the first 3 days.Review outcomes reported:Primary: not reportedSecondary: other behaviours promoted during counselling to maximise newborn health, e.g. skilled attendance for childbirth, birth preparedness (for home deliveries), immediate drying and wrapping of the baby, clean cord care and delayed bathing of the baby.\n\nFollow-up: outcome data are based on a post-intervention survey conducted by trained interviewers with women who had given birth in the past 12 months.\n\n\nNotes: Funders: the study was funded by the Bill & Melinda Gates Foundation through the Saving Newborn Lives program of Save the Children (www.savethechildren.org/\n\nprograms/health/saving-newborn-lives/), Unicef, the Laerdal Foundation and the Batchworth Trust.The study was part of INSIST (www.clinicaltrials.gov, NCT01022788), and was approved by the review boards of Ifakara Health Institute, the Medical Research Coordinating Committee of the National Institute for Medical Research, Tanzania Commission for Science and Technology, and the London School of Hygiene and Tropical Medicine, UK\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Background\nIn Sub-Saharan Africa over one million newborns die annually.\nWe developed a sustainable and scalable home-based counselling intervention for delivery by community volunteers in rural southern Tanzania to improve newborn care practices and survival.\nHere we report the effect on newborn care practices one year after full implementation.\nMethods\nAll 132 wards in the 6-district study area were randomised to intervention or comparison groups.\nStarting in 2010, in intervention areas trained volunteers made home visits during pregnancy and after childbirth to promote recommended newborn care practices including hygiene, breastfeeding and identification and extra care for low birth weight babies.\nIn 2011, in a representative sample of 5,240 households, we asked women who had given birth in the previous year both about counselling visits and their childbirth and newborn care practices.\nResults\nFour of 14 newborn care practices were more commonly reported in intervention than comparison areas: delaying the baby\u2019s first bath by at least six hours (81% versus 68%, OR 2.0 (95% CI 1.2-3.4)), exclusive breastfeeding in the three days after birth (83% versus 71%, OR 1.9 (95% CI 1.3-2.9)), putting nothing on the cord (87% versus 70%, OR 2.8 (95% CI 1.7-4.6)), and, for home births, tying the cord with a clean thread (69% versus 39%, OR 3.4 (95% CI 1.5-7.5)).\nFor other behaviours there was little evidence of differences in reported practices between intervention and comparison areas including childbirth in a health facility or with a skilled attendant, thermal care practices, breastfeeding within an hour of birth and, for home births, the birth attendant having clean hands, cutting the cord with a clean blade and birth preparedness activities.\nConclusions\nA home-based counselling strategy using volunteers and designed for scale-up can improve newborn care behaviours in rural communities of southern Tanzania.\nFurther research is needed to evaluate if, and at what cost, these gains will lead to improved newborn survival.\nTrial registration\nTrial Registration Number NCT01022788 (http://www.clinicaltrials.gov, 2009)\nBackground\nNeonatal mortality (death with the first 28\u00a0days of life) most commonly occurs in South East Asia and sub-Saharan Africa, in the first seven days of life, and at home.\nReductions in childhood mortality will stall unless neonatal survival improves.\nHence, the World Health Organization (WHO) and global partners in maternal and child health recommend several low-cost measures including preventive practices, such as clean delivery, thermal protection and early and exclusive breast feeding, and interventions to manage complications, such as resuscitation and management of infections.\nReductions in newborn deaths of 41 to 72% have been predicted from universal coverage of such measures.\nThere is renewed interest in community health workers\u2019 potential to increase coverage of these measures in communities served by primary health facilities.\nEvidence from proof-of-principle trials in Southeast Asia indicates that home-based counselling in pregnancy and shortly after childbirth, combined with community-based treatment or referral of sick babies, can result in higher coverage of recommended newborn care practices, leading to reductions in neonatal mortality of between 34 and 62%.\nReceiving a visit early in the postnatal period has been found to be associated with improved neonatal outcomes.\nTwo trials of community-based maternal and neonatal interventions in Southeast Asia implemented in programme settings have also reported increases in the practice of recommended behaviours, with one reporting a 15% reduction in neonatal mortality.\nIn Africa, a region suffering one million newborn deaths annually, only three similar interventions have been evaluated to date using an experimental design; all in programme settings.\nThe Newhints trial in Ghana assessed the effect of community volunteers visiting women at home during pregnancy and the week after childbirth to promote essential newborn-care practices, weigh and assess babies for danger signs, and refer as necessary.\nThere were increases in the practice of recommended newborn care behaviours, including care-seeking for newborns (77% of sick babies in Newhints zones were taken to a health facility versus 55% in comparison zones) and initiation of breastfeeding within one hour of birth (49% versus 41%), but limited impact on neonatal mortality (risk ratio (RR) 0.9 (95% confidence interval (95% CI) 0.8-1.1).\nIn Malawi, the MaiMwana study evaluated the effects of home-based counselling as well as another intervention - women\u2019s groups \u2013 on maternal, neonatal and child health outcomes, including neonatal mortality rates, using a cluster randomised factorial design.\nIn the whole trial, although areas receiving home-based counselling reported higher levels of exclusive breastfeeding for the first six months (20%) compared to areas without counselling (9%; odds ratio (OR) 2\u00b742 (95% CI 1\u00b748\u20133\u00b796)), there was no evidence of a reduction in neonatal mortality.\nIn South Africa the evaluation of the Goodstart programme in Durban reported nearly a doubling of rates of exclusive breastfeeding (RR 1.92 (95% CI: 1.59\u20132.33)) at 12\u00a0weeks of age following a programme of pregnancy and post-natal home visits by community health workers.\nImproving newborn survival is a current health priority for Tanzania to achieve the fourth millennium development goal.\nIn 2005 Tanzania had a neonatal mortality rate of 32/1000 live births, and the fourth highest number of neonatal deaths in sub-Saharan Africa.\nThe Improving Newborn Survival in Southern Tanzania (INSIST) project was conceived to develop, implement and evaluate a sustainable and scalable behaviour-change community intervention, with the aim of improving newborn survival in a region where neonatal mortality was higher than the national average.\nHere we report the effect of the intervention on newborn care behaviours in the community one year after full implementation.\nMethods\nThe study is detailed in the protocol, and summarised below.\nStudy design and area\nThe INSIST community intervention was implemented as a cluster-randomised trial in six districts of Southern Tanzania.\nBaseline data collected in five of those six districts in 2007 estimated the neonatal mortality rate at 34 per 1,000 live births (unpublished data).\nIntervention funding started in 2008.\nIn 2009 the area comprised 132 wards, 720 villages and 3,428 sub-villages and had a population of around 1.2 million.\nEach ward consists of an average of five villages, approximately 8,000 people, and 260 births per year.\nThe area has a wide mix of ethnic groups.\nCommon occupations include subsistence farming, fishing, and small-scale trading.\nMost rural roads are unpaved, some becoming impassable to motor vehicles in wet weather.\nThe public health system comprises a network of dispensaries, health centres and hospitals offering a varying quality of care.\nThe majority of health facilities are government-run; a health facility survey in the same districts in 2009 recorded four district hospitals, 15 health centres and 156 dispensaries.\nTwo regional hospitals just outside the study area provide referral care.\nImproving the quality of care in dispensaries and health centres for mothers and babies was originally intended to be part of the intervention, but resource limitations restricted this to implementation in just one study district.\nAt the time the study started the majority of women attended antenatal care (88%), around half (57%) delivered in a health facility and no formal system existed for postnatal checks.\nRandomisation\nIn 2009 the research team allocated half of the 132 wards to receive the home-based counselling intervention in addition to routine care (n = 65), and half to continue to receive routine care only (n = 67).\nThe 114 wards in the five districts with baseline data (Newala, Tandahimba, Lindi Rural, Ruangwa and Nachingwea) were randomised using implicit stratification to maximise balance in intervention and comparison groups.\nWe listed the 114 wards in order of district, division, tertile of baseline neonatal mortality, and population, splitting them into 57 pairs.\nWe allocated the wards in each \u2018pair\u2019 to intervention or control using random numbers generated by Microsoft Excel.\nThis is equivalent to 57 tosses of a coin: the scheme has 2**57 = 10**17 realisations, so is highly unconstrained.\nFor the district with no baseline data (Mtwara Rural) we listed the 18 wards by division, and then in alphabetical order within each division, and for each \u2018pair\u2019 of wards within this list we randomised the allocation of the wards in each \u2018pair\u2019 using random numbers generated by Microsoft Excel.\nThis scheme has 2**9 = 512 realisations.\nThere were no exclusion criteria for clusters, households, or women, after randomisation.\nAll villages in intervention wards recruited volunteers to implement the counselling intervention.\nThe nature of the intervention prevented blinding researchers, community members or health staff to the allocation.\nDesign and implementation of the community intervention\nThe intervention, branded Mtunze Mtoto Mchanga (\u201cprotect your newborn baby\u201d in Swahili), designed to be a sustainable and scalable part of the health system, was developed in 2008\u20139 on the basis of formative research.\nNewborn care behaviours selected for targeting through the community intervention were in line with WHO recommended newborn care practices, and jointly agreed with key national stakeholders including regional health leaders, the Ministry of Health and Social Welfare, UNICEF, WHO and the Paediatric Association of Tanzania.\nFollowing development and piloting, in the first half of 2010 over 800 women who volunteered and were not currently involved in other community activities were recruited from and by their communities (two per village in intervention wards).\nThey were trained for five days by their district health teams and followed-up in their villages after starting work as volunteers conducting home visits.\nAll volunteers were working by June 2010.\nVolunteers were supported through quarterly review meetings with district health leaders, monthly contacts with village executive officers, who facilitated the link with the community, and with health facility staff, who provided technical support.\nTwo full-time staff from Ifakara Health Institute co-ordinated the community intervention through planning review meetings, compiling and reviewing monitoring data, and distributing behaviour change communication materials to districts.\nFor every pregnant woman identified in her village, a volunteer was expected to make three visits to her home during pregnancy and two in the early neonatal period, with additional visits for small babies (Table\u00a01).\nThe counselling focussed on one-on-one interaction between the volunteer and mother.\nDiscussion with other family members involved in making decisions about childbirth and newborn care, including fathers and mothers-in-law, was encouraged.\nBehaviour change messages focused on hygiene during delivery, including gloves for those assisting childbirth, immediate and exclusive breastfeeding, and identification of and extra care for small babies.\nAdditional behaviours promoted included: birth preparedness, with messages about the importance of health facility delivery and of preparing clean cloths, soap, gloves, a clean blade, a clean cord tie, and money; delayed bathing of the baby; and putting nothing on the cord.\nAll counselling messages were introduced in pregnancy visits.\nPostnatal visits focused on reinforcing and supporting mothers to implement recommended practices directly applicable to the newborn.\nIn addition, during postnatal visits to babies born at home, volunteers were trained to measure foot size as a proxy for birth weight, to counsel the mother to practise skin-to-skin care for babies who were smaller than usual and to refer very small babies to hospital in the early postnatal visits.\nA picture-based card illustrating the counselling messages was used at each visit and left with each household to enable family members to aid retention of the information or, for those who were not present at the time of the volunteer visit, to receive the counselling messages.\nVolunteers used a locally-made doll (not left with the families) to demonstrate breastfeeding positioning and skin-to-skin care.\nVolunteers regularly reviewed antenatal care registers in order to identify new pregnant women in their village.\nSubsequent visits (except for the first postnatal) were scheduled at each counselling session.\nIf a volunteer first visited a woman late in her pregnancy the gaps between the scheduled visits were reduced accordingly.\nIf visits were missed, counselling messages at a subsequent visit were combined or adapted according to the schedule (Table\u00a01) for the time of the visit in pregnancy or the neonatal period.\nVolunteers asked family members to notify them immediately after the birth in order to conduct postnatal visits.\nTo support early postnatal home visits, facility staff gave mothers a delivery notification slip at discharge after delivery or when a baby was brought to a facility soon after birth, which staff advised families to take to the volunteer.\nSampling\nThe household survey sample size was based on the number of women age 13\u201349 who had given birth in the year preceding the survey, which the baseline survey suggested was approximately 10%.\nWe assumed one woman of reproductive age resided in each household and aimed to visit 40 households per ward (n = 131; one ward did not participate), which gave a sample size of 5,240 households and an estimated 524 reportable deliveries.\nThe study was powered to detect a 15 percentage point change in the practice of early breastfeeding initiation, with 95% confidence and 80% power, including a design effect of 1.5 to account for clustering.\nWe used multi-stage sampling to select households.\nIn stage one, for the districts with baseline data we selected one sub-village from within each ward with probability proportional to size (PPS).\nFor Mtwara Rural district, with no baseline data, we obtained numbers of households in each village, which serves as a proxy for population.\nWe ran the same PPS method, this time selecting villages.\nIn stage two, within each village a sub-village was selected by simple random sampling, and 40 households were selected for interview using a modified EPI-type sampling approach that gave all households an equal chance of selection.\nThe method, used by the research team in previous surveys, is detailed elsewhere and summarised here.\nIn the centre of each sub-village the supervisor threw a pen to choose a random direction.\n(S)he walked in the direction indicated until the edge of the sub-village, sketching a map of and numbering all the households passed.\nOne of these households was selected at random as the first household.\nAt this household, the supervisor threw a pen to choose a random direction, and walked in that direction until (s)he came to another household, which was the second household, and so on until 40 households were counted.\nIf there was a junction in the path, a pen was thrown again to select from the choices available.\nVillages were visited one day before the survey interviewers arrived, and an invitation letter left in each of the selected households.\nData collection, processing and quality control\nThe household questionnaire was developed from tools used by the Demographic and Health Surveys (DHS), Newhints and the baseline household survey.\nQuestions asked to household heads determined his/her occupation, household members and assets.\nFemale residents aged 13\u201349 years at the time of the survey were asked about their birth history.\nThose who had delivered a live baby in the year preceding the survey were asked about their pregnancy, delivery and newborn care practices, and receipt and content of any home-based counselling during pregnancy and the neonatal period.\nThe questionnaire was pre-tested in one sub-village using printed forms.\nPendragon Forms 4.0 software (http://pendragonsoftware.com/) was used to develop a modular questionnaire data entry template.\nFor data collection, the questionnaire was loaded onto Dell Axim X51 personal digital assistants (PDA)s with 64\u00a0MB RAM.\nThe PDA version of the questionnaire was piloted in one ward before the main survey started.\nData were collected in August and September 2011 by trained interviewers who visited selected households, sought written informed consent for survey participation, and recorded responses on PDAs.\nIf household heads refused to participate no other household members were approached.\nIf no household members were present at the time the interviewer visited, the household was visited again later the same day.\nHouseholds were not replaced in cases of refusal or absence.\nLogical checks and skip patterns took place at data entry.\nDigital data records were locked after leaving each household.\nData were downloaded to laptop computers and daily summary reports produced to evaluate completeness and consistency.\nField supervisors undertook a number of quality control activities.\nFirstly, each supervisor accompanied interviewers to three households each day.\nSecondly, they revisited households where interviewers had reported that there were no residents, or the household heads refused participation.\nLastly, two households were revisited daily and a small number of interview questions repeated, the responses to which were compared with those collected by the interviewer.\nData analysis\nData were analysed at the individual level using Stata v12.\nWe calculated means and proportions of respondent characteristics, intervention coverage, delivery characteristics and newborn care behaviours.\nTo estimate the size of the effect of the intervention, logistic regression analysis was used to calculate the ORs of women reporting behaviours in intervention wards compared with those from comparison wards, using svy commands to account for the clustered study design and multi-stage sampling.\nReceipt of the intervention was defined as reporting being visited by a volunteer who had used one of the Mtunze counselling tools (card or doll), to exclude other community health activities.\nA wealth index score, as a measure of socio-economic status, was constructed for each household using the first principal component of ten household assets and characteristics, namely ownership of a radio, bicycle, telephone, poultry, livestock and the home, household connection to an electricity supply, roofing material, cooking fuel and source of income.\nHouseholds were ranked according to this total wealth score and divided into quintiles.\nTo investigate the effect of the intervention on childbirth and newborn care practices we compared mothers\u2019 self-reported behaviours (Table\u00a02) for those who gave birth in the preceding year in intervention and comparison areas, using the allocation given at the sub-village level (intention-to-treat analysis).\nThe primary outcomes were breast feeding within an hour of delivery, birth attendants for home deliveries washing hands before childbirth or wearing gloves, and babies fed only breast milk in the first three days.\nSecondary outcomes were the other behaviours promoted during counselling to maximise newborn health, e.g. skilled attendance for childbirth, birth preparedness (for home deliveries), immediate drying and wrapping of the baby, clean cord care and delayed bathing of the baby.\nAlthough a key behaviour of the intervention, this study was not powered to detect changes in the levels of identification and provision of extra care for small babies.\nWe assumed that childbirth in health facilities took place on a clean surface, that the birth attendant had clean hands or wore clean gloves, and that the cord was cut with a clean blade and tied with a clean thread, so these behaviours were only asked about and reported for home deliveries.\nThe data analyst was masked to the cluster allocation until analysis was complete.\nEthical approval and consent procedures\nThe study was part of INSIST (http://www.clinicaltrials.gov, NCT01022788), and was approved by the review boards of Ifakara Health Institute, the Medical Research Coordinating Committee of the National Institute for Medical Research, Tanzania Commission for Science and Technology, and the London School of Hygiene and Tropical Medicine, UK.\nPrior written consent to approach village leaders was obtained from each district council.\nVillage and sub-village leaders gave verbal consent for data collection to proceed before any households were approached.\nThe head of each household gave written informed consent to participate.\nIn the absence of the household head, another adult resident was approached to give consent.\nIf no adult residents were present the household was revisited later in the day.\nIf a household head refused to participate or adults were absent no replacement households were sought.\nThe consenting adult resident was asked about the members of his/her household and the ownership of household assets.\nAll females age 13\u201349 in consenting households gave their individual verbal informed consent before being interviewed.\nResults\nRespondents\nWe visited 131 of 132 wards, as one ward did not participate in the survey.\nIn each of these wards we visited one sub-village.\nWe visited a total of 5,217 households.\nAlthough 5,240 households were expected, in ten sub-villages data were obtained from fewer than 40 households.\nSome sub-villages were small: in six sub-villages we obtained data from 39 households, in two 38 households, and in one 37 households.\nIn one sub-village data from ten households were lost.\nIn the households visited, 4,989 (96%) household heads agreed to participate, 2,491 in intervention areas and 2,498 in comparison areas (Figure\u00a01).\nThese households comprised a population of 19,475 people, of whom 4,976 were women aged 13\u201349.\nOf these, 4,157 (84%) were available for interview, 4,149 (83%) agreed to participate and 3,199 (78%) had ever delivered a baby.\nThere were 512 women (257 in intervention areas, 255 in comparison areas) from 128 of the sub-villages who had delivered 521 live babies (including nine pairs of twins) since 1st August 2010 and went on to answer detailed questions about their most recent birth, and who comprise the respondents in these analyses.\nBackground characteristics were similar in intervention and comparison areas.\nRespondents had a mean age of 28\u00a0years (standard deviation (sd) 7.2\u00a0years) and had completed a median of seven years of education (range 0\u201317 years, Table\u00a03).\nHousehold heads were mainly from the Makonde ethnic group.\nThere were 12 neonatal deaths in intervention areas and 9 in comparison areas (neonatal mortality rates of 47/1000 live births (95% CI 24\u201391) and 35/1000 live birth (95% 19\u201364) respectively, p = 0.521) in the year preceding the survey.\nCoverage of home-based counselling intervention\nSeventy-three percent (187/257) of women in intervention areas reported receiving a counselling visit, and seven percent (18/255) in comparison areas (Table\u00a04).\nWomen most commonly received their first counselling visit at five months gestation (sd 1.6\u00a0months).\nWomen reported receiving a mean of 2.4 counselling visits (standard error (SE) 0.1).\nMost commonly women reported receiving three visits in pregnancy (87/187, 47% of those reporting any visit, range 1\u20138) and one in the neonatal period (86/187, 46% of those reporting any visit, range 0\u20133).\nOf women reporting receiving visits, 14% (26/187) reported receiving the full complement of at least five visits.\nEighteen percent (34/187) of women reported receiving the first postnatal visit within two days of childbirth; most commonly it was received three days after birth (inter-quartile range 2\u20135 days).\nEleven of the 16 women in intervention areas who reported that their baby was smaller than normal at birth received counselling visits; none reported receiving more than two postnatal visits.\nNone of the 11 women in comparison areas who reported that their baby was smaller than normal at birth received counselling visits.\nChildbirth characteristics and newborn care behaviours\nThe majority of women gave birth in a facility (69%) or with a skilled attendant (71%).\nSimilar proportions of women reported childbirth in a health facility or with a skilled attendant in intervention and comparison areas (both OR 1.4 (95% CI 0.9-2.3)) (Table\u00a05).\nThe majority of women giving birth at home reported that each of the birth preparedness activities was undertaken, with little difference between intervention and comparison areas (Table\u00a06).\nMore women in intervention than comparison areas reported that the cord was tied with a clean thread (69% versus 39%, OR 3.4 (1.5-7.5), all used new thread).\nThere was little evidence of differences between the groups with regard to whether or not the birth attendant had clean hands or the cord was cut with a clean blade.\nA minority of women reported that their babies were dried (33%) or wrapped (20%) within five minutes of delivery, and reported rates were similar in intervention and comparison areas (Table\u00a07).\nThe majority of women reported delaying the baby\u2019s first bath by at least six hours, and this was more commonly done in intervention (81%) than comparison (68%) areas (OR 2.0 (95% CI 1.2-3.4)).\nAlthough breastfeeding within an hour of birth was reported by less than a third of women, with little evidence of a difference between intervention and comparison areas, exclusive breastfeeding in the first three days after delivery was reported by the majority of women in all areas, and more commonly in intervention (83%) than comparison (71%) areas (OR 1.9 (95% CI 1.3-2.9)).\nApplying nothing to the cord was commonly reported, more so in intervention (87%) than comparison (70%) areas (OR 2.8 (95% CI 1.7-4.6)).\nOne hundred and three respondents reported applying substances to the cord; most commonly oil (20%) and herbs (15%).\nNo adverse events as a result of the intervention were reported.\nDiscussion\nThis cluster-randomized controlled trial of a community-based home counselling intervention by volunteers found high levels of coverage for a programme designed for implementation at scale: around three-quarters of women in intervention areas received a counselling visit during pregnancy, and half in the early postnatal period.\nOne year after full implementation, in intervention areas, four recommended newborn care practices were more common than in comparison areas: delaying the baby\u2019s first bath by at least six hours, exclusive breastfeeding in the three days after birth, putting nothing on the cord, and, for home births, tying the cord with a clean thread.\nOther evaluations of similar interventions implemented in the programme setting have reported wide variations in intervention coverage.\nOur coverage of pregnancy visits (76%) was comparable with the Newhints study in Ghana (72%), higher than reported in Pakistan (63%) but lower than in a study from Bangladesh (reported receipt of at least one antenatal visit was >90% at most recent measurement).\nOur coverage of postnatal visits (47%) was lower than reported from Ghana (63%) and in Bangladesh (80%), but higher than reported in Pakistan (24%).\nThe MaiMwana study in Malawi reported 55% of women in intervention areas had received a counselling visit at any time when asked at one month post-delivery, compared to 78% in our study.\nDespite variations in the newborn care practices included in the published trials of similar interventions, some comparisons can be made with our findings.\nExclusive breastfeeding was reported by a majority of women in both the INSIST (during the three days after childbirth) and Newhints (at 26\u201332 days old) studies, with a slightly higher proportion reporting the practice in intervention than comparison areas in both cases (INSIST 83% versus 71%, OR 1.9 (95% CI 1.3-2.9), Newhints 86% versus 80%, RR 1\u2009\u00b7\u200910, 95% CI 1\u2009\u00b7\u200904\u20131\u2009\u00b7\u200916).\nA trial in Uttar Pradesh found much lower rates of pre-lacteal feeding in areas implementing community-based promotion of essential newborn care compared to comparison areas (38% versus 80%, rate ratio 0.49 (95% CI 0.42-0.57)).\nThe proportion of respondents reporting delaying the first bath until at least six hours after delivery was between 12 and 23 percentage points higher in intervention areas in the evaluations in Pakistan, Ghana and our study in Tanzania (Pakistan: 50% versus 27%, p = 0.008.\nGhana: 41% versus 29% rate ratio 1\u2009\u00b7\u200965, 95% CI 1\u2009\u00b7\u200927\u20132\u2009\u00b7\u200913.\nINSIST: 81% versus 68%, OR 2.0, 95% CI 1.2-3.4).\nWhile there is evidence of association between clean cord practices and improved neonatal survival several studies evaluating the impact of combined packages of care did not report changes in cord care practices, or presented only data on the use of a clean instrument to cut the cord; a behaviour already known to be practised by the vast majority of the population in our study area.\nThus there is little evidence from similar trials with which we can compare the increases in rates of applying nothing to the cord and use of clean cord ties found in this study.\nNational data show some similarities with and some differences from our study findings.\nThe Tanzania DHS 2010 found that 68% of deliveries in the Southern Zone took place in facilities, which is comparable with the proportion found in this survey (69%) and reflects a large increase since this study\u2019s baseline survey in 2007 (41%).\nNationally around half of babies were reported to have been breastfed within an hour of birth.\nThis is considerably higher than in our study.\nThe more specific recording of the timing of feeding initiation in our questionnaire compared to that used by the DHS may explain this difference, and has been discussed previously.\nFinding comparable rates of exclusive breastfeeding in the first three days between DHS 2010 and this study, where similar wording was used, supports this argument.\nThe coverage of some behaviours, such as wrapping and drying of the baby within five minutes of birth and breastfeeding within an hour of birth, remained low in intervention areas.\nFormative research in the study area suggested that the attention of birth assistants remained with the mother until the placenta was delivered in home births, which may explain these findings.\nFurther encouragement of facility deliveries, or the presence of an additional birth attendant to assist the baby at home deliveries may help improve thermal care of newborns.\nAlthough the proportion of women delivering at home who reported that their birth attendant had clean hands was higher in intervention than comparison areas, the difference was inconclusive.\nIt may be that the intervention had no effect on hand cleaning practiced by birth attendants, or there was insufficient power to detect the difference in the practice of this behaviour that related only to home births.\nReceipt of counselling visits may have been over-reported in intervention areas.\nVolunteer counsellors were residents of their village and pregnant women were identified to be approached to be visited for counselling through antenatal registers: women who chose not to receive counselling visits were still likely to have been aware of the programme in their community and may have reported receiving a counselling visit.\nBy randomising at ward level, the study reduced intervention \u2018leakage\u2019 to comparison areas: although 18 (7%) women in comparison areas reported receiving an intervention counselling visit, only four (2%) had a study counselling card.\nThese could have been women living close to intervention areas, or women who had moved to be with relatives for childbirth.\nSome of the reported counselling visits in both intervention and comparison areas could have been linked to other community activities.\nThis study showed improvements in newborn care behaviours with home-based counselling in programme settings, but the changes may have been greater with improved intervention implementation.\nFor example, early postnatal visits are associated with improved neonatal outcomes, but in our study fewer than half of women in intervention areas received a postnatal visit and a fifth of women reporting receiving any counselling visit were visited within two days of childbirth.\nThere are many possible reasons for this.\nVolunteers were trained by staff at the district level, so volunteers\u2019 knowledge of the counselling programme may have varied.\nMotivational activities and a supervision programme were included in the design of the counselling programme, which are known to help to maintain volunteer work standards and coverage.\nHowever, these may have not been sufficient or novel enough to sustain volunteers over long time periods.\nThese were reviewed regularly during implementation and steps taken to improve them where needed.\nA substantial change to the supervision procedures in 2011 increased the frequency of supervision, and could improve counselling coverage and quality, although it was implemented too late for its effect to be measured by this evaluation.\nIn the context of increasing health facility deliveries, early postnatal visits may have been difficult to complete and measures to facilitate them ineffective.\nFor analysis we assumed that facility birth would involve adequate basic obstetric care, including clean delivery surfaces, the cord being cut and tied with clean instruments, and immediate wrapping and drying of the baby.\nHowever health system weaknesses mean this may not always be so.\nEvidence from a health facility survey in the study area found many instances of missing or recently out of stock items for delivery, including examination gloves and cord ligatures, particularly in dispensaries.\nSuch weaknesses and the limited quality improvement work undertaken alongside the community intervention means care of mothers and newborns in health facilities is likely to remain a barrier to improved newborn survival.\nThere is some evidence from sub-Saharan Africa that clean delivery practices, early and exclusive breastfeeding and early skin-to-skin care are associated with improved newborn survival.\nWhile the target behaviours, and content and mode of delivery of the messages would need to be adapted to the local setting in order to implement a similar intervention beyond this region of Tanzania, this study contributes to the small but growing robust evidence base that home-based counselling implemented at scale in the community can improve newborn care practices in low-resource African settings with high levels of neonatal mortality.\nA meta-analysis of trials of home-visit strategies in Africa and Asia found overall a 12% (95% CI 5\u201318) reduction in neonatal mortality.\nTherefore future research needs to establish if the behaviour change reported here is sufficient to reduce neonatal mortality.\nFurthermore, the impact of the number and timing of counselling visits on mortality rates should be assessed.\nLastly, the quality, acceptability and cost effectiveness of the counselling intervention need to be evaluated to understand the process of behaviour change and sustainability of the intervention.\nConclusions\nA home-based counselling strategy to promote recommended newborn care implemented by volunteers and designed for scale within the health system can improve newborn care in rural communities in southern Tanzania.\nFurther research is needed to evaluate if, and at what cost, these gains will lead to improved newborn survival.\nTrial profile.\n\nFocus and timing of home visits for INSIST community intervention\nVisit | Timing | Key behaviours promoted | Additional behaviours promoted | Equipment\n1 | As soon as pregnant woman identified | \u2022 Birth attendant should wash hands and wear gloves | \u2022 Birth preparedness: preparing for facility delivery and saving money; and preparing in case of unexpected home delivery, preparing clean cloths, soap, clean blade for cutting & clean thread for tying cord, gloves for birth attendant | Counselling card\n2 | Four weeks after visit 1 | \u2022 Early and exclusive breastfeeding | \u2022 Check on birth preparedness issues from previous visit | Counselling card\n3 | At the beginning of 9th month of gestation | \u2022 Early and exclusive breastfeeding including position | \u2022 Check on birth preparedness issues from previous visits | Counselling card with doll\n\u2022 In case of home birth: | \u2022 Warmth: immediate drying and wrapping, delayed bathing, keep the vernix\n\u2003\u25cb Birth attendant should wash hands and wear gloves, including while tying and cutting the cord\n\u2003\u25cb Identification of low birth weight babies using foot size as a proxy | \u2022 Danger signs for sick newborns\n\u2003\u25cb Immediate referral for very small or premature babies, and those who don\u2019t cry | \u2022 In case of home birth, cord should be cut with clean blade and tied with clean thread\n\u2003\u25cb Skin to skin care for small babies\n4 | Day of delivery | \u2022 Observe breastfeeding and counsel on positioning | \u2022 Check on warmth and knowledge of danger signs (as above) | Counselling card \u2013 measure foot size using scale\n\u2022 Reminder of exclusive breastfeeding | \u2022 Put nothing on cord\n\u2022 In case of home birth:\n\u2003\u25cb Identification of low birth weight babies using foot size as a proxy\n\u2003\u25cb Immediate referral for very small or premature babies\n\u2003\u25cb Skin to skin care for small babies\n5 | Third day after delivery | \u2022 Observe breastfeeding and counsel on positioning | \u2022 Put nothing on the cord | Counselling card\n\u2022 Reminder of exclusive breastfeeding\n1st extra visit for small baby | Day after visit 5 | \u2022 Skin to skin until the baby doesn\u2019t want to be carried skin to skin | \u00a0 | Counselling card\n2nd Extra visit for small baby | Day after visit 6 | \u2022 Skin to skin until the baby doesn\u2019t want to be carried skin to skin | \u00a0 | Counselling card\n\n\nOutcome measures\nOutcome category | Practice | Timing of practice | Measured for which babies?\nPrimary | Baby breastfed within one hour of birth | Newborn care | All\nBirth attendant washed hands with soap before childbirth or wore gloves | Childbirth | Home birth\n\u00a0 | Baby fed only breast milk in the first three days | Newborn care | All\nSecondary | Childbirth in a health facility | Childbirth | All\n\u00a0 | Childbirth with a skilled attendant | Childbirth | All\n\u00a0 | Prepared soap | Childbirth | Home birth\n\u00a0 | Prepared new or washed cloth for drying baby | Childbirth | Home birth\n\u00a0 | Prepared cloth or mat for childbirth | Childbirth | Home birth\n\u00a0 | Cleaned floor where childbirth to take place | Childbirth | Home birth\n\u00a0 | Prepared new or washed cloth for wrapping | Childbirth | Home birth\n\u00a0 | Had plan in case of emergency childbirth | Childbirth | Home birth\n\u00a0 | Attendant had clean hands during childbirth | Childbirth | Home birth\n\u00a0 | Baby had cord cut with new or sterilised blade | Newborn care | Home birth\n\u00a0 | Baby had cord tied with new thread | Newborn care | Home birth\n\u00a0 | Baby dried <5\u00a0minutes after birth | Newborn care | All\n\u00a0 | Baby wrapped <5\u00a0minutes after birth | Newborn care | All\n\u00a0 | Baby bathed at least six hours after birth | Newborn care | All\n\u00a0 | Baby had nothing applied to umbilical cord | Newborn care | All\n\n\nRespondent characteristics\nCharacteristic | Intervention (I) wards (N = 257) | Comparison (C) wards (N = 255) | Percentage points difference (I-C)\n\u00a0 | n | % | n | % | \u00a0\nWealth quintile | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u20091 (Poorest) | 52 | 20 | 49 | 19 | 1\n\u2003\u20092 | 52 | 20 | 49 | 19 | 1\n\u2003\u20093 | 53 | 21 | 50 | 19 | 2\n\u2003\u20094 | 46 | 18 | 58 | 23 | -5\n\u2003\u20095 (Wealthiest) | 52 | 20 | 51 | 20 | 0\nEthnic group of household head | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u2009Makonde | 140 | 54 | 141 | 55 | -1\n\u2003\u2009Mwera | 82 | 32 | 72 | 28 | 4\n\u2003\u2009Makuwa | 8 | 3 | 13 | 5 | -2\n\u2003\u2009Yao | 10 | 4 | 7 | 3 | 1\n\u2003\u2009Other | 17 | 7 | 22 | 9 | -2\nYears of Education | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u20090-6 | 96 | 37 | 103 | 40 | -3\n\u2003\u20097 | 142 | 55 | 141 | 55 | 0\n\u2003\u20098-17 | 19 | 7 | 11 | 5 | 2\nAge | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u200915-19 | 32 | 12 | 39 | 15 | -3\n\u2003\u200920-24 | 70 | 27 | 83 | 33 | -6\n\u2003\u200925-29 | 58 | 23 | 45 | 18 | 5\n\u2003\u200930-34 | 40 | 16 | 50 | 20 | -4\n\u2003\u200935-39 | 38 | 15 | 26 | 10 | 5\n\u2003\u2009> = 40 | 19 | 7 | 12 | 5 | 2\n\n\nCoverage and implementation of home-based counselling intervention\nIntervention event | Intervention wards | Comparison wards | Percentage points difference (I-C) | OR (95% CI) | p\nn (N) | % | N (N) | % | \u00a0 | \u00a0\nWoman ever received an Mtunze visit | 187 (257) | 73 | 18 (255) | 7 | 66 | 35.2 (19.4-63.6) | <0.001\nWoman received an Mtunze visit during pregnancy | 187 (257) | 73 | 18 (255) | 7 | 66 | 35.2 (19.4-63.6) | <0.001\nWoman received an Mtunze visit after childbirth | 118 (257) | 46 | 8 (255) | 3 | 43 | 26.2 (11.6-59.3) | <0.001\n\n\nChildbirth characteristics, all deliveries\nChildbirth characteristic | Intervention wards | Comparison wards | Percentage points difference (I-C) | OR (95% CI) | p\nn (N) | % | n (N) | % | \u00a0\nIn a health facility | 187 (256) | 73 | 166 (254) | 65 | 8 | 1.4 (0.9-2.3) | 0.14\nWith a skilled attendant | 189 (255) | 74 | 171 (254) | 67 | 7 | 1.4 (0.9-2.3) | 0.16\n\n\nChildbirth practices, home deliveries\nPractice | Intervention wards | Comparison wards | Percentage points difference (I-C) | OR (95% CI) | p\nn (N) | % | n (N) | % | \u00a0\nPrepared soap | 63 (68) | 93 | 80 (88) | 91 | 2 | 1.3 (0.4-4.0) | 0.70\nPrepared new/washed cloth for drying baby | 61 (68) | 90 | 82 (88) | 93 | -3 | 0.6 (0.2-2.0) | 0.43\nPrepared cloth or mat for birth | 51 (53) | 96 | 69 (73) | 95 | 1 | 1.5 (0.2-10.2) | 0.69\nCleaned floor where birth to take place | 50 (67) | 75 | 66 (88) | 75 | 0 | 1.0 (0.5-2.0) | 0.96\nPrepared new/washed cloth for wrapping baby | 65 (69) | 94 | 80 (88) | 91 | 3 | 1.6 (0.5-5.3) | 0.42\nHad plan in case of emergency delivery | 49 (68) | 72 | 63(88) | 72 | 0 | 1.0 (0.5-2.2) | 0.95\nAttendant washed hands before childbirth or wore gloves | 55 (68) | 81 | 65 (88) | 74 | 7 | 1.5 (0.6-3.6) | 0.37\nBaby had cord cut with new or sterilised blade | 65 (68) | 96 | 81 (88) | 92 | 4 | 1.9 (0.5-7.8) | 0.39\nBaby had cord tied with clean thread | 46 (67) | 69 | 34 (87) | 39 | 30 | 3.4 (1.5-7.5) | 0.003\n\n\nNewborn care practices, all deliveries\nPractice | Intervention wards | Comparison wards | Percentage points difference (I-C) | OR (95% CI) | p\nn (N) | % | n (N) | %\nBaby dried <5\u00a0minutes after birth | 84 (253) | 33 | 85 (253) | 34 | -1 | 1.0 (0.6-1.5) | 0.89\nBaby wrapped <5\u00a0minutes after birth | 56 (255) | 22 | 45 (254) | 18 | 4 | 1.3 (0.8-2.1) | 0.26\nBaby bathed at least 6\u00a0hours after birth | 207 (255) | 81 | 173 (254) | 68 | 13 | 2.0 (1.2-3.4) | 0.007\nBaby breastfed within 1\u00a0hour | 68 (249) | 27 | 53 (251) | 21 | 6 | 1.4 (0.9-2.1) | 0.11\nBaby fed only breast milk in first 3\u00a0days | 206 (249) | 83 | 178 (250) | 71 | 12 | 1.9 (1.3-2.9) | 0.002\nNothing applied to umbilical cord | 222 (255) | 87 | 179 (255) | 70 | 17 | 2.8 (1.7-4.6) | <0.001\n",
        "label": "low",
        "id": "task4_RLD_test_494"
    },
    {
        "paper_doi": "10.1371/journal.pone.0028957",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Individually RCTDates of trial: between July and January, from 2000 to 2003.\n\n\nParticipants: 237 individuals aged from 3 to 70 years, in 5 villages for Afghan refugees in Pakistan.Inclusion: > 2 years of age, P. falciparum mono-infection, confirmed by slide, will be resident during entire follow-up period.Exclusions: pregnancy, signs of severe malaria, report of antimalarial drug in past 21 days, other serious disease\n\n\nInterventions: CQ: 3 days 25 mg/kg.CQ+PQ: CQ as in 1; PQ on day 3 (0.5 mg/kg).SP: 25(S)/1.25(P) mg/kg in single dose.SP+PQ: SP as in 3; PQ on same day (0.5 mg/kg).\n\n\nOutcomes: Clinical treatment failure (PCR non-adjusted and adjusted).Gametocytes on day 8.Gametocyte density on days 1 to 8 of follow-up.Genotyping of resistant strains for CQ and SP-specific mutations.\n\n\nNotes: Also included CQ + AS and SP + AS arms, compared with CQ +- PQ and SP +- PQ arms, respectively\n\n",
        "objective": "To assess the effects of single dose or short\u2010course PQ (or an alternative 8AQ) alongside treatment for people with P. falciparum malaria.",
        "full_paper": "Introduction\nAntimalarial resistance has led to a global policy of artemisinin-based combination therapy.\nDespite growing resistance chloroquine (CQ) remained until recently the official first-line treatment for falciparum malaria in Pakistan, with sulfadoxine-pyrimethamine (SP) second-line.\nCo-treatment with the gametocytocidal primaquine (PQ) is recommended for transmission control in South Asia.\nThe relative effect of artesunate (AS) or primaquine, as partner drugs, on clinical outcomes and gametocyte carriage in this setting were unknown.\nMethods\nA single-blinded, randomized trial among Afghan refugees in Pakistan compared six treatment arms: CQ; CQ+(single-dose)PQ; CQ+(3 d)AS; SP; SP+(single-dose)PQ, and SP+(3 d)AS.\nThe objectives were to compare treatment failure rates and effect on gametocyte carriage, of CQ or SP monotherapy against the respective combinations (PQ or AS).\nOutcomes included trophozoite and gametocyte clearance (read by light microscopy), and clinical and parasitological failure.\nFindings\nA total of 308 (87%) patients completed the trial.\nFailure rates by day 28 were: CQ 55/68 (81%); CQ+AS 19/67 (28%), SP 4/41 (9.8%), SP+AS 1/41 (2.4%).\nThe addition of PQ to CQ or SP did not affect failure rates (CQ+PQ 49/67 (73%) failed; SP+PQ 5/33 (16%) failed).\nAS was superior to PQ at clearing gametocytes; gametocytes were seen on d7 in 85% of CQ, 40% of CQ+PQ, 21% of CQ+AS, 91% of SP, 76% of SP+PQ and 23% of SP+AS treated patients.\nPQ was more effective at clearing older gametocyte infections whereas AS was more effective at preventing emergence of mature gametocytes, except in cases that recrudesced.\nConclusions\nCQ is no longer appropriate by itself or in combination.\nThese findings influenced the replacement of CQ with SP+AS for first-line treatment of uncomplicated falciparum malaria in the WHO Eastern Mediterranean Region.\nThe threat of SP resistance remains as SP monotherapy is still common.\nThree day AS was superior to single-dose PQ for reducing gametocyte carriage.\nTrial Registration\nClinicalTrials.gov bold>\nIntroduction\nAntimalarial drug resistance is an ongoing threat to malaria control.\nDespite widespread documented resistance, chloroquine remained widely used in Pakistan and Afghanistan for first line treatment of falciparum malaria at the time of this study (2000\u20132003).\nIn many other parts of the world including Pakistan and Afghanistan the antifolate drug sulfadoxine-pyrimethamine (SP) is comparatively effective against falciparum malaria in contrast to the situation in South East Asia and many African settings.\nIn Pakistan and Afghanistan transmission is seasonal, strong immunity seldom develops, infected individuals are symptomatic and it is thought that the majority seek treatment.\nThese conditions expose the majority of infections to antimalarial drugs and would be expected to exert strong selection for resistance.\nSP has a long plasma half-life which may further contribute to selection.\nIn such areas of low seasonal transmission the operational effectiveness of SP before resistance arises appears to be short.\nWhilst current effectiveness remains comparatively good in the region low level in vivo resistance to SP has been demonstrated.\nWide-scale adoption of effective SP based combinations, rather than SP monotherapy, which was being considered as an option to replace chloroquine at the time of the study, could delay the selection of resistance and prolong the useful life of SP across the subcontinent.\nIn areas of low or medium endemicity co-treatment of infections with a gametocytocidal drug may help to reduce transmission and may be particularly important when considering strategies for malaria elimination.\nIn South Asia, the standard policy has been to co-treat with a single dose of primaquine to reduce gametocyte carriage.\nPrimaquine is known to have poor efficacy as a direct treatment but has repeatedly been shown to be highly gametocytocidal.\nReduction in gametocyte carriage and infectivity to mosquitoes after artesunate treatment is widely documented.\nIt appears to be due to a combination of rapid clearance of asexual stages, direct activity on immature gametocytes (either killing them or preventing their maturation) and possibly reduction in the infectivity of mature gametocytes.\nUse of ACTs has been shown to reduce transmission and is recommended for epidemic response, as too is primaquine.\nThere is increasing interest in using primaquine with ACTs to accelerate gametocyte clearance.\nSP treatment of clinical infections can lead to high gametocyte loads which persist for 10 days or more with latent gametocytaemia detected by PCR for up to a year.\nA direct comparison of artesunate and primaquine on treatment failure and gametocyte carriage has not been conducted before.\nMany Afghan refugee communities in Pakistan have a history of falciparum malaria and are prone to outbreaks.\nA trial examining the comparative efficacy of chloroquine and SP combined with either artesunate or primaquine was therefore carried out in refugee villages in North-western Pakistan.\nMaterials and Methods\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nStudy area and population\nThe study recruited patients from five Afghan refugee villages within an 80 km radius of Peshawar, Khyber Province (formerly North West Frontier Province), Pakistan.\nThe 5 villages were established in the early 1980s and each has a history of falciparum transmission.\nThree (Adizai, Naguman and Yakka Ghund) are situated on the banks of the Kabul river, and two (Mohammed Khoja and Kotki) are situated to the south of Peshawar in Kohat district.\nMalaria transmission is seasonal with vivax malaria occurring from March to November and falciparum from July to December.\nParticipants were Afghan refugees who were permanently resident in the Pakistan villages.\nA minority might have acquired their infections in Afghanistan, but admission criteria required 4\u20136 weeks of follow-up which would have excluded the more regular travellers.\nStudy sites\nThe study was conducted through three sites with well-managed clinics run by Non-Governmental Organisations (NGO) and government agencies.\nSite 1 was Adizai village which also served the population of nearby Naguman.\nSite 2 was Yakka Ghund village.\nSite 3 was Kotki which also serviced Mohammed Khoja village.\nHealth staff from the NGO HealthNet TPO were seconded to each clinic to manage the recruitment and follow-up of patients.\nThe study took place over three malaria seasons from 2000 to 2003.\nIn the first season only site 1 was used, in the second season site 2 was added, and in the third season site 3 was added.\nThe stepped inclusion of sites resulted from unexpectedly low recruitment rates at the first study site, and is summarized in Table 1.\nAn earlier in vivo survey across all refugee camps along the length of the 1000 km North-South axis of Khyber Province showed little heterogeneity in resistance frequency to chloroquine and SP.\n(Note that SP is the only drug name abbreviated throughout the text.\nAbbreviations for chloroquine (CQ), primaquine (PQ) and artesunate (AS) are used when referring to specific study arms (e.g. CQ+PQ) but not at other points in the text.)\nStudy design and procedures.\nIndividuals presenting with clinical symptoms and diagnosed microscopically with falciparum malaria were referred to trial staff for further assessment.\nInclusion criteria were: over two years of age, not pregnant, P. falciparum mono-infection, more than 1 asexual parasite per 10 fields, no other serious disease, resident in the refugee village for the full period of follow-up, no verbal report of antimalarial use during the last 21 days, and no signs of severe malaria.\nIndividuals meeting the inclusion criteria and giving informed consent were allocated using randomization tables stratified by age and sex to the following treatment arms: chloroquine (CQ1) (25 mg/kg) over 3 days; CQ (25 mg/kg) over 3 days plus primaquine (PQ) (0.5 mg/kg) on the last day of treatment; CQ (25 mg/kg) over 3 days plus artesunate (AS) (4 mg/kg per day) over 3 days; sulfadoxine (25 mg/kg) and pyrimethamine (1.25 mg/kg) (SP) on day 0; SP (25:1.25 mg/kg) plus PQ (0.5 mg/kg) on day 0; SP (25:1.25 mg/kg) on day 0 plus AS (4 mg/kg per day) over 3 days.\nThe primaquine co-treatment regimens were based on recommendations for use of primaquine as a gametocytocide treatment for falciparum malaria.\nOn treatment days when any specific arm did not require a partner drug, a placebo was substituted.\nPatients were not allocated to all six treatment arms at each site (Table 1).\nDuring years two and three, at any one site, patients were allocated to either one of the three SP treatment arms or one of the three chloroquine treatment arms; only in year one were patients allocated to all six treatment arms in the original site.\nFor each of the study sites patients were stratified by age and sex and assigned consecutive patient numbers at enrolment.\nTreatment groups were pre-assigned to patient numbers using simple randomization lists generated using Excel (Microsoft Corp., Seattle, USA).\nThe assignment of treatment group to patient number was concealed until after enrolment.\nThe random allocation sequence was generated by two of the investigators (KK, MR) neither of whom enrolled or assessed patients.\nThe brands and manufacturers/suppliers were: SP (Fansidar \u00ae, Roche), chloroquine (Nivaquine \u00ae, Beacon), primaquine (IDA), artesunate (Plasmotrim \u2122, Mepha).\nAlthough chloroquine resistance was known to exist in the study area, chloroquine was the first line treatment policy of the Government of Pakistan and the UN refugee agency (UNHCR) at the time and was therefore included.\nThe trial manager, who allocated the treatment arms and administered treatment, was not blind to the study drugs and allocations.\nRecruited patients, microscopists and health workers were partially blinded; full blinding was not achieved given (i) the appearance of the different drugs, and (ii) the different times of follow up for the SP arms compared to the chloroquine arms.\nPatients were given directly observed treatment (by the trial manager), monitored for 30 minutes and re-dosed if vomiting occurred.\nHealth workers then recorded the other day 0 biomedical parameters.\nPatients were asked to return on each day of treatment (days 0, 1, 2) and on days 3, 7, 14, 21 and 28 post treatment.\nPatients in the three SP arms were also followed up on days 35 and 42 for detection of late recrudescence.\nOn each day of follow up thick and thin smears were taken, clinical symptoms recorded and blood spots collected on Whatman No. 1 filter paper for parasite typing and differentiation of recrudescent from new infections.\nHaematocrits and blood spots were taken either on the day of failure or on day 28.\nPatients not presenting at the clinic were followed up at home.\nCriteria for withdrawal were reported administration of additional anti-malarial drugs (protocol violation), emergence of any concomitant febrile illness that interfered with outcome classification (including a non-falciparum malaria, for which the withdrawn study patient would be given appropriate first line treatment), parasitaemia still present on day 7, or signs of severe malaria developing before completing the initial treatment course.\nPatients found to be parasitaemic on any day after day 3 were treated with SP (the official second line treatment) or with SP and mefloquine (Fansimef, Roche) for those whose initial treatment was SP based.\nSP resistant parasites were sensitive to the mefloquine component of Fansimef, which was used in lieu of mefloquine monotherapy which was unavailable in the study area.\nThose who developed severe malaria, severe anaemia or other complications were referred to Khyber Hospital, Peshawar for treatment.\nLaboratory tests.\nThick and thin blood smears were stained with 2% Giemsa solution.\nAll slides were read on the day of collection by a microscopist based at each site.\nDifferential diagnosis of vivax and falciparum (trophozoites and gametocytes) was by examination of thick and thin smears according to standard microscopy methods.\nTrophozoites and gametocytes were counted against 200 white blood cells (WBC) from the thick blood smear on the assumption of a WBC count of 8000/\u00b5l.\nA smear was declared negative if no parasites were seen after examining 100 fields.\nSlides from sites 2 and 3 were re-examined for accuracy by the site 1 microscopist.\nComparison was made between the parasite counts made by the three microscopists.\nThe mean variation in parasite count for trophozoites and gametocytes was less than 5%.\nOwing to electricity supply deficiency, the haematocrit microcentrifuge was only used at sites 1 and 3.\nOutcome measures.\nPatients completed the trial if treatment was administered fully and all follow-up appointments conducted, or if they failed treatment on any day of follow-up.\nThe primary endpoint of the trial was any clinical or parasitological failure up to day 28, although a subset of patients in the SP treatment arms was also followed until day 42.\nPatient outcomes were classified under the WHO parasitological classification system of sensitive (S) or resistant (RI, RII, RIII) infections with those whose outcomes were classified as S (sensitive) being treatment successes and those with outcomes of RI, RII or RIII being classified as treatment failures.\nThis classification system was still commonly used at the time of the study and the ethical clearance for this study (received in 2000) was based on a protocol using this classification system.\nPatients were also classified against the newer WHO treatment outcome system.\nPresenting the outcomes under both of these systems allows better comparability with data past and present.\nFor the WHO classification system standard definitions were applied: adequate clinical and parasitological response (ACPR), early treatment failure (ETF) or late treatment failure (LTF) which incorporated standard definitions of late clinical and parasitological response.\nThose cases with ETF or LTF were classified as treatment failures, and ACPR was classified as treatment success if they completed the follow-up period.\nOther parameters examined were time to fever resolution (with fever defined as axillary temperature \u226537.5\u00b0C), asexual parasite clearance, gametocyte clearance and gametocyte carriage on or after day 7.\nMolecular characterisation.\nPCR genotyping was conducted on a subset of samples to distinguish recrudescent from new infections using the protocol described by Brockman et al.\nP. falciparum genes msp-1, msp-2 and glurp were genotyped according to polymorphisms present at variable loci on day 0 and day of failure.\nApproximately 50% of cases were available for genotyping.\nPCR corrected outcomes were applied as a secondary analysis using redefined outcomes based on the number and frequency of true recrudescent infections and reclassifying cases with new infections as treatment successes, excluding those with indeterminate outcomes or negative PCR results.\nPCR testing was performed at the Shoklo Malaria Research Unit, Thailand.\nWe also conducted an analysis at the London School of Hygiene and Tropical Medicine (LSHTM), UK, for known resistance-associated mutations to chloroquine and antifolate drugs.\nSamples collected at enrolment were tested for known mutations in the pfcrt, pfmdr1, pfdhfr and pfdhps genes.\nPCR and sequence specific oligonucleotide probe assays were used to analyse nucleotide polymorphisms at pfcrt codon 76, pfmdr1 codons 86 and 184, pfdhfr codons 16, 51, 59 and 108 and pfdhps codons 436, 437, 581 and 613.\nSample size and data analysis\nThe sample size required to detect a difference in treatment failure or gametocyte carriage with 80% power and 95% confidence (two-sided significance level for type 1 error of 0.05) were calculated using the following predictions.\nIn the chloroquine arms (i) the estimated frequency of failure in chloroquine monotherapy arm was 30% and in each of the combination arms (CQ+PQ or CQ+AS) it was 10% or less; (ii) the estimated proportion of gametocyte positive individuals after 7 days in the chloroquine arm was 50% and in each of the combination arms it was 25%.\nIn the SP arms (i) the estimated frequency of failure in the SP arm was 10% and in each of the combination arms (SP+PQ or SP+AS) it was 1%; (ii) the estimated frequency of gametocyte positive patients after 7 days in the SP arm was 50% and in each combination arm it was 25%.\nThe estimated samples sizes were 64 per arm in the chloroquine arms and 121 per arm in the SP arms.\nFor gametocyte carriage the required sample sizes were 65 per arm.\nThe target sample sizes for the study were 65 for the chloroquine arms and 121 for the SP arms.\nTo allow for 15% loss to follow up the targets for recruitment were 76 for the chloroquine arms and 142 for the SP arms.\nThe primary aims of the study were to evaluate: 1) the relative efficacy of the combination drugs in providing parasite clearance with no recrudescence compared to monotherapy, and 2) the effect on gametocyte clearance.\nThe secondary aim was to determine the proportion of individuals in each arm with classifiable treatment failure.\nThe primary outcome was the proportion of individuals in each treatment group classified as a treatment failure during the 28 days follow-up compared to the respective monotherapy.\nThe odds of treatment failure between each of the study groups at day 28 were estimated after adjusting for potential confounders (age, sex, parasitaemia, PCV, study site) using logistic regression analysis or Mantel-Haenszel \u03c72 test.\nTreatment failure was also analysed using the definitions of ACPR and ETF or LTF, and parasitological classifications using the S-R scale in addition to the PCR corrected analysis.\nTime to event data (time to treatment failure) were analysed with Kaplan Meier survival analysis.\nFor assessing potential effects on gametocyte carriage, the primary outcome was the presence or absence of gametocytes on day 7 with a secondary analysis examining the presence or absence of gametocytes and the geometric mean gametocyte density on each day of follow-up.\nAll data were entered in Microsoft Excel (1997) by one data clerk and the database checked patient by patient against paper records by a second data clerk.\nAnalysis was performed using STATA 10.0 (STATA Corp, College Station, TX, USA).\nEthics Statement\nThe study protocol was approved by the ethics committees of the Pakistan Medical Research Council and LSHTM, UK.\nAll patients or their guardians gave written informed consent to participate.\nThe trial was registered under clinicaltrials.gov [NCT00959517].\nResults\nRecruitment and follow-up\nA total of 355 cases of microscopically confirmed falciparum malaria were enrolled into the study between July 2000 and December 2002.\nTable 2 shows enrolment characteristics and Figure 1 shows the trial profile.\nPatients were recruited to the chloroquine arms in line with recruitment targets.\nInsufficient numbers were recruited to the SP arms.\nThis was due to considerably lower than expected malaria cases occurring in these locations during the study period (in contrast to the years leading up to the study).\nAdditional sites were added over the three-year study period to increase recruitment but it was not possible to continue the trial for longer because of resource constraints.\nchloroquine arms met the recruitment targets because of the site allocations and unexpected pattern of case loads in the third year.\nCharacteristics of patients recruited into the 3 chloroquine arms were broadly similar, as were the three SP arms.\nThe SP+AS arm showed slightly higher asexual parasite densities on enrolment than other groups (Table 2).\nOf 355 patients enrolled, 38 (10.7%) were either withdrawn or lost to follow up leaving 317 for possible inclusion (Figure 1).\nOf these 292 (82%) were evaluable for parasitological outcomes because 25 patients classified as RII parasitological failure did not meet the definition of clinical failure, i.e. they did not have fever on day 3.\nLikewise, 308 (87%) were evaluable for clinical outcomes because 9 patients who had classifiable Early Treatment Failure did not fit the definition for parasitological failures (they did not reach day 7 of the trial).\nOf the 266 patients followed up for 42 days, 221 (83%) were evaluable for clinical and parasitological outcomes.\nAll treatments were well tolerated and no severe or serious adverse events were recorded during the study.\nClinical Outcomes.\nThe 28-day failure rate in the CQ and CQ+PQ arms was 81% and 73% respectively (Table 3, Figure 2A).\nThe addition of artesunate improved the treatment response, with a treatment failure rate of 27%.\nThe failure rates in the SP groups were 16% or less (Table 3) with the combination of SP+AS having the lowest failure rate (1/41 [2.4%]).\nNeither SP+AS nor SP+PQ arms were significantly different to the SP arm at the 28-day end point (Figure 2B).\nNone of the potential confounding factors examined (age, sex, parasitaemia, study site, PCV) were associated with treatment outcome so adjustments were unnecessary in the final regression analysis.\nCrude odds ratios are presented.\nClinical and parasitological outcomes are shown in Table 4.\nOnly 32/133 (24%) of the individuals who failed had fever on the day of failure; the majority of \u2018late treatment failures\u2019 therefore fell into the category of \u2018late parasitological failure\u2019.\nSome recent history of fever may have gone unreported.\nTrophozoite clearance time (days until negative smear) was lowest for the artesunate combination arms.\nAddition of primaquine to either chloroquine or SP did not appear to affect clearance times.\nClearance times for each arm (median (interquartile range)) were: CQ, 3 days (2\u20137); CQ+PQ, 3 days (2\u20137); CQ+AS, 2 days (1\u20132); SP, 2 days (2\u20133); SP+PQ, 2 days (1.5\u20133); SP+AS, 1 day (1\u20132).\nOf the 133 failures, 79 (47%) matched pairs were available for PCR evaluation of MSP-1, MSP-2 and GLURP.\nMatched pairs were collected on day 0 and on the day of failure.\nOutcomes of the PCR are shown in Table 5, which excludes those with indeterminate results.\nMost re-infections took place at one site (Site 3, Yakka Ghund) where malaria transmission was higher than in the other villages.\nTreatment outcomes were reclassified for those failures where matched pairs were available and used to give adjusted failure rates for the sample (Table 5).\nAlthough the number of failures in the chloroquine arms was reduced by the correction, the failure rates remained high at >50% for chloroquine monotherapy.\nThe PCR adjusted treatment failure rate of the CQ+AS combination (9%) was considerably lower that the in vivo rate (28%).\nIn the subset of patients followed-up for 42 days, recrudescence between day 28 and 42 was rare: only 8/317 (2.5%) of individuals classified as adequate clinical and parasitological response (ACPR) or sensitive (S) at day 28 went on to fail by day 42 (1 in the CQ arm, 2 in the CQ+PQ arm, 3 in the CQ+AS arm, 1 in the SP arm and 1 in the SP+PQ arm).\nFailure rates at day 42 were therefore similar to day 28.\nThe failures between 28-day and 42-day rates were not corrected by PCR and hence could have resulted from new infections.\nMolecular outcomes.\nMolecular analysis of drug resistance-associated alleles was conducted on samples taken on day 0 (Table 6).\nThe major marker of chloroquine resistance, pfcrt-76T, was fixed at 100%.\nMutations in pfmdr1 are known to modulate resistance to quinoline and other drugs.\nThe prevalence of chloroquine-resistance associated allele pfmdr1-86Y was 13.6% in this sample.\nMutations associated with pyrimethamine resistance (on the pfdhfr gene) were seen frequently; 108 N was found in all but one of 74 typable samples (98.6%) and 59R occurred in 66/76 (86.8%) samples.\nHowever, 51I was only detected in 5/75 (6.7%) samples.\nTherefore pfdhfr \u2018double mutants\u2019 were common (at codons 108+59) but \u2018triple mutants\u2019 were rare.\nInterestingly, we observed one sample with the pfdhfr mutations 16V+108T.\nThese are reportedly selected by the antifolate drug cycloguanil, rather than pyrimethamine.\nNo mutations on the dhps gene were seen.\nThese PCR results broadly match what would be expected from the clinical outcomes.\nThe pfdhr/pfdhps mutation profile suggests that the parasites would be killed by a full therapeutic dose of SP, but that some tolerance to lower doses existed.\nEffect on Gametocytes\nThe addition of artesunate or primaquine to chloroquine or SP reduced gametocyte carriage on day 7 (Table 3).\nCombining chloroquine or SP with artesunate succeeded in eliminating gametocytaemia from more individuals by day 7 compared to monotherapy or co-treatment with a single-dose of primaquine.\nFigure 3 shows the proportion of individuals with patent gametocytaemia over 28 days of follow-up.\nIn the chloroquine and SP arms, gametocyte carriage persisted at day 28 in >30% of individuals in the CQ arm and >70% of individuals in the SP arm.\nThe addition of primaquine reduced gametocyte carriage; the effect was more pronounced with chloroquine than with SP.\nThe proportion of gametocytaemic individuals was lowest in the artesunate arms and the difference was evident within two days of the start of treatment.\nPeak gametocyte densities occurred 7 days after the start of treatment (Figure 4).\nGametocyte density was higher after SP than after chloroquine treatment.\nBoth primaquine and artesunate reduced density to low levels by day 7.\nMost patients did not have gametocytaemia on day 0 (291/355 (82.0%) had no gametocytes on day 0).\nHowever, the presence or absence of sexual stage parasites on day 0 was an important explanatory variable in the secondary analysis; patients who presented with gametocytes on day 0 were more likely to be gametocytaemic on day 7 than individuals who were not gametocytaemic on day 0 regardless of treatment group (Mantel-Haenszel OR: 3.5 [95%CI: 2.0\u20136.4], p<0.001).\nWe therefore conducted further analysis comparing those who presented with gametocytes on day 0 and those who did not.\nThe artesunate treatments were more effective at preventing the emergence of gametocytaemia between day 0 and 7 than in removing established gametocytaemia already present on day 0 (Table 7).\nFor example, among the CQ+AS group who did have patent gametocytaemia on day 0, 8/9 were still gametocytaemic on day 7 after treatment.\nThis was not significantly different from the proportion gametocytaemic on day 7 (11/12) after treatment with chloroquine monotherapy (Fisher's exact test p\u200a=\u200a1.0).\nBy contrast, in the chloroquine arms of the patients who did not have patent gametocytaemia on day 0, only 11% (7/63) were gametocytaemic on day 7 when treated with CQ+AS as compared to 47/56 (84%) among those treated with chloroquine monotherapy (\u03c72 p<0.001).\nSimilar trends were evident in the SP arms treated with artesunate; prevalence of gametocytaemia on day 7 was significantly lower in the SP+AS group that was not gametocytaemic on day 0 than in the group that was (Fisher's exact test p<0.001).\nThis indicates that artesunate is less active against older gametocytes than against those newly emerged or immature forms that are not yet emerged.\nPrimaquine showed no such trend: the proportion of patients gametocytaemic on day 7 was not significantly different for patients who were or were not gametocytaemic on day 0.\nPrimaquine therefore appears to be active against gametocytes of all ages.\nPrimaquine was more effective at controlling gametocytaemia when combined with CQ than with SP, regardless of whether individuals were gametocytaemic at the start of treatment (MH-OR\u200a=\u200a0.22, P\u200a=\u200a0.001) (Table 7).\nBy contrast, artesunate seemed as effective when combined with chloroquine as it was when combined with SP (MH-OR\u200a=\u200a1.04, P\u200a=\u200a0.83).\nThough, as discussed, artesunate was more effective at preventing development of patent gametocytaemia than in clearing existing gametocytaemia.\nArtesunate was better than primaquine at preventing patent gametocytaemia among those that were not gametocytaemic at day 0 enrolment (CQ: OR\u200a=\u200a0.30, P<0.001; SP: OR\u200a=\u200a0.18, P<0.001).\nUsing the presence of gametocytes on day 7 as the secondary endpoint, adjusted for the presence of gametocytes at day 0, both primaquine and artesunate with chloroquine or SP were more effective than their respective monotherapies at reducing the presence of gametocytes (Table 8).\nCompared to monotherapy with chloroquine or SP, co-treatment with artesunate was negatively associated with having gametocytes on day 28 whether the accompanying drug was chloroquine or SP (Table 8).\nThe presence of gametocytes on day 0 was not associated with the presence of gametocytes on day 28 (MH-OR, adjusted for treatment arm: 1.0, p\u200a=\u200a1.0) suggesting that older gametocytes had cleared by day 28, regardless of treatment.\nDiscussion\nThe present study was designed to inform decision-making at several levels: to guide UNHCR on appropriate treatment in the Afghan refugee communities in Pakistan, to build evidence for national treatment policy in the Pakistan Directorate of Malaria Control and in the Afghanistan Ministry of Public Health, to guide the WHO Eastern Mediterranean Region Office (EMRO) on regional treatment recommendations, and to build evidence for international epidemic response guidelines.\nThe findings of this study, when presented at an inter-country meeting of national malaria control managers coordinated by WHO/EMRO in 2004, contributed to a shift in policy from chloroquine monotherapy to ACT in South and West Asia, influencing the defining of SP+AS as the first line treatment of choice for the WHO Eastern Mediterranean Region.\nIndia is the latest neighbouring country to adopt SP+AS as first line treatment with the support of WHO.\nThe SP+AS combination is also effective against vivax malaria, although not as an approved treatment.\nSP monotherapy was effective in the study population, with treatment failure rates remaining in the 5\u201310% range, similar to that seen in other studies in the region.\nMolecular studies in the region show that genetic markers for SP resistance are still at low levels in the parasite population.\nHowever, the South East Asian experience of rapid rise in SP resistance in settings of similar endemicity points to SP resistance spreading rapidly if ACT policy is not implemented rigorously in both public and private sectors across the region.\nThe complete failure of chloroquine (which at the time of the study remained the officially sanctioned first line treatment) convinced policy makers of the need to redefine treatment practices in South Asia.\nThe new data demonstrated higher levels of resistance than in previous studies dating back to the 1980s and 1990s and emphasised the need for change to national treatment guidelines.\nThe addition of artesunate to chloroquine did reduce treatment failure rates, but if this regimen was used as policy it would, in effect, constitute artesunate monotherapy in the majority of infections and accelerate the development of artesunate tolerance, as reported in South East Asia.\nA constraint on the study design was that not all 6 arms could be included at each of the three study sites.\nSite is therefore a potential confounding covariate that cannot be fully controlled for in the data analysis.\nDifferences in transmission could, for example, affect the risk of re-infection between sites but this was corrected for by the PCR analysis.\nAt each site in any one year each triplet of treatment arms (all the SP arms or all the chloroquine arms) were examined at the same time, and it is comparisons within these triplets which are of greatest interest.\nThe study did not complete its target sample size for the SP arms because of low recruitment rates.\nPCR corrected outcomes were not attainable for all failures, and higher numbers of patients were reinfected at one site where only the chloroquine arms were being tested.\nLow rates of transmission at the other sites make it unlikely that failing cases were due to reinfection.\nThe secondary analysis with PCR-corrected outcomes did not change the overall conclusions on the suitability of the combinations for revised treatment policy.\nTreatment with either the clinically effective SP monotherapy or the failing chloroquine monotherapy resulted in persistence of gametocytaemia into the second and third weeks in the majority of individuals.\nThis effect was more evident for SP than for chloroquine, SP being well known for high rates of gametocyte carriage post treatment.\nThis contrasts with situations of low-frequency chloroquine resistance where the proportion of gametocytaemic patients post treatment also tends to be lower.\nThe effect of artemisinin derivatives on reducing gametocyte carriage is already documented.\nArtesunate also proved highly effective in reducing the prevalence of gametocytaemia in the present study.\nRegimens combining artesunate with chloroquine or SP saw marked reductions in gametocyte carriage and few cases had persisting gametocytaemia.\nAs demonstrated with SP+AS in The Gambia the SP+AS and CQ+AS combinations appear to have limited activity against mature circulating gametocytes.\nCo-treatment with a single dose of primaquine was more effective against older gametocyte infections.\nBoth primaquine and, more markedly, artesunate reduced the odds of persisting gametocytes from that seen post treatment with either chloroquine or SP monotherapy.\nPrimaquine is more rapidly excreted than artesunate and this may account for the fact that the single dose of primaquine used in this study had a lower impact on gametocyte carriage than the artesunate regimens.\nThe usefulness of primaquine as a gametocytocidal treatment may be improved by administering it after the ACT course.\nIn our study, a key factor in the clearance of gametocytes was the presence or absence of gametocytes on day 0.\nThose who presented with gametocytes were more likely to have gametocytes on day 7, an effect which was independent of the treatment given.\nAlthough the majority of patients (80%) presented without gametocytaemia, artesunate appeared to be less active against these older parasites whereas primaquine appeared to be effective against all ages.\nThe proportion presenting with gametocytes may vary according to background transmission levels.\nIf gametocytes persist after treatment with ACTs, the effect on transmission in areas where a high proportion of cases present with gametocytes may prove suboptimal and justify the simultaneous use of primaquine.\nPrimaquine is more rapidly excreted than artesunate and this may account for the fact that the single dose of primaquine used in this study had a lower impact on gametocyte carriage than the artesunate regimens.\nThe usefulness of primaquine as a gametocytocidal treatment may be improved by administering it after the ACT.\nFor treatment policy to have a major impact on transmission several criteria need to be met: i) transmission is low to moderate; ii) the majority of people use public health rather than private facilities, or effective private sector interventions ensure adherence to policy; iii) public health facilities correctly prescribe the approved regimen, and iv) patients take the full course.\nThese conditions are largely met in the refugee villages of Pakistan.\nThe region is characterized by low endemicity and low immunity, well supported health care facilities are available in the Afghan refugee villages, and diagnosis of malaria to species level is maintained to a high standard.\nThe health facilities are well utilized by the Afghan refugees and Pakistanis living nearby.\nThe effect of this policy is evident in the refugee populations: since 2005 falciparum malaria has virtually disappeared in all but a handful of the refugee villages [International Rescue Committee, Final Report, unpublished].\nThis reduction is attributed to prompt and accurate diagnosis and wider access to ACTs as prevention interventions were minimal.\nThe operational data from this region and elsewhere suggest that drugs showing faster gametocyte clearance in clinical trials do help to reduce transmission.\nHowever, it is also important to note the results of recent research which suggest that drawing conclusions about the potential impact on transmission is not straightforward, for the following reasons:\nThe effect of drugs on gametocyte carriage cannot be fully determined if only standard light microscopy (LM) is used.\nResearch in the last decade shows that LM gives starkly different indications of gametocyte prevalence and densities compared to other techniques and that sub-microscopic levels of gametocytes likely play an important role in transmission, perhaps particularly in low transmission settings.\nGametocytes that are readily identified in blood following treatment may or may not be viable.\nGametocytes may not be infective to mosquitoes either because they are recently emerged (naturally, gametocytes are not infectious to mosquitoes in the first 1\u20132 d of emergence) or because the effect of the drug regimens have rendered them \u2018sterile\u2019.\nIt is unclear how the density of gametocytes in the blood links to potential for transmission success to blood feeding mosquitoes.\nBousema & Drakeley summarise the available data from membrane feeding studies showing that whilst there is an overall correlation, the variations within it are confusing; high density gametocytaemia can lead to minimal mosquito infection and very low densities lead to reasonable rates of infection.\nEven if solid data show that a particular drug does reduce the number of gametocytes which successfully infect, and lead to infectious, mosquitoes; the effect of this on malaria transmission cannot be stated with certainty.\nThis effect would depend on what proportion of the gametocyte reservoir these patients would otherwise make up.\nThe comparative performance of artesunate and primaquine on gametocyte carriage in this study did not influence policy.\nThe decision to use SP+AS was based on treatment efficacy and the need to protect against SP resistance through the use of an ACT combination.\nThe fact that SP+AS led to reduced gametocyte carriage compared to the previous policy of chloroquine montherapy was reassuring, and raised the possibility of the regimen reducing transmission too.\nSP+AS is now the first line treatment policy in Pakistan, Iran, Afghanistan and India.\nThe continuing collection of molecular data on SP resistance in this region will be vital, given that the treatment efficacy of artesunate could mask the emergence of SP resistance by acting alone to provide a degree of clinical cure in the presence of a failing companion drug.\nThis risk could be reduced by introduction of a co-blister of SP plus artesunate.\nTrial profile.\nKaplan Meier survival analysis showing cumulative probability of failure in CQ treatment groups (A) and SP treatment groups (B).\nPercentage of patients (+CI) carrying gametocytes on specified days after treatment in CQ (A) and SP (B) treatment arms.\nGeometric mean gametocyte density (+CI) after treatment in chloroquine (A) and SP treatment arms (B).Note the different scales on the Y-axis. Where the CI bars are absent, only 1 patient had gametocytes.\n\nSummary of treatment arms tested at each of the study sites over the three transmission seasons.\n | Study site\nTransmission season (July\u2013January) | Site 1 (Adizai) | Site 2 (Yakka Ghund) | Site 3 (Kotki)\nSeason 1: 2000\u20132001 | CQ, CQ+PQ, CQ+AS,SP, SP+PQ, SP+AS |  | \nSeason 2: 2001\u20132002 | SP, SP+PQ, SP+AS | CQ, CQ+PQ, CQ+AS | \nSeason 3: 2002\u20132003 | CQ, CQ+PQ, CQ+AS | SP, SP+PQ, SP+AS | CQ, CQ+PQ, CQ+AS\n\n\nEnrolment characteristics of the treatment groups.\nVariable | CQ | CQ+PQ | CQ+AS | SP | SP+PQ | SP+AS\nNumber enrolled | 76 | 76 | 74 | 45 | 40 | 44\nNumber evaluable at day 28 (%) | 68 (89) | 67 (88) | 67 (91) | 41 (91) | 33 (82) | 41 (93)\nAge [median (IQR) years] | 12 (8\u201318) | 12 (8\u201320) | 12 (8\u201320) | 17 (9\u201327) | 14 (7\u201325) | 18.5 (9.5\u201330)\nPercentage female | 42 | 37 | 50 | 33 | 43 | 45\nWeight [median (IQR) kg] | 29 (20\u201347) | 30 (20\u201353) | 33 (20\u201345) | 48 (25\u201358) | 42 (22\u201355) | 41 (21\u201357)\nTemperature [mean (SD) \u00b0C] | 37.3 (1.0) | 37.5 (1.2) | 37.4 (1.2) | 37.5 (1.0) | 37.5 (1.2) | 37.5 (1.5)\nTemperature \u226537.5\u00b0C on presentation [n (%)] | 33 (43) | 34 (45) | 34 (46) | 21 (47) | 17 (43) | 23 (52)\nPCV [mean (SD) % haematocrit]1 | 42.9 (9.7) | 40.8 (3.9) | 41.5 (4.3) | 44.2 (7.7) | 45.5 (6.9) | 44.3 (5.1)\nPCV<30% [n (%)]1 | 0 | 0 | 0 | 1 (2.2) | 0 | 0\nAsexual parasite density [geometric mean (95% CI) per \u00b5l] | 5161 (3536\u20137535) | 5263 (3647\u20137595) | 7366 (4972\u201310,915) | 7600 (5185\u201311,140) | 8091 (4236\u201315,454) | 12,134 (7757\u201318,982)\nGametocyte positive [n (%)] | 12 (15.8) | 15 (19.7) | 9 (12.2) | 14 (31.1) | 7 (17.5) | 7 (15.9)\nGametocyte density [geometric mean (95% CI) per \u00b5l] | 1.3 (0.4\u20132.6) | 1.7 (0.7\u20133.4) | 0.6 (0.2\u20131.2) | 4.1 (1.4\u201310.1) | 1.6 (0.3\u20134.1) | 1.2 (0.2\u20132.9)\n\nNotes: (1) PCV was not recorded for all patients; a microcentrifuge was only available at one of the 3 clinics. For PCV percentages in the 6 treatment groups: CQ n\u200a=\u200a10; CQ+PQ n\u200a=\u200a19; CQ+AS n\u200a=\u200a13; SP n\u200a=\u200a19; SP+PQ n\u200a=\u200a15, SP+AS n\u200a=\u200a15.\n\nClinical and parasitological failure rates at 28 days follow-up and for gametocytaemia at day 7.\n | Day 28 clinical or parasitological failure | Day 7 gametocytaemia\n | No./Total (%) | Odds Ratio (95% CI) | P | No./Total (%) | Odds Ratio (95% CI) | P\nCQ | 55/68 (81) | Reference | - | 52/67 (78) | Reference | -\nCQ+PQ | 49/67 (73) | 0.64 (0.29\u20131.4) | 0.3 | 27/70 (39) | 0.18 (0.09\u20130.4) | <0.001\nCQ+AS | 19/67 (28) | 0.09 (0.04\u20130.2) | <0.001 | 12/72 (17) | 0.06 (0.02\u20130.1) | <0.001\nSP | 4/41 (10) | Reference | - | 37/43 (86.1) | Reference | -\nSP+PQ | 5/33 (16) | 1.6 (0.41\u20136.7) | 0.5 | 24/36 (66.7) | 0.32 (0.10\u20131.0) | 0.046\nSP+AS | 1/41 (2) | 0.23 (0.02\u20132.2) | 0.2 | 9/44 (20.5) | 0.04 (0.01\u20130.1) | <0.001\n\n\nClinical and parasitological outcomes after 28 days follow-up: n (%).\n | CQ | CQ+PQ | CQ+AS | SP | SP+PQ | SP+AS\nClinical outcomes | N\u200a=\u200a56 | N\u200a=\u200a58 | N\u200a=\u200a67 | N\u200a=\u200a41 | N\u200a=\u200a30 | N\u200a=\u200a40\nAdequate clinical response | 13 (23) | 18 (31) | 48 (72) | 37 (90) | 28 (93) | 40 (100)\nEarly treatment failure | 9 (16) | 7 (12) | 0 | 1 (2) | 1 (3) | 0\nLate clinical failure | 2 (4) | 4 (7) | 1 (2) | 0 | 0 | 0\nLate Parasitological Failure | 32 (57) | 35 (50) | 18 (26) | 3 (7) | 1 (3) | 0\nParasitological outcomes | N\u200a=\u200a63 | N\u200a=\u200a65 | N\u200a=\u200a67 | N\u200a=\u200a40 | N\u200a=\u200a32 | N\u200a=\u200a41\nS | 13 (21) | 18 (28) | 48 (72) | 37 (93) | 28 (88) | 40 (98)\nRI | 35 (56) | 33 (51) | 19 (28) | 3 (7) | 1 (3) | 0\nRII | 13 (21) | 11 (16.9) | 0 | 0 | 3 (9) | 1 (2)\nRIII | 2 (3) | 3 (5) | 0 | 0 | 0 | 0\n\n\nOutcomes of PCR analysis, and projected adjusted failure rates by treatment group excluding indeterminate results.\n | PCR Results [n (%)] | Failure rates [n/N (%)]\n | Reinfection | Recrudescent | Negative | Total | in vivo1 | PCR adjusted 2\nCQ | 6 (21) | 20 (69) | 3 (10) | 29 | 55/68 (81) | 42/68 (62)\nCQ+PQ | 4 (21) | 11 (58) | 4 (21) | 19 | 49/67 (73) | 36/67 (54)\nCQ+AS | 7 (64) | 3 (27) | 1 (9) | 11 | 19/67 (28) | 6/67 (9)\nSP | 0 | 2 (100) | 0 | 2 | 4/41 (10) | 4/41 (10)\nSP+PQ | 0 | 2 (100) | 0 | 2 | 5/33 (16) | 5/33 (16)\nSP+AS | - | - | - | 0 | 1/41 (2) | 1/41 (2)\n\nNotes: (1) Taken from Table 3 for 28 day failures; (2) Adjusted by the ratio of PCR recrudescent to PCR re-infected cases to estimate number of true recrudescent cases among the observed in vivo failures.\n\nDrug resistance alleles in a sub-set collected at enrolment.\nLocus | Allele | Number | (%)\nPfcrt | 76 K | 0 | \n | 76 T | 63 | (100)\nPfmdr1 | 86 N | 76 | (86)\n | 86 Y | 12 | (14)\n | 184 Y | 22 | (27)\n | 184 F | 60 | (73)\nDHFR | 16 A | 76 | (99)\n | 16 V | 1 | (1)\n | 50/51 CN1 | 70 | (93)\n | 50/51 C1 | 5 | (7)\n | 59 C | 10 | (13)\n | 59 R | 66 | (67)\n | 108 N | 73 | (99)\n | 108 T | 1 | (1)\nDHPS | 436/437 SA | 43 | (100)\n | 581 A | 41 | (100)\n | 613 A | 43 | (100)\n\n\nNumbers and percentages of individuals with gametocytes on day 7 post treatment.\n | No (%) parasitaemic on day 7\n | Amongst those who had gametocytes on day 0 | Amongst those who did not have gametocytes on day 0\nCQ | 11/12 (91.7%) a, 1 | 47/56 (83.9%) a, 1\nCQ+PQ | 7/14 (50%) b, 1 | 21/56 (37.5%) b, 1\nCQ+AS | 8/9(88.9%) a, 1 | 7/63(11.1%) c, 2\nSP | 14/14 (100%) a, 1 | 25/29 (86.2%) a, 1\nSP+PQ | 5/7 (71.4%) b, 1 | 23/30 (76.7%) a, 1\nSP+AS | 5/7 (71.4%) b, 1 | 5/37 (13.5) b, 2\n\nChloroquine or SP treatments in the same column that share the same superscript letter are not significantly different. Treatments in the same row that share the same numeric superscript are not significantly different.\n\nEffect of treatment on gametocyte carriage at day 7 and day 28, for those patients who were gametocytaemic on enrolment.\n | AOR for presence of gametocytes on day 7, (95% CI), p-value 1 | AOR for presence of gametocytes on day 28, (95% CI), p-value 1\nCQ | Reference | Reference\nCQ+PQ | 0.15 (0.07\u20130.34), p<0.001 | 0.42 (0.08\u20132.2), p\u200a=\u200a0.3\nCQ+AS | 0.05 (0.02\u20130.12), p<0.001 | 0.04 (0.004\u20130.39), p\u200a=\u200a0.006\nSP | Reference | Reference\nSP+PQ | 0.37 (0.12\u20131.16), p\u200a=\u200a0.09 | 0.14 (0.05\u20130.4), p<0.001\nSP+AS | 0.04 (0.01\u20130.13), p<0.001 | 0.02 (0.005\u20130.1), p<0.001\n\nNotes: (1) AOR: Odds ratios using logistic regression analysis adjusted for presence of gametocytes on day 0 (CQ and SP arms analysed separately).",
        "label": "low",
        "id": "task4_RLD_test_917"
    },
    {
        "paper_doi": "10.1186/1475-2875-12-363",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: cRCT\n\nUnit of allocation: village (paired on the basis of geographical location)\n\nNumber of units: 10:10\n\nLength of follow-up: 1 year\n\nOutcome assessment: passive case detection at local health centre. Children were followed up in 2 postintervention cross-sectional surveys: after 5 months and at the end of the study (10 months).\n\nAdjustment: cluster adjustment using intraclass correlation coefficient of 0.048, between-cluster variation 0.006, and within-cluster variation 0.12\n\n\nParticipants: Number of participants: 8395\n\nInclusion criteria: children under 10 years of age\n\n\nInterventions: Intervention: bed net (n = 4066)\n\nInsecticide and dosage: deltamethrin (25 mg/m2)Retreatment: not stated\n\nUsage: not stated\n\nControl: no net (n = 4109)\n\n\nOutcomes: Outcomes measured:Malaria prevalenceMalaria incidenceNumber of deathsNumber of cases of severe malaria\n\n\nNotes: Study location: Rakhine State, Western Myanmar, in 2 areas: Dabhine and MyothugyiEIR: not statedMalaria transmission: predominantly low with pockets of intense transmissionMain vectors: not stated% P vivax cases: 52%, and 2% P falciparum/P vivax mixe\n\n",
        "objective": "The primary objective of this review was to assess the impact of ITNs on mortality and malaria morbidity, incorporating any evidence published since the previous update into new and existing analyses, and assessing the certainty of the resulting evidence using GRADE.",
        "full_paper": "Background\nInsecticide-treated bed nets (ITN) reduce malaria morbidity and mortality consistently in Africa, but their benefits have been less consistent in Asia.\nThis study\u2019s objective was to evaluate the malaria protective efficacy of village-wide usage of ITN in Western Myanmar and estimate the cost-effectiveness of ITN compared with extending early diagnosis and treatment services.\nMethods\nA cluster-randomized controlled trial was conducted in Rakhine State to assess the efficacy of ITNs in preventing malaria and anaemia in children and their secondary effects on nutrition and development.\nThe data were aggregated for each village to obtain cluster-level infection rates.\nIn total 8,175 children under 10\u00a0years of age were followed up for 10\u00a0months, which included the main malaria transmission period.\nThe incidence and prevalence of Plasmodium falciparum and Plasmodium vivax infections, and the biting behaviour of Anopheles mosquitoes in the area were studied concurrently.\nThe trial data along with costs for current recommended treatment practices were modelled to estimate the cost-effectiveness of ITNs compared with, or in addition to extending the coverage of early diagnosis and treatment services.\nResults\nIn aggregate, malaria infections, spleen rates, haemoglobin concentrations, and weight for height, did not differ significantly during the study period between villages with and without ITNs, with a weighted mean difference of \u22122.6 P. falciparum episodes per 1,000\u00a0weeks at risk (95% Confidence Interval \u22127 to 1.8).\nIn areas with a higher incidence of malaria there was some evidence ITN protective efficacy.\nThe economic analysis indicated that, despite the uncertainty and variability in their protective efficacy in the different study sites, ITN could still be cost-effective, but not if they displaced funding for early diagnosis and effective treatment which is substantially more cost-effective.\nConclusion\nIn Western Myanmar deployment of ITNs did not provide consistent protection against malaria in children living in malaria endemic villages.\nEarly diagnosis and effective treatment is a more cost effective malaria control strategy than deployment of ITNs in this area where the main vector bites early in the evening, often before people are protected by an ITN.\nBackground\nMalaria is a major cause of morbidity and mortality in Myanmar.\nMalaria control activities in this country have been concentrated on early, mostly clinical, diagnosis and treatment.\nLimited availability of curative services in remote areas, the difficulties in accessing malarious areas, and the relatively high costs of effective treatment of multi-drug resistant malaria, have compromized malaria control efforts.\nEffective malaria control activities are needed, and there is a recent substantial increase in donor support for these activities.\nLevels of chloroquine and sulphadoxine-pyrimethamine resistance in Plasmodium falciparum have been high in this region for decades.\nSubsidies for highly effective artemisinin combination treatment (ACT) have considerably increased their availability in recent years.\nUnfortunately artemisinin monotherapies are still widely available providing continued selective pressure on emerging artemisinin resistance in the east of the country.\nRegular and systematic indoor spraying of residual insecticides was stopped in 1993 and is now only used for special situations (outbreaks, new settlements).\nLargely ignored by the outside world until 2007, since then there has been an over fifty-fold increase in external donor funding for malaria control in Myanmar, the majority of which has been spent on insecticide-treated mosquito nets (ITN).\nITNs are an appealing approach to the control of malaria.\nThey repel and kill malaria vectors, and they prevent sporozoite-bearing anopheline mosquitoes from biting the occupants lying under or near the bed net.\nITNs have been shown to exert a mass effect reducing vector populations in villages with extensive use and high transmission.\nITNs are usually greatly appreciated by their occupants, because they prevent all kinds of insect bites, and they are very safe.\nOverall use of ITNs prevented approximately one in five childhood deaths in sub-Saharan Africa.\nThese substantial benefits have rightly led to the inclusion of ITNs as a central component of many malaria control initiatives throughout the world.\nIn endemic areas of South and South-East Asia, where malaria transmission is lower and unstable, mobile young adult males are particularly affected, and vector behaviour is different, the results of ITN efficacy studies are mixed, and much less clear than in Africa.\nThis is not because of insecticide resistance, although some vector species do exhibit low-grade pyrethroid resistance, but because of the behaviour of the vectors which often bite early in the evening or morning outdoors or away from dwellings, and the corresponding behaviour of humans.\nITN always provide some protection against malaria \u2013 the critical question is how much?\nAll individuals in malaria endemic areas should ideally have ITN, even if they provide only modest benefit, but whilst there are financial constraints, choices need to be made.\nIt is necessary therefore to assess the efficacy of ITN in the South and South-East Asian region at a local level to assist decisions on their wider-scale deployment in relation to alternative proven highly effective malaria control measures (i.e. early diagnosis and effective treatment: EDAET).\nThe effectiveness of ITN in reducing malaria morbidity and mortality depends on several factors, including the level of malaria endemicity and the behaviour and immunity of the population, the climate, the acceptance and usage of the ITN by the population and, crucially, the biting behaviour of the main anopheline vectors.\nA cluster-randomized controlled trial was carried out in Western Myanmar, to evaluate the protective efficacy of village-wide usage of insecticide-impregnated ITNs on malaria incidence and prevalence, anaemia, and the development of children.\nThe anopheline vector abundance and biting behaviour were also studied, and data were collected on population sleeping habits, to assess the potential of ITNs to reduce man-mosquito contact in this region; these findings are described in the accompanying paper.\nWhile effective diagnosis and treatment facilities were an ethical and research necessity in the study setting, these are often not present in routine settings.\nBudgetary limitations may result in ITN programmes competing with EDAET services for limited resources.\nA cost-effectiveness analysis was therefore carried out to assess the costs and benefits of ITN when compared with those of EDAET when neither has been implemented, and to assess the incremental cost-effectiveness of introducing one of the interventions when the other is already in place.\nMethods\nStudy area and population\nThe study was conducted in Rakhine State, Western Myanmar.\nThe monsoon season is from May to October and yearly rainfall is very high (+/\u2212 5000\u00a0mm/year).\nMalaria transmission occurs throughout the year and peaks during the post monsoon (November-January), and sometimes in the early monsoon (June-July) periods (Figure\u00a01).\nThe transmission intensity varies from predominantly low to pockets of intense transmission, depending on the location, and there are considerable variations over short geographic distances and also between years and between seasons.\nSymptomatic malaria occurs at all ages but is seen predominantly in children.\nAccess to effective treatment has historically been very poor.\nOutbreaks of falciparum malaria can spread over several townships, or may be limited to a single village at other times.\nIn 1994, M\u00e9d\u00e9cins Sans Fronti\u00e8res - Holland (MSF-H) began a malaria control programme in Rakhine State in cooperation with the Myanmar VBDC (Vector Borne Disease Control) department.\nThe programme focussed on early diagnosis and treatment, through support of 30 fixed field clinics with laboratories and riverboat \u2018mobile malaria clinics\u2019.\nPlasmodium falciparum was responsible for approximately 80% of the malaria infections in patients who presented to these clinics.\nDrug resistance in P. falciparum in the area was studied in 1995; high levels of resistance were found with treatment failure rates of 82% to chloroquine and 67% to sulphadoxine-pyrimethamine, whereas mefloquine was very effective (93% cure rate).\nSince 1996 all patients with falciparum malaria have been treated with a combination of mefloquine and artesunate.\nBetween 1997 and 2007 over one million patients (approximately 43% of whom were children under 10\u00a0years of age) received treatment for slide confirmed malaria, supported by this programme.\nIt was unclear whether ITN should be deployed as a priority so the effectiveness of ITN was studied between May 1998 and February 1999, covering the peak malaria transmission season, in villages around two Rural Health Clinics (RHC), in Dabhine and Myothugyi, two village-tracts located in the townships Sittwe and Maungdaw, which are approximately 90\u00a0km apart.\nThe study villages in the Dabhine area are all within 5 miles from the Bay of Bengal.\nThis is a coastal plain area without hills or forest, where rice and other crops are cultivated.\nCommon breeding sites for Anophelines are paddy fields, brackish water in streams and tide pools and saline ponds used for prawn culture which are close to villages, swamps, and small ponds used for growing watercress.\nThe area was classified generally as meso- to hyperendemic for malaria.\nA malaria blood-slide survey, performed in September 1995, found a prevalence of 34% for P. falciparum and 11% for Plasmodium vivax, and a spleen prevalence of 36% among primary school children.\nMyothugyi is a densely populated area situated 5\u20137 miles from the Bay of Bengal.\nThe area is characterized by rice-fields and partly forested hills.\nPotential breeding sites for Anophelines are freshwater creeks, ponds and stagnant water in rice fields, as well as brackish water in tide pools, prawn-breeding ponds and small pools and streams at foothills.\nIt has been classified generally as a low meso-endemic malaria area.\nA malaria blood-slide survey in January 1996, 10 kilometres south of Myothugyi, found a prevalence of 13% for P. falciparum and 3% for P. vivax and none of the primary schoolchildren examined had palpable splenomegaly.\nThese surveys, and the substantial burden of malaria in children, prompted discussions with the community and authorities as to what control measures could be applied.\nIn view of the uncertainties over ITN efficacy and their limited availability at the time, and after consultation with the authorities and village leaders, it was decided jointly that an evaluation was needed to set priorities before consideration of wide-scale deployment.\nIt was agreed that if there was any evidence of benefit, then after the evaluation, ITNs would be given to all villages included in the evaluation.\nStudy design, randomization and sample size calculation\nIdeally, a study population is randomized on an individual or household basis.\nHowever, the use of insecticide-impregnated bed nets (ITN) has an impact on the vector populations in the nearby environment and, therefore, on malaria transmission and risk, not only in the household itself, but also among people living nearby, including those who do not have ITN (the \u201cmass\u201d effect).\nTherefore, the unit chosen for randomization in this study was the village, and this was agreed with the village leaders in consultation with the community.\nThe primary outcome measure was the incidence and prevalence of malaria in children.\nIn December 1997 and January 1998 a population survey was done in the study areas.\nOutreach Workers (ORW) were trained to register all families, number houses, and to collect demographic data on family size and number of children under 10\u00a0years of age.\nDuring this period, a sample of the population of children under 10 was checked for parasitaemia and splenomegaly.\nThe results of the pre-intervention survey were used to determine the number of clusters and the sample sizes needed for the larger study.\nThe pre-intervention malaria survey was performed on 1,088 children (Table\u00a01).\nThe intra-class correlation coefficient (ICC) calculated on the basis of the pre-intervention data was 0.048, with a between-cluster variation of 0.006 and within-cluster variation of 0.12.\nThe prevalence of P. falciparum parasitaemia by microscopy averaged 17% (range 0 \u2013 53%).\nMicroscopy was quality assured throughout the study with independent cross-checking.\nThe number of village-clusters and their size needed for a valid comparative study was calculated using the methods described by Thompson.\nIn order to detect a 50% reduction in the incidence of falciparum malaria, with 80% power at a 5% significance level, a total of 10 village pairs were required with an average of 420 children per village, amounting to a total sample size of 8,400 children.\nInitially, 22 villages were informed about the study-procedures and were invited to participate, but after further discussions two villages declined to join.\nFinally 20 villages (clusters) were paired, and for each pair one was selected randomly (using a computer generated random number) to receive impregnated bed nets (ITNs), while the other acted as the control village.\nMatching was done according to geographical location so that the intervention and control villages were nearby (villages were 1\u20132 miles apart).\nThis minimized differences in the environment in and around each village pair (i.e. proximity to foothills, forest fringe, prawn-ponds, rice-fields, waterways), which could have had a strong influence on vector populations and consequently on the prevalence and incidence of malaria.\nThe travel-time to the health clinic and its influence on the number of patient visits, and therefore on malaria incidence, were comparable for intervention and control villages, as the distances were matched equally.\nITN distribution\nBetween the 26th and 29th of April 1998, all households in the intervention villages received a number of ITNs, according to the number of family members and were instructed explicitly about the correct use of the nets (Figure\u00a02).\nMore than 5000 ITN were distributed.\nThe ITN (green colour, polyester, size 130 \u00d7 180 \u00d7 150 cm (11.6\u00a0m2) or 190 \u00d7 180 \u00d7 150 cm (14.5\u00a0m2), Siam-Dutch Co, Thailand) were already impregnated with deltamethrin (25\u00a0mg/m2) by the manufacturer.\nAll households of the control villages received ITNs after the study was completed in May 1999.\nIncidence surveillance\nMalaria morbidity surveillance consisted of passive case detection.\nEach family from both study groups received a specific registration card at the beginning of the study and anyone with complaints of fever was urged to visit the RHC and bring the registration card.\nIn both study areas a RHC operated by staff from the Department of Health, was supported by doctors, microscopists, and out-reach workers from MSF-H. Microscopy was quality checked on a regular basis.\nThe RHCs were open seven days a week.\nOut-reach workers regularly visited the villages and informed the population about using the clinics in case of illness.\nIn the RHC blood-smears of patients with fever or a history of fever were examined for malaria parasites.\nData-files were kept for all children under 10\u00a0years coming from the villages under study.\nPatients with falciparum malaria were treated with a single dose of mefloquine 15\u00a0mg base/kg and artesunate 4\u00a0mg/kg (then the current treatment).\nVivax malaria was treated with chloroquine (10\u00a0mg base/kg on day 0 and 1, 5\u00a0mg/kg on day 2), followed by primaquine (0.25\u00a0mg base/kg/day for 14\u00a0days).\nPatients who were not part of the study were also provided with medical services.\nNo other significant health services were available in the areas under study.\nCross-sectional surveys\nAfter distribution of ITN, children were followed up for weight, height and haemoglobin at three time-points: at the beginning of the study (May 11 - June 10, 1998), after 5\u00a0months (14 Sept- 9 Oct. 1998) and at the end of the study (25 Jan - 22 Feb 1999) (Figure\u00a02).\nDuring the last survey a blood-smear for the detection of malarial parasitaemia was taken as well.\nDuring the surveys people were asked about the usage of ITN and whether they had washed their ITN.\nEthics statement\nThe study was discussed in detail with community representatives of each village and with officials from the Ministry of Health.\nThere was general agreement on the need for an evaluation but no local ethical review committee or relevant formal organizational ethics review process was available before the study started in 1998.\nThe reasons for the study were discussed and it was explained that individual or community decision to participate or not would not in any way jeopardize anti-malarial screening, treatment, or subsequent deployment of ITN.\nVillages were included in the study only after extensive discussions and full approval of the village representatives.\nMany study participants were illiterate, and for these individuals fully informed witnessed verbal consent was obtained from adults and the parents of children involved in the study.\nThe Myanmar National Health authorities granted approval for this study.\nStatistical analysis\nMalariometric data\nThis study was designed as a cluster randomized trial and, therefore, the data were aggregated for each village (i.e. cluster) to obtain cluster-level infection rates.\nThis approach is appealing because the cluster is both the unit of randomization and the unit of analysis.\nThe villages with bed nets were then compared to the other villages using a weighted paired t test.\nThe results of blood smears, collected at the clinics, were used to calculate the incidence density of P. falciparum, defined as the number of malaria episodes per 1,000 child-weeks at risk.\nThe number of weeks at risk for each child was calculated from the start of the study till the end of the study (the last malaria prevalence survey) or the final date they were seen before they moved or died.\nIf children moved in the first half of the study (before the 2nd prevalence survey), the weeks at risk could be not calculated and these children were, therefore, excluded from the analysis of malaria episodes per 1,000 child weeks at risk.\nIf children moved in the second half of the study, the weeks at risk were taken from the first half of the study.\nChildren who had falciparum malaria were treated with mefloquine and artesunate and were considered not at risk of reinfection because of post-treatment prophylaxis for four weeks following treatment.\nThree categorical variables were also calculated, in which each child was classified to have one or more infections with either P. falciparum, P. vivax or \u201call species of malaria (i.e. any malaria)\u201d between 11 May 1998 and 27 February 1999.\nThis was analysed on an intention-to-treat basis.\nData from the cross-sectional surveys were also aggregated to obtain cluster-level prevalences of malaria, anaemia, and malnutrition.\nCost-effectiveness analysis\nFour scenarios were modelled for the costs and benefits of malaria control using ITN and/or EDAET as compared with a baseline of no intervention.\nThe study data on incidence of P. falciparum cases in the individual villages were used, with Poisson distributions assigned to each of these independently, to capture the uncertainty and variability in these findings.\nIncidence data for vivax cases per village were not available and the proportion of children with at least one case of vivax over the study period was, therefore, adapted as a proxy for incidence (likely to be an under-estimate of actual incidence), and converted to the rate per 1,000 person weeks.\nThe costs of diagnosis, treatment and ITN were adapted to present day (2013) recommended strategies and their current costs (i.e. rapid tests for the diagnosis of malaria prior to treatment, the current recommended dosage of artemether-lumefantrine for falciparum malaria, continued use of chloroquine with primaquine for vivax malaria, and the use of long-lasting insecticide treated nets (LLIN) instead of ITN).\nA cost for village malaria workers was also included, that would be required for the EDAET strategy at $1.05 per child per annum (adapted from a study of VMWs in Cambodia).\nThe costs of treatment are based on the currently recommended dose regimen of 3\u00a0days of artemether-lumefantrine (current first line policy); the provider unit costs in Myanmar in 2013 for a full treatment course for children weighing between 5-14\u00a0kg was $0.47 and for children weighing 15-24\u00a0kg it was $0.94.\nThe cost of one chloroquine tablet was $0.021 and for primaquine it was $0.015.\nThe cost of RDTs was estimated at $0.8.\nAn extra 15% was added to the costs of diagnostics and treatment for shipping and wastage.\nThe cost of LLIN including delivery was assumed to be $10 based on a recent review of LLIN costs which conformed with estimates from donors currently active in Myanmar.\nIt was assumed that each ITN is shared by 1.8 people on average (while ITN can be shared by two children a family of five will receive three nets) and that they have a useful life of three years.\nThese costs were applied for each 1,000 person-weeks.\nClinical cases of falciparum malaria were converted to disability-adjusted life-years (DALYs) assuming a mortality rate of 0.1% in treated cases and 1% in untreated P falciparum cases, and a mortality rate of 0.01% and 0.05% in vivax malaria respectively.\nModelling also assumed a one-week total duration of illness for uncomplicated treated malaria and four weeks for untreated malaria.\nResults were generated by running 10,000 Monte Carlo simulations sampling from the probability distributions and calculating the mean costs and effects for each strategy.\nCost-effectiveness acceptability curves were generated to summarize the parameter uncertainties and identify which strategy was most likely to be cost-effective across a range of willingness to pay thresholds.\nResults\nCharacteristics of study groups\nDuring December 1997 and January 1998 data were collected from 1,088 children (13.3% of the total study population) on the prevalence of malaria and palpable spleens.\nThere was considerable variability in the proportions of children with P. falciparum and/or P. vivax parasitaemia, and palpable spleens among the clusters (Table\u00a01).\nIn May 1998, immediately after the ITN were distributed (but before ITN could have had an impact on malaria related pathology), base-line demographic and clinical data of the study children were gathered.\nA total of 8,175 children were recruited from twenty clusters.\nThe age and sex distributions of children were similar for the intervention (4,066 children) and control villages (4,109 children), as were the variables \u201cweight for height\u201d, and the proportions of children with anaemia (Hgb\u2009<\u200910.0\u00a0g/dL).\nIncidence of malaria\nDuring the study-period (10 May 1998 to 28 February 1999) [42\u00a0weeks], 2,408 children visited the clinics with complaints of fever and were confirmed by microscopy to have malaria; 1,102 P. falciparum (46%), 1,257 P. vivax (52%), and 49 mixed i.e. with both species (2%).\nOf the children infected with P. falciparum who visited the clinics, 15.5% came twice with a P. falciparum infection (32 from ITN clusters and 117 children from NN clusters), 1.8% came three times (two from ITN cluster and 15 from NN clusters), and three children (NN cluster) came four times with a P. falciparum infection during the 10-month follow up period.\nThe number of P. falciparum episodes per 1,000 child-weeks at risk was calculated for each cluster (Table\u00a02, Figure\u00a03).\nThere was considerable heterogeneity in the effects.\nEight out of the ten ITN villages showed some protective efficacy against falciparum malaria compared to their control village while the remaining two village pairs showed the opposite trend.\nThe study design (by cluster) provides a limited number of data points so further investigation by stratifying clusters by transmission and other characteristics was not feasible.\nThe protective effect of ITN appeared to increase in both relative and absolute terms with the incidence of malaria in the control village in each cluster pair (Figure\u00a03).\nHowever the weighted difference between the intervention and control villages for the incidence density of P. falciparum was not statistically significant (weighted mean difference of \u22122.6 episodes per 1,000\u00a0weeks at risk, (95% Confidence Interval; -7 to 1.8); p\u2009=\u20090.22).\nTwelve children (1% of acute falciparum malaria) developed severe falciparum malaria; in Dabhine, six severe patients were reported from the NN-villages and two from the ITN-villages.\nIn Myothugyi, one severe patient was reported from the NN-villages and three from the ITN-villages.\nThus, overall seven children developed severe malaria in NN villages and five in ITN villages.\nSeven out of 10 village pairs showed a protective efficacy of ITN for both one or more falciparum malaria episodes and one or more vivax malaria episodes while the remaining three village pairs showed the opposite trend.\nThe overall proportion of children having had one or more malaria episode was 16.3% in ITN villages and 22.5% in NN villages, a weighted difference of \u22126.4% (95% CI; -21.2% to 8.4%), p\u2009=\u20090.35 (Table\u00a03, Figure\u00a04).\nFor P. falciparum this was 8.4% (341/4066) in ITN villages and 15.2% (618 /4109) in NN control villages respectively, a weighted difference of \u22126.9% (95% CI; \u201318.9% to 5.1%), p\u2009=\u20090.23, and for P. vivax, 9.9% (401/4066) in ITN villages versus 12.3% (505/4109) in NN villages, a weighted difference of \u22122.6% (95% CI; -10.9% to 5.7%), p\u2009=\u20090.25 (Table\u00a03).\nAge-groups\nThere was no significant trend for falciparum malaria-infection by age.\nRestricting the analysis to children under five years of age, similar results were observed compared to the whole study group.\nVivax malaria was more common among younger children.\nRelapse of vivax malaria cannot usually be distinguished reliably from incident infection, although first infections of life are by definition incident.\nAmong children under one year of age, 143 (17,5%) had at least one vivax infection during the 10\u00a0month study period (13.7% in ITN villages and 21.0% in NN villages) and children in their second year of age had 185 (19.4%) infections (18.1% in ITN villages and 20.7% in NN villages).\nAfter the second year the proportion of children with an episode of vivax malaria decreased steadily by age and only 3.7% of 9-year-old children (3.8% in ITN villages and 3.7% in NN villages) had an episode of vivax malaria in the same period.\nCross-sectional surveys\nOf the 8175 children registered, 7764 (95%) were present at all three \u2018prevalence\u2019 surveys.\nOf the other 411 children, 46 children reportedly died during the study-period (overall child mortality 5.62% in the 42\u00a0weeks), (20 from NN villages and 26 from ITN villages; p\u2009>\u20090.2) and 365 had moved out of the area or were absent during one or two of the follow-up surveys (NN 142, ITN 223).\nMalaria prevalence\nThe prevalences of children with malaria and splenomegaly were measured at the end of the study (January \u2013 early February 1999), which coincided with the end of the peak transmission season.\nA total of 7,828 children were present at this survey.\nThere was considerable variation in malaria prevalence, both for P. falciparum and P. vivax, and also in the splenomegaly prevalences among the villages (Table\u00a04, and in further detail in Additional files 1, 2, 3 and 4).\nThe overall prevalence of P. falciparum was 5.7% (225) in NN villages and 4.2% (161) in ITN villages.\nSome protective efficacy from impregnated bed nets was seen in six village pairs, while four village pairs showed the opposite effect; overall difference - 1.9% (95% CI; -5.8% to 2.0%), p\u2009=\u20090.30 (Additional files 5, 6 and 7).\nFor P. vivax the overall prevalence was 14.2% (564) in NN and 11.6% (446) in ITN villages.\nFour village pairs suggested protection by ITN while five did not, and in one village pair, both clusters, had no vivax malaria at all; overall difference\u2009=\u2009\u22122.8% (95% CI; - 8.5%, 2.9%); p\u2009=\u20090.30.\nFor splenomegaly prevalence, six village pairs suggested protection by ITN, while four did not; overall difference\u2009=\u2009\u2212 7.3% (95% CI; - 20.1%, 5.5%), p\u2009=\u20090.23.\nThe overall splenomegaly prevalence was 14% in bed net villages and 21% in non-bed net villages.\nAnaemia and malnutrition\nThe prevalences of anaemia (Hb <10\u00a0g/dL) and malnutrition (Wt/Ht; Z-score < \u22122) at baseline in the ITN and NN villages were similar.\nThe weighted mean difference in the prevalence of anaemia was \u22125.2% (ITN-NN; 95% CI: -13.8%, 3.4%) and the weighted mean difference in the prevalence of malnutrition was 0% (ITN-NN; 95% CI: -4%, +4%).\nThe mean haemoglobin level of children during the first cross-sectional survey was lower in Myothugyi (8.85\u00a0g/dL; village means varying from 7.84 to 9.24\u00a0g/dL) than in Dabhine (9.70\u00a0g/dL; village means varying from 9.15 to 10.02\u00a0g/dL).\nThe proportion of anaemic children (Hb\u2009<\u200910\u00a0g/dL) decreased during the study period in all but two clusters (1 NN and 1 ITN village).\nAt the final cross-sectional survey, the decrease in the proportion of children with anaemia in bed net villages was more pronounced than in control villages in 7 cluster pairs while 3 pairs showed the opposite trend (Table\u00a05).\nThe prevalences of anaemia at the end of the study were not significantly different comparing NN and ITN clusters (weighted mean difference of - 3.3% (95% CI; - 11.3, 4.7), p\u2009=\u20090.38 (Additional file 8).\nThe proportion of children with moderate acute malnutrition (Wt/Ht; Z-score\u2009<\u2009\u22122) decreased over the study period in 14 of the 20 villages; 8 ITN villages and 6 NN villages (Table\u00a05).\nAt the final nutrition survey the decrease of malnutrition in ITN villages was greater than in control villages in seven cluster pairs, while three pairs showed a greater decrease in control villages.\nThe malnutrition prevalences at the end of the study were not significantly different comparing NN and ITN clusters, either for all children under 10\u00a0years (\u2212 1.7% (95% CI; - 3.8, 0.4), p\u2009=\u20090.094), or for the children under five years of age (\u2212 1.8% (95% CI; - 5.9, 2.3), p\u2009=\u20090.35) (Additional file 9).\nCost-effectiveness analysis of ITN and EDAET\nThe average weight of a child under 10\u00a0years of age presenting to the malaria programme in Rakhine State was 12.7\u00a0kg, which puts the current cost of the drugs for the treatment of falciparum malaria for an average person in this study area at $0.62 and for vivax malaria at approximately $0.3.\nThe model output suggests that in the absence of ITN or EDAET, 1.4 DALYs are accumulated per 1,000 person-weeks at no cost to the provider.\nThe use of ITN would add a cost of $35 and reduce the number of DALYs to 0.69, or an incremental cost-effectiveness ratio (ICER) of $51/DALY averted.\nThe use of EDAET alone would cost $23 per 1000 person-weeks and reduce DALYs to 0.19, or an ICER of $19 per DALY averted (i.e. nearly three times less).\nWhile both options are considered cost-effective, the implementation of ITN instead of EDAET would incur higher costs and avert less than a third of the number of DALYs.\nWhere EDAET is already in place the introduction of ITN would avert an additional 0.2 DALYs at a cost of $148 per DALY averted.\nThe cost-effectiveness acceptability curves (Figure\u00a05) illustrate these results, and show that if policy makers are willing to pay over approximately $280 per DALY averted, the combination of both ITN and EDAET is likely to be cost-effective.\nIf resources are more constrained then EDAET alone will always dominate the use of ITN alone.\nThis analysis derives from the data in this study and, therefore, specifically to children in this area of Western Myanmar.\nEDAET is likely to be equally effective in adults, but ITNs are likely to be less effective in the adult population who are often outside dwellings in the early evening, and go to sleep later and leave the home earlier than children.\nThus the comparative economic advantage of EDAET over ITN is likely to be even greater in adults and, therefore, the overall advantage of EDAET over ITN for the population overall is predicted to be greater than reported here.\nThis analysis assumes no impact of EDAET on incidence, although a comparison of the baseline and post-intervention surveys in the control villages suggests that EDAET might itself be reducing transmission.\nThis assumption does not impact on the incremental cost/gain of either of the strategies compared with each other, but it could potentially underestimate the relative advantage of EDAET as compared with a baseline of doing nothing, while overestimating that of ITN.\nDiscussion\nThe results of this study suggest that ITNs provided some protection for children against malaria in most villages in this area of Rakhine State, but they did not in the others, and the overall result was that there was no significant benefit evident from their deployment.\nMalaria episodes were less common overall in villages with ITN than in the control group, because differences in favour of ITN were greater than differences in favour of no nets (Figures\u00a03 and 4), but when these data were aggregated per cluster, systematic use of ITN was not found to reduce either falciparum or vivax malaria significantly.\nImportantly they also did not attenuate the adverse consequences of malaria on anaemia or growth.\nThis study focussed on children, as they are the most vulnerable age group for malaria in this area.\nThis provided the highest estimate of potential protective efficacy and thus benefit from ITN, as children go to bed earlier and they sleep longer than adults (particularly relevant in this area with early evening outdoor biting vectors), and they are less likely to travel far from the house.\nAs all villagers in the ITN clusters were provided with a net, any positive outcome could have been further enhanced through a mass insecticidal effect on the malaria vectors.\nIn these villages malaria affected children more than adults but malaria in the South East Asian region is often a disease particularly affecting young mobile adults who work in the forest and do not tend to use nets, so any benefits demonstrated here are likely to be greater than in the total population at risk in the region.\nAdherence with instructions for correct use of ITN was generally good, but as the benefits were small, incorrect use did not lead to more malaria.\nEven though the ITN and control clusters were selected randomly, the malaria prevalence before the study was higher in the study villages than in the control villages.\nThis might indicate a higher endemicity, which could have influenced the incidence and final prevalence results of the ITN clusters, but transmission intensity varies greatly per year and the differences were small and unlikely to make a material difference to the results.\nThis study was tightly controlled, with a high coverage and user rate compared to the normal context of use.\nDespite this no significant overall benefit from ITN deployment in terms of malaria protection could be demonstrated.\nThe proportion of children with anaemia decreased both in villages provided with ITNs and in the control villages over the duration of the study.\nThis is probably largely because diagnosis and effective treatment of malaria were provided in all villages, and because children with anaemia were treated promptly, which emphasizes the benefit provided by prompt and effective anti-malarial treatment in such areas of low unstable transmission.\nThe most likely explanation for this rather disappointing result from ITN deployment in this malaria endemic area of Western Myanmar is the early evening biting pattern and strong preference for outdoor biting of most malaria vectors in this area, as in many other areas of South-East Asia.\nFocussing on the human bite catches between July 1998 and January 2000 (n\u2009=\u20092,895), the overall peak biting time was between 6\u00a0pm and 7\u00a0pm, and over half of the anophelines (51%) were caught before 8\u00a0pm.\nThis would differentiate any benefits of ITNs between children, who are often in or near an ITN at this time, and adults who often are not.\nAs the behaviour of malaria vectors differs so much, particularly in Asia, entomological information is essential before ITNs are deployed.\nThese somewhat negative results contrast with the widespread positive perception of the uniform value of ITNs in malaria control.\nA systematic review of ITN evaluations, which included 22 studies from Africa, Asia and South America, concluded that ITNs reduce childhood morbidity and mortality and that ITNs should be employed in all malarious areas.\nOver the past decade ITNs have been taken up enthusiastically as a key component of many malaria control programmes, but caution is needed in interpreting the findings from bed net studies with different designs and analytical methods.\nAccording to Lengeler, of 29 identified studies conducted in Asia, only four studies used correct statistical procedures for their trials and were included in the meta-analysis.\nAnother study did use correct statistical procedures, but was not included because it studied only pregnant women.\nOf these five \u201cstatistically correct\u201d studies, three showed reduced morbidity while the other two did not.\nIn other studies, non-comparable control groups were used or ITN were allocated at a village-level, whereas end-points were calculated for individuals.\nThere is no doubt concerning the consistent and large benefits provided by ITNs in Africa which fully justifies current deployment initiatives.\nIn contrast, the epidemiology of malaria in Asia is extremely heterogeneous.\nTransmission is highly seasonal and unstable and intensities vary greatly over short distances, and also show great variation from year to year.\nIt is not surprising that ITNs are relatively ineffective in areas of unstable malaria where the principal malaria vectors bite outdoors early in the evening, before people go in or near their ITN.\nThis and previous studies argue against generalising from ITN studies in Africa to the rest of the malaria affected world.\nWhere the main malaria vectors bite mainly indoors after adults and children have gone to sleep, ITNs should be very effective, but in many places, this is not the case.\nThis is not to say that ITNs provide no benefit in these places\u2013 they do protect against malaria, albeit sometimes to a small extent, and if cost were no obstacle then everyone in malaria endemic areas should certainly be provided with an ITN.\nBut funds for malaria control are usually limited and ITN deployment is often included uncritically in many malaria control efforts in Asia in preference to other control measures.\nEarly diagnosis and effective treatment reduces malaria morbidity and mortality, but in a large proportion of patients in south and Southeast Asia, the diagnosis of malaria is still made clinically and, therefore, incorrectly.\nA lack of resources means that most patients with malaria in Myanmar and in adjacent countries under similar constraints still receive inadequate treatment.\nMany patients take a few tablets of artesunate until they feel better.\nSometimes chloroquine is still used which is inexpensive but ineffective.\nArtesunate-based combination treatment is a proven highly effective treatment, which significantly reduces malaria transmission in low-transmission areas.\nWithout accounting for these transmission blocking properties, the present economic model indicates that effective diagnostic and treatment (EDAET) services are less costly and substantially more effective than ITN, when comparing each of these to a baseline of no intervention, which is still the reality in some areas in Myanmar.\nOnce EDAET is implemented effectively, ITNs can still provide additional benefit.\nDuring the extended interval since this study was carried out there has been a consistent and substantial decline in malaria incidence across the region, including many areas in Myanmar even when allowing for improved reporting rates.\nWhile the Rakhine state still accounts for approximately half the burden in Myanmar, the incidence rates reported in this study are likely to be higher than those that would be documented today in this region.\nThis could further strengthen the argument in favour of EDAET.\nLower incidence of malaria will imply that protective measures which require the same costs applied to the entire population will be relatively less cost-effective than those that target infected individuals such as diagnosis and treatment.\nITNs are very popular with international donor agencies.\nMillions of dollars are spent on ITN programmes in this populous malaria endemic region.\nITNs are both provided free by donor organizations or promoted through social marketing, encouraging the population to purchase their own ITN.\nIn this malaria endemic region of Myanmar, the economic modelling based on the study data and contemporary prices suggests that early-diagnosis and effective treatment is substantially more cost-effective than deployment of LLINs as a malaria control measure.\nThis argues for careful evaluation in each region before large-scale deployment.\nThe findings here provide some evidence of a protective efficacy of ITNs in areas of higher endemicity within this region, suggesting that they can be a cost-effective intervention in the context of improving child health.\nIf, however, donor funding is targeting specifically the control of malaria and ultimately its elimination from the region it seems likely that improving preventive, diagnostic and curative interventions for all ages and particularly in adult, often migrant males, could offer better returns on investment.\nConclusion\nMalaria infections, palpable splenomegaly, haemoglobin concentrations, and weight for height, did not differ significantly during the study period between villages with and without ITNs.\nThe limited efficacy of ITNs may be explained by the biting behaviour (peak biting time between 1800 and 1900\u00a0hours, mainly outdoors) of the most common Anopheles mosquito vectors.\nGiven the lack of significant efficacy and relatively high costs of ITNs, the first priority in implementation of malaria control interventions in this area should be the provision of effective diagnosis and treatment.\nWhere EDAET services are already in place and sufficient budgets are available then the use of ITN can be cost-effective.\nClimate and seasonal malaria incidence in Rakhine State in 1998\u20131999. Upper. Monthly rainfall and minimum/maximum temperature, average of Sittwe and Maungdaw Townships (data obtained from township weather stations). Lower. Monthly number of malaria-patients visiting 6 malaria clinics in the region (not only at the study sites).\nTime frame of the study of insecticide treated mosquito nets and entomological surveys  between 1995 and 2000 in the study areas.\nIncidence rate ratio of P.falciparum for ITN versus control (NN) villages by incidence rate (per 1000\u00a0weeks) in the control villages.\nRisk difference in malaria (ever) in ITN villages by proportion compared with malaria in control (NN) villages.\nCost-effectiveness acceptability curves for the four options, accounting for the uncertainty due to the variability in the different study sites and varying levels of willingness to pay per DALY averted.\n\nPlasmodium falciparum, Plasmodium vivax and splenomegaly prevalences in ITN and control villages during the pre-study malaria survey (December 1997)\n\u00a0 | Intervention villages | Control villages\n\u00a0 | N | M.S. | Pf (%) | Pv (%) | Spleen (%) | N | M.S. | Pf (%) | Pv (%) | Spleen (%)\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 707 | 125 | 37 (30) | 35 (28) | 79 (63) | 796 | 102 | 38 (37) | 33 (32) | 68 (67)\nPair 2 | 551 | 89 | 26 (29) | 33 (37) | 47 (53) | 679 | 48 | 9 (19) | 7 (15) | 16 (34)\nPair 3 | 462 | 37 | 8 (22) | 10 (27) | 16 (42) | 298 | 49 | 11 (22) | 6 (12) | 30 (61)\nPair 4 | 223 | 33 | 7 (21) | 8 (24) | 9 (28) | 184 | 32 | 4 (13) | 16 (50) | 19 (59)\nPair 5 | 74 | 18 | 3 (17) | 4 (22) | 5 (28) | 87 | 15 | 8 (53) | 1 (7) | 6 (40)\nSubtotal | 2017 | 302 | 81 (27) | 90 (30) | 156 (52) | 2044 | 246 | 70 (28) | 63 (26) | 139 (57)\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 78 | 16 | 3 (19) | 1 (6) | 6 (38) | 47 | 15 | 3 (20) | 0 (0) | 1 (7)\nPair 7 | 714 | 96 | 1 (1) | 0 (0) | 15 (16) | 819 | 105 | 0 (0) | 0 (0) | 30 (29)\nPair 8 | 492 | 54 | 0 (0) | 1 (2) | 13 (24) | 602 | 70 | 2 (3) | 1 (1) | 6 (9)\nPair 9 | 659 | 57 | 4 (7) | 4 (7) | 7 (12) | 650 | 82 | 1 (1) | 0 (0) | 1 (1)\nPair 10 | 153 | 25 | 10 (40) | 2 (8) | 8 (32) | 120 | 20 | 7 (35) | 1 (5) | 4 (20)\nSubtotal | 2096 | 248 | 18 (7) | 8 (3) | 49 (19) | 2238 | 292 | 13 (4) | 2 (1) | 42(14)\nTotal | 4113 | 550 | 99 (18) | 98 (18) | 205 (37) | 4282 | 538 | 83 (15) | 65 (12) | 181 (34)\n\nN\u2009=\u2009number of children\u2009<\u200910\u00a0yrs (data from local authorities).\nMS\u2009=\u2009number of children of whom a malaria smear was taken.\nPf\u2009=\u2009P. falciparum prevalence (including mixed infections), Pv\u2009=\u2009P. vivax prevalence (including mixed infections).\n\nFalciparum malaria incidence per 1,000\u00a0weeks exposure in ITN and control villages (from 10 May 1998 to 28 February 1999)\n\u00a0 | ITN villages | Control villages | ITN Protective efficacy\n\u00a0 | N | Pf episodes per 1,000\u00a0weeks exposure | N | Pf episodes per 1,000\u00a0weeks exposure | Rate ratio | Rate difference\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 677 | 4.04 | (97 /24034) | 715 | 15.25 | (376 /24662) | 0.26 | \u221211.21\nPair 2 | 522 | 7.37 | (137 /18592) | 653 | 3.51 | (83 /23625) | 2.09 | +3.86\nPair 3 | 429 | 1.57 | (24 /15304) | 280 | 9.38 | (91 /9697) | 0.17 | \u22127.81\nPair 4 | 208 | 2.83 | (21 /7411) | 171 | 8.03 | (48 /5980) | 0.35 | \u22125.20\nPair 5 | 67 | 2.13 | (5 /2343) | 71 | 13.32 | (33 /2477) | 0.16 | \u221211.19\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 74 | 3.68 | (9 /2447) | 44 | 3.84 | (6 /1562) | 0.96 | \u22120.16\nPair 7 | 737 | 0.23 | (6 /26632) | 786 | 0.46 | (13 /28427) | 0.50 | \u22120.23\nPair 8 | 482 | 0.06 | (1 /17344) | 595 | 0.97 | (21 /21637) | 0.06 | \u22120.91\nPair 9 | 653 | 1.61 | (37 /22973) | 631 | 0.57 | (13 /22852) | 2.82 | +1.04\nPair 10 | 140 | 2.90 | (14 /4824) | 107 | 7.68 | (29 /3775) | 0.38 | \u22124.78\nTotal | 3,989 | 2.47 | (351 /141903) | 4053 | 4.93 | (713 /144694) | weighted paired t-test p\u2009=\u20090.216\n\n\nProportion of children (%) with one or more malaria infections during the study\n\u00a0 | \u00a0 | \u00a0 | P. falciparum- ever | P. vivax - ever | Any malaria \u2013 ever\n\u00a0 | ITN (N) | Control (N) | ITN | NN | Risk ratio | Risk difference | ITN | NN | Risk ratio | Risk difference | ITN | NN | Risk ratio | Risk difference\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 695 | 740 | 89 | 305 | 0.31 | - 28.4 | 151 | 243 | 0.66 | - 11.2 | 214 | 429 | 0.53 | - 27.2\n\u00a0 | \u00a0 | \u00a0 | (12.8) | (41.2) | \u00a0 | \u00a0 | (21.7) | (32.9) | \u00a0 | \u00a0 | (30.8) | (58.0) | \u00a0 | \u00a0\nPair 2 | 527 | 662 | 126 | 80 | 1.98 | + 11.8 | 120 | 72 | 2.09 | + 11.9 | 215 | 139 | 1.94 | + 19.8\n\u00a0 | \u00a0 | \u00a0 | (23.9) | (12.1) | \u00a0 | \u00a0 | (22.8) | (10.9) | \u00a0 | \u00a0 | (40.8) | (21.0) | \u00a0 | \u00a0\nPair 3 | 441 | 285 | 26 | 81 | 0.21 | - 22.5 | 46 | 64 | 0.43 | - 13.8 | 67 | 120 | 0.36 | - 26.9\n\u00a0 | \u00a0 | \u00a0 | (5.9) | (28.4) | \u00a0 | \u00a0 | (10.4) | (22.5) | \u00a0 | \u00a0 | (15.2) | (42.1) | \u00a0 | \u00a0\nPair 4 | 213 | 176 | 24 | 46 | 0.43 | - 14.8 | 32 | 60 | 0.44 | - 19.1 | 53 | 89 | 0.49 | - 25.7\n\u00a0 | \u00a0 | \u00a0 | (11.3) | (26.1) | \u00a0 | \u00a0 | (15.0) | (34.1) | \u00a0 | \u00a0 | (24.9) | (50.6) | \u00a0 | \u00a0\nPair 5 | 67 | 72 | 5 | 29 | 0.19 | - 32.8 | 6 | 33 | 0.20 | - 36.8 | 11 | 49 | 0.24 | - 51.7\n\u00a0 | \u00a0 | \u00a0 | (7.5) | (40.3) | \u00a0 | \u00a0 | (9.0) | (45.8) | \u00a0 | \u00a0 | (16.4) | (68.1) | \u00a0 | \u00a0\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 77 | 44 | 8 | 4 | 1.14 | + 1.3 | 6 | 0 | 3.35 | + 5.4 | 12 | 4 | 1.71 | + 6.5\n\u00a0 | \u00a0 | \u00a0 | (10.4) | (9.1) | \u00a0 | \u00a0 | (7.7) | (0) | \u00a0 | \u00a0 | (15.6) | (9.1) | \u00a0 | \u00a0\nPair 7 | 745 | 792 | 6 | 13 | 0.05 | - 0.8 | 1 | 4 | 0.20 | - 0.4 | 7 | 16 | 0.45 | - 1.1\n\u00a0 | \u00a0 | \u00a0 | (0.8) | (1.6) | \u00a0 | \u00a0 | (0.1) | (0.5) | \u00a0 | \u00a0 | (0.9) | (2.0) | \u00a0 | \u00a0\nPair 8 | 489 | 598 | 1 | 23 | 0.05 | - 3.6 | 6 | 13 | 0.56 | - 1.0 | 7 | 34 | 0.25 | - 4.3\n\u00a0 | \u00a0 | \u00a0 | (0.2) | (3.8) | \u00a0 | \u00a0 | (1.2) | (2.2) | \u00a0 | \u00a0 | (1.4) | (5.7) | \u00a0 | \u00a0\nPair 9 | 664 | 632 | 38 | 13 | 2.71 | + 3.6 | 28 | 6 | 4.20 | + 3.2 | 59 | 17 | 3.30 | + 6.2\n\u00a0 | \u00a0 | \u00a0 | (5.7) | (2.1) | \u00a0 | \u00a0 | (4.2) | (1.0) | \u00a0 | \u00a0 | (8.9) | (2.7) | \u00a0 | \u00a0\nPair 10 | 148 | 108 | 18 | 24 | 0.55 | - 10.0 | 5 | 4 | 0.92 | - 0.3 | 19 | 26 | 0.53 | - 11.3\n\u00a0 | \u00a0 | \u00a0 | (12.2) | (22.2) | \u00a0 | \u00a0 | (3.4) | (3.7) | \u00a0 | \u00a0 | (12.8) | (24.1) | \u00a0 | \u00a0\nTotal | 4066 | 4109 | 341 | 618 | \u00a0 | \u00a0 | 401 | 505 | \u00a0 | \u00a0 | 664 | 923 | \u00a0 | \u00a0\n\u00a0 | \u00a0 | \u00a0 | (8.4) | (15.0) | \u00a0 | \u00a0 | (9.9) | (12.3) | \u00a0 | \u00a0 | (16.3) | (22.5) | \u00a0 | \u00a0\nWeighted paired t-test | \u00a0 | \u00a0 | p\u2009=\u20090.23 | p\u2009=\u20090.25 | p\u2009=\u20090.35\n\nP. falciparum-ever includes P. falciparum\u2009+\u2009mixed, P. vivax-ever includes P. vivax\u2009+\u2009mixed. ITN: insecticide treated nets, NN: no nets.\n\nPlasmodium falciparum, Plasmodium vivax and splenomegaly prevalences (%) in ITN and control clusters during the peak season at the end of the study (January 1999)\n\u00a0 | N | N | P. falciparum | P. vivax | Spleen\n\u00a0 | ITN | NN | ITN | NN | Prevalence difference | ITN | NN | Prevalence difference | ITN | NN | Prevalence difference\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 659 | 708 | 31 (5) | 107 (15) | - 10.4 | 183 (28) | 280 (40) | - 11.8 | 170 (26) | 432 (61) | - 35.2\nPair 2 | 508 | 644 | 39 (8) | 35 (5) | + 2.3 | 120 (24) | 135 (21) | + 2.7 | 155 (31) | 131 (20) | + 10.2\nPair 3 | 413 | 268 | 17 (4) | 21 (8) | - 3.7 | 57 (14) | 83 (31) | - 17.2 | 63 (15) | 90 (34) | - 18.3\nPair 4 | 203 | 169 | 10 (5) | 6 (4) | + 1.4 | 23 (11) | 25 (15) | - 3.5 | 56 (28) | 74 (44) | - 16.2\nPair 5 | 62 | 71 | 3 (5) | 7 (10) | - 5.0 | 23 (37) | 22 (31) | + 7.1 | 23 (38) | 31 (44) | - 5.9\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 65 | 43 | 14 (22) | 3 (7) | + 14.6 | 7 (11) | 1 (2) | + 8.4 | 11 (17) | 4 (9) | + 7.6\nPair 7 | 717 | 763 | 2 (0) | 5 (1) | - 0.4 | 0 (0) | 0 (0) | 0 | 6 (1) | 23 (3) | - 2.2\nPair 8 | 462 | 580 | 0 (0) | 6 (1) | - 1.0 | 1 (0) | 4 (1) | - 0.7 | 1 (0) | 9 (2) | - 1.3\nPair 9 | 627 | 620 | 5 (1) | 3 (0) | + 0.3 | 9 (1) | 2 (0) | + 1.1 | 11 (2) | 3 (0) | + 1.3\nPair 10 | 143 | 103 | 40 (28) | 32 (31) | - 3.1 | 23 (16) | 12 (12) | + 4.4 | 37 (26) | 23 (22) | + 3.5\nTotal | 3859 | 3969 | 161 (4) | 225 (6) | \u00a0 | 446 (12) | 564 (14) | \u00a0 | 533 (14) | 818 (21) | \u00a0\nWeighted Paired t-test | \u00a0 | \u00a0 | Pf : -1.9% (\u22125.8%, 2.0%), | Pv : -2.8% (\u22128.5%, 2.9%), | Spleen : -7.3% (\u221220.1%, 5.5%),\np-value\u2009=\u20090.30 | p-value\u2009=\u20090.30 | p-value\u2009=\u20090.23\n\nP. falciparum% and P. vivax% indicate the percentage of children with P. falciparum or P. vivax parasitaemia, both values include mixed infections. ITN: insecticide treated nets, NN: no nets.\n\nPrevalence of anaemia and malnutrition during the 1st and 3rd cross-sectional surveys\nMalnutrition (Wt/Ht < \u22122Z1)\n\u00a0 | ITN clusters | NN clusters\n1st prevalence | 3rd prevalence | Change | 1st prevalence | 3rd prevalence | Change | Relative change\nPair 1 | 87/692 (13) | 67/651 (10) | \u22122.3% | 98/728 (13) | 76/701 (11) | \u22122.6% | \u22120.3%\nPair 2 | 61/516 (12) | 58/506 (11) | \u22120.4% | 86/651 (13) | 102/634 (16) | +2.9% | 3.3%\nPair 3 | 72/435 (17) | 71/413 (17) | +0.6% | 51/277 (18) | 43/266 (16) | \u22122.5% | \u22123.1%\nPair 4 | 62/210 (30) | 28/201 (14) | \u221215.6% | 38/169 (22) | 32/167 (19) | \u22123.3% | 12.3%\nPair 5 | 17/64 (27) | 6/61 (10) | \u221216.7% | 11/71 (15) | 6/70 (9) | \u22126.9% | 9.8%\nPair 6 | 8/77 (10) | 2/65 (3) | \u22127.3% | 3/42 (7) | 5/42 (12) | +4.8% | 12.1%\nPair 7 | 105/735 (14) | 92/701 (13) | \u22121.2% | 158/780 (20) | 92/760 (12) | \u22128.2% | \u22127.0%\nPair 8 | 60/486 (12) | 56/450 (12) | +0.01% | 67/585 (11) | 81/572 (14) | +2.7% | 2.7%\nPair 9 | 87/652 (13) | 55/620 (9) | \u22124.5% | 64/627 (10) | 78/612 (13) | +2.5% | 7.0%\nPair 10 | 39/148 (26) | 8/140 (6) | \u221220.6% | 10/104 (10) | 7/100 (7) | \u22122.6% | 18.0%\nTotal | 598/4015 (15) | 443/3808 (12) | 586/4034 (15) | 586/4034 (15) | 522/3924 (13) | \u22121.2% | 2.1%\nAnaemia (Hb <10\u00a0g/dL)\n\u00a0 | ITN clusters | NN clusters\n\u00a0 | 1st prevalence | 3rd prevalence | Change | 1st prevalence | 3rd prevalence | Change | Relative change\nPair 1 | 362/695 (52) | 299/659 (45) | \u22126.7% | 440/739 (60) | 413/708 (58) | \u22121.2% | 5.5%\nPair 2 | 303/527 (57) | 235/508 (46) | \u221211.2% | 332/662 (50) | 305/644 (47) | \u22122.8% | 8.4%\nPair 3 | 228/441 (52) | 182/413 (44) | \u22127.6% | 152/285 (53) | 125/268 (47) | \u22126.7% | 0.9%\nPair 4 | 133/213 (62) | 88/203 (43) | \u221219.1% | 110/176 (63) | 59/169 (35) | \u221227.6% | - 8.5%\nPair 5 | 46/67 (69) | 22/62 (35) | \u221233.2% | 29/72 (40) | 33/71 (46) | +6.2% | 39.4%\nPair 6 | 76/77 (99) | 52/65 (80) | \u221218.7% | 42/44 (95) | 40/43 (93) | \u22122.4% | 16.3%\nPair 7 | 485/745 (65) | 602/716 (84) | +19.0% | 657/792 (83) | 558/761 (73) | \u22129.6% | \u221228.6%\nPair 8 | 355/489 (73) | 307/461 (67) | \u22126.0% | 532/597 (89) | 473/579 (82) | \u22127.4% | \u22121.4%\nPair 9 | 579/664 (87) | 443/627 (71) | \u221216.5% | 540/632 (85) | 446/619 (72) | \u221213.4% | 3.1%\nPair 10 | 142/148 (96) | 108/143 (76) | \u221220.4% | 104/108 (96) | 95/103 (92) | \u22124.1% | 16.3%\nTotal | 2709/4066 (67) | 2338/3857 (61) | \u22126.0% | 2938/4107 (72) | 2547/3965 (64) | \u22127.3% | \u22121.3%\n\n1 Weighted paired t test 3rd prevalence Wt/Ht - 2 Z score, p-value\u2009=\u20090.094. Relative change\u2009=\u2009(Prev 3 - Prev 1) NN cluster - (Prev 3 - Prev 1) ITN cluster. ITN: insecticide treated nets, NN: no nets.",
        "label": "low",
        "id": "task4_RLD_test_896"
    },
    {
        "paper_doi": "10.1097/qad.0b013e32833e77c9",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Trial design: cluster-RCTUnit of randomization: a 'community' comprising a health clinic, its catchment population and its secondary schoolsNumber of clusters: 30Data collection: a representative survey of 18-22 year olds in study communities 4 years after the intervention. This included a questionnaire, HIV-1, HSV2, and a pregnancy testLength of follow-up: 4 yearsAdjustment for clustering: yes\n\n\nParticipants: Target group: Form 2 pupils (median age 15 years)Sample size: 6791Exclusions: none stated\n\n\nInterventions: The interventionDid the target group receive sexuality education? Yes.How many sessions? Not clear. Reported as an \"in-school 3-year curriculum and 1-year 24 session out-of school programme\".Who delivered the sessions? 'Professional peer educators' (PPEs) - i.e. school leavers who were selected, trained, and supervised and worked in the community for 8 to 10 months.What was the content of the sessions? HIV prevention activities using adapted 'MEMA kwa Vijana' programme with additional materials from 'Talktime', 'Mopani', 'Auntie Stella' and 'Young People We Care' which included self-awareness, communication, self-belief and gender.What additional components were there? A 22-session community programme targeting parents and community stakeholders aimed at improving communication between parents and children and support for adolescent reproductive health. A 5-day residential training programme for clinic nurses to improve accessibility for adolescents.Were condoms distributed free? No.Control group: no intervention (delayed intervention until 2007)\n\n\nOutcomes: Outcomes included in this review:HIV prevalence;HSV2 prevalence;current pregnancy;self-reported sexual debut;use of condoms at last sex.Outcomes not included in this review:knowledge and attitudes around sexual behaviour;reported sexual behavior including multiple sexual partners;use of pregnancy prevention methods with first, last, or any partner;self reported symptoms of STDs.\n\n\nNotes: Country: ZimbabweSetting: rural districtsStudy dates: 2003 to 2007Study sponsors: National Institute of Mental Health, DfID Zimbabw\n\n",
        "objective": "To evaluate the effects of school\u2010based sexual and reproductive health programmes on sexually transmitted infections (such as HIV, herpes simplex virus, and syphilis), and pregnancy among adolescents.",
        "full_paper": "Background\nHIV prevention among young people in southern Africa is a public health priority.\nThere is little rigorous evidence of the effectiveness of different intervention approaches.\nWe describe findings of a cluster randomised trial of a community-based, multi-component HIV and reproductive health intervention aimed at changing social norms for adolescents in rural Zimbabwe.\nMethods\nThirty rural communities were randomised to early or deferred implementation of the intervention in 2003.\nImpact was assessed in a representative survey of 18\u201322 year olds after 4 years.\nParticipants self-completed a questionnaire and gave a dried blood spot sample for HIV and HSV-2 antibody testing.\nYoung women had a urinary pregnancy test.\nAnalyses were by intention-to-treat and were adjusted for clustering.\nFindings\n4,684 18\u201322 year olds participated in the survey (97.1% of eligibles, 55.5% female).\nJust over 40% had been exposed to \u2265 10 intervention sessions.\nThere were modest improvements in knowledge and attitudes among young men and women in intervention communities, but no impact on self-reported sexual behaviour.\nThere was no impact of the intervention on prevalence of HIV or HSV-2 or current pregnancy.\nWomen in intervention communities were less likely to report ever having been pregnancy.\nInterpretation\nDespite an impact on knowledge, some attitudes and on reported pregnancy, there was no impact of this intervention on HIV or HSV-2 prevalence, further evidence that behavioural interventions alone are unlikely to be sufficient to reverse the HIV epidemic.\nThe challenge remains to find effective HIV prevention approaches for young people in the face of continued and unacceptably high HIV incidence, particularly among young women.\nBackground\nRecent surveillance suggests that 2.5 million people become infected with HIV annually.\nForty percent are aged 15\u201324 years.\nOf the six million HIV positive young people in Sub-Saharan Africa, 76% are female.\nThe United Nations General Assembly Special Session on HIV/AIDS declared that by 2005 90% of youth aged 15\u201324 should have the information, education, and life-skills needed to reduce their chances of HIV infection (no country achieved this) and that there should be a 25% reduction in new HIV infections in the worst affected countries (six countries achieved this).\nEncouragingly the proportion of young people who first have sex aged <15 years appears to be declining.\nIn 2004 the Joint United Nations Programme on HIV/AIDS (UNAIDS) commissioned systematic reviews on the effectiveness of HIV prevention interventions for young people.\nThey found good evidence that school-based interventions can reduce reported sexual risk taking.\nThere was also evidence that providing training to make health clinics more \u2018youth friendly\u2019 increases clinic usage.\nHowever, there remains sparse data to support the implementation of community-based approaches, which aim to change societal norms to support individual behaviour change.\nOverall few trials of adolescent HIV prevention interventions with objective biomedical endpoints were identified and one of the key recommendations of the review was that more rigorous research should be funded and conducted.\nWe report the results of a cluster randomised trial to determine the effectiveness of a community-based multi-component HIV prevention intervention for young people, conducted in rural Zimbabwe between 2003 and 2007.\nThe trial had two primary endpoints: prevalence of HIV and of HSV-2 was compared between clusters in early and deferred intervention arms.\nSecondary endpoints included pregnancy prevalence, and reported knowledge, behaviour and attitudes.\nMethods\nStudy population\nThe trial was conducted in 30 communities in seven districts in South-Eastern Zimbabwe (map shown figure 1 suppl data).\nCommunities were randomised to early intervention implementation (2003) or delayed implementation (2007) using restricted randomisation.\nA community comprised a rural clinic, its catchment population and its secondary schools.\nThe Regai Dzive Shiri intervention\nThe intervention has been described in detail elsewhere.\nIt was delivered to young people, parents and clinic staff and was theoretically based in social learning theory and the stages of change model.\nIt aimed to achieve change in societal norms within communities.\nThe intervention had three integrated components:\nThe youth programme for in- and out-of-school youth, was delivered by carefully selected, trained and supervised Zimbabwean school leavers in the year between leaving school and starting university.\nThese professional peer educators (PPEs) lived and worked in the intervention communities for 8\u201310 months of the year.\nThey used theoretically-based materials delivered using participatory methods which aimed to enhance knowledge and develop skills.\nDuring the trial, secondary school attendance in Zimbabwe dropped as a result of the political and economic challenges facing the country and so intervention activities shifted from schools to communities.\nBoth in and out-of-school youth attended activities throughout.\nThe peer educator model was developed by Students Partnership Worldwide (www.SPW.org).\nThe programme for parents and community stakeholders, a 22 session community-based programme, which aimed to improve knowledge about reproductive health, to improve communication between parents and their children and to improve community support for adolescent reproductive health.\nA training programme for nurses and other staff working in rural clinics, which aimed to improve accessibility of clinics for young people.\nStandard HIV prevention activities were implemented through the District AIDS Action Committees across all communities.\nNo other HIV prevention activities aimed specifically at youth were conducted in either arm of the trial during the period of implementation.\nImpact evaluation\nThe original intention was to determine the impact of the intervention within a cohort of Form 2 pupils (9th school year) attending trial secondary schools (n=82).\nAll Form 2 pupils were invited to join the cohort and to participate in the baseline survey (in 2003); 6791 pupils, median age 15 years, participated (87% of eligibles); crude HIV prevalence was 0.8% (95%CI:0.6\u20131.0) and HSV-2 prevalence was 0.2% (95%CI:0.1\u20130.3).\nIt was intended to follow this cohort for four years till 2007.\nHowever during our interim survey in 2006, we found that there had been considerable out-migration (46%).\nThose who remained were of lower risk than those who had left (HIV prevalence among remaining cohort was 1.2% (95%CI:0.7\u20131.9%)).\nWe also conducted a representative population-based survey of 3,960 18\u201321 year olds (91% of eligibles) in 2006 to see if the well publicised decline in HIV incidence in Zimbabwe was likely to reduce the power of the trial.\nCommunity HIV prevalence in females was 8.3% (95%CI:7.0\u20139.8%) and in males was 3.3% (95%CI:2.6\u20134.1).\nRevised sample size calculations suggested that the trial would have considerably reduced power if we followed the original cohort, but that a cross-sectional population-based survey of 18\u201322 year olds would provide >95% power to detect a 40% reduction in HIV prevalence and 80% power to detect a 30% reduction.\nAs the intervention had become increasingly community-based, assessing intervention impact using a population-based survey rather than among the original school-going cohort seemed logical.\nAs a result it was decided at a meeting of investigators and the data and safety monitoring board chair to change the design accordingly.\nFor the final survey, we selected six enumeration areas (EAs) (Census Bureau geographical areas \u2248100 households) in each trial community.\nThe EAs were purposively selected to ensure that sites where intervention activities took place in that community were included (clinics, schools, community centres).\nEach trial community comprised approximately 50 EAs suggesting that around 12% (i.e. those from 6/50 EAs) of 18\u201322 year olds living in trial communities as originally defined were eligible for inclusion in the final survey.\nOf note while the age of the cohort spanned 11 years, the age of participants surveyed at the end of the trial spanned 5 years.\nWithout outmigration from communities, the proportion of original cohort members being included in the final survey was unlikely to be more than 7%.\nAll 18\u201322 year olds who lived in the 180 EAs selected were eligible for inclusion.\nSurvey procedures\nParticipants provided written consent, and completed a questionnaire in two stages; (i) using audio-self administered questionnaire (audio-SAQ), (ii) using audio-computer-assisted-survey-instrument (ACASI)).\nACASI was used for collection of particularly sensitive data and for all questions that required use of complex skip patterns.\nData collected using audio-SAQ were double-entered onto a Microsoft Access password-protected database.\nAll participants were asked to provide a finger-prick blood sample.\nWomen gave a urine sample for pregnancy testing.\nLaboratory procedures\nBlood samples were collected onto filter paper and tested for HIV-1 antibody in Harare using a validated testing algorithm.\nAll specimens were tested using two ELISA tests (Vironostika\u00ae HIV Microelisa System BioMerieux, Inc., NC and AniLabsytems EIA kit (AniLabsystems Ltd, Finland), with additional western blot for discrepant results.\nDried blood spot samples were tested for antibodies to HSV-2 using a type-specific HSV-2 assay (Focus HerpeSelect EIA, Focus Technologies, CA) with the index for diagnosing positive samples raised to >3.4 to improve specificity.\nUrine samples were tested on site for pregnancy using Cortez OneStep hCG Rapidip InstaTest\u00ae.\nStatistical analyses\nStatistical analyses were conducted using Stata 10 (Stata Corp., TX), and were stratified by sex.\nThe primary analysis was intention-to-treat.\nContinuous variables were categorized using recognized cut-off values or dichotomized at the median value.\nAttitudinal and knowledge outcomes were based on questions in 11 domains (suppl data Table 1).\nBinary variables were created, for those participants who answered all questions \u2018correctly\u2019 in each domain.\nUnivariate and multivariate analyses were performed using generalized estimating equations (GEE) with exchangeable correlation matrix and robust standard errors to account for cluster randomisation.\nImpact of the intervention programme on primary and secondary endpoints was assessed by performing an unadjusted GEE analysis comparing the outcomes between study arms.\nCrude odds ratios were adjusted a priori for age, marital status, education and the strata used for randomisation, and further adjusted for any other potential confounding variables showing an imbalance between study arms.\nCox regression, with robust standard errors to allow for clustered design, was used to explore the association between intervention status and age of sexual debut.\nA sub-group analysis was undertaken to assess whether impact varied with intensity of intervention exposure, which compared the impact of the intervention among those participants who had attended Regai Dzive Shiri trial schools and had lived in trial communities over the period of intervention delivery.\nAnalyses were pre-planned and the analytical plan submitted to the data and safety monitoring board before analysis was undertaken.\nEthical approval\nThe trial was approved by the Medical Research Council of Zimbabwe and the ethics committees of University College London Hospitals and the London School of Hygiene & Tropical Medicine.\nResults\nOverall 4,684 of 4,822 (97.1%) eligible individuals identified, participated in the final survey (Figure 1); 2593 (56%) were female.\nThe two trial arms were well balanced (Table 1).\nWomen were more likely than men to report having been married (45.6% vs. 7.6%), and to having married younger.\nMen had lived in communities longer than women and were better educated.\nWomen were more likely to have left school due to pregnancy or marriage.\nOrphaning and poverty were widespread.\nIntervention exposure\nAround 54% of participants (61% of males and 46% of females) had attended a trial school.\nFew participants in intervention communities went to comparison schools or vice versa (<3%) (Table 1).\nOverall 30% (n=695) of survey participants in the intervention arm (48% of males and 15% of females) reported attending an intervention school when a PPE was present and so are likely to have received the in-school intervention (supplementary data Figure 2).\nOverall 20% (n=478) of the intervention arm reported attending \u2265 10 out-of-school youth sessions; 9% (n=212) had attended both the in- and out-of-school intervention, 41% (n=961) had attended either the in-school and/or out-of-school youth programme.\nOne third of males and 61% of females surveyed did not report receiving any intervention (ie had not attended a trial school or the out-of school intervention).\nImpact of the intervention on knowledge and condom self-efficacy\nIn males there was an increase in knowledge related to sexually transmitted disease (STD) acquisition (AOR=1.32;95%CI:1.08\u20131.61) and pregnancy prevention (AOR=1.59;95%CI:1.27\u20131.99) in the intervention arm but not for HIV acquisition (Table 2a).\nThere was no effect on reported self-efficacy.\nIn women there was an increase in knowledge related to STD acquisition (AOR=1.45;95%CI:1.17\u20131.79) and pregnancy prevention (AOR=1.32;95%CI:1.14\u20131.55) in the intervention arm but again not for HIV acquisition (Table 2b).\nThere was a modest impact on reported self-efficacy.\nImpact of the intervention on reported attitudes and behaviours\nAmong men there was no impact on attitudes relating to relationship control overall, although there was an impact on certain items within the scale.\nThere was also no overall impact of the intervention on gender empowerment among men, but again there was on certain items.\nThe intervention did have an impact on women\u2019s attitudes to both relationship control (AOR=1.34;95%CI:1.11\u20131.63) and to gender empowerment (AOR=1.32 95%CI:1.05\u20131.66).\nSexual behaviour\nFemales were more likely to report having had sex than males (53% vs. 42%).\nMedian age of partner at first sex was 24 for females and 17 for males.\nSexually active males reported more lifetime partners (median=2 vs. 1) and more partners in the last 12 months (median 2 vs. 1) than females.\nMales were more likely to report condom use at last sex (81% vs. 59%).\nThere was no effect of the intervention on any of these behavioural outcomes in men or women (Table 2).\nAge at sexual debut was the same across trial arms (adjusted Hazard ratio for males 1.02[95%CI:0.90\u20131.15] and females 0.94[95%CI:0.86\u20131.04].\nThere was also no effect on reported use of pregnancy prevention.\nImpact of the intervention on clinic attendance\nThere was no effect of the intervention on any aspect of clinic attendance (Table 2a and 2b).\nWomen in the intervention arm were more likely to report that they would go to a clinic if they needed to access contraception (AOR=1.33;95%CI:1.05\u20131.69).\nImpact of the intervention on the prevalence of biological outcomes\nEighteen HIV and 119 HSV-2 results were indeterminate (similarly distributed between trial arms).\nIndeterminate results were assumed negative.\nHIV prevalence was 1.5% in males and 7.7% in females, HSV-2 prevalence was 1.6% in males and 10.8% in females; both HIV and HSV-2 prevalence increased steeply with age (supplementary data Figure 3).\nThere was no effect of the intervention on the primary endpoints of HIV or HSV-2 prevalence in either males (Table 3a) or females (Table 3b).\nThe proportion of women with a positive pregnancy test did not differ significantly between arms.\nThere was a reduction in reported current or past pregnancies in the intervention arm among all women (AOR=0.64;95%CI:0.49\u20130.83) and among married women (AOR=0.65; 95%CI:0.49\u20130.87).\nAmong unmarried women, there were non-significant reductions in the number of current pregnancies, and the number of reported past, current and unwanted pregnancies.\nUnmarried women in intervention communities were at significantly lower risk of any pregnancy1 (AOR=0.55;95%CI:0.32\u20130.95) than those in comparison communities.\nThere was no effect of the intervention on reporting of symptoms of STDs in either men or women.\nImpact of the intervention by intensity of intervention exposure\nA subgroup analysis was carried out among participants who attended an RDS trial school and lived in the trial community throughout the duration of the intervention (Table 4).\nEffects of the intervention on knowledge and attitudinal outcomes in this subgroup tended to be somewhat larger than in the full study population, but there was no evidence of an increased impact on the primary outcomes of HIV or HSV-2 in either males or females.\nDiscussion\nThis is one of the first reported trials of a community-based multi-component behavioral intervention aimed at changing social norms for young people in Africa, that had biological endpoints as its primary outcomes.\nThere was no effect of the intervention on any of these including HIV prevalence, HSV-2 prevalence or current pregnancy as measured by pregnancy test.\nThere was an effect on some secondary endpoints including an increase in knowledge related to STD acquisition and pregnancy prevention in intervention communities and a positive effect on some attitudes relating to relationship control and to gender empowerment.\nAmong young women, there was also an increase in reported self-efficacy.\nThere was no effect of the intervention on reported sexual behaviour, reported clinic use or reported use of pregnancy prevention in males or females in intervention communities.\nImportantly there was also no evidence to suggest that young people taking part in the intervention were more sexually \u2018promiscuous\u2019 a concern raised by adult community members and policy makers within Zimbabwe (and elsewhere).\nDespite the change in knowledge and reported attitudes, there was no difference in effect on HIV or HSV-2 prevalence in males or females, nor was there a difference in prevalence of positive pregnancy test between arms of the trial.\nThere was a significant reduction in reported pregnancies among women.\nDespite the widely reported fall in HIV incidence in Zimbabwe, these trial data clearly illustrate that HIV acquisition continues to be a serious public health problem with annual incidence of 4% in young women and 0.8% in young men, the higher rates of infection in women likely reflecting the greater age disparity between themselves and their partners.\nThis trial was carefully designed and implemented.\nThe intervention was theoretically-based and fulfilled all the criteria for likely effectiveness identified by the systematic review of community-based interventions for young people in developing countries.\nQualitative and process evaluation data (not shown) suggested that the intervention model was popular with both adults and young people and that the highly skilled and motivated PPEs were an inspiration to young people within the communities where they worked.\nAlthough such careful selection, training and supporting of PPE is expensive (total cost US$ 400/peer/annum), each peer reaches hundreds of young people and adults within the community in which they work and is able to provide ongoing motivation to key people within the community to become actively involved in HIV prevention.\nHad it been shown to be effective this intervention model had the potential to go to scale, as part of a system of national service for young people.\nThe effect of the intervention was assessed in the wider community rather than just among intervention recipients so the modest intervention effects that we detected are likely to have been diluted.\nOf note the effects on knowledge and attitudes were of similar magnitude to those found with the MEMA kwa Vijana trial which assessed impact in a cohort of intervention recipients.\nWhile we found no evidence of dose-response, however our data on intervention exposure showed considerable inconsistency, which limits our ability to detect this.\nExposure misclassification also reduced our ability to discern the extent to which the intervention impact diffused beyond intervention recipients into the wider population, although a feature of both the out-of-school youth and parents\u2019 intervention was to encourage participants to engage friends and relatives in discussions around HIV prevention.\nClearly improving knowledge and changing attitudes of young people are important endpoints in their own right.\nWhile the rates of comprehensive knowledge (participants responding to all items correctly) were low, when responses to individual items were examined separately (data not shown) levels were somewhat better at 50\u201360%.\nChanging attitudes relating to gender issues is thought to be a particularly important prerequisite to changing the HIV risk environment and it is encouraging that the intervention was able to make modest gains in this regard, particularly among young women.\nHowever, beliefs relating to the rights of women are deeply entrenched, and will consequently be difficult to change.\nEncouragingly there was some evidence of an effect on reported past pregnancy in married women and all pregnancies in unmarried women.\nWhile self-reported data on pregnancy need to be interpreted with caution, there was no evidence of reporting bias for current pregnancy (actual and reported pregnancy were highly correlated).\nWe also demonstrated that pregnancy prevention knowledge was increased in both men and women and that young women in the intervention arm were more likely to report that they would be able to go to the clinic to access contraception.\nUnlike antibiotic treatment for STIs, contraception was consistently available at clinics throughout the trial.\nSeveral studies in Africa, including this one, have found that pregnancy is of more immediate concern to young women than HIV.\nTo date there have been relatively few published trials of HIV prevention interventions among young people in Africa that have used objective endpoints.\nMEMA kwa Vijana, a trial of a school-based intervention supported by a youth-friendly clinic intervention in Tanzania found that intervention recipients had increased knowledge as well as reductions in reported unsafe sexual behaviour.\nThere was however no effect on biological outcomes.\nA follow-up survey conducted 5 years after the trial found that while some knowledge, attitude and behavioural change persisted, again there was no evidence of long-term impact on biological endpoints.\nJewkes et al conducted a cluster randomised trial of the \u201cStepping Stones\u201d community-based intervention in rural South Africa and were able to show an effect on knowledge, attitudes and behaviour as well as a reduction in incidence of HSV-2 that was of borderline significance.\nThere were some limitations to intervention implementation which may have undermined its potential effectiveness.\nFirstly, in Zimbabwe as in many other countries there is ambivalence about promoting condoms to young people and mention of condoms was not permitted in schools.\nMany churches continue to denigrate their effectiveness.\nWhile our intervention was able to tackle this issue to some extent by educating young people and adults within communities, young people were aware of these conflicting opinions and this may have undermined intervention impact.\nThe recent HIV prevention series in the Lancet called for better scaling up of interventions that are known to be effective.\nFinding ways to convince and educate policy makers and other stakeholders that adolescent interventions in general and condom promotion in particular do no harm is clearly going to be key to successful scale-up.\nA further limitation to intervention implementation was the high level of mobility of young people.\nRelatively few survey participants were exposed to both in and out-of-school programmes.\nOnly 55% of young women and 68% of young men had lived in trial communities throughout the four years of the trial; women were more likely to leave or enter the community around the time of marriage than men.\nWhile this sub-optimal exposure of survey participants (as opposed to those living in communities during the course of the trial) reduces the likelihood of demonstrating intervention effectiveness, it also likely reduces the possibility that a critical mass of young people were reached, sufficient to bring about change in community norms to reduce sexual risk taking.\nQuantifying the extent to which the high mobility could have affected intervention effectiveness or our measurement of it is not possible with the data collected.\nThis high mobility emphasises the need to bring interventions to scale to ensure that people who move from one community to another continue to be able to access HIV prevention programmes and are not disadvantaged by their mobility.\nOrphans and women who are married (two groups with high rates of HIV) are those most likely to have moved and are also those most in need or HIV prevention education.\nIn addition, the political and economic climate in Zimbabwe during the trial made many aspects of intervention implementation and evaluation challenging.\nFor example, it became difficult to implement the intervention within secondary schools for political reasons.\nThis coincided with a fall in school attendance for economic reasons.\nWe therefore shifted delivery of the intervention into the community from 2005.\nIn addition the clinics suffered from high rates of staff attrition; trained staff were frequently lost and those remaining were increasingly overstretched.\nThe STD drugs supply at clinics was poor.\nIn focus group discussions, young people cited lack of drug availability as a reason for non-attendance.\nPopulation mobility also affected intervention evaluation and resulted in a change in the design of the trial from a cohort to serial cross-sectional evaluation which diluted intervention effects.\nThe rationale for altering the design was not only based on the high rate of out-migration from communities but also on the fact that the intervention had become increasingly community-based.\nWhile this clearly affected our ability to detect intervention effectiveness it is not possible to say whether the trial outcomes would have been different if it had been implemented under different circumstances.\nThis trial is one of several that contributed to a recently published systematic review of behavioural interventions for HIV prevention.\nDisappointingly none of the behavioural interventions included in that review had an impact on HIV endpoints either because the interventions were ineffective or because of trial design/implementation issues.\nA consensus seems to be emerging however, that behavioural interventions alone are unlikely to be sufficient to reverse the HIV epidemic and that it is likely that combination approaches which integrate behavioural, biomedical and structural components will be more effective at a population level.\nFinding ways to implement those HIV prevention interventions that are known to be effective (such as male circumcision, HIV testing and counselling, condom promotion) while looking for innovative ways to combine or layer them, in addition to searching for novel intervention approaches, is the challenge for the next generation of HIV prevention research.\nThe urgency of meeting this challenge could not be more clearly exemplified than by the unacceptably high HIV incidence among the young Zimbabweans in this trial, particularly the young women.\nTrial design\n\nCharacteristics of final evaluation survey participants\nCharacteristic | Male n (%) | Female n (%)\n\nControl (n=1001) | Intervention (n=1078) | Control (n=1352) | Intervention (n=1241)\nAge:\n18 years | 364 (36.4) | 388 (36.0) | 515 (38.1) | 441 (35.5)\n19\u201320 yrs | 356 (35.6) | 355 (32.9) | 422 (31.2) | 373 (30.1)\n21\u201322 yrs | 281 (28.1) | 335 (31.1) | 415 (30.7) | 427 (34.4)\n\nReligion:\nCatholic | 192 (19.2) | 208 (19.3) | 240 (17.8) | 230 (18.5)\nAnglican | 281 (28.1) | 279 (25.9) | 345 (25.5) | 322 (26.0)\nApostolic | 203 (20.3) | 212 (19.7) | 315 (23.3) | 266 (21.4)\nPentecostal | 91 (9.1) | 92 (8.5) | 173 (12.8) | 149 (12.0)\nOther/none | 219 (21.9) | 278 (25.8) | 263 (19.4) | 265 (21.4)\nMissing | 15 (1.5) | 9 (0.8) | 16 (1.2) | 9 (0.7)\n\nEver married | 72 (7.2) | 84 (7.8) | 599 (44.3) | 579 (46.7)\nMissing | 9 (0.9) | 8 (0.7) | 6 (0.4) | 3 (0.2)\n\nMarried aged \u2264 16 years | 10 (1.0) | 14 (1.3) | 228 (16.7) | 239 (19.3)\nMissing | 29 (40.3) | 30 (35.7) | 138 (23.0) | 104 (18.0)\n\nLived in community \u2265 5 years | 692 (69.1) | 738 (68.5) | 760 (56.2) | 672 (54.2)\nMissing | 94 (9.4) | 81 (7.5) | 156 (11.5) | 141 (11.4)\n\nLevel of education:\nNone/primary only | 106 (10.6) | 118 (10.9) | 201 (14.9) | 180 (14.5)\nF1\u20132 | 118 (11.8) | 142 (13.2) | 181 (13.4) | 187 (15.1)\nF3\u20134 | 635 (63.4) | 661 (61.3) | 825 (61.0) | 752 (60.6)\nF5 or higher | 137 (13.7) | 149 (13.8) | 135 (10.0) | 118 (9.5)\nMissing | 5 (0.5) | 8 (0.7) | 10 (0.7) | 4 (0.3)\n\nOrphan status:\nNon-orphan | 498 (49.7) | 566 (52.6) | 718 (53.1) | 666 (53.7)\nLost one/both parents | 494 (49.4) | 493 (45.7) | 622 (46.0) | 565 (45.5)\nMissing | 9 (0.9) | 19 (1.8) | 12 (0.9) | 10 (0.8)\n\nSocio-economic status:\nCannot afford soap to wash clothes | 209 (20.9) | 244 (22.6) | 278 (20.6) | 268 (21.6)\nMissing | 47 (4.7) | 67 (6.2) | 54 (4.0) | 55 (4.4)\n | \nChild/children in house receiving external assistance1 | 181 (18.1) | 236 (21.9) | 225 (16.6) | 197 (15.9)\nMissing | 6 (0.6) | 12 (1.1) | 5 (0.4) | 4 (0.3)\n | \nAdult in house skipped meal in last week | 162 (16.2) | 203\u201318.8 | 254 (18.8) | 222 (17.9)\nMissing | 8 (0.8) | 7 (0.7) | 8 (0.6) | 3 (0.2)\nParticipant gone day without food in last week | 148 (14.8) | 176 (16.3) | 204 (15.1) | 174 (14.0)\nMissing | 8 (0.8) | 7 (0.7) | 3 (0.2) | 4 (0.3)\n\nAttended RDS study school:\nControl school | 623 (62.2) | 22 (2.0) | 693 51.3) | 35 (2.8)\nIntervention school | 22 (2.2) | 661 (61.3) | 45 (3.3) | 569 (45.8)\nNon-RDS School | 210 (21.0) | 234 (21.7) | 348 (25.7) | 409 (33.0)\nNo secondary education | 119 (11.9) | 138 (12.8) | 238 (17.6) | 206 (16.6)\nMissing | 27 (2.7) | 23 (2.1) | 28 (2.1) | 22 (1.8)\n\nExternal assistance includes financial, food, education assistance provided by government or aid.\n\nImpact of the intervention on population prevalence of knowledge, attitudinal and behavioural outcomes \u2013 males\nEndpoint | Prevalence1 | Crude | Adjusted2\nControl (N=1001) | Intervention (N=1078)\n\nn/N | (%) | n/N | % | OR | OR | [95% CI]\nKnowledge and self-efficacy (% responding \u201ccorrectly\u201d to questions)\nHIV acquisition (3 questions ) | 229/1000 | (22.9) | 264/1074 | (24.6) | 1.10 | 1.09 | [0.88\u20131.35]\nSTD acquisition (2 questions) | 407/1000 | (40.7) | 502/1074 | (46.7) | 1.29 | 1.32 | [1.08\u20131.61]\nPregnancy prevention (2 questions) | 261/995 | (26.2) | 380/1073 | (35.4) | 1.54 | 1.59 | [1.27\u20131.99]\nCondom self-efficacy (3 questions) | 448/989 | (45.3) | 524/1067 | (49.1) | 1.12 | 1.18 | [0.94\u20131.48]\nSexual refusal self-efficacy (2 questions) | 638/964 | (66.2) | 661/1031 | (64.1) | 0.91 | 0.92 | [0.74\u20131.14]\nHIV-testing self-efficacy (3 questions) | 616/990 | (62.2) | 685/1065 | (64.3) | 1.09 | 1.08 | [0.89\u20131.30]\n\nAttitudes - Control over sex (% responding \u201ccorrectly\u201d to questions)\nAll responses \u201ccorrect\u201d (10 questions) | 38/912 | (4.2) | 54/977 | (5.5) | 1.36 | 1.44 | [0.90\u20132.32]\n\u2265 7/10 questions responded to \u201ccorrectly\u201d 3 | 525/912 | (57.6) | 598/977 | (61.2) | 1.16 | 1.18 | [0.94\u20131.48]\nControl around sexual refusal (3 questions) | 229/954 | (24.0) | 277/1023 | (27.1) | 1.17 | 1.22 | [1.02\u20131.47]\nControl around sexual partners (4 questions) | 323/934 | (34.6) | 363/997 | (36.4) | 1.08 | 1.08 | [0.87\u20131.32]\nSafe sex and condoms (2 questions) | 342/956 | (35.8) | 411/1024 | (40.1) | 1.20 | 1.20 | [0.95\u20131.52]\n\nAttitudes - Jewkes scale: Gender empowerment (% responding \u201ccorrectly\u201d to questions)\n\u2265 4/8 responses \u201ccorrect\u201d3 | 490/946 | (51.8) | 546/1010 | (54.1) | 1.09 | 1.12 | [0.93\u20131.35]\nRight to refuse sex (2 questions) | 465/968 | (48.0) | 542/1038 | (52.2) | 1.18 | 1.20 | [0.98\u20131.46]\nRights within marriage (2 questions) | 14/966 | (1.4) | 27/1041 | (2.6) | 1.81 | 1.79 | [1.05\u20133.04]\n\nControl over life & future\nHave long range goals | 845/991 | (85.3) | 931/1070 | (87.0) | 1.16 | 1.19 | [0.94\u20131.51]\n\nReported Sexual Behaviour (reported on ACASI)\nEver had sex | 402/974 | (41.3) | 442/1038 | (42.6) | 1.07 | 1.04 | [0.87\u20131.24]\nSexual debut 17 or younger4 | 189/974 | (19.4) | 201/1038 | (19.4) | 1.01 | 1.01 | [0.78\u20131.31]\nTwo or more lifetime partners4 | 278/974 | (28.5) | 303/1038 | (29.2) | 1.04 | 1.03 | [0.80\u20131.31]\nTwo or more partners in last 12m4 | 117/789 | (14.8) | 109/818 | (13.3) | 0.89 | 0.86 | [0.59\u20131.26]\nDid not use condom at last sex 4 | 179/971 | (18.4) | 202/1035 | (19.5) | 1.08 | 1.03 | [0.83\u20131.29]\n\nReported Pregnancy prevention\nNo pregnancy prevention used with first partner 5 | 172/420 | (41.0) | 179/459 | (39.0) | 0.92 | 0.90 | [0.69\u20131.17]\nNo pregnancy prevention used with last partner 5 | 175/420 | (41.7) | 179/459 | (39.0) | 0.89 | 0.87 | [0.64\u20131.17]\nNo pregnancy prevention used with any partner 5 | 130/420 | (31.0) | 133/459 | (29.0) | 0.91 | 0.87 | [0.63\u20131.21]\n\nClinic attendance and perceptions of staff\nBeen to the clinic in the last 12 months | 447/999 | (44.7) | 482/1075 | (44.8) | 0.99 | 0.99 | [0.76\u20131.29]\nNever worry that clinic staff will tell others purpose of my visit6 | 252/399 | (63.2) | 281/426 | (66.0) | 1.13 | 1.10 | [0.81\u20131.51]\nAlways seen in private, never worry that other patients will know purpose of my visit 6 | 300/399 | (75.2) | 314/426 | (73.7) | 0.89 | 0.87 | [0.66\u20131.14]\nWould go to clinic for treatment if had discharge from penis | 756/986 | (76.7) | 845/1062 | (79.6) | 1.18 | 1.19 | [0.90\u20131.57]\n\nDenominators vary depending on missing values\nAdjusted for a priori confounders (age, strata, marital status & education)\nCut-off set at median number of \u201ccorrect\u201d responses\nReference category includes not reporting the characteristic and does not exclude those who have never had sex\nRestricted to those who reported ever having had sex (includes those who reported non-consensual sex, anal sex, or sex when too drunk to say no)\nRestricted to those who visited the clinic in the last 12 months\n\nImpact of the intervention on population prevalence of knowledge, attitudinal and behavioural outcomes \u2013 females\nEndpoint | Prevalence1 | Crude | Adjusted2\nControl (N=1352) | Interventio (N=1241)\n\nn/N | % | n/N | % | OR | OR | [95% CI]\nKnowledge and self-efficacy (% responding \u201ccorrectly\u201d to questions)\nHIV acquisition (3 questions ) | 233/1351 | (17.2) | 246/1241 | (19.8) | 1.19 | 1.16 | [0.92\u20131.45]\nSTD acquisition (2 questions) | 464/1350 | (34.4) | 524/1239 | (42.3) | 1.45 | 1.45 | [1.17\u20131.79]\nPregnancy prevention (2 questions) | 355/1351 | (26.3) | 404/1239 | (32.6) | 1.36 | 1.32 | [1.14\u20131.55]\nCondom self-efficacy (3 questions) | 311/1335 | (23.3) | 339/1223 | (27.7) | 1.27 | 1.22 | [1.01\u20131.48]\nSexual refusal self-efficacy (2 questions) | 887/1329 | (66.7) | 847/1215 | (69.7) | 1.16 | 1.17 | [0.95\u20131.43]\nHIV-testing self-efficacy (3 questions) | 897/1335 | (67.2) | 872/1222 | (71.4) | 1.22 | 1.22 | [1.03\u20131.44]\n\nAttitudes - Control over sex (% responding \u201ccorrectly\u201d to questions)\nAll responses \u201ccorrect\u201d (10 questions) | 47/1181 | (4.0) | 60/1091 | (5.5) | 1.42 | 1.36 | [0.87\u20132.14]\n\u2265 7/10 questions responded to \u201ccorrectly\u201d 3 | 586/1181 | (49.6) | 616/1091 | (56.5) | 1.34 | 1.34 | [1.11\u20131.63]\nControl around sexual refusal (3 questions) | 304/1274 | (23.9) | 301/1162 | (25.9) | 1.12 | 1.16 | [0.95\u20131.43]\nControl around sexual partners (4 questions) | 373/1231 | (30.3) | 378/1137 | (33.2) | 1.15 | 1.14 | [0.91\u20131.43]\nSafe sex and condoms (2 questions) | 406/1272 | (31.9) | 430/1162 | (37.0) | 1.25 | 1.24 | [1.03\u20131.48]\n\nAttitudes - Jewkes scale: Gender empowerment (% responding \u201ccorrectly\u201d to questions)\n\u2265 4/8 responses \u201ccorrect\u201d 3 | 569/1268 | (44.9) | 596/1157 | (51.5) | 1.31 | 1.32 | [1.05\u20131.66]\nRight to refuse sex (2 questions) | 585/1309 | (44.7) | 576/1192 | (48.3) | 1.18 | 1.17 | [0.95\u20131.44]\nRights within marriage (2 questions) | 33/1315 | (2.5) | 31/1201 | (2.6) | 1.04 | 1.19 | [0.74\u20131.91]\n\nControl over life & future\nHave long range goals | 1126/1334 | (84.4) | 1054/1232 | (85.6) | 1.10 | 1.10 | [0.88\u20131.38]\n\nReported Sexual Behaviour (reported on ACASI)\nEver had sex | 681/1289 | (52.8) | 648/1217 | (53.2) | 1.01 | 0.83 | [0.61\u20131.13]\nSexual debut 17 or younger 4 | 298/1289 | (23.1) | 295/1217 | (24.2) | 1.01 | 1.02 | [0.80\u20131.28]\nTwo or more lifetime partners 4 | 138/1289 | (10.7) | 142/1217 | (11.7) | 1.12 | 1.11 | [0.79\u20131.56]\nTwo or more partners in last 12m 4 | 35/1102 | (3.2) | 27/957 | (2.8) | 0.89 | 0.91 | [0.56\u20131.47]\nDid not use condom at last sex 4 | 514/1282 | (40.1) | 498/1209 | (41.2) | 1.04 | 0.93 | [0.72\u20131.20]\n\nReported Pregnancy prevention\nNo pregnancy prevention used with first partner 5 | 372/696 | (53.4) | 352/667 | (52.8) | 0.97 | 0.97 | [0.76\u20131.25]\nNo pregnancy prevention used with last partner 5 | 369/696 | (53.0) | 361/667 | (54.1) | 1.04 | 1.04 | [0.77\u20131.40]\nNo pregnancy prevention used with any partner 5 | 345/696 | (49.6) | 329/667 | (49.3) | 0.98 | 0.99 | [0.74\u20131.30]\n\nClinic attendance and perceptions of staff\nBeen to the clinic in the last 12 months | 782/1340 | (58.4) | 729/1238 | (58.9) | 1.01 | 0.98 | [0.76\u20131.28]\nNever worry that clinic staff will tell others purpose of my visit6 | 472/706 | (66.9) | 447/661 | (67.6) | 1.03 | 1.04 | 0.80\u20131.36]\nAlways seen in private, never worry that other patients will know purpose of my visit 6 | 556/706 | (78.8) | 517/661 | (78.2) | 0.96 | 0.96 | [0.72\u20131.28]\nAble to go to the clinic if I needed to get contraception | 933/1294 | (72.1) | 928/1195 | (77.7) | 1.36 | 1.33 | [1.05\u20131.69]\n\nDenominators vary depending on missing values\nAdjusted for a priori confounders (age, strata, marital status & education)\nCut-off set at median number of \u201ccorrect\u201d responses\nReference category includes not reporting the characteristic and does not exclude those who have never had sex\nRestricted to those who reported ever having had sex (includes those who reported non-consensual sex, anal sex, or sex when too drunk to say no)\nRestricted to those who visited the clinic in the last 12 months\n\nImpact of the intervention on population prevalence of biological outcomes males\nEndpoint | Prevalence1 | Crude | Adjusted2\nControl (N=1001) | Intervention (N=1078)\n\nn | % | n | % | OR | [95% CI] | OR | [95% CI]\nReported symptoms of STDs\nEver had symptoms of STD 3 | 145/974 | 14.9 | 157/1038 | 15.1 | 1.02 | [0.79\u20131.32] | 0.98 | [0.76\u20131.25]\nSought treatment for STD symptoms 3,4 | 72/145 | 49.7 | 74/157 | 47.1 | 0.89 | [0.49\u20131.62] | 0.82 | [0.44\u20131.53]\nGenital discharge prevalence | 83/950 | 8.7 | 95/1023 | 9.3 | 1.08 | [0.77\u20131.51] | 1.09 | [0.81\u20131.46]\nGenital warts or sores prevalence | 84/950 | 8.8 | 84/1013 | 8.3 | 0.95 | [0.65\u20131.40] | 0.92 | [0.67\u20131.27]\nPrevalence of any symptom of STD | 367/991 | 37.0 | 407/1060 | 38.4 | 1.06 | [0.88\u20131.27] | 1.06 | [0.90\u20131.24]\n\nPrimary biological outcomes\nHIV infection | 13/1001 | 1.3 | 18/1078 | 1.7 | 1.28 | [0.68\u20132.41] | 1.20 | [0.66\u20132.18]\nHSV-2 infection | 15/1001 | 1.5 | 19/1078 | 1.8 | 1.13 | [0.65\u20131.96] | 1.23 | [0.69\u20132.18]\n\nDenominators vary depending on missing values\nAdjusted for a priori confounders (age, strata, marital status & education)\nReported on ACASI\nAmong those who reported symptoms of STDs on ACASI\n\nImpact of the intervention on population prevalence of biological outcomes females\nEndpoint | Prevalence1 | Crude | Adjusted2\nControl (N=1352) | Intervention (N=1241)\n\nn | % | n | % | OR | [95% CI] | OR | [95% CI]\nPregnancy and reported pregnancy\nAll women (n=2593)\n\u2003Currently pregnant5 | 109/1349 | 8.1 | 95/1237 | 7.7 | 0.94 | [0.69\u20131.28] | 0.92 | [0.70\u20131.19]\n\u2003Reported unwanted pregnancy | 183/1324 | 13.8 | 159/1218 | 13.0 | 0.93 | [0.70\u20131.23] | 0.88 | [0.69\u20131.12]\n\u2003Reported past or current pregnancy | 572/1346 | 42.5 | 517/1235 | 41.9 | 0.97 | [0.75\u20131.27] | 0.64 | [0.49\u20130.83]\n\u2003Reported aborted pregnancy | 31/1332 | 2.3 | 36/1224 | 2.9 | 1.30 | [0.85\u20132.00] | 1.26 | [0.82\u20131.94]\n\u2003Any evidence of pregnancy (incl.currently pregnant5) | 600/1352 | 44.4 | 541/1241 | 43.6 | 0.97 | [0.74\u20131.27] | 0.64 | [0.49\u20130.83]\n\nUnmarried women (n=1406)\n\u2003Currently pregnant5 | 20/745 | 2.7 | 11/656 | 1.7 | 0.63 | [0.30\u20131.31] | 0.66 | [0.32\u20131.36]\n\u2003Reported unwanted pregnancy | 24/731 | 3.3 | 13/648 | 2.0 | 0.61 | [0.24\u20131.53] | 0.54 | [0.19\u20131.54]\n\u2003Reported past or current pregnancy | 37/743 | 5.0 | 21/655 | 3.2 | 0.64 | [0.32\u20131.28] | 0.60 | [0.27\u20131.31]\n\u2003Reported aborted pregnancy | 8/737 | 1.1 | 8/648 | 1.2 | 1.07 | [0.46\u20132.52] | 0.98 | [0.42\u20132.25]\n\u2003Any evidence of pregnancy (incl.currently pregnant5) | 58/747 | 7.8 | 31/659 | 4.7 | 0.59 | [0.36\u20130.95] | 0.55 | [0.32\u20130.95]\n\nMarried women (n=1178)\n\u2003Currently pregnant5 | 89/598 | 14.9 | 84/578 | 14.5 | 0.99 | [0.74\u20131.33] | 1.02 | [0.78\u20131.35]\n\u2003Reported unwanted pregnancy | 158/587 | 26.9 | 145/567 | 25.6 | 0.93 | [0.68\u20131.26] | 0.93 | [0.72\u20131.19]\n\u2003Reported past or current pregnancy | 533/597 | 89.3 | 495/577 | 85.8 | 0.72 | [0.54\u20130.95] | 0.65 | [0.49\u20130.87]\n\u2003Reported aborted pregnancy | 22/589 | 3.7 | 27/573 | 4.7 | 1.30 | [0.77\u20132.20] | 1.20 | [0.63\u20132.26]\n\u2003Any evidence of pregnancy (incl.currently pregnant5) | 540/599 | 90.2 | 509/579 | 87.9 | 0.79 | [0.60\u20131.06] | 0.70 | [0.53\u20130.93]\n\nReported symptoms of STDs\nEver had symptoms of STD3 | 222/1289 | 17.2 | 209/1217 | 17.2 | 1.00 | [0.80\u20131.25] | 0.97 | [0.79\u20131.19]\nSought treatment for STD symptoms3,4 | 100/222 | 45.0 | 93/209 | 44.5 | 0.98 | [0.67\u20131.43] | 0.91 | [0.62\u20131.35]\nGenital discharge prevalence | 160/1297 | 12.3 | 139/1191 | 11.7 | 0.94 | [0.71\u20131.23] | 0.91 | [0.70\u20131.19]\nGenital warts or sores prevalence | 112/1280 | 8.8 | 83/1164 | 7.1 | 0.80 | [0.59\u20131.09] | 0.78 | [0.57\u20131.05]\nPrevalence of any symptom of STD | 482/1336 | 36.1 | 411/1231 | 33.4 | 0.89 | [0.73\u20131.08] | 0.86 | [0.72\u20131.02]\n\nPrimary biological outcomes\nHIV infection | 98/1352 | 7.2 | 101/1241 | 8.1 | 1.15 | [0.78\u20131.69] | 1.15 | [0.81\u20131.64]\nHSV-2 infection | 132/1352 | 9.8 | 148/1241 | 11.9 | 1.26 | [0.91\u20131.74] | 1.24 | [0.93\u20131.65]\n\nDenominators vary depending on missing values\nAdjusted for a priori confounders (age, strata, marital status & education)\nReported on ACASI\nAmong those who reported symptoms of STDs on ACASI\nBased on result of pregnancy test\n\nSub-analysis restricted to survey participants who attended a Regai Dzive Shiri trial school and had lived in the community for the duration of the intervention (i.e. 5 years or more)\nEndpoint | Male | Female\nControl | Intervention | Crude | Adjusted1 | Control | Intervention | Crude | Adjusted1\n | \n% | % | OR | OR | [95% CI] | % | % | OR | OR | [95% CI]\nParticipants who had lived in trial community 5 years or more and attended an RDS trial school\nn | 485 | 519 |  |  |  | 493 | 399 |  |  | \nHIV | 1.4 | 1.5 | 1.07 | 0.91 | [0.35\u20132.34] | 3.8 | 6.5 | 1.77 | 1.65 | [0.90\u20133.03]\n\n |  |  |  |  |  |  |  |  |  | \nHSV-22 | 0.8 | 1.4 | 1.40 | 1.34 | [0.51\u20133.53] | 5.9 | 7.5 | 1.30 | 1.21 | [0.71\u20132.05]\nPregnancy |  |  |  |  |  | 5.9 | 5.3 | 0.90 | 0.83 | [0.50\u20131.35]\nAny evidence of pregnancy (incl.currently pregnant3) |  |  |  |  |  | 33.5 | 31.1 | 0.87 | 0.49 | [0.29\u20130.84]\n\nKnowledge and self-efficacy (% responding \u201ccorrectly\u201d to questions)\nHIV acquisition (3 questions ) | 25.0 | 25.5 | 1.03 | 1.01 | [0.73\u20131.41] | 15.8 | 22.3 | 1.56 | 1.52 | [0.98\u20132.37]\nSTD acquisition (2 questions) | 43.6 | 50.4 | 1.31 | 1.30 | [1.00\u20131.68] | 35.9 | 41.6 | 1.29 | 1.23 | [0.91\u20131.66]\nPregnancy prevention (2 questions) | 26.0 | 41.9 | 2.05 | 2.05 | [1.51\u20132.77] | 26.8 | 36.6 | 1.59 | 1.56 | [1.18\u20132.07]\n\nAttitudes - Control over sex (% responding \u201ccorrectly\u201d to questions)\n\u2265 7/10 questions responded to \u201ccorrectly\u201d4 | 61.2 | 63.3 | 1.08 | 1.07 | [0.76\u20131.50] | 53.0 | 60.8 | 1.38 | 1.37 | [1.04\u20131.80]\nControl around sexual refusal (3 questions) | 26.9 | 30.2 | 1.17 | 1.19 | [0.93\u20131.52] | 26.9 | 33.2 | 1.35 | 1.48 | [1.17\u20131.87]\nControl around sexual partners (4 questions) | 36.7 | 37.6 | 1.06 | 1.02 | [0.82\u20131.26] | 34.9 | 36.9 | 1.09 | 1.06 | [0.79\u20131.44]\nSafe sex and condoms (2 questions) | 37.5 | 40.2 | 1.11 | 1.11 | [0.82\u20131.50] | 31.8 | 39.5 | 1.38 | 1.35 | [0.98\u20131.85]\n\nAttitudes - Jewkes scale: Gender empowerment (% responding \u201ccorrectly\u201d to questions)\n\u2265 4/8 responses \u201ccorrect\u201d4 | 49.4 | 56.9 | 1.34 | 1.40 | [1.05\u20131.87] | 44.1 | 56.1 | 1.62 | 1.58 | [1.20\u20132.10]\nRight to refuse sex (2 questions) | 48.5 | 53.8 | 1.24 | 1.24 | [0.97\u20131.59] | 44.0 | 49.5 | 1.25 | 1.20 | [0.91\u20131.59]\n\nAdjusted for a priori confounders (age, strata, marital status & education)\nAdjusted OR obtained using logistic regression with robust standard errors to allow for clustering\nBased on result of pregnancy test\nCut-off set at median number of \u201ccorrect\u201d responses",
        "label": "unclear",
        "id": "task4_RLD_test_177"
    },
    {
        "paper_doi": "10.4314/ahs.v13i2.44",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: DesigncRCTAllocation of clusters10 villages randomized to intervention, 9 to control\n\n\nParticipants: 558 children younger than age 5\n\n\nInterventions: Primarily education\n\n\nOutcomes: Any STH\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Design and implementation of participatory hygiene and sanitation transformation (PHAST) as a strategy to control soil-transmitted helminth infections in Luweero, Uganda\nDesign and implementation of participatory hygiene and sanitation transformation (PHAST) as a strategy to control soil-transmitted helminth infections in Luweero, Uganda\nAfrican Health Sciences\nAfr H. Sci.\nBackground:\nThe study is a continuation of a research carried out in Luweero district in Uganda 1 .\nIt investigated whether PHAST was a suitable tool for reducing transmission of soil transmitted helminths.\nPHAST means Participatory Hygiene and Sanitation Transformation; a participatory approach that uses visual tools to stimulate the participation of people in promotion of improved hygiene and sanitation.\nObjective: To assess the effect of PHAST on intestinal helminth transmission in children under five years.\nMethods: Three phases namely; (1) Baseline survey (2) PHAST intervention (3) Follow up were conducted.\nDuring Phase 1, the subjects' stool samples were examined for presence of helminthic ova and questionnaires administered.\nIn Phase 2, PHAST was conducted only in experimental villages.\nAll subjects in the experimental and control villages were treated thrice with Albendazole.\nDuring Phase 3, all steps of Phase 1 were repeated.\nResults: There was an overall reduction in the prevalence of children infected with helminths after PHAST intervention.\nAlso, comparison of pre-intervention and post-intervention multivariate results indicates that the likelihood of children getting infected with helminths reduced in most of the experimented variables.\nConclusion: Health stakeholders should utilize PHAST approach to sensitize communities on the importance of hygiene to curb soil-transmitted helminth infections.\nIntroduction\nIntestinal helminth infections have been reported in many parts of Uganda including Luweero district [1][2][3][4] .\nThe district has had inadequate investigations on the problem, yet it has had the infections for quite a long time 3 ; it was therefore necessary to investigate a way of controlling its transmission.\nThus, this study was conducted to assess the effect of PHAST intervention on intestinal helminth infections in children less than 5 years.\nPHAST is a participatory approach, which was developed to encourage people to analyze their own situation and identify key problems, decide what things need to be improved, plan how they are going to do it and then act.\nFor many years, conventional messages on hygiene and sanitation had been known and largely understood by people.\nHowever, these messages had not translated into significant improvement in good hygiene practices.\nIn 1993, WHO and the Regional Water and Sanitation Group for East and Southern Africa (RWSG-ESA) initiated the PHAST strategy to address this concern 5 .\nAmong the PHAST tools that were developed, 4 were purposely selected as most appropriate for the current study to cover 5 of the essential steps to community planning namely; Sanitation ladder, Three-Pile Sorting, Faecal-Oral disease Transmission Routes and Barriers (FTRB), and Tippy Tap 6 .\nSanitation ladder (set of pictures showing various methods of excreta disposal) was selected for problem identification.\nParticipants arranged pictures in form of ladder, from the worst practice/latrine to the best; identifying their own situation and looking at advantages of moving up the ladder.\nThree-Pile Sorting was for analyzing problems and selecting options; participants sorted out 30 pictures of hygiene/sanitation related situations depending on whether they considered them \"good\", \"bad\" or \"in-between\" giving reasons in each case.\nFTRB was for planning for solutions; participants organized a set of pictures basing on what they knew about faecal-oral transmission routes and then worked out how to block the routes using common barrier pictures.\nTippy tap was for planning for new facilities and behaviour change; this was a hand washing facility demonstrated to participants; they were encouraged to make and use it in their homes.\nMonitoring was done after every training session and family members assessed their progress using pictorial monitoring forms.\nDespite the fact that PHAST had been employed in some programmes such as RUWASA, effects of the use of participatory methods had not been systematically monitored or documented.\nRecorded effects were merely anecdotal and it lacked baseline surveys to prove its effectiveness 5,7 .\nThis study therefore tested the approach as no studies had been carried out to assess its impact.\nMethods\nThe study was a randomized community intervention trial with pre-and post-intervention phases.\nThe study was implemented in 3 phases.\nPhase 1 was a baseline cross-sectional descriptive survey that investigated the prevailing helminth status described in already published paper 1 .\nTwo sub-counties, selected by simple random sampling, consisted of 4 parishes from which 19 study villages were studied.\nStool samples from 727 eligible children were examined for presence of different types of helminth ova using Kato-Katz 8 technique.\nSemi-structured questionnaires were also administered to parents/ guardians and inspection of households conducted to assess their hygiene status.\nDuring Phase 2, the four parishes were randomly assigned 10 experimental and 9 control villages with 357 and 370 children respectively.\nPHAST health education was carried out thrice among the experimental group (parents/guardians) only.\nAfter each training session, the respondents' households were visited to reinforce what had been discussed during the training.\nAt each visit, household members freely discussed and identified their household sanitation and hygiene status they had attained plus the ones they were aiming at using pictorial monitoring forms.\nIn addition, all the children were treated with a single oral dose of Albendazole depending on age once every 3 months; those below 2 years were given one 200-mg tablet whereas those between 2 and 5 years two 200-mg tablets.\nIt was a directly observed therapy.\nDuring Phase 3, all steps and procedures in Phase 1 were repeated.\nThe results were analyzed using univariate and bivariate analyses.\nChi-square test was used and the level of significance set to 95% level.\nThe relationship between the variables was established as statistically significant when found to be equal to or less than 0.05 9 .\nThe odds ratios (OR) of the children were determined using the odds ratio statistic in a 2x2 analysis.\nMultivariate analysis was further used to investigate how helminth infection was related to more than one variable at a time while controlling for confounders.\nA binary logistic regression model was used to obtain the adjusted OR.\nResults\nTable 1 indicates that the prevalence rate of children infected with helminth ova which was 27.6% (201/ 727) at baseline reduced to 16.5% after PHAST intervention.\nComparison of Phase 1 and 3 preintervention (pr) and post-intervention (po) multivariate results for different variables also indicate that the odds of children getting infected with helminth ova reduced after PHAST intervention in all the variables except three; these either had an increase or a constant Odds Ratio (tables 2a and 2b).\nIf OR after PHAST is lower than OR before PHAST, there was a reduction in helminth infections.\nThe variables that indicated the most significant reduction were; condition of latrines, respondents' hand washing after handling children's faeces and keeping of pigs.\nThese variables were OR -Odds ratio CI -Cornfield 95% confidence limits for OR # -Respondents did not wash hands * -Most significant OR reductions after PHAST intervention \u00a7 -Children pushing against the ground as a way of cleaning themselves after visiting significantly associated with helminth infection before the intervention but were not associated after the intervention.\nFor instance, there was nearly a threefold reduction in OR noted among helminth-infected children who were living in homes with poorly maintained latrines compared to those in homes with fairly maintained latrines (OR pr = 1.90; 95% CI = 1.17 -3.10 Vs OR po = 0.74; 95% CI = 0.30 -1.80).\nFurthermore, of the children infected with helminths, the likelihood of those who lived in homes with pigs getting infected also decreased by a half after the intervention compared to those in homes without pigs, (OR pr = 1.73; 95% CI = 1.17 -2.58 Vs OR po = 0.82; 95% CI = 0.48 -1.39).\nTable 3 indicates that generally, there was a statistically significant decline in prevalence in the control group (chi square =16.90, p= 0.00).\nVariable\nPre-intervention -\nOR -Odds ratio CI -Cornfield 95% confidence limits for OR\n# -Respondents did not wash hands * -Most significant OR reductions after PHAST intervention \u00a7 -Children pushing against the ground as a way of cleaning themselves after visiting\nThere was also a statistically significant decline in the prevalence in experimental group (Chi-square = 6.41, p= 0.01).\nA difference in the decline in prevalence for the control and experimental groups was noted but not statistically significant (The control group declined 6 percent points more than the experimental group (Chi-square = 1.83, p > 0.05).\nDiscussion\nEffect of PHAST intervention on helminthic infections\nThe overall prevalence rate of helminth infections reduced after the intervention in Phase 3.\nThe likelihood of children getting infected with helminths also reduced in almost all the variables.\nThe observed reduction in these aspects is likely to be attributed to PHAST intervention.\nAs stipulated by Simpson-Herbert et al., 1997 6 , tools and techniques that were used during PHAST participatory health education in the experimental group could have stimulated participation even among participants who did not know how to read.\nBy taking them through the various steps and activities, it was possible to help them understand that poor hygiene and sanitation behaviours and practices are the principal causes of many preventable diseases, such as helminthiasis.\nWhereas conventional health education, which is normally delivered by didactic teaching, is aimed at reducing transmission and re-infection by encouraging healthy behaviour, it is not normally presented as a user friendly package involving participants.\nUNDP recommends that participatory methods create a non-threatening environment in which all participants regardless of class, age and sex can express their views freely 10 .\nThis freedom was observed during PHAST sessions where almost every participant was able to identify their own problems and plan for solutions in a relaxed atmosphere.\nEvidently, PHAST approach was an eye opener to the respondents as noted by a reduction in odds ratio after the intervention in most aspects.\nReduction, for instance, was noted in the risk of children acquiring infections through respondents' hand washing with soap after visiting the latrine and after handling children's faeces.\nReduction in the likelihood of children getting infected with worms in homes with mud floors indicates that, despite the fact that floors remained uncemented, participants cleaned them better after PHAST exposure.\nDecrease in the risk of children acquiring infections for those living in homes with soiled compounds could be attributed to respondents' becoming aware that indiscriminate dumping of faeces was harmful and therefore ensured that it was properly disposed of.\nThe improvement in the type of latrine, latrine floor and condition was linked with the newly constructed or improved latrines observed during Phase 2 monitoring.\nThis in turn resulted in the reduction of risk of children acquiring worm infections.\nThe decline in odds of children acquiring worm infection in homes keeping pigs could be attributed to respondents who could have restricted the roaming of pigs after the intervention, thus reducing on indiscriminate defaecation.\nAccording to Scott 11 , the association with pig ownership is intriguing giving the continuing interest in zoonotic potential of transmission of Ascaris from pigs to humans.\nThe results of respondents who had difficulties in accessing water show that, despite the problem, they could have started using it sparingly for improving hygiene practices.\nHowever, an increase in the risk of children acquiring worm infections for those who cleaned their anal region by pushing themselves against the ground (sliding) after visiting the latrine, could be attributed to the fact that it was not easy for the parents/guardians to teach their kids the proper way of cleaning themselves as they were either busy with domestic chores or away from home whenever they defaecated.\nThe fact that the risk of children getting infected for those living in homes with refuse dumps increased whereas that of children in homes without hand washing facilities (tippy taps) remained the same after intervention could imply that more home visits were still required to sensitize respondents to clean up compounds and to produce more of such facilities.\nIn view of the findings discussed above that there were significant changes in hygiene related behaviour after the intervention, it is possible that within the short run, chemotherapy has a greater effect on worm prevalence generally than the PHAST intervention as indicated in Table 3.\nThe increase in hygiene related behaviour in various variables, however, suggests that PHAST intervention would have a greater effect on worm prevalence but in a long run.\nIn this respect, it should be noted that least decline was recorded in Kasala parish (2.4%) where PHAST sessions were least attended and highest decline in Kisimula (15%) where there was good attendance yet the parish had had highest prevalence (38.8%) prior intervention.\nConclusion\nIt is possible that higher reductions would have been registered if PHAST activities were conducted for a longer period.\nSwitzerland et al., 2010 12 reports that PHAST is a process that takes time.\nThe study findings, therefore, raise questions on how long and how frequently PHAST has to be implemented in order to be fully effective.\nThis requires further research.\nAccording to Loevinsohn 13 , successful health education depends on using a few messages, of proven benefit, repeatedly, and in many fora.\nTable 1 : Comparison of Phases 1 and 3 helminth infections -Phase 3 percentages are based on a total of 92 infected cases, 101 includes double infectionsThere was a high drop rate basically due to migration from study area; hence the difference in the two study populations during Phase 1 & 3 (727 and 558 respectively). Phase 3 data contains both experimental and control groups.\nType of worm | No. of cases before intervention No. of cases after intervention\n | -Phase 1 (n = 727) |  | -Phase 3 (n = 558) | \n | No. of infections | (%) | No. of infections \u00de (%)\nAncylostoma duodenale /Necator americans | 165 | (82.1) | 64 | (71.1)\nAscaris lumbricoides | 38 | (18.9) | 10 | (11.1)\nTrichuris trichiura | 14 | (7.0) | 22 | (24.4)\nHymenolepis nana | 1 | (1.0) | 0 | 0\nEnterobius vermicularis | 2 | (0.5) | 5 | (5.6)\nTotal | 220 | 27.6 | 101 | 16.5\n\u00de\nTable 2a : Comparison of pre-intervention and post-intervention multivariate results\nVariable | Pre-intervention -Phase 1 | Post-intervention -Phase 3\n | Adjusted OR (95% CI) | Adjusted OR (95% CI)\nLevel of Education |  | \nNone | 1.06 (0.70 -1.63) | 0.63 (0.34 -1.17)\nAt least primary education |  | \nRespondents' hand washing after latrine | \nNothing # /Water only | 1.03 (0.55 -1.94) | 0.48 (0.21 -1.08)\nWashing with soap |  | \nDo after handling children's faeces |  | \nNothing # /Water only | 1.79 (1.03 -3.11) | 0.87* (0.46 -1.63)\nWashing with soap |  | \nCleaning of child after latrine |  | \nSliding  \u00a7 | 0.46 (0.30 -0.70) | 1.85 (1.05 -3.26)\nLeaves/Bathing |  | \nType of floor of house |  | \nMud | 1.14 (0.62 -2.09) | 0.60 (0.24 -1.51)\nCemented |  | \nSoiling the compound |  | \nYes | 1.05 (0.68 -1.64) | 0.50 (0.14 -1.78)\nNo |  | \nPresence of refuse dumps |  | \nYes | 1.12 (0.73 -1.71) | 1.71 (0.61 -4.81)\nNo |  | \nStatus of compound |  | \nPoorly maintained | 1.23 (0.79 -1.90) | 0.80 (0.37 -1.71)\nFairly maintained |  | \nEasy |  | \nTable 2b : Comparison of pre-intervention and post-intervention multivariate results\n | Phase 1 | Post-intervention -Phase 3\n | Adjusted OR (95% CI) | Adjusted OR (95% CI)\nKeep pigs |  | \nYes | 1.73 (1.17 -2.58) | 0.82* (0.48 -1.39)\nNo |  | \nType of latrine |  | \nTNR  \u2021 | 1.16 (0.76 -1.80) | 0.73 (0.41 -1.29)\nTIL \u00b1 |  | \nLatrine floor |  | \nLogs/Mud | 1.32 (0.76 -2.32) | 1.09 (0.47 -2.53)\nConcrete |  | \nCondition of latrine |  | \nPoor | 1.90 (1.17 -3.10) | 0.74* (0.30 -1.80)\nFair |  | \nHand washing facility |  | \nNo | 0.78 (0.38 -1.61) | 0.79 (0.40 -1.59)\nYes |  | \nAccessibility to water |  | \nDifficult | 1.28 (0.82 -1.99) | 0.85 (0.47 -1.55)\nEasy |  | \nTable 3 : Comparison of prevalence of helminth infections of the study groups for Phase 1 \u00a5 and Phase 3 \u00a9\nStudy groups Prevalence (%) | Prevalence (%) | Percent Decline | \n | Per Parish | Per study | Per | Per study | Per Parish | Per study | Chi-Square \u20ac\n |  | Arm | Parish | Arm |  | Arm | (P-value)\nControl | (24.6) | 29.7 | (10.2) | 15.9 | 14.4 | 14.5 | 16.90\n |  |  |  |  |  |  | (0.00)\n | (34.2) |  | (20.7) |  | 13.5 |  | \nExperimental (12.3) | 23.5 | (9.9) | 17.1 | 2.4 | 8.5 | 6.41\n | (38.8) |  | (23.8) |  | 15.0 |  | (0.01)\n |  | 27.6 |  | 16.5 |  | 11.1 | 1.83 0.18\n\u00a5 Phases 1 -Pre-intervention \u00a9 Phases 3 -Post-intervention \u2022 Mantel -Haenszel's Chi-square used",
        "label": "unclear",
        "id": "task4_RLD_test_800"
    },
    {
        "paper_doi": "10.1155/2018/3629643",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallelEthics and informed consent: yes\nFollow-up period: 1 yearSample size estimate: noITT analysis: yes, number randomised: 51, number analysed: 51Funding: not reportedPreregistration: not reported\n\n\nParticipants: Location: Korea; single-centre (hospital)\nIntervention group: 30, control group: 21Mean age: intervention group 38.77 +- 1.68, control group 41.38 +- 10.92\nInclusion criteria: > 20 years, acute multi-tissue hand injury of moderate severity (assessed by HISS score 21-50), underwent reconstruction within 3 days after injury by two surgeons.\nExclusion criteria: history of impaired motor function, injury to the peripheral nerves and/or vessels distal to the wrist, or a bone fracture requiring transarticular fixation with a Kirchner (K) wire, a congenital hand deformity, an operation history on the same hand, and underlying diseases including autoimmune diseases such as rheumatoid arthritis or systemic lupus erythematosus or those taking medications that could influence wound healing.\n\n\nInterventions: Aim/s: to compare outcomes in patients with acute hand injury who were managed with or without NPWT after reconstructive surgery.\n Group 1 (NPWT) intervention: NPWT (CuraVAC, CGBio, Seongnam-si, Gyeonggi-do, Korea) applied at a pressure of 75 mmHg in continuous mode and secondary dressing including Vaseline gauze.Group 2 (control) intervention: conventional dressing, including vaseline gauze was applied over the closed skin using polyurethane foam with a compressible elastic bandage, and a short arm splint was applied in a functional position; dressing and NPWT were changed every 3 days.\nStudy date/s: January 2013 to December 2016\n\n\nOutcomes: SSI/infectionHaematomaWound disruption (dehiscence)Validity of measure/s: unclear what definition was used for infectionTime points: 1 month and 1 year\n\n\nNotes: \n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "In this study, we compared outcomes in patients with acute hand injury, who were managed with or without negative pressure wound therapy (NPWT) after reconstructive surgery.\nAll of the patients who sustained acute and multitissue injuries of the hand were identified.\nAfter reconstructive surgery, a conventional dressing was applied in Group 1 and NPWT was applied in Group 2.\nThe dressing and NPWT were changed every 3 days.\nThe mean age and Hand Injury Severity Scoring System score of both groups were not significantly different.\nDisabilities of the Arm, Shoulder, and Hand (DASH) scores were evaluated 1 month after all the sutures were removed and 1 year postoperatively, which were both significantly lower in Group 2.\nApplying NPWT to the hand promoted wound healing by reducing edema, stabilizing the wound, and providing immobilization in a functional position.\nEarly wound healing and decreased complications enabled early rehabilitation, which led to successful functional recovery, both objectively and subjectively.\n1. Introduction\nA significant proportion of hand injury cases are multiple faceted and heavily contaminated and involve composite soft tissue and bone injuries due to the complexity of the anatomy and function of the hand.\nAs a result, hand injuries are often difficult to manage promptly and require multiple staged serial treatment.\nOn the other hand, functional recovery is as important as structural reconstruction and resurfacing in hand injuries, as the hand is a functional unit.\nEarly exercise and rehabilitation improve functional recovery; therefore, wound healing should be achieved as soon as possible.\nNegative pressure wound therapy (NPWT) is a good alternative not only for management during the preoperative period of early reconstruction, but also for early recovery after reconstruction.\nNPWT has been widely used for almost every type of wound, from acute traumatic wounds to chronic intractable wounds.\nIt generates a subatmospheric pressure of 50\u2212150\u2009mmHg in either a continuous or an intermittent mode.\nAlthough the exact mechanism is undefined, the effects of NPWT are to remove excess fluid and debris, improve tissue perfusion, and promote wound healing by enhancing formation of granulation tissue and decreasing the size of the wounds.\nIn this study, we compared outcomes in patients with acute hand injury who were managed with or without NPWT after reconstructive surgery.\n2. Materials and Methods\nThis study was approved by the Institutional Review Board.\nAll of the data were analyzed anonymously and according to the principles in the 1975 Declaration of Helsinki, revised in 2008.\nThe study was a prospective open trial.\nAll of the adult patients (>20 years) who sustained acute multitissue injury of the hand from January 2013 to December 2016 were enrolled with the following criteria.\nThe patients included in this study sustained acute hand injury of a similar severity, as assessed by a Hand Injury Severity Scoring System (HISS) score of 21\u221250 (Table 1), which is defined as a moderate severity level II injury, and underwent reconstruction within 3 days after injury by two surgeons.\nPatients with a medical history of impaired motor function, injury to the peripheral nerves and/or vessels distal to the wrist, or a bone fracture requiring transarticular fixation with a Kirchner (K) wire, a congenital hand deformity, an operation history on the same hand, and underlying diseases including autoimmune diseases such as rheumatoid arthritis or systemic lupus erythematosus or those taking medications that could influence wound healing were excluded from the study.\nInformed consent was obtained from patients who met the inclusion criteria before randomization.\nPatients were randomly assigned to the control or experimental group following a simple randomization procedure (computerized random numbers) achieved using opaque envelopes.\nReconstruction was performed according to the injury on a case-by-case basis.\nBone fractures including fractures of the phalangeal, metacarpal, and carpal bones were fixed with K-wires averting articular surfaces, and the tip of the K-wire was closely cut and embedded under the skin.\nOpen reduction and ligament repair were performed as required for dislocated joints.\nTendons were repaired accordingly for tendon rupture or avulsion injuries.\nThe skin lacerations were closed primarily, and skin and soft tissue defects were reconstructed with local flaps or a skin graft.\nA silastic drain was inserted before closure.\nAfter reconstruction, a conventional dressing was applied over the closed skin using polyurethane foam with a compressible elastic bandage, and a short arm splint was applied in a functional position in Group 1 (control group).\nBy contrast, NPWT (CuraVAC\u00ae, CGBio, Seongnam-si, Gyeonggi-do, Korea) was applied at a pressure of 75\u2009mmHg in continuous mode in Group 2 (experimental group).\nThe secondary dressing for Group 2, including Vaseline gauze, was applied before NPWT.\nThe dressing and NPWT were changed every 3 days.\nIn both groups, when the skin was completely healed within 2 weeks after the injury, the dressing or NPWT was removed, followed by the sutures.\nPhysical therapy was started under consultation with the rehabilitation medicine department after wound healing, and the allocation information to each group was not provided to reduce bias.\nPhysical therapy was performed twice weekly for 4 weeks with individual home-exercise instruction.\nData were collected from the patient's medical records and radiographs.\nThe baseline characteristics collected were age, sex, date of injury, injury site, and the HISS score.\nTime to recover over 90% of the full range of motion (ROM) compared to the normal values of full flexion and extension was analyzed for every interphalangeal and metacarpal joint.\nIn addition, the Disability of the Arm, Shoulder, and Hand (DASH) score was evaluated 1 month after suture removal, when the skin was completely healed and 1 year postoperatively.\nEvaluated complications were hematoma, infection, wound disruption, or a secondary operation.\nComparisons between the two groups were performed using chi-square test and Fisher's exact test.\nA p value < 0.05 was considered statistically significant.\n3. Results\nWe identified 51 patients (17 females and 34 males; age: from 21\u201361 years; mean age: 39.8 years) with acute hand injuries who met the study inclusion criteria.\nA total of 21 patients received conventional dressing using polyurethane foam and a short arm splint, and 30 patients received NPWT.\nThe mean age of Group 1 was 41.4 (range: 22\u201361) years and that of Group 2 was 39.9 (range: 21\u201361) years.\nThe mean HISS score of Group 1 was 33.6 (range: 21\u201350) and that of Group 2 was 35.7 (range, 21\u201350).\nNo significant differences were observed in patient demographics or HISS scores between the two groups.\nDASH scores were evaluated 1 month after all of the sutures were removed and 1 year postoperatively.\nThe scores at 1 month averaged 33.14 (range: 18.3\u201348.3) in Group 1 and 22.67 (range: 5.8\u201340.1) in Group 2 (p = 0.031).\nThe score at 1 year averaged 22.08 (range: 14.9\u201331.9) in Group 1 and 20.99 (range: 5.1\u201332.0) in Group 2 (p = 0.667).\nThe hand joints recovered >90% of the full ROM at 46.9 (range: 30\u201361) days after injury in Group 1 and at 33.3 (range, 22\u201358) days in Group 2 (p = 0.022).\nThere were five complications: two hematomas and one infection were treated conservatively by drainage and antibiotics in Group 1, and two wound macerations in Group 2 healed conservatively without additional surgery.\nNo difference in complications was observed between the two groups.\nThe statistical comparisons between the two groups are presented in Table 2.\nCase 1. A 59-year-old male visited the emergency room after his hand had been smashed in a heavy rolling machine.\nAll of the dorsal skin on the hand was avulsed with multiple ruptures of the extensors (Figure 1).\nThe ruptured second and third extensor digitorum communis and fifth extensor digitorum minimi were repaired, the avulsed skin envelope was tension- free repaired, and a silastic drain was inserted (Figure 2).\nThen the whole dorsal side of the hand except the fingers was covered with NPWT (Figure 3).\nNPWT was changed every 3 days.\nFull ROM was achieved without restriction of daily activity 4 weeks after suture removal (Figure 4).\nCase 2. A 54-year-old male suffered a multitissue injury of the right second to fifth fingers in a press machine accident.\nThe third proximal phalangeal bone was fractured in an avulsed manner.\nThe second and third flexor tendons were also ruptured, and multiple skin defects occurred on the volar side of the hand (Figure 5).\nThe fractured bone was reduced while repairing the ruptured flexor tendons, and the lacerations and skin defects were repaired with skin grafts.\nThen, the whole volar side of the hand was covered with NPWT, which was changed every 3 days (Figure 6).\nThe wound was healed 2 weeks later, and the sutures were removed.\nAfter 2 months of rehabilitation and physical therapy, the patient was able to use his hands freely with full flexion and extension and returned to work (Figure 7).\n4. Discussion\nNPWT was first reported in 1993 and was introduced as \u201cvacuum-assisted closure\u201d for wound control and treatment by Morykwas et al. in 1997.\nSince then, NPWT has been widely used not only for chronic nonhealing wounds, but also for acute traumatic injuries.\nIts effectiveness is thought to be due to decreased bacterial count, increased tissue perfusion, removal of exudates, and promotion of granulation tissue formation, all of which promote wound healing.\nThe NPWT system consists of foam connected to a vacuum pump through a connecting tube, and the whole system is covered with a semiocclusive dressing.\nThe application of NPWT has been expanded from managing and protecting the wound and preparing for final reconstruction to improving skin graft outcomes and patient comfort and thereby reducing cost.\nNPWT has been mostly used in patients undergoing hand surgery with soft tissue defects associated with trauma, burns, or infection.\nThe effective use of NPWT in preparing soft tissue defects before reconstruction has been well described, and favorable results have been achieved in patients with bone, tendon, or nerve exposure.\nOn the other hand, use of NPWT after reconstruction has only been reported in selective cases.\nMost commonly, NPWT has been applied after skin grafting.\nNPWT stabilizes the graft and promotes adherence of the skin graft, which improves graft take.\nThe hand is a functional and mobile unit with a complex anatomy.\nThus, reconstruction of hand injuries should focus not only on resurfacing with healthy soft tissue, but also on maintaining good muscle strength and flexibility of the tendons without adhesion.\nMany joints of the hand require early rehabilitation of the ROM to prevent contracture.\nNPWT can be applied after reconstruction and offers several advantages.\nIt can be used instead of conventional polyurethane foam and short arm splint dressings.\nNPWT simplifies the dressing while stabilizing the hand.\nUse of NPWT splints of the hand in a functional position, and the hand can be molded into the desired functional position before applying suction.\nThe absence of a splint allows easy visualization of the position and status of the hand.\nMoreover, NPWT with foam allows only minimal motion of the joints, functioning as a sort of dynamic splint.\nNPWT as a partial dynamic splint has two advantages.\nOne is that NPWT helps decrease swelling, which leads to better overall hand function.\nHand wounds can remain swollen for some time following injury and become more swollen after reconstruction.\nReduced swelling fosters early recovery of tissues, which leads to early rehabilitation.\nAs in our cases, NPWT can also function as a negative drainage tool through the space using a silastic drain.\nA second advantage is that the minimal motion of the joint protects against severe contracture of the hand.\nThese advantages explain the superior results of NPWT compared to a conventional dressing with a splint.\nOne of the limitations of using NPWT is that the pressure might compress the microvessels in the soft tissue, compromise the vascularity of the tissue, and decrease tissue perfusion.\nApplying NPWT could be of concern to surgeons, particularly in the hand, where blood circulation is limited to certain vessels and thin, pliable soft tissue with a weaker cushion effect.\nSome surgeons hesitate to use NPWT on the hands because of fear of restricted movement, difficulty of the application, and leakage due to the complex shape of the hand.\nTherefore, some modifications of commercially available NPWT have been reported, using gauze instead of a foam sponge or a sealing bag instead of semiocclusive dressing coverage.\nThese modifications are suitable adjustments for hand injuries; however, they were reported as certain indicated cases and required surgical adjustment on a case-by-case basis.\nThe pressure applied through the NPWT foam evenly distributes the mechanical force to the wound.\nMorykwas et al. tested various suction pressures from 0 to 400\u2009mmHg and found that 125\u2009mmHg was optimal for increasing local blood flow.\nCurrent recommendations state that 50\u2013150\u2009mmHg of negative pressure is acceptable.\nIn our cases, the hand was placed in the most functional position, and the drain was connected to suction power and set to 75\u2009mmHg.\nAlthough 125\u2009mmHg is the standard pressure for NPWT, similar effects can be achieved at lower pressures.\nIn addition, some reports have demonstrated that tissue pressure increases beneath the NPWT in all types of wounds, is directly proportional to the amount of suction applied, and is most pronounced in circumferential dressing.\nPrevious authors have reported that increased pressure results in 17% decreased perfusion when circumferential NPWT is applied with a suction pressure of 125\u2009mmHg.\nThe theories regarding the mechanism of action of NPWT suggest that compression of tissue decreases perfusion and concurrent hypoxia is a stimulus for angiogenesis.\nIn addition, tissue hypoxia results in release of nitric oxide and local vasodilation.\nOn the other hand, concerns regarding the safety of NPWT on tissues with compromised perfusion have also been raised.\nWe are aware that there is a potential risk for perfusion from the compression effects of a circumferentially applied NPWT dressing on the hand.\nWe did not find any evidence of reduced vascularity or compromised tissue perfusion as a result of using NPWT for hand injuries.\nBy contrast, we noticed a significant reduction in edema.\nThe compression provided by NPWT likely forces edema away from injured tissues.\nThis ultimately results in decreased interstitial pressure, decreased compression of the vessels, and improved oxygen and nutrient supply.\nThese results are likely to be the most important contributions of NPWT.\nThe HISS is the most commonly used measure to clinically assess hand injury severity.\nIt is evaluated by scoring the severity of each hand segment from skin, bones, motor function, and nerve injury.\nThe total score is determined by adding the point values of hand injury severity, then classifying it according to the score obtained, expressed as grades I\u2013IV.\nIn this study, we excluded patients with concomitant peripheral nerve injury because in this study, we compared the functional outcomes of patients with acute hand injury with or without NPWT after surgery, and nerve injury can interfere with the results of the functional outcome.\nIn addition, we only included patients with hand injuries and HISS scores of 21\u201350, which is grade II, and those who underwent reconstruction within 2 weeks after the injury.\nPatients with HISS scores > 50 and who were severely injured are often difficult to manage and require several staged treatments that could not be performed within 2 weeks; therefore, they were excluded.\nA quantitative assessment of hand dysfunction is much more difficult.\nAmong the most commonly used scales is the DASH scale, a 30-item questionnaire that evaluates symptoms and physical function with a five-response option for each item.\nThe DASH score is determined by calculating the circled responses.\nIt produces a brief, self-administered measure of symptoms and functional status.\nThe only limitation is that the DASH is a subjective measurement that represents hand function but does not fully correlate with objective functional recovery.\nTherefore, we first evaluated the DASH scale to represent personal symptoms and subjective situations.\nThen, we also evaluated objective functional recovery by determining the period when recovery of ROM was >90%.\nA 90% recovery of ROM is almost full recovery of function, which enables daily activity, the end of rehabilitation, and the return to social life.\nIn Group 2, the DASH scores were lower and the number of ROM recovery days was fewer compared to those in Group 1; these differences were statistically significant.\nAlthough the DASH score at postoperative 1 year was not different, the results suggest that NPWT was essential for early and fast recovery of hand function.\nIn the cases we described above, NPWT was successfully used to treat challenging hand injuries.\nComplications such as tissue loss, dehiscence, infection, or hematoma can have serious effects on the functional outcome.\nThe use of NPWT on very thin skin flaps or over skin grafts, where there is a concern for hematoma, perfusion, and skin survival, was particularly useful.\nIn addition, commercially available NPWT is increasingly evolving.\nThe foam sponge has become thinner, more flexible, and customized to the defect; the connecting tube is slender and length-adjustable; and the vacuum pump system has become smaller and more easily portable.\nAs NPWT maintains the injured hand in a stable state and can be changed every 3 days, most patients can be discharged and followed in the outpatient clinic, which is more convenient for patients and reduces hospital stay and costs.\nIn our experience of treating acute hand injures, NPWT is quick and easily applied.\nNPWT promotes wound healing by reducing edema, stabilizing the wound, and providing immobilization in a functional position.\nEarly wound healing and decreased complications enabled early rehabilitation, which lead to a successful functional recovery, both objectively and subjectively.\nA 59-year-old male suffered multitissue injury of the right hand, including all of the dorsal skin and extensors.\nEmergency operations including tendon repair and wound closure were performed.\nThe wound was covered with negative pressure wound therapy (NPWT) immediately after surgery.\nFull range of motion was achieved 4 weeks after surgery.\nA 54-year-old male with multitissue injury of the right hand visited the emergency room. Flexors of the second and third fingers were ruptured, with an avulsion bone fracture and multiple skin defects.\nThe patient was treated with negative pressure wound therapy (NPWT).\nFull range of motion was achieved 2 months after surgery, without restriction of daily motion.\n\nHand Injury Severity Scoring System (HISS).\n\u2009 | Score | Sum | Severity\nIntegumental injuries | 0\u201340 | <20 | I: minute\nBone injuries | 0\u20139 | 21\u201350 | II: medium\nImpairment of motor function | 0\u201316 | 51\u2013100 | III: severe\nNerve injury | 0\u201334 | >100 | IV: major\n\n\nComparison between two groups.\n\u2009 | Group 1(conventional dressing) | Group 2(NPWT) | p value\nAge (years) | 41.38 \u00b1 10.92 | 38.77 \u00b1 1.68 | 0.375\n\nHISS score | 33.57 \u00b1 1.86 | 35.73 \u00b1 1.55 | 0.377\n\nTime to recover over 90% of the ROM(days) | 46.90 \u00b1 2.05 | 33.30 \u00b1 1.51 | 0.022\n\nDASH score at one month | 33.14 \u00b1 1.68 | 22.67 \u00b1 1.43 | 0.031\n\nDASH score at one year | 22.08 \u00b1 2.03 | 20.99 \u00b1 1.91 | 0.667\n\nComplications | 3 | 2 | 0.383\nHematoma | 2 | 0\nInfection | 1 | 0\nWound disruption | 0 | 2\n\nNPWT: negative pressure wound therapy; HISS: Hand Injury Severity Scoring System; ROM: range of motion; DASH: Disability of the Arm, Shoulder, and Hand questionnaire.",
        "label": "high",
        "id": "task4_RLD_test_722"
    },
    {
        "paper_doi": "10.1111/j.1365-3156.2004.01198.x",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Cluster-RCTUnit of randomizations: householdIntra-cluster correlation coefficient factor of 0.04.Trial duration: 7 months between August 1999 and February 2000.\n\n\nParticipants: Adults and children living in malaria-endemic regionsParticipants were not screened at start for P. vivax.\n\n\nInterventions: Topical repellent - Mosbar soap (20% DEET + 0.5% permethrin) versus placebo lotionCo-interventions: noneTreatment arms:- Mosbar soap (20% DEET + 0.5% permethrin) arm: 67 households (618 participants)- Placebo arm: 60 households (530 participants)\n\n\nOutcomes: - Participants with clinical malaria confirmed through blood smears or rapid diagnostic tests (P. falciparum or P. vivax); and- Recorded adverse events.\n\n\nNotes: Trial was conducted with Afghan refugees in malaria-endemic region of Pakistan.Funded by HealthNet International's Malaria and Leishmaniasis control and research programme\n\n",
        "objective": "To assess the impact of topical repellents, insecticide\u2010treated clothing, and spatial repellents on malaria transmission.",
        "full_paper": "DEET mosquito repellent provides personal protection against malaria: a household randomized trial in an Afghan refugee camp in Pakistan\nDEET mosquito repellent provides personal protection against malaria: a household randomized trial in an Afghan refugee camp in Pakistan\nTropical Medicine & International Health\nTropical Med Int Health\nSynthetic repellents based on di-ethyl 3-methyl benzamide (DEET) are a popular method of obtaining protection from mosquitoes and yet clear evidence for a protective effect against malaria has hitherto never been convincingly demonstrated.\nA household randomized trial was undertaken among a study population of 127 families (25%) in an Afghan refugee village in Pakistan to compare the efficacy of repellent soap (Mosbar TM containing 20% DEET and 0.5% permethrin) vs. a placebo lotion.\nCases of falciparum and vivax malaria were detected by passive case detection at the camp's clinic.\nAt the end of the 6 month trial 3.7% (23 of 618) of individuals in the Mosbar group had presented with one or more episodes of falciparum malaria compared with 8.9% (47 of 530) of the placebo group (odds ratio 0.44, 95% CI 0.25-0.76).\n16.7% of the Mosbar group (103 of 618) presented with vivax malaria compared with 11.7% (62 of 530) of the placebo group, and thus no effect was shown against vivax malaria (odds ratio 1.29, 95% CI 0.86-1.94).\nA considerable proportion of individuals (22%) had presented with vivax malaria during the 7 months leading up to the trial and thus any intervention effect would be partially masked by relapsed infections.\nThe distribution of mosquitoes among households was broadly similar between Mosbar and placebo groups.\nThe repellent was popularly received and very few side-effects were reported.\nThere is a case for giving repellents more prominence in public health as a preventive measure in regions where vectors bite in the early evening or in emergency situations such as epidemics or newly established refugee camps.\nIntroduction\nDEET (di-ethyl 3-methyl benzamide) is an effective and well-known mosquito repellent, as testified by several decades of entomological research (McCabe et al. 1954;Schreck 1977;Curtis et al. 1990;Gupta & Rutledge 1994;Barnard 2000).\nDespite the intuitive appeal of synthetic repellents there have been surprisingly few attempts at demonstrating protection against malaria and no trial has been fully successful.\nAn early study in which DEET was distributed to one member of a pair of Indian villages produced no evidence for a reduction in transmission (Vittal & Limaye 1984).\nA study on schoolchildren in Tanzania did not show a significant effect (Curtis et al. 1994).\nA third study on Karen pregnant women in Thailand was thwarted by an unexpected decline in transmission caused by other factors (McGready et al. 2001).\nA community randomized trial in South America showed no effect on malaria incidence rates (Kroeger et al. 1997).\nA recent outbreak of malaria in a village in South Africa did subside after distribution of DEET repellent among the affected population (Durrheim & Govere 2002) but it is not clear whether this was due to other undetermined factors as there was no control group.\nWhereas an unequivocal epidemiological demonstration in favour of repellents is still lacking, there can be no doubt that insecticide-treated nets (ITN) are in many regions of the tropics a highly effective method of malaria prevention (Lengeler & Snow 1996;Lengeler 1998).\nWe share the conviction that ITN constitute the most feasible method of obtaining protection against malaria and one that needs to be made much more widely available.\nBy comparison, skin repellents may appear a false economy or a distraction from the main public health challenges of improving ITN Tropical Medicine and International Health volume 9 no 3 pp 335-342 march 2004 \u00aa 2004 Blackwell Publishing Ltd usage and establishing early diagnosis and treatment.\nBut in certain situations repellents may have a significant role as a supplement to ITN: for example, in regions where the vectors bite in the early evening (Lindsay et al. 1998;Pates et al. 2002) or as a response to temporary upsurges of malaria because of local outbreaks, political emergencies, or refugee displacements (McGready et al. 2001;Rowland & Nosten 2001).\nIn Afghanistan and Pakistan there is a potential market for a good mosquito repellent.\nMalaria vectors such as Anopheles stephensi, A. culicifacies, A. nigerrimus and A. pulcherrimus bite in the evening as well as at night (Reisen & Aslamkhan 1978;Rowland et al. 2002a).\nRepellents are not commonly used in Afghanistan at present, but neither were ITN before the social marketing drives of the last decade (Rowland et al. 2002b) and repellents may give added protection (Pates et al. 2002).\nIt is not clear whether, through appropriate health education and product promotion, the practice of using repellents might be successfully instilled.\nThe aim of the study was to investigate the efficacy and acceptability of a commercially available mosquito repellent, Mosbar TM repellent 'soap' (Yap 1986;Frances 1987) as a technique for malaria prevention in refugee settings.\nThis paper describes a cluster randomized trial in an Afghan refugee camp in Pakistan in which households were randomly assigned to receive the repellent or a placebo, and the impact of the intervention measured through monitoring of clinical episodes of Plasmodium falciparum and P. vivax malaria over a period of 6 months.\nMaterials and methods\nStudy area and malaria transmission\nThe trial was carried out in Adizai refugee settlement (population 3945) near Peshawar, in North-West Frontier Province, Pakistan.\nThe settlement was established in the early 1980s on waterlogged land on the banks of the Kabul river and is endemic for malaria.\nThe houses are made from mud and each building is surrounded by a courtyard and perimeter wall in a style typical of rural Pakistan and Afghanistan.\nA health centre run by the non-governmental organization HealthNet International provides free microscopical diagnosis and treatment of malaria to the 510 refugee families living in the camp.\nTransmission of malaria occurs from late June to November.\nThe vectors include A. culicifacies, A. stephensi, A. nigerrimus, and A. pulcherrimus, among others.\nMosquito biting starts after dusk, peaks around 9-11 p.m. then declines gradually through the night (Rowland et al. 2004).\nAccording to health centre records, vivax malaria increases from March to June, reaches a peak in July-August, and then declines.\nSignificant numbers of falciparum malaria cases are first observed in August and reach a peak in October-November.\nTransmission is unstable, particularly that of falciparum malaria.\nAccording to passive case detection records, the ratio of vivax to falciparum infections in 1999 was 5:1 and incidence was 0.80 malaria episodes per person year.\nRepellent\nMosbar is a commercially available repellent formulation (Yap 1986;Frances 1987).\nThe active ingredients are DEET (20%) and permethrin (0.5%), a pyrethroid insecticide with some repellent action.\nUnlike other repellents Mosbar is a 'portable' solid formulation considered to be more appropriate for use in developing countries (Rozendaal 1997).\nThe bar is used by making a lather in water which is applied to wetted skin.\nThe foam dries to an invisible film, much like standard liquid formulations of DEET, and should not be rinsed off after drying.\nThus, the term 'repellent soap' is somewhat misleading.\nThe repellent was obtained from a Zimbabwean manufacturer at a cost of US $0.40 per bar.\nPreliminary tests showed that a single application to face and exposed skin would give several hours' protection -confirming the findings of earlier studies (Frances 1987;Kroeger et al. 1997) -and that one bar would last an individual several weeks of nightly use.\nStudy design and procedure\nThe study took place over 7 months between August 1999 and February 2000.\nEach house compound in the camp had been assigned a unique number for census purposes.\nSample size calculations were based on P. falciparum incidence of 12% in the control group as indicated by the previous year's incidence rate, the assumption of an intracluster correlation of 0.04, and a treatment effect of 50% as a result of the intervention.\nTo receive 80% power to detect a difference in P. falciparum rates of 6%, assuming a drop out rate of 10%, we would require 1100 subjects to be enrolled into this study.\nBy applying simple randomization 13% (67 of 510) of households were allocated to the repellent soap group and a similar proportion (12%, 60 of 510) to the placebo control.\nIt was assumed that an intervention group of this magnitude would not unduly affect transmission rates in the remainder of the population and hence the study would give an accurate measure of personal protection.\nThe rationale behind using households rather than individuals was twofold: it was logistically easier and more acceptable to administer the intervention at the household level, and the risk of different treatments being shared within households, which would have seriously interfered with a trial based on individual randomization, was averted.\nThe placebo formulation was a moisturizing lotion imported from UK, unknown to Afghan refugees and shown in pre-tests to have no repellent effect on mosquitoes.\nBoth the lotion and the repellent bars were issued without packaging or identifying labels.\nHeads of families were invited to group meetings and informed about the aims of the study, namely to determine whether repellents protected against malaria and which of two products was better than the other.\nInformed consent was obtained from each head of family on behalf of other family members.\nNames and details of each family were reconciled with the clinic's family records.\nWomen and family members received health education in the home and were shown how to apply the repellent or placebo lotion according to which group they had been allocated.\nWe advised parents not to use Mosbar or the lotion on children <5 years of age to avoid the risk of discomfort caused by involuntary rubbing into eyes.\nThis accords with the advice of the US Environmental Projection Agency which concluded after reviewing 20 studies on the toxicology of DEET (USEPA 1998) that the repellent does not cause unreasonable risks but might be restricted to protect young children with skin sensitivity (Barnard 2000).\nFamilies of the placebo group were given the same health information as families of the Mosbar group.\nEach family was given a month's supply of product, and was visited fortnightly for topping up and monitoring of usage and side-effects.\nThe village health workers implementing the trial were blinded as to which product was the true repellent as both products were described as such.\nNo other malaria preventive intervention (e.g. nets, chemoprophylaxis) was being implemented in this camp at this time.\nFor baseline comparisons, clinic records of malaria infections were compiled for the period January-July 1999 leading up to the start of the trial.\nThis provided a suitable database for showing the history of previous infections.\nCases of vivax and falciparum malaria were diagnosed by microscopy using a process of passive case detection at the clinic.\nAt the start of the trial study families were encouraged to use the clinic in the event of febrile illness but did not receive any preferential treatment over other inhabitants.\nClinic workers responsible for treatment (medical officers) and diagnosis (microscopists) were blind to the allocation of families to the two study groups (they too assumed both products were effective) and were not informed about which families were in the trial.\nConfirmed cases of falciparum and vivax malaria were treated with 25 mg chloroquine per kg body weight and falciparum treatment failures were treated with sulphadoxinepyrimethamine.\nPrimaquine was not used in treatment.\nRecording of malaria in trial families started on 9 August and ended on 31 January.\nA non-registered private microscopist was present in the camp, but the clinic's family registration system indicated that over 90% of families attended the HealthNet clinic.\nAfghans sometimes self-treat with chloroquine, but this was not monitored during the present study.\nEntomological surveillance was conducted in 30 sentinel households randomly selected from each group, once in July before the repellent/placebo products were issued and again in October.\nA mosquito catch was made by space spraying a living room and animal shelter of each house with a commercial aerosol pyrethroid after closing all exits and covering the floor with sheets.\nThe night before the space spraying was done a CDC light trap collection was made outdoors close to where people were sleeping (the majority of people sleep outdoors from May to October).\nCollections were identified to species and scored.\nTowards the end of the study, 20 heads of family from each group and their wives were interviewed about their own and their family's experiences.\nThe structured questionnaire included questions about usage, perceived effectiveness, acceptability, convenience and side-effects.\nEthical clearance was obtained from Pakistan Medical Research Council and the LSHTM Ethics Committee.\nAnalysis\nThe trial was primarily designed to determine whether the repellent could be considered an effective intervention against falciparum and vivax infections, with efficacy measured in terms of whether or not an individual became infected.\nAll malaria episodes used in the analysis occurred 3 weeks after the start of the trial (the assumed incubation period).\nIt was not possible to distinguish new infections of vivax malaria from relapsed infections.\nAll members of every household that entered the trial were included in the analysis.\nThis intention-to-treat approach was an attempt to negate the opportunity for bias to enter the trial.\nFor the analysis, we used the method of Generalized Estimating Equations (GEEs), using the robust variance estimator and an exchangeable working correlation matrix (Liang & Zeger 1986).\nThis approach, an extension of standard multiple logistic regression, adjusted for the effect of clustering at the household level.\nBias due to relapse was minimized by including a binary covariate in the model that indicated whether an individual had contracted malaria during the period leading up to the trial.\nSubgroup analyses have been presented, with inferences based on statistical tests of interaction.\nAll efficacy analyses used an intent-to-treat approach.\nNo losses to follow-up were reported throughout the duration of the trial, with the reason for this being that Adizai is a relatively stable refugee camp.\nEntomological outcomes were analysed using Poisson regression analysis.\nResults\nA total of 127 households and 1148 individuals were enrolled into the study.\nThis represented 29% of the camp population.\nA summary of study characteristics upon admission is presented in Table 1.\nThe age distribution and the number of individuals with previous malaria infections during the 7 months leading up to the trial were comparable across the treatment groups.\nThe same conclusion is reached when data is presented at the household level.\nThe intra-cluster correlation coefficient for P. falciparum is 0.038.\nAs expected, only a small number of P. falciparum infections occurred during the pre-intervention period of January-July because this was the low season for malaria transmission.\nTable 2 presents a simple summary at the individual level, unadjusted for household clustering, by species of malaria.\nIrrespective of treatment or malaria species, the percentage infected initially increased with age, reaching a peak in children aged between 5 and 10 years, and then decreased with age.\nFor P. falciparum, the percentage of infections was lower in the Mosbar group than in placebo group.\nA clear difference was shown between the two interventions for P. falciparum malaria across individuals with 1, 2 or 3+ infections, with the percentage of infections in these categories more than double for individuals on placebo than for those on Mosbar.\nFor P. vivax the percentage of infections were slightly higher in the Mosbar group.\nThe percentage of P. falciparum infections peaked in October for the placebo group (19 of 511, 3.7%) and in November for the Mosbar group (12 of 611, 2.0%).\nThe later peak in the Mosbar group may have been due to some individuals ceasing to use repellent in late-October, despite continuing vulnerability to infection, when mosquito numbers began to decline and people started sleeping indoors.\nThe pattern of P. vivax infections was similar for both interventions, with a gradual reduction in the number of infections from September 1999 to January 2000,  followed by an increase in February as a result of relapsed infections.\nFindings with the household as the unit of analysis were consistent with that shown at the individual level.\nThere were more households with at least one episode of P. falciparum malaria in the placebo group (27 of 60, 45%) than in the Mosbar group (18 of 67, 27%).\nFor P. vivax, 66% of households in the Mosbar group (44 of 67) had at least one infection compared with 52% in the placebo group (31 of 60).\nBaseline covariates that may predict subsequent malarial infection were included in the model to increase the validity of estimates of treatment effects.\nResults are presented in Table 3.\nWhen expressed in terms of protective efficacy [100(1 ) odds ratio)%], the odds of being infected with P. falciparum malaria were reduced by 56% (95% CI 24-75%, P \u00bc 0.004) in the Mosbar group.\nSimilar conclusions are reached when the effect of Mosbar was modelled unadjusted for the baseline covariates (protective efficacy \u00bc 53%, 95% CI 20-73%).\nAs expected, previous P. falciparum infections were a strong indicator for later P. falciparum infections.\nThere was evidence of a treatment-age interaction with the effect of Mosbar use being significant only among individuals aged between 5 and 20 years (test of interaction P \u00bc 0.04).\nThere was no evidence that the effect of Mosbar use varied by previous P. falciparum infection or gender (tests of interaction P \u00bc 0.63 and P \u00bc 0.26 respectively).\nThere was no evidence of an effect of Mosbar use on P. vivax infections (P \u00bc 0.226), although the estimated odds of being infected appeared to be higher in individuals on Mosbar than placebo.\nSimilar conclusions are reached when the effect of Mosbar soap was modelled unadjusted for the baseline covariates (P \u00bc 0.193).\nIn depth interviews with 20 Mosbar group families confirmed that 19 families used the repellent regularly and considered it effective.\nSixteen families considered Mosbar easy to apply, but 10 considered oil-based formulations easier to use.\nNineteen considered the smell to be acceptable.\nOne person complained of skin irritation and stopped using the repellent.\nInterviews with 20 placebo group families confirmed that 12 families considered the lotion ineffective as a repellent, citing mosquito biting and disrupted sleep as the two main problems.\nBut all continued to use it and all considered it easy to apply and the aroma pleasant.\nA total of 2225 mosquitoes were captured in space spray collections.\nAnopheles culicifacies comprised 47%, A. stephensi 25%, Culex spp. 24%, and A. fluviatilus, A. annularis, A. maculatus and A. subpictus made up the remaining 4%.\nCulicines were more abundant in the July pre-intervention survey (average 13.4 per house) than in October survey (average 7.7 per house), whereas Anophelines were much more abundant in October (52.4 per house) than pre-intervention (6.8 per house).\nCulicines were equally abundant in living quarters and animal shelters, whereas anophelines were seven times more abundant in animal shelters than living quarters.\nDuring the intervention significantly fewer mosquitoes were caught in the light traps close to where people of the Mosbar group were sleeping outdoors, although curiously the number of culicines were significantly higher in this group (Table 4).\nThere were significantly more culicine mosquitoes in the living quarters of Mosbar households than in placebo group households during October survey.\nHowever, this may have been a chance effect because people were mainly sleeping outdoors during this time.\nThere was no difference between Mosbar and placebo group households in the density of mosquitoes in animal shelters either before or during the intervention.\nDiscussion\nThis placebo controlled trial appears to be the first clear demonstration of a skin repellent providing protection against falciparum malaria, the potentially fatal species of malaria.\nThe protective effect was greatest among those at greatest risk of malaria: individuals between 5 and 20 years of age.\nThe absence of effect in children under five was most likely due to parents responding to our expressed concern about the possibility of skin sensitivity in young children.\nThis inference may be drawn from the results of an earlier intervention trial involving a different method of protection (permethrin treated top-sheets and blankets) in the same camp in which we placed no restriction on age but did observe a protective effect in children aged 0-5 years (Rowland et al. 1999).\nIn the present study no effect of Mosbar was observed in adults over 20 years old.\nAgain this result is consistent with the earlier treated blanket study and may reflect reluctance or conservatism among adults to adopt new preventive measures for personal use, compounded by a greater immunity to new infections.\nBut older children clearly benefited from use of repellents and are an appropriate group to target, perhaps via concerned parents.\nOur purpose was to measure absolute efficacy, as protection against malaria had been never before demon-strated using a skin repellent, and to extend user choice in personal protection.\nThis necessitated comparison with a placebo.\nThe trial was designed to be double-blind initially, but we realized that blinding would gradually be lost as users of placebo lotion discovered the formulation was ineffective.\nCompliance to the allocated treatment might at that point become lower in the placebo lotion group and lead to a dilution of the 'placebo effect' which could, in turn, inflate the true Mosbar effect.\nThis was a limitation of the study.\nIn reality, compliance and demand for the lotion was unexpectedly high throughout the study, due partly to the pleasing effect of the lotion on the skin and partly to some continuing faith in its effectiveness.\nThe question remains whether any dilution of the placebo effect could have impacted upon the trial.\nBecause the parasitological outcome was assessed although passive case detection and study participants had to take the initiative to go to the health centre, members of the lotion group might have felt more inclined to go with minor febrile ailments if they believed these to be malaria whereas members of the repellent group might not have felt so inclined if they thought they were protected.\nIf malaria attacks were self-limiting this might indeed lead to a proportion of malaria cases in the repellent group going undetected.\nHowever, for this line of argument to hold we would expect to see a more notable treatment effect in the more-immune older groups as only among these would malaria be self-limiting (in this region, where malaria is hypoendemic or mesoendemic, development of immunity is only partial and is more observed in adults -see Table 2 for evidence of this).\nFurthermore, if this line of argument held up we would expect to see a greater treatment effect against the more benign and self-limiting vivax malaria than was observed.\nThe more serious falciparum malaria is rarely self-limiting in this region and usually grows worse unless treated, particularly in children.\nChildren with this disease would be motivated to attend the clinic regardless of which group they were allocated, and it was among children (over 5 years) of course that the greatest treatment effect was observed.\nAnother possible source of bias might result from microscopists being less rigorous in slide reading if they knew the blood film came from a Mosbar user.\nHowever, because microscopists were blind to the allocation, this source of bias is considered unlikely.\nThe trial showed that Mosbar use was effective in reducing P. falciparum infections in an Afghan refugee population in which there was no use of ITNs.\nThe intervention and placebo groups appeared to be comparable at baseline, as indicated by similar mosquito densities in living rooms or animal shelters, and the similar rates of malaria infection in the lead up to the trial.\nThe absence of effect against P. vivax is unexplained, and was surprising considering the two groups showed similar rates of infection in the 7 month lead up.\nThe number of vivax cases in Adizai in 1999 reached a peak 2 months before the start of the trial, which was timed primarily to coincide with the season of falciparum transmission from September to December.\nA recent case-control study in nearby Afghanistan was deliberately started 2 months earlier on 1 July at the beginning of the vivax transmission season and this study did demonstrate a negative association between Mosbar use and vivax infections after adjusting for ITN use and other covariates (Rowland et al. 2004).\nHad the cluster randomized trial started earlier an effect against vivax might have been observed.\nA high proportion of observed cases in the two groups were probably relapses of infections contracted before the trial, as indicated by the odds ratio of 5.6 for previous infections, and these would have masked any treatment effect.\nA significant reduction in anopheline density was observed in the vicinity of individuals in the Mosbar group sleeping outdoors.\nAn effect on indoor resting anopheline density (in living rooms and animal shelters) was not observed nor was one expected because of the outdoor sleeping habit.\nThe significantly higher density of culicines in living rooms and light traps of the Mosbar group during the trial remains a puzzle.\nThe sample size would need to be increased and the entomological surveys would need to be repeated at regular intervals during the trial before drawing any firm conclusion.\nThe overall benefit to be gained from using Mosbar should be weighed against the potential risks.\nSafety concerns exist as to the long-term use of DEET particularly on young children (USEPA 1998).\nDEET repellent is cost-effective over the short-term, but ITN are more practical and cost-effective when the need for protection extends over more than one season.\nMosbar remained popular over a single season but some users said they preferred oil-based formulations and adherence long-term is unknown.\nFor reasons of safety and cost we advocate that DEET-based repellents be restricted as a short-term intervention in situations such as epidemics or temporary crises such as newly established refugee camps, where ITN may be too costly, inappropriate for existing living conditions or unobtainable in sufficient numbers fast enough.\nThe suitability of repellents in emergencies or epidemics should be weighed against other potential short-term interventions such as insecticide-treated blankets or tarpaulins (Rowland et al. 1999;Graham et al. 2002).\nRepellents would be more useful where vectors bite in early evening (i.e. Asia more than Africa) or where it is too hot to use insecticide-treated blankets.\nRepellents are less bulky than any of these alternatives and stockpiles could be air-freighted and distributed quickly.\nPotential drawbacks with repellents are the need for self-discipline and some health education initially, unlike the other interventions which are needed anyway for purposes of warmth or shelter and whose use is guaranteed.\nAn argument sometimes raised against the use of repellents is the possibility that infective mosquitoes might simply be diverted to feed on unprotected individuals, particularly if the vector species are strongly anthropophilic.\nThis possibility still needs to be tested.\nBut in regions where vector species are seasonal, bite all night, and are partially zoophilic, such as South Asia, we would expect combined use of ITN and repellents (in adults and children over 3 years) to provide all-night protection with little risk of diversion to unprotected neighbours.\nRepellents need to be taken more seriously as a public heath measure.\nTable 1 Study characteristics at baseline\n | Mosbar | Placebo lotion\nNumber of households | 67 | 60\nNumber of individuals | 618 | 530\nMedian household size (IQR) | 9 (5) | 9 (3)\nMedian age in years (IQR) | 12 (21) | 13 (23)\nGender (female %) | 48.1 | 50.6\nPreviously infected with malaria* |  | \nIndividual level |  | \nFor Plasmodium falciparum |  | \nYes (%) | 11 (1.8) | 7 (1.3)\nNo (%) | 607 (98.2) | 523 (98.7)\nFor Plasmodium vivax |  | \nYes (%) | 137 (22.2) | 122 (23.0)\nNo (%) | 481 (77.8) | 408 (77.0)\nHousehold level |  | \nFor Plasmodium falciparum |  | \nYes (%) | 9 (13.4) | 7 (11.7)\nNo (%) | 58 (86.6) | 53 (88.3)\nFor Plasmodium vivax |  | \nYes (%) | 54 (80.6) | 45 (75.0)\nNo (%) | 13 (19.4) | 15 (25.0)\nIQR, inter-quartile range. |  | \n* This includes any infections up to 7 months prior to the trial starting in August 1999.\nOne or more individual pre-intervention infections within a household are classified as 'yes'.\n* This includes any infections up to 7 months prior to the trial starting in August 1999.\nTable 2 Summaries of malaria infection, including count of infections, categorized by age and species of malaria, using the individual as the unit of analysis\nIndividual level | Mosbar (%) Placebo lotion (%)\nFor Plasmodium falciparum |  | \nTotal number infected (%) | 23 (3.7) | 47 (8.9)\nAge category (year) |  | \n<5 | 6 (5.5) | 5 (5.3)\n5-10 | 8 (6.2) | 19 (17.3)\n10-20 | 2 (1.3) | 11 (9.7)\n>20 | 7 (3.1) | 12 (5.7)\nNumber of individuals with |  | \nNo infections | 595 (96.3) | 483 (91.1)\n1 infection | 15 (2.4) | 29 (5.5)\n2 infections | 7 (1.1) | 12 (2.3)\n3+ infections | 1 (0.2) | 6 (1.1)\nFor Plasmodium vivax |  | \nTotal number infected (%) 103 (16.7) | 62 (11.7)\nAge category (year)* |  | \n<5 | 25 (22.9) | 15 (16.0)\n5-10 | 37 (28.5) | 25 (22.7)\n10-20 | 20 (13.0) | 11 (9.7)\n>20 | 20 (9.0) | 11 (5.2)\nNumber of individuals with |  | \nNo infections | 515 (83.3) | 468 (88.3)\n1 infection | 85 (13.8) | 48 (9.0)\n2 infections | 13 (2.1) | 11 (2.1)\n3+ infections | 5 (0.8) | 3 (0.6)\n* There was one more case of malaria (Plasmodium vivax on | \nMosbar) but the age of the infected individual was missing, and so | \ndoes not appear here in this cross-classification. | \nTable 3 Logistic regression analysis show-\nTropical Medicine and International Health |  |  | volume 9 no 3 pp 335-342 march 2004\nM. Rowland et al. DEET mosquito repellent |  |  |  | \ning associations between treatment effect and malaria infection, adjusted by baseline covariates (previous infection, gender | Variables included in model | Plasmodium falciparum Odds ratio (95% CI) | P-value | Plasmodium vivax Odds ratio (95% CI) | P-value\nand age) | Treatment |  |  |  | \n | Placebo lotion | 1 |  | 1 | \n | Mosbar soap | 0.44 (0.25, 0.76) | 0.004 | 1.29 (0.86, 1.94) | 0.226\n | Previous infection |  |  |  | \n | No | 1 |  | 1 | \n | Yes | 9.86 (3.74, 26.02) | <0.001 | 5.64 (3.82, 8.34) | <0.001\n | Gender |  |  |  | \n | Female | 1 |  | 1 | \n | Male | 0.80 (0.52, 1.23) | 0.303 | 1.14 (0.83, 1.59) | 0.417\n | Age (years) |  |  |  | \n | <5 | 1 |  | 1 | \n | 5-10 | 2.11 (0.93, 4.76) |  | 1.26 (0.82, 1.92) | \n | 10-20 | 0.89 (0.36, 2.18) |  | 0.62 (0.40, 0.94) | \n | >20 | 0.82 (0.37, 1.83) | 0.004 | 0.42 (0.27, 0.67) | <0.001\n | The method of Generalized Estimaing Equations (GEEs) used in this logistic regression\n | analysis, adjusts for the effect of clustering at the household level. | \n\u00aa 2004 Blackwell Publishing Ltd\nTable 4 Mosquito density per room as determined by space spray and light trap catches in houses before (July) and during (October) the intervention\nTropical Medicine and International Health |  |  | volume 9 no 3 pp 335-342 march 2004\n | Anopheles stephensi | Anopheles culicifacies | Culex\n | Placebo | Mosbar | Placebo | Mosbar | Placebo Mosbar\nSurvey | AM GM AM GM Density ratio | AM GM AM GM Density ratio | AM GM AM GM Density ratio\nLiving room |  |  | \nBefore 0.3 0.2 0.2 0.2 0.67 (0.11, 3.99) 1.2 0.8 0.1 0.1 0.09* (0.01, 0.71) 5.5 3.6 5.8 2.6 1.05 (0.72, 1.52) | \nDuring 1.5 0.7 3.3 1.2 2.20* (1.19, 4.05) 2.3 0.8 3.8 1.7 1.65 (0.98, 2.77) 1.4 0.9 7.6 4.0 5.4* (3.07, 9.60) | \nAnimal shelter |  |  | \nBefore 1.2 0.7 0.9 0.6 0.72 (0.35, 1.47) 3.7 1.6 4.3 1.9 1.18 (0.83, 1.67) 8.3 4.8 7.3 3.9 0.87 (0.67, 1.12) | \nDuring 14.9 4.6 16.5 4.1 1.11 (0.93, 1.33) 30.1 7.0 29.3 5.3 0.97 (0.85, 1.11) 3.8 2.8 2.5 1.9 0.67 (0.44, 1.01) | \nLight trap |  |  |  | \nBefore 0.04 0.01 0.08 0.02 2.00 (0.18, 22.0) 0.04 0.01 0.16 0.04 4.00 (0.45, 35.8) 1.80 0.37 2.44 0.43 1.36 (0.92, 1.99) | \nDuring 0.68 0.16 0.28 0.08 0.41* (0.17, 0.99) 0.32 0.09 0.24 0.06 0.75 (0.26, 2.16) 0.36 0.08 0.80 0.15 2.22* (1.01, 4.88) | \nAM, arithmetic mean density per room; GM, geometric mean density per room. | \nFifteen sentinel houses were monitored in each intervention group. Density ratios were calculated using Poisson regression analysis. | \n* P < 0.05. |  |  | \nAcknowledgements\nAuthors wish to thank the staff of Adizai clinic, microscopist Naeem Sherpao, community health workers Zahoor and Isra, and entomologists Mushtaq Ahmad and Mohammed Kamal for their support.\nHealthNet International's malaria control and research programme is supported by the European Commission (DG1), the United Nations High Commissioner for Refugees, and WHO/UNDP/World Bank Special Programme for Research and Training in Tropical Diseases.\nMR and JL are supported by the UK Department for International Development and the Gates Malaria Partnership, and CC by the Medical Research Council.\nHowever, none of these donors can accept responsibility for any information provided or views expressed.",
        "label": "unclear",
        "id": "task4_RLD_test_300"
    },
    {
        "paper_doi": "10.1186/1471-2458-11-475",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Study design: cluster randomised controlled trialStudy duration: recruitment occurred May to July 2006. The parent meeting was delivered eight times during July and August 2006. Follow-up was 3 monthsStudy arms: intervention - educational session plus leaflet; control - leaflet only\n\n\nParticipants: Setting:Country and income level: England (high; 3 of 33 wards, selected to represent high-, medium-, and low-deprived geographical areas)Degree of regional development: not describedParticipants:ParentsInclusion and exclusion criteria for participation in study: Clusters - primary healthcare centres with at least two medical practitioners selected by highest deprivation score. Childcare centres selected by largest size. Parents - English literate, with a child eligible for the first or second dose MMR vaccination (first dose was given at 13 months and the second dose between four and five and half years of age, so target age range was six months to five years.)Categorisation (mothers, fathers, parents, expectant parents, guardians): parentsNumber randomised to intervention: healthcare centres N = 3; childcare centres N = 3; parents N = 71 (N = 68 completed baseline survey)Number randomised to control: healthcare centres N = 3; childcare centres N = 3; parents N = 71 (N = 67 completed baseline survey)Total number randomised: healthcare centres N = 6; childcare centres N = 6; parents N = 142Number lost to follow-up, withdrew from intervention: N = 23 did not receive intervention; N = 13 lost to follow-up for final time pointNumber lost to follow-up, withdrew from control: N = 7 lost to follow-up for final time pointAge range: intervention 34.07 years +- 5.43; control 34.06 years +- 5.52Gender: intervention female N = 67; male N = 4. control female N = 67; male N = 4Ethnicity: intervention - White British 68 (95.8%); other 3 ( 4.2%). control - White British 68 (95.8%); other 3 ( 4.2%)Level of education or literacy: intervention - left school at 16 years: 24 (33.8%); left school at 18 years: 10 (14.1%); achieved degree or higher: 37 (52.1%). control - left school at 16 years 25: (35.2%); left school at 18 years 10: (14.1%); achieved degree or higher: 36 (50.7%)Socioeconomic status: not describedChildrenAge range and categorisation (infants, preschool-aged children, school aged children): preschool-aged childrenMean age +- SD of first (youngest) child eligible, months: intervention 25.73 +- 14.66. control 19.77 +- 11.69Gender: not stated\n\n\nInterventions: Intervention purpose: to provide parents with the opportunity to discuss MMR with other parents who are making an MMR decision; to provide information about MMR from a variety of perspectives; to introduce and practice one approach to supporting parents to ask questions about MMR of their healthcare practitionerDeliverer: co-facilitated by a researcher and a parent facilitator. Parent facilitators were all female and recruited from local communitiesFormat or delivery mode: parent meeting including three components: provision of balanced information, a group discussion, and a coaching exercise. Mean of 6 participants per meetingContent of communication: three parts. Group discussion provided opportunity to discuss any issues about MMR with other parents who were also making an MMR decision; Q&A session provided opportunity to ask questions of the immunisation nurse specialist; Coaching exercise introduced and practiced one approach to supporting parents to ask questions about MMR in the primary care consultation (role playing exercise using question prompt sheet)Vaccine or vaccines delivered or described: MMRDirection of communication: interactiveGroup or individual: groupWhere the intervention took place: non-healthcare venues (e.g. community centres) close to participating healthcare centres and childcare organisationsTraining required for intervention: parent facilitators received one half-day trainingTheoretical basis for intervention: in line with fundamental tenets of health promotion, i.e. based on an 'engagement' model of communication where a key goal is empowermentCost of intervention: not describedIntervention quality: content and delivery of parent meeting informed by interviews with parents, systematic review of parents' decision-support needs, and two focus groups with parentsFidelity and integrity: session was co-facilitated by a researcher at each meeting, which potentially kept the sessions consistent. 23/71 parents in intervention group did not attend an intervention meeting, so intervention was not delivered as intended to all those randomised to intervention groupDetails of control, usual, or routine care: MMR leaflet onlyDetails of co-interventions in all groups: all participants received 'MMR - your questions answered' leaflet. Authors noted this leaflet was meant to be equivalent to usual care, but turned out to be more comprehensive than usual care (parents reported they were not normally given leaflets that were meant to be mandatory).\n\n\nOutcomes: Primary outcomes measured:Vaccination statusDefinition of immunisation status used by authors: receipt of MMR ('Since the study started, have you taken your child to have the combined MMR vaccine?')Description of outcome assessment tool: postal questionnaireTiming of outcome assessment: three months post-interventionKnowledgeDefinition: knowledge about MMRDescription of outcome assessment tool: postal questionnaire, 11 questions (score 0 to 11)Timing of outcome assessment: one week and three months post-interventionAttitudes or beliefsDefinition: attitude towards MMR, beliefs about vaccine necessity, concerns about MMRDescription of outcome assessment tool: one question about attitude (rate 1 to 7); four items assessing necessity beliefs (score 4 to 20); four items assessing concern beliefs (score 4 to 20)Timing of outcome assessment: one week and three months post-interventionIntention to vaccinateDefinition: intention to vaccinate child with MMRDescription of outcome assessment tool: postal questionnaire, 3 items on a 7-point scale, averaged over the three itemsTiming of outcome assessment: one week and three months post-interventionAdverse effects (anxiety)Definition: anxiety from interventionDescription of outcome assessment: Short form State-Trait Anxiety Inventory (STAI) tool with 6 items scored 1 to 4. Total score multiplied by 20/6. A normal score is 34 to 36.Timing of outcome assessment: one week and three months post-interventionSecondary outcomes measured: none\n\n\nNotes: Sample size: \"To achieve 80% power to detect a standardised effect size of 0.67 on the primary outcome of decisional conflict, using a two-sided t-test with significance level of 0.05 and an estimated intracluster correlation coefficient (ICC) of 0.05 (giving a design effect of 1.5 based on an average of 11 parents per cluster) required a sample size of 108 parents (54 in each group). Predicting 25% attrition, 73 parents were required in each group. Parent numbers were not balanced across the clusters. Based on our previous research, we estimated recruiting 12 parents per week over three months.\"Contact with author: authors contacted to clarify sequence generation process and to identify any published protocol or trial record. No protocol available, and authors could not recall randomisation process but confirmed it was conducted by a statistician.Other: authors conducted qualitative research alongside, determining that the intervention was feasible to deliver in non-healthcare, community venues, and it was acceptable to parents, with the majority expressing positive view\n\n",
        "objective": "To assess the effects of face\u2010to\u2010face interventions for informing or educating parents about early childhood vaccination on vaccination status and parental knowledge, attitudes and intention to vaccinate.",
        "full_paper": "Background\nIn the UK public concern about the safety of the combined measles, mumps and rubella [MMR] vaccine continues to impact on MMR coverage.\nWhilst the sharp decline in uptake has begun to level out, first and second dose uptake rates remain short of that required for population immunity.\nFurthermore, international research consistently shows that some parents lack confidence in making a decision about MMR vaccination for their children.\nTogether, this work suggests that effective interventions are required to support parents to make informed decisions about MMR.\nThis trial assessed the impact of a parent-centred, multi-component intervention (balanced information, group discussion, coaching exercise) on informed parental decision-making for MMR.\nMethods\nThis was a two arm, cluster randomised trial.\nOne hundred and forty two UK parents of children eligible for MMR vaccination were recruited from six primary healthcare centres and six childcare organisations.\nThe intervention arm received an MMR information leaflet and participated in the intervention (parent meeting).\nThe control arm received the leaflet only.\nThe primary outcome was decisional conflict.\nSecondary outcomes were actual and intended MMR choice, knowledge, attitude, concern and necessity beliefs about MMR and anxiety.\nResults\nDecisional conflict decreased for both arms to a level where an 'effective' MMR decision could be made one-week (effect estimate = -0.54, p < 0.001) and three-months (effect estimate = -0.60, p < 0.001) post-intervention.\nThere was no significant difference between arms (effect estimate = 0.07, p = 0.215).\nHeightened decisional conflict was evident for parents making the MMR decision for their first child (effect estimate = -0.25, p = 0.003), who were concerned (effect estimate = 0.07, p < 0.001), had less positive attitudes (effect estimate = -0.20, p < 0.001) yet stronger intentions (effect estimate = 0.09, p = 0.006).\nSignificantly more parents in the intervention arm reported vaccinating their child (93% versus 73%, p = 0.04).\nConclusions\nWhilst both the leaflet and the parent meeting reduced parents' decisional conflict, the parent meeting appeared to enable parents to act upon their decision leading to vaccination uptake.\nBackground\nIn the UK public concern about the safety of the combined measles, mumps and rubella [MMR] vaccine as a consequence of Wakefield et al's now discredited research continues to impact on MMR coverage even though the sharp decline in uptake has begun to reverse.\nCurrent first and second dose MMR uptake rates in England are 84% and 77% respectively, short of the 95% required for population immunity.\nAs a consequence, there is a pool of unimmunised children susceptible to the diseases, reflected in persistent localised measles outbreaks with an epidemic predicted in the near future.\nIn some European countries and in the USA, childhood immunisation is mandatory yet MMR vaccine refusal has increased, similarly leading to measles outbreaks.\nRepeated assertions by the Department of Health for England and Wales and the U.S Centers for Disease Control and Prevention that the MMR vaccine is safe have had limited effect on allaying parents' concerns in some sections of the community.\nMore than ten years after the publication of Andrew Wakefield's now discredited findings, there is some evidence that parent trust in MMR has improved, yet significant numbers continue to lack confidence in making an MMR decision and many criticise what is perceived to be the poor quality of information provided.\nA recent systematic review identified the decision support needs of parents making child health decisions (including immunisation); these related to three themes: (i) a need for timely, consistent up-to-date evidence based information tailored to the individual child, delivered in a variety of formats from trustworthy sources; (ii) a need to talk with others facing the same decision to share experiences; and (iii) a need to be in control of their level of preferred involvement in the decision-making process.\nThis suggests that interventions, informed by parents' expressed needs of what they would find helpful to make informed decisions about MMR are required.\nHowever, in spite of a large literature describing the factors that influence parents' decisions whether to vaccinate their child with MMR, evaluations of decision support interventions in this context are limited.\nInformed by a systematic review and an interview study with parents we developed and evaluated an evidence-based, parent-centred, multi-component intervention to support informed parental decision-making for the MMR vaccine.\nThis parent-centred approach is consistent with Western health policy in which a clinician-patient partnership is emphasised.\nIt is also in line with the fundamental tenets of health promotion that is based on an 'engagement' model of communication where a key goal is empowerment.\nThe intervention was designed to supplement routine UK primary care service for childhood vaccination whereby parents are invited to take their child, free of charge, for all immunisations on the National Health Service Routine Childhood Immunisation Schedule.\nChildhood vaccination is not mandatory in the UK.\nIn this paper we report the findings of a cluster randomised controlled trial to evaluate a parent-centred, multi-component intervention to support informed decision-making for MMR.\nThe effectiveness data are presented here.\nAcceptability of the intervention is reported elsewhere.\nMethods\nSetting and Participants\nThe study was approved by the Local Research Ethics Committee (06/Q1107/25) on 18 May 2006.\nIt was located in three of 33 wards (electoral district) in Leeds, England (approx. 770,000 population).\nUsing the Index of Multiple Deprivation these wards were selected to represent low, medium and high deprived geographical areas [mean scores: low = 6.89; medium = 29.22; high = 55.07].\nPrimary healthcare centres employing at least two medical practitioners were purposively selected based on their low income scheme index [LISI, 39] scores.\nFor example, in the high deprived ward, we approached centres with the most deprived practice population first (i.e. highest LISI score).\nChildcare organisations in the same wards were approached on the basis of size, the largest first.\nEleven (of 15) healthcare centres and six (of eight) childcare organisations were invited to participate.\nSix primary care centres and six childcare organisations agreed.\nWithin these providers the target sample was parents who were English literate with a child eligible for the first or second dose MMR vaccination.\nAt the time of the study, in the UK, the first dose was given at 13 months and the second dose between four and five and half years of age.\nThe target age range was, therefore, six months to five years.\nLetters were sent to eligible parents on providers' registers.\nParents replied to the research team and were telephoned for screening and recruitment.\nRecruitment occurred May to July 2006.\nDesign and Intervention\nThe design was a cluster randomised controlled trial design with two arms: intervention and control.\nThis design was chosen to reduce the potential risk of contamination between arms.\nThe six healthcare centres and six childcare organisations were matched in pairs based on their ward (three pairs of healthcare centres, three pairs of childcare organisations).\nOne of each pair was randomly allocated to the intervention, the other to the control arm.\nA researcher not involved in the study and blind to the identity of clusters performed the randomisation using a sealed envelope procedure.\nThe study researcher (RP) was blind to arm assignment when screening and recruiting parents, and sending out the baseline questionnaire.\nStatisticians (WH, RW) saw blinded data.\nParents were blind to arm assignment at recruitment and in completing the baseline questionnaire.\nParents allocated to the intervention arm were invited to attend one two-hour parent meeting, co-facilitated by a researcher (CJ, FMC, RP) and a parent.\nThree parent facilitators (all women) were recruited from local communities.\nThey received one half-day training.\nIn advance of the meeting parents were sent an information leaflet ('MMR your questions answered',).\nThe content and delivery of the parent meeting (see Table 1) was informed by interviews with 69 parents and a systematic review of parents' decision support needs.\nThis was refined in two focus groups with local parents.\nThe meeting included three components: provision of balanced information, a group discussion and a coaching exercise.\nParents in the control arm were sent the same MMR leaflet.\nMeasures\nParent characteristics (e.g. age, ethnicity) were collected by telephone at recruitment.\nPrimary and secondary outcomes were collected by postal questionnaire prior to randomisation (T1), one week post-intervention (T2) and three months post-intervention (T3).\nThe questionnaire was developed in collaboration with an expert in the field of health decision-making (HB) and piloted with five parents, though no changes were made.\nImpact of the parent meeting and MMR leaflet\nThe primary outcome measure was decisional conflict as measured by the Decisional Conflict Scale.\nThis generic measure is a 16-item scale to assess people's perceptions about the quality of their decision-making process; it has five sub-sections for being informed, clear about their values, degree of support, uncertainty with the choice, effectiveness of their decision.\nIt has demonstrated test-retest reliability, construct and predictive validity in the patient health decision-making context.\nScores range from 1 (no decisional conflict) to 5 (extremely high decisional conflict).\nScores lower than two are associated with 'implementing decisions', higher scores are interpreted as 'decision delay or feeling unsure about implementation'.\nHigh Cronbach alpha coefficients of 0.95 (T1), 0.92 (T2) and 0.94 (T3) were achieved.\nSecondary outcomes were self-reported measures of the decision (actual and intended actions), attitude towards MMR and beliefs about the MMR options, knowledge and anxiety.\nThe MMR decision was measured at three months post-intervention using a self-report item 'Since this study started, have you taken your child to have the combined MMR vaccine?'\nIn addition, a measure of intended choice was developed using three items measured on a 7-point scale e.g.\n'I intend to give my child the combined MMR vaccine at the recommended ages' (definitely do not-definitely do).\nThese three items were measured at all three time-points.\nResponses were averaged over the three items.\nCronbach alpha coefficients of 0.84 (T1), 0.79 (T2) and 0.90 (T3) were obtained.\nKnowledge about MMR and the measles disease was measured using multiple choice items developed for the purposes of this study using Department of Health for England and Wales literature.\nThe measure was not validated.\nThe number of questions answered correctly were summed to produce a total knowledge score (maximum 11).\nAttitude towards MMR was measured on a 7-point scale.\nParents responded to the statement 'For me to give my child the combined MMR vaccine at the recommended ages would be' on three semantic differential evaluative endpoints (1 to 7); e.g. extremely bad/extremely good.\nResponses were averaged over the three items.\nCronbach alpha coefficients were 0.78 (T1), 0.73 (T2) and 0.80 (T3).\nThese intended choice and attitude items have demonstrated validity and reliability and were informed by guidelines on measuring health cognitions.\nParents' beliefs about the MMR options were assessed using a modified version of the Beliefs about Flu Vaccination Questionnaire.\nThe measure was not validated for use in this context.\nFour items assessed parents' beliefs about the necessity of MMR e.g.\n'Without the combined MMR vaccine, my child could get very ill from measles, mumps or rubella' and four items assessed parents' concerns about MMR e.g.\n'Giving my child the combined MMR vaccine worries me'.\nAll items were scored on a 5-point scale (strongly disagree-strongly agree).\nItems for each sub-scale were summed.\nTotal scores for the two scales range from 4 to 20 with higher scores representing stronger beliefs in the necessity for, and concerns about, MMR.\nCronbach alpha coefficients for the necessity sub-scale were 0.70 (T1), 0.63 (T2) and 0.70 (T3).\nReliability was not improved by eliminating any items.\nCronbach alpha coefficients for the concerns sub-scale were 0.77 (T1), 0.75 (T2) and 0.78 (T3).\nAnxiety was measured to ensure that the parent meeting and MMR leaflet did not evoke anxiety in parents.\nWe used the short form STAI.\nSix items were used e.g. 'I feel calm', 'I am tense' and were scored on a 4-point scale (not at all-very much).\nThe positive items (e.g. calm) were reverse scored and all six items were summed.\nThe total score was multiplied by 20/6.\nA normal score is 34 to 36.\nHigh Cronbach alpha coefficients of 0.81 (T1), 0.86 (T2) and 0.84 (T3) were obtained.\nSample size\nTo achieve 80% power to detect a standardised effect size of 0.67 on the primary outcome of decisional conflict, using a two-sided t-test with significance level of 0.05 and an estimated ICC of 0.05 (giving a design effect of 1.5 based on an average of 11 parents per cluster) required a sample size of 108 parents (54 in each group).\nPredicting 25% attrition 73 parents were required in each group.\nParent numbers were not balanced across the clusters.\nBased on our previous research we estimated recruiting 12 parents per week over three months.\nAnalysis\nAn intention to treat analysis was conducted.\nThis trial design was clustered within centres (healthcare centres, childcare organisations) and had repeated measures.\nThe number of clusters and parents within each cluster were small in respect to multilevel modelling.\nTo explore the potential effectiveness of the intervention on the primary outcome (decisional conflict) longitudinal analysis was used.\nThis accounted for the multilevel structure of the data, with outcome measures collected at different time points within parent data.\nWe were interested in exploring how decisional conflict changed over time with respect to covariates of interest, namely arm, focal MMR decision, parent characteristics (age, ethnicity, marital status, education, relationship to child, if have older child) and intended choice, knowledge, attitude, beliefs and anxiety at recruitment.\nA normal model was used, using MLwiN 2.10 beta 5 to perform these analyses.\nDue to missing values, owing to non-completion of some questionnaire items, complete case analysis corresponded to only 65% of the data.\nOf these 92 parents, 44 (48%) were in the intervention arm and 48 (52%) were in the control arm.\nMissing values appeared to be at random.\nMultiple imputation was undertaken in Stata 10.0 to account for this.\nFive imputed datasets were generated using the results from linear regression analyses.\nPrior to undertaking multiple imputations seven parents were excluded (n = 3 intervention; n = 4 control) as they had not completed any study questionnaires, providing only parent characteristics data at recruitment.\nThe best fit model for the complete case data was fitted to each imputed dataset and compared with the best fit model for those data.\nAll imputed datasets agreed on the importance of the significant variables in the complete case model and results were found to be similar, thus indicating that minimal bias was introduced due to missing values.\nAggregated results from the five imputed datasets using 135 participants (142 minus 7) are presented.\nConfidence intervals were calculated using the widest values to allow for errors generated through the imputation.\nTwo sided significance tests and an alpha level of 0.05 were used throughout.\nRepeated measures ANOVAs were computed for the secondary outcome measures using multilevel modelling using the aggregated results from the five imputed datasets.\nChi squared analysis explored MMR uptake by arm.\nResults\nIntervention delivery\nThe parent meeting was delivered eight times during July and August 2006 in non-healthcare venues (e.g. community centres) close to participating healthcare centres and childcare organisations.\nFour daytime and four evening meetings were organised.\nCr\u00e8che facilities were provided at three daytime meetings.\nForty one parents attended a parent meeting, 23 did not.\nParents attending the meeting did not differ from those not attending the meeting in their characteristics or in their decisional conflict levels at baseline.\nThe mean number of parents attending was 6 (range: 2 to 10).\nOne meeting had less than four parents attending.\nClusters and participants\nParticipant flow through the study is presented in Figure 1.\nThe two arms were equivalent on all but one cluster characteristic.\nMean list size for the childcare organisations was larger for the control arm (Table 2).\nOf 1447 eligible parents invited, 150 (10%) consented to participate.\nEight parents did not meet the inclusion criteria (did not have an 'actual' MMR decision to make at that time).\nRecruitment of 142 parents fell short of the 146 target allowing for 25% attrition, but the required sample size of 108 parents was achieved because of a less than anticipated drop out.\nThe two arms were equivalent on all parent characteristics (Table 2).\nThere was a difference in the age of the first (youngest) eligible child and therefore in the MMR decision parents were making.\nOne third of parents in the intervention arm were making a first dose decision compared with almost two thirds in the control arm.\nDose decision was therefore modelled in the analysis.\nImpact of the parent meeting and MMR leaflet\nWas the parent meeting associated with a reduction in decisional conflict?\nMean decisional conflict by arm over time is presented in Figure 2.\nAt recruitment parents in both arms reported levels of decisional conflict above two indicating that they had sufficient conflict about the choice to interfere with making the MMR decision effectively.\nAt one week post-intervention mean decisional conflict had decreased for both arms to below two; and remained below two at three months post-intervention.\nTime was significantly associated with decisional conflict.\nThere was no significant association between arm and decisional conflict at any time point (see Table 3).\nIn short, post-intervention, parents could implement an effective decision irrespective of arm allocation.\nThe greatest reduction in decisional conflict occurred at one-week post-intervention.\nFocal MMR decision (first or second dose) was not significantly associated with decisional conflict i.e. perceived decisional conflict about this choice reduced over time for parents making first or second dose decisions.\nWas the parent meeting associated with MMR decision?\nSixty six parents provided self-report data about their MMR choice.\nOf these 66, 29 (44%) were in the intervention arm and 37 (56%) were in the control arm.\nThe remaining parents did not have children who were invited for vaccination within the study time period.\nNinety three percent of parents in the intervention arm reported taking their child for the vaccination compared to 73% in the control arm.\nThis difference was statistically significant (\u03c72 (1, N = 66) = 4.43, 95% confidence interval 3.1% to 37.2%, p = 0.04).\nWas the parent meeting associated with changes in parents' intended choice, knowledge, attitudes, beliefs or anxiety?\nTime by group mean scores, 95% confidence intervals and significance levels for secondary outcomes are presented in Table 4.\nSmall changes in the predicted direction were evident for the intervention arm for knowledge, intended choice, attitudes, and beliefs.\nHowever repeated measures ANOVAs revealed no significant time by arm effects.\nMean anxiety remained below or within the normal range suggesting that neither the parent meeting nor the MMR leaflet evoked anxiety in parents.\nWhich parent characteristics and cognitions were associated with changes in decisional conflict?\nBaseline parent characteristics and outcome measures (irrespective of arm allocation) associated with changes in decisional conflict were: whether the parent had an older child; intended choice, attitude and concern beliefs (see Table 3).\nIf a parent had previously made an MMR decision for an older child, decisional conflict decreased over time by 0.25 compared to a parent who had not previously made an MMR decision.\nParents' concerns about the potiential adverse consequences of MMR at recruitment were significantly associated with changes in decisional conflict over time.\nThe more concerned parents were, decisional conflict increased by 0.07.\nFor each additional point increase in attitude i.e. the more positive parents were about MMR, decisional conflict decreased by 0.20.\nFor each additional point increase in intended choice, i.e. the stronger parents' intentions were to vaccinate, decisional conflict increased by 0.09.\nDiscussion\nIn response to a continuing lack of confidence amongst many UK parents facing a decision about MMR and informed by our earlier work we developed and evaluated a parent-centred, multi-component intervention, delivered in community-based (non-healthcare) venues.\nWe believe this to be the first study to evaluate a multi-component intervention to support informed decision making for MMR.\nSome study limitations should be acknowledged.\nParent numbers were not balanced across the clusters thus reducing statistical power.\nHowever it seems unlikely that this would have changed the non-significant time by arm effect for decisional conflict.\nOnly a small number of parents provided self-report data about their choice thus the study may have been under powered on this secondary outcome.\nThe study was based in one city and only 10% of parents invited to take part did so.\nDue to the Data Protection Act we cannot determine if they differ to non-responders.\nHowever the immunisation policy in Leeds mirrors UK policy and the sample was consistent with other MMR research that identifies parents who find it difficult to make this decision.\nFinally, whilst complete case analysis was undertaken on just 65% of the data we believe that minimal bias was introduced.\nThe intervention was feasible to deliver in non-healthcare, community venues and it was acceptable to parents, with the majority expressing positive views.\nParents were equally positive about the MMR leaflet.\nOur measure of decisional conflict showed a statistically significant decrease over time for both the intervention and control arms to a level where an informed decision for MMR could be made.\nThe positive effect of the MMR leaflet on decisional conflict observed in the control arm was unexpected.\nThe inclusion of the leaflet was to reduce possible bias from a 'Hawthorne' effect.\nFurther, we considered the provision of a leaflet reflected usual vaccination practice.\nHowever, parents reported during meetings and in questionnaires that, contrary to stated local vaccination policy, leaflets were not routinely provided.\nFor some parents this may have been because their child was not invited for MMR vaccination at the time of the study.\nNevertheless, parents reported that this leaflet was more helpful in addressing their concerns about MMR compared to the usual Department of Health information at that time.\nConsequently instead of comparing our intervention with a control (usual care) we were comparing two different decision support interventions, a decision support leaflet versus participation in a parent meeting and a decision support leaflet.\nWe were, therefore, unable to identify the independent effects on perceptions about the decision process of the parent meeting from usual care as originally intended.\nThis study did find that parents in the intervention arm were significantly more likely to report taking their child for the MMR vaccination than parents in the control arm.\nThis suggests that providing information in a well-designed leaflet may be insufficient to lead to subsequent changes in the final choice (i.e. taking the child for immunisation).\nEnabling parents to act on their informed decisions may require a more pro-active approach than increasing knowledge and enabling clarification of their values.\nIn this study, it seems likely that the parent meetings provided sufficient decision support to enable parents to act on their decision, possibly illustrating a greater values-choice outcome.\nParents making 'proxy' decisions about MMR on behalf of a child may require additional decision support where vaccination/non-vaccination consequences of regret and blame may be common and where media-induced controversy has adversely affected public trust in government and medical authorities.\nIn this vaccination context, the more proactive of the two decision support interventions was associated with more children receiving the MMR vaccination.\nWe suggest that the concerns parents felt about this choice were met more fully by both interventions in this study than those parents who receive standard invitation letters for, and advice about, MMR in the UK.\nInterestingly, for both groups the observed improvements in informed decision-making occurred for parents making first and second dose decisions.\nPrevious research would suggest that the first dose decision (for a first child) is the most anxiety-provoking and that parents may, therefore, experience greater uncertainty.\nOur study suggests that parents may benefit from concerted decision support for both doses.\nWe also found that parents who had not previously made an MMR decision for an older child, those who were less positive in their attitude and more concerned about MMR had higher decisional conflict.\nThese parents could usefully be targeted for decision support.\nOur finding that parents with strong intentions to vaccinate had higher decisional conflict suggests targeting parents who are approaching vaccination in the near future rather than those for whom vaccination and its potential consequences are more remote.\nThe majority of decision support research focuses on preference sensitive decisions for which the 'best' decision is considered to be unclear and dependent on personal values.\nImmunisation is considered to be an 'effective' health decision for which the weight of the scientific evidence would typically lead a health professional to recommend a particular course of action.\nThis perhaps explains why governments and health professionals have historically adopted the 'knowledge deficit model' approach of simply providing information and reassurance to parents.\nIrrespective of how childhood immunisation 'delivery systems' are organised, there is evidence from international literature that parents' decision support needs are generally the same.\nMoreover, the below target MMR uptake rates in many countries suggests that reliance on a passive information-giving approach has limited effect.\nComprehensive decision support, as provided in our parent meeting offers a potential solution.\nConclusions\nWhilst both the leaflet and the parent meeting reduced parents' decisional conflict, the parent meeting appeared to enable parents to act upon their decision leading to vaccination uptake.\nWe are now testing the effectiveness of a web based MMR decision aid for parents that could be made easily accessible, for example in public places where parents frequent such as schools, libraries, community centres as well as in the waiting rooms of healthcare centres.\nParticipant flow.\nMean decisional conflict by arm over time. Intervention/Control Note. Scores lower than two are associated with 'implementing decisions', higher scores are interpreted with decision delay or feeling unsure about implementation.\n\nOverview of parent meeting\nTime | Facilitator | Content | Aims\n15 minutes | Parent facilitator and Researcher | WELCOMEIntroductionsOutline aims of meetingGo through programmeAgree ground rules for meeting | Of meetingTo provide parents with the opportunity to discuss MMR with other parents who are making an MMR decisionTo provide information about MMR from a variety of perspectivesTo introduce and practice one approach to supporting parents to ask questions about MMR of their healthcare practitioner\n30 minutes | Parent facilitator | GROUP DISCUSSIONAim of sessionReminder of ground rulesIntroductory question - would anyone like to tell us what you hoped to get out of this parent meeting today?Discussion | Of sessionTo provide parents with the opportunity to discuss any issues about MMR with other parents who are also making an MMR decision\n30 minutes | Immunisation Nurse Specialist | QUESTION AND ANSWER SESSIONAim of sessionParents ask questions of the immunisation nurse specialistParents are alerted to resources that they can take away | Of sessionTo provide parents with the opportunity to ask questions of the immunisation nurse specialist\n35 minutes | Researcher | COACHING EXERCISEAim of sessionBrief input on the importance of raising questions about MMR with a healthcare practitionerIntroduce and discuss the question prompt sheetRole play exercise using the question prompt sheetBrief discussion on usefulness of the question prompt sheet and role play | Of sessionTo introduce and practice one approach to supporting parents to ask questions about MMR in the primary care consultation\n10 minutes | Parent facilitator and Researcher | CLOSE OF MEETINGThank parents and provide overview of next stage of research study | \n\n\nBaseline characteristics of clusters and parents by arm\nCharacteristics | Intervention arm | Control arm\nPrimary healthcare centres, n | 3 | 3\n\nChildcare organisations, n | 3 | 3\n\nMean healthcare centre parent list size | 216 | 210\n\nMean childcare organisation parent list size | 19 | 30\n\nMean Low Income Scheme Index scorea | 10 | 11\n\n |  | \n\nParents, n | 71 | 71\n\nMean age \u00b1 SD, yrs | 34.07 \u00b1 5.43 | 34.06 \u00b1 5.52\n\nEthnicity, n (%) |  | \nWhite British | 68 (95.8%) | 68 (95.8%)\nOther | 3 ( 4.2%) | 3 ( 4.2%)\n\nMarital status, n (%) |  | \nSingle or living with partner | 27 (38.0%) | 13 (18.3%)\nMarried or re-married | 40 (56.4%) | 57 (80.2%)\nSeparated/Divorced/Widowed | 4 .(5.6%) | 1 (1.5%)\n\nRelationship to eligible child, n (%) |  | \nMother | 67 (94.4%) | 67 (94.4%)\nFather | 4 (5.6%) | 4 (5.6%)\n\nEducation |  | \nLeft school at 16 years | 24 (33.8%) | 25 (35.2%)\nLeft school at 18 years | 10 (14.1%) | 10 (14.1%)\nAchieved Degree or higher | 37 (52.1%) | 36 (50.7%)\n\nHave older child |  | \nYes | 36 (50.7%) | 36 (50.7%)\nNo | 35 (49.3%) | 35 (49.3%)\n\nFirst (youngest) child eligible, n (%) First dose MMR decision | 23 (32.4%) | 44 (62.0%)\nSecond dose MMR decision | 48 (67.6%) | 27 (38.0%)\nMean age \u00b1 SD of first (youngest) child eligible, months | 25.73 \u00b1 14.66 | 19.77 \u00b1 11.69\n\nSecond youngest child eligible, n (%) |  | \nFirst dose MMR decision | 1 ( 4%) | 0 (0.00%)\nSecond dose MMR decision | 24 (96%) | 22 (100.0%)\n\nMean age \u00b1 SD of second youngest child eligible, months | 50.56 \u00b1 17.13 | 49.32 \u00b1 21.41\n\nNote. N = 12 clusters, N = 142 parents. aLow Income Scheme Index score is based on the percentage of prescribed items exempt from a prescription charge due to low income of the patient.\n\nCoefficients for the longitudinal multilevel model of decisional conflict on potential covariates\nModel variables | Effect Estimate | 95% CI | p-value\nTime-1 week post-intervention | -0.54 | -0.67 to -0.41 | <0.001\n\nTime-3 months post-intervention | -0.60 | -0.73 to -0.47 | <0.001\n\nArm-intervention | 0.07 | -0.11 to 0.25 | 0.215\n\nMMR decision-2nd dose | -0.05 | -0.24 to 0.14 | 0.310\n\nOlder child | -0.25 | -0.42 to -0.07 | 0.003\n\nIntended choice | 0.09 | 0.02 to 0.17 | 0.006\n\nAttitude | -0.20 | -0.30 to -0.10 | <0.001\n\nConcern beliefs | 0.07 | 0.04 to 0.10 | <0.001\n\nNote. N = 135. Results are presented per one point increase of decisional conflict.\n\nDescriptive data for secondary outcomes\n |  | Intervention | Control | \nOutcome | Time point | Mean | 95% CI | Mean | 95% CI | p-valuea\n\nKnowledgeb | T1 | 6.38 | 6.00- 6.77 | 5.97 | 5.53- 6.41 | 0.253\n\n | T2 | 8.22 | 7.84- 8.60 | 7.83 | 7.52- 8.15 | \n\n | T3 | 7.30 | 6.93- 7.66 | 7.08 | 6.80- 7.36 | \n\nIntended choicec | T1 | 5.93 | 5.54- 6.32 | 5.10 | 4.60- 5.60 | 0.605\n\n | T2 | 6.03 | 5.67- 6.39 | 5.44 | 4.96- 5.92 | \n\n | T3 | 6.34 | 6.00- 6.68 | 5.58 | 5.06- 6.09 | \n\nAttituded | T1 | 4.98 | 4.66- 5.30 | 4.52 | 4.17- 4.88 | 0.786\n\n | T2 | 5.19 | 4.89- 5.48 | 4.77 | 4.44- 5.10 | \n\n | T3 | 5.26 | 4.96- 5.55 | 4.68 | 4.33- 5.04 | \n\nNecessity beliefse | T1 | 16.88 | 16.22-17.53 | 16.73 | 15.99-17.47 | 0.578\n\n | T2 | 17.63 | 16.97-18.28 | 17.18 | 16.49-17.87 | \n\n | T3 | 17.43 | 16.87-18.00 | 17.21 | 16.46-17.96 | \n\nConcern beliefse | T1 | 9.83 | 8.92-10.73 | 11.00 | 10.04-11.96 | 0.939\n\n | T2 | 8.92 | 8.15- 9.70 | 10.15 | 9.07-11.22 | \n\n | T3 | 8.66 | 7.81- 9.50 | 10.06 | 9.04-11.09 | \n\nAnxietyf | T1 | 32.50 | 29.63-35.37 | 34.20 | 31.52-36.89 | 0.219\n\n | T2 | 30.89 | 28.00-33.77 | 33.78 | 30.43-37.13 | \n\n | T3 | 31.46 | 28.49-34.43 | 33.39 | 30.09-36.68 | \n\nNote. N = 135. aSignificance value for time by arm interaction. Values range from b0 (no knowledge) to 11 (good knowledge); c1(definitely do not intend) to 7 (definitely do intend); d1 (extremely negative attitude) to 7 (extremely positive attitude); e4 (not at all necessary/concerned) to 20 (very necessary/concerned); f20 (low anxiety) to 80 (high anxiety).",
        "label": "unclear",
        "id": "task4_RLD_test_35"
    },
    {
        "paper_doi": "10.3390/ijerph9113806",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 2043 individuals, of which 440 were children < 5, from 260 householdsInclusion criteria: households were eligible if there was at least one child < 5\n\n\nInterventions: Plastic Biosand filter (117 households, 1012 individuals)Primary drinking supply (143 households, 1031 individuals)\n\n\nOutcomes: Diarrhoeal incidenceMicrobiological water quality\n\n\nNotes: Location: six rural communities, Tamale, GhanaLength: three months follow-upPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "A randomized controlled trial of the plastic BioSand filter (BSF) was performed in rural communities in Tamale (Ghana) to assess reductions in diarrheal disease and improvements in household drinking water quality.\nFew studies of household water filters have been performed in this region, where high drinking water turbidity can be a challenge for other household water treatment technologies.\nDuring the study, the longitudinal prevalence ratio for diarrhea comparing households that received the plastic BSF to households that did not receive it was 0.40 (95% confidence interval: 0.05, 0.80), suggesting an overall diarrheal disease reduction of 60%.\nThe plastic BSF achieved a geometric mean reduction of 97% and 67% for E. coli and turbidity, respectively.\nThese results suggest the plastic BSF significantly improved drinking water quality and reduced diarrheal disease during the short trial in rural Tamale, Ghana.\nThe results are similar to other trials of household drinking water treatment technologies.\n1. Introduction\nMany communities, especially in rural sub-Saharan Africa, still face significant challenges to provide access to improved drinking water sources and are struggling to meet the Millennium Development Goals for water and sanitation.\nIn Ghana, it is estimated that 93% of the urban population and 76% of the rural population have access to improved water and about 17% of urban and less than 10% of rural population have access to improved sanitation.\nThe lack of access to improved water and sanitation contribute significantly to diarrheal disease in the population.\nSpecifically, the Ghana Demographic and Health Survey suggests that diarrheal disease is a significant cause of morbidity and mortality in children less than five years of age, with 1 in 5 having reported diarrheal disease in the two weeks preceding the survey.\nIn the Northern region of Ghana, the two-week prevalence of diarrheal disease was almost 33%.\nAn immediate solution to address the lack of access to safe water is household water treatment (HWT), which allows households to treat drinking water at the point of consumption to improve its quality.\nStudies of HWT have shown that it can reduce the risk of diarrheal disease by 35% or more for a wide range of technologies in different settings and populations.\nOne technology that shows promise for areas where water has higher turbidity is the Hydraid\u00ae BioSand Water filter (plastic BSF), originally invented by David Manz.\nThis filter, which functions similarly to a slow sand filter, has been modified for intermittent use.\nThe most studied BSFs have been those having concrete housings.\nThe BSF\u2019s advantages are many, including a simple design, durable materials, local fabrication of the concrete housing, and provision of abundant quantities of water.\nThere have been four peer-reviewed published trials examining the health impact of the concrete BSF.\nThese trials suggest that use of the concrete BSF can reduce diarrheal disease by 50% or greater.\nWhile the BSF most often implemented is a concrete version, it has a relatively slow rate of daily production, can weigh as much as 500 lbs.\nwhen fully installed and can be difficult to transport to remote locations.\nA version of the BSF having a plastic housing has recently been produced and may overcome the problems in production, distribution and transport of the concrete filter.\nThe plastic BSF is light in weight, stackable and can be produced rapidly in large quantities by injection molding.\nThe plastic version of the BSF tested in this study is licensed by Manz, manufactured by Cascade Engineering and has specific depth and design parameters.\nIts sand filter bed has a smaller surface area than that of the concrete filter and it has tapered rather than straight side walls (see Figure 1).\nFurthermore, there is limited evidence of how well it will work in the field to improve water quality and reduce diarrheal disease risks, especially for waters with high turbidity.\nTo address the lack of field evidence of the performance of the plastic BSF to improve drinking water quality and reduce diarrheal disease in turbid surface water used by a population with a high diarrheal disease burden, a cluster randomized controlled trial (RCT) was performed in rural communities located in the Northern Region, Tamale, Ghana.\nThis study is one of three RCTs simultaneously performed on the plastic BSF in different geographic regions (including Cambodia and Honduras).\nThis is the first trial of the plastic BSF in this region of the world and only the second BSF RCT in Sub-Saharan Africa.\nUnique to this region and due to water scarcity, the communities often rely on water stored in \u201cdugouts\u201d, which are shallow surface water impoundments,.\nThe purpose of the RCT was to document the ability of the plastic BSF to improve water quality for both fecal indicator bacteria and turbidity and to reduce diarrheal disease in a setting where the population relies heavily on contaminated, highly turbid surface water for drinking and where the lack of access to water and sanitation are likely to be contributing significantly to morbidity and mortality in children under five years of age.\n2. Materials and Methods\n2.1. Research Setting, Study Population, and Participant Recruitment\nThis study (a cluster RCT) of the plastic BSF was conducted in six rural communities in Tamale, Ghana.\nAll villages and the water quality testing laboratory were located within 25 km of the city of Tamale (see Figure 2).\nField data collection took place between May to December 2008.\nThe six study communities and their households were selected based on the following criteria: child under the age of five years old, stored drinking water in the home, use of surface water as their primary drinking water source, did not spend most of the day selling goods in Tamale, were within 60 minutes from Tamale during the rainy season, and households agreed to participate.\nStudy design and protocols were approved by the Institutional Review Board of the University of North Carolina (IRB #08-0063) and the Ethical Review Committee of the Ghana Health Service.\nThe initial study was powered to detect a 25% reduction in diarrheal disease between the two groups based on an initial prevalence of diarrheal disease of 5%, 80% power and, \u03b1 = 0.05.\nWe also took into account the clustering of diarrhea within individuals and households and assumed four months of follow up visits as well as four people per household.\nBased on these sample size parameter values, we estimated the need for approximately 100 households in each study arm.\nPrior to recruitment, village elders were approached and informed about the study.\nIf the village elders were interested in participating, individual households were then asked to participate.\nHouseholds were excluded from the study if they did not have a child less than five years of age and/or did not want to participate.\nHousehold recruitment began on 10 May 2008 and was finalized on 16 June 2008, with informed consent obtained during the initial household visit.\nThe purpose of the initial cross-sectional recruitment questionnaire was to collect data on diarrheal disease prevalence in the households, risk factors of diarrheal disease, main drinking water sources, education levels of household members, access to sanitation and presence and type of any drinking water treatment practices.\nAccess to sanitation was assessed via questionnaire and a visual inspection of the facility, if it was present.\nLack of access was characterized by the absence of any type of latrine, pour flush toilet or other appropriate sanitation technology.\nThe initial cross-sectional recruitment was completed in six communities and a total of 260 households were recruited.\nAfter the initial recruitment, a baseline period of observation was performed prior to intervention with the plastic BSF.\nThe purpose of the baseline data collection period was to characterize and compare diarrheal disease and water quality between what would become the randomly selected intervention (plastic BSF) and control (no plastic BSF) villages.\n2.2. Intervention\nData collection for the longitudinal portion of the prospective cohort began on 17 June 2008 and was finished on 23 December 2008.\nHouseholds were visited at one week intervals and asked questions about diarrheal disease and water management practices in the home.\nDuring this time period, household water quality was also monitored periodically for total coliforms, E. coli and turbidity.\nThe randomization of villages and installation of plastic BSFs took place during the last week of August and first week of September 2008.\nBased on discussions with village leaders, randomization at the household level was deemed unacceptable by the majority of the villages.\nTherefore, the randomization was performed at the village-level.\nNumbers were assigned to the six villages and three numbers were selected from a random number generator to be the intervention villages.\nDue to the unequal size of the villages, more households were selected into the control group.\nAt the time of randomization, the six villages ranged in size from 14 to 70 households with children <5 years of age.\nAs a result of randomization, three villages with 70, 58 and 14 households were selected into the control group and three villages with 58, 24 and 33 households were selected into the intervention group (and received the plastic BSF).\nThe intervention phase of the study included weekly household observations from September 2008 to December 2008, with a potential of 15 weeks of household observations completed during this time period.\nAll households in the intervention and control groups continued to provide detailed information on weekly levels of diarrheal disease.\nFor household water quality analysis, turbidity was measured every two weeks and bacterial concentrations were measured five times during the intervention period, usually every two weeks.\n2.3. Diarrheal Disease Surveillance\nA consistent system of diarrheal disease surveillance was developed where one person in the household was identified as the primary respondent during the recruitment interview.\nThis person was surveyed weekly about diarrheal disease for all members of the household.\nUsing the following questions \u201cHas anyone in your house had diarrhea in the past one week?\u201d and \u201cIf yes, how many times did that person go in one 24-hour period?\u201d, the primary respondents were asked to verbally report any occurrence of diarrhea in the household within the last 7 days.\nAdditional questions on stool appearance (including the appearance of blood), duration of symptoms of diarrhea and use of treatment were also asked.\nIf the case of diarrhea was ongoing at the time of the visit, the case was asked about during the next visit to determine if it had resolved as well as to determine the duration of the case.\nOverall, there were 12 possible visits prior to plastic BSF installation as the intervention and 15 possible visits after plastic BSF installation, for a total of 27 potential weeks of observation for diarrheal disease surveillance.\n2.4. Water Quality Sample Collection and Analysis\nFrom the 260 households that were enrolled, samples of drinking water were taken during household visits from both the control and plastic BSF household groups.\nDuring the plastic BSF intervention, control households provided a sample of water used for drinking and BSF households providing three water samples at each visit: pre-filtered or untreated water, water directly from the plastic BSF outlet tube, and plastic BSF-treated water that had been stored for drinking.\nWater samples were collected by field staff directly into 500 mL sterile plastic collection bottles.\nThese samples were stored on ice and transported to the World Vision Laboratory at Savelugu, where they were immediately processed (within six hours of collection).\nAll samples were tested for total coliforms and E. coli using the IDEXX Colilert\u00ae Quantitray 2000\u00ae system (IDEXX Laboratories, Westbrook, ME, USA).\nMost probable numbers (MPN) for total coliforms and E. coli were determined using the IDEXX provided MPN table.\nTurbidity was tested using the Hach 2100P Portable Turbidimeter (Hach Company, Loveland, CO, USA).\n2.5. Data Analysis\nData from the initial cross-sectional questionnaire were used to compare the plastic BSF and control groups.\nPearson chi-squared tests were used to assess the proportion in each group for the following variables: access to sanitation, main drinking water source, educational attainment, having multiple children <5 years of age in the household, and reported drinking water treatment practices.\nThe effect of the plastic BSF on diarrheal disease was determined by comparing the longitudinal prevalence of diarrheal disease for all participants in each group, intervention (received BSF) and control (no BSF) using longitudinal prevalence ratios (LPR) generated from Poisson regression.\nIn order to classify a case of diarrheal disease, we used the World Health Organization (WHO) definition of three or more loose or watery stools in at least a 24-hour period.\nTo adjust for clustering within households and the villages, multi-level Poisson regression was performed and all data reported are based on the LPR from the regression model adjusted for clustering.\nTo fit the model, days with diarrheal disease (as counts) and person-days of observation were totaled by individual participant.\nThe multi-level model was used to adjust for clustering because individuals belonged to households which belonged to villages.\nAll diarrheal disease data analysis was performed in Stata 10.0 (Stata, StataCorp, College Station, TX, USA).\nBacterial concentration and turbidity data were log10 transformed and analyzed in Microsoft Excel and Stata 10.0 for graphical presentation and means testing.\nThe bacterial and turbidity reductions achieved by the plastic BSF were calculated as log10 reductions: log10 reduction = log10 influent \u2013 log10 effluent (Equation 1).\nFiltered drinking water quality of plastic BSF households was compared both for water taken directly from the filter and for stored filtered water and compared to untreated water from BSF and control households.\nPaired and unpaired t-tests were used to compare geometric mean log10 E. coli concentrations and geometric mean turbidities between plastic BSF and control water samples.\n3. Results\n3.1. Study Enrollment and Completion\nDuring the initial cross-sectional recruitment interview, six villages and 260 households were recruited into the study.\nThree villages were later randomized into the BSF intervention group and a total of 117 plastic BSFs were installed in separate households.\nThree villages were selected to remain as the control villages.\nAll control households were asked to continue normal water management practices for the intervention period.\nAlthough this randomization resulted in a small number of clusters, this small number of villages was selected to facilitate weekly follow-up visits.\nA larger number of villages would have made it logistically challenging to complete all visits within one week (the desired diarrheal disease recall period).\nPrior to randomization, during the baseline period, nine households (3.4%) dropped out of the study.\nDuring the plastic BSF intervention period, a total of seven households (2.3%) dropped out of the study; two households from the control group and five households from the plastic BSF intervention group.\n3.2. Baseline Characteristics and Group Comparability\nA total of 1012 people in the 117 households randomized to the plastic BSF-intervention villages and a total of 1031 people in the 143 households randomized to the control villages were compared.\nShown in Table 1 and Table 2 are characteristics of the plastic BSF and control groups based on data collected during the initial cross-sectional questionnaire.\nWhen the two groups were compared based on the age and proportion of males and females in various age groups, the two groups differed (statistically) in the proportion of those that were <2, 2\u20134 and >4 years of age, although the proportions were similar.\nThe proportion of males to females and the number of respondents that reported currently attending school were not significantly different between BSF and control households.\nPlastic BSF and control group characteristics regarding water, sanitation, hygiene, and other household level characteristics are presented in Table 2.\nAn overwhelming majority of households reported using surface water (collected from earthen dams, called \u201cdugouts\u201d) for drinking water in both dry and rainy seasons (71\u201398%).\nThese dams or dugouts are typically shallow areas with slightly raised banks that capture rain or runoff water during the rainy season.\nThis is the most common source for drinking water in this region of the country.\nFewer control households (71%) compared to BSF households (94%) reported using surface water from earthen dams during the rainy season, a difference that was statistically significant.\nHouseholds that reported using a source other than earthen dams during the rainy season reported using rainwater.\nAlmost all households lacked access to any type of sanitation (97 and 98%, respectively for control and plastic BSF households).\nThe two groups were not found to be significantly different when compared for other water and sanitation or demographic variables listed in Table 2, such as the practice of cloth sieving for water treatment, the proportion of households with at least one person currently attending school or households with more than one child less than five years of age.\n3.3. Diarrheal Disease\nIn order to examine the impact of the plastic BSF on diarrheal disease of participants, we compared the longitudinal prevalence between the two groups, plastic BSF intervention and control (no BSF) for the age groups of <2 years, all <5 years and all ages, both prior to the intervention and after installation of the plastic BSF as the intervention (Table 3).\nBefore intervention, households that were randomly selected to receive plastic BSFs experienced slightly lower longitudinal prevalence of diarrheal disease than control households for all categories of age groups, a difference that was not statistically significant (adjusted LPR for all ages: 0.98, 95% CI: 0.23\u20133.94).\nDuring the plastic BSF intervention period, longitudinal prevalence of diarrheal disease of all age groups was significantly lower in the plastic BSF intervention group than in the control group.\nFor example, for all ages, the BSF intervention group had 0.40 times the longitudinal prevalence of reported diarrhea as the control group (95% CI: 0.05, 0.80).\nThis observation suggests significant protection from diarrheal disease by the plastic BSF during the four month intervention period.\nWhen stratified by age group, the difference in longitudinal prevalence of diarrheal disease between participants in control and BSF households was even greater in children less than five years of age than in all age groups, with an adjusted LPR of 0.26, (95% CI: 0.07\u20130.89), corresponding to an estimated 74% reduction in diarrheal disease for plastic BSF participants compared to control participants.\nThe level of diarrheal disease reduction for BSF participants compared to control participants was only slightly lower for children less than two years of age compared to all participants combined, with an (adjusted LPR of 0.37 (95% CI: 0.15, 0.90)), corresponding to an estimated 63% diarrheal disease reduction.\n3.4. Water Quality Analysis\nHousehold drinking water quality was compared over the entire study period for plastic BSF and control households.\nThe geometric mean concentrations of E. coli and mean turbidities of household drinking water for the baseline and intervention periods are compared in Table 4.\nBefore the intervention, plastic BSF households and control households had water with very high geometric mean concentrations of E. coli and turbidity: 724 and 832 MPN E. coli /100 mL, respectively (control and plastic BSF).\nLikewise, households in both groups had drinking water with high geometric mean turbidities; 95 and 85 NTU for control and plastic BSF, respectively.\nNeither E. coli nor turbidity levels of household water were statistically significantly different between the two groups prior to the plastic BSF intervention (two sample t-test for turbidity (p = 0.23) and for E. coli, (p = 0.22)).\nAfter the plastic BSF intervention, households with the plastic BSF demonstrated appreciable improvements in drinking water quality compared to the control household without the BSF (Table 4).\nDuring the intervention period, E. coli concentrations were similar in the water that was collected and stored in the household before treatment (490 and 437 MPN/100mL for control and BSF groups, respectively; p value = 0.27).\nFor the plastic BSF group, E. coli levels in water from the plastic BSF were significantly lower than those in the stored untreated water, (16 MPN/100mL and 437 MPN/100mL, in BSF direct filtrate and untreated stored water, respectively, p value < 0.0001).\nPlastic BSF treated and stored water was also significantly improved based on E. coli levels compared to untreated stored water in BSF households (76 MPN/100 mL and 437 MPN/100 mL, respectively; p < 0.0001).\nHowever, plastic BSF treated and stored water had significantly higher E. coli concentrations than water taken directly from the plastic BSF outlet, with 76 MPN/100 mL and 16 MPN/100 mL, respectively; p < 0.0001.\nAfter the plastic BSF was installed, both the control and plastic BSF intervention groups had collected stored water with lower turbidity as compared to the baseline phase (near 100 NTU prior to intervention and <50 NTU for both groups during the intervention phase).\nHowever, plastic BSF households had significantly lower turbidity for water treated by the plastic BSF (15 NTU compared to 27 NTU for control households).\nUnlike E. coli levels, turbidity levels of water did not change significantly during storage.\nControl and plastic BSF households were asked (at each household visit) whether or not they had performed treatment to their water.\nPrior to plastic BSF intervention, households reported sieving the water through a cloth prior to drinking for 97% of all observations.\nBoiling and chlorine were cited as treatment for 0.2% and 0.15% of the observations, respectively.\nDuring the intervention period, the control group reported high levels of cloth-sieving (93% of all observations) but no additional treatment.\nThe plastic BSF group did not report any additional treatment beyond filtration with the plastic BSF during the intervention period.\nIn an attempt to measure adherence to the plastic BSF intervention, households reported the frequency of weekly plastic BSF use and use of the plastic BSF filtered water for drinking.\nDuring the weekly observations, none of the plastic BSF intervention households reported not using the BSF to filter water in the previous 7 days.\nIn addition, in very few observations (3% all household observations in plastic BSF households) users report not drinking the plastic BSF filtered water.\n3.5. Plastic BSF Performance\nThe plastic BSF achieved a geometric mean 97% reduction of E. coli when untreated water in BSF households was compared to water direct from the BSF outlet.\nHowever, the E. coli reduction was much lower (85%) when comparing untreated water in BSF households to BSF-treated and stored water.\nA geometric mean reduction of 67% for turbidity was found comparing untreated water to water directly from the BSF; a similar reduction was found comparing untreated to BSF treated and stored water (66% geometric mean reduction).\nWhen compared on a categorical basis of order of magnitude concentration ranges as a basis for categorizing risk levels posed by the water, plastic BSF treated water had significantly fewer samples in high risk E. coli concentration categories as compared to untreated source water (p < 0.0001, Pearson\u2019s chi-squared test).\nAs shown in Figure 3, 44% and 15% of water samples taken directly from the plastic BSF outlet and the plastic BSF treated and stored water, respectively, had less than 10 MPN E. coli/100mL (considered low risk water) as compared to only 1.5% of water in BSF households prior to treatment.\nFurthermore, 77 and 56% of samples taken from the plastic BSF outlet directly or plastic BSF treated and stored, respectively, had fewer than 100 MPN E. coli/100mL (considered moderate risk) as compared to only 12% of all samples in this category prior to plastic BSF treatment in plastic BSF intervention households.\nOverall, there was significant improvement in categorical concentrations of E. coli for plastic BSF treated drinking water and plastic BSF treated and stored water compared to untreated water.\n4. Discussion\nTo our knowledge, this is the first study to assess the ability of plastic BSF to reduce the longitudinal prevalence of diarrheal disease in Ghana.\nThere are relatively few studies examining household water filtration in communities in this region of the world and in water sources as turbid as the ones found in this study.\nWe documented a significant reduction (60%) in diarrheal disease for households who were asked to use a plastic BSF compared to control households that were given no BSF during the study period.\nOne similar study, performed in Kenya with a concrete housing intermittently operated slow sand filter, found a 54% reduction in diarrhea prevalence.\nThey also found that the reduction of diarrheal disease was substantially higher (77%) when comparing households only using unimproved surface water sources for their drinking water supply.\nSimilarly high reductions (80%) in diarrheal disease were also reported in a randomized controlled trial of a ceramic water filter in South Africa and Zimbabwe.\nThe results from our study suggest reductions in diarrheal disease that are consistent with similarly designed trials on household water filters in other countries in this region.\nThe results from this study are also similar to the diarrheal disease reduction results we documented (as part of the three country trial of the plastic BSF) from an RCT in Cambodia, where diarrheal reductions in households using the plastic BSF were similar (59%) to those reported here during a five-month intervention trial.\nIn Honduras, the diarrheal disease reduction results of a plastic BSF RCT were similar in magnitude to those reported here although this was not statistically significant.\nOverall, the results for diarrheal disease reduction in BSF households compared to control households observed in this study suggest greater or comparable reductions compared to those obtained from trials of concrete BSFs in other regions of the world such as Cambodia and the Dominican Republic, which demonstrated 47%, and 54% reductions in diarrhea illness risks, respectively, in children <5 years of age.\nThese results also compare favorably to peer-reviewed studies of other HWT technologies such as chlorine disinfection, which has been found to achieve reductions in diarrheal disease of about an average 30% in many regions around the world.\nResearchers, however, have recently questioned the validity of results of unblinded, randomized controlled trials lacking a placebo such as this one, suggesting that there may be a significant source of bias where households with the intervention may under-report diarrheal disease.\nBecause this is the first study of the plastic BSF in Northern Ghana, we did not include a placebo filter.\nFurthermore, due to the high turbidity of the surface water used as drinking water by households, it would have been difficult to blind participants to the treatment.\nBoisson et al. recently performed a placebo controlled trial of a household filter in the Democratic Republic of Congo and experienced considerable challenges in designing and implementing a neutral filter as a placebo.\nWe are unaware of any study that successfully employed a placebo filter at point-of-use in any developing country setting.\nAdditional research on the plastic BSF and other HWT technologies should attempt to measure health impacts in more objective ways that can help to eliminate bias.\nThese may include incorporating diagnostic procedures to detect intestinal infections or including anthropometric measurements of children for longer-term health outcomes such as was performed in a recent trial of solar disinfection in Kenya.\nHowever, even child anthropometry can be subject to measurement error and may not completely eliminate the question of bias.\n4.1. Effect of the Plastic BSF on Household Drinking Water Quality\nDuring the four months of intervention with the plastic BSF, we documented significant improvements in household water quality for both E. coli and turbidity of plastic BSF treated water.\nBecause the lack of access to improved drinking water was high in these communities (71\u201398% used surface water which was collected from earthen dams) and the water accessed was highly microbiologically contaminated and often very turbid, this study provides important evidence regarding the potential application of the plastic BSF for locations where safe water access is limited.\nDuring the study, the plastic BSF demonstrated an average 97% reduction of E. coli in BSF-treated water, which is similar to reductions seen both in the laboratory and in the field for concrete BSFs and other filtration technologies.\nFor the untreated water of this study, when E. coli concentrations were \u22651000 MPN/100mL, the plastic BSF averaged 99% reduction (data not shown).\nAs demonstrated in laboratory studies of a similarly designed plastic BSF, protozoan parasites and bacterial removals can be as high as \u226599%.\nVirus removals may not be as robust (<90%) and may be more dependent on the conditions of the biological activity and the biofilm in the BSF.\nUnder the conditions examined during our study, we expect bacterial and protozoan pathogens to be effectively removed by the plastic BSF at levels similar to the removals of the bacterial indicators (97% or more).\nFurthermore, there is potential for virus removals greater than what has been documented in the laboratory due to the potential for virus attachment to particles in the turbid raw water and the likelihood of robust biofilms as the result of this highly turbid raw water.\nRotavirus have been documented as an important pathogen in the region and further study on the potential for this BSF to remove viruses under the conditions of this study is warranted.\nAlthough there were significant reductions of E. coli in household drinking water samples as a result of treatment with the plastic BSF, there was also evidence of recontamination during storage of plastic BSF-treated water (Figure 3, Table 4).\nThis type of recontamination has been documented in previous studies of the concrete BSF.\nAs shown in Figure 1, although the water may leave the outlet tube of the plastic BSF relatively uncontaminated, during storage of treated water there are ample opportunities for bacterial contamination to be re-introduced via hands, dippers and even the storage container itself.\nThe opportunity for bacterial recontamination has been documented for other treatment options that do not provide a residual disinfectant, such as boiling.\nFor these types of technologies, additional training regarding safe and hygienic storage of treated water should be included to reduce bacterial recontamination.\nWhile the turbidity reduction in household drinking water by the plastic BSF in the intervention group was significant compared to untreated water and to control household water turbidity, the average turbidity of treated drinking water of 14 NTU was still higher than the WHO suggested limit of 5 NTU.\nFew studies have examined treatment of water in regions where the main source of drinking water is highly turbid surface water.\nCrump et al. demonstrated a significant reduction in diarrheal disease, bacterial contamination and turbidity for households using a combined flocculant-disinfectant in Kenya, with turbidities reduced from >100 NTU to <5 NTU in 50\u201370% of water samples treated.\nMwabi et al. reported that a range of locally produced point-of-use water filters, including BSFs, consistently reduced the turbidities of surface waters with an average turbidity of about 40 NTU to <5 NTU in South Africa.\nAlso important to consider was that turbidity decreased for both control and plastic BSF households during the intervention period.\nThis may be due to water quality improvements at the point of collection, possibly associated with changes in rainfall.\nLimitations of this study include randomization at the village level which resulted in a small number of clusters, having participants unblinded to the exposure and lack of a placebo.\nAdditional limitations include use of self-reported diarrhea, with a 7-day recall of diarrhea, as the health outcome measure and the relatively short duration of the intervention.\nHowever, such limitations are common in other studies of HWT technologies such as the ceramic water filter and chlorination.\nThe short duration and frequent observations of this study are important limitations of its design because evidence suggests that many point-of-use household water interventions show decreased performance effectiveness over time.\nA recent study investigated the impact of household survey on health behaviors and found that the act of survey itself may impact respondents\u2019 behaviors.\nTherefore, studies of longer duration and those that are designed to eliminate household survey impact on respondents\u2019 behaviors should also be considered for the BSF in future studies.\nDesign and conduct of studies that evaluate implementation programs extending over longer periods of time would also be more informative of the longer term impacts of the plastic BSF on correct and consistent daily use, water quality and health.\nDespite the aforementioned limitations in this study, the results can still be compared to other rigorous and widely cited studies of HWT technologies and their impact on the diarrheal disease risks of the users.\nIn particular, these results confirm those of past RCTs of the concrete BSF by documenting that the plastic BSF has the ability to reduce the longitudinal prevalence of diarrheal disease and significantly improve drinking water quality, even in a cohort of households that is primarily using unimproved, turbid and microbiologically contaminated surface water from earthen dams for their main source of drinking water.\n5. Conclusions\nTo our knowledge, this is only one of two known health impact studies on the performance of the BSF in Sub-Saharan Africa and is the first on the plastic BSF on this continent.\nPositive results were found for the ability of the plastic BSF filter to reduce diarrheal disease and improve water quality in communities of the Northern Region of Ghana using highly turbid and microbiologically contaminated surface water collected from earthen dams as their drinking water source.\nHowever, more research is necessary to document plastic BSF performance, continued and effective use and overall sustainability after installation and in the absence of intensive sampling or other monitoring.\nTypical set-up of the plastic BSF in a household in rural Tamale, Ghana (2008).\nMap a (1:500,000) indicating the location for the Randomized Controlled Trial of the Plastic Biosand Filter in Tamale, Ghana (2008).\nComparison of categorical E. coli concentrations in different water samples from plastic BSF households during intervention period of the RCT of the plastic BSF in rural Tamale, Ghana (2008).\n\nAge (as of May 2008), sex and education status of participants in control (no BSF) and intervention (BSF) households during a randomized controlled trial of the plastic BSF in rural Tamale, Ghana 2008.\nIndividual Variables | Control | Intervention | p values\n | (N = 1031) | (N = 1012) | Pearson \u03c72 test\nAge | N (%) | N (%) | \nParticipants \u2265 5 years old | 809 (78.5) | 794 (78.5) | \nParticipants 2\u20134 | 146 (14.3) | 116 (11.5) | \nParticipants <2 | 76 (7.4) | 102 (10.2) | 0.027 a\nSex |  |  | \nMale (\u22655) | 436 (42) | 391 (39) | \nMale (<5) | 113 (11) | 116 (11) | 0.37 a\nFemale (\u22655) | 373 (36) | 401 (40) | \nFemale (<5) | 108 (11) | 104 (10) | 0.12 a\nEver or currently attending school | 261 (25.3) | 260 (25.7) | 0.85\n\na Pearson chi-squared test performed for proportions in all categories of age and gender comparing people in BSF and control groups.\n\nSelected characteristics regarding demographics, water and sanitation for control (no BSF) and intervention (BSF) households in a randomized controlled trial of the plastic BSF in rural Tamale, Ghana, June\u2013December 2008.\nHousehold Level Variables | Control | Intervention | p values\n(N = 143) | (N = 117) | (\u03c72 test)\nUse surface water during dry season | 95.0% | 98.3% | 0.16\nUse surface water during rainy season | 70.6% | 94.0% | <0.001\nReported sieving drinking water through cloth | 96.5% | 96.6% | 0.97\nAt least one person attending school in household | 70.4% | 74.8% | 0.39\nLack access to sanitation a | 97.1% | 98.3% | 0.67\nMulti-child household b | 56.7% | 55.6% | 0.86\n\na Lack of access was characterized by the absence of any type of latrine or pour flush toilet.\nb Household has at least two children less than five years of age participating in the study.\n\nAdjusted longitudinal prevalence ratios for diarrheal disease, stratified by age, during the pre-intervention and intervention phases of a randomized controlled trial of the plastic BSF in rural Tamale, Ghana (2008).\nData collection Period | Age stratum | Unadjusted LP a\u2014Control Villages | Unadjusted LP a\u2014Plastic BSF Villages | Adjusted LPR (95% CI)\nBaseline (May\u2013August 2008) b  | All | 0.024 | 0.020 | 0.98 (0.23, 3.95) c\n<2 years of age | 0.081 | 0.10 | 1.56 (0.25, 9.83) d\n<5 years of age | 0.078 | 0.074 | 1.38 (0.19, 10.17) d\nPlastic BSF Intervention (September\u2013December 2008) | All | 0.012 | 0.0063 | 0.40 (0.05, 0.80) c\n<2 years of age | 0.028 | 0.015 | 0.37 (0.15, 0.90) d\n<5 years of age | 0.034 | 0.018 | 0.26 (0.07, 0.89) d\n\na LP\u2014unadjusted longitudinal prevalence which was calculated as the total number of days with diarrheal disease over the total number of days observed; b Period of observation in villages prior to randomization and plastic BSF installation; c Longitudinal prevalence ratio and 95% confidence interval with plastic BSF as exposure adjusted for adjusted for categorical age of participant and clustering of diarrheal disease within household and villages; d Longitudinal prevalence ratio and 95% confidence interval with plastic BSF as exposure adjusted for clustering of diarrheal disease within household and villages.\n\nGeometric mean E. coli concentrations and turbidities of household drinking waters for the control and plastic BSF groups before and after plastic BSF intervention in a randomized control trial in rural Tamale, Ghana (2008).\nBaseline (May\u2013August 2008) | Plastic BSF Intervention (September\u2013December 2008)\nWater Quality Parameter | Control HH Water | Plastic BSF HH Water | Control HH Water | Plastic BSF Untreated Water b | Plastic BSF Direct Filtrate | Plastic BSF Stored Filtrate\nE. coli (95%CI) a | 724 (631\u2013851) | 832 (724\u2013977) | 490 (426\u2013549)  | 437 (380\u2013501)  | 16 (13\u201320) | 76 (62\u201391)\n(N = 516) | (N = 424) | (N = 587) | (N = 385) | (N = 382) | (N = 381)\nNTU | 95 (83\u2013109)  | 85 (74\u201398)  | 25 (23\u201327) | 47 (42\u201351)  | 15 (13\u201317) | 15 (14\u201318)\n(N = 523) | (N = 430) | (N = 787) | (N = 527) | (N = 524) | (N = 527)\n% E. coli Reductions c | -- | -- | -- | -- | 97 | 85\n% NTU Reductions c | -- | -- | -- | -- | 67 | 66\n\na Geometric mean and 95% confidence interval for E. coli (MPN/100mL) and turbidity of household (HH) drinking water; b Untreated water refers to water that was taken from households prior to any treatment. This is the assumed contamination level before the BSFs were used to treat the water.; c\nE. coli and NTU reduction calculated as log10 reduction = log10 influent \u2013 log10 effluent and then the log10 reductions were transformed into %.",
        "label": "unclear",
        "id": "task4_RLD_test_686"
    },
    {
        "paper_doi": "10.1371/journal.pone.0001023",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Individually RCTDates of trial: June to Sept 2006.\n\n\nParticipants: 108 children with fever > 37.5degC or history of fever in last 48 hours and P. falciparum mono-infection 500 to 100,000/mL.Age three to 15 years.Both sexes.Site: Mynuzi health centre, North-Eastern Tanzania, a hyperendemic area with rainy seasons in March to June and October to DecemberExclusion criteria: Hb < 8, inability to take drugs orally, known hypersensitivity to meds, reported anti-malarial treatment in last 2 weeks, evidence of chronic disease or acute infection other than malaria, domicile outside trial area, signs of severe malaria, eligible for other malaria studies.\n\n\nInterventions: AS+SP: AS: 4 mg/kg once daily for 3 days; SP: S 25 mg/kg and P: 1.125 mg/kg.AS+SP+PQ: As above for AS and SP plus PQ base 0.75 mg/kg on the third day.\n\n\nOutcomes: Proportion of people with gametocytes (by microscopy) days 1, 4, 8, 15, 29 and 43 (reported as 0, 3, 7, 14, 28, and 42).Proportion with gametocytes (by PCR), same time points.Gametocyte density by PCR.AUC for gametocyte presence.Adverse events.Adequate clinical and parasitological response.Haemoglobin.\n\n\nNotes: Hb outcome assessed with respect to G6PD variant\n\n",
        "objective": "To assess the effects of single dose or short\u2010course PQ (or an alternative 8AQ) alongside treatment for people with P. falciparum malaria.",
        "full_paper": "Background\nP. falciparum gametocytes may persist after treatment with sulphadoxine-pyrimethamine (SP) plus artesunate (AS) and contribute considerably to malaria transmission.\nWe determined the efficacy of SP+AS plus a single dose of primaquine (PQ, 0.75 mg/kg) on clearing gametocytaemia measured by molecular methods.\nMethodology\nThe study was conducted in Mnyuzi, an area of hyperendemic malaria in north-eastern Tanzania.\nChildren aged 3\u201315 years with uncomplicated P. falciparum malaria with an asexual parasite density between 500\u2013100,000 parasites/\u00b5L were randomized to receive treatment with either SP+AS or SP+AS+PQ.\nP. falciparum gametocyte prevalence and density during the 42-day follow-up period were determined by real-time nucleic acid sequence-based amplification (QT-NASBA).\nHaemoglobin levels (Hb) were determined to address concerns about haemolysis in G6PD-deficient individuals.\nResults\n108 individuals were randomized.\nPfs25 QT-NASBA gametocyte prevalence was 88\u201391% at enrolment and decreased afterwards for both treatment arms.\nGametocyte prevalence and density were significantly lower in children treated with SP+AS+PQ.\nOn day 14 after treatment 3.9% (2/51) of the SP+AS+PQ treated children harboured gametocytes compared to 62.7% (32/51) of those treated with SP+AS (p<0.001).\nHb levels were reduced in the week following treatment with SP+AS+PQ and this reduction was related to G6PD deficiency.\nThe Hb levels of all patients recovered to pre-treatment levels or greater within one month after treatment.\nConclusions\nPQ clears submicroscopic gametocytes after treatment with SP+AS and the persisting gametocytes circulated at densities that are unlikely to contribute to malaria transmission.\nFor individuals without severe anaemia, addition of a single dose of PQ to an efficacious antimalarial drug combination is a safe approach to reduce malaria transmission following treatment.\nTrial Registration\nControlled-Trials.com ISRCTN61534963\nIntroduction\nThe majority of anti-malarial drug treatments target the asexual blood stages of Plasmodium falciparum that are responsible for clinical disease and death.\nSexual stage parasites, gametocytes, can also be present in infected individuals and are responsible for the transmission of the parasite to mosquitoes.\nDrugs specifically targeting these sexual stage parasites may affect the spread of malaria in the human population.\nAnti-gametocyticidal drugs are used by several countries to prevent onward transmission from clinical malaria cases and have also been evaluated by mass drug administration to reduce malaria transmission in communities.\nArtemisinin-based combination therapies (ACT) are advocated as first-line antimalarial treatment because of their high treatment efficacy and beneficial effects on malaria transmission.\nAlthough ACT efficiently reduces microscopic levels of gametocytes, submicroscopic gametocytes (detected by molecular analysis) may persist after treatment and allow post-treatment malaria transmission.\nThe implementation of ACT may have a beneficial influence on malaria transmission in the general population but ACT may not be sufficient to completely prevent post treatment malaria transmission.\nPrimaquine (PQ) may be of added value in attempts to block malaria transmission as part of a mass drug administration.\nPQ is an 8-aminoquinolone that is widely used for the treatment of P. vivax malaria and actively clears mature P. falciparum gametocytes.\nAlthough there is no consensus about which drug is the most potent gametocytocidal drug, artesunate (AS) may predominantly inhibit gametocyte development while PQ may accelerate gametocyte clearance.\nIn combination with sulphadoxine-pyrimethamine (SP) and AS, PQ was found to be safe and highly efficacious in clearing asexual parasites and P. falciparum gametocytes detected by microscopy.\nThe efficacy of this combination on submicroscopic gametocytaemia is unknown and, in general, information on PQ use in Africa is scarce.\nPrior to the wide-scale introduction of ACT, a single dose of PQ following first line antimalarial treatment was recommended by the World Health Organisation to reduce malaria transmission in low endemic areas.\nAlthough several countries adopted this recommendation, there is concern for negative haemolytic side effects in individuals who are glucose-6-phosphate-dehydrogenase (G6PD) deficient.\nHere, we determine the safety and efficacy of SP+AS plus a single dose of PQ on clearing submicroscopic levels of P. falciparum gametocytaemia in an area of hyperendemic malaria in north eastern Tanzania.\nPossible haemolytic effects of PQ were determined in relation to G6PD status.\nMethods\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nThe trial was registered at Current Controlled Trials; ISRCTN61534963; http://www.controlled-trials.com/ISRCTN61534963/.\nRegistration was done after patient recruitment started due to communication problems.\nParticipants\nThis study was conducted in the period July through September 2006 in Mnyuzi, a rural village in the Tanga Region, north eastern Tanzania.\nMalaria transmission intensity is high with an estimated entomological inoculation rate (EIR) of 91 infectious bites per person per year.\nThe rainfall pattern is bimodal, with a long rainy season between March and June, and a short rainy season between October and December.\nThe study protocol was approved by the ethics committees of Kilimanjaro Christian Medical Centre, the Tanzanian National Institute for Medical Research (NIMR/HQ/R.8a Vol. XIII/446) and the London School of Hygiene and Tropical Medicine (#4097).\nParticipants were recruited from among children consulting the Mnyuzi health centre and who were resident within a 10 kilometres radius.\nInformed consent was obtained form the child's parents or guardians prior to inclusion.\nChildren aged 3\u201315 years with a temperature >37.5C\u00b0 or a history of fever within the last 48 hours and with P. falciparum mono-infection at a density between 500\u2013100,000 parasites/\u00b5L were eligible for recruitment.\nExclusion criteria were: a haemoglobin (Hb) concentration measured by Hemocue\u00ae below 8g/dL, inability to take drugs orally, known hypersensitivity to any of the drugs given, reported treatment with antimalarial chemotherapy in the past 2 weeks, evidence of chronic disease or acute infection other than malaria, domicile outside the study area, signs of severe malaria and eligibility for other malaria studies conducted in the region.\nInterventions\nParticipants enrolled were randomized to one of the two treatment regimes:\nSulphadoxine (25 mg/kg) and pyrimethamine (1.25 mg/kg) as a single dose (SP; Fansidar\u00ae, Roche, Switzerland) plus artesunate (AS), 4 mg/kg once daily for three days (Arsumax\u00ae, Sanofi-aventis, France) plus placebo once on the third day (Organon, The Netherlands);\nSP plus AS plus primaquine (base; department of clinical pharmacology, Radboud University Nijmegen Medical Centre, the Netherlands) as a single dose on the third day (0.75 mg/kg).\nPrimaquine capsules were produced following regulations of the European Pharmacopeia.\nTreatment was administered by staff at the recruitment clinic.\nEach child was observed for 30 minutes after treatment, a replacement dose was given in case of vomiting.\nNone of the children had repeated vomiting.\nParacetamol (10 mg/kg) was given until symptoms had subsided.\nIn case of parasitological treatment failure, rescue treatment with mefloquine was administered (Lariam\u00ae, Roche, Switzerland; 15 mg/kg on first day and 10 mg/kg on second day).\nAll staff engaged in the trial were blinded as to the treatment group of each child, apart from the study physician who administered medication.\nObjectives\nOur primary objectives were to determine the effect of SP+AS and SP+AS+PQ on submicroscopic P. falciparum gametocyte prevalence and density and to determine the safety of a single dose of PQ in glucose-6-phosphate-dehydrogenase (G6PD) deficient children.\nOutcomes\nThe primary outcomes were gametocyte prevalence and density by real-time nucleic acid sequence-based amplification (QT-NASBA).\nThe secondary outcome was haemoglobin concentration following treatment.\nOther outcomes that we evaluated were microscopic gametocyte prevalence, treatment efficacy and the occurrence of side effects.\nParticipants were encouraged to attend the recruiting clinic at day 1, 2, 3, 7, 14, 28 and 42 after enrolment and at any time the child became unwell.\nOn each day of follow-up, tympanic temperature was measured by electric thermometer and a finger prick blood sample was used for haemoglobin (Hb) measurement using a Hemocue photometer (Angelholm, Sweden), a microscopic slide, a 50 \u00b5L-blood sample for real-time nucleic acid sequence-based amplification (QT-NASBA) and a filter paper sample.\nThe presence of symptoms suggestive of anaemia (fatigue, weakness, dizziness, headache, heart palpitations) or allergic drug reactions (rash) was assessed verbally during every follow-up visit.\nField assistants visited the homes of children who failed to show up to collect additional samples.\nBlood smears were stained for 10 minutes with 10% Giemsa and screened for asexual parasites and gametocytes at enrolment and on day 3, 7, 14, 28 and 42 after treatment.\nAll slides were double-read by experienced microscopists and were declared negative if no parasites were observed in 100 microscopic fields.\nReadings were compared for validation and slides giving discordant results were read by a third reader.\nThe majority result was taken as final in the case of positive versus negative results and geometric mean of the two closest values for density discordants .\nAsexual parasites and gametocytes were counted against 200 and 500 white blood cells, respectively and converted to parasites/\u00b5L by assuming a density of 8000 white blood cells/\u00b5L blood.\nP. falciparum parasite detection by QT-NASBA was performed as described previously.\nBriefly, nucleic acids were extracted from 50 \u00b5L-blood samples with initial RNA extraction carried out in the field following the original Guanidine isothiocyanate (GuSCN) RNA extraction method until the nucleic acids were bound to silica dioxide particles.\nAt this point, samples were stored at \u221220\u00b0C prior to completion of the extraction and QT-NASBA analysis.\nQT-NASBA was performed on a NucliSens EasyQ analyser (bioM\u00e9rieux, Boxtel, the Netherlands) for Pfs25 mRNA.\nThe Pfs25 QT-NASBA is gametocyte-specific and has a detection limit of 10\u2013100 gametocytes/mL.\nNuclisens Basic kits (bioM\u00e9rieux, Boxtel, the Netherlands) were used for amplification according to the manufacturers instructions.\nA standard dilution series of in vitro cultured mature NF54 gametocytes was included in each run to ascertain gametocyte density.\nMolecular Genotyping\nDNA extraction and PCR genotype analysis\nDNA extraction from bloodspots on filter paper was carried out by the chelex-100 method as described by Wooden et al. with some modifications described in Pearce et al.\n.\nIn brief, all samples were extracted in a 96-well plate format.\nThe bloodspot was first soaked in phosphate-buffered saline (PBS) with 0.5% saponin overnight and was then washed twice in 1 ml of PBS.\nThe samples were then boiled for 8 min in 100 \u00b5L of H2O and 50 \u00b5L of 20% chelex suspension in distilled water (pH 9.5) and centrifugated at 5000 rpm for 10 minutes.\n1 \u00b5L of supernatant was used in the PCR reactions.\nGenotyping for MSP-1, MSP-2 and glucose-6-phosphate-dehydrogenase (G6PD) deficiency\nTo differentiate between recrudescent parasites i.e. those persisting from the initial infection and parasites from a new infection, a nested PCR amplification of the polymorphic regions of P. falciparum genes msp1 (K1, MAD20 and RO33) and msp2 (IC1 and FC27) was performed as described by Snounou et al .\nThis PCR was performed for follow-up samples with microscopically confirmed parasitaemia.\nThe PCR products (10 \u00b5L) were run in electrophoresis on 2\u20132.5% metaphor agarose gels in 1xTBE buffer, stained with ethidium bromide and then visualized in UV trans-illumination.\nThe procedure of Cattamanchi et al. was followed in that indeterminate samples for which a majority of novel bands appeared for the post-treatment infection were scored as new infections.\nG6PD deficiency was determined by screening human DNA for single nucleotide polymorphisms in the G6PD gene (G202A, A376G) by a simple high throughput method using PCR, sequence specific oligonucleotide probes (SSOPs) and ELISA-based technology.\nIndividuals with no G202A mutation were classified G6PD B, heterozygotes for the G202A mutation were classified G6PD A and homozygote or hemizygote (males) for the G202A mutation were classified G6PD A-.\nSample size\nThe primary endpoint used for sample size calculation was Pfs25 QT-NASBA gametocyte prevalence after treatment.\nAssuming a gametocyte prevalence in the SP+AS group of 50% on day 14 after treatment, a sample size of 50 individuals per group would allow over a power of 80% to detect a reduction in gametocyte prevalence to 20% in the SP+AS+PQ group, allowing for 5% drop-out and using a significance level of 0.05.\nThis sample size also allowed us to detect a two-fold reduction in gametocyte prevalence during the entire period of follow-up in longitudinal data analyses, assuming an average Pfs25 QT-NASBA gametocyte prevalence of 58% in SP+AS treated children, and a maximum correlation between observations of the same individual of 0.30.\nRandomization\nThe randomization sequence was generated in Stata 8.0 (Stata Corporation, Texas, USA) using restricted randomization with a block size of 20.\nTreatment allocation was determined by opening pre-prepared randomization envelopes in sequence by the study physician.\nThe same physician was involved in participant selection and clinical evaluation.\nParasite carriage by microscopy and Pfs25 QT-NASBA and haemoglobin concentrations were determined by technicians who were unaware of the treatment allocated to study participants.\nStatistical methods\nTherapeutic outcome was classified as early parasitological treatment failure (ETF), late treatment failure (LTF), re-infection or adequate clinical and parasitological response (ACPR).\nHaemoglobin concentrations during follow-up were expressed as a percentage of the enrolment concentration.\nTo quantify the effect of treatment on gametocyte densities, we determined the area under the curve (AUC) of Pfs25 QT-NASBA gametocyte density versus time.\nThis measure incorporates both the magnitude and the duration of gametocyte carriage and was described by M\u00e9ndez et al.\n.\nThe AUC from days 0\u201342 was calculated as: AUC\u200a=\u200a[(3\u22120)\u00d7(g0+g3)/2+(7\u22123)\u00d7(g3+g7)/2+(14\u22127)\u00d7(g7+g14)/2+(28\u221214)\u00d7(g14+g28)/2+(42\u221228)\u00d7(g28+g42)/2]/42; where gd represents Pfs25 QT-NASBA gametocyte density on day d. Gametocyte negative samples were included as zeroes.\nThe measure was scaled by 42 so that it represents AUC per day and this was transformed by log10.\nMicroscopic and QT-NASBA parasite densities were analysed after log10-transformation.\nBecause we were interested in clearance of gametocytes of the original infection, slides from individuals in which PCR analysis defined a new infection were excluded from analyses on post-treatment gametocyte prevalence and density.\nProportions were compared using the chi-squared statistic for a 2-by-2 contingency table.\nNormally-distributed continuous variables were compared using the Student t-test.\nVariables that were not normally distributed were compared using the Wilcoxon rank-sum test.\nMultiple linear regression models were used in case of continuous variables to adjust for potential confounding factors such as asexual parasite and gametocyte density at enrolment.\nMultiple logistic regression models with Generalized Estimating Equations (GEE) were used to test the influence of treatment on dichotomous variables with multiple observations per participant, such as gametocyte prevalence during follow-up.\nEstimates were adjusted for potential confounding factors and a random effect was included in the models to allow for correlations within individuals.\nRegression coefficients (\u03b2) were calculated for continuous dependent variables and odds ratios (OR) for dichotomous dependent variables, both with 95% confidence intervals (95% CI).\nStatistical analyses were performed using SPSS for Windows 12.0 (SPSS Inc., Chicago, USA) and Stata 8.0 (Stata Corporation, Texas, USA).\nResults\nRecruitment and Participant Flow\nA total of 108 children were randomised over the two treatment arms (figure 1).\nThis number exceeds the original sample size (100) because of a high number of patients appearing on the last day of enrolment.\nTwo children (1.9%) were lost for evaluation during the 42-day follow-up period, one in each treatment arm.\nParasite density, gametocyte prevalence, haemoglobin concentrations, fever prevalence and the proportion of G6PD-deficient children at enrolment were not different between the treatment regimens (table 1).\nAdequate clinical and parasitological response (ACPR) on day 42 after treatment was observed in 71.7% (38/53) of the children treated with SP+AS and in 67.9% (36/53) of those treated with SP+AS+PQ (\u03c72\u200a=\u200a1.70; p\u200a=\u200a0.64)(table 2).\nOutcomes and Estimation: Gametocyte carriage after treatment\nMicroscopic gametocyte prevalence at enrolment was 26.4% (14/53) for children treated with SP+AS and 18.9% (10/53) for children treated with SP+AS+PQ (table 1).\nMicroscopic gametocyte prevalence decreased after treatment in both treatment arms (figure 2A) and was significantly lower in individuals treated with SP+AS+PQ compared to those treated with SP+AS when the entire period of follow-up was considered, after adjustment for enrolment gametocyte prevalence (GEE: OR\u200a=\u200a0.17 (95% CI\u200a=\u200a0.049\u20130.57), p\u200a=\u200a0.004).\nNo microscopic gametocyte carriage was seen on day 7 and 14 after treatment with SP+AS+PQ.\nEnrolment gametocyte prevalence defined by Pfs25 QT-NASBA was 88.2% (45/51) for SP+AS treated individuals compared to 90.6% (48/53) for SP+AS+PQ treated children (table 1).\nPredictably, children with microscopically confirmed gametocytes at enrolment had a significantly higher median Pfs25 QT-NASBA gametocyte density (46.6 gametocytes/\u00b5L; IQR 8.9\u2013346.5) compared to those gametocyte-free by microscopy (7.5 gametocytes/\u00b5L; IQR 3.1\u201367.5)(Wilcoxon Rank-Sum test z\u200a=\u200a\u22122,01, p\u200a=\u200a0.04).\nPfs25 QT-NASBA gametocyte density was not initially defined as outcome measure, but densities were compared between the treatment arms post-hoc.\nAt the time of enrolment Pfs25 QT-NASBA gametocyte density was 28.8 gametocytes/\u00b5L (IQR 6.9\u2013109.9 gametocytes/\u00b5L) for SP+AS treated children compared to 17.5 gametocytes/\u00b5L (IQR 1.1\u201376.9) for SP+AS+PQ treated children (table 1).\nDuring follow-up Pfs25 QT-NASBA gametocyte prevalence decreased for both SP+AS and SP+AS+PQ treated children (figure 2B and table 3).\nAfter adjustment for enrolment Pfs25 QT-NASBA gametocyte density, Pfs25 QT-NASBA gametocyte prevalence was lower in individuals treated with SP+AS+PQ during the entire period of follow-up (GEE: OR\u200a=\u200a0.27 (95% CI\u200a=\u200a0.18\u20130.40), p<0.001).\nThe Pfs25 QT-NASBA gametocyte density in gametocyte positive samples was consistently lower for SP+AS+PQ treated children during follow-up (table 3).\nThe average duration of gametocyte carriage was summarised in the area under the curve of Pfs25 QT-NASBA gametocyte density versus time (table 4).\nThe area under the curve was significantly lower for SP+AS+PQ treated children (t-test t\u200a=\u200a3.28, p <0.001), as were the number of sampling times when gametocytes were detected (GEE: OR\u200a=\u200a0.27 (95% CI\u200a=\u200a0.18\u20130.40), p<0.001) and the geometric mean gametocyte density in gametocyte positive samples (GEE: \u03b2\u200a=\u200a\u22120.40 (95% CI\u200a=\u200a\u22120.74\u2212\u22120.07), p\u200a=\u200a0.02).\nMost treatment failures or re-infections appeared after day 14 (table 2).\nIt is therefore appropriate to focus on the first two weeks after treatment to determine the effect of treatment in the absence of treatment failure or re-infection.\nOn day 14 after SP+AS treatment, 32 (62.7%) individuals were gametocyte positive with a mean gametocyte density of 4.5 gametocytes per \u00b5L (IQR 1.8\u201337.1)(table 3).\nOnly two individuals (3.9%) were positive by Pfs25 QT-NASBA on day 14 after treatment with SP+AS+PQ with low gametocyte densities of 0.067 and 0.094 gametocytes per \u00b5L.\nHaemoglobin concentrations after treatment\nTo address concerns about haemolysis associated with PQ use in G6PD deficient individuals, Hb concentration was assessed at enrolment and during follow-up.\nMedian Hb at enrolment was 10.5 g/dL (IQR 9.4\u201311.8) for SP+AS treated children and 10.8 g/dL (9.8\u201311.8) for SP+AS+PQ treated children (table 1).\nWhile Hb relative to enrolment concentration gradually increased in the weeks after treatment with SP+AS, it first decreased for SP+AS+PQ treatment and remained lower up to day 14 (figure 3).\nThe relative decrease was most pronounced on day 7 after SP+AS+PQ treatment when Hb concentration was 5.2% lower (95% CI \u22128.4\u2013\u22121.8) than on enrolment (paired t-test: t\u200a=\u200a2.86; p\u200a=\u200a0.006).\nWhen only children with the G6PD B variant were considered, the relative decrease in Hb concentration on day 7 after treatment with SP+AS+PQ was no longer statistically significant (paired t-test: t\u200a=\u200a1.57; p\u200a=\u200a0.12).\nThe reduction in Hb shortly after SP+AS+PQ treatment was most pronounced in children with the G6PD A- variant (figure 4) although numbers were too small to allow statistical comparisons.\nNone of the children developed clinical symptoms related to anaemia or an Hb below 5g/dL.\nThe Hb concentrations on day 28 and day 42 after treatment were equal to or greater than that at enrolment for all G6PD categories (figure 4).\nEight children experienced \u226520% reduction in Hb concentration on day 7 relative to that at enrolment, compared to none in the SP+AS treatment arm.\nThe range of Hb concentration in these eight individuals was 5.4\u201310.4 g/dL on day 7 after SP+AS+PQ treatment.\nTwo of the children with \u226520% reduction in Hb concentration on day 7 had the G6PD A- variant (25%), one had the G6PD A variant (12.5%) and the other five had G6PD B variant (62.5%).\nNone of the children reported symptoms suggestive of anaemia or allergic drug reactions during follow-up.\nDiscussion\nInterpretation\nThis study shows that a single dose of primaquine (PQ) is of significant additive value in clearing gametocytes in an area of high malaria endemicity in Tanzania.\nOnly 3.9% of children treated with SP+AS+PQ had gametocytes on day 14 after successful treatment compared to 62.7% for SP+AS treated children.\nGametocytes persisting on day 14 after SP+AS+PQ treatment circulated at densities well below the theoretical threshold for mosquito infection.\nGametocyte carriage after treatment with SP+AS persisted for more than one month and more than two-thirds of the treated individuals harboured gametocytes on day 14.\nThese data confirm previous findings.\nThe addition of a single dose of PQ to this regime significantly reduced gametocyte carriage.\nWhen microscopy was used as tool to detect gametocytes, no gametocytes were observed seven days after the initiation of treatment with SP+AS+PQ.\nOn day 28 after treatment, gametocytes were detected in only one individual.\nSubmicroscopic gametocyte prevalence decreased to 3.9% (2/51) on day 14 after initiation of treatment but increased at subsequent follow up time points.\nThis increase is most likely due to recrudescence of infection or new infections that were undetected by microscopy.\nTreatment failure rates were similar to those reported in recent drug efficacy studies in east Africa and re-infection rates were high in this area of intense malaria transmission.\nAs a consequence, 28.3% of the SP+AS treated children and 32.1% of the SP+AS+PQ treated children experienced microscopically confirmed treatment failure or re-infection during follow-up.\nTrue treatment failure and re-infection rates may be higher as not all asexual parasites will be detected by microscopy.\nA high persistence of submicroscopic asexual parasites after apparently successful drug treatment has been described, as well as the acquisition of new sub-patent infections.\nGametocytes that newly developed from new or persisting asexual infections are therefore likely to be responsible for the increase in gametocyte prevalence after day 14 that has been reported previously.\nFocusing on the first two weeks after initiation of treatment, PQ seems to decrease the number of gametocytes rapidly to a level where onward transmission may be arrested completely.\nAlthough analyses on gametocyte densities should be considered as post-hoc analyses and we did not directly determine post-treatment malaria transmission, it is clear that the transmission potential is reduced in children treated with SP+AS+PQ.\nGametocyte prevalence and density were lower in children treated with SP+AS+PQ and gametocytes persisting on day 14 after treatment circulated at densities below 0.1 gametocyte/\u00b5L.\nWhen assuming an average mosquito blood meal size of 2\u20133 \u00b5L, these concentrations are unlikely to result in mosquito infection, although infectiousness of very low gametocyte densities can not be excluded.\nGeneralisability\nThe rationale for using PQ in P. falciparum infections is to reduce post-treatment infectivity (as routinely practiced in several countries in Asia and the Americas) and as a quarantine treatment to reduce the spread of (drug resistant) parasites.\nThe aim of the current study was not to determine if PQ should be routinely added to current ACT regimens.\nThe objective was to explore if PQ is a valuable component in mass drug administration studies that aim to reduce malaria transmission.\nMass drug administration (MDA) studies are typically conducted in areas of low malaria transmission intensity and also include asymptomatic parasite carriers.\nThe likelihood of re-infection is lower in these areas and the effect of PQ can be expected to be larger because enrolment gametocyte densities are lower.\nIn MDA studies, drugs are administered to asymptomatic individuals, raising safety concerns over the haemolytic effect of PQ on individuals with G6PD mutations.\nDue to the potential protective effect of G6PD-deficiency from malaria, G6PD deficiency is probably selected in malaria endemic regions, in a similar manner as for other haemoglobinopathies.\nIn our population, 6.5% (7/107) of the children had the A- variant of G6PD deficiency.\nWe observed that the addition of PQ resulted in a statistically significant but transient reduction in haemoglobin levels, as was previously reported for PQ administered at a curative dose (0.5 mg/kg, 14 days) to individuals with the African Variant (A-) of G6PD-deficiency.\nThe observed reduction in Hb concentration indicates a small but genuine risk of PQ use in individuals who are anaemic prior to drug administration.\nThe risk to the individual patient has to be weighted against the potential benefit of a reduced malaria transmission.\nIn the current study, we excluded individuals with an Hb<8 g/dL.\nThe risk of anaemia is likely to be increased in individuals with lower pre-treatment Hb concentrations, which is important since Hb is not routinely determined prior to treatment.\nIn case of mass-administration of SP+AS+PQ, anaemic individuals should preferably be identified prior to treatment and excluded from PQ treatment.\nThis can be done by Hemocue\u00ae which allows a rapid and reliable assessment of anaemia in the field.\nAreas of low and seasonal malaria transmission are those that are most likely to benefit from MDA.\nIn these areas, the prevalence of severe anaemia prior to the malaria season is likely to be low.\nIn these circumstances we consider the addition of a single dose of PQ to SP+AS to be a safe approach to reduce post-treatment malaria transmission.\nOverall evidence\nThis is the first study that determines the effect of PQ on submicroscopic gametocyte densities and our findings are in line with previous studies that where PQ efficiently cleared microscopic gametocyte concentrations.\nThe addition of a single dose of PQ had no beneficial influence on the clearance of asexual parasites, as was described previously.\nProfile of the study\nGametocyte prevalence by microscopy (A) and Pfs25 QT-NASBA (B).Gametocyte prevalence for SP+AS (closed diamonds, solid line) and SP+AS+PQ (open triangles, broken lines) treated children. Bars indicate the 95% confidence intervals around the proportions. * indicates a statistically significant difference between the two treatment arms.\nHaemoglobin concentration following treatment.Concentrations are expressed relative to that at enrolment for SP+AS (closed diamonds, solid line) and SP+AS+PQ (open triangles, broken lines). Bars indicate the 95% confidence intervals around the proportions. * indicates a statistically significant difference between the two treatment arms.\nRelative haemoglobin concentration after treatment with SP+AS+PQ for different G6PD genotypes.Haemoglobin concentration relative to enrolment for children without the G202A mutation (G6PD genotype B; black diamonds, n\u200a=\u200a39), heterozygotes (G6PD genotype A; white triangles, n\u200a=\u200a9) and homozygotes or hemizygotes (G6PD genotype A-: grey diamonds, n\u200a=\u200a4). Each individual measurement is shown; lines indicate the median value.\n\nCharacteristics of the study population at enrolment\n | SP+AS | SP+AS+PQ\nN | 54 | 54\nAge, median (IQR) | 5 (3\u20138) | 5.5 (3\u201310)\nSex, % male (n/N) | 46.3 (25/54) | 55.6 (30/54)\nTemperature, % fever (>37.5\u00b0C) (n/N) | 30.2 (16/53) | 35.8 (19/53)\nAsexual parasite density GM/\u00b5L (IQR) | 4,759 (959\u201322,248) | 7,379 (2,334\u201325,304)\nMicroscopic gametocyte prevalence, % (n/N) | 26.4 (14/53) | 18.9 (10/53)\nPfs25 QT-NASBA gametocyte prevalence, % (n/N) | 88.2 (45/51) | 90.6 (48/53)\nPfs25 QT-NASBA gametocyte density, GM/\u00b5L (IQR) * | 28.8 (6.9\u2013109.9) | 17.5 (1.1\u201376.9)\nG6PD, % (n/N)\nB | 68.5 (37/54) | 75.5 (40/53)\nA | 25.9 (14/54) | 17.0 (9/53)\nA- | 5.6 (3/54) | 7.5 (4/53)\nHaemoglobin concentration, g/dL, median (IQR) | 10.5 (9.4\u201311.8) | 10.8 (9.8\u201311.8)\n\nIQR\u200a=\u200ainterquartile range; GM\u200a=\u200ageometric mean; G6PD: B\u200a=\u200ano G202A mutation; A\u200a=\u200aheterozygotes, single G202A mutation; A-\u200a=\u200ahomozygote or hemizygote (males), only G202A mutations. *For gametocyte carriers only.\n\nTreatment outcome for the different treatment regimens on day 14, 28 and 42.\n | SP+AS | SP+AS+PQ\nNumber evaluated | 53 | 53\nDay 14 Treatment outcome, % (n)\nACPR | 98.1 (52) | 100.0 (53)\nETF | 0 (0) | 0\nLTF | 0 (0) | 0\nRe-infection | 1.9 (1) | 0\nIndeterminate | 0 (0) | 0\nDay 28 Treatment outcome, % (n)\nACPR | 77.4 (41) | 83.0 (44)\nETF | 0 (0) | 0 (0)\nLTF | 3.8 (2) | 5.7 (3)\nRe-infection | 15.1 (8) | 9.4 (5)\nIndeterminate | 3.8 (2) | 1.9 (1)\nDay 42 Treatment outcome, % (n)\nACPR | 71.7 (38) | 67.9 (36)\nETF | 0 (0) | 0 (0)\nLTF | 3.8 (2) | 9.4 (5)\nRe-infection | 17.0 (9) | 13.2 (7)\nIndeterminate | 7.5 (4) | 9.4 (5)\n\nACPR\u200a=\u200aadequate clinical and parasitological response, ETF\u200a=\u200aearly treatment failure, LTF\u200a=\u200alate treatment failure. Indeterminate\u200a=\u200aindeterminate due to PCR-failure (n\u200a=\u200a1) or missing filter paper DNA samples (n\u200a=\u200a8).\n\n\nPfs25 QT-NASBA gametocyte prevalence and density after treatment with SP+AS and SP+AS+PQ.\n | SP+AS | SP+AS+PQ\n | Gametocyte prevalence, % (n/N) | Gametocyte density/\u00b5L, GM (IQR) | Gametocyte prevalence, % (n/N) | Gametocyte density/\u00b5L, GM (IQR)\nDay 0 | 88.2% (45/51) | 28.8 (6.9\u2013109.9) | 90.6% (48/53) | 17.5 (1.1\u201376.9)\nDay 3 | 80.8% (42/52) | 25.5 (4.2\u2013132.9) | 53.8% (28/52) | 2.9 (0.6\u201336.6)\nDay 7 | 71.7% (38/53) | 8.0 (1.8\u201359.1) | 15.7% (8/51) | 0.3 (0.02\u20132.4)\nDay 14 | 62.7% (32/51) | 4.5 (1.8\u201337.1) | 3.9% (2/51) | 0.07&0.09\nDay 28 | 41.9% (18/43) | 10.9 (1.0\u201370.0) | 10.4% (4/46) | 3.8 (0.2\u20132.4)\nDay 42 | 28.2% (11/39) | 17.4 (1.2\u2013191.4) | 16.3% (5/40) | 7.9 (0.5\u201354.3)\n\nGM\u200a=\u200ageometric mean Pfs25 QT-NASBA gametocyte density per microlitre for gametocyte carriers only; IQR\u200a=\u200ainterquartile range\n\nEffect of treatment on gametocyte carriage during follow-up\n | SP+AS | SP+AS+PQ | p-value\nMean AUC of gametocyte density/\u00b5L versus time, (IQR) | 11.1 (2.2\u201353.8) | 1.5 (0.3\u20138.8) | <0.001*\nNumber of sampling times when gametocytes were detected, % (n/N) | 64.4 (186/289) | 32.4 (95/293) | <0.001\u2020\nGM gametocyte density/\u00b5L on days when gametocytes were detected, (IQR) | 15.8 (4.1\u201385.8) | 5.8 (0.8\u201355.1) | 0.02\u2020\n\nAUC\u200a=\u200aarea under the curve; GM\u200a=\u200ageometric mean; IQR\u200a=\u200ainterquartile range\nAdjusted for gametocyte density at enrolment; \u2020 adjusted for correlations between observations from the same individual",
        "label": "unclear",
        "id": "task4_RLD_test_951"
    },
    {
        "paper_doi": "10.1186/s40608-014-0021-5",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Setting: university conference room, USADesign: randomised controlled trialRecruitment: on a university campus through announcements in lectures and through electronic bulletin boardsAllocation to groups: randomised using a random number generator\n\n\nParticipants: Overweight or obese women (n = 62)Mean age: 21.87 (SD 3.03), range 18-33Ethnicity: 45.16% Hispanic or Latino; 27.42% Black/African American; 4.84% Caribbean non-Hispanics; 8.06% Asian/Pacific Islander; 3.23% white Non-Hispanic; 9.68% mixed race; 1.61% don't know/not sureMean BMI: 28.42 kg/m2 (SD 3.10)Education: 82% of all participants had a high school degree/equivalency, some college or a 2-year college degree\n\n\nInterventions: Intervention 1: menu with energy (kcal) information (n = 20)Intervention 2: menu with energy (kcal) information and with exercise equivalents (n = 20)Control: menu with no energy information (n = 22)\n\n\nOutcomes: Energy (kcal) consumption during a meal; measured by weighing leftover food\n\n\nNotes: Food offered was fast food from Burger King. Participants attended twice for baseline and intervention meal.Subgroup analysis was conducted by the study author for restrained vs unrestrained eaters. A repeated measures analysis of variance was conducted. Data were combined from intervention 1 and 2 and compared with the control. This study was a thesis dissertatio\n\n",
        "objective": "To assess the impact of nutritional labelling for food and non\u2010alcoholic drinks on purchasing and consumption of healthier items. Our secondary objective was to explore possible effect moderators of nutritional labelling on purchasing and consumption.",
        "full_paper": "Background\nBetter techniques are needed to help consumers make lower calorie food choices.\nThis pilot study examined the effect of menu labeling with caloric information and exercise equivalents (EE) on food selection.\nParticipants, 62 females, ages 18-34, recruited for this study, ordered a fast food meal with menus that contained the names of the food (Lunch 1 (L1), control meal).\nOne week later (Lunch 2 (L2), experiment meal), participants ordered a meal from one of three menus with the same items as the previous week: no calorie information, calorie information only, or calorie information and EE.\nResults\nThere were no absolute differences between groups in calories ordered from L1 to L2.\nHowever, it is noteworthy that calorie only and calorie plus exercise equivalents ordered about 16% (206\u00a0kcal) and 14% (162\u00a0kcal) fewer calories from Lunch 1 to Lunch 2, respectively; whereas, the no information group ordered only 2% (25\u00a0kcal) fewer.\nConclusions\nMenu labeling alone may be insufficient to reduce calories; however, further research is needed in finding the most effective ways of presenting the menu labels for general public.\nBackground\nPoint-of-purchase menu labeling, particularly at fast food restaurants, has been of special interest in the fight against obesity.\nAs fast food consumption has been correlated with obesity and other negative health outcomes, these food outlets are being targeted for change.\nIn 2010, the US federal health care reform bill was signed into law and includes a requirement that restaurant chains with at least 20 outlets nationwide post calorie labels on menu boards.\nIt has been suggested that knowledge of the calories contained in foods is essential to choosing and consuming an energy-balanced diet.\nWhile consumer polls show a desire for calorie information at the point-of-purchase in restaurants, research on the actual effects has shown mixed results.\nOne possible reason for inconsistent effectiveness may be the lack of understanding of the value of a calorie or the lack of a reference amount for a calorie.\nExercise equivalents have been discussed by nutrition experts as a potential method to inform consumers about calorie values.\nExercise equivalents are defined as the amount of time doing particular physical activities that would be needed to burn off calories in foods.\nFor example, burning off a 300-calorie hamburger would require about 75\u00a0minutes of walking, after expending the calories needed for daily subsistence.\nExercise equivalents could potentially simplify food and/or restaurant nutrition labels, increase understanding of calories and of energy imbalance, and facilitate a decrease in overall energy intake.\nLiterature exploring the use of exercise equivalents is limited.\nA new study by Dowray et al., explored the potential effect of exercise equivalents on menu labels.\nThis study was a web-based survey, and asked participants to \u201cimagine they are in a fast food restaurant\u201d, and order a meal from an online menu.\nParticipants were randomly assigned to see one of four menus (calories only, calories and number of minutes to walk to burn off that amount of calories, calories and number of miles to walk off that amount of calories, or no information).\nThe results from this study are significant; calories were significantly different based on menu type (p\u2009=\u20090.02), with the calories and exercise equivalents in mileage group ordering significantly less calories than the other three groups (p\u2009=\u20090.0007).\nAdditionally, 82% of their participants reported a preference for exercise equivalents on menu labeling.\nThis study shows a positive impact in using exercise equivalents to aid in the understanding of a calorie, and potential to help consumers order lower calorie food items.\nThese findings are consistent with an earlier study, which showed that exercise equivalents helped reduce purchases of sugar-sweetened beverages among low-income black adolescents.\nA study by Fitch et al. assessed the influence of calorie information versus exercise equivalents on food selection amongst adolescents and adults who ate at fast food restaurants regularly.\nThis study indicated that calorie labels were preferred to exercise equivalents overall (71%), and some cited the latter as demotivating; however, the Fitch study has limitations.\nThey tested the impression of exercise equivalents rather than their actual effect, examined exercise equivalents as an alternative to, rather than addition to calories, and had a predominantly white sample population, many of whom were not overweight or obese.\nThey also found that exercise equivalents had a more favorable impression among non-whites than whites, and, along with another study, among younger persons.\nThe current study aimed to test the actual effect of exercise equivalents on fast food point-of-purchase behaviors.\nResearch has shown that non-white, overweight and obese individuals are more likely to consume fast food and thus be at increased risk of negative health outcomes.\nFor this reason, we recruited young, predominantly non-white overweight and obese women for our study and presented them with exercise equivalents alongside calories at the point-of-choice.\nWe compared the effect of the exercise equivalents with the provision of simple caloric information or no information at all.\nAdditionally, we sought to evaluate the impact of restrained eating on point of purchase and consumption behaviors.\nThe current study was a pilot.\nWhile the researchers acknowledge the small sample size as a limitation, the goal of the study was to test a new design: the potential use of exercise equivalents for public health outreach, with hopes that other researchers can utilize for future studies on this topic.\nThinking of new ways to promote healthy behaviors at the point-of-purchase is important for nutrition researchers, educators, and policymakers.\nWe believe the novelty of our experimental design, with emphasis on using exercise equivalents as a nutrition intervention for at risk individuals, furthers thought on point-of-purchase interventions.\nMethods\nStudy overview\nA three-group repeated-measures experimental study was conducted to determine whether providing information about calories and exercise equivalents at the point-of-choice for a fast food meal would decrease calories ordered or consumed among overweight and obese 18-34-year-old women at a public university in southern Florida in 2009, and to investigate any correlation with consumption with prior dieting history, qualified as dietary restraint in this study.\nAll participants were asked to participate in two sessions during a two-week period.\nThe Florida International University Institutional Review Board approved this study.\nAll persons gave their informed consent prior to inclusion in the study.\nStudy participants\nA total of 62 overweight or obese female participants were recruited on a south Florida college campus.\nTelephone and in-person screening determined whether participants met the inclusion criteria: female, age 18-34 years old, BMI at least 25 and less than 40, as calculated from researcher-measured height and weight, ate fast food at least \u201coccasionally\u201d, and able to read and speak English.\nParticipants were also screened at this time for dietary restraint for randomization into the three study groups.\nPersons were excluded for dieting in the last three months; requiring a special diet such as vegetarian, kosher, or accommodating a food allergy or health condition; being pregnant or giving birth in last year; having a chronic disease such as heart disease or diabetes; having current self-reported depression, self-reported alcohol or drug abuse, or eating disorder; being a health major; not typically eating lunch; and participating in a previous food-related study.\nExclusion criteria were set to ensure participants were healthy and able to partake in a food-related study and to help avoid any bias gained from previous food-related studies.\nIn order to help further blind participants to the menu manipulation aspect of the study, participants were told that the purpose of the study was to \u201cbetter understand fast food meal choices\u201d.\nExperimental design\nParticipants attended two meal sessions, Lunch 1 and one week later Lunch 2.\nThe food choices were from a fast food restaurant located on the university campus.\nThe restaurant is part of a national chain specializing in hamburgers and French fries.\nThe foods were in their original portion-controlled wrappers or packaging, which allowed the researcher to easily record choices made by participants.\nThe study took place in a controlled setting within the university, at a private conference room in the University\u2019s Graham Center nearby the student union where the students normally eat.\nIncentives for the participants included $5 for completion of the screening questions, the two free lunches, and a $20 gift card at each lunch.\nAt the start of each Lunch, participants were given a menu.\nThe paper menus were in a similar format to menu boards at fast food restaurants.\nThe food items were those available for lunch at Burger King on the dates of the experiment.\nThe participants were able to choose entr\u00e9es (e.g. Hamburger, Whopper, TenderGrill, BK Veggie Burger or TenderGrill), a garden salad, side dishes (i.e., fries, onion rings), condiments (ketchup, mayonnaise, fat free ranch dressing, or honey mustard dressing) and a drink (i.e., water, Coca-Cola, diet Coca-Cola, or apple juice).\nThe observer recorded the quantity of the food ordered and eaten by using a digital food scale, weighing the remaining portions and using a measuring cup for the liquids.\nThe researcher ensured that all participants had finished eating and had left the study site prior to weighing and measuring left-over foods and drinks.\nCalories consumed were derived by taking food waste and weighing on a digital scale and calculating total calories eaten by the following formula: Total Calories For Food Item Chosen\u2013Food Waste = Calories Consumed.\nLunches were served from 11:30\u00a0a.m. until 3:00\u00a0p.m., and participants made appointments at 30-minute increments.\nAll participants were told in advance that they would not be able to leave the study site with any leftover food, to limit the possibility that participants would order more food than they intended to consume.\nDuring Lunch 1, participants were given a menu, similar in format to menu boards at fast food restaurants.\nParticipants were able to order any foods and beverages from the menu, which listed only names of items, no calories or exercise equivalents.\nThe experimental manipulation took place one week later at Lunch 2.\nAll participants were randomly assigned to one of three groups.\nEach group received different information on their menus: no information on calorie or exercise equivalents, calories only, or calories and exercise equivalents.\nColumn headers for the exercise equivalents and calories described the numbers (\u201cminutes to burn off food in walking\u201d, \u201ccalories\u201d), as did labels after each values.\nThe exercise equivalence of calories was calculated based on an intensity level of 3.3 METs for walking at the moderate pace of 3.0 mph on a firm surface, and a body weight of 160 pounds.\nData collection\nAt the research table, participants completed standardized questions on the following demographic information: age, marital status, education, income, race, religion, and whether the participant was a smoker.\nBody Mass Index was assessed by the investigators at the research table using a standardized height and weight measurement procedure as outlined in Third National Health and Nutrition Examination Survey (NHANES III) Anthropometric Procedures Manual.\nDietary restraint was determined using the TFEQ.\nScores on the TFEQ restraint sub-scale range from 0 to 21, with restrained eaters defined as those who have a score of 13 or above.\nParticipants were blocked by restraint in order to test whether unrestrained and restrained eaters responded differently, since restraint has been shown to influence food choice and the reading of nutrition labels, and were then randomly assigned to one of three study groups.\nStatistical analysis\nANOVA and chi-square tests were conducted to compare demographic information by study group.\nBoth the foods ordered and the foods consumed were analyzed.\nWithin each study group, a paired t-test was conducted to test for the change in calories ordered or consumed from Lunch 1 to Lunch 2.\nThe subsequent change by study group was calculated as the mean (plus or minus the standard error) of the changes from Lunch 1 to Lunch 2 for each of the group\u2019s individual members.\nProportionate change for calories order and calories consumption from Lunch 1 to Lunch 2 were calculated as the mean (plus or minus the standard error) of the proportionate changes of each of the group\u2019s individual members.\nThe t-tests were used to examine whether the proportionate changes are significant.\nAnalysis of covariance (ANCOVA) was conducted using General Linear Model in SPSS 17.0 statistical software (SPSS Inc., 2008) to test for differences between study groups in difference from Lunch 1 to Lunch 2 in calories ordered or consumed.\nThe effect size (partial eta squared) and observed power (using alpha\u2009=\u20090.05) are also calculated using SPSS.\nTwo general linear models were created, both controlling for age, BMI, and dietary restraint, and with study group as the fixed factor.\nFor model 1, the response variable was the difference from Lunch 1 to Lunch 2 in total calories ordered, and an additional covariate was calories ordered in Lunch 1 (control meal).\nFor model 2, the response variable was the difference from Lunch 1 to Lunch 2 in total calories consumed, and an additional covariate was calories consumed in Lunch 1.\nBoth models assumed that the response variables were continuous, residuals were normally distributed, and the subjects were independent.\nTotal number of items ordered was also compared by study group.\nResults\nStudy participants had a mean age of 21.9\u2009\u00b1\u20093.03\u00a0years and BMI of 28.4\u2009\u00b1\u20093.10.\nSeventy-three percent (n\u2009=\u200945) were black or Hispanic and 63% (n\u2009=\u200939) were unrestrained eaters (Table\u00a01).\nDemographic information was comparable across study groups (all p values\u2009>\u20090.05) (Table\u00a01).\nAll study groups decreased the number of calories ordered from Lunch 1 to Lunch 2 (Table\u00a02).\nWhile the current study is under-powered to ascertain statistical non-significance or significance, calorie only and calorie plus exercise equivalents ordered about 16% (206\u00a0kcal) and 14% (162\u00a0kcal) fewer calories from Lunch 1 to Lunch 2, respectively; whereas, the no information group ordered only 2% (25\u00a0kcal) fewer (Table\u00a02).\nIn all study groups, both restrained and unrestrained eaters had an average decrease in number of calories ordered from Lunch 1 to Lunch 2, with the exception of restrained eaters in the group receiving no calorie or exercise equivalent information at Lunch 2.\nThe greatest proportionate decrease in calories ordered was seen in restrained eaters in the calories-only group (24.7% decrease; p\u2009=\u20090.05).\nUnrestrained eaters in the calories and exercise equivalents group ordered an average of 275 fewer calories at Lunch 2 compared with Lunch 1, with a proportionate decrease of 14.0%, although this was not statistically significant (p\u2009=\u20090.24).\nUnrestrained eaters in the calories and exercise equivalents group had greater absolute and proportionate decreases in calories ordered from Lunch 1 to Lunch 2 than unrestrained eaters in the other two study groups.\nAdditional analyses examining number of items ordered revealed no significant differences by study group (data not shown).\nDuring the exit questionnaire, 57 participants (92%) said they believed that a combination of calories and exercise equivalents would influence the foods they ordered at a fast food restaurant.\nDiscussion\nConsumption of fast foods is common in the US.\nTo reduce negative effects and mitigate public health disparities in food environments, interventions may be especially critical in populations of persons who eat at fast food restaurants.\nCalorie information presented at point-of-purchase is a relatively new concept nationwide; however, research has shown mixed results at the point of purchase.\nThere is a potential for exercise equivalents as a supplemental guide to novice calorie counters and those unaware of the negative health implications in consuming fast food.\nWhile the current study was underpowered, we believe the novelty of the design of the experiment, the emphasis on utilizing exercise equivalents for a potential nutrition intervention for at risk individuals, will heighten awareness for future researchers on the need for further investigating point of purchase interventions, specifically exercise equivalents.\nThere have been several real-world studies that have shown an impact on calories ordered using sales data.\nFor example, using a randomization design, Roberto et al. reported that calorie information on restaurant menu did reduce the total amount of calories people ordered and consumed.\nAnother quasi-experimental design study examined the sales data before and after provision of point-of-selection nutrition labels found that the nutrition labels reduced average energy content of entr\u00e9e purchased without reducing overall sales.\nAdditionally, using data from Starbucks, Bollinger et al. found that mandatory calorie posting in chain restaurants resulted in 6% decrease in calories per transaction.\nDumanovsky et al. conducted a cross-sectional survey and assessed consumer purchases in 2007, before caloric information was mandated by chain restaurants, and again in 2009, after the menu labeling legislation was passed.\nAlthough they did not find an overall change in calories consumed, they did observe a significant decrease in the calories consumed at specific chain restaurants including McDonald\u2019s, Au Bon Pain and KFC.\nWith the rollout of the new law mandating fast food restaurants list caloric value for all menu items pending, understanding the potential implications is important.\nCalorie information at the point-of-purchase for restaurants has been required by law for chain restaurants in New York City since 2008, in California, Oregon and Maine since 2009 and has also been adopted in many other cities and counties.\nA recent study by Krieger et al. is one of the first to investigate the effect of the nationwide menu labeling bill.\nThis cross-sectional study surveyed fast food patrons both before the menu label regulation was implemented, and again 18\u00a0months later, post-regulation.\nInterestingly, they found a significant decrease in calories ordered in coffee and taco establishments, but not in burger and sandwich shops; and a decrease in calories ordered by women, but not men.\nThe effectiveness of nutrition labels on point-of-choice food purchasing has provided mixed results.\nSimilar to the present study, prior studies that have looked at point-of-purchase at fast food restaurants and the other at nutrition labels have also failed to show statistical significance.\nThis study provided a real-world setting was created to measure actual point-of-purchase behavior.\nThe unique strength of this study is that the study design provides a potential alternative or addition to the soon-to-be implemented national menu labeling law as a public health intervention.\nThis study illustrates a novel design to test the effectiveness of adding exercise equivalents to provide a frame of reference for consumers.\nUsing exercise equivalents on food labels and food served away from home could provide consumers with a context for the term, \u201ccalorie\u201d, and, thus, contribute to the understanding of the nutrition labels for better food choice and selection.\nPresentation of caloric information of fast food translated into exercise equivalents did not have a statistically significant impact on the food choices of overweight and obese women who were restrained eaters or unrestrained eaters.\nHowever, unrestrained eaters presented with calorie information and exercise equivalents combined had a larger decrease in calories ordered compared to those with calorie information only and for those with no information.\nThe impact of calorie information with exercise equivalents on unrestrained eaters should be further examined as unrestrained eaters generally do not deliberately attempt to limit their food intake.\nThere are several limitations in this exploratory study.\nThe small study sample size is a major limitation of this study.\nAdditionally, the study was limited to female college students thus limiting its generalizability.\nAnother limitation is that individuals were getting food at no cost, which might have influenced the total number of food items, and hence amount of calories chosen.\nThe average calories for the foods chosen for Lunch 1 and Lunch 2 were 1215.16 and 1087.50, respectively.\nDumanovsky and colleagues (2009) established baseline data on mean calorie intake at Burger King of 926.2.\nParticipants in the current study chose approximately 225 more calories per meal on average than participants in Dumanovsky study, perhaps because the food was free.\nThis study did not collect pre- and post-intervention food diaries.\nIt is possible that participants who chose lower calorie foods during the intervention may have increased their intake later in the day to compensate.\nAdditionally, there is a potential limitation in using exercise equivalents, or calories alone, to promote lower calorie food choices as opposed to nutritionally dense options.\nLower calorie foods do not necessarily make a food nutritionally \u201cbetter\u201d than another.\nDespite this, although nutritionally and calorically dense foods such as tree nuts, avocado, and fatty fish are touted as an important part of a healthful diet, they are not commonly offered at fast food restaurants.\nAs such, exercise equivalents and calories listed can potentially provide meaningful reference points for fast-food patrons.\nConclusions\nThe current study presented an intervention designed to improve the effectiveness of calorie information on point-of-choice.\nThe concept of menu labeling, exercise equivalents and other point-of-purchase messages are a potentially useful way to reach consumers at the point of their food decision.\nThis research, combined with previous studies, suggest that in addition to calorie labeling information, there is a need for further research of point-of-purchase interventions to find the most effective ways of presenting the menu labels for general public.\nAlthough this study was not powered to see statistical differences, the concept that behaviors may differ based on calorie and exercise information should be further explored in a larger study.\n\n\nParticipant characteristics by study group, in a group of overweight or obese women\n\n | Study group for menu type at Lunch 2 (experiment meal)\nNo calorie or exercise equivalent information | Calories only | Calories and exercise equivalents | p1\nTotal | N | % | N | % | N | % | \n22 | 100% | 20 | 100% | 20 | 100% | \nAge (years; mean, SD) | 21.9\u2009\u00b1\u20093.5 |  | 21.6\u2009\u00b1\u20092.3 |  | 22.2\u2009\u00b1\u20093.2 |  | 0.82\nWeight (pounds; mean, SD) | 167.9\u2009\u00b1\u200926.5 |  | 171.2\u2009\u00b1\u200926.6 |  | 165.6\u2009\u00b1\u200925.8 |  | 0.79\nBMI (kg/cm2; mean, SD) | 27.9\u2009\u00b1\u20093.1 |  | 28.7\u2009\u00b1\u20093.0 |  | 28.7\u2009\u00b1\u20093.3 |  | 0.64\nRace/Ethnicity (N) |  |  |  |  |  |  | 0.90\nHispanic/Latino | 8 | 36% | 10 | 50% | 10 | 50% | \nBlack/African American | 7 | 32% | 5 | 25% | 5 | 25% | \nOther | 7 | 32% | 5 | 25% | 5 | 25% | \nDietary restraint2 |  |  |  |  |  |  | 0.66\nRestrained | 7 | 32% | 7 | 35% | 9 | 45% | \nUnrestrained | 15 | 68% | 13 | 65% | 11 | 55% | \n\n\n1Using ANOVA for age, weight and BMI, and Chi-square test for dietary restraint and race/ethnicity.\n\n2Classified using restraint subscale of TFEQ; score of <13 indicates restrained eater, >\u2009=\u200913 indicates unrestrained eater.\n\n\nCalories ordered and consumed (mean \u00b1 SE) by meal and study groups\n\n | Model 1: Calories ordered (mean \u00b1 SE) by meal1\nStudy group | n | Lunch 13 | Lunch 2 | Difference | Proportionate change (%)4 | p | Cohen\u2019s d\nNo calorie or exercise equivalent information | 22 | 1,201.4\u2009\u00b1\u2009100.0 | 1,176.1\u2009\u00b1\u200999.5 | -25.2\u2009\u00b1\u200995.2 | 9.3\u2009\u00b1\u200911.6 | 0.43 | 0.17\nCalories only | 20 | 1,282.8\u2009\u00b1\u200989.7 | 1,077.0\u2009\u00b1\u2009114.0 | -205.8\u2009\u00b1\u2009110.6 | -14.4\u2009\u00b1\u20097.3 | 0.06 | 0.45\nCalories and exercise equivalents | 20 | 1,162.8\u2009\u00b1\u2009141.1 | 1,000.5\u2009\u00b1\u200998.2 | -162.3\u2009\u00b1\u2009132.5 | 1.6\u2009\u00b1\u200913.3 | 0.90 | 0.02\n | Model 2: Calories consumed (mean \u00b1 SE) by meal2\nStudy Group | n | Lunch 13 | Lunch 2 | Difference | Proportionate change (%)5 | p | Cohen\u2019s d\nNo calorie or exercise equivalent information | 22 | 986.6\u2009\u00b1\u200984.1 | 995.4\u2009\u00b1\u200991.5 | 8.8\u2009\u00b1\u200983.9 | 10.7\u2009\u00b1\u200911.6 | 0.36 | 0.2\nCalories only | 20 | 1,059.6\u2009\u00b1\u200972.7 | 898.8\u2009\u00b1\u200987.6 | -160.7\u2009\u00b1\u2009106.3 | -9.3\u2009\u00b1\u200910.7 | 0.39 | 0.2\nCalories and exercise equivalents | 20 | 840.9\u2009\u00b1\u200988.6 | 841.3\u2009\u00b1\u200982.0 | 0.5\u2009\u00b1\u200976.9 | 11.9\u2009\u00b1\u200913.7 | 0.40 | 0.2\n\n\n1ANCOVA p-value\u2009=\u20090.43, controlling for age, BMI, race, dietary restraint, and calories ordered at Lunch 1; partial eta squared\u2009=\u20090.03, observed power\u2009=\u20090.19.\n\n2ANCOVA p-value\u2009=\u20090.31, controlling for age, BMI, race, dietary restraint, and calories consumed at Lunch 1; Partial eta squared\u2009=\u20090.04, observed power\u2009=\u20090.25.\n\n3All persons received menus with no calorie or exercise equivalent information at Lunch 1.\n\n4Overall mean \u00b1 SE of individual proportionate changes, each calculated as (calories ordered in Lunch 2-calories ordered in Lunch 1)/calories ordered in Lunch 1.\n\n5Overall mean \u00b1 SE of individual proportionate changes, each calculated as (calories consumed in Lunch 2-calories consumed in Lunch 1)/calories consumed in Lunch 1.",
        "label": "low",
        "id": "task4_RLD_test_129"
    },
    {
        "paper_doi": "10.3390/tropicalmed4040141",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Design\ncNON-RCT*\nAllocation of clusters1 village randomized* to intervention, 1 to control\n\n\nParticipants: 527 individuals ages 3 to 70\n\n\nInterventions: Broad multiple\n\n\nOutcomes: Any STH\n\n\nNotes: *The study may have used a random mechanism to allocate the intervention, but there was only 1 intervention area compared to 1 control area, so randomization in this case not likely to have reduced confounding or imbalances\n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Many latrine campaigns in developing countries fail to be sustained because the introduced latrine is not appropriate to local socio-economic, cultural and environmental conditions, and there is an inadequate community health education component.\nWe tested a low-cost, locally designed and constructed all-weather latrine (the \u201cBALatrine\u201d), together with community education promoting appropriate hygiene-related behaviour, to determine whether this integrated intervention effectively controlled soil-transmitted helminth (STH) infections.\nWe undertook a pilot intervention study in two villages in Central Java, Indonesia.\nThe villages were randomly allocated to either control or intervention with the intervention village receiving the BALatrine program and the control village receiving no program.\nSTH-infection status was measured using the faecal flotation diagnostic method, before and eight months after the intervention.\nOver 8 months, the cumulative incidence of STH infection was significantly lower in the intervention village than in the control village: 13.4% vs. 27.5% (67/244 vs. 38/283, p < 0.001).\nThe intervention was particularly effective among children: cumulative incidence 3.8% (2/53) for the intervention vs. 24.1% (13/54) for the control village (p < 0.001).\nThe integrated BALatrine intervention was associated with a reduced incidence of STH infection.\nFollowing on from this pilot study, a large cluster-randomised controlled trial was commenced (ACTRN12613000523707).\n1. Introduction\nThe global prevalence of infection with soil-transmitted helminths (STH) remains high, with 1.5 billion people infected worldwide, many of them children.\nOver two thirds of STH infections are in Asia, mostly in Southeast Asia.\nThe prevalence of STH infection in Indonesia is high at 45\u201365%, with areas having poor sanitation reaching 80% prevalence.\nIn Central Java, research into STH infection among elementary school children by Laksono and later the Health Department, found an infection prevalence of 84\u201396%.\nMore recently, a cross-sectional survey in Semarang, Central Java, found a prevalence of 34% among a cohort of 6466 people aged between two and 93 years.\nAnthelmintic drugs aimed at reducing morbidity are effective, but only temporarily, with a cure often followed by subsequent reinfection.\nIn rural areas, open defecation coupled with poorly constructed or inadequate latrines allows STH eggs to spread infection.\nTherefore, for long-term prevention, improved sanitation and community education are essential.\nA recent systematic review and meta-analysis concluded that \u201cintegrated control approaches emphasizing health education and environmental sanitation are needed to interrupt transmission of STH\u201d.\nIn particular, interventions that improve the hygienic disposal of faeces to reduce soil and/or water contamination have been identified as a key strategy to control transmission and prevent related diseases.\nIn Indonesia, open defecation is common, with 55% of the poorest and 18% of the richest households practicing open defecation.\nIn 2010, less than 40% of the people in rural areas had improved latrines, defined as facilities that hygienically separate human excreta from human contact.\nThe country did not reach its Millennium Development Goal of 75% sanitation coverage by 2015.\nIn rural areas, which include 118 million people, or 46.3% of the country\u2019s population, it has been estimated that 47% of the population have improved latrines, 12% shared latrines, 12% other unimproved latrines and 29% no latrines (i.e., practice open defecation).\nCompounding the problems caused by the lack of improved latrines is inappropriate hygiene-related behaviour, particularly related to hand washing, with 2007 National baseline data indicating that less than a quarter (23.2%) of the population had appropriate hand-washing behaviour.\nThe aim of this study was to develop and test an integrated approach to the prevention of STH infection and reduction of both transmission and reinfection.\nOur intervention included anthelmintic drugs, the construction and adoption of improved latrines, and effective education regarding hygienic and sanitary behaviour.\n2. Methods\n2.1. Study Design\nThis study was conducted in two villages, Palemon and Cepoko, in the Gunungpati sub-district of the city of Semarang, Central Java, Indonesia (see Figure 1), from July 2011 to June 2012.\nA random selection was made from these villages regarding which one should receive the integrated intervention and which one should be the control, by researchers who had no prior knowledge or contact with the villagers or village officials.\nThe two villages though similar in size and local topography, were not in close proximity to each other.\nThe study area is wooded and hilly and most of the houses, often made of local brick, are constructed by the householders themselves.\nMore than half of the households in the study villages did not have their own latrines.\nA randomly selected cohort (control: n = 244; intervention: n = 283) was followed over the eight-month duration of the study.\nA questionnaire was administered at baseline and follow-up to all village residents, with the eligibility criterion of being over two years of age.\nParticipants also provided two stool samples for parasitological examination.\nFollowing the baseline survey, all residents (regardless of infection status) were treated with anthelmintic medication and the incidence of STH infection was assessed at follow-up.\nFor ethical reasons, participants who were found to be STH positive at follow-up were re-treated.\n2.2. Ethics \nEthical approval was given by the Semarang City authorities (ref. 070/613/IV/2011), and from the Human Research Ethics Committee at Griffith University (ref. PBH/17/11/HREC).\n2.3. Procedure\nOur study procedure reflected the integrated model previously described (see also Figure 2), comprising chemotherapy, a locally constructed latrine (the \u201cBALatrine\u201d) and community health education.\nFollowing WHO Guidelines, a single oral dose of Albendazole (400mg) was administered immediately after the baseline survey.\nThe BALatrine is designed for resource poor rural communities and emergency situations, to be made by local people using local materials.\nTesting for cultural acceptance was conducted in the field through pilot studies in Pekalongan in Central Java and BALatrines were also used in an emergency refugee camp during the 2010 eruption of the Mt Merapi volcano, where they were proven appropriate for the level of technology available in the village context.\nThe BALatrine is a relatively simple squat latrine (Figure 3) that can be constructed by village residents using inexpensive materials.\nWhen water for flushing is available, a U-bend (\u2018goose-neck\u2019) water closet can be attached.\nWhen water is not available, such as during the dry season, the latrine can be used in a dry-pit configuration (with removal of the U-bend attachment), with a lid to isolate it from insects and to prevent odours from escaping.\nThus, it can function despite seasonal changes in water supply.\nBesides being inexpensive (cost at the time of the study was $50 USD for the latrine and local building materials; equating to approximately $80\u201390 at the time of publication), for people with limited financial and educational resources it is easy to copy.\nIts construction and use reflect critical resource, environmental and technical issues and due to local village input it also overcomes some major disincentives embodied in conventional latrines by being culturally familiar, simple and easy to use.\nThe community health education programme is an essential adjunct to latrine construction.\nIn the intervention village, all residents were given health education regarding hygiene, sanitation, and prevention of STH infections.\nThis health education component was delivered via community meetings in each village.\nAll village residents were invited and meetings were held in the village meeting hall.\nThe health education/health promotion component of the intervention was implemented through a two-hour village-wide mobilization meeting, which formed the project launch and was designed to mobilize households by consciousness-raising and provision of information about parasite infection and burden of STH.\nSubsequently, a series of small group workshops took place with the villagers in order to describe the BALatrine construction in detail and how to plan, construct, use, and maintain their latrines, as well as to discuss STH disease pathways.\nThe content of the health education programme comprised information about the dangers of STH infections and, using illustrated leaflets, how the transmission of STH infections can be prevented by the construction of latrines and with appropriate hygiene-related behaviours.\n2.4. Measurements and Analyses\nThe primary outcome measures of the integrated intervention were STH infection status at baseline and at follow-up.\nSTH infection status was measured through laboratory analysis of stool samples collected from each participant at baseline and at follow-up eight months after the BALatrines were constructed.\nThe samples were analysed microscopically for the presence of helminth eggs, according to the Willis-Mollay Flotation technique.\nAfter preparing the samples, a cover slip was placed over each sample tube and left for 10 minutes.\nAfter 10 minutes, a drop of eosin solution (2%) was added to a glass slide onto which the cover slip was then placed and observed using a light microscope at 10 \u00d7 magnification.\nA positive sample was where at least one STH egg was identified.\nA face-to-face questionnaire (the Helminth Education and Latrine Project (HELP) questionnaire) was also administered at baseline and follow-up and this provided information about villagers\u2019 demographic attributes.\nWe also assessed local village environmental contamination with faeces.\nThese findings have been published elsewhere.\nData were analysed using SPSS Version 22 (IBM, New York, United States), Microsoft Excel and the tools at Open Source Epidemiologic Statistics for Public Health.\nDifferences between participants in control and intervention villages were analysed using the unpaired t-test and the Pearson\u2019s chi-squared (\u03c72) test.\nLogistic regression analysis was performed and odds ratios calculated.\nTo compare the intervention village with the control village, we report both crude odds ratios and adjusted for age and sex.\n3. Results \n3.1. Characteristics of the Participants\nThere were 527 participants in the study at baseline, 244 in the control village and 283 in the intervention village.\nTheir ages ranged from three to 70 years, with a mean \u00b1 SD of 29.4 \u00b1 16.1 years in the control village and 32.2 \u00b1 17.9 years in the intervention village (Table 1; for the age difference between villages, p = 0.06).\nIn both villages, similar proportions of the participants had completed elementary education (control, 211/244, 86.5%; intervention, 253/283, 89.4%).\nIn the control village, 98.8% of the residents had a monthly income below 1 M Indonesian Rupiah (IDR) or about US$70, whereas in the intervention village the comparable percentage was 97.2% (p = 0.017).\nIn the control village, 38.5% of the residents lived in a home in which all floor spaces were dry, whereas in the intervention village the comparable percentage was 54.1% (p < 0.001).\nIn the intervention village greater than 90% of households adopted the BALatrine, as measured at the follow up.\n3.2. STH-Infection Status\nAt baseline, the prevalence of STH infection was almost the same in the two villages (Table 1).\nAt follow-up, the cumulative incidence of infection was much lower in the intervention village than in the control village (Table 2, 13.4% vs. 27.5%, p < 0.001).\nAfter adjustments for age and gender, the benefit of the intervention was still clear (adjusted odds ratio = 0.38, 95% CI 0.25\u20130.60, Table 2).\nThe intervention was particularly effective in children (adjusted odds ratio = 0.12, 95% CI 0.03\u20130.56, Table 2).\n4. Discussion\nImproving access to adequate sanitation is a critical step toward the sustainable interruption of STH transmission.\nYet this is an ongoing challenge in low resource settings where public sewerage system infrastructure rarely exists, on-site sanitation systems are either improperly designed or poorly functioning, and open defecation is seen as culturally acceptable, especially in rural areas.\nOften people must rely on shared sanitation facilities, which have been shown to increase the risk of adverse health outcomes compared to individual household latrines.\nCompounding the issue is limited access to materials and a lack of technical expertise to build or improve latrines.\nThe BALatrine was specifically designed to overcome many of these challenges whilst also taking into consideration cultural appropriateness and acceptability.\nUsing simple technology and locally sourced, inexpensive materials, the BALatrine is cheap, easy to build and maintain and adaptable for wet and dry conditions.\nBuilding the latrines through community mobilisation also helps keep the costs down and enables the householders themselves to take ownership over the latrines and their maintenance and consequently, benefit from improved health, helping alleviate the cycle of poverty that is often associated with STH infections.\nIn the present study, we evaluated the effectiveness of an integrated BALatrine intervention at reducing human worm burden through a pilot study in two villages in Central Java.\nPeople in the intervention village were 2.6 times less likely to be infected following the BALatrine-based intervention than those in the control village indicating that the BALatrine is associated with a reduced worm burden.\nHowever, it is important to note that we did not resurvey participants after baseline deworming to determine the efficacy of the Albendazole treatment nor did we differentiate between STH species, which are known to respond differently to treatment.\nWe have based our interpretation of the results on the assumption that treatment was effective at temporarily reducing infections to zero.\nConsequently, it is possible that the effect seen could be a result of differential cure rates between villages based on their STH profile at baseline.\nHowever, a prevalence study we conducted the following year (manuscript under review) across 16 villages in two neighbouring subdistricts of Semarang, including Gunung Pati, revealed Ascaris lumbricoides as the predominant species (mean prevalence of 26% vs. 7.9% and 1.8% for hookworm and Trichuris trichuria, respectively).\nWe therefore believe that it is not unrealistic to assume that our study villages also had similar burdens of each of the STH species at baseline and that treatment would be similarly effective across the two sites.\nOur study also found that nearly all households adopted the BALatrine suggesting a strong willingness among villagers to improve sanitation and a desire to improve the health of their families.\nAccess to improved latrines does not guarantee their use, however, particularly over time when old habits or cultural preferences can be difficult to overcome.\nIntegrating health education and promotion programmes to improve peoples\u2019 understanding and knowledge of the link between open defecation and ill-health and WASH-related behaviours, is therefore extremely important.\nPeople also need to understand the importance of properly cleaning and maintaining their latrines, particularly as this is associated with higher latrine use.\nIn the present study, a community health education programme was administered prior to the construction and installation of the latrines and aimed to raise awareness, improve hygiene behaviours and motivate the villagers to build and continue to use their new latrines, which may have led to the high uptake observed in this study.\nHowever, we did not assess the impact of this program on participants\u2019 knowledge and behaviour change or assess latrine use over time, which are key limitations of this study.\nIt is well established that chemotherapy alone will not break the transmission cycle.\nRecent studies have also shown that programmes solely focusing on WASH have limited effect on STH incidence and may provide no additional impact compared to mass drug administration programmes alone.\nIn contrast, health education and promotion programmes can be highly effective at reducing the incidence of STH infections if designed appropriately such as the highly successful \u201cMagic Glasses\u201d study, which resulted in a 50% reduction in STH infections.\nHowever, sustained reinforcement of health messages is required in order to increase their effectiveness over the longer term.\nUltimately, eliminating STH will be best achieved through integrated control programmes.\nThe current study adds to the growing body of research into the impact of integrated control programs on soil-transmitted helminthiases and demonstrates that sanitation interventions can be effective at reducing worm burden when designed appropriately for the local context and combined with health education and promotion.\nIn conclusion, our findings provide \u201cproof of principle\u201d that the BALatrine-based intervention is effective in preventing STH infection.\nWe will now undertake a full-scale randomized controlled trial and contribute much needed evidence based on WASH and STH.\nMap of Gunung Pati subdistrict in Semarang, Central Java (source: Wikipedia Indonesia, 2017).\nFlowchart of the study.\nSchematic presentation of the BALatrine.\n\nBaseline characteristics of participants.\nVillage Status | Control | Intervention\nVillage  | Cepoko | Palemon\nSample Size | 244 | 283\nMean Age (years) | 29.4 | 32.2\nPrevalence of STH infection: % (95% CI) | 21.7% (16.5\u201326.9) | 25.8% (20.7\u201330.9)\nSex Ratio (F/M) | 141/103 | 151/132\nPrevalence of STH infection by Sex (F/M)  | 22.0%/21.4% | 20.5%/31.8%\n\n\nInfection rates in the control and intervention villages.\nVariable | Control | Intervention | Odds Ratio | Odds Ratio\nCrude | p Value | Adjusted * | p Value\nAll participants | n = 244 | n = 283 |  |  |  | \nPrevalence of infection at baseline: % (95% CI) | 21.7 (16.5\u201326.9) | 25.8 (20.7\u201330.9) | - | - | - | -\nCumulative incidence of infection at follow-up: % (95% CI) | 27.5 (21.9\u201333.1) | 13.4 (9.5\u201317.4) | 0.41 (0.26\u20130.64) | <0.001 | 0.38 (0.25\u20130.60) | <0.001\nChildren | n = 54 | n = 53 |  |  |  | \nPrevalence of infection at baseline: % (95% CI) | 18.8 (8.2\u201328.9) | 18.9 (8.3\u201329.4) | - | - | - | -\nCumulative incidence of infection at follow-up: % (95% CI) | 24.1 (12.7\u201335.5) | 3.8 (0.0\u20138.9) | 0.12 (0.03\u20130.58) | 0.01 | 0.12 (0.03\u20130.56) | 0.01\n\n* The model for all participants was adjusted for age and sex. The model for children (<14 years) was adjusted for gender only.",
        "label": "high",
        "id": "task4_RLD_test_815"
    },
    {
        "paper_doi": "10.1155/2018/3629643",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallelEthics and informed consent: yes\nFollow-up period: 1 yearSample size estimate: noITT analysis: yes, number randomised: 51, number analysed: 51Funding: not reportedPreregistration: not reported\n\n\nParticipants: Location: Korea; single-centre (hospital)\nIntervention group: 30, control group: 21Mean age: intervention group 38.77 +- 1.68, control group 41.38 +- 10.92\nInclusion criteria: > 20 years, acute multi-tissue hand injury of moderate severity (assessed by HISS score 21-50), underwent reconstruction within 3 days after injury by two surgeons.\nExclusion criteria: history of impaired motor function, injury to the peripheral nerves and/or vessels distal to the wrist, or a bone fracture requiring transarticular fixation with a Kirchner (K) wire, a congenital hand deformity, an operation history on the same hand, and underlying diseases including autoimmune diseases such as rheumatoid arthritis or systemic lupus erythematosus or those taking medications that could influence wound healing.\n\n\nInterventions: Aim/s: to compare outcomes in patients with acute hand injury who were managed with or without NPWT after reconstructive surgery.\n Group 1 (NPWT) intervention: NPWT (CuraVAC, CGBio, Seongnam-si, Gyeonggi-do, Korea) applied at a pressure of 75 mmHg in continuous mode and secondary dressing including Vaseline gauze.Group 2 (control) intervention: conventional dressing, including vaseline gauze was applied over the closed skin using polyurethane foam with a compressible elastic bandage, and a short arm splint was applied in a functional position; dressing and NPWT were changed every 3 days.\nStudy date/s: January 2013 to December 2016\n\n\nOutcomes: SSI/infectionHaematomaWound disruption (dehiscence)Validity of measure/s: unclear what definition was used for infectionTime points: 1 month and 1 year\n\n\nNotes: \n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "In this study, we compared outcomes in patients with acute hand injury, who were managed with or without negative pressure wound therapy (NPWT) after reconstructive surgery.\nAll of the patients who sustained acute and multitissue injuries of the hand were identified.\nAfter reconstructive surgery, a conventional dressing was applied in Group 1 and NPWT was applied in Group 2.\nThe dressing and NPWT were changed every 3 days.\nThe mean age and Hand Injury Severity Scoring System score of both groups were not significantly different.\nDisabilities of the Arm, Shoulder, and Hand (DASH) scores were evaluated 1 month after all the sutures were removed and 1 year postoperatively, which were both significantly lower in Group 2.\nApplying NPWT to the hand promoted wound healing by reducing edema, stabilizing the wound, and providing immobilization in a functional position.\nEarly wound healing and decreased complications enabled early rehabilitation, which led to successful functional recovery, both objectively and subjectively.\n1. Introduction\nA significant proportion of hand injury cases are multiple faceted and heavily contaminated and involve composite soft tissue and bone injuries due to the complexity of the anatomy and function of the hand.\nAs a result, hand injuries are often difficult to manage promptly and require multiple staged serial treatment.\nOn the other hand, functional recovery is as important as structural reconstruction and resurfacing in hand injuries, as the hand is a functional unit.\nEarly exercise and rehabilitation improve functional recovery; therefore, wound healing should be achieved as soon as possible.\nNegative pressure wound therapy (NPWT) is a good alternative not only for management during the preoperative period of early reconstruction, but also for early recovery after reconstruction.\nNPWT has been widely used for almost every type of wound, from acute traumatic wounds to chronic intractable wounds.\nIt generates a subatmospheric pressure of 50\u2212150\u2009mmHg in either a continuous or an intermittent mode.\nAlthough the exact mechanism is undefined, the effects of NPWT are to remove excess fluid and debris, improve tissue perfusion, and promote wound healing by enhancing formation of granulation tissue and decreasing the size of the wounds.\nIn this study, we compared outcomes in patients with acute hand injury who were managed with or without NPWT after reconstructive surgery.\n2. Materials and Methods\nThis study was approved by the Institutional Review Board.\nAll of the data were analyzed anonymously and according to the principles in the 1975 Declaration of Helsinki, revised in 2008.\nThe study was a prospective open trial.\nAll of the adult patients (>20 years) who sustained acute multitissue injury of the hand from January 2013 to December 2016 were enrolled with the following criteria.\nThe patients included in this study sustained acute hand injury of a similar severity, as assessed by a Hand Injury Severity Scoring System (HISS) score of 21\u221250 (Table 1), which is defined as a moderate severity level II injury, and underwent reconstruction within 3 days after injury by two surgeons.\nPatients with a medical history of impaired motor function, injury to the peripheral nerves and/or vessels distal to the wrist, or a bone fracture requiring transarticular fixation with a Kirchner (K) wire, a congenital hand deformity, an operation history on the same hand, and underlying diseases including autoimmune diseases such as rheumatoid arthritis or systemic lupus erythematosus or those taking medications that could influence wound healing were excluded from the study.\nInformed consent was obtained from patients who met the inclusion criteria before randomization.\nPatients were randomly assigned to the control or experimental group following a simple randomization procedure (computerized random numbers) achieved using opaque envelopes.\nReconstruction was performed according to the injury on a case-by-case basis.\nBone fractures including fractures of the phalangeal, metacarpal, and carpal bones were fixed with K-wires averting articular surfaces, and the tip of the K-wire was closely cut and embedded under the skin.\nOpen reduction and ligament repair were performed as required for dislocated joints.\nTendons were repaired accordingly for tendon rupture or avulsion injuries.\nThe skin lacerations were closed primarily, and skin and soft tissue defects were reconstructed with local flaps or a skin graft.\nA silastic drain was inserted before closure.\nAfter reconstruction, a conventional dressing was applied over the closed skin using polyurethane foam with a compressible elastic bandage, and a short arm splint was applied in a functional position in Group 1 (control group).\nBy contrast, NPWT (CuraVAC\u00ae, CGBio, Seongnam-si, Gyeonggi-do, Korea) was applied at a pressure of 75\u2009mmHg in continuous mode in Group 2 (experimental group).\nThe secondary dressing for Group 2, including Vaseline gauze, was applied before NPWT.\nThe dressing and NPWT were changed every 3 days.\nIn both groups, when the skin was completely healed within 2 weeks after the injury, the dressing or NPWT was removed, followed by the sutures.\nPhysical therapy was started under consultation with the rehabilitation medicine department after wound healing, and the allocation information to each group was not provided to reduce bias.\nPhysical therapy was performed twice weekly for 4 weeks with individual home-exercise instruction.\nData were collected from the patient's medical records and radiographs.\nThe baseline characteristics collected were age, sex, date of injury, injury site, and the HISS score.\nTime to recover over 90% of the full range of motion (ROM) compared to the normal values of full flexion and extension was analyzed for every interphalangeal and metacarpal joint.\nIn addition, the Disability of the Arm, Shoulder, and Hand (DASH) score was evaluated 1 month after suture removal, when the skin was completely healed and 1 year postoperatively.\nEvaluated complications were hematoma, infection, wound disruption, or a secondary operation.\nComparisons between the two groups were performed using chi-square test and Fisher's exact test.\nA p value < 0.05 was considered statistically significant.\n3. Results\nWe identified 51 patients (17 females and 34 males; age: from 21\u201361 years; mean age: 39.8 years) with acute hand injuries who met the study inclusion criteria.\nA total of 21 patients received conventional dressing using polyurethane foam and a short arm splint, and 30 patients received NPWT.\nThe mean age of Group 1 was 41.4 (range: 22\u201361) years and that of Group 2 was 39.9 (range: 21\u201361) years.\nThe mean HISS score of Group 1 was 33.6 (range: 21\u201350) and that of Group 2 was 35.7 (range, 21\u201350).\nNo significant differences were observed in patient demographics or HISS scores between the two groups.\nDASH scores were evaluated 1 month after all of the sutures were removed and 1 year postoperatively.\nThe scores at 1 month averaged 33.14 (range: 18.3\u201348.3) in Group 1 and 22.67 (range: 5.8\u201340.1) in Group 2 (p = 0.031).\nThe score at 1 year averaged 22.08 (range: 14.9\u201331.9) in Group 1 and 20.99 (range: 5.1\u201332.0) in Group 2 (p = 0.667).\nThe hand joints recovered >90% of the full ROM at 46.9 (range: 30\u201361) days after injury in Group 1 and at 33.3 (range, 22\u201358) days in Group 2 (p = 0.022).\nThere were five complications: two hematomas and one infection were treated conservatively by drainage and antibiotics in Group 1, and two wound macerations in Group 2 healed conservatively without additional surgery.\nNo difference in complications was observed between the two groups.\nThe statistical comparisons between the two groups are presented in Table 2.\nCase 1. A 59-year-old male visited the emergency room after his hand had been smashed in a heavy rolling machine.\nAll of the dorsal skin on the hand was avulsed with multiple ruptures of the extensors (Figure 1).\nThe ruptured second and third extensor digitorum communis and fifth extensor digitorum minimi were repaired, the avulsed skin envelope was tension- free repaired, and a silastic drain was inserted (Figure 2).\nThen the whole dorsal side of the hand except the fingers was covered with NPWT (Figure 3).\nNPWT was changed every 3 days.\nFull ROM was achieved without restriction of daily activity 4 weeks after suture removal (Figure 4).\nCase 2. A 54-year-old male suffered a multitissue injury of the right second to fifth fingers in a press machine accident.\nThe third proximal phalangeal bone was fractured in an avulsed manner.\nThe second and third flexor tendons were also ruptured, and multiple skin defects occurred on the volar side of the hand (Figure 5).\nThe fractured bone was reduced while repairing the ruptured flexor tendons, and the lacerations and skin defects were repaired with skin grafts.\nThen, the whole volar side of the hand was covered with NPWT, which was changed every 3 days (Figure 6).\nThe wound was healed 2 weeks later, and the sutures were removed.\nAfter 2 months of rehabilitation and physical therapy, the patient was able to use his hands freely with full flexion and extension and returned to work (Figure 7).\n4. Discussion\nNPWT was first reported in 1993 and was introduced as \u201cvacuum-assisted closure\u201d for wound control and treatment by Morykwas et al. in 1997.\nSince then, NPWT has been widely used not only for chronic nonhealing wounds, but also for acute traumatic injuries.\nIts effectiveness is thought to be due to decreased bacterial count, increased tissue perfusion, removal of exudates, and promotion of granulation tissue formation, all of which promote wound healing.\nThe NPWT system consists of foam connected to a vacuum pump through a connecting tube, and the whole system is covered with a semiocclusive dressing.\nThe application of NPWT has been expanded from managing and protecting the wound and preparing for final reconstruction to improving skin graft outcomes and patient comfort and thereby reducing cost.\nNPWT has been mostly used in patients undergoing hand surgery with soft tissue defects associated with trauma, burns, or infection.\nThe effective use of NPWT in preparing soft tissue defects before reconstruction has been well described, and favorable results have been achieved in patients with bone, tendon, or nerve exposure.\nOn the other hand, use of NPWT after reconstruction has only been reported in selective cases.\nMost commonly, NPWT has been applied after skin grafting.\nNPWT stabilizes the graft and promotes adherence of the skin graft, which improves graft take.\nThe hand is a functional and mobile unit with a complex anatomy.\nThus, reconstruction of hand injuries should focus not only on resurfacing with healthy soft tissue, but also on maintaining good muscle strength and flexibility of the tendons without adhesion.\nMany joints of the hand require early rehabilitation of the ROM to prevent contracture.\nNPWT can be applied after reconstruction and offers several advantages.\nIt can be used instead of conventional polyurethane foam and short arm splint dressings.\nNPWT simplifies the dressing while stabilizing the hand.\nUse of NPWT splints of the hand in a functional position, and the hand can be molded into the desired functional position before applying suction.\nThe absence of a splint allows easy visualization of the position and status of the hand.\nMoreover, NPWT with foam allows only minimal motion of the joints, functioning as a sort of dynamic splint.\nNPWT as a partial dynamic splint has two advantages.\nOne is that NPWT helps decrease swelling, which leads to better overall hand function.\nHand wounds can remain swollen for some time following injury and become more swollen after reconstruction.\nReduced swelling fosters early recovery of tissues, which leads to early rehabilitation.\nAs in our cases, NPWT can also function as a negative drainage tool through the space using a silastic drain.\nA second advantage is that the minimal motion of the joint protects against severe contracture of the hand.\nThese advantages explain the superior results of NPWT compared to a conventional dressing with a splint.\nOne of the limitations of using NPWT is that the pressure might compress the microvessels in the soft tissue, compromise the vascularity of the tissue, and decrease tissue perfusion.\nApplying NPWT could be of concern to surgeons, particularly in the hand, where blood circulation is limited to certain vessels and thin, pliable soft tissue with a weaker cushion effect.\nSome surgeons hesitate to use NPWT on the hands because of fear of restricted movement, difficulty of the application, and leakage due to the complex shape of the hand.\nTherefore, some modifications of commercially available NPWT have been reported, using gauze instead of a foam sponge or a sealing bag instead of semiocclusive dressing coverage.\nThese modifications are suitable adjustments for hand injuries; however, they were reported as certain indicated cases and required surgical adjustment on a case-by-case basis.\nThe pressure applied through the NPWT foam evenly distributes the mechanical force to the wound.\nMorykwas et al. tested various suction pressures from 0 to 400\u2009mmHg and found that 125\u2009mmHg was optimal for increasing local blood flow.\nCurrent recommendations state that 50\u2013150\u2009mmHg of negative pressure is acceptable.\nIn our cases, the hand was placed in the most functional position, and the drain was connected to suction power and set to 75\u2009mmHg.\nAlthough 125\u2009mmHg is the standard pressure for NPWT, similar effects can be achieved at lower pressures.\nIn addition, some reports have demonstrated that tissue pressure increases beneath the NPWT in all types of wounds, is directly proportional to the amount of suction applied, and is most pronounced in circumferential dressing.\nPrevious authors have reported that increased pressure results in 17% decreased perfusion when circumferential NPWT is applied with a suction pressure of 125\u2009mmHg.\nThe theories regarding the mechanism of action of NPWT suggest that compression of tissue decreases perfusion and concurrent hypoxia is a stimulus for angiogenesis.\nIn addition, tissue hypoxia results in release of nitric oxide and local vasodilation.\nOn the other hand, concerns regarding the safety of NPWT on tissues with compromised perfusion have also been raised.\nWe are aware that there is a potential risk for perfusion from the compression effects of a circumferentially applied NPWT dressing on the hand.\nWe did not find any evidence of reduced vascularity or compromised tissue perfusion as a result of using NPWT for hand injuries.\nBy contrast, we noticed a significant reduction in edema.\nThe compression provided by NPWT likely forces edema away from injured tissues.\nThis ultimately results in decreased interstitial pressure, decreased compression of the vessels, and improved oxygen and nutrient supply.\nThese results are likely to be the most important contributions of NPWT.\nThe HISS is the most commonly used measure to clinically assess hand injury severity.\nIt is evaluated by scoring the severity of each hand segment from skin, bones, motor function, and nerve injury.\nThe total score is determined by adding the point values of hand injury severity, then classifying it according to the score obtained, expressed as grades I\u2013IV.\nIn this study, we excluded patients with concomitant peripheral nerve injury because in this study, we compared the functional outcomes of patients with acute hand injury with or without NPWT after surgery, and nerve injury can interfere with the results of the functional outcome.\nIn addition, we only included patients with hand injuries and HISS scores of 21\u201350, which is grade II, and those who underwent reconstruction within 2 weeks after the injury.\nPatients with HISS scores > 50 and who were severely injured are often difficult to manage and require several staged treatments that could not be performed within 2 weeks; therefore, they were excluded.\nA quantitative assessment of hand dysfunction is much more difficult.\nAmong the most commonly used scales is the DASH scale, a 30-item questionnaire that evaluates symptoms and physical function with a five-response option for each item.\nThe DASH score is determined by calculating the circled responses.\nIt produces a brief, self-administered measure of symptoms and functional status.\nThe only limitation is that the DASH is a subjective measurement that represents hand function but does not fully correlate with objective functional recovery.\nTherefore, we first evaluated the DASH scale to represent personal symptoms and subjective situations.\nThen, we also evaluated objective functional recovery by determining the period when recovery of ROM was >90%.\nA 90% recovery of ROM is almost full recovery of function, which enables daily activity, the end of rehabilitation, and the return to social life.\nIn Group 2, the DASH scores were lower and the number of ROM recovery days was fewer compared to those in Group 1; these differences were statistically significant.\nAlthough the DASH score at postoperative 1 year was not different, the results suggest that NPWT was essential for early and fast recovery of hand function.\nIn the cases we described above, NPWT was successfully used to treat challenging hand injuries.\nComplications such as tissue loss, dehiscence, infection, or hematoma can have serious effects on the functional outcome.\nThe use of NPWT on very thin skin flaps or over skin grafts, where there is a concern for hematoma, perfusion, and skin survival, was particularly useful.\nIn addition, commercially available NPWT is increasingly evolving.\nThe foam sponge has become thinner, more flexible, and customized to the defect; the connecting tube is slender and length-adjustable; and the vacuum pump system has become smaller and more easily portable.\nAs NPWT maintains the injured hand in a stable state and can be changed every 3 days, most patients can be discharged and followed in the outpatient clinic, which is more convenient for patients and reduces hospital stay and costs.\nIn our experience of treating acute hand injures, NPWT is quick and easily applied.\nNPWT promotes wound healing by reducing edema, stabilizing the wound, and providing immobilization in a functional position.\nEarly wound healing and decreased complications enabled early rehabilitation, which lead to a successful functional recovery, both objectively and subjectively.\nA 59-year-old male suffered multitissue injury of the right hand, including all of the dorsal skin and extensors.\nEmergency operations including tendon repair and wound closure were performed.\nThe wound was covered with negative pressure wound therapy (NPWT) immediately after surgery.\nFull range of motion was achieved 4 weeks after surgery.\nA 54-year-old male with multitissue injury of the right hand visited the emergency room. Flexors of the second and third fingers were ruptured, with an avulsion bone fracture and multiple skin defects.\nThe patient was treated with negative pressure wound therapy (NPWT).\nFull range of motion was achieved 2 months after surgery, without restriction of daily motion.\n\nHand Injury Severity Scoring System (HISS).\n\u2009 | Score | Sum | Severity\nIntegumental injuries | 0\u201340 | <20 | I: minute\nBone injuries | 0\u20139 | 21\u201350 | II: medium\nImpairment of motor function | 0\u201316 | 51\u2013100 | III: severe\nNerve injury | 0\u201334 | >100 | IV: major\n\n\nComparison between two groups.\n\u2009 | Group 1(conventional dressing) | Group 2(NPWT) | p value\nAge (years) | 41.38 \u00b1 10.92 | 38.77 \u00b1 1.68 | 0.375\n\nHISS score | 33.57 \u00b1 1.86 | 35.73 \u00b1 1.55 | 0.377\n\nTime to recover over 90% of the ROM(days) | 46.90 \u00b1 2.05 | 33.30 \u00b1 1.51 | 0.022\n\nDASH score at one month | 33.14 \u00b1 1.68 | 22.67 \u00b1 1.43 | 0.031\n\nDASH score at one year | 22.08 \u00b1 2.03 | 20.99 \u00b1 1.91 | 0.667\n\nComplications | 3 | 2 | 0.383\nHematoma | 2 | 0\nInfection | 1 | 0\nWound disruption | 0 | 2\n\nNPWT: negative pressure wound therapy; HISS: Hand Injury Severity Scoring System; ROM: range of motion; DASH: Disability of the Arm, Shoulder, and Hand questionnaire.",
        "label": "unclear",
        "id": "task4_RLD_test_721"
    },
    {
        "paper_doi": "10.1371/journal.pntd.0001044",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: randomized, non-blinded study.Participants: adult participants with chronic strongyloidiasis.Length of follow-up: participants were followed up 2 weeks after treatment initiation, then 1 month, 3 months, 6 months, 9 months, and 1 year post-treatment.Monitoring and diagnostics: infection with Strongyloides stercoralis was ascertained using the direct smear, formol-ether concentration method, and modified Koga agar plate culture method.\n\n\nParticipants: Number of participants: 90 adult participants with chronic Strongyloides infection were recruited. There were 10 participants with HIV co-infections, with 3 HIV-positive participants randomized to albendazole, 2 HIV-positive participants randomized to single dose ivermectin, and 5 HIV-positive participants randomized to double dose ivermectin.Inclusion criteria: adult participants with characteristic rhabditiform larvae of S. stercoralis present on faecal microscopy.Exclusion criteria: history of allergic reaction to either study medication, treatment within the month prior to the trial with any drug known to have anti-Strongyloides activity, pregnancy, or lactation, and any suggestion of disseminated strongyloidiasis.\n\n\nInterventions: Intervention: Group 1: ivermectin delivered as a single dose of 200 ug/kg; Intervention. Group 2: 2 doses of ivermectin (200 ug/kg) delivered 2 weeks apart. For the purpose of this analysis we considered both groups that received ivermectin together.Control: participants received 7 days of albendazole (800 mg per day).\n\n\nOutcomes: Outcomes included in this review: incidence of adverse events defined as \"symptoms or signs that developed after the study drug administration and had not been reported prior to the administration of the first dose of the antihelmintic.\"Other trial outcomes: treatment cure (defined as clinical improvement (if symptomatic before treatment) and the absence of rhabditiform larvae in the stool at day 14 of treatment and throughout the follow-up period) and treatment failure (defined as the presence of larvae two weeks after initiation of treatment or the reappearance of larvae during follow-up).\n\n\nNotes: Location: Siriraj Hospital, ThailandParticipant helminth status: ascertainedParticipant ART status: it was unspecified if any individuals were receiving ART treatment at the start or during the trial.Author contact: we requested additional data regarding the incidence of adverse events in the HIV-positive participants specifically from the trial authors, who provided this information\n\n",
        "objective": "To evaluate the effects of deworming drugs (antihelminthic therapy) on markers of HIV disease progression, anaemia, and adverse events in children and adults.",
        "full_paper": "Background\nStrongyloidiasis, caused by an intestinal helminth Strongyloides stercoralis, is common throughout the tropics.\nIt remains an important health problem due to autoinfection, which may result in hyperinfection and disseminated infection in immunosuppressed patients, especially patients receiving chemotherapy or corticosteroid treatment.\nIvermectin and albendazole are effective against strongyloidiasis.\nHowever, the efficacy and the most effective dosing regimen are to be determined.\nMethods\nA prospective, randomized, open study was conducted in which a 7-day course of oral albendazole 800 mg daily was compared with a single dose (200 microgram/kilogram body weight), or double doses, given 2 weeks apart, of ivermectin in Thai patients with chronic strongyloidiasis.\nPatients were followed-up with 2 weeks after initiation of treatment, then 1 month, 3 months, 6 months, 9 months, and 1 year after treatment.\nCombination of direct microscopic examination of fecal smear, formol-ether concentration method, and modified Koga agar plate culture were used to detect strongyloides larvae in two consecutive fecal samples in each follow-up visit.\nThe primary endpoint was clearance of strongyloides larvae from feces after treatment and at one year follow-up.\nResults\nNinety patients were included in the analysis (30, 31 and 29 patients in albendazole, single dose, and double doses ivermectin group, respectively).\nAll except one patient in this study had at least one concomitant disease.\nDiabetes mellitus, systemic lupus erythrematosus, nephrotic syndrome, hematologic malignancy, solid tumor and human immunodeficiency virus infection were common concomitant diseases in these patients.\nThe median (range) duration of follow-up were 19 (2\u201376) weeks in albendazole group, 39 (2\u201374) weeks in single dose ivermectin group, and 26 (2\u201374) weeks in double doses ivermectin group.\nParasitological cure rate were 63.3%, 96.8% and 93.1% in albendazole, single dose oral ivermectin, and double doses of oral ivermectin respectively (P\u200a=\u200a0.006) in modified intention to treat analysis.\nNo serious adverse event associated with treatment was found in any of the groups.\nConclusion/Significance\nThis study confirms that both a single, and a double dose of oral ivermectin taken two weeks apart, is more effective than a 7-day course of high dose albendazole for patients with chronic infection due to S. stercoralis.\nDouble dose of ivermectin, taken two weeks apart, might be more effective than a single dose in patients with concomitant illness.\nTrial Registration\nClinicalTrials.gov NCT00765024\nAuthor Summary\nStrongyloidiasis, caused by an intestinal helminth Strongyloides stercoralis, is common throughout the tropics.\nWe conducted a prospective, clinical study to compare the efficacy and safety of a 7-day course of oral albendazole with a single dose of oral ivermectin, or double doses, given 2 weeks apart, of ivermectin in Thai patients who developed this infection.\nPatients were regularly followed-up after initiation of treatment, until one year after treatment.\nNinety patients were studied (30, 31 and 29 patients in albendazole, single dose, and double doses ivermectin group, respectively).\nThe average duration of follow-up were 19 (range 2\u201376) weeks in albendazole group, 39 ( range 2\u201374) weeks in single dose ivermectin group, and 26 ( range 2\u201374) weeks in double doses ivermectin group.\nParasitological cure rate were 63.3%, 96.8% and 93.1% in albendazole, single dose oral ivermectin, and double doses of oral ivermectin respectively.\nNo serious adverse event associated with treatment was found in any of the groups.\nTherefore this study confirms that both a single, and a double dose of oral ivermectin taken two weeks apart, is more effective than a 7-day course of high dose albendazole for patients with chronic infection due to S. stercoralis.\nIntroduction\nInfection with the intestinal helminth Strongyloides stercoralis remains a common problem throughout the tropics, including Thailand.\nIt is estimated that 30 to 100 million people are infected worldwide.\nMost infected individuals are asymptomatic or developed minimally symptomatic chronic infection through autoinfection.\nPotentially fatal disseminated infections, due to an acceleration of the autoinfection cycle, are seen in immunocompromised patients, such as those with concurrent human T-lymphotropic virus-1 (HTLV-1) infection, or those on corticosteroid therapy.\nOther recognized risk factors for disseminated strongyloidiasis include malignancies especially lymphoma, organ transplantation and diabetes mellitus.\nGastrointestinal symptoms associated with strongyloidiasis include diarrhea, abdominal discomfort, nausea/vomiting and anorexia.\nThe diagnosis of strongyloidiasis should be suspected if there are clinical signs and symptoms, or eosinophilia.\nDefinitive diagnosis of strongyloidiasis is usually made on the basis of detection of larvae in the stool.\nThe combination of diagnostic approaches such as repeated direct microscopic examination of fecal smear, fecal concentration methods such as formol-ether concentration (FEC), and modified Koga agar plate culture have been used to improve the likelihood of detecting this parasite.\nIn the past, the treatment of choice for strongyloidiasis has been thiabendazole, but this drug has unpleasant side effects and is no longer available.\nAlbendazole, another broad-spectrum antihelmintic agent, was previously shown to be effective against S. stercoralis .\nMore recent reports suggest ivermectin, a macrolide-like agent developed primarily for the treatment of onchocerciasis, is as effective as thiabendazole and superior to albendazole against intestinal strongyloidiasis.\nAlthough a single dose of ivermectin 200 microgram/kilogram body weight (\u00b5g/kg) was shown to be effective in uncomplicated chronic strongyloidiasis, repeated treatment at two or three week intervals was thought to be necessary to eliminate larvae generated by autoinfection.\nA preparation of oral ivermectin licensed for human use has recently become available in Thailand.\nHowever, albendazole remains the most widely used antiparasitic drugs for the treatment of this infection in this country.\nThe purpose of the present study was to assess the safety and efficacy of a single dose of ivermectin (200 \u00b5g/kg), or two doses of ivermectin given 2 weeks apart, and a 7-day course of high dose albendazole for the treatment of chronic strongyloidiasis in adult patients who were at high risk of hyperinfection or disseminated infection.\nMaterials and Methods\nStudy design and ethics\nThis was a prospective open-label, randomised, controlled study conducted between July 2008 and April 2010 at Siriraj Hospital, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand.\nThe study was approved by the Ethical Committee on Research Involving Human Subjects, Siriraj Hospital, Faculty of Medicine, Mahidol University, Thailand.\nAll patients were informed about the purpose of the trial and gave written informed consent before enrollment.\nThe study enrollment was stopped in December 2009 after 100 eligible patients had been recruited.\nPatients\nAdult patients (>18 years) were recruited from Siriraj Hospital if characteristic rhabditiform larvae of S. stercoralis were present on fecal microscopy.\nExclusion criteria included a history of allergic reaction to either study medication, treatment within the month prior to the study with any drug known to have anti-strongyloides activity, pregnancy or lactation and any suggestion of disseminated strongyloidiasis.\nTreatment\nComputer generated, simple, random allocation sequences were prepared for 3 study groups by the investigator team.\nThese were sealed in an opaque envelope and numbered.\nThe investigator (YS) assigned study participants to their respective treatment group after opening the sealed envelope.\nOnce an eligible patient was identified and informed consent was obtained, the patient was randomly allocated to one of the following group (1\u22361\u22361 ratio):\nIvermectin-I group: a single oral dose of 200 \u00b5g/kg (Vermectin\u00ae, Atlantic Laboratories Co, Ltd., Thailand).\nIvermectin-II group: two oral doses of 200 \u00b5g/kg of ivermectin (Vermectin\u00ae, Atlantic Laboratories Co, Ltd., Thailand) given 2 weeks apart.\nAlbendazole group: oral albendazole (Albatel\u00ae, TO Chemical, Thailand) 400 mg twice daily for 7 days.\nStudy Procedures\nBaseline evaluation included history, detailed physical examination, and laboratory investigations such as complete blood count (CBC), urinalysis, and biochemistry.\nPatients were requested to collect two consecutive fecal samples at every hospital visit.\nThe coprodiagnosis for the detection of S. stercoralis larvae using direct smear, formol-ether concentration method, and modified Koga agar plate culture method was performed for each patient at the Infectious Diseases and Tropical Medicine Laboratory, Division of Infectious Diseases and Tropical Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Thailand.\nPatients were required to make seven hospital visits to complete the study: at baseline evaluation and initiation of treatment, at 2 weeks after initiation of treatment, then at 1 month, 3 months, 6 months, 9 months, and 1 year after treatment.\nPatients who completed 1 year of follow-up were invited for further follow-up visits every 3 months or at their convenience.\nOutcome measures\nEfficacy\nThe efficacy of treatment was analyzed on a modified intention to treat, and a per-protocol basis.\nModified intention to treat analysis was based on the number of patients who were randomized and received treatment.\nPer-protocol analysis was based on the number of patients who completed the treatment and were followed up as planned.\nAnalyses of adverse events were on the modified intention to treat basis.\nThe outcome was evaluated according to the following definitions; \u201cCure\u201d was defined as clinical improvement (if symptomatic before treatment) and the absence of rhabditiform larvae in the stool at day14 of treatment and throughout the follow-up period.\n\u201cFailure\u201d was defined as the presence of larvae two weeks after initiation of treatment or the reappearance of larvae during follow-up.\n\u201cAdverse events\u201d were defined as symptoms or signs that developed after the study drug administration and had not been reported prior to the administration of the first dose of the antihelmintic.\nSample Size and Statistical Methods\nAssuming the therapeutic efficacy of albendazole to be 60% and of both regimens of ivermectin treatment to be 90%, with alpha error 0.5, and 80% power, it was calculated that 22 patients would be needed in each study group.\nThe lost to follow up rate was assumed to be 20%.\nTherefore 27 patients would be needed in each study group.\nDemographic data, results of investigations, stool examination at baseline and follow up visits were recorded in the database.\nAll statistical analyses were performed using SPSS, version 17.5.\nPearson chi square or Fisher's exact tests were used to compare rates and proportions, as appropriate.\nMann-Whitney U tests were used to analyze continuous variables that were not normally distributed.\nIndependent sample t- tests were used to compare normally distributed variables, taking a probability of less than 5% as the level of significance.\nKaplan-Meier plot and Cox proportional hazard model were constructed to identify independent risk factors for treatment failure.\nResults\nOne hundred and fifty one patients had detectable rhabditiform larvae of S. stercoralis on fecal microscopy during the study period.\nOne hundred patients were enrolled (36, 32, and 32 patients in albendazole, ivermectin-I, and ivermectin-II groups respectively).\nTen patients were excluded from analysis because they did not receive or complete the study treatment (3 in albendazole group, 2 in ivermectin-II group), or they were lost to follow-up immediately after treatment (3 in albendazole group, 1 each in ivermectin-I and ivermectin-II respectively).\nOverall, 90 patients were eligible for the modified intention to treat analysis.\nDetail of the total number of enrollment, randomization, follow-up and inclusion in the final analysis comparing among the three treatment groups is shown in Figure 1.\nThe demographic data, concomitant diseases, baseline clinical and laboratory investigations are shown in Table 1 and 2.\nAll except one patient had an associated medical problem, including concurrent other parasitoses.\nThese patients also had abnormal serum aspartate aminotranferase (AST) and alanine aminotransferase (ALT) levels prior to entering the study due to their underlying conditions.\nThe intensity of initial infection of the three study groups was similar, i.e. S. stercoralis larvae were found from the direct fecal examination in 24 (80%), 25 (80.6%), and 28 (96.6%) in albendazole, Ivermectin-I, and ivermectin\u2013II groups, respectively (P\u200a=\u200a0.123).\nLarvae were also detected from modified Koga agar plate culture in 22/26 (84.6%) patients in the albendazole group, in 22/26 (84.6%) patients in the ivermectin-I group, and in 24/29 (82.8%) patients in the ivermectin-II group, respectively (P\u200a=\u200a0.976).\nDiarrhea was detected in half of the patients and it was relieved after treatment in most patients.\nAbnormal bowel movement at second week of follow-up was reported in 4 patients in the albendazole group, 2 patients in the ivermectin-I group, and 3 patients in the ivermectin-II group, respectively (P\u200a=\u200a0.641).\nS. stercoralis larvae were detected in one patient in the ivermectin-I group at third month of follow-up.\nIn ivermectin-II group, S. stercoralis larvae were detected at second week prior to the second dose of ivermectin in two patients.\nNo patients had reinfection/relapse after the second dose of ivermectin treatment.\nIn albendazole treated patients, S. stercoralis larvae were detected at second week of follow-up in 2 patients, at first month of follow-up in 2 patients, between 3\u20136 months of follow-up in 3 patients, and between 6\u201312 months of follow-up in 4 patients.\nAll of the relapses/ reinfections found during follow-up were clinically inapparent.\nParasitologically, parasite elimination was documented in 19 (63.3%) albendazole treated patients, in 30 (96.8%) single-dose ivermectin treated patients, and 27 (93.1%) two-dose ivermectin treated patients (P\u200a=\u200a0.006) (Table 3).\nCox regression analysis showed that albendazole treated group had 14.7 times (95%CI 1.8\u2013111.9), and 5.7 times (95%CI 1.3\u201325.7) higher risk for reinfection/ relapse of strongyloidiasis than ivermectin-I and ivermectin-II group, respectively.\nKaplan- Meier Plot compares the parasitological cure rate between these study groups is shown in figure 2.\nNo hyperinfection syndrome or disseminated infection was found among these patients during the study period.\nS. stercoralis larvae were detected after treatment using FEC in 8 patients, and by modified Koga agar plate culture only in 6 of them.\nAll patients with relapse/reinfection were retreated with two doses of ivermectin in two weeks apart.\nOverall albendazole and ivermectin were well tolerated.\nTransient elevation of AST, and ALT levels was detected in one patient in ivermectin-II group.\nThe AST and ALT levels returned to normal 2 weeks after the second dose of ivermectin treatment.\nSevere nausea and vomiting was reported in one patient in the albendazole group.\nFifteen patients died after enrollment (5 patients in each treatment group).\nCauses of death were not related to the study drugs, and were considered to be due to an underlying disease or its complications (solid tumor in 5, hematologic malignancies in 3, diabetes mellitus, or systemic lupus erythrematosus (SLE), or hypertension with complications such as myocardial infarction or sepsis in 7 patients).\nThe median duration from enrollment to death was 2 weeks (range 2\u201314 weeks) in the albendazole group, 5 weeks (range 2\u201338 weeks) in the ivermectin-I group, and 2 weeks (range 1\u201327 weeks) in the ivermectin-II group, respectively.\nDiscussion\nStrongyloidiasis remains a significant health problem in many developing countries, mainly due to the potential for lethal disseminated disease.\nGastrointestinal symptoms associated with strongyloidiasis found in this study included diarrhea, abdominal discomfort, nausea/vomiting and anorexia.\nChronic infection with S. stercoralis was clinically inapparent in half of the patients at enrollment, and in all of relapses/ reinfections found during follow up.\nPeripheral eosinophilia (>500 eosinophils/\u00b5L.) was detected in half of the patients at enrollment.\nS. stercoralis larvae were detected after treatment using FEC in 8 patients, and by modified Koga agar plate culture in 6 patients.\nThis information confirmed that fecal examination, including culture and/or serology, every 3\u20136 months of follow-up should be recommended for early detection and treatment of latent infection to prevent hyperinfection or disseminated disease in these patients.\nResults of this study corroborate the results from previous randomised controlled studies on the higher efficacy of ivermectin compared to various dosage regimens of albendazole for treating chronic strongyloidiasis.\nA summary of results from these previous controlled trials of ivermectin treatment for chronic strongyloidiasis is shown in Table 4.\nAlthough these studies were conducted in different geographical areas and population groups, i.e. in children and adults, they were considered to be within a community-based setting, such as schools or primary care clinic.\nThe duration of follow-up varied from 3 weeks to 12 months.\nThe present study was conducted in a tertiary hospital.\nThe majority of patients had known risk factors for disseminated strongyloidiasis, and approximately one-third of them received corticosteroid or chemotherapy.\nResults of this study confirmed that ivermectin was also effective in this population who were at high risk of severe infection.\nAlbendazole remains an option of treatment for chronic strongyloidiasis in many countries in South East Asia, where oral ivermectin is not widely available.\nCure rates of a regimen consisting of albendazole 400 mg daily for three to five days varied from 38\u201387% in those without underlying diseases.\nIn this study, the cure rate was found to be 63% when a 7-day course of high dose albendazole was used.\nThe efficacy of albendazole varied widely even when the same dose and duration of treatment was used.\nDifferences in duration of follow-up examinations could be one explanation, and re-infection from the environment may also be a factor when the efficacy is monitored for an extended period in endemic areas.\nThe study which reported the highest cure rate (87%) was conducted in Okinawa, Japan, where the chance of re-infection from the environment was less likely to occur compared to other studies conducted in endemic areas.\nTwo patients in the ivermectin-II group had detectable S. stercoralis larvae in the second week prior to the second dose of ivermectin treatment.\nOne patient in ivermectin-I group also had detectable S. stercoralis larvae 3 months after treatment.\nThis observation supports the recommendation that repeated doses of ivermectin should be the preferred treatment in patients with chronic strongyloidiasis who have an underlying or concomitant illness.\nThe limitation of this study was the high loss to follow-up rates over time.\nHigh mortality associated with the concomitant illnesses was an unavoidable cause of concern in this study.\nThe median duration of follow up was 19 weeks in albendazole group, 39 weeks in ivermectin-I, and 26 weeks in ivermectin-II group.\nThe non-significant shorter duration of follow up found in albendazole treatment group was due to the significant higher rate of treatment failure compared to ivermectin.\nHowever, the study still had sufficient power to detect a difference between albendazole and ivermectin treatments.\nThis study, however, was too small to detect any but the most severe and common side- effects of both albendazole and ivermectin.\nOnly one of albendazole treated patients and one treated with ivermectin had transient changes in transaminases, a well-recognized and reversible adverse event.\nIn conclusion, this clinical study confirms that both a single and a double dose of oral ivermectin taken at a two-week interval is more effective than a 7-day course of high dose of albendazole for patients with chronic infection due to S. stercoralis.\nTotal number of enrollment, randomization, follow-up, and inclusion in the final analysis comparing among three treatment groups.\nKaplan-Meier Plot comparing the parasitological cure among the albendazole, ivermectin-I, and ivermectin-II treatment groups over one year follow-up period.\n\nDemographic and baseline clinical features of the three study groups.\nCharacteristics | Albendazol(N\u200a=\u200a30) | Ivermectin-I(N\u200a=\u200a31) | Ivermectin \u2013II(N\u200a=\u200a29) | P-value\nMale: female | 21\u22369 | 21\u223610 | 19\u223610 | 0.934\nMedian (range) age, yr | 54 (23\u201381) | 51 (29\u201377) | 52 (25\u201378) | 0.273\nMedian (range) weight, kg | 59 (37\u201380) | 57 (37\u201385) | 59 (44\u201373) | 0.75\nConcomitant illnesses, n (%) |  |  |  | 0.607\n- None | 1 | 0 | 0 | \n- Diabetes mellitus | 8 | 6 | 5 | \n- NS/SLE | 4 | 5 | 4 | \n- AIDS/HIV infection | 3 | 2 | 5 | \n- Hematological malignancy | 2 | 3 | 4 | \n- Solid tumor | 3 | 1 | 6 | \n- Rheumatologic diseases | 2 | 1 | 1 | \n- Chronic kidney disease | 3 | 2 | 2 | \n- Alcohol drinker | 3 | 1 | 1 | \n- Others | 6 | 12 | 4 | \nImmunosuppressive drug, n (%) | 10 (33.3) | 11(35.5) | 11 (37.9) | 0.934\nConcomitant parasitoses* |  |  |  | 0.380\n- Hookworm infection | 2 | 1 | 0 | \n- Opisthorchiasis | 2 | 1 | 2 | \n- Enterobious infection | 0 | 1 | 0 | \n- Entamoeba histolytica infection | 0 | 1 | 0 | \n- Isospora belli infection | 0 | 0 | 1 | \n\n*Diagnosis obtained by ova or cyst found from fecal examination.\n\nComparison of symptoms related to chronic strongyloidiasis and baseline laboratory results.\nParameters | Albendazol(N\u200a=\u200a30) | Ivermectin-I(N\u200a=\u200a31) | Ivermectin \u2013II(N\u200a=\u200a29) | P-value\nSymptoms associated with strongyloidiasis, n (%) |  |  |  | \n- Diarrhea | 14 (46.7) | 11 (35.5) | 16 (51.6) | 0.609\n- Abdominal pain | 3 (10) | 4 (12.9) | 6 (20.7) | 0.483\n- Nausea/vomiting | 4 (13.3) | 4 (12.9) | 6 (20.7) | 0.650\nLaboratory test, mean (SD) |  |  |  | \n- Hematocrit, % (35\u201345) | 32.7 (7) | 35.4 (6) | 32 (8) | 0.132\n- Eosinophil count, \u00d7106/L (<500) | 967(1,239) | 1,203(2,714) | 554(1,781) | 0.366\n- Total eosinophil >500/\u00b5L, n (%) | 14 (46.7) | 18 (58.1) | 13 (44.8) | 0.535\n- AST, U/L (0\u201337) | 45(43) | 45(60) | 38(34) | 0.805\n- ALT, U/L (0\u201340) | 38(31) | 42(38) | 38(47) | 0.924\n- Creatinine, mg/dL (0.8\u20131.2) | 1.2(0.5) | 1.1(0.9) | 0.8(0.8) | 0.933\n\n\nOutcome of treatment among the three study groups.\nParameters | Albendazol(N\u200a=\u200a30) | Ivermectin-I(N\u200a=\u200a31) | Ivermectin-II(N\u200a=\u200a29) | P-value\nDuration of follow up |  |  |  | \n- Median (range), weeks | 19(2\u201376) | 39(2\u201374) | 26 (2\u201374) | 0.248\nOutcome: Parasitological responses |  |  |  | 0.006\n- Elimination, n (%) | 19 (63.3) | 30 (96.8) | 27 (93.1) | \n- Failure | 11 (36.7) | 1 (3.2) | 2 (6.9) | \n- Persistence at 2 week | 2 | 0 | 2 | \n- Relapse /reinfection | 9 | 1 | 0 | \n\n\nSummary of published controlled trials of oral ivermectin treatment for chronic strongyloidiasis.\nComparative Drug, Dosage Regimen | Durationof follow-up | N | Cure,N (%) | Author,year, [Ref]\n1. Albendazole 400 mg/d - 3days2. Ivermectin 150\u2013200 \u00b5g/kg, single dose | 30 days | 2429 | 9 (38)24 (83) | Datry A,1994 \n1. Thiabendazole 50 mg/kg/day - 3 days2. Ivermectin 200 \u00b5g/kg, single dose3. Ivermectin 200 \u00b5g/kg, - 2 days | 7 days, then1, 3, 6months | 191618 | 18(94.7)16(100)18(100) | Gann PH,1994 \n1. Albendazole 400 mg/d -3 days2. Ivermectin 200 \u00b5g/kg, single dose | 3 weeks | 149152 | 67(45)126(82.9) | Marti H,1996 [18]\n1. Pyrvinium pamoate 5 mg/kg/d -3 days2. Albendazole 400 mg/d - 3days3. Ivermectin 6 mg 2 doses- 2 weeks apart | 2 weeks,then 6, 12months | 608467 | 14 (23.3)65 (77.4)65 (97) | Toma H,2000 \n1. Albendazole 400 mg/d - 5 days2. Ivermectin 150\u2013200 \u00b5g/kg, single dose | 30 days | 3378 | 26(78.8)77(98.7) | Nontasut P,2005 \n",
        "label": "low",
        "id": "task4_RLD_test_966"
    },
    {
        "paper_doi": "10.1186/1471-2458-13-256",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: DesigncNON-RCTAllocation of clusters4 schools allocated to intervention, 3 to control\n\n\nParticipants: 341 children ages 6 to 7\n\n\nInterventions: Single WASH aspect\n\n\nOutcomes: Any STH\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Background\nChild health in many low- and middle-income countries lags behind international goals and affects children\u2019s education, well-being, and general development.\nLarge-scale school health programmes can be effective in reducing preventable diseases through cost-effective interventions.\nThis paper outlines the baseline and 1-year results of a longitudinal health study assessing the impact of the Fit for School Programme in the Philippines.\nMethods\nA longitudinal 4-year cohort study was conducted in the province of Camiguin, Mindanao (experimental group); an external concurrent control group was studied in Gingoog, Mindanao.\nThe study has three experimental groups: group 1\u2014daily handwashing with soap, daily brushing with fluoride toothpaste, biannual deworming with 400\u00a0mg albendazole (Essential Health Care Program [EHCP]); group 2\u2014EHCP plus twice-a-year access to school-based Oral Urgent Treatment; group 3\u2014EHCP plus weekly toothbrushing with high-fluoride concentration gel.\nA non-concurrent internal control group was also included.\nBaseline data on anthropometric indicators to calculate body mass index (BMI), soil-transmitted helminths (STH) infection in stool samples, and dental caries were collected in August 2009 and August 2010.\nData were analysed to assess validity of the control group design, baseline, and 1-year results.\nResults\nIn the cohort study, 412 children were examined at baseline and 341 1\u00a0year after intervention.\nThe baseline results were in line with national averages for STH infection, BMI, and dental caries in group 1 and the control groups.\nChildren lost to follow-up had similar baseline characteristics in the experimental and control groups.\nAfter 1\u00a0year, group 1 showed a significantly higher increase in mean BMI and lower prevalence of moderate to heavy STH infection than the external concurrent control group.\nThe increases in caries and dental infections were reduced but not statistically significant.\nThe results for groups 2 and 3 will be reported separately.\nConclusions\nDespite the short 1-year observation period, the study found a reduction in the prevalence of moderate to heavy STH infections, a rise in mean BMI, and a (statistically non-significant) reduction in dental caries and infections.\nThe study design proved functional in actual field conditions.\nCritical aspects affecting the validity of cohort studies are analysed and discussed.\nTrial registration\nDRKS00003431 WHO\nUniversal Trial Number U1111-1126-0718\nBackground\nThe health and education of children are a public good that lies at the core of government policies and programmes.\nThe Millennium Development Goals have encouraged significant resource allocation to these two sectors, which are closely related to long-term poverty reduction and development, and much progress has been made.\nStill, many low- and middle-income countries are unlikely to reach the health- and education-related targets to which they have committed themselves.\nThe 2011 United Nations Millennium Development Goals Report clearly states, \u201cDespite real progress, we are failing to reach the most vulnerable\u201d.\nAs a low-middle-income country, the Philippines is one such case: there are persistent high levels of preventable diseases among children in addition to poor primary education indicators.\nIll health is the main reason for school absenteeism and dropouts (about 40% of dropouts are due to illness; toothache is the most common reason for absenteeism).\nFilipino children mainly suffer from a few widespread diseases: diarrhoea, pneumonia, and respiratory infections are the leading cause of death (82,000 children in the 5- to 12-year age group die from these every year), and 54% of children are infested with soil-transmitted helminths (STH).\nOne-third of children are stunted, and 17% have a below-normal body mass index (BMI).\nThe prevalence of dental caries is extreme: more than 97% of 6-year-old children suffer from tooth decay; it usually goes untreated, which leads to very high rates of dental infection (85% of 6-year olds show such signs).\nEssential health care programme in the Philippines\nSchool health programmes have the potential to contribute significantly to preventing and controlling key diseases in children, particularly if the programmes are able to systematically reach mass numbers of children, thereby producing benefits for both health and education.\nThe Philippine Essential Health Care Program (EHCP), also known as the Fit for School Programme, aims to address some of the major diseases that affect the Filipino child population through simple, evidence-based, integrated approaches.\nThe EHCP currently targets more than 2 million children in the Philippines.\nThe EHCP was developed by the Health and Nutrition Center of the Department of Education Central Office and the Philippine National Institutes of Health.\nHowever, it operates in the greater context of official development assistance linking Philippine and European universities.\nAs a result, the EHCP was developed in close cooperation with the following organizations: the German Development Cooperation (GIZ); the World Health Organization (WHO) Collaborating Center for Prevention of Oral Diseases at the University of Jena in Germany; the former WHO Collaborating Center on Oral Health Care and Future Scenarios in Nijmegen, the Netherlands; and Xavier University in Cagayan de Oro, Philippines.\nthe EHCP is part of a bigger context of official development assistance linking in Philippine and European universities.\nThe EHCP aims to improve child health and development by institutionalization of three preventive interventions within public elementary schools in the Philippines: daily supervised handwashing with soap and clean water; daily supervised brushing with fluoride toothpaste; and biannual deworming via mass drug administration.\nThis innovative approach is conceptually based on the Fit for School Action Framework which outlines the principles of simplicity, scalability, and sustainability.\nThe approach was awarded by the Worldbank, UNDP and the WHO for innovation in global health in 2009.\nFit for School health outcome study\nThe Fit for School Health Outcome Study (FITHOS) is a longitudinal cohort survey whose objective is to provide data on the impact of established interventions that are part of the EHCP and implemented in schools using the manpower and financial resources of the government school system.\nThe FITHOS is financed by GIZ and conducted by local institutions (such as the regional Department of Education Health and Nutrition Unit in Cagayan de Oro).\nThis study is part of efforts to improve local capacity in applied research, and it aims to provide evidence for informed management and policy decisions.\nThe FITHOS is a large study, and it includes a range of health and health-related parameters, e.g. quality of life, school days lost because of illness, and education performance.\nFurthermore, two interventions toward controlling dental caries are also part of the FITHOS in order to assess their impact and feasibility in the school context.\nThe details of those dental interventions, however, are not presented in this paper owing to lack of space and will be reported in a separate publication.\nWe will introduce and discuss the methodology of this community trial and briefly present the baseline and 1-year data as the direct health outcomes of the three integrated interventions of the EHCP.\nMethods\nStudy design\nThe study applies a longitudinal cohort design over a period of 4\u00a0years with an external concurrent control group and an internal (within-school) non-concurrent control group (Figure\u00a01).\nThe study started in 2009 with baseline data collection among first-grade students (6\u20137\u00a0years old) of public elementary schools on the island province of Camiguin.\nSubsequent annual data collection has been scheduled for August of every following year up to 2013 (with optional extension until 2014).\nCamiguin was selected because of its comparably low migration rate and stable socioeconomic indicators.\nBased on a local government decision to implement the EHCP in the entire province, it was not possible to assign a control group with children not participating in the EHCP in that province.\nAn external concurrent control group was therefore identified in Gingoog, Misamis Oriental, Northern Mindanao.\nThe local government authorities of Gingoog had no intention of implementing the EHCP during the study period, and the local child population was similar to that in Camiguin in terms of health and socioeconomic characteristics.\nPublic elementary schools participating in the study (experimental and control groups) were selected based on two criteria: (1) location along a highway or no more than 1 kilometre from a highway; and (2) no problems related to law and order in the surrounding community.\nThe calculation of the study cohort size was based on the following assumptions: a 50% reduction in the prevalence of moderate to heavy STH infections; an assumed caries prevented fraction (using the Decayed-, Missing-, Filled Surfaces Index [DMFS] control \u2013 caries DMFS intervention / caries DMFS control \u00d7 100%) of 30% in permanent first molars over the study period; and an anticipated dropout rate of 40% during that period.\nThough a group of around 100 children would have allowed for statistical power of 80% and 95% confidence intervals, a group size of 200 children at baseline was considered safer and sufficient to cover all eventualities.\nFor this sample size, four public elementary schools in Camiguin (of a total of 56 schools on the island) were randomly assigned to the EHCP experimental group based on the above two criteria.\nThe Department of Education Health and Nutrition Unit selected three schools in Gingoog as concurrent control schools (not participating in the EHCP).\nChildren with systemic medical conditions and other chronic infectious diseases, such as tuberculosis, were excluded from the study.\nExperimental groups\nThe study protocol defined three experimental groups, as detailed below.\nHowever, this paper reports only on the baseline and 1-year results of experimental group 1.\nExperimental groups 2 and 3 were used to test additional interventions in controlling caries, and those results will be presented in a separate paper.\nExperimental group 1\nThis group participates in the EHCP, which includes the following: daily supervised handwashing with soap and clean water (as a scheduled group activity); daily supervised brushing with a fluoride toothpaste (0.3\u00a0ml; 1,450\u00a0ppm free available fluoride, scheduled group activity); and biannual deworming with a single dose of albendazole (400\u00a0mg) as a mass drug administration at school.\nThese interventions are implemented by education staff (teachers for daily tasks, school health nurses of the Department of Education for orientation and supervision) as defined in the EHCP protocol.\nExperimental group 2\nThe second experimental group participates in the EHCP (like experimental group 1), and it has access to Oral Urgent Treatment (as defined by WHO), in which treatment is offered twice a year at every school for children suffering from toothache due to advanced dental caries.\nThis on-demand treatment includes tooth extractions, drainage of abscesses, and drug administration in selected cases.\nExperimental group 3\nThe third experimental group participates in the EHCP (like experimental group 1), and it uses a high-concentration fluoride gel (Elmex Gel\u00ae, Gaba GmbH L\u00f6rrach, Germany in a dispenser, releasing 0.3\u00a0ml per usage; total fluoride concentration 12,500\u00a0ppm) once a week instead of regular toothpaste.\nChildren use normal toothpaste on the other 4 school days.\nThe fluoride gel is applied on the toothbrush and children brush with it after normal brushing for 2\u00a0minutes.\nControl groups\nExternal concurrent control group (Group C1)\nLocated in Gingoog City, this cohort of children receives a standard health education programme as defined by the Department of Education.\nIt consists of an annual physical examination, biannual deworming carried out by school nurses, the distribution of a single (10-ml) commercial toothpaste sachet, a toothbrush, and an oral health message at the beginning of the school year, and health education as part of the regular school curriculum (Figure\u00a01a).\nInternal non-concurrent control group (Group C2)\nThis group consists of children in the participating EHCP schools in grades 2 and 4 whose data were collected during the baseline assessment.\nSince the intervention child cohort from the baseline moves from grade 1 to grade 5 during the study period, the data from children in grades 2 and grade 4 are used to serve as a control for the experimental group (Figure\u00a01b).\nExaminer training and calibration\nOne week before the start of the study, a 2-day examiner training session was conducted.\nSchool nurses were trained in standardized collection of stool samples, obtaining height and weight data, and regular calibration of weighing scales.\nOral examinations were conducted by three dentists after 2\u00a0days of theoretical and clinical training in the diagnosis methodology.\nOne of the dentists had participated in the last national oral health survey and was used as the gold standard.\nTo assess reproducibility, 7.5% of children were examined twice.\nAll examiners were blind to the different groups.\nExperienced examiners used a standard questionnaire to collect personal and socio-demographic data.\nAnthropometric measures\nTrained school nurses took all measurements according to standard guidelines.\nChildren removed their shoes and stood upright for measurement of height using a portable stadiometer (Seca, Hamburg, Germany) to the nearest 0.5\u00a0cm.\nTheir weight was measured with a portable weighing scale to the nearest 0.1\u00a0kg (Detecto\u00ae Cardinal Scale Manufacturing Co., Webb City, USA).\nGenerally the children were only lightly dressed, so no adjustments were made for clothing.\nAnthropometric measuring equipment was re-calibrated at the beginning of each day and after every 10th child using a plastic bag filled with 6\u00a0kg of sand.\nHeight and weight were used to compute BMI for age.\nThe BMI of each child was calculated as the body weight in kilos divided by the height in metres squared\u2014weight (kg)/height (m2).\nThe results were grouped as normal, below-, and above-normal BMI according to the sex- and age-related cut-off points of Cole et al..\nParasitological examination\nEach child submitted a stool sample, which was labelled, coded, and sent daily (by courier) to the laboratory of the National Institutes of Health, University of the Philippines, in Manila.\nSamples were examined to determine the prevalence and intensity of STH infection using the Kato-Katz method.\nCut-off points defined by WHO were used to classify light-, moderate-, and heavy-intensity infections.\nFor quality control of parasitological examinations, 10% of all slides were randomly selected and re-examined by a reference microscopist.\nOral examination\nAn assessment was made of the prevalence of dental caries and oral infections.\nThe children brushed their teeth before the examination.\nOral examinations were performed in the schoolyard in the open air, with children lying supine on benches taken from a classroom.\nMouth mirrors (lighted mouth mirror Mirrorlight\u2122, Kudos, Hong Kong) and a CPI ball-end probe were used as examination tools to score caries according to the WHO basic methods for epidemiological oral health surveys Decayed-, Missing-, Filled Index (DMF).\nTeeth with early stages of caries, but where the ball-end probe was unable to enter, were not scored as caries and excluded from analysis.\nOral infections were recorded according to criteria for the PUFA index : PUFA measures the consequences of untreated dental caries, such as open pulp, ulceration, fistula and abscess, with PUFA values representing permanent teeth and pufa values primary teeth.\nStatistical analysis\nThe data were analysed using SAS 9.1 software (SAS Institute, USA).\nInter-examiner reproducibility for caries score and dental infection (PUFA) at baseline and at follow-up examinations was calculated using kappa statistics.\nThe reproducibility of parasitological examinations was assessed with sensitivity and specificity values.\nThe outcome variables in the longitudinal design are presented as mean increments: percentages for prevalence data and means for caries score (DMFS and PUFA) and BMI (Table\u00a01).\nThe outcome variables in the cross-sectional design are presented as means: percentages for prevalence data and means for caries scores and BMI (Table\u00a02).\nFor differences between mean (increment) data, Student\u2019s t test was applied.\nFor differences between percentages of increment data below normal BMI, the chi-square test was applied with 2 \u00d7 4 cells, where each group was divided in: normal remains normal, below normal remains below normal, normal moves to below normal, and below normal moves to normal.\nFor differences between percentages of increment data of moderate to heavy STH infections the chi-square test was applied with 2 \u00d7 4 cells, where each group was divided in: heavy remains heavy, low remains low, heavy moves to low, and low moves to heavy.\nEthical considerations\nThe study protocol was reviewed and approved by the Institutional Review Board of the Kinaadman Research Center of Xavier University in Cagayan de Oro, Philippines, and it fully complies with the Philippine National Ethical Guidelines for Health Research and the Code of Conduct of the German Development Cooperation GIZ, the financing organization.\nWritten consent was obtained from parents or caregivers of children participating in the study.\nThe study is registered with the German Clinical Trial Register DRKS (DRKS00003431) and the WHO system (WHO Universal Trial Number U1111-1126-0718).\nBased on the Institutional Review Board recommendation, the external concurrent control group of the study also receives a standard intervention (see above) so that no child participating in the study is deprived of potential programme benefits.\nResults\nReproducibility assessments\nThe kappa values for inter-examiner reliability for oral examinations were 0.91 and 0.93 for baseline and 1-year data collection, respectively.\nSensitivity and specificity for the Kato\u2013Katz test, both at baseline and 1-year data collection, were an average 84.6% sensitivity (confidence interval 68.8\u201393.6%) and 96.8% specificity (confidence interval 81.5\u201399.8%) for the diagnosis of STH infections.\nBaseline data collection\nA total of 200 children (mean age 6.47\u00a0years, 52% male) were examined in the experimental group and 212 children (mean age 6.37\u00a0years, 47.1% male) in the external concurrent control group.\nThe baseline data of the experimental cohort and external concurrent control cohort did not show statistically significant differences except for the prevalence of moderate to heavy STH infections (Table\u00a03).\nCharacteristics of children lost to follow-up\nIn all, 32 children were lost to follow-up in the experimental group and 39 in the external concurrent control group.\nMore boys dropped out than girls; otherwise, the socio-demographic and clinical parameters of the dropouts were similar to those of the children at baseline in both groups.\nLongitudinal design\u2014comparison between experimental group 1 and external concurrent control group after 1 year\nIn all, 168 children (mean age 7.56\u00a0years, 50.6% male) were examined in the experimental group and 173 in external concurrent control group (mean age 7.39\u00a0years, 45.7% male).\nThe mean BMI in experimental group 1 increased whereas it remained unchanged in the control group.\nThe increase in mean BMI was higher (statistically significant) in the experimental group than in the control group (Table\u00a01).\nThe prevalence of children with low BMI in the experimental group decreased whereas it increased in the control group.\nIn both groups, the prevalence of moderate to heavy STH infection decreased, but it was more pronounced in the control group.\nIncreases in the DMFS (with no changes in the missing or filled index component) and PUFA indexes for permanent first molars over 1\u00a0year were lower in the experimental than in the control group (Table\u00a01).\nThe prevented fraction of increases in these indexes was 17% and 42%, respectively.\nThe latter approximates to statistical significance (p value = 0.068).\nCross-sectional design\u2014comparison between experimental group 1 and internal non-concurrent control group\nTable\u00a02 presents the data of the 168 children (grade 1) of the experimental group after 1\u00a0year and the data of 133 children (grade 2) whose data were collected at baseline.\nData comparison between the experimental group 1 and the internal non-concurrent control group did not reveal any statistically significant difference.\nDiscussion\nHealth outcome assessment of school health programmes\nLarge-scale, low-cost intervention programmes such as the EHCP have the potential to achieve population-level health effects.\nDeveloping and applying data related to public health programmes in low- and middle-income countries is an important basis for improving health.\nIn this context, primary research examining the value of health interventions among poor communities and providing crucial knowledge for informed local policy decisions is of great relevance.\nField research in low-resource settings often faces challenges related to poor logistics and local resources as well as limited available local staff; thus, research is often organized and undertaken by institutions of the global North.\nBy contrast, in the context of Official Development Assistance, the present research had a strong focus on involving and advancing local research capacities.\nThis study examined a school health programme that is part of a bilateral development project between the Philippines and Germany.\nThe broad participation of local universities and the involvement of Department of Education health have facilitated an increase in the capacity to conduct field research.\nIt is hoped that this will help promote effective health strategies among policymakers in low-resource countries.\nIt was not intended that this study would produce representative results for the entire child population participating in the EHCP throughout the country.\nThat would have required a complex sampling procedure, a much larger setup, an exponentially higher budget, and a huge management structure.\nNevertheless, the various sociodemographic and health-related parameters of the study sample were very similar to those reported in other large-scale or national child surveys in the Philippines.\nThe chosen study design provides highly valuable insights related to programme effectiveness under real-life conditions.\nStudy design and methodology\nRandomized control trials are considered the gold standard for clinical research.\nHowever, such a design is often difficult if a large-scale public health programme covers the entire study population or all schools of a province, which is the case with Camiguin Island.\nThere, the authorities did not allow a stepped-wedge randomized control trial since they wanted to introduce the programme simultaneously on the entire island.\nRandomization of children within a school was impossible and would have been unethical.\nSince an internal concurrent control school in the same province was not possible, an external concurrent control group was selected despite the problem of inherent selection bias.\nSchoolchildren in this control group received the traditional health education-based intervention carried out by school health personnel and the deworming programme.\nIn addition, an internal non-concurrent control group was included in the present study.\nThis design allowed for internal comparison and supported the data collected from the concurrent external control group.\nWith regard to the internal non-concurrent control group, evaluation of the outcomes was limited to cross-sectional comparison since it was not a cohort followed over time.\nThe inclusion of an external concurrent control group allowed the assessment of increases in disease parameters.\nGenerally speaking, cohort studies have a number of advantages but also significant limitations and sources of bias.\nIt has been suggested that four critical areas be examined when assessing the validity of a cohort study.\n1. Selection bias: Here the essential question is whether intervention and control groups are similar in all important aspects except for the intervention.\nThe children of the experimental cohort were selected on the basis of attending randomly selected schools, whereas the external concurrent control schools were assigned by the regional Department of Education Office.\nThe analysis of sociodemographic data between the experimental and control groups revealed a high degree of conformance\u2014both between each other and in relation to national averages.\nMoreover, in terms of disease burden, as measured with the different health indicators, the intervention and control cohorts were very similar\u2014except for worm load, which will be discussed in more detail below.\n2. Information bias: The study design tried to reduce information bias by limiting the choice of indicators to a few essential ones that were not too complex to measure.\nThe high consistency of examiner results and the constant re-checking through double examinations and duplicate tests indicate a low information bias.\nThe reliability of the caries diagnosis focused on reproducibility with reference to the gold standard examiner; therefore, inter-examiner kappa values were presented.\nThe examiners were blind as to the different groups, although it is probably realistic to assume that the examiners would soon have discovered that the control schools were located in a province where the EHCP did not exist.\nAppraising the viability, effectiveness, and appropriateness of the screening methods used followed the guidelines of the UK National Screening Committee.\n3. Loss to follow-up: This is a potential source of selection bias and a crucial issue for the validity of cohort studies.\nDropout resulted in the loss of 32 children in the experimental group and 39 children in the external control group.\nIt is known that dropout rates are highest for the transition between grades 1 and 2, with a national average of 14.5% of children dropping out of school.\nThe observed loss-to-follow-up rates in the intervention and control groups are thus not surprising.\nThe characteristics of children lost to follow-up in both the experimental and concurrent control group are very similar, which indicates a negligible selection bias.\n4. Confounding factors: Socioeconomic factors are usually major confounding factors for health outcomes, with poverty being the strongest determinant of health.\nThe study used a pragmatic approach to assess socioeconomic status by means of proxy measures, such as asking whether there was a TV set at home.\nCareful questioning is required in this context to avoid information bias.\nEven the seemingly simple question about the number of siblings can be difficult to answer for children who are not used to differentiating between siblings and cousins and other relatives within an enlarged family.\nReassessment of these questions after 1\u00a0year helped to trace at least some bias.\nAfter 1\u00a0year, the mean number of siblings decreased slightly by 0.11 and the number of children whose homes had a TV increased by 3.9%.\nCompliance with study protocol\nCompliance with study protocol can be a major confounding factor and needs to be carefully assessed.\nThe results related to deworming are a good example of problems stemming from logistics and lack of compliance to the protocol.\nDespite careful advance planning and orientation of study staff and participating school personnel, a deworming activity was conducted prior to baseline data collection in the intervention schools.\nAfter consultation with the National Institute of Health, it was decided that the data collection be postponed by 6\u00a0weeks to give time for partial re-infection.\nOwing to the complex logistics of the study, it was not possible to delay the baseline data collection any further.\nThis is the reason for the lower prevalence at baseline of moderate to heavy STH infection in the intervention group (17.4%) compared with the external concurrent control group (31.1%).\nThe observed difference of about 15% in the prevalence of moderate to heavy STH infection after deworming is in line with published data.\nThe fact that the reduction in the prevalence of STH infection after 1\u00a0year was significantly higher in the external concurrent control group also seems to be related to the deworming of the intervention group weeks before the baseline assessment.\nFurthermore, without knowledge of the study team, there was an 8-week delay in starting toothbrushing activities owing to problems with government procurement of supplies.\nConsiderations related to BMI\nBMI was included in the FITHOS as a derivative health indicator of the EHCP intervention for the following reasons:\n\u2022 An association between severe caries (PUFA) and low BMI was found in a representative sample of 12-year-old Filipino schoolchildren.\n\u2022 Treatment of severe caries in 5-year-old Filipino children resulted in a significant increase in BMI, but the magnitude of the effect caused by using fluoride toothpaste is unknown.\n\u2022 Medication against STH infection in combination with daily handwashing has a potential beneficial effect on BMI; however, the extent of this combined effect is not known.\nWe therefore anticipated an effect of the EHCP intervention on BMI, though the overall effect was uncertain.\nAs a result, a power estimation for BMI was not possible.\nOne-year analysis and the power of the study\nReduction in the prevalence of moderate to heavy STH infections after 1\u00a0year of medication has been found to be around 50%.\nAfter 2\u20133\u00a0years, daily school-based brushing with fluoride toothpaste has been reported to result in caries reduction of over 30%.\nThis affects the DMF scores, but for the new PUFA score the magnitude of reduction is not yet known.\nThe estimated power for STH and DMFS data in the FITHOS is based on a 4-year study period, and therefore this interim analysis after 1\u00a0year provides insufficient power for caries data.\nHowever, the main aim of this first paper on the FITHOS is to present the methodology and aspects of the study design.\nBaseline and 1-year results\nAfter 1\u00a0year, no differences could be identified in the caries status between the experimental and internal non-concurrent control groups, which may have been due to biological spreading.\nWith longer observation periods of the cohort, the masking effect of biological spreading will gradually disappear, allowing any intervention effect to become apparent.\nIn the longitudinal cohort design, despite the short evaluation period, positive trends became evident: an increase of the mean BMI; a reduction in the prevalence of moderate to heavy STH infection; and a reduction (non-significant) in caries.\nThe confounding effect of the unplanned deworming of the intervention group prior to the baseline data collection will have to be carefully observed with subsequent data collections over following years.\nSo far, the control group design has worked well, with little baseline differences among the groups across all indicators\u2014except for the deworming-related indicators for the aforementioned reasons.\nBased on this publication and the emerging data as the study continues, other FITHOS papers will report on progress and elucidate aspects of health equity, effectiveness of oral disease prevention, and quality of life.\nIn combination with a costing study, the health outcome data provide the basis for future cost-effectiveness and cost-benefit analyses.\nTogether with the survey protocol, the key lessons from the FITHOS can help in designing templates for studies in similar countries on ways to adopt the Fit for School concept.\nSuch future multi-country research may contribute to filling gaps in essential knowledge on effective school health programmes worldwide.\nConclusions\nThe emphasis of this paper was on presenting the survey methodology and discussing the relevance of this design.\nThe FITHOS has demonstrated that methodology and study design are crucial in collecting viable data on health outcomes where the objective is enhancing programme management, political decision making, advocacy, and donor accountability.\nThis study has demonstrated positive trends in the health impacts of the EHCP after just 1\u00a0year of implementation.\nThe key health outcomes thus far are related to handwashing and deworming, resulting in a lower prevalence of moderate to heavy STH infections and an increase in the mean BMI, and to toothbrushing, which has tended to reduce caries.\nIt is hoped that the data over following years will confirm and substantiate these trends by building on the solid survey protocol and functioning logistic support.\nLongitudinal cohort design with an external concurrent control group and an internal non-concurrent control group. a. Data collection & comparisons for concurrent external control group. b. Data collection & comparisons for non-concurrent internal control group.\n\nMean (\u00b1 se) baseline data, 1-year data, and incremental data for experimental group and the external concurrent control group\nIndicators | Experimental group 1 | External concurrent control group (C1) | Difference between increments (p-value)\n\u00a0 | n=168 | n=173 | \u00a0\n\u00a0 | Baseline | 1-year | increment | Baseline | 1-year | increment | \u00a0\nMean BMI | 14.70 (0.11) | 14.88 (0.13) | 0.18 (0.06) | 14.65 (0.11) | 14.62 (0.11) | \u22120.03 (0.05) | Student-t p<0.01\nPrevalence of below normal BMI | 29.2% (3.5) | 27.8% (3.5) | \u22121.4% | 31.8% (3.5) | 37.6% (3.7) | 5.8% | Chi-square NS\nPrevalence of moderate to heavy STH infection | 17.2% (2.9) | 10.7% (2.4) | \u22126.5% | 32.0% (3.5) | 17.3% (2.9) | \u221214.7 | Chi-square p<0.001\nMean DMFS in permanent first molars | 0.82 (0.12) | 1.54 (0.17) | 0.72 (0.10) | 1.12 (0.16) | 1.99 (0.24) | 0.87 (0.14) | Student-t NS\nMean PUFA in permanent first molars | 0.060 (0.02) | 0.137 (0.03) | 0.077 (0.02) | 0.087 (0.03) | 0.220 (0.05) | 0.133 (0.03) | Student-t NS P = 0.068\n\n\nMean (\u00b1 se) data for experimental group after 1\u00a0year and the internal non-concurrent control group (grade 2 at baseline examination)\nIndicators | Experimental group (1) | Internal non-concurrent control group (Group C2) | Difference between groups (p-value)\n\u00a0 | n = 168 | n = 133 | \u00a0\nMean age | 7.56 (0.04) | 7.47 (0.04) | Student-t NS\n% of boys | 50.6% (3.9) | 46.6% (4.3) | Chi-square NS\nMean number of siblings | 3.12 (0.16) | 3.07 (0.16) | Student-t NS\nPrevalence of TV ownership | 70.4% (3.5) | 76.7% (3.7) | Chi-square NS\nMean BMI | 14.88 (0.13) | 14.86 (0.12) | Student-t NS\nPrevalence of children categorized as below normal BMI | 27.8% (3.5) | 22.6% (3.6) | Chi-square NS\nPrevalence of children with moderate to heavy STH infection | 10.7% (2.4) | 12.4% (2.9%) | Chi-square NS\nMean DMFS of permanent first molars | 1.54 (0.18) | 1.53 (0.20) | Student-t NS\nMean PUFA of permanent first molars | 0.137 (0.03) | 0.188 (0.05) | Student-t NS\n\n\nMean (\u00b1 se) baseline data for experimental group (1) and external concurrent control group (C1)\nIndicators | Experimental group (Group 1) | External concurrent control Group (Group C1) | Difference between groups (p-value)\n\u00a0 | n = 200 | n = 212 | \u00a0\nMean age | 6.47 (0.04) | 6.37 (0.04) | Student\u2013t NS\n% of boys | 52.0% (3.5) | 47.1% (3.4) | Chi-square NS\nMean number of siblings | 3.34 (0.16) | 3.10 (0.13) | Student\u2013t NS\nPrevalence of TV ownership | 67.5% (3.3) | 70.3% (3.1) | Chi-square NS\nMean BMI | 14.73 (0.10) | 14.64 (0.09) | Student\u2013t NS\nPrevalence of children categorized as below normal BMI | 28.5% (3.2) | 31.6% (3.2) | Chi-square NS\nPrevalence of children with moderate to heavy STH infection | 17.4% (2.9) | 31.1% (3.2) | Chi-square p = 0.0013\nMean DMFS of permanent first molars | 0.80 (0.11) | 1.16 (0.15) | Student\u2013t NS\nMean PUFA of permanent first molars | 0.065 (0.02) | 0.090 (0.03) | Student\u2013t NS\nMean dmft primary dentition | 7.74 (0.30) | 8.27 (0.31) | Student\u2013t NS\nMean pufa primary dentition | 3.14 (0.20) | 3.11 (0.17) | Student\u2013t NS\nPrevalence dmft>0 | 97.0% (1.2) | 97.2% (1.1) | Chi-square NS\nPrevalence DMFT>0 for permanent first molars | 37.0% (3.4) | 42.0% (3.4) | Chi-square NS\n",
        "label": "high",
        "id": "task4_RLD_test_867"
    },
    {
        "paper_doi": "10.21147/j.issn.1000-9604.2020.04.06",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Accrual: June 2011 to December 2015Single centrePhase of trial: 3Study design: open-label RCTCountry or countries where the trial was conducted: ChinaMedian follow-up: 57.3 months\n\n\nParticipants: Age: median 49, range 22-64 yearsNodal status of breast cancer: 37% node positive, 63% node negativeAdjuvant or neoadjuvant: adjuvantNotable exclusion criteria: none\n\n\nInterventions: Arm 1: paclitaxel 150 mg/m2 + carboplatin AUC3 every 2 weeks for 8 cyclesArm 2: epirubicin 80 mg/m2 and cyclophosphamide 600 mg/m2 every 2 weeks for 4 cycles followed by paclitaxel 175 mg/m2 every 2 weeks for 4 cycles\n\n\nOutcomes: Primary3-year DFS rate, defined as the date of randomisation to the date of the first local/distant recurrence (in the absence of other primary malignancies)SecondaryOS, defined as the time from randomisation to death due to any causeToxicity, according to NCI-CTCAE, version 3.0\n\n\nNotes: Trial registration record: NCT01378533All randomised participants were included in analysis.Study did not report assessing the proportional hazards assumption.Funding considerations: funded by the National Key Research and Development Program of China and the Chinese Academy of Medical Science Initiative for Innovative Medicine\n\n",
        "objective": "To evaluate the benefits and harms of platinum\u2010based chemotherapy as adjuvant and neoadjuvant treatment in people with early triple\u2010negative breast cancer.",
        "full_paper": "Objective\nThe objective of this open-label, randomized study was to compare dose-dense paclitaxel plus carboplatin (PCdd) with dose-dense epirubicin and cyclophosphamide followed by paclitaxel (ECdd-P) as an adjuvant chemotherapy for early triple-negative breast cancer (TNBC).\nMethods\nWe included Chinese patients with high recurrence risk TNBC who underwent primary breast cancer surgery.\nThey were randomly assigned to receive PCdd [paclitaxel 150 mg/m2 on d 1 and carboplatin, the area under the curve, (AUC)=3 on d 2] or ECdd-P (epirubicin 80 mg/m2 divided in 2 d and cyclophosphamide 600 mg/m2 on d 1 for 4 cycles followed by paclitaxel 175 mg/m2 on d 1 for 4 cycles) every 2 weeks with granulocyte colony-stimulating factor (G-CSF) support.\nThe primary endpoint was 3-year disease-free survival (DFS); the secondary endpoints were overall survival (OS) and safety.\nResults\nThe intent-to-treat population included 143 patients (70 in the PCdd arm and 73 in the ECdd-P arm).\nCompared with the ECdd-P arm, the PCdd arm had significantly higher 3-year DFS [93.9% vs. 79.1%; hazard ratio (HR)=0.310; 95% confidence interval (95% CI), 0.137\u22120.704; log-rank, P=0.005] and OS (98.5% vs. 92.9%; HR=0.142; 95% CI, 0.060\u22120.825; log-rank, P=0.028).\nWorse neutropenia (grade 3/4) was found in the ECdd-P than the PCdd arm (47.9% vs. 21.4%, P=0.001).\nConclusions\nPCdd was superior to ECdd-P as an adjuvant chemotherapy for early TNBC with respect to improving the 3-year DFS and OS.\nPCdd also yielded lower hematological toxicity.\nThus, PCdd might be a preferred regimen for early TNBC patients with a high recurrence risk.\nIntroduction\nTriple-negative breast cancer (TNBC) is characterized by the lack of expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor 2 (HER2).\nTNBCs, which account for approximately 12%\u221220% of all invasive breast cancers, are resistant to endocrine and HER2-targeted therapy; their aggressive behavior and poor prognosis make them one of the most challenging cancers to treat.\nPostoperative adjuvant therapy for early breast cancer, which is an important part of comprehensive treatment, can reduce the risk of recurrence and metastasis.\nAt present, since the use of polygenic prognostic detection in domestic hospitals is low, the decision of adjuvant treatment for early breast cancer patients is relatively conservative.\nThe clinical application of an anthracycline sequential taxane regimen and aromatase inhibitors has also reached an expert consensus.\nSystemic chemotherapy is generally recommended by guidelines and is, thus, currently considered as a mainstay of TNBC management.\nHowever, the proposed chemotherapy regimens remain controversial.\nIn routine clinical practice, anthracycline and taxane-containing regimens are the most commonly used systemic cytotoxic regimens for TNBC patients.\nAdding platinum to neoadjuvant chemotherapy regimens not only substantially increases the pathological complete response (pCR) rate but may also improve the event-free survival (EFS) or overall survival (OS) of TNBC patients according to previous trials.\nPlatinum-based neoadjuvant\u00a0chemotherapy\u00a0may be recommended as an option in\u00a0TNBC\u00a0patients with the cost of higher hematological toxicity incidence.\nHowever, there is limited direct evidence regarding an appropriate platinum-based adjuvant chemotherapy.\nFurthermore, determination of the optimal regimen balancing well-tolerated adverse toxicity with high efficacy is difficult.\nUnderlying genetic conditions appear to play an important role in TNBC.\nBRCA1-positive tumors show distinct clinic pathological characteristics.\nSeventy percent of all BRCA1-positive breast cancers and up to 23% of BRCA2 carriers have a TNBC phenotype.\nTNBC tumors with germline BRCA (gBRCA) mutation are associated with a better response to DNA-damaging systemic regimens such as the platinum agents.\nDose-dense chemotherapy (i.e., a chemotherapy regimen in which each cycle has a shortened treatment interval) is associated with significant improvements in survival and has been considered for use in the adjuvant setting for TNBC.\nWith granulocyte colony-stimulating factor (G-CSF) support, dose-dense chemotherapy regimens at the optimal dose have been permitted at two-week intervals rather than the conventional three-week cycle in early breast cancer regimens.\nData supporting platinum-based adjuvant regimens for TNBC are scarce and are based mostly on retrospective research.\nGiven the lack of well-established prospective or randomized studies, we conducted this study to compare the efficacy and safety of dose-dense paclitaxel plus carboplatin (PCdd) with those of the commonly used dose-dense epirubicin and cyclophosphamide followed by paclitaxel (ECdd-P) as adjuvant chemotherapy treatment in Chinese TNBC patients with high recurrence risk.\nMaterials and methods\nStudy design\nThis was a randomized, open-label, single-center study conducted in Chinese females with TNBC at high recurrence risk.\nThe study was approved by the Independent Ethics Committee of the National Cancer Center/Cancer Hospital (No. CH-BC-012).\nAll interventions were performed in accordance with the Declaration of Helsinki, guidelines of the International Conference for Harmonization/Good Clinical Practice.\nThe study was registered with the ClinicalTrials.gov (No. NCT01378533).\nParticipants\nAll participating patients provided written informed consent.\nFemale patients aged 18\u221265 years who had undergone primary breast surgery for confirmed ER-negative, PR-negative, and HER2-negative breast cancer were eligible.\nER, PR and HER2 status were determined by immunohistochemistry (IHC) on patients\u2019 tumor sections.\nThe IHC cutoff for ER-negative and PR-negative status was 1% or less positive tumor cells with nuclear staining.\nHER2-negative status was determined by IHC by giving a score of 0 or 1 or by the absence ofHER2 amplification (HER2/CEP17 ratio <2.0 and HER2 copies <4.0) upon fluorescence in situ hybridization (FISH) analysis.\nER, PR and HER2 analyses were performed centrally in a single laboratory of National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences.\nPatients were selected with positive axillary lymph or with other high-risk factors for recurrence (e.g., age <35 years, grade III disease, and intravascular cancer embolus).\nFurther details regarding this study protocol are available in the Supplementary Table S1.\nRandomization and masking\nThe patients were randomly assigned to receive either the PCdd or the ECdd-P regimen.\nSimple randomization was conducted with no stratification factors and was carried out by using\u00a0random allocation sequence.\nThe patients, medical staff, and investigators were aware of treatment allocation and assessing outcomes.\nProcedures\nPatients in both study arms received treatment in two-week cycles.\nPatients assigned to the PCdd arm received paclitaxel 150 mg/m2 on d 1 plus carboplatin AUC=3 on d 2 for 8 cycles.\nPatients assigned to the ECdd-P arm received epirubicin 80 mg/m2 divided in 2 d and cyclophosphamide 600 mg/m2 on d 1 for 4 cycles followed by paclitaxel 175 mg/m2 on d 1 for 4 cycles.\nProphylactic G-CSF 3 \u00b5g/kg was administered during each cycle according to European Society for Medical Oncology (ESMO) and American Society of Clinical Oncology (ASCO) guidelines in the dose-dense setting.\nToxicities were managed through dose delays of up to 3 weeks, and dose reductions were permitted in the following events: grade 4 hematological, grade 3 or 4 non-hematological, or other protocol-specified toxic effects.\nSafety was monitored with adverse events (AEs) reports, physical examinations, regular laboratory tests and electrocardiogram assessments at the end of each cycle until the 30th day of the last follow-up cycle.\nOutcomes\nThe primary efficacy endpoint was the 3-year disease-free survival (DFS) rate, which was calculated from the date of randomization to the date of the first local/distant recurrence (in the absence of other primary malignancies).\nSecondary objectives included OS and safety.\nOS was defined as the time from randomization to death due to any cause.\nWe analyzed the DFS and OS in patients who received at least one dose of the study treatment (intention-to-treat population, ITT).\nIn the safety analysis, we evaluated the numbers and proportions of patients in each treatment arm who had any AEs, delay of chemotherapy, and dose reduction.\nAEs were graded using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 (NCI-CTCAE, version 3.0).\nIn addition, we conducted exploratory subgroup analyses according to age (\u226440 vs. >40 years), Ki-67 index (\u226430 vs. >30), tumor size (<2 cm vs. \u22652 cm), nodal status (negative vs. positive), and surgery-chemotherapy interval (<30 dvs. \u226530 d) to investigate whether the treatment effect varied by subgroup.\nSample size computation\nThe sample size was calculated based on the primary endpoint, i.e., 3-year DFS rate.\nAssuming an approximate higher proportion of 0.10 as a primary outcome in PCdd regimen (results of our preliminary clinical research demonstrated the proportion achieving 3-year DFS in the ECdd-P regimen was 80.0%), an overall sample size of 133 participants (66 in the ECdd-P arm and 67 in the PCdd arm) was calculated to achieve 80.0% power with an alpha level at 0.05, with a 5% dropout rate in each control/treatment arm.\nSince the censoring proportion during the course of the study might be higher than expected; therefore, the sample size was increased to 143 patients to ensure the target number of events would be reached in a reasonable time frame.\nStatistical analysis\nAll statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) software (Version 22.0; IBM Corp., New York, USA).\nData were presented as numbers (%) or as the mean standard deviation.\nFrequency tables were analyzed by using the\u00a0\u03c72 test.\nThe survival analysis was estimated using the Kaplan-Meier product-limit method in the ITT population.\nThe hazard ratio (HR) and 95% confidence interval (95% CI) were estimated using the Cox proportional hazard model.\nPatients not showing progression were censored at the study cutoff date.\nThe multivariable Cox model was used for subgroup analysis to explore the influence of clinical characteristics on the 3-year DFS.\nThe safety analysis set included all randomized patients who received at least one dose of the study treatment and underwent at least one post-baseline safety assessment.\nA P value of <0.05 was considered statistically significant.\nResults\nPatients\nFrom June 2011 to December 2015, 143 patients were randomly enrolled in the PCdd arm (n=70) or the ECdd-P arm (n=73).\nAfter excluding 11 patients [treatment discontinuation due to tumor progression (n=1), withdrawal after chemotherapy (n=3), and lost to follow-up (n=7)], 132 patients who completed the planned eight cycles of chemotherapy were included in the per-protocol analysis (Figure 1).\nThe data cutoff for the primary analysis was November 30th, 2018.\nBaseline characteristics were balanced between arms (Table 1).\nAll enrolled patients had an Eastern\u00a0Cooperative\u00a0Oncology\u00a0Group (ECOG) performance status score of 0\u22121, and 62.2% were postmenopausal.\nThe median age was 49 (range, 22\u221264) years.\nIn total, 111 patients (77.6%) were aged older than 40 years.\nMost patients had stage II or III disease (n=92, 64.3%), and 90.9% of patients had invasive ductal carcinoma.\nMore than 42% had T2\u2212T4 tumors, and 53 patients (37.1%) were clinically node positive.\nMore than 75% of patients had a Ki-67 proliferation index >30%.\nSurvival outcomes\nAs of cutoff date, the median duration of follow-up was 57.3 (range, 1.2\u221298.6) months, with 58.1 months in the PCdd arm and 56.1 months in the ECdd-P arm.\nIn total, 98 patients (74.2%) were followed up over 4 years.\nDuring the study period, 23 relapse events were recorded, 5 in the PCdd arm and 18 in the ECdd-P arm.\nMost events (96.2%) were observed during the first 3 years after first diagnosis.\nIn the full analysis of ITT population, patients had significantly fewer DFS events in the PCdd arm than in the ECdd-P arm (5 vs. 18; HR=0.310; 95% CI, 0.137\u22120.704; log-rank, P=0.005).\nThe 3-year DFS was 93.9% (95% CI, 88.2%\u221299.6%) in the PCdd arm and 79.1% (95% CI, 69.7\u221288.5%) in the ECdd-P arm.\nThe Kaplan-Meier curves for DFS remained separated for the rest of the 3-year follow-up (Figure 2).\nData on OS were immature.\nEight patients died in the ECdd-P arm, whereas only one died in the PCdd arm; all deaths were cancer related.\nPreliminary data showed a potential trend on a higher 3-year OS rate in the PCdd arm (98.5% vs. 92.9%; HR=0.142; 95% CI, 0.060\u22120.825; log-rank, P=0.028) (Figure 3).\nSubgroup analyses showed a consistent DFS benefit in the PCdd arm, with the difference reaching statistical significance in the following subgroups: age >40 years (HR=4.31; 95% CI, 1.42\u221213.11; P=0.010), Ki-67 index >30% (HR=3.80; 95% CI, 1.08\u221213.36; P=0.038), and clinically evaluated lymph nodes (HR=5.73; 95% CI, 1.28\u221225.65; P=0.022) ( Figure 4).\nAEs\nOverall, both regimens were well tolerated with manageable AEs.\nThere were more patients who experienced chemotherapy delay [25 (35.7%) vs. 23 (31.5%), P=0.361] and dose reduction [16 (22.9%) vs. 14 (19.2%), P=0.369] in the PCdd arm than in the ECdd-P arm, but the difference was not significant (Table 2).\nThe most frequent AEs were neutropenia, nausea and emesis.\nThe incidence of grade 3 or 4 neutropenia was significantly higher in the ECdd-P arm than that in the PCdd arm [35 (47.9%) vs. 15 (21.4%), P=0.001], while the incidence of other grade 3 and 4 AEs was similar between the two arms.\nThere was also no significant difference in the incidence of peripheral neuropathy between the two arms (Table 3).\nNo death or life-threatening event was recorded during the study or within 30 days after the last cycle of treatment.\nDiscussion\nThis open-label, randomized study achieved its primary endpoint, with a statistically significant difference in the 3-year DFS rate in patients randomized to receive PCdd as adjuvant chemotherapy for high-risk early TNBC vs. ECdd-P (93.9% vs. 79.1%; HR=0.310; 95% CI, 0.137\u22120.704; log-rank P=0.005).\nFurther, PCdd was better tolerated than ECdd-P, with fewer hematological toxicities (grade 3/4) (21.4% and 47.9%, respectively).\nCollectively, these results indicate that PCdd might be an appropriate regimen for TNBC.\nPCdd not only is superior to ECdd-P as adjuvant chemotherapy with respect to improving the 3-year DFS and OS rates but also yields lower chemotherapy-related toxicities in early TNBC patients regardless of theBRCA mutation status.\nThus, PCdd might be a beneficial standard adjuvant regimen for early TNBC patients at a high recurrence risk, as indicated herein by the clinically meaningful improvement in survival and safety.\nTo our knowledge, this is an innovative randomized clinical study to evaluate the efficacy of a dose-dense carboplatin-based regimen in the adjuvant setting for TNBC with high recurrence risk.\nTNBC may be more sensitive to platinum-based regimens.\nCarboplatin increased the pCR rate from 41% to 54% in the CALGB40603 trial and from 36.9% to 53.2% in the GeparSixto trial.\nIn the GeparSixto study, the improved pCR rate significantly increased the 3-year DFS rate from 76.1% to 85.8% (HR=0.56; 95% CI, 0.33\u22120.96; P=0.024).\nHowever, in the CALGB40603 study, the 5-year distant recurrence-free interval was 76.3% with no significant difference.\nThe randomized phase III clinical trial EA 1131 (NCT02445391) has also been designed to prove the efficacy of adjuvant cisplatin or carboplatin following neoadjuvant chemotherapy in patients with residual TNBC.\nThe BrighTNess study has also confirmed that carboplatin-containing regimen appears to have a favorable risk-to-benefit profile for patients with high-risk TNBC in the neoadjuvant setting.\nHowever, the clinical benefit of adjuvant carboplatin in TNBC has not been well-investigated.\nFor an adjuvant scenario, a retrospective, single-center study in a Swiss breast cancer center reported a 5-year relapse-free survival (RFS) of 90% in patients treated with carboplatin.\nIn the present study, the PCdd regimen achieved significantly better survival benefit (3-year DFS and OS rates) for TNBC patients in the adjuvant setting compared with historical data from standard chemotherapy regimens (60%\u221280% with taxane-based regimens, 65%\u221285% with anthracycline- and taxane-based therapy, and 83.7% with anthracycline-based chemotherapy plus bevacizumab).\nA dose-dense regimen has been hypothesized to minimize residual tumor burden compared to dose escalation and serve as a more effective method for high-risk breast cancer.\nIn the CALGB9741 trial, the 4-year DFS rate was 82% in the dose-dense group.\nA previous study from our institution also compared the epirubicin and cyclophosphamide followed by paclitaxel (EC-P) or epirubicin plus paclitaxel (EP) dose-dense group and the EP regular group regarding postoperative adjuvant treatment for high-risk breast cancer.\nThe dose-dense group had higher 3-year RFS rates (84.1% vs. 80.0%, P=0.501) and OS rates (95.6% vs. 90.0%, P=0.153).\nOur trial is a novel prospective study showing significant improvements in the 3-year DFS and OS rates by using a dose-dense anthracycline-free platinum-based adjuvant chemotherapy regimen for TNBC regardless of the BRCA mutation status.\nThe 3-year DFS (93.9%) and OS (98.5%) rates in the PCdd arm were also superior to those of a dose-dense regimen reported by the Early Breast Cancer Trialists\u2019 Collaborative Group (EBCTCG).\nAlthough the survival data in our study are immature at present, a relatively long follow-up time will allow us to report a beneficial trend in OS.\nIn addition, these data are comparable to previous data on anthracycline- and taxane-based dose-dense regimens.\nBecause the TNBC phenotype is closely associated with hereditary breast cancer, the administration of platinum-based regimens has received a new impetus.\nHowever, in the Chinese population, BRCA1/2 mutations are prevalent in only 10.5% of TNBC patients younger than 50 years.\nThe benefit of adjuvant carboplatin in TNBC with BRCA1/2 mutation(s) is still controversial.\nThe GeparSixto trial showed that carboplatin is more effective in TNBC patients; however, a secondary analysis of the GeparSixto demonstrated that TNBC patients without BRCA1 and BRCA2 germline mutations would also benefit from the addition of carboplatin, which increased the DFS rate (85.3% in the carboplatin group and 73.5% in the non-carboplatin group; HR=0.53; 95% CI, 0.29\u22120.96; P=0.04).\nThe BRCA1/2 mutation status plays an important role for tumor identification in TNBC patients with higher response rate of platinum-based neoadjuvant therapy.\nHowever, other studies have shown that the clinical use of the homologous recombination deficiency (HRD) test may also have the potential to identify patients with TNBC that may respond to the treatment of DNA damage, in excess of those currently identified by gBRCA1/2 mutational screening.\nIt has been suggested that tumors carrying gBRCA mutations may be sensitive to DNA-damaging chemotherapeutic drugs, including platinum.\nIn the present study, we found that for early TNBC patients, the addition of carboplatin to paclitaxel was superior to epirubicin plus paclitaxel with respect to the 3-year DFS among BRCA1/2 unselected patients.\nTo analyze the trends in adjuvant regimens for TNBC and to explore the factors influencing efficacy, we demonstrated that patients aged >40 years, with Ki-67 index >30%, and clinically evaluated lymph nodes were found to have a survival advantage from the PCdd regimen.\nFuture refinement of platinum-sensitive subgroups for targeting specific tumor biomarkers in TNBC is warranted ().\nWith respect to tolerance, previous trials showed a high incidence of AEs and an increasing discontinuation rate for dose-dense chemotherapy of TNBC.\nThe PCdd regimen, which yields fewer adverse toxicities, may be considered a better alternative for the high-risk group of patients in our study, particularly for older patients.\nThe toxicity profile in our study was as anticipated: gastrointestinal toxic effects were more common in the PCdd arm, while grade 3/4 hematological toxicity was more common in the ECdd-P arm.\nAll gastrointestinal toxic effects were manageable and self-limiting.\nThese findings indicate that the PCdd regimen can be recommended to reduce unnecessary toxicities.\nOur study has some limitations, including its small sample size and the potential investigator bias from a single-center institutional experience.\nFurther, we had limited statistical power to show a significant OS benefit.\nA longer follow-up time is necessary, and the median OS should be further evaluated.\nIn addition, given the financial and technical limitations during the study period, the BRCA mutation status was not analyzed to identify whether the gBRCA subgroup will benefit from the PCdd regimen.\nFurther prospective trials to evaluate other platinum-based regimens in the adjuvant setting for TNBC are warranted, particularly to define a sensitive population.\nAn ongoing phase III trial in National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences (NCT03876886 at http://ClinicalTrials.gov) might provide further insight to evaluate the incorporation of platinum in the adjuvant setting, to detect HRD, and to identify specific TNBC patients who might benefit from carboplatin-based therapy.\nConclusions\nPCdd not only is superior to ECdd-P as adjuvant chemotherapy with respect to improving 3-year DFS and OS rates but also yields lower chemotherapy-related toxicities in early TNBC patients regardless of the BRCA mutation status.\nThus, PCdd might be a beneficial standard adjuvant regimen for early TNBC patients at a high recurrence risk, with clinically meaningful improvement in survival and safety data.\nFlow diagram of study design. ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nKaplan-Meier plot of disease-free survival (DFS). Cross marks indicate censored observations. Data for the intention-to-treat population. Hazard ratio (HR), 0.310, 95% confidence interval (95% CI), 0.137\u22120.704; Log-rank P=0.005; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nKaplan-Meier plot of overall survival (OS). Cross marks indicate censored observations. Data for the intention-to-treat population. Hazard ratio (HR), 0.142, 95% confidence interval (95% CI), 0.060\u22120.825, Log-rank P=0.028; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nSubgroup analyses of disease-free survival (DFS). The analyses of two arm patients were stratified for modified intention-to-treat population in clinically relevant subgroups. ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin; SCI, surgery-chemotherapy interval; HR, hazard ratio; 95% CI, 95% confidence interval.\n\nSynopsis of study protocol\nItem | Description\nER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor 2; ANC, absolute neutrophil count; AST, aspartate aminotransferase; ALT, alanine aminotransferase; AUC, area under the curve; 5-HT, 5-hydroxytryptamine; G-CSF, granulocyte colony-stimulating factor.\nStudy ID | CH-BC-012\nStudy title | Randomized phase III trial comparing dose-dense epirubicin and cyclophosphamide followed by paclitaxel with paclitaxel plus carboplatin as adjuvant therapy for triple-negative\u00a0breast cancer\nProtocol date | 4/20/2011\nTrial stage principal | Phase III\nInvestigator | Binghe Xu, M.D. & PhD. National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College. Email: xubinghe@medmail.com.cn;\nQing Li, B.S.Med. National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College. Email: cheryliqing@126.com\n\nParticipating study left | National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing 100021, China\nObjectives | To compare the efficacy and safety of dose-dense epirubicin and cyclophosphamide (ECdd) followed by paclitaxel (P) with dose-dense paclitaxel plus carboplatin (PCdd) as adjuvant therapy for patients with triple-negative breast cancer (TNBC) at high risk of recurrence\nPrimary objective:\n\u2022 Compare 3-year disease-free survival (DFS) of early TNBC patients at high risk treated with PCdd to those treated with ECdd-P regimens\nSecondary objectives:\n\u2022 Compare 3-year overall survival (OS) in the same population\n\u2022 Compare the toxicity of the PCdd to the ECdd-P in patients with TNBC at high risk of recurrence\n\nStudy population | Patients with early TNBC at high risk of recurrence\nStudy design | This is a single-left, open label, randomized, comparative phase III trial. The trial includes two groups: ECdd-P and PCdd.\nEligible participants will be randomly assigned in a 1:1 ratio to the PCdd group or the ECdd-P group. Randomization was conducted with no stratification factors. Eligible patients will be continually enrolled into the study until the total number of patients reached the planned sample size. The patients, medical staff and investigators were aware of treatment allocation. Sample size was determined based on a superiority test of 3-year DFS rate. To detect a difference of an approximate higher proportion of 0.10 between the two regimens (result of our preliminary clinical research demonstrated the proportion surviving in the ECdd-P regimen was 80.0%), an overall sample size of 133 subjects (66 in the ECdd-P arm and 67 in the PCdd arm) was calculated to achieve 80.0% power at a one-sided 0.050 significance level, with a 10% dropout rate (5% in each control/treatment arm). The accrual pattern across time periods was uniform (all periods equal). Primary and secondary efficacy analyses include the intent-to-treat (ITT) population of all randomly assigned patients. The safety analysis population includes all patients who received at least one dose of treatment.\n\nEligibility | Inclusion criteria: 1) Patient must accept the primary breast surgery; 2) Patients with histologically confirmed ER (\u2212), PR (\u2212) and HER2 (\u2212),i.e., <1% positive tumor cells with nuclear staining in IHC and no HER2 overexpression; 3) Positive axillary lymph nodes; negative axillary lymph node with age <35 years or III grade or intravascular cancer embolus; 4) Age between 18 years to 65 years; 5) Able to give informed consent; 6) Patients with an Eastern Cooperative Oncology Group (ECOG) performance score of 0 or 1; 7) Not pregnant, and on appropriate birth control if of child-bearing potential; 8) Adequate bone marrow reserve with ANC >1.5\u00d710 9/L and platelets >100\u00d710 9/L; 9) Adequate renal function with serum creatinine <2.0\u00d7 the upper limit of normal; 10) Adequate hepatic reserve with serum bilirubin <2.0\u00d7 the upper limit of normal, AST/ALT <2\u00d7 the upper limit of normal, and alkaline phosphatase < 5\u00d7 the upper limit of normal. Serum bilirubin >2.0 is acceptable in the setting of known Gilbert\u2019s syndrome; and 11) No active major medical or psychosocial problems that could be complicated by study participation.\nExclusion criteria: 1) Received neo-adjuvant therapy; 2) cardiac dysfunction documented by an ejection fraction less than the lower limit of the facility normal by multi-gated acquisition (MUGA) scan, or 45% by echocardiogram; 3) uncontrolled medical problems; 4) evidence of active acute or chronic infection; 5) pregnant or breast feeding; or 6) hepatic, renal or bone marrow dysfunction as detailed above.\n\nSample size calculation | The target sample size was calculated based on the primary endpoint, i.e., 3-year DFS rate. To detect a difference of 0.13 between the two regimens (result of our preliminary clinical research demonstrated the proportion surviving in the ECdd-P regimen was 80.0%), an overall sample size of 133 subjects (66 in the ECdd-P arm and 67 in the PCdd arm) was calculated to achieve 80.0% power at a one-sided 0.050 significance level. The accrual pattern across time periods was uniform (all periods equal). The proportion of drop out in the control and treatment group was 0.1000 (each 0.05).\nRandomization | Upon meeting the eligibility criteria, patients will be randomised under concealment, by the study lead investigator (Cancer Hospital, Chinese Academy of Medical Sciences), according to prespecified randomisation number lists to receive ECdd-P or PCdd.\nTreatment | Administration: Patients in both study groups received treatment in 14-day cycles. Patients assigned to the PCdd arm received paclitaxel 150 mg/m2 on d 1 plus carboplatin AUC=3 on d 2 for 8 cycles. Patients assigned to the ECdd-P arm received epirubicin 80 mg/m2 divided in 2 d and cyclophosphamide 600 mg/m2 on d 1 for 4 cycles followed by paclitaxel 175 mg/m2 on d 1 for 4 cycles. Prophylactic antiemetic measures, including 5-HT3 receptor antagonists, and dexamethasone, were allowed. Premedication with dexamethasone and histamine antagonists was administered before paclitaxel to prevent hypersensitivity reactions. Prophylactic G-CSF 3 \u00b5g/kg in d 5\u22129 was given for each chemotherapy cycle.\n\nSafety assessments and dose modifications | Safety assessments included 12-lead electrocardiograms, vital sign taking and clinical laboratory evaluations every cycle. Adverse events (AEs) were recorded at each treatment cycle until 28 follow-up d after the end of study visit. Toxicity was graded by using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 (NCI-CTCAE, version 3.0). Febrile neutropenia was managed according to institutional treatment guidelines in China. Toxicities were managed through dose delays of up to 3 weeks, and dose reductions were permitted in the following events: grade 4 hematological, grade 3 or 4 non-hematological, or other protocol-specified toxic effects.\nStudy drugs | Drug: epirubicin, cyclophosphamide, paclitaxel, carboplatin, G-CSF epirubicin 80 mg/m2 iv divide in 2 d cyclophosphamide 600 mg/m2 iv d 1 G-CSF 3 \u00b5g/kg in d 5\u22129 q14d \u00d74 cycles paclitaxel 175 mg/m2 iv d 1 G-CSF 3 \u00b5g/kg in d 5\u22129 q14d \u00d74 cycles paclitaxel 150 mg/m2 iv d 1 carboplatin AUC=3 iv d 2 G-CSF 3 \u00b5g/kg in d 5\u22129 q14d \u00d78 cycles.\n\nConcomitant medications | 1. Antiemetics can be prescribed to patients who are vomiting due to administration of treatment drug(s);\n2. Patients experiencing peripheral neuropathy can be treated with neurotropic supplements such as duloxetine, vitamin B, etc.;\n3. Analgesics can be used for patients who have pain affecting quality of life;\n4. Patients with constipation, diarrhea, or other conditions can be treated using appropriate medication for their respective condition;\n5. Prophylactic antiemetic measures, including 5-HT3 receptor antagonists, and dexamethasone, were allowed.\n6. Premedication with dexamethasone and histamine antagonists was administered before paclitaxel to prevent hypersensitivity reactions.\n\nOutcome measures | Primary outcome measure:\nThe primary endpoint is 3-year DFS rate. DFS was calculated from the date of randomization to the date of the first local/distant recurrence (without second primary malignancies), according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.\nSecondary outcome measures:\nSecondary endpoints include 3-year OS (defined as the time from randomization to death due to any cause) and safety of the treatment. Toxicity was graded by using the NCI- CTCAE, version 3.0.\n\nSafety parameters | AEs, vital signs and clinical laboratory tests\nStatistical analysis | All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) software (Version 22.0; IBM Corp., New York, USA). Data on clinical characteristics, chemotherapy, recurrence, and survival were analyzed. Data were presented as the number (%) or the mean standard deviation. Continuous variables were compared using the Student\u2019s t test, while categorical variables were compared using the \u03c72 or Fisher\u2019s exact test.\nThe proportion of patients remaining event-free over time will be displayed using the Kaplan-Meier method and analyzed using a two-sided log-rank test. All statistical tests were two-sided, and a P value of <0.05 was considered statistically significant.\nThe safety population will include all patients who received at least one dose of treatment. For safety analysis, AEs will be coded using the Medical Dictionary for Regulatory Activities (MedDRA). Analysis of AEs will be based on treatment-emergent adverse events (TEAEs). TEAEs are AEs not present prior to medical treatment, or are already present and worsen either in intensity or frequency following treatment. The incidence rate of TEAEs will be described according to system organ class (SOC) and preferred term (PT). Meanwhile, serious AEs (SAEs) and AEs leading to study discontinuation will be similarly summarized and tabulated. Laboratory tests will be analyzed using descriptive statistical analysis.\n\nFollow-up | All treated patients will be followed-up with once every 3 months to collect survival information for DFS and OS. Patients who discontinue treatment due to any causes will be followed-up with once every 3 months until disease recurrence or death. After disease recurrence, patient follow up can be conducted by phone or as general clinical visits until death.\n\n\nBaseline characteristics of patients with triple-negative breast cancer\nVariable | ECdd-P arm (N=73) [n (%)] | PCdd arm (N=70) [n (%)] | P\nECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; MRM, modified radical mastectomy; BCS, breast conservative surgery; SLN, simple mastectomy and sentinel lymph node biopsy; SCI, surgery chemotherapy interval.\nAge [mean (range)] (year) | 46 (26\u221264) | 49 (22\u221263) | 0.216\n\u3000\u226440 | 20 (27.4) | 12 (17.1) | 0.163\n\u3000>40 | 53 (72.6) | 58 (82.9) | \nMenopause at diagnosis |  |  | \n\u3000Post-menopause | 50 (68.5) | 39 (55.7) | 0.124\n\u3000Pre-menopause | 23 (31.5) | 31 (44.3) | \nPathology |  |  | 0.114\n\u3000IDC | 63 (86.3) | 67 (95.7) | \n\u3000ILC | 2 (2.7) | 0 (0) | \n\u3000Other type | 8 (11.0) | 3 (4.3) | \nTumor size (cm) |  |  | 0.179\n\u3000<2 | 27 (37.0) | 34 (48.6) | \n\u3000\u22652 | 46 (63.0) | 36 (51.4) | \nLymph node metastasis |  |  | 0.604\n\u3000Yes | 29 (39.7) | 24 (34.3) | \n\u3000No | 44 (60.3) | 46 (65.7) | \nIntravascular cancer embolus |  |  | 0.167\n\u3000Yes | 16 (21.9) | 10 (14.3) | \n\u3000No | 57 (78.1) | 60 (85.7) | \nNuclear grade |  |  | 0.999\n\u3000Grade 1, 2 | 23 (31.5) | 22 (31.4) | \n\u3000Grade 3 | 50 (68.5) | 48 (68.6) | \nKi-67 |  |  | 0.108\n\u3000\u226430 | 12 (16.4) | 20 (28.6) | \n\u3000>30 | 61 (83.6) | 50 (71.4) | \nTNM stage |  |  | 0.104\n\u3000I | 24 (32.9) | 27 (38.6) | \n\u3000II/III | 49 (67.1) | 43 (61.4) | \nType of surgery |  |  | 0.309\n\u3000MRM | 57 (78.1) | 54 (77.1) | \n\u3000BCS | 13 (17.8) | 9 (12.9) | \n\u3000SLN | 3 (4.1) | 7 (10.0) | \nRadiotherapy |  |  | 0.141\n\u3000Yes | 42 (57.5) | 33 (47.1) | \n\u3000No | 31 (42.5) | 37 (52.9) | \nSCI (d) |  |  | 0.609\n\u3000<30 | 47 (64.4) | 42 (60.0) | \n\u3000\u226530 | 26 (35.6) | 28 (40.0) | \n\n\nTreatment exposure in TNBC patients treated with ECdd-P/PCdd chemotherapy\nVariables | n (%) | P\nECdd-P Arm (N=73) | PCdd Arm (N=70)\nTNBC, triple-negative breast cancer; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nFollow-up time [Median (range)] (month) | 56.1 (2.8\u221298.6) | 58.1 (1.2\u221276.6) | 0.320\nNumber of chemotherapy cycles |  |  | \n\u3000Total | 573 | 552 | \n\u3000Median | 8 (3\u22128) | 8 (2\u22128) | 0.783\nDelay of chemotherapy |  |  | 0.361\n\u3000Yes | 23 (31.5) | 25 (35.7) | \n\u3000No | 50 (68.5) | 45 (64.3) | \nDose reduction |  |  | 0.369\n\u3000Yes | 14 (19.2) | 16 (22.9) | \n\u3000No | 59 (80.8) | 54 (77.1) | \n\n\nCommon adverse events in TNBC patients treated with ECdd-P/PCdd chemotherapy\nAdverse events | n (%) | P*\nECdd-P arm (n=73) | PCdd arm (n=70)\nA patient could have experienced more than one specific toxicity. TNBC, triple-negative breast cancer; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin; AST, aspartate aminotransferase; ALT, alanine aminotransferase; TBIL, total bilirubin; CRE, creatinine; *, P values for differences in two arms are tested by \u03c72 test or Fisher exact test.\n\nHematologic toxicities | Grade 1/2 | Grade 3/4 | Grade 1/2 | Grade 3/4 | Grade 3/4\n\u3000Anemia | 28 (38.4) | 0 (0) | 14 (20.0) | 0 (0) | \u2212\n\u3000Leukopenia | 39 (53.4) | 26 (35.6) | 39 (55.7) | 12 (17.1) | 0.010\n\u3000Neutropenia | 30 (41.1) | 35 (47.9) | 31 (44.3) | 15 (21.4) | 0.001\n\u3000Thrombocytopenia | 8 (11.0) | 0 (0) | 9 (12.9) | 2 (2.9) | 0.238\nNon-hematologic toxicities | Grade 1/2 | Grade 3/4 | Grade 1/2 | Grade 3/4 | Grade 3/4\n\u3000Alopecia | 36 (49.3) | 8 (11.0) | 32 (45.7) | 4 (5.7) | 0.204\n\u3000Stomatitis | 38 (52.1) | 0 (0) | 29 (41.4) | 0 (0) | \u2212\n\u3000Nausea emesis | 65 (89.0) | 0 (0) | 56 (80.0) | 1 (1.4) | 0.490\n\u3000Diarrhea | 5 (6.8) | 1 (1.4) | 1 (1.4) | 0 (0) | 0.490\n\u3000Mucositis/cutaneous | 3 (4.1) | 1 (1.4) | 1 (1.4) | 0 (0) | 0.490\n\u3000Peripheral neuropathy | 28 (38.4) | 1 (1.4) | 31 (44.3) | 4 (5.7) | 0.170\n\u3000Foot and hand syndrome | 6 (8.2) | 0 (0) | 1 (1.4) | 0 (0) | \u2212\n\u3000Myalgia/arthralgia | 12 (16.4) | 1 (1.4) | 11 (15.7) | 0 (0) | 0.490\n\u3000Asthenia | 8 (11.0) | 1 (1.4) | 6 (8.6) | 0 (0) | 0.490\n\u3000Allergic | 1 (1.4) | 0 (0) | 3 (4.3) | 0 (0) | \u2212\n\u3000Cardiac toxicity | 3 (4.1) | 0 (0) | 2 (2.9) | 0 (0) | \u2212\n\u3000ALT elevation | 25 (34.2) | 3 (4.1) | 19 (27.1) | 1 (1.4) | 0.326\n\u3000AST elevation | 30 (41.1) | 0 (0) | 26 (37.1) | 0 (0) | \u2212\n\u3000TBIL elevation | 29 (39.7) | 0 (0) | 26 (37.1) | 0 (0) | \u2212\n\u3000CRE elevation | 3 (4.1) | 0 (0) | 7 (10.0) | 0 (0) | \u2212\n",
        "label": "unclear",
        "id": "task4_RLD_test_7"
    },
    {
        "paper_doi": "10.1186/s40814-019-0432-7",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Cluster-RCT\n\n\nParticipants: Participants were 103 pregnant women of over 34 weeks gestation who met the eligibility criteria in a community setting involving 10 villages (sources of data: 103 mother-neonate pairs). Inclusion criteria: \"The study village was selected if it had recorded at least ten births within the past 3 months and had one or more active VHWs and the village leaders committed to study participation and implementation. Meanwhile, the health centres were included if they had participated previously in community trials or research. Pregnant women of over 34 weeks gestation were recruited over a 3-month period and followed up for 3 months postnatally\".Exclusion criteria: \"The exclusion criteria included temporary visitors\" (defined as any pregnant mother found as a visitor in the home who did not stay afterward). Other clinical criteria (e.g. malaria in pregnancy, previous caesarean section) were not used as exclusion criteria in this pilot study due to the nature of the intervention.\n\n\nInterventions: Intervention: \"Pregnant women in intervention villages were provided with ABHR (Alsoft V, Saraya East Africa Co. Ltd.) at recruitment. The recruiting research midwives provided the ABHR free of charge to each woman in a 1-l bottle for use while at home, along with a refillable 100-mL bottle for use while travelling. The recruiting midwives trained each woman in the intervention villages on the use of ABHR, the basic hand rub steps, and the 'three moments for community neonatal hand hygiene', developed by the study team for the pilot trial. This was adopted from the WHO '5 Moments for Hand Hygiene'. The three moments for community neonatal hand hygiene instructions were printed on a poster with a pictorial illustration, which was given to the participants as instructions to display in a visible area and follow in their homes. The poster was available in both English and the local language (Lumasaba). Participants were also given Maama Kits which consisted of basic supplies, i.e. sterile gloves, plastic sheets, cord ligature, razor blades, tetracycline, cotton, gauze, soap, and sanitary pads (usual care)\".\nControl: \nControl villages: current standard care practice. Women in the control villages received the current\nstandard care of Maama Kits for delivery and the usual antenatal education. At the time of recruitment, the recruiting midwives advised the women to deliver at health facilities. The midwives encouraged women to attend postnatal checks and immunisation clinics at 6-24 h, 1-2 weeks, 6 weeks, 10 weeks, and 14 weeks at the nearby health facilities in line with the local guidelines. The Maama Kit consisted of basic supplies, i.e. sterile gloves, plastic sheets, cord ligature, razor blades, tetracycline, cotton, gauze, soap, and sanitary pads.\n\n\nOutcomes: 1) infant sepsis (physician-defined and microbiological)2) infant mortality3) rate of positive infections using the WHO's IMCI screening criteria for infection4) early neonatal death5) late neonatal death6) mother-reported infant infection7) hospitalisation\n\n\nNotes: Sponsor: Medical Research Council/WellcomeTrust/DfID (Global Health Trials Scheme)Country: UgandaSettting: \"The study was conducted in ten clusters (villages) around two community health centres (health centre three (HCIII) and health centre four (HCIV)). The HCIII was surrounded by three control villages, while the HCIV was surrounded by five intervention and three control villages. The clusters were rural villages located in Mbale region, Eastern Uganda. A map of the villages surrounding the targeted community health centres in Mbale district was reviewed by the research team in collaboration with the village health team workers (VHWs). Ten villages were selected from a map of 28 villages if they were within the catchment area of the participating community health centres, were not served by another community health centre, and had a village health worker.\"\n The study also adjusted for clustering effects in the analysis.Contact: Ditai J, INSTITUTION Sanyu Africa Research Institute (SAfRI), Mbale, UgandaEMAIL: J.Ditai@safri.ac.ugADDRESS:1 Sanyu Africa Research Institute (SAfRI), Mbale Regional Referral Hospital, Pallisa-Kumi Road Junction, P.O Box 2190, Mbale, Uganda2 Sanyu Research Unit, Department of Women's and Children's Health, Liverpool Women's Hospital, University of Liverpool This study was funded by Medical Research Council/WellcomeTrust/DfID (Global Health Trials Scheme).\n\n",
        "objective": "To determine the effectiveness of different hand hygiene agents for preventing neonatal infection in both community and health facility settings.",
        "full_paper": "Background\nAlcohol-based hand rub (ABHR) is widely used in both health and social facilities to prevent infection, but it is not known whether supplying it for regular perinatal use can prevent newborn sepsis in African rural homes.\nOur study piloted a cluster randomised trial of providing ABHR to postpartum mothers to prevent neonatal infection-related morbidity in the communities.\nMethods\nWe conducted a pilot parallel cluster randomised controlled trial across ten villages (clusters) in rural Eastern Uganda.\nPregnant women of over 34\u2009weeks\u2019 gestation were recruited over a period of 3\u2009months.\nBoth clusters received the standard of care of antenatal health education, Maama Kit, and clinic appointments.\nIn addition, women in the intervention villages received ABHR, instructions on ABHR use, a poster on the \u2018three moments of hand hygiene\u2019, and training.\nWe followed up each mother-baby pair for 3\u2009months after birth and measured rates of consent, recruitment, and follow-up (our target rate was more than 80%).\nOther measures included ABHR use (the acceptable use was more than four times a day) and its mode of distribution (village health workers (VHWs) or pharmacy), acceptability of study protocol and electronic data capture, and the use of WHO Integrated Management of Childhood Illness (IMCI) tool to screen for newborn infection.\nResults\nWe selected 36% (10/28) of villages for randomisation to either intervention or control.\nOver 12\u2009weeks, 176 pregnant women were screened and 58.5% (103/176) were eligible.\nAll, 100% (103/103), eligible women gave consent and were enrolled into the trial (55 intervention and 48 control).\nAfter birth, 94.5% (52/55) of mothers in the intervention and 100% (48/48) of mothers in the control villages were followed up within 72\u2009h. Most, 90.9% (50/55), of the mothers in the intervention villages (96.2% of live births) and 95.8% (46/48) of mothers in the control villages (95.9% of live births) were followed up at 3\u2009months.\nIn intervention villages, the average hand rub use was 6.6 times per day.\nVHWs accounted for all ABHR stock, compared to the pharmacy that could not account for 5\u2009l of ABHR.\nThe screening tool was positive for infection among a third of babies, i.e. 29.2% (14/48) in the intervention villages versus 31.4% (16/51) in the control villages.\nVHWs completed the first four questions of IMCI screening tool with ease and accuracy.\nThere were no adverse reactions with the ABHR.\nConclusion\nIt is feasible to conduct a cluster-randomised controlled trial (cRCT) of the provision of ABHR to postpartum mothers to prevent neonatal infection-related morbidity in the community in resource-poor settings.\nOur results indicate that home recruitment promotes excellent follow-up and retention of participants in community trials.\nThe intervention was safe.\nThis pilot study informed the substantial changes necessary in the larger cRCT, including a change in the primary outcome to a composite outcome considering multiple methods of infection detection.\nA\u00a0large BabyGel cluster randomised controlled trial is now required.\nTrial registration\nISRCTN67852437, registered March 02, 2015\nTrial funding\nMedical Research Council/WellcomeTrust/DfID (Global Health Trials Scheme)\nBackground\nGlobally, 45% of deaths in children under 5\u2009years occur in the neonatal period with nearly 90% occurring in sub-Saharan Africa and South Asia.\nIn Uganda, the neonatal mortality rate is 27 deaths per 1000 live births and has not changed for the past decade while the post-neonatal mortality rate is 16 deaths per 1000 live births.\nIn Eastern Uganda, neonatal mortality is higher than the country average at 34 per 1000 live births,, with 77% presenting with infection-related symptoms.\nThough pneumonia, diarrhoeal diseases, and sepsis are the leading infectious causes of deaths in children under 5\u2009years annually, the neonatal cause-of-death distribution differs between the early (0\u20136\u2009days of life) and late (7\u201328\u2009days of life) periods and varies with neonatal mortality rate level.\nPreterm birth (40.8%) and intrapartum complications (27.0%) account for most early neonatal deaths in all regions of the world while infections cause nearly half of late neonatal deaths, the target for this study.\nIn Uganda, 33% of newborns present with a fever, 9% with symptoms of acute respiratory infections, and 20% experience diarrhoea.\nMost newborn infections and deaths occur in the community and are frequently unreported to the health sector.\nEvidence shows that these infant infections are diseases of poverty, associated with poor home environments, remoteness, hunger, undernutrition, and lack of access to essential services.\nSome infections are transmitted directly from the mother\u2019s genital tract at the time of birth, including streptococci, staphylococci, and Escherichia coli, while most infections are transmitted from toilets, animals, gardens, or other unclean areas through carers\u2019 hands, resulting in neonatal tetanus, skin infections, pneumonia, diarrhoea, or septicaemia.\nThe baby\u2019s umbilical stump is a particular risk especially with traditional practices that increase chances of cord infection, such as the umbilical application of baby powder, soil, or manure.\nHandwashing with soap is a simple and sustainable measure that results in a large reduction in hand contamination, even when used with unclean water.\nBirth attendant and maternal handwashing have been associated with reductions in neonatal mortality.\nAn estimated 40% reduction in neonatal sepsis deaths relate to newborn care practices at home.\nA systematic review concluded that evidence for the effect of clean birth and postnatal newborn care practices on neonatal mortality was of low quality.\nThe importance of handwashing in preventing infection-related deaths has led WHO to develop guidelines for hand hygiene both within health care settings and in the community.\nHowever, studies show widespread non-adherence to the household guidelines, often due to lack of water and or washing facilities.\nGlobally, less than 20% wash their hands after defecation compared to the only 10% in Uganda.\nUganda thus ranks among the ten countries with the poorest handwashing behaviour, and access to water is not guaranteed for many regions in the country including Mbale.\nAn alternative could, therefore, be alcohol-based hand rub (ABHR) which is produced locally in Uganda from sugar cane.\nIt costs only US$0.025 (\u00a30.02) per cleansing and is active against a broad range of Gram-positive and Gram-negative aerobic bacteria, fungi, and enveloped and non-enveloped viruses known to cause diarrhoea and lower respiratory tract infections in early childhood.\nABHR requires no infrastructure and can be easily distributed to any recruited Ugandan household.\nIt is also quicker than handwashing with soap and has been suggested to increase compliance.\nIt is therefore rapidly scalable, either by including it with birthing kits or through its provision at the time of birth.\nAlthough ABHR was added to the 19th WHO Model List of Essential Medicines, in support for hand hygiene, there is insufficient evidence for the prevention of infections in early infancy.\nNo policies as yet recommend the use of ABHR for routine community postnatal prevention of newborn infections.\nIn the BabyGel study, we hypothesise that the addition of ABHR to birthing kits would enable mothers to provide effective hand hygiene for the first three postnatal months and lower the risk of young infant infections.\nTo assess whether village leaders and pregnant women are willing to participate in the study (criteria: a participation rate of at least 80%)\nTo test the integrity, feasibility, and acceptability of the study protocol including questionnaires, information sheets, data management systems, and methods for the identification of community neonatal infections (iterative assessment and refinement of processes)\nTo evaluate the use of the WHO IMCI criteria for a community screening of possible infection as a primary outcome (criteria: assessment of the completed IMCI screening tool).\nTo determine the locally appropriate mechanism for the distribution of ABHR through a comparison of the relative benefits of pharmacy-based and VHW-based distribution mechanisms (an iterative assessment)\nTo assess whether participants use ABHR once recruited into the BabyGel trial (acceptable criteria: mean use of at least four times/day)\nTo assess the level of contamination between intervention and control clusters (acceptable criteria: mean alcohol hand rub use of more than once a day in control clusters).\nTo estimate the intracluster correlation coefficient to inform the sample size calculation for a large cluster randomised controlled trial\nThis study was undertaken to pilot a cluster randomised trial design for the provision of ABHR antenatally for use by postpartum mothers and other baby carers in the prevention of infection-related neonatal morbidity and mortality in these rural communities.\nThe feasibility aims were as follows:\nOther published findings from the BabyGel pilot trial include the acceptability of ABHR for community use, the optimising informed consent for trial participation in Uganda, and the newborn moments of community hand hygiene.\nMethods\nStudy design\nIn this feasibility study, we piloted an open two-arm parallel cluster randomised controlled trial of the provision of ABHR to postpartum mothers to prevent neonatal infection-related morbidity in the home.\nWe used the definitions and methodology for feasibility and pilot studies as recommended in the \u2018CONSORT 2010 statement: extension to randomised pilot and feasibility trials\u2019.\nThe pilot trial was locally approved by the Mbale Regional Hospital Institutional Review Committee (REIRC IN\u2013COM 011/2015), registered with the Uganda National Council for Science and Technology, and prospectively registered (ISRCTN67852437).\nCriteria for site selection\nThe study village was selected if it had recorded at least ten births within the past 3\u2009months and had one or more active VHWs and the village leaders committed to study participation and implementation.\nMeanwhile, the health centres were included if they had participated previously in community trials or research.\nStudy setting\nThe study was conducted in ten clusters (villages) around two community health centres (health centre three (HCIII) and health centre four (HCIV)).\nThe HCIII was surrounded by three control villages, while the HCIV was surrounded by five intervention and three control villages.\nThe clusters were rural villages located in Mbale region, Eastern Uganda.\nA map of the villages surrounding the targeted community health centres in Mbale district was reviewed by the research team in collaboration with the village health team workers (VHWs) (Fig.\u00a01).\nTen villages were selected from a map of 28 villages (Fig.\u00a02) if they were within the catchment area of the participating community health centres, were not served by another community health centre, and had a village health worker.\nPilot randomisation\nThe ten villages were assigned to either intervention or control by simple random sampling in a central office in Mbale in a 1:1 ratio by a person who did not draw the map.\nThis represented a variety of distances of villages from each other, from market areas, from the health centres, and from the control villages, thus allowing contamination to be effectively assessed.\nIt was not possible to blind women, health workers, local council leaders, or researchers, due to the nature of the intervention.\nCluster consent\nAfter allocation, researchers met the local council leader(s) of each village to seek consent for participation of their village in the pilot trial and the pilot trial allocation.\nStudy timeline\nVillages were selected in November 2014, 9\u2009months before\u00a0recruitment of the first participant\u00a0for appropriate trial preparation purposes.\nThe recruitment of participants into the trial was time bound.\nThe research midwives recruited eligible women for a period of 12\u2009weeks between August 2015 and November 2015, while each individual woman together with her baby was followed up in her home by research midwives until 12\u2009weeks after childbirth.\nThe overall pilot trial follow-up of participants continued until May 2016 when the last mother-baby pair had their final study assessments at 12\u2009weeks (90\u2009days) after birth.\nParticipants\u2019 selection criteria\nPregnant women were eligible for the study if they had a pregnancy with an estimated gestation of over 34\u2009weeks during the recruitment period and were resident in the participating villages.\nThe exclusion criteria included temporary visitors (defined as any pregnant mother found as a visitor in the home and will not stay afterward).\nOther clinical criteria (e.g. malaria in pregnancy, previous caesarean section) were not used as exclusion criteria in this pilot study due to the nature of the intervention.\nParticipant recruitment\nEach village health worker (VHW) identified pregnant women from his/her village by direct contact or review of the village-specific monthly pregnancy and birth register.\nAlso, midwives identified pregnant women turning up for their antenatal clinics at the two participating health facilities from the list of study villages.\nThe VHWs or facility midwives then notified a member of the research team about the potential participant.\nThe research team then visited the woman\u2019s home to confirm eligibility, obtain informed consent, and conduct detailed study-specific assessments.\nThe research midwife screened each woman for eligibility and established the gestational age using a gestational wheel and the self-reported last normal menstrual period (LNMP).\nThe research team recruited eligible women into the pilot trial for a period of 12\u2009weeks.\nThey provided informed consent for trial participation and follow-up.\nThose who declined were given the Maama Kit, and their data was not included.\nThe Maama Kit was devised by the Ministry of Health in Uganda to ensure childbirth is conducted in a clean environment.\nThe Maama Kit consists of basic supplies, i.e. sterile gloves, plastic sheets, cord ligature, razor blades, tetracycline, cotton, gauze, soap, and sanitary pads.\nInterventions: BabyGel intervention villages\nPregnant women in intervention villages were provided with ABHR (Alsoft V, Saraya East Africa Co. Ltd.) at recruitment.\nThe recruiting research midwives provided the ABHR free of charge to each woman in a 1-l bottle for use while at home, along with a refillable 100-ml bottle for use while travelling.\nThe recruiting midwives trained each woman in the intervention villages on the use of ABHR, the basic hand rub steps, and the \u2018three moments for community neonatal hand hygiene\u2019 (Fig.\u00a03), developed by the study team for the pilot trial.\nThis was adopted from the WHO \u20185 Moments for Hand Hygiene\u2019.\nThe three moments for community neonatal hand hygiene instructions were printed on a poster with a pictorial illustration, which was given to the participants as instructions to display in a visible area and follow in their homes.\nThe poster was available in both English and the local language (Lumasaba).\nThe instructions on the poster recommended hand rub before touching the baby, before clean or aseptic procedures by birth attendants, and daily wiping of cord end with the ABHR three times a day until it falls off.\nIt also included, after any body fluid exposure risk such as after the mother or carer using the toilet, touching any surfaces and exposure to child faeces.\nThe midwives encouraged women to apply the ABHR based on this policy.\nIn addition to the training and the poster, the recruiting research midwives provided further instructions to the women.\nThis included women commencing the ABHR at the time of recruitment until 3\u2009months after childbirth, each woman performing a whole body wipe of the baby with ABHR within 4\u2009h of birth, and every woman encouraging any carer or family member (including children) to use the ABHR with the \u2018three moments\u2019 instructions used within and away from their homes.\nThe research midwives instructed women from three of the five intervention villages to obtain refills from their VHWs (VHW-based distribution) if the need arose, while women from the other two intervention villages refilled from the pharmacy at the health centre IV (pharmacy-based distribution).\nThis allowed us to compare the two modes of ABHR distribution.\nAlongside the above intervention, midwives offered the current standard care practice (described below) as was provided for the control villages.\nAfter follow-up for the first 14 participants after birth within the intervention clusters, we recognised that some mothers reported forgetting to use the ABHR as indicated, while some who had homebirths had not taken their newborns to the health facilities for immunisation.\nHowever, in almost every household studied, there were already school-going children present.\nWe hence introduced a system of the \u2018expert child\u2019 (a child in each house tasked with the responsibility of reminding carers to use the ABHR) to improve ABHR adherence and remind the mother to notify the research team of any sick baby to strengthen the multifaceted approach of the intervention.\nIn this pilot trial, we limited our communication to the one-to-one teaching of the participants with the use of the poster.\nThis was a small-scale study in which the behaviour change was limited to the participating women and their immediate families.\nMass communication or media strategies were, therefore, not necessary.\nControl villages: current standard care practice\nWomen in the control villages received the current standard care of Maama Kits for delivery and the usual antenatal education.\nAt the time of recruitment, the recruiting midwives advised the women to deliver at health facilities.\nThe midwives encouraged women to attend postnatal checks and immunisation clinics at 6\u201324\u2009h, 1\u20132\u2009weeks, 6\u2009weeks, 10\u2009weeks, and 14\u2009weeks at the nearby health facilities in line with the local guidelines.\nThese visits also included information on the importance of hand hygiene in line with WHO guidelines and promote the usual practice of \u2018dry care\u2019 of the umbilical cord.\nThe research midwives discouraged women in the control villages from applying any local substances to the cord including home-made saline but use \u2018dry care\u2019, keeping the cord clean and dry.\nTrial follow-up procedures\nAt the end of baseline data collection, the follow-up study visit dates for day 1 and day 90 were auto-generated from the ODK system on the mobile smartphone, based on the expected date of delivery.\nThese scheduled dates were written on the follow-up cards which were given to mothers.\nOn the actual follow-up dates, the research assistant captured the exact dates for day 1 and day 90 assessment on the follow-up card.\nThis enabled some women to remind or call the research team to confirm if they were approaching day 90.\nThe day 90 follow-up schedule from the electronic system guided the research team when to visit the woman.\nThe participant, health centre midwives, or VHWs notified the research team of the birth immediately after birth.\nThe research team also monitored the scheduled expected date of delivery and contacted the participant regularly to see whether they had given birth.\nThe team then visited the woman to collect data as soon as possible after birth.\nVHWs were encouraged to visit mothers in their homes twice weekly in the first 4\u2009weeks after birth, then once a week thereafter.\nAt each visit, VHWs were asked to complete the WHO IMCI screening tool (Fig.\u00a04).\nBabies with a positive WHO IMCI screen were referred to the nearest participating health facility for assessment.\nMeanwhile, mothers with concerns about their babies were also advised to take their babies directly to the health centres where they were screened for infection using the WHO IMCI screening criteria and assessed by a clinician.\nThe research team was notified of any positive WHO IMCI screen by either the VHW or participant or health facility midwife/clinician.\nThe paediatrician based at Mbale hospital was available for consultation for any sick infants presenting at the community health facility and or being admitted.\nThe research team continued to follow up each woman and her baby(ies) until 90\u2009days postnatally.\nThe research team encouraged VHWs to remind the mothers in intervention villages to use the ABHR and return for refills as required.\nThe research assistants also encouraged the use of the ABHR during any follow-up phone calls and on the day 1 follow-up visit.\nOutcome measurements\nPilot trial process outcomes\nElectronic data capture system\nThe research midwives collected data using the Open Data Kit system (ODK; University of Washington, USA) loaded onto handheld Samsung Galaxy S4 mobile phones.\nThe research team carried these phones into the community for entering data directly from the participants.\nAfter completion by a research assistant, three approvers performed quality checks (i.e. the data collected by the research assistant was reviewed and approved by the study coordinator followed by the co-investigator, then data manager).\nOnce approved, the data on the phone automatically synchronised with the server at the Liverpool School of Tropical Medicine either while in the community or upon return to the research office where there was adequate Wi-Fi signal or internet.\nThe research team discussed the structure, content of the questions, format, and font size of the text on the ODK phone after every data collection visit.\nSpecific versions of questions were developed for data collection at baseline, follow-up at day 1 and day 90 after birth, and any interim visit.\nAt baseline visit.\nThe research midwife collected baseline demographics, hand hygiene exposure, and current obstetric history.\nThe midwife also collected ABHR details including quantity dispensed and how to obtain the refill, for participants from the intervention clusters.\nDay 1 follow-up visit data included maternal outcomes (antenatal and intrapartum) and neonatal outcomes (age, birth weight, IMCI screening for infection).\nDay 90 follow-up visit data included neonatal outcomes (medical history, IMCI screening for infection), hand hygiene practices, and ABHR use.\nParticipant-specific outcomes\nPrimary outcome\nThe primary outcome was the number of positive infants on the WHO IMCI screening tool with modifications informed by the YICSSG algorithm.\nThe rates of infection were assessed according to the study-specific neonatal and young infants\u2019 outcomes screening algorithm (Fig.\u00a05).\nThe research midwives assessed newborns in their homes at 1\u20132\u2009days and 3\u2009months postnatally and applied the WHO IMCI screening tool.\nThe research team encouraged the VHWs to complete the WHO screening tool whenever in contact with the mother-baby pair.\nFurther, the research team asked the staff at health facilities to complete the WHO IMCI screening tool if the baby was presented to the facility without notice of the study team.\nAny baby screening positive for infection at these visits was referred to Mbale hospital for appropriate management by the paediatrician.\nThe blood sample was collected from all admitted babies in the hospital for malaria blood smear, complete blood count, culture, and sensitivity.\nSecondary outcomes\nInfection-related infant mortality\nThese were sepsis deaths in the first 90\u2009days of life, assessed using verbal autopsy.\nInfant mortalities at 24\u2009h, 7\u2009days, 4\u2009weeks, and 3\u2009months of life (all cause and infective) were assessed.\nMaternal quality of life at 3\u2009months after childbirth\nThe research midwife assessed every woman using the WHO Quality of Life Scale Brief Version (WHOQOL-BREF 1998).\nThis tool was translated into the local language (Lumasaba) for use and validation in this community.\nABHR use\nThe research midwives recorded the volume of ABHR dispensed to each participant on the e-data capture system at the time of recruitment and at end of study through review of the dispensing and accountability logs.\nThe staff at the health facility pharmacy and VHW in intervention villages maintained dispensing and accountability logs of ABHR distributed to participants.\nThe study team then conducted an inventory of the distributed ABHR to participants from the designated health facility and the VHW\u2019s home.\nAt the last follow-up visit, the research assistant measured the total volume of the ABHR remaining in the bottles to establish the total volume of ABHR used by each participant.\nAdverse events\nThe research team collected data on any adverse events related to the ABHR among the newborns.\nThe research team reported any event to the mother or child that required hospitalisation and was life-threatening or a congenital abnormality as a severe adverse event (SAE) irrespective as to whether it is expected or potentially related to the study intervention.\nRisk of contamination\nThe research midwives collected data on the ABHR use and knowledge of ABHR, during the follow-up of all mothers in the control villages.\nCommunity advisory board\nA community advisory board (CAB) was set up consisting of a religious leader, a woman representative, a teacher, the local council (LC) 1 chairperson, a VHW, a clinical officer, a motorbike ambulance cyclist, a midwife, the health facility in charge, a doctor, and the Mbale District health officer (chair).\nTwo CAB meetings were held; the first one immediately after the start of the study and the last one at the end of the study.\nDuring the meetings, the researchers explored the views of these members about the study and its interventions.\nSample size\nThe main objective of this pilot trial was to test the feasibility of providing ABHR antenatally for use by mothers and other baby carers after birth to prevent infection-related neonatal morbidity and mortality in the communities.\nAs this was a pilot trial designed to examine the methodological and procedural uncertainties of a future full-scale trial, we did not undertake a formal sample size calculation.\nInstead, the data collection period was time limited.\nWith an average of 3 to 4 deliveries per month in the study villages, in a 3-month pilot, therefore, we expected to recruit 90\u2013120 women.\nThis would allow us to assess background infection rates among newborns which we could use for a formal sample size calculation for the full cluster randomised trial (since funded by European and Developing Countries Clinical Trials Partnership (EDCTP)).\nData analysis\nFor this pilot trial, the samples were compared as if individually randomised (i.e. ignoring possible clustering within villages).\nThe baseline characteristics of the study mothers are summarised as mean values with their standard deviations and ranges for continuous measures, or as frequency counts and corresponding percentages for categorical measures.\nSimilar statistics are used to summarise all outcome measures.\nAs the study was primarily concerned with determining the feasibility of providing ABHR to prevent infection-related neonatal morbidity, the differences between the two arms are also presented alongside their 95% confidence intervals (calculated using exact binomial methods for categorical measures).\nNo formal statistical comparisons (p values) were calculated.\nThe intracluster correlation coefficients (ICC) for possible infant infections among infants at 3\u2009months after birth was also calculated.\nResults\nAcceptability of intervention\nCluster consent\nAll, 100% (10/10), local council leaders of the randomised villages gave informed consent for the participation of their village in the study and study allocation.\nParticipants\u2019 consent\nAll, 100% (103/103), eligible women gave individual informed consent to participate in the BabyGel pilot trial.\nHowever, one woman in the intervention village withdrew her consent from the study after birth, due to pressure from the husband about the ABHR.\nCommunity advisory board\nAll, 100% (12/12), CAB members chaired by the District Health office approved and accepted the pilot trial and the intervention in this area.\nScreening and recruitment outcome\nOver a 3-month period from 20 August 2015 to 30 November 2015, 176 pregnant women were screened and 103 met the eligibility from the 10 villages.\nAll, 100% (103/103), eligible women were recruited into the pilot trial.\nOf the 41.5% (73/176) excluded from the trial, 71 did not meet the inclusion criteria of gestation age less than 34\u2009weeks and 2 others had only visited the homes for a temporary period.\nOf the 103 participants recruited, 55 were allocated and received the BabyGel intervention, while 48 were allocated and received the usual care.\nFigure\u00a02 shows the CONSORT flow diagram for the study.\nDemographics\nThe participants\u2019 demographic characteristics are summarised in Table\u00a01.\nBoth types of clusters were evenly matched.\nThe study participants were mostly married (77.1% (37/48) for control compared to 78.2% (43/55) for intervention).\nAbout half (45.8% (22/48) for control compared to 56.4% (31/55) for intervention) did not complete primary education level and were either housewives (47.9% (23/48) for control compared to 45.5% (25/55) for intervention) or peasant farmers (37.5% (18/48) for control compared to 36.4% (20/55) for intervention), although those in the intervention villages were more likely to have homes with tap water (either personal or shared) and more likely to have no latrine (Table\u00a01).\nThey generally lived in \u2018mud and wattle\u2019 houses with mud floors (47.9% (23/48) for control compared to 45.5% (25/55) for intervention) and corrugated iron roofs (87.5% (42/48) for control compared to 89.1% (49/55) for intervention), and obtained water from a shared tap or borehole (72.9% (35/48) for control compared to 90.9% (50/55) for intervention).\nMost (91.7% (44/48) for control compared to 80.0% (44/55) for intervention) used non-ventilated pit latrines without the facility for handwashing.\nDespite this, over 80% of women reported washing their hands more than 50% of times after either urinating or defecating, with about 87% using soap and water (87.5% (42/48) for control compared to 87.3% (48/55) for intervention).\nOnly half (47.9% (23/48) for control compared to 54.5% (30/55) for intervention) of the pregnancies were planned; however, all reported attending antenatal clinics.\nMost women, who reported unplanned pregnancies, wished to have had appropriate child spacing; though, they were not using family planning.\nFollow-up outcome\nAll, 100% (103/103), mothers in both clusters were followed up immediately after birth either in their homes or from health facilities.\nThe majority of women had live births (94.5% (52/55) in intervention compared to 100% (491/48) in control) and the research midwife completed the planned assessments for the immediate study visit (day 1 after childbirth) study visit.\nThree women in intervention clusters had stillbirths and did not have the immediate assessment performed.\nAt three months (90\u2009days) after birth, most mothers with live babies were followed up and completed the final assessment (90.9%, 50/55) mother-baby pairs in intervention compared to (95.8%, 46/48) mothers with (95.9%, 47/49) live babies in control villages.\nThis follow-up rate is commensurate with our predetermined acceptable follow-up criteria of above 80% both within 24\u2009h and three months after birth.\nOf the seven women not followed-up on day 90, three had stillbirths, one had relocated to Kampala, one withdrew consent after childbirth, and two had neonatal deaths (Fig. 2 and Table\u00a02).\nBirth outcomes\nThere were no major differences in clinical outcomes between both types of clusters for mothers and babies (Table\u00a02).\nSpontaneous vaginal births were common in both types of clusters with only two caesarean sections in each arm.\nThere were slightly more babies born at health facilities in the intervention arm (78.4%, 40/51) than the control arm (64.6%, 31/48), and the births were assisted by a local nurse/midwife (84.0% (42/51) intervention vs 79.2% (38/48) control) or doctors in case of caesarean sections (4.0% (2/51) intervention vs 10.4% (5/48) control).\nThere were homebirths assisted by a relative, neighbour, while others were born en route to the facility or in the traditional birth attendants\u2019 homes.\nStudy protocol\nOverall, the study protocol was feasible and acceptable; minor adjustments were made as the pilot progressed and errors were identified.\nThe electronic data capture worked well, except for one participant whose initials changed during the follow-up visit which caused confusion with another participant.\nThe research assistants believed that this error might have happened due to the small size of the handheld mobile smartphone (Samsung Galaxy S4).\nThree of the questions in the case report form had their text gradually revised to improve on context, which gradually improved data collection as the study progressed.\nThe research assistants and those performing quality checks of the data (study coordinator, co-investigator, and data manager) all had no login errors onto the ODK smartphone system.\nFurther, there were no extra logins required for synchronising data; the phone data automatically and easily synchronised whenever the phone connected to the internet.\nABHR distribution\nThere were more women assigned to receive ABHR refills through the VHWs (70.9%; 39/55) than through the pharmacy (29.1%; 16/55), thus explaining the increased refills in that group (61.8% (34/55) compared to the pharmacy 21.8% (12/55)) (Table\u00a03).\nFifty women received a mean of two additional litres per participant (each received 1.1\u2009l at the start).\nThe mean number of refills was hence two with a mean volume of 3.1\u2009l per participant.\nThe women in both modes of distribution who did not return for a refill had stillbirths (three), had relocated (three), or had withdrawn from the study (one).\nVHWs were very systematic and accurately accounted for all the ABHR using the dispensing logs, compared to pharmacy workers who were unable to account for 5\u2009l at the end of the study.\nInformal enquiry into the underlying reasons revealed that there were five midwives dispensing the ABHR at the pharmacy, and some midwives would dispense without completing the dispensing logs.\nSome pharmacy midwives also reported using the ABHR for their routine care, and some mothers reported being sent back from the pharmacy without a refill.\nABHR use\nThe average frequency of ABHR use in the 5 intervention villages was 6.6 times per infant per day, but half stopped using it for at least 3\u2009days continuously during the 3\u2009months; most reported that they had simply forgotten to use it (Table\u00a03).\nRisk of contamination (criteria: ABHR use in control sites less once a day)\nThere was no routine postnatal ABHR use reported in any of the control sites, except for one woman who was provided with a chlorhexidine gel from her local health facility.\nShe had not applied it on the cord and was advised to use ABHR instead.\nA separate local study started towards the end of the BabyGel pilot.\nIn this study, the health facility midwives provided chlorhexidine gel to every woman delivering in their facilities and advised mothers to apply the gel on the umbilical cord until it dropped off.\nPrimary outcome\nThe a priori primary outcome was the rate of positive infection screen using the WHO\u2019s IMCI screening criteria for infection (Table\u00a04).\nThe screening tool was positive for infection among a third of babies, i.e. 29.2% (14/48) in the intervention villages versus 31.4% (16/51) in the control villages.\nVHWs completed the first four questions of IMCI screening tool with ease and accuracy (Fig.\u00a04).\nHowever, the VHWs had challenges counting the baby\u2019s respiratory rate or taking body temperature, both of which were recordable vital signs on the IMCI screening tool.\nThey either left those questions blank (with no record entered) or filled a value that was wrong.\nThough they found challenges in taking temperature and counting respiratory rate and pulse rate, the VHWs made the judgement of high temperature through a touch of the baby with their palm and the fast breathing through experience or observation in the change of breathing from usual normal and completed the yes or no boxes.\nFurther, the VHWs frequently did not attend to the mothers to complete the screening tool according to the proposed schedule in the protocol.\nNeonatal infection rate\nFew babies actually had a clinical diagnosis of infection made in the study health centres or hospital (control 16.7% (8/48) compared to the intervention 21.6% (11/51); Table\u00a04).\nThe highest rates of identification of possible infant infection were through direct questioning of mothers.\nA half of women in the intervention arm (51.0%, 26/51) reported that their babies had suffered infections compared to two fifths in the control arm (39.6%, 19/48).\nSome babies received antibiotics from the local pharmacies and private clinics (control 16.7% (8/48) compared to intervention 27.5% (14/51)) but were neither screened nor reported to the study facilities.\nOverall, more than half (control 56.3% (27/48) compared to the intervention 54.9% (28/51)) of babies had some evidence of non-malaria infection either reported by mothers, evidenced in medical records, confirmed in the hospital, or screened positive on the WHO IMCI tool.\nThere was no corresponding data collected centrally against which to compare.\nOnly a half of all the sick babies had a positive IMCI screening tool (control 29.2% (14/48) compared to the intervention 31.4% (16/51)).\nAdverse events\nThere were no reported adverse reactions to the ABHR.\nThere were, however, 15 severe adverse events (SAEs) unrelated to ABHR.\nThese consisted of three stillbirths and one baby born with a cleft lip and palate (which was later repaired) in the intervention villages.\nIn the control villages, there were two neonatal deaths: one occurred 16\u2009h after birth due to asphyxia and another at 9\u2009days after birth due to severe sepsis.\nThe other SAEs were prolonged hospitalizations in both clusters.\nVerbal autopsy\nThere were five verbal autopsy assessments performed by research midwives before all investigators reviewed the completed assessment forms.\nOne was an early neonatal death that occurred 16\u2009h after birth, and the likely cause of death was asphyxia with no signs or symptoms of sepsis.\nHowever, in the late neonatal death that occurred 9\u2009days after birth, the autopsy revealed the cause of death to be severe sepsis.\nThe rest of the verbal autopsies were performed on the three stillbirths, and no cause of death was found.\nIntracluster correlation coefficients\nThe intracluster correlation coefficients (ICC) for possible infant infection in the first 90 postnatal days using the WHO IMCI screen were 0.238 in the control villages versus 0.171 in the intervention villages, and 0.171 for all\u00a0clusters considered together.\nThe corresponding values for the composite measure of infection (anyone with a positive screen, maternal report, or clinical diagnosis) were 0.179 in the control villages, 0.194 in the intervention villages, and 0.180 for both clusters together.\nThe important changes to the methods after pilot trial commencement\nThe introduction of the expert children was an important change in the trial.\nThese reinforced the mothers and other carers or visitors to use ABHR based on the three moments for hand hygiene poster.\nThey further encouraged mothers to take their newborns to the clinics for immunisation.\nOur planned follow-up schedule was that research team would only follow up participants on day 1 and day 90 after birth.\nHowever, the research team often visited the mothers outside the planned follow-up schedule, especially during recruitment of another eligible mother or follow-up of other mothers after birth for their scheduled dates of study appointment within the same or nearby village.\nDiscussion\nThis pilot trial demonstrated that it was feasible to conduct a cluster randomised controlled trial of antenatal distribution of ABHR to prevent infections among newborns in the communities.\nThe absence of any adverse reactions suggests that ABHR was safe to the newborns and communities.\nThis pilot trial supports the conduct of a larger cluster randomised controlled trial in regard to cluster randomisation, village consent, screening, recruitment, individual informed consent rates, high fidelity, and acceptability of the ABHR intervention which was far above the set predetermined acceptable criteria of more than 80%.\nIn this pilot trial, we demonstrated the relevance of home recruitment in promoting follow-up and retention of participants in community trials.\nRecruiting participants from their homes, with the GPS coordinates automatically captured onto the ODK system, made follow-up of participants simple even by another research assistant (who simply followed coordinates).\nThere was no loss to follow-up.\nThe well-established network of village health team workers contributed to the ability to fully recruit pregnant women in this trial in each village.\nOnly 176 pregnant women were required to be screened to fully recruit into this pilot trial, which was 1.7 times the enrolled sample.\nThis translates into achievable recruitment rate for the large main cluster randomised controlled trial in the region.\nWe have demonstrated how the use of the VHWs for the distribution of ABHR refills worked efficiently.\nCommunity health workers have been demonstrated to be an important bridge to health care in other studies.\nHowever, pharmacy distribution was difficult with some mothers being refused ABHR and some ABHR missing from the pharmacies.\nThe study, further, confirms the efficient use of the ODK electronic data collection system in the field as seen in other African studies.\nThe reportedly high handwashing rate at baseline in both types of clusters could be explained by the behavioural changes following the study information, and possibly a cholera outbreak in the region that occurred prior the recruitment period and might have left a number of people informed about the need for handwashing.\nThe pilot trial showed that not all pregnancies end up into live births.\nSome ended up into stillbirths and neonatal deaths in these African settings in line with the existing literature.\nIn the planned main cRCT, considerations should be made to exclude stillbirths and neonatal deaths from further follow-up after death due to the sensitivity of their situation.\nFurther, these women might benefit from bereavement support services.\nThere is a value in supporting bereaved families through neonatal death and beyond.\nThe training of mothers in the intervention clusters to use the ABHR based on the three moments for community neonatal hand hygiene was a novel part of this pilot trial.\nIt aimed at reinforcing the practice of hand rub for critical moments in the newborn\u2019s life.\nRelated training based on the WHO\u2019s 5 Moments of Hand Hygiene has been reported to improve hand hygiene compliance in the health facilities.\nFuture research would include investigating the perspectives of women and baby carers towards these three moments of hand hygiene and whether this influences hand hygiene compliance in the homes.\nThis pilot provided useful practical insights to explore further in the conduct of the main trial.\nOne example is the creation of an \u2018expert child\u2019 within the home to improve adherence to the intervention.\nThough the expert children encouraged the use of the ABHR by anyone handling the baby and reminded the mother to take the baby to the health facility for immunisation and in the event of infant fever or any other concerns, there is not enough evidence in support of their effectiveness to promote home adherence to medical interventions.\nWe, hence, would propose exploring this further in the main trial design and any other African community settings.\nPeer mentors and support groups have been found to improve adherence to HIV/AIDS interventions in the adult population, but it is not yet established whether the expert children would have a related benefit in a large community hand hygiene intervention.\nThe WHO IMCI screening tool was used in this pilot trial to describe the primary outcome of newborn infection.\nThe pilot trial revealed difficulties with the IMCI screening tool as a measure of the primary outcome.\nIt was difficult for the VHWs to accurately and completely fill the screening tool, with many of their screening forms having the wrong or missing vitals for temperature, pulse, and respiratory rate.\nRoutinely, the VHWs are not routinely trained in measuring these vital signs and many have little formal education.\nIn the future trial, therefore, we recommended that the VHWs complete as much of the screening tool as they could and refer any suspected sick babies directly to the facilities where the screening tool would be accurately completed by skilled health workers/midwives.\nThe screening tool also missed babies where the mother sought care through informal routes such as local pharmacies, drug shops, or private clinics with no designated staff for the infection screening process.\nIn this economically deprived rural community, many women take their babies for care at local pharmacies and private clinics.\nA reliance on the attendance of babies at the government clinics for the primary outcome, therefore, misses a number of infants with infection.\nFor a future main trial, these difficulties could be avoided by carefully providing good training of frontline health workers in the use of the WHO IMCI criteria for the detection of infection in young infants.\nIn other settings, therefore, with better trained VHWs, the IMCI may still be an appropriate outcome.\nThis was a small pragmatic intervention study which is feasible to conduct, and its aim was not to cause lasting social change or implement a programme.\nHowever, there is some evidence that the trial intervention may have affected care-seeking behaviour.\nIn the intervention group, there was a doubling of the number of women receiving antibiotics for their babies or being hospitalised, and more women reporting infant infection.\nThough this is in line with\u00a0the theory of behaviour\u00a0change framework, it could benefit exploring it\u00a0further in the main trial.\nIn this open study, where women are having daily contact with the intervention, it produces a daily reminder of the importance of reducing infection-related morbidity in the babies.\nThis could reduce infection-related morbidity through preventative means (for example increasing the frequency of handwashing after toilet use) or by seeking earlier care for the infant at the first sign of any fever or any other symptom.\nHowever, it could also emphasise to women the danger of infection and lead to an increase in care-seeking behaviour and antibiotic use.\nIt is critical therefore to use an objective outcome measure for all women.\nUse of the IMCI screen as an outcome would have provided this as it is independent of the stage of the illness and would have partially negated the effect of early care seeking.\nAlternatives would have been clinician-diagnosed or microbiologically proven infection.\nThese would have had the same benefits but require massive resources.\nMaternally reported infant antibiotic use may be an alternative primary outcome and has the benefit of picking up babies with infection, irrespective of where care was sought.\nAlthough it is likely to overestimate the rate of true infections (whether diagnosed by health care worker or microbiologically), it reflects a mixture of infection-related symptoms of the infants and maternal care-seeking behaviour.\nIt is also a very important public health outcome given the actual and social cost of antibiotic use.\nHowever, a move to antibiotic use alone for the primary outcome measure would place the study at risk of ascertainment bias as women in the ABHR arm potentially seek care for their infants at an earlier stage.\nThis bias could be eliminated through the use of a placebo hand rub in the control arm.\nThis would allow a change of design to an individualised randomised trial, but there are obvious practical and ethical difficulties with the design and use of a placebo ABHR.\nThus, we propose that, for the main trial, the primary outcome is a composite of infant infections in the first 90\u2009days of life, defined as any one or more of (i) diarrhoea; (ii) lower respiratory tract infections; (iii) omphalitis; (iv) IMCI danger sign(s), all verified by health worker; (v) hospitalisation; or (vi) death.\nEach of the above composite items will also be secondary outcomes in their own right.\nThese are meaningful clinical endpoints that are feasible to collect and are relatively unambiguous.\nThese outcome measures for the proposed main trial have been selected, to reflect the diversity of detection methods for sepsis as learned from the pilot trial.\nEach component of the primary outcome measure will be individually interrogated to ensure no illogical effects within the composite.\nDiarrhoea has long been regarded as a disease of poverty and is closely associated with poor hand hygiene of carers and unsanitary home environments.\nLower respiratory tract infections are commonly spread to infants through droplets on hands (rather than by aerosol).\nBoth are therefore preventable with good hand hygiene.\nThis would accelerate adoption and/or optimization of prevention products for poverty-related diseases in sub-Saharan Africa for use in pregnant women, newborns, and/or children.\nSample size in this trial was time-bound to accommodate the number of participants who would ably be followed up within the timelines of the 6\u2009months.\nThe results of this pilot trial allow the sample size estimation for a future cluster randomised controlled trial comparing ABHR and usual care.\nThe intracluster coefficient (ICC) estimate from this pilot trial was 0.17 (95% CI 0\u20130.65); however, a larger study reported ICC estimates from five similar context cluster randomised trials predominantly in the range 0.01 to 0.10, which is considered more realistic than the pilot trial ICC.\nWe hence propose to use this ICC instead of the pilot trial one for the future trial.\nTo detect a reduction in the infection rate of \u2265\u200925% for control group rates down to 15% if ICC\u2009\u2264\u20090.01, or to detect a reduction in the infection rate of \u2265\u200933% for control group rates down to 5% if ICC\u2009\u2264\u20090.001 with a 90% power, 5932 participants are required.\nWe have hence ignored multiple births for this calculation.\nThis sample size should be achievable over a 2-year period of recruitment and across 72 villages.\nThe sample size calculations are based on a primary endpoint of severe infection, with estimated rates of 5\u201330%.\nLimitations of the study\nThis pilot trial was limited in the estimate of potential ABHR effect due to the small sample size, as one would expect in a pilot and feasibility study.\nA larger cRCT is needed to draw conclusions about the effect of ABHR on the prevention of newborn infections in the community.\nAlthough research midwives and clinicians objectively evaluated the newborns and mothers during the face-to-face encounters, we were only able to use self-reported assessments for a number of baseline variables and outcome measures.\nThis included handwashing practices, ABHR acceptability, use of ABHR, and incidence of newborn infections from participants.\nThis may have led to subjective bias in a number of ways.\nWe assumed that women\u2019s reports were correct and accurate and that women used the ABHR correctly instead of disposing it or using it for other purposes.\nConclusion\nThis pilot trial results suggest that it is feasible to conduct a cluster randomised controlled trial (cRCT) of the provision of ABHR to postpartum mothers for the prevention of neonatal infection-related morbidity in the community in resource-poor settings.\nOur results indicate that home recruitment promotes excellent follow-up and retention of participants in community trials.\nThis study informed the substantial changes necessary in the larger cRCT.\nOur study enabled us to calculate ICC for the sample size calculation of a future cRCT.\nWe also recommend a change in the primary outcome of our study to a composite outcome considering multiple methods of infection detection in a larger cRCT.\nThe intervention was safe, and a large BabyGel cluster randomised controlled trial is now required.\nThe village map showing the villages drawn and the distribution of participants in each village for the BabyGel pilot trial. A map of villages around the community health facilities drawn locally. The 10 BabyGel study villages were selected from the above map. There is also a Google Earth map showing the distribution of all the study participants from each respective village. Each number (e.g. 0412) is the assigned identifier for a participant in her household. The first two digits of the number is a study village number (e.g. 04) and the preceding two digits represent the consecutive number for each participant as recruited in each village. The first five (01 to 05) were intervention while the last five (06 to 10) were control villages. This distribution of participants shows clearly the careful selection of villages to observe the effects of contamination in this study\nCONSORT diagram showing the flow of participants through each stage of the BabyGel pilot cluster randomised trial. This CONSORT flow chart illustrates the screening and randomisation of clusters and the flow of participants in the study\nThe \u2018three moments for community neonatal hand hygiene\u2019 poster developed for the BabyGel pilot trial. This shows an illustrative and diagrammatic representation of the key moments of hand hygiene for newborns in the community\nAdapted WHO IMCI screening tool for infection. This tool was administered to newborns at particular encounters as a screening tool for infection\nIllustration of the method for the identification of infection-related infant outcomes in the BabyGel pilot trial. Infants participating in the study had the IMCI screening tool administered either at routine research visits or when their mothers brought them to participating hospitals or health centres. The group who screened positive were referred to the paediatric team at Mbale Hospital for clinical care, and along with those who died, composed the primary outcome for the study. Those who screened negative returned to the community. At the hospital, those who had screened positive had both a clinical and bacterial diagnosis performed; these were collected as secondary outcomes\n\nDemographic/background characteristics (baseline assessment)\n | Control group | Intervention group | Both groups combined\nSample size | 48 | 55 | 103\nAge (mean, s.d., range) | 24.8 (5.6) [15\u201337] | 25.0 (5.7) [15\u201338] | 24.9 (5.7) [15\u201338]\nMarital status\n\u2003Single (n, %) | 10 (20.8) | 12 (21.8) | 22 (21.4)\n\u2003Married (n, %) | 37 (77.1) | 43 (78.2) | 80 (77.7)\n\u2003Widowed (n, %) | 1 (2.1) | 0 | 1 (1.0)\nHighest level of education attained\n\u2003No formal education (n, %) | 2 (4.2) | 1 (1.8) | 3 (2.9)\n\u2003Did not complete primary ed. (n, %) | 22 (45.8) | 31 (56.4) | 53 (51.5)\n\u2003Completed primary (PLE) (n, %) | 16 (33.3) | 15 (27.3) | 31 (30.1)\n\u2003Completed ordinary (UCE) (n, %) | 4 (8.3) | 5 (9.1) | 9 (8.7)\n\u2003Completed advanced (n, %) | 3 (6.3) | 2 (3.6) | 5 (4.9)\n\u2003Completed tertiary (n, %) | 1 (2.1) | 1 (1.8) | 2 (1.9)\nPrimary occupation\n\u2003Unemployed (n, %) | 6 (12.5) | 3 (5.5) | 9 (8.7)\n\u2003Housewife (n, %) | 23 (47.9) | 25 (45.5) | 48 (46.6)\n\u2003Student (n, %) | 0 | 3 (5.5) | 3 (2.9)\n\u2003Peasant farmer (n, %) | 18 (37.5) | 20 (36.4) | 38 (36.9)\n\u2003Businesswoman (n, %) | 0 | 2 (3.6) | 2 (1.9)\n\u2003Professional (n, %) | 1 (2.1) | 0 | 1 (1.0)\n\u2003Other (n, %) | 0 | 2 (3.6) (hotel management, teacher) | 2 (1.9)\nHouse roof type\n\u2003Iron sheet (n, %) | 39 (81.3) | 52 (94.5) | 91 (88.3)\n\u2003Grass thatched (n, %) | 9 (18.8) | 3 (5.5) | 12 (11.7)\nHouse floor type (more than one response possible)\n\u2003Mud (n, %) | 42 (87.5) | 49 (89.1) | 91 (88.3)\n\u2003Brick (n, %) | 10 (20.8) | 11 (20.0) | 21 (20.4)\n\u2003Cement/stone/tile (n, %) | 5 (10.4) | 6 (10.9) | 11 (10.7)\n\u2003Other (n, %) | 1 (2.1) (bricks and mud) | 0 | 1 (1.0)\nThe main water source for home\n\u2003Open water well (n, %) | 12 (25.0) | 1 (1.8) | 13 (12.6)\n\u2003Piped and tapped to home (n, %) | 1 (2.1) | 4 (7.3) | 5 (4.9)\n\u2003Shared tap/borehole (n, %) | 35 (72.9) | 50 (90.9) | 85 (82.5)\nType of latrine used in home\n\u2003No latrine (use bushes) (n, %) | 3 (6.3) | 8 (14.5) | 11 (10.7)\n\u2003Non-ventilated pit (n, %) | 44 (91.7) | 44 (80.0) | 88 (85.4)\n\u2003Ventilated improved pit (VIP) (n, %) | 1 (2.1) | 3 (5.5) | 4 (3.9)\nIf latrine, type of handwashing facility\n\u2003No facility (n, %) | 32 (71.1) | 38 (80.9) | 70 (76.1)\n\u2003Near latrine (n, %) | 13 (28.9) | 4 (8.5) | 17 (18.5)\n\u2003Away from latrine (n, %) | 0 | 5 (10.6) | 5 (5.4)\nAnimals/poultry reared/kept (n, %) | 45 (93.8) | 51 (92.7) | 96 (93.2)\n\u2003Cows and goats (n, %) | 35 (77.8) | 37 (72.5) | 72 (75.0)\n\u2003Poultry (n, %) | 41 (91.1) | 49 (96.1) | 90 (93.8)\n\u2003Other (n, %) | 12 (26.7) | 11 (21.6) | 23 (24.0)\n\u2003\u2003Pigs (n, %) | 11 (24.4) | 11 (21.6) | 22 (22.9)\n\u2003\u2003Ducks (n, %) | 0 | 1 (2.0) | 1 (1.0)\n\u2003\u2003Rabbits (n, %) | 1 (2.2) | 0 | 1 (1.0)\nTimes (out of 10) hands washed when\n\u2003Urinating (median[range]) | 5 [0\u201310] | 4 [0\u201310] | 5 [0\u201310]\n\u2003\u20030 (n, %) | 8 (16.7) | 9 (16.4) | 17 (16.5)\n\u2003\u20031\u20134 (n, %) | 13 (27.1) | 19 (34.5) | 32 (31.1)\n\u2003\u20035\u20139 (n, %) | 11 (22.9) | 12 (21.8) | 23 (22.3)\n\u2003\u200310 (n, %) | 16 (33.3) | 15 (27.3) | 31 (30.1)\n\u2003Defecating (median [range]) | 10 [1\u201310] | 10 [4\u201310] | 10 [1\u201310]\n\u2003\u20030 (n, %) | 0 | 0 | 0\n\u2003\u20031\u20134 (n, %) | 9 (18.8) | 1 (1.8) | 10 (9.7)\n\u2003\u20035\u20139 (n, %) | 12 (25.0) | 9 (16.4) | 21 (20.4)\n\u2003\u200310 (n, %) | 27 (56.3) | 45 (81.8) | 72 (69.9)\nLast 10 handwashes, used\n\u2003Water alone (n, %) | 6 (12.5) | 7 (12.7) | 13 (12.6)\n\u2003Water and bar soap (n, %) | 42 (87.5) | 48 (87.3) | 90 (87.4)\nPregnancy planned (n, %) | 23 (47.9) | 30 (54.5) | 53 (51.5)\nExisting medical conditions\n\u2003Asthma (n, %) | 2 (4.2) | 0 | 2 (1.9)\n\u2003Cardiac disease (n, %) | 0 | 0 | 0\n\u2003Coagulation disorder (n, %) | 0 | 0 | 0\n\u2003Congenital abnormalities (n, %) | 0 | 0 | 0\n\u2003Diabetes (type 2) (n, %) | 1 (2.1) | 0 | 1 (1.0)\n\u2003High blood pressure (n, %) | 0 | 0 | 0\n\u2003Malaria (n, %) | 6 (12.5) | 14 (25.5) | 20 (19.4)\n\u2003Tuberculosis (n, %) | 0 | 0 | 0\n\u2003HIV\n\u2003\u20030 (n, %) | 32 (66.7) | 39 (70.9) | 71 (68.9)\n\u2003\u20031 (n, %) | 7 (14.6) | 8 (14.5) | 15 (14.6)\n\u2003\u20032 (n, %) | 9 (18.8) | 8 (14.5) | 17 (16.5)\n\u2003STD (n, %) | 1 (2.1) | 0 | 1 (1.0)\n\u2003Other (n, %) | 0 | 0 | 0\n\n\nBirth outcome (mother)\n | Control group | Intervention group | Difference (95% CI)\nSample size | 48 | 55 | 103\nBirth outcome\n\u2003Singleton (n, %) | 47 (97.9) | 55 (100.) | 2.1 (not calculable)\n\u2003Twin (n, %) | 1 (2.1) | 0\nOutcome at initial assessment\n\u2003Baby/babies survived (n, %) | 47 (97.9) | 51 (92.7) | \u2212\u20095.2 (\u2212\u200913.2:2.8)\n\u2003Stillbirth (n, %) | 0 | 3 (5.5) | \n\u2003Baby died within 24\u2009h (n, %) | 1 (2.1) | 0 | \u2013\n\u2003Mother withdrew consent (n, %) | 0 | 1 (1.8) | \n\u2003Mother lost to follow-up after initial assessment (n, %) | 0 | 1 (2.1) | \u2013\n\u2003Baby died after 10\u2009days (n, %) | 1 (2.1) | 0 | \nRevised sample size for initial postnatal assessment | 48* | 51 | \nMode of birth\n\u2003Spontaneous vaginal birth (n, %) | 46 (95.8) | 49 (96.1) | 0.2 (\u2212\u20097.5:8.0)\n\u2003Caesarean section (n, %) | 2 (4.2) | 2 (3.9)\nPlace of birth\n\u2003Home (n, %) | 7 (14.6) | 6 (11.8) | \u2212\u20092.8 (\u2212\u200916.2:10.5)\n\u2003En route (n, %) | 2 (4.2) | 1 (2.0) | \u2212\u20092.2 (\u2212\u20099.0:4.6)\n\u2003Health centre (n, %) | 31 (64.6) | 40 (78.4) | 13.8 (\u2212\u20093.8:31.5)\n\u2003Hospital (n, %) | 4 (8.3) | 2 (3.9) | \u2212\u20094.4 (\u2212\u200913.9:5.0)\n\u2003Other (n, %) | 4 (8.3) | 2 (3.9) | \u2212\u20094.4 (\u2212\u200913.9:5.0)\nPerson assisting birth (n, %) | 48 (100.) | 50 (98.0) | \u2212\u20092.0 (not calculable)\n\u2003Relative (n, %) | 6 (12.5) | 9 (17.6) | 5.5 (\u2212\u20098.7:19.7)\n\u2003Traditional birth attendant (n, %) | 1 (2.1) | 3 (6.0) | 3.9 (\u2212\u20093.8:11.6)\n\u2003Nursing assistant (n, %) | 1 (2.1) | 0 | \u2212\u20092.1 (not calculable)\n\u2003Midwife/nurse (n, %) | 38 (79.2) | 42 (84.0) | 4.8 (\u2212\u200910.5:20.2)\n\u2003Other (n, %) | 5 (10.4) | 2 (4.0) | \u2212\u20096.4 (\u2212\u200916.6:3.8)\n\u2003\u2003Doctor | 2 | 2 | \u2013\n\u2003\u2003Neighbour | 1 | 0 | \u2013\n\u2003\u2003Retired midwife | 1 | 0 | \u2013\n\u2003\u2003Trained birth attendant | 1 | 0 | \u2013\nSex of baby\n\u2003Male | (n, %) | 21 (43.8) | 22 (43.1) | \u2212\u20090.6 (\u2212\u200920.2:18.9)\n\u2003Female | (n, %) | 27 (56.3) | 29 (56.9)\nBirthweight | Mean (s.d.) | 3.3 (0.6) [2.0\u20134.5]1 | 3.4 (0.4) [2.8\u20134.6]2 | 0.11 (\u2212\u20090.20:0.42) (p\u00a0=\u20090.478)\n\n1n\u00a0=\u200925\n2n\u00a0=\u200924\n*Mothers\u2009=\u200947, but as 1 mother had twin birth, babies\u2009=\u200948\n\nABHR distribution in intervention villages (n\u00a0=\u20095) with participants (N\u00a0=\u200955)\n | Via VHW | Via pharmacy | Overall | Difference (95% CI)\nVillages assigned | 60.0% (3/5) | 40.0% (2/5) | 100% (5/5) | 20 (\u221267.7:107.7)\nWomen wished to receive refills (self-reported) | 70.9% (39/55) | 29.1% (16/55) | 100% (55/55) | 41.8 (15.4:68.2)\nWomen actually refilled (self-reported at follow-up) | 65.5% (36/55)2 | 25.5% (14/55)3 | 90.9% (50/55) | 40 (12.4:67.6)\nABHR volume dispensed to women (litres; mean, range) | 3.7 (2.8:4.1) | 2.5 (2.0: 3.1) | 3.1 (2.0: 4.1) | 1.2 (1.09:1.31)\nABHR use/day (self-reported; mean, range) | 7.0 (6.0:10) | 6.2 (5.0:8.1) | 6.6 (6.0:10.0) | 0.8 (04.46:1.14)\nABHR accountability\n\u2003ABHR volume delivered (litres) | 40 | 20 | 60 | \u2013\n\u2003ABHR accounted for (litres) | 40 | 15 | 55 | \u2013\n\n1One village had one recruit, who had a stillbirth and did not do any ABHR refills\n2Three mothers did not return for ABHR refills due to either stillbirth (1), study withdrawal (1), or relocation (1)\n3Two mothers did not return for ABHR refills due to stillbirths (2)\n\nInfant infection outcomes\nPrimary outcome | Control group | Intervention group | Difference (95% CI)\nn\u00a0=\u200948 | n\u00a0=\u200951\nScreen positive on IMCI form at any point within 90\u2009days after birth | 14 (29.2%) | 16 (31.4%) | 2.2 (\u2212\u200915.9:20.3)\nSepsis deaths | 1 (2.1%) | 0 | \u2212\u20092.1 (not calculable)\nClinically diagnosed infection in health centre or hospital | 8 (16.7%) | 11 (21.6%) | 4.9 (\u2212\u200910.5:20.3)\nMother reported infant infection | 19 (39.6%) | 26 (51.0%) | 11.4 (\u22128.1:30.9)\nInfant received antibiotics | 8 (16.7%) | 14 (27.5%) | 10.8 (\u2212\u20095.4:26.9)\nInfant was hospitalised | 3 (6.3%) | 6 (11.8%) | 5.5 (\u2212\u20095.7:16.7)\nTotal with any evidence of non-malaria infection (positive screen, maternal report, or clinical diagnosis) | 27 (56.3%) | 28 (54.9%) | \u2212\u20091.3 (\u2212\u200920.9:18.2)\n",
        "label": "low",
        "id": "task4_RLD_test_43"
    },
    {
        "paper_doi": "10.1186/1471-2458-11-438",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Setting: cinema, the NetherlandsDesign: quasi-randomised controlled trialRecruitment: announcements in local newspapers, radio, and on the Internet. Other recruitment methods included posting flyers in mailboxes and handing out flyersAllocation to group: allocated according to evening available\n\n\nParticipants: 89 participants. Mean age of 50.44 (SD 12.35), 26.4% were male, 33% were overweight or obese, 50.5% had moderate educational level and 41.4% had high\n\n\nInterventions: Intervention: large poster with portion size and caloric guidelines for daily amounts (GDA) information on soft drinks (n = 48)Control: no label; different portion sizes for soft drinks were displayed indicating only the amount of millilitres that each cup contained (n = 41)\n\n\nOutcomes: Soft drink consumed (mL) during film was calculated by electronic weighing of leftovers\n\n\nNotes: Participants could choose between five portion sizes (200 mL, 250 mL, 400 mL, 500 mL and 750 mL cups). The study took place on two subsequent evenings during which participants could order free soft drinks. Authors were contacted to request information about the energy content of the soft drinks, but this information was not forthcoming. Information on study funding was not reporte\n\n",
        "objective": "To assess the impact of nutritional labelling for food and non\u2010alcoholic drinks on purchasing and consumption of healthier items. Our secondary objective was to explore possible effect moderators of nutritional labelling on purchasing and consumption.",
        "full_paper": "Background\nLarge soft drink sizes increase consumption, and thereby contribute to obesity.\nPortion size labelling may help consumers to select more appropriate food portions.\nThis study aimed to assess the effectiveness of portion size and caloric Guidelines for Daily Amounts (GDA) labelling on consumers' portion size choices and consumption of regular soft drinks.\nMethods\nA field experiment that took place on two subsequent evenings in a Dutch cinema.\nParticipants (n = 101) were asked to select one of five different portion sizes of a soft drink.\nConsumers were provided with either portion size and caloric GDA labelling (experimental condition) or with millilitre information (control condition).\nResults\nLabelling neither stimulated participants to choose small portion sizes (OR = .75, p = .61, CI: .25 - 2.25), nor did labelling dissuade participants to choose large portion sizes (OR = .51, p = .36, CI: .12 - 2.15).\nConclusions\nPortion size and caloric GDA labelling were found to have no effect on soft drink intake.\nFurther research among a larger group of participants combined with pricing strategies is required.\nThe results of this study are relevant for the current public health debate on food labelling.\nBackground\nThe mean portion size of soft drinks has increased in the past decades, and over time larger portion sizes have been added to the product lines.\nSoft drinks have been recognized as potentially important contributors to obesity and it has been demonstrated that serving larger soft drink portions results in increased beverage consumption.\nNext to the availability of larger portion sizes, 'portion distortion' might stimulate the consumption of increasingly larger amounts of soft drinks.\nNutrition labelling could help consumers to make healthy choices, and many different formats of nutrition labels, varying in design and complexity, are currently being used.\nGuidelines for Daily Amounts (GDA labelling) is one example and gives consumers standards against which they can evaluate the number of calories that a food or drink serving provides.\nIn the UK, GDA labelling was introduced by many manufacturers and retailers in 1998, whereas in continental Europe, GDA's are gradually gaining acceptance.\nOther labelling formats that have been implemented internationally are for instance the Multiple Traffic Light system, the Heart Symbol, and the Choices logo.\nPortion size labelling could both be a promising and feasible intervention to help consumers to select appropriate portion sizes.\nEspecially in Europe, portion size labelling is currently not a widespread practice, and a standard format does not yet exist.\nHowever, a pilot study on the most effective format for portion size labelling indicated that providing consumers with a reference portion size was the most promising format.\nAll in all, portion size information combined with caloric GDA labelling may help consumers choose appropriate portion sizes and moderate the effects of portion distortion in a complex food environment that provides several and large portion sizes.\nThe few experimental studies that have explored the effectiveness of portion size labelling on consumption provide inconclusive results.\nAlso, previous studies have shown that once food is served, people find it difficult to regulate their intake.\nIt is therefore important to assess the impact of labelling both on portion size choices as well as on consumption.\nFurthermore, it is important to assess the impact of labelling in more realistic settings than the laboratory.\nThe aim of the present study was to assess the impact of portion size and caloric GDA labelling on consumers' portion size choices and consumption of regular soft drinks.\nMethods\nBrief Overview\nThe study, that took place on two subsequent evenings, employed an experimental between subject design with an experimental condition with portion size and caloric GDA labelling (second evening) and a control condition (first evening).\nIn both conditions, participants could choose between five portion sizes (200, 250, 400, 500 and 750 millilitre cups).\nThese portion sizes were selected as being representative of the portion sizes currently available in the Netherlands.\nThe experimental manipulation consisted of information displayed near the bar where participants ordered their drinks.\nIn the experimental condition, portion sizes were presented on a display with both the number of portions each cup represented and the caloric GDA information (see Figure 1).\nPortion sizes were based on guidelines from the Netherlands Nutrition Centre (an institution funded by the Dutch government that provides information and education about healthy nutrition) that defines one portion of soft drink as 225 millilitres.\nHowever, as 225 millilitre cups were not the market standard, the 250-millilitre cup was designated as the reference portion.\nThe smallest size was labelled as 0.8 portions, and the largest size was labelled as three portions.\nIn the control condition, different portion sizes for soft drinks were displayed indicating only the amount of millilitres that each cup contained.\nThe comprehensibility of both the portion size and the caloric GDA information was pretested with satisfying results.\nRecruitment Procedures\nParticipants were recruited through announcements in local newspapers, radio, and on the internet.\nOther recruitment methods included posting flyers in mailboxes and handing out flyers.\nPotential participants were told that a marketing study was conducted into consumers' attitudes towards cinemas.\nParticipants, unknowing of the study conditions, could choose the evening that was most convenient for them to participate.\nThe true purpose of the study was not revealed until the conclusion of the experiment.\nParticipants were considered eligible if they were between 21 and 65 years of age.\nParticipants received a gift voucher worth \u20ac10, -.\nParticipants\nThere were 101 participants in the study.\nAfter excluding participants who had not ordered a soft drink (n = 12), the experimental condition consisted of 48 participants and the control condition consisted of 41 participants.\nOverall, participants' mean age was 50.44 (12.35), 26.4% were male and 33% were overweight or obese.\nSee Table 1 for further details.\nStudy Procedures and Data Collection\nA cinema was chosen as the location for the study because it is a setting in which a diverse range of people can be found.\nUpon arrival at the cinema, participants received the first questionnaire and were assigned a unique number and asked to write it on each questionnaire.\nParticipants self-completed the first questionnaire in the lobby, before the beginning of the film.\nThis questionnaire consisted of spurious questions about the participants' previous cinematic experiences and mood.\nIn addition, this questionnaire contained one control question measuring thirst using a visual analogue scale ranging from 0 (not at all thirsty) to 10 (very thirsty).\nSubsequently, participants were invited to the bar for a free regular soft drink.\nAfter participants received their drinks, they were invited to watch the film.\nWhen the movie was finished, participants were asked to fill out the second questionnaire.\nThe second questionnaire consisted of items that were to be used as control variables in the data analyses.\nThe questionnaire began by asking participants what they thought was the true purpose of the study.\nSubsequently, they were asked a number of questions regarding their soft drink consumption (i.e. general consumption frequency, and whether they made a habit of drinking diet or regular soft drinks).\nAdditionally, participants were asked if they had seen the displays situated above the bar.\nA number of 5-point Likert items regarding the participants' opinions of the display material and their self-reported impact of the labelling were also included.\nTo measure the participants' self-reported impact of the labelling, they were asked to rate on a 5-point Likert scale whether labelling had affected their portion size choices and soft drink consumption.\nFurthermore, participants were asked whether labelling had made them aware of appropriate soft drink portion sizes\nAdditionally, the dietary restraint and external disinhibition scales derived from the Dutch Eating Behaviour Questionnaire (DEBQ,) were included in the second questionnaire.\nDietary restraint (i.e. the deliberate restriction of energy intake with the intent to decrease or maintain weight) was measured with a scale consisting of ten 5-point Likert items (e.g. 'Do you try to eat only a little when you want to eat a lot?') with \u03b1 = .91.\nExternal disinhibition (i.e. overeating in response to external food-related cues such as sight and smell of attractive food) was measured with ten 5-point Likert items (e.g. 'If food smells yummy, do you eat a lot of it?') with \u03b1 = .83.\nThe questionnaire also contained questions on gender, age, height, and body weight.\nWhen the participants had completed the second questionnaire, they were asked to mark their participant number on their cup, and to hand in their cups and questionnaires to the research assistants.\nIf soft drink remained in the cups, this amount was weighed afterwards.\nThis study was approved by the VU Medical Centre's Institutional Review Board.\nWritten informed consent was obtained from all subjects.\nData analysis\nLogistic regression and chi square analyses were run in order to assess the impact of labelling on participants' portion size choices.\nSince we considered it relevant to assess 1) whether labelling had an effect on selecting reference portion sizes of soft drinks, and 2) whether labelling had an effect on selecting one of the two largest soft drink sizes, the data were dichotomized and coded in two different ways.\nFirst, the portion size choices were dichotomized in order to assess whether labelling stimulated participants to choose the reference portion size or smaller (i.e. 250 or 200 millilitres).\nTherefore, participants' portion size choices were either coded as the reference portion size or smaller, or as being larger than the reference portion size.\nSecond, portion size choices were dichotomized in order to assess the effect of labelling on discouraging participants from choosing one of the two largest portion sizes (i.e. 500 or 750 millilitres).\nData were either dichotomized as choosing one of the two largest portion sizes, or not choosing one of the two largest portion sizes.\nTo assess the impact of portion size and caloric GDA labelling on soft drink consumption and to assess the self-reported impact of labelling, General Linear Model procedures were used.\nThe dependent variable was either the amount of soft drink consumed, or the self-reported impact of labelling on 1) size choice, 2) soft drink consumption or 3) portion size awareness.\nBecause we randomized the study conditions instead of the individual participants, we could not rule out differences in background characteristics that are likely to be related to choice and consumption behaviour of soft drinks.\nTherefore, both the logistic analyses and the General Linear Models were adjusted for these variables (i.e. age, gender, BMI, external disinhibition, dietary restraint, thirst, and a preference for diet versus regular soft drinks).\nResults\nOn the whole, 59.8% of the participants had noticed the displays (68.8% in the experimental and 48.7% in the control condition, \u03c72 (1) = 3.59, p = .06).\nImpact of Labelling on Choice and Consumption Behaviour\nOverall, 37.5% chose the reference amount or smaller.\nA chi square analysis did not show a significant difference between both conditions, see Figure 2.\nThe logistic regression analyses indicated that portion size labelling did not increase the likelihood of choosing the reference portion size or smaller (OR = .75, p = .61, CI: .25 - 2.25).\nFurthermore, portion size labelling did not dissuade participants to choose one of the two largest portion sizes (OR = .51, p = .36, CI: .12 - 2.15).\nFinally, no significant effects of labelling were found on soft drink consumption (experimental condition: Mean = 376.30, SD = 125.40, control condition: Mean = 382.14 SD = 147.60), F (1, 71) = .39, p = .50.\nSelf-reported Impact of Labelling\nWith respect to the participants' self-reported impact of labelling, results showed no differences between both conditions on portion size choices, F (1, 46) = 2.31, p = .14.\nHowever, a significant interaction effect was found between labelling and gender, F (1, 46) = 6.66, p = .01.\nSpecifically, for women the self-reported impact on choice behaviour was slightly higher in the experimental condition (Mean = 2.76, SD = 1.48) than in the control condition (Mean = 2.20, SD = 1.58).\nWhereas, for men the self-reported impact was lower in the experimental condition (Mean = 1.50, SD = .71) compared to the control condition (Mean = 3.20, SD = 1.48).\nFinally, no significant results were found for the self-reported impact of labelling on consumption F (1, 47) = .15, p = .70 or on portion size awareness, F (1, 47) = .17, p = .68.\nDiscussion\nThis study was one of the first experimental studies that are known to us, that assessed the impact of portion size and caloric GDA labelling on consumers' regular soft drink portion size choices, their intake of soft drinks, and their self-reported awareness of portion sizes.\nThe study results did not demonstrate significant effects of portion size labelling on increasing the likelihood of selecting one of the reference sizes or decreasing the likelihood of selecting the largest sizes.\nWith respect to the latter however, it is relevant to note that the OR of selecting one of the largest sizes was lower in the experimental condition than in the control condition.\nA lack of power might explain that this result was not significant.\nTherefore, we conclude that portion size labelling did not have an effect on selecting reference portion sizes of soft drink, and that further research is needed to assess the impact of labelling on selecting large portion sizes.\nWith respect to the self-reported impact of portion size labelling on portion size choices, it seemed that labelling had a neutral effect on women, but a detrimental impact on men.\nAlthough this gender difference was not found for participants' actual consumption, this finding is partly in line with other studies showing that women generally attach greater importance to healthy eating than men and report more health information seeking behaviour.\nIt is therefore recommended to further study gender differences in consumers' responses to labelling.\nAn important factor that might explain that that we found no effect for GDA labelling is that a large majority of the participants indicated that they never or seldomly drank regular soft drinks.\nConsequently, this could make portion size and caloric GDA labelling less relevant for them.\nIn order to assess whether labelling was more effective among participants who reported drinking soft drinks regularly, the logistic analyses were also run among this subgroup of participants.\nDue to a lack of power these results could not be tested for significance, but the OR's did not indicate that portion size labelling had a beneficial impact on portion size choices (results not shown).\nIn this study we were interested in the effect of portion size labelling on portion size choices, as opposed to the replacement of regular products by diet products.\nDiet soft drinks were therefore unavailable and, as a result, participants who only drank diet soft drinks might have refused the free regular soft drink.\nAnother consequence is that we could not test the potential effect of portion size labelling on the selection of diet soft drinks instead of regular soft drinks.\nIn addition, participants did not have to purchase their drinks, obviating the cost of the drink from affecting portion size choice.\nIt is unclear how pricing would affect the impact of labelling.\nOn the one hand, free soft drinks might have stimulated participants to select larger portion sizes than they would normally have if they had to pay.\nOn the other hand, point of purchase settings employ value size pricing to stimulate consumers to choose large portion sizes too.\nNevertheless, it would be interesting to assess the impact of portion size and GDA labelling combined with proportional pricing (i.e. eliminating beneficial pricing for large portion sizes by keeping the price per millilitre consistent).\nAlso, about 40% of the participants in both conditions had not noticed the displays.\nWe chose to include all participants in the analyses, regardless of whether they had seen the displays.\nThe reason for this was that the results from these analyses would be more generalizable to real world settings in which people often oversee nutrition labels.\nIt is nevertheless worth mentioning that when the logistic regression analyses were run solely on participants who had seen the displays, comparable OR's were found (results not shown).\nAnother issue is that with respect to the participants' BMI, in this study we had to rely on self-reported data that might have suffered from a social desirability bias and under-reporting.\nWe expect that the amount of underreporting was approximately the same for both conditions, but random measurement errors resulting from the self-reported data might still have caused some residual confounding.\nLast, some researchers have suggested that multiple exposures (i.e. seeing the labels more often) may be required in order for labelling to become effective.\nFurther research on portion size and caloric GDA labelling among a larger number of participants is necessary to draw more definitive conclusions.\nIt is suggested to conduct studies that provide participants with multiple exposures to labelling and studies in which labelling is combined with pricing strategies.\nFuture studies might benefit from more objective methods to define the participants' BMI.\nLast, it is recommended to gain more insight into gender differences related to labelling.\nConclusions\nPortion size and caloric GDA labelling were found to have no effect on regular soft drink portion size choices and intake.\nFurther research with multiple exposures combined with pricing strategies among a larger number of people who have a habit of drinking regular soft drinks is recommended.\nDisplay material in the experimental condition.\nCup size choices (in %) in both study conditions1. 1\u03c72 (4) = 3.58, p = .47.\n\nParticipant characteristics\n |  | Total sample (n = 89) | Experimental condition (n = 48) | Control condition (n = 41)\n |  | Mean (SD) or % | Mean (SD) or % | Mean (SD) or %\n\nAge |  | 50.44 (12.35) | 50.12 (12.17) | 50.82 (12.71)\nSex (female) |  | 73.6 | 68.8 | 79.5\nThirst |  | 6.36 (2.87) | 6.28 (2.73) | 6.46 (3.06)\nDietary restraint |  | 2.92 (.74) | 3.00 (.68) | 2.82 (.79)\nExternal disinhibition |  | 2.81 (.52) | 2.84 (.54) | 2.80 (.50)\nEducational level\n | Low | 8 | 8.3 | 7.7\n | Moderate | 50.5 | 45.9 | 56.4\n | High | 41.4 | 45.8 | 35.9\nWeight status1\n | Underweight2 | 3.2 | 2.2 | 2.6\n | Healthy weight3 | 63.8 | 68.9 | 53.8\n | Overweight4 | 27.7 | 26.7 | 33.3\n | Obese5 | 5.3 | 2.2 | 10.3\nSoft drink consumption frequency\n | Never | 47.1 | 56.3 | 35.9\n | Seldom | 32.2 | 27.1 | 38.5\n | Sometimes | 13.8 | 12.5 | 15.4\n | Often | 5.7 | 2.1 | 10.3\n | Very often | 1.1 | 2.1 | 0\nHabitually drinks regular soft drink (when drinking soft drink) | 55.3 | 54.3 | 56.4\nInferred that study was about soft drink consumption and health. | 13.4 | 15.6 | 10.8\nHad seen display | 59.8 | 68.8 | 48.7\n\n1In the Netherlands, 35% of the population are considered to be overweight and 11% are obese\n2 BMI < 18.50\n3 BMI 18.50-24.99\n4 BMI 25.00-29.99\n5 BMI \u2265 30.00\nNote: No significant differences were found with respect to age, sex, BMI, dietary restraint, external disinhibition, thirst, and educational level between participants in the experimental and in the control condition.",
        "label": "high",
        "id": "task4_RLD_test_134"
    },
    {
        "paper_doi": "10.5021/ad.2013.25.1.17",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Trial designMulticentre, randomised, double-blind, half-sided trialTrial registration numberNot reportedSettingMulticentre; it does not specifically mention location however all authors are from South Korea and the trial was approved by Institutional Review Boards at 5 South Korean hospitals.Date trial conductedNot reportedDuration of trial participation15 daysAdditional design detailsNoneInclusion criteriaPatients with moderate-severe symmetrical eczematous skin lesionsExclusion criteriaPatients < 4 years oldPatients currently undergoing treatment with systemic glucocorticoids, antibiotics or immunosuppressive agentsPatients treated with UV radiationPatients with other chronic non-eczematous skin diseases, also those with infectious dermatosesPatients with a chronic medical illness such as diabetesPatients who were pregnant or lactatingPatients with skin lesions involving the face or genital areaPatients with other severe dermatoses or scarsNotesPrior to the start of the trial, participants taking a systemic corticosteroid had a washout period of 4 weeks, and participants who applied a TCS had a washout period of 1 week.\n\n\nParticipants: Total number randomised175 participants (350 sides of the body)AgeFor the 159 participants who were analysed the mean age was 32.32 years (SD 19.86, range 5-79).SexOf 159 participants who contributed data there were 76 male (47.80%) and 83 female (52.20).Race/ethnicityNot reportedDuration of eczemaPaper states that 25 (15.72%) participants had \"past skin disease history\", whilst 134 (84.28%) did not.Severity of eczemaBaseline IGA of clinical response was 7.46 +- 3.11 in the mometasone furoate group, whilst 7.51 +- 3.18 in the methylprednisolone aceponate group. The index was calculated from assessment of 4 signs/symptoms: erythema, vesiculation, pruritus, and burning/pain where the physician rated each parameter on a 0-3 scale: 0 = no symptoms, 1 = mild, 2 = moderate and 3 = severe. The paper does not describe how this was calculated; possibly it was summed for each participant and a mean calculated.Filaggrin mutation statusNot reportedNumber of withdrawals15 participants were excluded due either to violation of protocols or adverse reactions, and 1 participant was excluded due to a screening criteria violation. It is unclear which group they were allocated to.NotesNone\n\n\nInterventions: Run-in detailsNAGroupsMometasone furoate cream; applied in a multi-lamellar emulsion cream to 1 side of the body once daily for 2 weeks. Concurrent treatment: noneMethylprednisolone aceponate cream; applied to 1 side of the body once daily for 2 weeks. Concurrent treatment: noneAdherenceNot reportedCo-interventionsThe mometasone furoate preparation also contained multi-lamellar emulsions (paper suggests this can aid restoration of the barrier function of the skin) and hexylene glycol (an antimicrobial excipient).NotesThe concentrations of the 2 TCS preparations were not stated in the paper. Therefore, the methylprednisolone preparation was assumed to be 0.1% and mometasone assumed to be 0.1%. This was  because these were the only concentrations identified in the reference sources that we used to identify potency.\n\n\nOutcomes: Adverse events (included those deemed treatment-related or not) at up to day 15*TEWL: measured by Tewameter(r) to evaluate epidermal permeability barrier function (Courage & Khazaka, Cologne, Germany). The TEWL improvement ratio was calculated as: TEWL improvement ratio (%) = [(TEWLday1-TEWLdayn)/ TEWLday1] x 100 (%) at day 1, 4, 8, and 15.IGA. The IGA index was calculated from assessment of 4 signs/symptoms: erythema, vesiculation, pruritus, and burning/pain where the physician rated each parameter on a 0-3 scale: 0 = no symptoms, 1 = mild, 2 = moderate and 3 = severe. The IGA improvement ratio was calculated as: IGA improvement ratio (%) = [(IGAday1-IGAdayn)/IGAday1] x 100 (%) at day 1, 4, 8, and 15.*VAS for pruritus: improvement of pruritus after treatment was scored subjectively, \"using 10 visual analogue scales that the patients scored\". We assume this means participants marked the severity of the itch on a 10 mm VAS. The VAS improvement ratio was calculated as: VAS improvement ratio (%) = [(VASday1-VASdayn)/VASday1] x 100 (%) at day 1, 4, 8, and 15.**denotes relevance to this review\n\n\nFunding source: None stated\n\n\nDeclarations of interest: None declared\n\n\nNotes: The mometasone furoate preparation also contained multi-lamellar emulsions (the paper suggests this can aid restoration of the barrier function of the skin) and hexylene glycol (an antimicrobial excipient)\n\n",
        "objective": "To establish the effectiveness and safety of different ways of using topical corticosteroids for treating eczema.",
        "full_paper": "Background\nTopical application of corticosteroids also has an influence on skin barrier impairment.\nPhysiological lipid mixtures, such as multi-lamellar emulsion (MLE) containing a natural lipid component leads to effective recovery of the barrier function.\nObjective\nThe purpose of this study was to conduct an evaluation of the therapeutic efficacy and skin barrier protection of topical mometasone furoate in MLE.\nMethods\nA multi-center randomized, double-blind, controlled study was performed to assess the efficacy and safety of mometasone furoate cream in MLE for Korean patients with eczema.\nThe study group included 175 patients with eczema, who applied either mometasone furoate in MLE cream or methylprednisolone aceponate cream for 2 weeks.\nTreatment efficacy was evaluated using the physician's global assessment of clinical response (PGA), trans-epidermal water loss (TEWL), and visual analogue scale (VAS) for pruritus.\nPatients were evaluated using these indices at days 4, 8, and 15.\nResults\nComparison of PGA score, TEWL, and VAS score at baseline with those at days 4, 8, and 15 of treatment showed a significant improvement in both groups.\nPatients who applied mometasone furoate in MLE (74.8%) showed better results (p<0.05) than those who applied methylprednisolone aceponate (47.8%).\nThe TEWL improvement ratio was higher in the mometasone furoate in MLE group than that in the methylprednisolone aceponate group, and VAS improvement was also better in the mometasone furoate in MLE group.\nConclusion\nMometasone furoate in MLE has a better therapeutic efficacy as well as less skin barrier impairment than methylprednisolone aceponate.\nINTRODUCTION\nTopical steroids are an important and common treatment modality for eczematous skin disorders.\nThe choice of topical steroid varies based on skin disease severity, potency, and delivery vehicle used.\nAppropriate use of a topical steroid makes it possible to treat an eczematous disorder effectively.\nHowever, they are many notable side effects such as skin atrophy, acneiform eruptions, hypertrichosis, hypopigmentation, and development of a cutaneous infection.\nSuch topical glucocorticoid side effects make many patients hesitant to use topical steroids.\nEfforts to minimize the side effects of topical steroids have been attempted.\nOne approach is to develop novel steroids such as non-halogenated double-ester-type glucocorticoids.\nAnother technique is to use topical steroid together with an anti-atrophogenic substance.\nIn particular, topical materials that not only minimize side effects of topical steroid but enhance the physiological lipid mixture have been investigated.\nAnother important aspect of eczema treatment is normalizing the defective skin barrier, because impaired skin barrier function is an important factor in the pathogenesis of eczema.\nSkin barrier function is affected by multiple factors, including downregulation of filaggrin and locogrin, reduced ceramide levels, increased proteolytic enzyme levels, and enhanced trans-epidermal water loss (TEWL).\nThe epidermal barrier is composed of a combination of corneocytes and intercellular lipids.\nThe stratum corneum (SC) provides a mechanical protection to the skin.\nIt functions as a barrier to water loss and permeation of soluble substances from the environment.\nRegulation of permeability, desquamation, antimicrobial peptide activity, toxin exclusion, and selective chemical absorption are all primary functions of the extracellular lipid matrix.\nIn contrast, mechanical reinforcement, hydration, cytokine-mediated initiation of inflammation, and protection from ultraviolet (UV) damage are all provided by corneocytes.\nAmong various factors, hydrophobic intercellular lipids are the most important factor for skin barrier function, and diminished ceramides create a leaky barrier.\nParticularly in the case of atopic dermatitis (AD), abnormal barrier function results from disruption of the multi-lamellar structure due to a significant reduction in the amount of intercellular lipid ceramides in the SC.\nA disrupted skin barrier can be replenished by a physiological lipid mixture.\nMulti-lamellar emulsions (MLEs) containing pseudoceramide has a multi-lamellar structure similar to intercellular lipids of the SC.\nLee et al. reported that patients with AD treated only with a MLE improved more than those applying a commercial moisturizing cream in an objective assessment and subjective satisfaction scores for symptoms and signs.\nAhn et al. reported on a co-application of MLE for topical steroid-protected skin barrier function, reinforcing the skin barrier permeability.\nTopical application of ceramides results in restoration of barrier function by reducing TEWL.\nEffective enhancement of skin barrier function with a physiological lipid mixture has been reported recently.\nChamlin et al. reported on alleviation of childhood AD using a ceramide-dominant, barrier-repair lipid, and attributed the improvement seen in their patients to a normalization of barrier function, which in turn dampened the cytokine cascade that initiates and sustains AD.\nMometasone furoate is a potent topical steroid with proven efficacy, similar to betametasone.\nMometasone furoate contains hexylene glycol, which has antimicrobial properties.\nThe effects of hexylene glycol on microorganisms may potentially be beneficial for treating eczematous disorders, with effects on microorganisms, possibly leading to better treatment and a longer relapse-free period.\nIt is also a safe and effective method for a long-term use to treat chronic, recurrent disease.\nThis study was performed to evaluate the clinical efficacy of this physiological lipid mixture as a vehicle of mometasone furoate in patients with eczema.\nWe designed a multi-center, randomized, double-blind controlled assessment to compare mometasone furoate in MLE with methylprednisolone aceponate.\nMATERIALS AND METHODS\nPatients\nPatients with eczema and moderate to severe manifestations were enrolled.\nAll patients showed eczematous skin lesions, presenting symmetrically.\nPatients <4 years old were excluded.\nThe following patients were also excluded from the study: patients currently undergoing treatment with systemic glucocorticoids, antibiotics or immunosuppressive agents; those treated with UV radiation; those with other chronic non-eczematous skin diseases; those with infectious dermatoses, chronic medical illness such as diabetes; those pregnant or lactating; those with skin lesions involving the face or genital area; and those with other severe dermatoses or scars.\nWe used a wash-out period for patients who had undergone treatment with topical and systemic corticosteroids.\nPatients taking a systemic corticosteroid had a wash-out period of 4 weeks.\nPatients who applied a topical corticosteroid had a wash-out period of 1 week.\nThis study was approved by the Institutional Review Boards at Incheon St. Mary's Hospital, Kangnam Sacred Heart Hospital, Seoul National University Hospital, Severance Hospital, and Kwandong University Myongji Hospital.\nStudy design\nAfter informed consent was obtained, the patients were randomly assigned to apply mometasone furorate in MLE on one side and apply methylprednisolone aceponate on the other side.\nThe topically applied formulations were mometasone furoate in MLE and methylprednisolone aceponate.\nTogether on each randomly assigned side, mometasone furoate in MLE was applied on one side for 2 weeks, and methylprednisolone aceponate was applied on the other side for 2 weeks.\nMometasone furoate in MLE or methylprednisolone aceponate was applied topically to skin lesions once daily by all subjects.\nThe study included a baseline visit before treatment, and on days 1, 4, 8, and 15 after treatment initiation.\nPatients were observed and assessed by one physician on days 1, 4, 8, and 15 throughout the trial.\nThe physician's global assessment of clinical response (PGA) score was adopted for an objective assessment of the clinical response to treatment.\nThe PGA index was calculated from scales of erythema, vesiculation, pruritus, and burning/pain.\nEach parameter was judged on a 0~3 scale: 0=no symptoms, 1=mild, 2=moderate, and 3=severe.\nClinical efficacy was assessed by the PGA improvement ratio.\nThe PGA improvement ratio was calculated as:\nPGA improvement ratio (%)=[(PGAday1-PGAdayn)/PGAday1]\u00d7100 (%)\nThe TEWL was measured by Tewameter\u00ae to evaluate epidermal permeability barrier function (Courage & Khazaka, Cologne, Germany) on days 1, 4, 8, and 15.\nClinical efficacy of the improved skin barrier function was assessed by the TEWL improvement ratio.\nThe TEWL improvement ratio was calculated as:\nTEWL improvement ratio (%)=[(TEWLday1-TEWLdayn)/TEWLday1]\u00d7100 (%)\nIn addition, improvement of pruritus after treatment was assessed subjectively, using 10 visual analog scales that the patients scored.\nThe visual analogue scale (VAS) improvement ratio was calculated as:\nVAS improvement ratio (%)=[(VASday1-VASdayn)/VASday1] \u00d7100 (%)\nAll adverse events were recorded, and whether they were treatment related or not was also noted.\nStatistical analysis\nMcnemar's t-test was used to compare the PGA improvement ratio on days 1, 4, 8, and 15 of treatment.\np<0.05 were regarded as statistically significant.\nA paired t-test and the Wilcoxon signed-rank test were used to verify significant differences in TEWL and VAS scores between groups A and B during the follow-up period.\nRESULTS\nSummary of patients\nA total of 175 patients were initially enrolled.\nFifteen patients were excluded due either to violation of protocols or adverse reactions, and one patient was excluded due to a screening criteria violation.\nIn total, 159 patients were analyzed (76 males and 83 females; age range, 5~79 years; mean age, 32.32\u00b119.86, mean\u00b1standard deviation years old).\nNo clinically significant differences were observed in the PGA score, TEWL, or VAS between the mometasone furoate in MLE group and the methylprednisolone aceponate group.\nBasal demographic characteristics of the study groups and basal results of the PGA, TEWL, and VAS scores are summarized in Table 1 and 2.\nClinical efficacy of mometasone furoate in MLE\nWe performed a data analysis on patients who completed 15 days of treatment to assess efficacy.\nComparison of the PGA score improvement ratio at baseline with days 4, 8, and 15 of treatment showed a significant increase for all follow-up periods.\nAt baseline, the mean PGA score was 7.46\u00b13.11 in the mometasone furoate in MLE group and 7.51\u00b13.18 in the methylprednisolone group.\nAfter 15 days, the PGA improvement ratio in the mometasone furoate in MLE group was 82.62\u00b121.62%, and that in the methylprednisolone acetonate group was 68.32\u00b124.05% (p\u22640.0001).\nThe PGA improvement ratio for days 4, 8, and 15 is summarized in Fig. 1.\nTEWL improvement in the mometasone furoate in MLE group\nAt baseline, the baseline TEWL score in the mometasone furoate in MLE group was 33.73\u00b122.47 g/h/m2 and 33.05\u00b121.96 g/h/m2 in the methylprednisolone acetonate group.\nNo significant difference was observed between the groups.\nAfter 15 days of treatment, the TEWL improvement ratio increased in both groups for all follow-up periods.\nAfter 15 days of treatment in the mometasone furoate in MLE group, the TEWL improvement ratio, which was 48.30\u00b168.04% at baseline, was statistically significant; and in the methylprednisolone acetonate group, the TEWL improvement ratio increased 32.74\u00b150.07% from baseline after 15 days.\nAlthough the TEWL improvement ratio increased in both groups, the TEWL improvement ratio in the mometasone furoate in MLE group was superior to that in the methylprednisolone acetonate group (p\u22640.0001).\nThe TEWL improvement ratio for days 4, 8, and 15 is summarized in Fig.\n2. The intergroup differences were statistically significant at every point.\nVAS improvement in the mometasone furoate in MLE group\nThe subjective VAS score was measured at every visit to evaluate improvement of pruritus.\nThe initial VAS score was 5.83\u00b12.31 in the mometasone furoate in MLE group and 5.99\u00b12.29 in the methylprednisolone aceponate group (p>0.05).\nThe VAS improvement ratio increased in both groups for all follow-up periods.\nAfter 15 days of treatment, the VAS improvement ratio score increased 83.28\u00b123.47% from baseline in the mometasone furoate in MLE group and 75.41\u00b127.24% in the methylprednisolone group.\nIn addition, pruritus showed a more significant improvement in the mometasone furoate in MLE group than that in the methylprednisolone group (p\u22640.0001).\nThe VAS improvement ratio at days 4 and 8 also showed an increase from baseline; however, no statistically significant intergroup difference was observed (p=0.2117, p=0.1131).\nThe VAS improvement ratio on days 4, 8, and 15 is summarized in Fig.\n3. The difference was statistically significant at day 15 (p\u22640.0001).\nAdverse effects\nAn itching sensation was observed in two patients (1.15%) who applied mometasone furoate in MLE and in 4 patients (2.30%) who applied methylprednisolone aceponate.\nUrticaria was observed in one patient (0.57%).\nHerpes virus infection was observed in one patient (0.57%) at a non-applied site.\nFive patients were excluded from the study due to pruritus and urticaria.\nDISCUSSION\nBoth mometasone furoate in MLE and methylprednisolone aceponate were effective treatments for eczema.\nStatistically significant differences in the PGA score were observed between the values at baseline and at day 15 of treatment.\nPatients who applied mometasone furoate in MLE showed better results in the PGA improvement ratio than those who applied methylprednisolone aceponate.\nBoth agents are potent group II corticosteroids, and they had shown similar efficacy in previous studies.\nHowever, mometasone furoate in MLE showed superior efficacy, likely due to the effects of the physiological lipid mixture on enhanced skin barrier function.\nThis was supported by the results showing TEWL improvement, as it mirrors the skin barrier function.\nThe TEWL improvement ratio was also higher in those who applied mometasone furoate in MLE.\nProlonged treatment with a topical steroid creates structural defects in the epidermis, which has been attributed to a disturbance in epidermal differentiation and thinning of the SC.\nA disrupted skin barrier due to abnormalities in intercellular lipid lamellae, which are thought to mediate transcutaneous water loss, results in an increase in TEWL score.\nThe VAS improvement ratio was higher with the physiological lipid mixture.\nApplying mometasone furoate in MLE resulted in significant improvement in clinical symptoms and signs of eczematous disorder.\nSubjectively, patients felt less of an itching sensation over the treatment period.\nTopical corticosteroids not only have antiproliferative effects but also suppress differentiation of the epidermal layer, resulting in defects in the epidermis.\nTherefore, long-term use of topical glucocorticoid causes weakening of the skin barrier.\nMany studies have been performed to minimize the side effects of glucocorticoids.\nOne attempt involves reinforcing the skin barrier function using a physiological lipid mixture.\nMan MQ et al. reported that topical application of a physiological lipid mixture results in accelerated recovery of barrier function, whereas an incomplete lipid mixture may inhibit the normal recovery response.\nAdditionally, topical application of an MLE containing a pseudoceramide results in significantly decreased TEWL and skin pH.\nAhn et al. reported that concurrent application of MLE with steroid significantly reduces skin atrophy induced by steroid.\nThe physiological lipid mixture using a vehicle allows for the penetration of topical steroid into the skin and enhances delivery.\nMometasone furorate in MLE showed a more beneficial effect than that of methylprednisolone aceponate, which has similar potency but not the physiological lipid mixture.\nAs such, the treatment period for a topical steroid would be shortened using the same potency.\nThe side effects of topical steroids are dependent on duration and frequency of application.\nIt is possible that the lower steroid potency causes similar effects to a higher potency steroid.\nWe suggest that a physiological lipid mixture as a topical steroid vehicle enhances the effects of a topical steroid and diminishes the side effects.\nUse of moisturizers is an important treatment method to improve skin hydration, even when overt disease is not observed.\nDaily treatment with topical moisturizers prevents exacerbation of skin lesions by reducing elevated TEWL.\nThe MLE-containing moisturizers are effective for hydration of eczema and have a safety profile.\nA physiological lipid mixture using pseudoceramides is presented in an orthorhombic lipid phase, which chiefly exists on the human SC lipid, while non-physiological lipid mixture moisturizers only show a liquid crystalline phase and a hexagonal phase.\nAn increase in TEWL results in skin dehydration, and xerosis cutis itself aggravates eczematous disease and makes pruritus more severe.\nPruritus, the most common and important symptom of an eczematous skin disorder, causes a vicious itching-scratching cycle.\nTherefore, rapid amelioration of pruritus shortens disease duration and enhances treatment compliance.\nIn recent studies, a physiological lipid mixture using pseudoceramides not only restored the skin barrier but also had an anti-inflammatory effect.\nKang et al. reported that topical application of MLE using pseudoceramides results in diminished mRNA expression of interleukin (IL)-4 and tumor necrosis factor-\u03b1 in murine atopic dermatitis-like skin lesions.\nIL-4 and IL-13 play a major role in allergic reactions.\nA suppressed immune response diminishes the cytokine cascade involved in the pathogenesis of pruritus.\nIn our study, only a few local side effects were observed such as pruritus and urticaria.\nThe intensity of the itching sensation was mild, and the duration was a few hours to a day.\nAs such, local side effects rarely required discontinuation of treatment.\nIn conclusion, mometasone furoate in MLE was a more effective treatment for moderate to severe skin eczema than that of methylprednisolone.\nAs mometasone furoate in MLE induced improvement in skin barrier function, and we recommend more studies using of physiological lipid mixtures as a vehicle for topical therapeutic agents.\nThe physician's global assessment of clinical response (PGA) improvement ratio at day 4, 8, 15. Comparison of the PGA improvement ratio between themometasone furoate in multi-lamella emulsion (MLE) group and the methylprednisolone group. A significant difference was observed in the PGA improvement ratio at all follow-up periods (p\u22640.0001).\nThe trans-epidermal water loss (TEWL) improvement ratio. Comparison of the TEWL improvement ratio between the mometasone furorate in multi-lamella emulsion (MLE) group and the methylprednisolone group. A significant difference was observed in the TEWL improvement ratio at all follow-up periods (p\u22640.0001).\nThe visual analog scale (VAS) improvement ratio. Comparison of the VAS improvement ratio between the mometasone furorate in multi-lamella emulsion (MLE) group and the methylprednisolone group. A significant difference was observed in the VAS improvement ratio at day 15 (p\u22640.0001).\n\nBase demographic characteristics of the study group\nValues are presented as mean\u00b1standard deviation or number (%).\n\nBaseline results of PGA, TEWL, and VAS score\nValues are presented as mean\u00b1standard devation. PGA: physician's global assessment of clinical response, TEWL: transepidermal water loss, VAS: visual analog scale, MLE: multilamellar emulsion.",
        "label": "unclear",
        "id": "task4_RLD_test_456"
    },
    {
        "paper_doi": "10.1371/journal.pone.0009696",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Parallel arm cluster-RCT conducted in Mirzapur, Bangladesh, between Dec 2003 and Dec 2006.\n\n\nParticipants: Sample size: 12 clusters (21,140 individuals randomised, 10,700 women with at least 1 pregnancy during 10 preceding months analysed).Clusters: rural unions surrounding an urban central union (excluded from the study) served by a 750 bed private referral-level hospital.Individuals: all married women of reproductive age (i.e. 15-49 years) in the intervention arm were eligible for enrolment. Women in the survey were eligible if they had had a pregnancy outcome in the last 3 years.\n\n\nInterventions: Target: health system (addition of home visits).Arm 1: 2 home visits (12-16 and 32-34 weeks); they were given a labour card for women to present upon arrival at hospital for delivery and 3 postnatal visits on days 2, 5 and 8. CHWs facilitated free-of-charge transfer of ill neonates to hospital.The purpose of the antenatal component of the intervention was to increase uptake of ANC (3 visits taking place at home or at a health centre or satellite clinic - distinct from the 2 antenatal CHW home visits), tetanus toxoid vaccination, general pregnancy and newborn care education, and birth preparedness (including delivery at a health facility).Arm 2: standard ANC.\n\n\nOutcomes: Trial primary outcomes: antenatal and immediate newborn care behaviours, knowledge of danger signs, care seeking for neonatal complications, and neonatal mortality.Review outcomes reported:Primary: not reported.Secondary: ANC coverage (at least 1 visit), health facility delivery, IPT for malaria, neonatal mortality.\n\nFollow-up: data collection at delivery and during pre and postnatal home visits. Endline survey Jan - May 2006, before the end of the trial.\n\n\nNotes: Funders: The Wellcome Trust: Burroughs Wellcome Fund Infectious Disease Initiative 2000 and the Office of Health, Infectious Diseases and Nutrition, Global Health Bureau, United States Agency for International Development (USAID) through the Global Research Activity Cooperative agreement with the Johns Hopkins Bloomberg School of Public Health (award HRN-A-00-96-90006-00). Support for data analysis and manuscript preparation was provided by the Saving Newborn Lives program through a grant by the Bill & Melinda Gates Foundation to Save the Children-US. The study was registered at clinicaltrials.gov, No. NCT00198627\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Background\nTo evaluate a delivery strategy for newborn interventions in rural Bangladesh.\nMethods\nA cluster-randomized controlled trial was conducted in Mirzapur, Bangladesh.\nTwelve unions were randomized to intervention or comparison arm.\nAll women of reproductive age were eligible to participate.\nIn the intervention arm, community health workers identified pregnant women; made two antenatal home visits to promote birth and newborn care preparedness; made four postnatal home visits to negotiate preventive care practices and to assess newborns for illness; and referred sick neonates to a hospital and facilitated compliance.\nPrimary outcome measures were antenatal and immediate newborn care behaviours, knowledge of danger signs, care seeking for neonatal complications, and neonatal mortality.\nFindings\nA total of 4616 and 5241 live births were recorded from 9987 and 11153 participants in the intervention and comparison arm, respectively.\nHigh coverage of antenatal (91% visited twice) and postnatal (69% visited on days 0 or 1) home visitations was achieved.\nIndicators of care practices and knowledge of maternal and neonatal danger signs improved.\nAdjusted mortality hazard ratio in the intervention arm, compared to the comparison arm, was 1.02 (95% CI: 0.80\u20131.30) at baseline and 0.87 (95% CI: 0.68\u20131.12) at endline.\nPrimary causes of death were birth asphyxia (49%) and prematurity (26%).\nNo adverse events associated with interventions were reported.\nConclusion\nLack of evidence for mortality impact despite high program coverage and quality assurance of implementation, and improvements in targeted newborn care practices suggests the intervention did not adequately address risk factors for mortality.\nThe level and cause-structure of neonatal mortality in the local population must be considered in developing interventions.\nPrograms must ensure skilled care during childbirth, including management of birth asphyxia and prematurity, and curative postnatal care during the first two days of life, in addition to essential newborn care and infection prevention and management.\nTrial Registration\nClinicaltrials.gov NCT00198627\nIntroduction\nNeonatal mortality declined by approximately 20% over the last decade in Bangladesh, however, the rate of decline was less than in the postneonatal and 1\u20134 year-old periods.\nNeonatal deaths now account for almost half of under-5 child deaths in Bangladesh and efforts to reduce neonatal mortality are crucial to achieving Millennium Development Goal 4 for child survival.\nSince 90% of births and most neonatal deaths still occur at home, community-level interventions must be introduced while linking with the healthcare system for treatment of life-threatening newborn illness.\nSeveral recent community-based trials of packages of maternal and neonatal interventions in low resource settings in South Asia have shown statistically significant reductions in neonatal mortality, employing a variety of healthcare delivery approaches.\nThe focus of the interventions, however, has been primarily on averting deaths due to serious infections.\nHome-based health education and routine neonatal assessment and antibiotic treatment of serious infections by community health workers (CHWs) decreased mortality in rural India and rural northeastern Bangladesh, although regulatory approval for and availability of CHWs for home-based treatment of illness is lacking in most settings.\nA preventive maternal and neonatal care behavior change management program implemented by CHWs through home visits as well as community mobilization also reported a mortality reduction of 54% in a very high mortality area of Uttar Pradesh, India.\nLady health workers in Hala, Pakistan, promoted essential maternal and newborn care through home visits, community education group sessions, and linkages with local traditional birth attendants (TBAs), resulting in a 28% reduction in mortality.\nStudies without home-based interventions also reported mortality reductions of about 30% through community-based participatory interventions in Nepal and by improving TBAs' clean delivery practices and strengthening their linkages with primary health facilities in Larkana, Pakistan.\nTo provide cost-effective essential preventive and curative services in low-resource settings, strategies must take into account the risk factors for and causes of mortality, the quality and accessibility of the health care system, and community perception and acceptance of the interventions.\nCommunity-based preventive care coupled with basic management of childhood illness and facilitated referral by CHWs is a potentially effective model where access to quality health care at facilities can be ensured.\nWe developed a preventive service delivery strategy in a rural area of central Bangladesh with good access to facility-based care to promote household newborn care practices through home visits by CHWs, and conducted routine, home-based illness surveillance coupled with facilitated referral of sick newborns to health facilities.\nA cluster-randomized controlled trial was conducted to examine its impact on knowledge and practice of newborn care and neonatal mortality.\nMethods\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nStudy Population and Design\nProjahnmo-Mirzapur was a cluster-randomized, controlled intervention trial of a preventive and curative maternal-neonatal healthcare package, in which was nested surveillance for community-acquired neonatal bacteremia..\nThe trial was implemented in Mirzapur, a sub-district of Tangail district, Dhaka division, Bangladesh, located 2 hours by car from the capital city of Dhaka, during January 2004\u2013 December 2006.\nThe neonatal mortality rate (NMR) was estimated at 24 per 1000 live births in 2002.\nThe area was served by Kumudini Hospital \u2013 a 750-bed, private, referral-level hospital, located in a central urban union which was excluded from the study.\nThe remaining population of about 292,000 was divided into 12 rural unions, which were randomly allocated to either comparison or intervention arm using a computer-generated pseudo-random number sequence without stratification or matching (Figure 1).\nBlinding was unachievable given the nature of the intervention.\nNewborns in the comparison arm received the usual health services provided by the government, non-governmental organizations and private providers.\nIn the intervention arm, each union had six CHW areas, each of which consisted of approximately 4000 population served by one CHW.\nThe CHW-to-population ratio was similar to the primary healthcare worker-to-population ratio in the Bangladesh government health system, thus facilitating sustainability and scalability of the healthcare delivery strategy.\nAll married women of reproductive age (i.e., 15\u201349 years) in the intervention arm were eligible for enrolment, and were administered informed verbal consent by the CHW in their area.\nDesign and Implementation of Interventions\nCommunity-level interventions were developed based on findings from formative research on newborn care practices in the study population, conducted during November 2002\u2013 April 2003.\nInformation on pregnancy, delivery, immediate newborn care and care seeking for newborn illness was collected through 26 unstructured interviews with women, husbands, mothers-in-law, and TBAs, and through semi-structured, in-depth interviews with 40 women and/or family members and 54 healthcare providers, including TBAs, health workers of BRAC or village associations, and village doctors.\nFindings from formative research were used to design the communications and negotiation approach to promote safe and clean delivery and preventive, household newborn care practices (Table 1).\nCHWs were trained for 36 days on pregnancy surveillance, counseling and negotiation skills, essential newborn care, neonatal illness surveillance and management of illness based on a clinical algorithm adapted from Integrated Management of Childhood Illness.\nAfter initial training and evaluation, routine monitoring and refresher training were provided each fortnight.\nFurther information on recruitment, characteristics, training and monitoring of CHWs is presented elsewhere.\nIn addition, TBAs serving in the intervention unions (n\u200a=\u200a84) attended a two-day orientation session on the aims and activities of the project, essential newborn care practices, and indications for referral of newborns and mothers.\nTable 1 presents detailed information on interventions provided by CHWs in the intervention arm.\nCHWs identified pregnancies in their population through bimonthly household pregnancy surveillance.\nBirth and newborn care preparedness (BNCP) was promoted by CHWs through two antenatal home visits scheduled at 12\u201316 and 32\u201334 weeks of gestation.\nCHWs gave a labor notification card to each woman with instructions for a family member to seek out and present the card to the CHW when the pregnant woman started into labor.\nCHWs, notified by the card, attended the delivery whenever possible, or visited the mother and newborn infant as early as possible in the postnatal period.\nCHWs conducted three additional postnatal visits on days 2, 5 and 8 to promote preventive newborn care practices and to identify and refer sick neonates to Kumudini Hospital.\nDuring each of the postnatal visits, CHWs completed a standardized newborn assessment form, identified the presence of serious illnesses requiring referral to Kumudini Hospital \u2013 including illness indicative of infections, potentially requiring antibiotic treatment \u2013 and made referral to the hospital according to the clinical algorithm.\nCHWs' classification of neonates with illness had high validity compared to physicians' classification.\nUse of the clinical algorithm by CHWs during routine household surveillance was also validated in identifying severely ill neonates needing urgent referral to the hospital and those who subsequently died.\nTo eliminate potential barriers to care seeking for illness, CHWs facilitated transport, if necessary, for neonates needing referral-level evaluation at Kumudini Hospital, and all care at the hospital was free-of-charge for referred neonates.\nThe mean travel time to the hospital was about one hour, and formative research suggested positive community perception of the quality of care at the hospital.\nIf the family refused to be referred, the CHW continued to encourage referral but managed the neonate in the home according to the algorithm, without use of injectable antibiotics.\nData\nIn order to examine intervention effectiveness, baseline and endline surveys were conducted in both study arms, using comparable questionnaires.\nPrimary outcome measure was neonatal mortality; secondary outcomes included antenatal and immediate newborn care behaviours, knowledge of danger signs, and care seeking for neonatal complications.\nThe surveys included all households with recently delivered women (RDW) (i.e., women who had a pregnancy outcome in the three calendar years before the survey), and collected household wealth and basic demographic information from all household members.\nTo measure the mortality outcome, a hypothesized 40% reduction during the intervention period, a total sample size of 14,872 neonates was required, based on the baseline NMR of 28 per 1000 live births, power of 80% and an estimated design effect of 2.55 derived from the baseline data.\nGiven a crude birth rate of 27 per 1000 population, 7884 live births were expected per year, and the surveys collected life-time pregnancy history from all eligible RDW at both baseline and endline.\nWe anticipated that, after three months of initial intervention scale-up, a two-year period of enrolment during which the implementation of the intervention was stabilized would be sufficient.\nIn addition, for all identified neonatal deaths during a defined period (see below), verbal autopsy data, including signs and symptoms of illness leading to deaths, were collected by separate interviewers who were trained in verbal autopsy data collection for six days.\nTo measure indicators of care practice and knowledge, the surveys also collected knowledge (K) of maternal and newborn care practices, household practice (P) of maternal-newborn preventive and curative care behaviours; and program coverage (C) among RDWs who had a pregnancy outcome in the last 12 months before each survey, hereafter referred to as KPC-RDW.\nAt baseline, all eligible KPC-RDW were interviewed, while a sample of KPC-RDW was interviewed during the endline survey.\nThe endline sample of KPC-RDW was randomly selected within each union, based on a sample size calculation to provide estimates for all KPC indicators assuming 50% prevalence with \u00b16% precision and response rate of 85% for each union.\nBaseline household listing and mapping was conducted during March \u2013 June 2003, and households with at least one RDW who had a pregnancy outcome between 2000 and 2002 were identified.\nThe baseline survey was conducted during April \u2013 July 2003.\nResponse rates were 86.9% (14532/16725) among all RDW and 92.4% among all KPC-RDW (4636/5015).\nVerbal autopsy data were collected during September \u2013 December 2003, on average 14.8 months (SD: 3.6, n\u200a=\u200a109) after the death, for neonatal deaths among those born in 2002 (response rate 88.6%, 109/123).\nThe intervention was introduced and scaled up to the entire study area during December 2003\u2013 February 2004.\nImplementation continued through December 2006.\nEndline enumeration of households with RDW who had a pregnancy outcome between 2003 and 2005 was conducted during December 2005\u2013 April 2006.\nThe endline survey was conducted during January \u2013 May 2006, before the end of the trial, in order to maintain community cooperation and minimize potential end-of-project effect in outcome measurement.\nThe response rate was 87.8% (14731/16771) among all RDW.\nThe KPC-RDW response rate was 94.0% (3519/3744).\nFor all neonatal deaths among those born during the intervention period (2004\u20132005), verbal autopsy information was collected during April \u2013 August 2006 (response rate 86.4%, 222/257).\nThe mean interval between a death and verbal autopsy data collection was 16.5 months (SD: 8.1, n\u200a=\u200a222).\nIn addition, two interim adequacy surveys of knowledge and practice were conducted to monitor the coverage or adequacy of the intervention, and to guide adjustments in the implementation to optimize coverage and quality of the intervention.\nRandom samples of households were selected from the baseline household listing for the two adequacy surveys.\nSample size was calculated to provide estimates for selected KPC indicators, assuming 50% prevalence, \u00b110% precision, and response rate of 85% for each union.\nThe first and second adequacy surveys were conducted during December 2004\u2013 January 2005 and August \u2013 September 2005, respectively.\nIn total, 1141 and 1213 women who had a pregnancy outcome in the 12 months before each survey were enumerated in the first and second adequacy surveys, respectively.\nResponse rates were 82.7% (1141/1380) for the first, and 86.5% (1194/1380) for the second adequacy survey.\nInformed verbal consent was administered by survey interviewers for all participants.\nStatistical Analysis\nWe analyzed the two adequacy surveys and the endline survey to estimate coverage changes in three consecutive 8-month periods in the intervention arm.\nAnalyses were restricted to pregnancies which ended during the following 8-month periods, to avoid overlap between surveys: April \u2013 November 2004 (from adequacy survey 1), December 2004\u2013 July 2005 (from adequacy survey 2), and August 2005\u2013 March 2006 (from the endline survey).\nCoverage of the program was assessed in three areas: antenatal (whether a CHW visited the home at least once during pregnancy), delivery (whether a CHW attended at delivery), and postnatal (whether a CHW assessed a neonate at least once within the first 2, 7, and 28 days of life, respectively, and, among those who received postnatal visits, the mean time of first visit and the mean number of visits).\nThe baseline and endline surveys were analyzed to assess changes in three main outcomes in both comparison and intervention arms: reported maternal and newborn care practices, knowledge of maternal and newborn danger signs of illness, and neonatal mortality, controlled for basic demographic and socioeconomic characteristics.\nWe first estimated means and proportions of RDW with selected background characteristics, by study arm and survey, including mother's age at birth (<20 years, 20\u201329 years, and \u226530 years), mother's educational attainment (< primary school completion vs. \u2265 primary school completion), and household wealth status.\nA household wealth index score, based on the pooled data of baseline and endline surveys, was constructed using principal component analysis of household assets.\nHouseholds in each survey were ranked based on the index score and categorized into quintiles.\nThe lowest and highest quintiles were classified as poor and rich, respectively, relative to the three middle quintiles.\nAntenatal and neonatal care practices were measured among KPC-RDW.\nThe last pregnancy was used as an index pregnancy if there were two or more pregnancies within the 12-month recall period.\nA woman was considered to have received routine antenatal care from a qualified provider (distinct from BNCP home visits by CHWs) if she had received \u22651 antenatal check-up either at a medical facility (i.e., satellite clinic, Union Health and Family Welfare Centre \u2013 a primary health facility serving approximately 20,000 population in the union, Upazila health complex \u2013 a first-level referral public hospital in each sub-district, qualified doctor's chamber, clinic or hospital) or by a qualified provider (i.e., doctor, nurse, Family Welfare Visitor \u2013 health personnel at a Union Health and Family Welfare Centre, or medical assistant).\nAmong all home-born live births, seven selected immediate essential newborn care variables were measured, including sterile cord cut (i.e., the cord was cut by either a blade which was boiled before use or a blade from a clean delivery kit); drying/wiping the baby before delivery of the placenta; wrapping the baby before delivery of the placenta; delaying the first bath to the third day of life or later; initiating breastfeeding within one hour after delivery; breastfeeding prior to giving any food or liquid; and not applying anything to the cord immediately after cutting and tying it.\nCare seeking to a qualified provider (defined above) was measured among all neonates who had signs of complications based on maternal report.\nThe baseline and endline surveys collected information on 11 and 19 complication signs, respectively, and we restricted analyses to neonates with \u22651 of 10 signs collected in both surveys (Table 2).\nKnowledge of maternal and neonatal danger signs was measured among KPC-RDW, using unprompted binary knowledge variables of 10 antenatal, 11 childbirth, 9 postpartum maternal, and 16 neonatal danger signs (Table 3).\nFour composite knowledge score variables were constructed for antenatal (range [0\u201310]), childbirth (0\u201311), postnatal maternal (0\u20139) and neonatal danger signs (0\u201316), by adding un-weighted positive answers for each of the individual signs within the category.\nComposite variables were treated as having missing values if the respondent had not completed all questions in each category.\nTo investigate differential changes in knowledge and practices, we conducted intention-to-treat analyses at the study arm level, using difference-in-difference test with interaction terms for time (baseline vs. endline) and study arm (comparison vs. intervention).\nWe estimated predicted mean of each knowledge or practice indicator by time and study arm and compared the change between baseline and endline by study arm, controlling for maternal and household background characteristics described above.\nLinear probability regression models were used to test the null hypothesis that the difference-in-difference was zero.\nRobust standard errors were adjusted for clustering on each union.\nNeonatal mortality was examined using pregnancy history by all RDW.\nWe assessed mortality data quality by examining distributions of the monthly number of live births, the monthly number of neonatal deaths, and age at deaths by year.\nWe included live births in two-calendar-year-periods, January 2001\u2013 December 2002 from the baseline survey and January 2004\u2013\nDecember 2005 from the endline survey, to control for the potential seasonal effect on mortality and to eliminate the 11-month pre-intervention period (January \u2013 November 2003) included in the 3-year pregnancy history recall period for the endline survey.\nWe estimated NMR and 95% confidence intervals (CI) by time and study arm.\nWe used a survival-time model with a Weibull survival distribution to estimate relative hazard of mortality between the study arms at baseline and at endline, adjusted for child sex and background characteristics described above.\nRobust standard errors adjusted for clustering on each union.\nFinally, verbal autopsy data were analyzed to estimate cause-specific neonatal mortality rate by time and study arm, using neonatal deaths among those born in 2002 (baseline), and in 2004\u20132005 (endline).\nWe applied a standard hierarchical algorithm to assign one primary cause out of the seven major causes of neonatal mortality in the order of: congenital malformation, tetanus, preterm birth, birth asphyxia, birth injury, sepsis or pneumonia, and diarrhea.\nA p-value of 0.05 was considered statistically significant, and all analyses of KPC indicators were adjusted for sampling weight.\nSTATA 9.0 statistical software (Stata Corporation, College Station, TX, USA) was used for all analyses.\nThe study was approved by the Committee on Human Research at the Johns Hopkins Bloomberg School of Public Health, the Ethical Review Committee and the Research Review Committee at ICDDR,B, the Ethical Review Committee at Dhaka Shishu Hospital and the Ethical Review Committee at Oxford University.\nThe study was registered at clinicaltrials.gov, No. NCT00198627.\nResults\nEnrolment\nA total of 9987 women of reproductive age had 5031 pregnancy outcomes in the intervention arm, including 302 miscarriages, 113 stillbirths and 4616 live births (Figure 2).\nIn the comparison arm, 11153 women of reproductive age had 5669 pregnancy outcomes, including 319 miscarriages, 109 stillbirths and 5241 live births.\nThere were no differences in the rates of miscarriage and stillbirth between the two arms.\nEnrolment rates did not vary across unions.\nCoverage of the Intervention\nIn the intervention arm, percent of pregnant women receiving \u22651 BNCP visit reached over 90% during the first survey period (April \u2013 November 2004) and remained comparable in the subsequent two 8-month survey periods (Table 4).\nSimilar rates were seen for receipt of two BNCP visits.\nPercent of home-deliveries attended by CHWs was 12% during April \u2013 November 2004, increased to 20% during December 2004\u2013 July 2005, but remained at 14% during August 2005\u2013 March 2006.\nPercent of home-born newborns assessed by CHWs within the first two and seven days of life improved from 54% to 69% and from 66% to 74%, respectively, over the survey periods.\nAmong those who were assessed by CHWs at least once during the first 28 days of life, the average timing of the initial assessment decreased and the mean number of assessments increased over the periods.\nPractice\nIndicators of maternal and newborn care practice and knowledge were similar in the intervention and comparison arms at baseline.\nAdjusted for significant improvement in background socioeconomic characteristics in each arm (Table 5), proportions of women who received \u22651 routine antenatal check-up (distinct from antenatal BNCP visits by CHWs) from a qualified provider and took antenatal iron supplements increased significantly in the intervention arm (reaching 69% and 56%, respectively), but not in the comparison arm (49% and 43%, respectively) (Table 6).\nThere was no change in the proportions of women who received \u22651 tetatus toxoid immunization during pregnancy in either study arm (approximately 75%), however, the proportion of women who received \u22652 tetanus toxoid immunizations during pregnancy decreased in both study arms (from about 55% to about 40%), likely associated with national shortage of the vaccine.\nPercent of women who delivered at a health facility remained small, but increased significantly more in the intervention (from 12 to 20%) than the comparison (from 13% to 17%) arm (Table 6).\nAmong all home-born neonates, sterile cord cut, delaying the first bath, early breastfeeding initiation and breastfeeding before any food or liquid increased in both arms, but the increases were substantially larger in the intervention arm than in the comparison arm, reaching about 80% or more (Table 6).\nImmediate drying and immediate wrapping of the baby improved only in the intervention arm, reaching about 14%.\nFinally, among neonates who had \u22651 of the 10 selected complication signs, care seeking from a qualified provider increased significantly more in the intervention arm (from 31% to 56%) than in the comparison arm (from 27% to 35%).\nKnowledge\nUnprompted knowledge of maternal and neonatal danger signs increased significantly in both study arms between the baseline and endline, adjusted for improvement in background socioeconomic characteristics (Table 5).\nHowever, the improvements in the intervention arm were significantly larger than those in the comparison arm (Table 6).\nNevertheless, intervention-arm women identified only about three signs among 15 neonatal danger signs at the endline, and recognition improved only in selected individual neonatal signs, including redness around or discharge from the umbilicus, body cold/shivering, skin lesions, and convulsion (results not shown).\nMortality\nNMR estimates did not vary significantly by time or study arm (Table 7).\nNMR was 24.8 (95% CI: 20.7\u201329.4) and 27.9 (95% CI: 23.5\u201332.8) in the comparison arm at baseline and endline, respectively, and was 25.2 (95% CI: 21.0\u201330.1) and 24.0 (95% CI: 19.8\u201329.0) in the intervention arm at baseline and endline, respectively.\nAdjusted mortality hazard ratio in the intervention arm, compared to the comparison arm, was 1.02 (95% CI: 0.80\u20131.30) at baseline and 0.87 (95% CI: 0.68\u20131.12) at endline.\nVerbal autopsy data ascertainment rate did not vary significantly by sex, age at death, or study arm (results not shown).\nCause-specific neonatal mortality rates did not differ by time or study arm.\nThe most common causes of death during the intervention period (2004\u20132005) in the intervention and comparison areas combined were birth asphyxia (109/222, 49%), prematurity (58/222, 26%) and infection (26/222, 12%) (Table 7).\nDiscussion\nThis cluster randomized controlled trial of a package of maternal and newborn healthcare interventions successfully achieved good coverage of antenatal (\u223c90%) and postnatal (\u223c70%) home visits by CHWs, and significantly improved several key newborn care practices and care seeking for newborn complications from qualified providers.\nKnowledge of maternal and newborn danger signs also improved, although to a limited extent.\nHowever, there was no evidence for an impact of the intervention on neonatal mortality.\nThese results are in contrast to several recent trials which decreased neonatal mortality in various settings in South Asia, and also contrasts with a large-scale program evaluation in rural India where lack of mortality impact seemed to stem from inadequate implementation and insufficient coverage of the interventions.\nOur program coverage for both the antenatal and postnatal components was comparable with levels achieved in other effective trials.\nIn addition, we had strict quality assurance of implementation through regular supervision of CHWs and through intensive monitoring of quality of program implementation through household \u201cadequacy\u201d surveys; data from the surveys was used to identify potential areas for improvement in program implementation, and to guide adjustments in intervention delivery to optimize program impact.\nIn Sylhet, Bangladesh, we achieved a 34% reduction in mortality through a similar package of interventions, supervision and monitoring; further analysis of that program revealed that a 64% reduction in mortality was seen among the newborns who were visited within the first two days of life whereas no mortality impact was found among those who were visited only after the two days.\nCoverage of the first visit within the two days, however, was similar in Sylhet (62%) and Mirzapur (69%), and the magnitude of changes in care practices were also similar.\nThus, factors other than reaching families with the intervention must be considered to explain the lack of mortality impact in this study and to guide future strategies to reduce mortality in moderate mortality settings such as Mirzapur.\nLack of evidence of mortality impact can be due to lack of power to test our hypothesis that the intervention would result in a 40% reduction in mortality in the intervention arm \u2013 a level of reduction that had been observed in other efficacy trials and that we thought would be needed to compel policy and program change in Bangladesh.\nGiven the lower number of live births in the study area than anticipated during study design, we did not achieve our enrolment target of 14,872 births.\nWe speculated that a number of factors contributed to this, including declining fertility in rural areas of Bangladesh, an overestimated initial population size, and potential omission of live births in the retrospective pregnancy history.\nIn particular, preliminary results from the Mirzapur Demographic Surveillance Systems since 2007 suggest that the initial population of 292,000 in 2003 was likely overestimated by about 18%, while the annual number of live births during 2004 and 2005 recorded in the retrospective birth history data was about 5% lower than the prospective demographic surveillance results.\nExpanding the study area or extending the intervention period would have been an option to achieve the target enrolment.\nHowever, the catchment area could not be extended in order to ensure access to Kumudini Hospital; and, there was no compelling reason to continue the trial longer than planned, due to the lack of evidence for a downward mortality trend in the intervention arm using program implementation data.\nIn Sylhet, for example, a non-significant downward trend in mortality was observed within 6 month after the intervention started, and a significant program effect on mortality was observed 2 years after the initial intervention introduction.\nIn addition, improvement in care seeking for illness with qualified providers at Kumudini Hospital by families in both study arms, coupled with the provision of quality, life-saving care for any who reached the hospital, likely contributed to the lack of mortality impact of the intervention relative to the comparison area.\nMost importantly, however, our results highlight that local epidemiology, including levels and causes of mortality in the community, must be taken into careful account during intervention design.\nAs NMR decreases, particularly below about 30 per 1000 live births, the cause structure of mortality and, thus, the relative importance of various risk factors for mortality changes.\nIn most other community-based trials, baseline NMR exceeded 45 per 1000 live births, and serious infections, including sepsis, pneumonia, and tetanus, likely accounted for >40% of neonatal deaths.\nAlthough our intervention was designed to address the major causes of mortality in neonates, it was most robust for the prevention and management of infections.\nIn the Mirzapur population, however, nearly 60% of deaths were due to birth asphyxia or prematurity, and the program had limitations in reaching households at the critical times (i.e., during labour, childbirth and immediately after delivery) to address these conditions, and the CHWs lacked the necessary tools and skills to effectively address these conditions.\nCHWs attended <20% of home deliveries, largely due to difficulties in receiving timely notification of labour onset and in travelling to the home to intervene during delivery, given their population catchment area which extended over four villages and, particularly at night, strong discouragement from CHW families for travelling outside the village out of safety concerns.\nTBAs attended most home deliveries (97%) but in spite of brief but focused training in clean delivery, immediate newborn care, and danger sign recognition and referral, they lacked the capabilities to provide skilled care at birth, including resuscitation of birth asphyxiated newborns.\nSome evidence suggests, however, that TBA training in resuscitation is a potentially effective intervention.\nMoreover, recent reviews and meta-analyses suggest that TBAs have some potential for promoting antenatal care, detecting obstetric complications, referring women to skilled obstetric care and positively impacting stillbirths and neonatal outcomes.\nWe found, however, that the numbers and diversity of TBAs in the community made it challenging to train, supervise and manage them to uniform standards of care, and that TBAs and CHWs infrequently encountered a newborn that required bag-and mask resuscitation, which further complicates attempts to train and equip them to provide effective resuscitation in the community.\nCurrent policy in Bangladesh does not promote TBA training programs, however, and implementation of newborn resuscitation outside health facilities is challenging.\nThus, skilled attendance at delivery remains a key policy and program priority for reducing both neonatal and maternal mortality in Bangladesh.\nIn addition to skilled care at delivery, early postnatal care is also critical for reducing mortality in moderate neonatal mortality settings, considering the preponderance of early deaths due to prematurity, birth asphyxia, and, to a lesser extent, vertically acquired sepsis.\nAlthough our overall coverage of postnatal care was good, only 18% and 33% of neonates who died within the first day and the first week of life, respectively, were visited by CHWs prior to the death.\nMoreover, among newborns who were assessed by CHWs and found to be ill, only 54% complied with referral to hospital and compliance with referral was 30% less likely in the first week of life, despite attempts to eliminate major care seeking barriers \u2013 danger sign recognition, access to the hospital and cost.\nThus, emphasis must be placed on community mobilization and empowerment, and on greater understanding of and development of improved approaches to overcome social and financial barriers to referral compliance and care seeking at facilities, especially in the first week of life and in settings where cultural seclusion after birth remains a social norm.\nMoreover, as NMR is reduced below about 30 per 1000, reliance on community-based care is likely to be inadequate to address the needs of extremely preterm infants, who often need additional interventions beyond essential newborn care interventions (e.g., breastfeeding, warmth and hygiene, and emollient therapy), including corticosteroid administration to the mother prior to delivery, surfactant therapy at birth, and assisted ventilation such as continuous positive airway pressure.\nSkilled attendance at facility-based deliveries, along with adaptation of these additional interventions for implementation in first-level facilities in low resource settings, can help to ensure their coverage.\nEmerging evidence suggests that in addition to understanding and overcoming social barriers to care seeking at facilities, programs to address financial barriers may also provide a powerful stimulus to families to access skilled care for delivery and immediate postnatal care at health facilities.\nFinally, for treatment of serious neonatal infections, community-based case management is a viable alternative to facility-based care even where access to quality health care at facilities can be ensured.\nIn Sylhet, Bangladesh, while only 34% of referrals of sick newborns to hospital by CHWs were complied with, another 43% accepted injectable antibiotic treatment at home.\nNeonates in each treatment group had a significantly reduced hazard of mortality, compared to sick neonates who received no treatment or treatment from unqualified providers, indicating that with the addition of home-based treatment, approximately three-fourths of sick neonates received effective curative antibiotic treatment preventing death, a substantial improvement over what was achieved in Mirzapur, where we did not offer home-based treatment with injectable antibiotics.\nIn summary, for optimal survival improvement in low resource populations with moderate NMR, the intervention design must include a clear pathway to survival that links risk factors with causes of mortality, and identifies locally contextualized approaches to risk reduction.\nAs community-based interventions mature and NMR comes down, programs must ensure, in addition to essential newborn care; skilled care during childbirth, including interventions to prevent and manage birth asphyxia and respiratory distress syndrome in preterm infants; and high coverage of curative postnatal care in the first two days of life.\nBarriers to care seeking for illness must also be addressed.\nWhere poor care seeking at referral-level hospitals exists during the early neonatal period, adaptation of interventions for extremely preterm infants for use at community clinic level must be prioritized, and consideration given to inclusion of home-based treatment of serious infections integrated into community case management strategies for childhood infections.\nDistribution of study unions (clusters), Mirzapur sub-district, Tangail district, Bangladesh.Red circle: Union Head Quarter. Star: Kumudini Hospital. Light blue line: River/Beel. Pink shade: Intervention Area, Purple shade: Comparison Area.\nTrial profile for measurement of neonatal mortality.*Participants are women of reproductive age (15\u201349).\n\nAntenatal (birth and newborn care preparedness) and postnatal interventions at home by community health workers.\nPRENATAL: Two home visits scheduled at 12\u201316 weeks and 32\u201334 weeks to:\n1. Promote antenatal care, including:\n(1) Making three antenatal care visits from a health centre or a satellite clinic\n(2) Receiving two doses of tetanus toxoid vaccine\n(3) Procuring adequate iron-folic acid (IFA) supplementation\n(4) Eating extra food\n(5) Care seeking for the following maternal danger signs:\n\u2003- Prolonged labor\n\u2003- Hemorrhage\n\u2003- Fever\n\u2003- Convulsion\n\u2003- Edema of the face, hands or legs, or\n\u2003- Blurred vision\n2. Promote birth planning, including:\n(1) Planning for delivery at a health facility\n(2) If facility is not feasible, choosing a trained birth attendant; preparing the site of delivery in the house; obtaining birth kit or boiling the blade and the pieces of thread; planning for emergency transport; and saving money for emergency\n3. Distribute: clean delivery kit, obtained from Bangladesh Rural Advancement Committee (NGO) free-of-charge, at the second antenatal visit for use by birth attendant\n4. Promote newborn-care preparedness, including:\n(1) Choosing a household member to take care of the newborn right after birth\n(2) Drying and wrapping the baby from head to toe soon after delivery and before the delivery of placenta; using 2 pieces of cloth to wrap the newborn; holding the baby at all times during and immediately after the delivery; avoiding any contact of the newborn with the floor; not keeping the newborn in an unclean or cold place; applying gentle stimulation or refer for resuscitation of the newborn if he/she does not breathe immediately after birth; and practicing wrapping the baby using a doll during CHW visits\n(3) Feeding colostrum to the newborn; initiating breastfeeding immediately after birth; practicing exclusive breastfeeding up to six months; and feeding the newborn frequently in the proper position day and night\n(4) Delaying bathing of the newborn for 72 hours\n(5) Umbilical area care: keeping the cord clean and dry; and avoiding applying anything to the umbilical stump\n(6) Monitoring the baby for signs of infection; and seeking care immediately from CHW or health facility if the newborn has any of the following danger signs:\n\u2003- No cry or breathing at birth,\n\u2003- Convulsions\n\u2003- Unconsciousness\n\u2003- Difficulty breathing\n\u2003- Feeling hot or cold to the touch\n\u2003- Skin pustules or blisters\n\u2003- Umbilical pus or redness\n\u2003- Weak, abnormal or absent cry\n\u2003- Lethargic or less than normal movement\n\u2003- Yellow colour of the body, or\n\u2003- Feeding problem\nPOSTNATAL: Four home visits on postnatal days 0, 2, 5, and 8 to:\n1. Reinforce newborn care messages provided through prenatal visits\n2. Provide counseling for routine breastfeeding and for breastfeeding difficulties\n3. Surveillance of newborn illness: Identify sick neonates based on a clinical algorithm. For identified sick neonates, recommend referral-level evaluation at Kumudini hospital or, if referral fails, continue monitoring according to the clinical algorithm.\n\n\nDefinition of neonatal complications used to measure conditional care seeking: baseline and endline survey.\nBaseline survey* | Endline survey\u2020\n1. Fever | 1. Fever (temp more than 101F)\n2. Trouble breathing | 2. Difficulty in breathing or fast breathing\u2021\n3. Jaundice | 3. Jaundice\n4. Diarrhea | 4. Diarrhea\n5. Umbilical infection or discharge | 5. Pus in the umbilicus or redness of the umbilicus\u2021\n6. Convulsion | 6. Convulsion\n7. Stopped breast feeding | 7. Poor feeding or unable to suck\n8. Body became excessive cold | 8. Hypothermia (temp 95.5\u201397.5 F)\n9. Retention of urine | 9. Doesn't pass urine\n10. Unconsciousness | 10. Unconscious\n\n*Persistent vomiting was included in the baseline survey.\nFollowing additional signs were included in the endline survey: (1) red eye/passage of pus from eyes, (2) skin lesion with infection, (3) baby doesn't cry/breath, (4) chest in drawing, (5) doesn't pass stool, (6) cold/cough, and (7) others.\nListed as 2 separate signs in the survey questionnaire.\n\nIndividual signs included in the prenatal, labor/delivery, and postpartum danger sign knowledge scores.\n | Danger signs\nPrenatal | 1. Severe headache\n | 2. Blurred vision\n | 3. Fetal movement absent\n | 4. High blood pressure\n | 5. Edema of the face/swelling\n | 6. Edema of the hands/leg swelling\n | 7. Convulsions\n | 8. Excessive vaginal bleeding\n | 9. Severe lower abdominal pain\n | 10. Leaking fluid (meconium stained)\nLabor/delivery | 1. Excessive vaginal bleeding\n | 2. Foul-smelling discharge\n | 3. High fever\n | 4. Baby's hand or feet coming out first\n | 5. Baby is in abnormal position\n | 6. Prolong labor (>12 hours)\n | 7. Retained placenta\n | 8. Rupture uterus\n | 9. Cord prolapse\n | 10. Cord around neck\n | 11. Convulsion\nPostpartum | 1. Excessive vaginal bleeding\n | 2. Foul-smelling discharge\n | 3. High fever\n | 4. Inverted nipples\n | 5. Tetanus\n | 6. Retained placenta\n | 7. Severe abdominal pain\n | 8. Convulsions\n | 9. Engorged breasts/swelling of breasts\nNeonatal | 1. Poor feeding or unable to suck\n | 2. Diarrhea\n | 3. Redness around the cord\n | 4. Red eye/discharging eyes\n | 5. Difficult breathing\n | 6. Yellow coloration of the skin/jaundice\n | 7. Hypothermia/shivering\n | 8. Blisters on skin/Skin lesion\n | 9. Baby doesn't cry\n | 10. Fever\n | 11. Unconscious\n | 12. Fast breathing\n | 13. Chest indrawing\n | 14. Doesn't pass urine\n | 15. Doesn't pass stool\n | 16. Convulsions\n\nOne additional prenatal danger signs (high fever) and 1 additional newborn danger sign (cold/cough) were included in the endline survey. We excluded those in creating the knowledge scores in order to maintain comparability across surveys.\n\nChanges in program coverage by community health workers in the intervention arm, among women who had a pregnancy outcome in the 8-month period before each survey.\nService | Adequacy survey 1 |  | Adequacy survey 2 |  | Endline survey | \n | (Apr 2004\u2013Nov 2004) |  | (Dec 2004\u2013Jul 2005) |  | (Aug 2005\u2013Mar 2006) | \nAntenatal BNCP, among all pregnant women | (N\u200a=\u200a565) |  | (N\u200a=\u200a564) |  | (N\u200a=\u200a1096) | \nHome visit at least once | 91.3 | (89.0\u201393.7) | 87.0 | (84.2\u201389.8) | 93.0 | (91.5\u201394.6)\nHome visits twice | 86.9 | (84.1\u201389.7) | 83.8 | (80.8\u201386.9) | 91.0 | (89.3\u201392.7)\nDelivery, among all women who delivered at home | (N\u200a=\u200a447) |  | (N\u200a=\u200a385) |  | (N\u200a=\u200a800) | \nLabor notification to CHWs | 28.8 | (24.6\u201333.0) | 44.4 | (39.4\u201349.3) | 34.8 | (31.5\u201338.1)\nDelivery attendance by CHWs | 12.0 | (9.0\u201315.0) | 20.0 | (16.0\u201324.0) | 13.8 | (11.4\u201316.2)\nPostnatal, among all women who delivered live births at home | (N\u200a=\u200a433) |  | (N\u200a=\u200a379) |  | (N\u200a=\u200a790) | \nHome visits at least once during postnatal day 0\u201327 | 75.5 | (71.4\u201379.5) | 83.7 | (79.9\u201387.4) | 79.7 | (76.9\u201382.5)\nHome visits at least once during postnatal day 0\u20136 | 65.6 | (61.1\u201370.1) | 77.7 | (73.4\u201381.9) | 73.8 | (70.7\u201376.9)\nHome visits at least once during postnatal day 0\u20131 | 53.6 | (48.8\u201358.3) | 73.6 | (69.1\u201378.1) | 69.1 | (65.9\u201372.4)\nTime of first home visit (day)\u2020 | 2.4 | (1.9\u20133.0) | 1.6 | (1.0\u20132.2) | 1.5 | (1.1\u20131.8)\nTotal number of home visits\u2020 | 2.6 | (2.5\u20132.8) | 2.8 | (2.6\u20133.0) | 3.2 | (3.0\u20133.3)\n\nBNCP: Birth and newborn care preparedness; CHWs: Community health workers.\n\u2020Among those who had at least 1 postnatal home visit during postnatal days 0\u201327: n\u200a=\u200a325 (Adequacy survey 1); n\u200a=\u200a318 (Adequacy survey 2); and n\u200a=\u200a628 (Adequacy survey 1).\n\nMaternal demographic and household economic characteristics by study arm and time, among all women had live births between 2001\u20132002 (baseline) and between 2004\u20132005 (endline).\n | Comparison |  |  |  | Intervention |  |  | \n | Baseline |  | Endline |  | Baseline |  | Endline | \n | (n\u200a=\u200a5166) |  | (n\u200a=\u200a5143) |  | (n\u200a=\u200a4822) |  | (n\u200a=\u200a4498) | \n | % | (95% CI) | % | (95% CI) | % | (95% CI) | % | (95% CI)\nMother's age at birth (year) |  |  |  |  |  |  |  | \nmean | 25.2 | (25.0\u201325.3) | 25.0 | (24.8\u201325.1) | 25.4 | (25.2\u201325.5) | 25.3 | (25.1\u201325.4)\n<20 years | 14.4 | (13.4\u201315.4) | 15.6 | (14.6\u201316.6) | 12.7 | (11.7\u201313.6) | 13.1 | (12.2\u201314.2)\n\u226530 years | 20.2 | (19.1\u201321.3) | 17.9 | (16.9\u201319.0) | 20.3 | (19.2\u201321.4) | 18.7 | (17.6\u201319.9)\n\u226535 years | 6.2 | (5.6\u20136.9) | 6.5 | (5.8\u20137.2) | 6.3 | (5.7\u20137.0) | 6.5 | (5.8\u20137.3)\nMaternal education |  |  |  |  |  |  |  | \nEver attended school | 61.7 | (60.4\u201363.1) | 72.4 | (71.1\u201373.6) | 64.7 | (63.3\u201366.1) | 75.6 | (74.4\u201376.9)\nCompleted primary school | 48.9 | (47.5\u201350.3) | 58.1 | (56.7\u201359.4) | 52.0 | (50.6\u201353.5) | 62.5 | (61.1\u201363.9)\nCompleted high school | 14.6 | (13.6\u201315.6) | 20.0 | (18.9\u201321.1) | 16.5 | (15.4\u201317.6) | 21.9 | (20.7\u201323.1)\nHousehold wealth* |  |  |  |  |  |  |  | \nWealth index score | \u22120.19 | (\u22120.2\u20130.1) | 0.44 | (0.4\u20130.5) | \u22120.36 | (\u22120.4\u20130.3) | 0.44 | (0.4\u20130.5)\nPoor | 22.7 | (21.6\u201323.9) | 15.0 | (14.0\u201316.0) | 25.7 | (24.5\u201327.0) | 14.7 | (13.7\u201315.8)\nRich | 17.6 | (16.6\u201318.7) | 26.2 | (25.0\u201327.4) | 15.7 | (14.6\u201316.7) | 26.5 | (25.2\u201327.8)\n\n*Wealth index created based on pooled baseline and endline surveys, using principal component analysis of durable goods, electricity, toilet facility, sources of drinking water, and housing materials. Poor refers to the lowest quintile of the wealth index and rich is the highest quintile of the wealth index.\n\nAdjusted predicted mean of knowledge and practice indicators by study arm and time, among women who had a pregnancy outcome in the 1-year period before each survey.*\n\n | Comparison |  | Intervention |  | \n | Baseline | Endline | Baseline | Endline | \n | (May 2002\u2013 Jul 2003) | (Feb 2005\u2013 Apr 2006) | (May 2002\u2013 Jul 2003) | (Feb 2005\u2013Apr 2006) | \nPRACTICE (percent of target population practicing a behavior) |  |  |  |  | \nPrenatal care and birth preparedness among all women |  |  |  |  | \nNumber of women | 2644 | 1759 | 2371 | 1732 | \nHad \u22651 antenatal care visits from a qualified provider\u2020 | 47.8 | 49.1 | 47.4 | 68.8 | \u2225\nReceived \u22651 tetanus immunization | 76.4 | 73.8 | 77.1 | 77.5 | \nReceived \u22652 tetanus immunizations | 56.9 | 41.0 | 54.7 | 39.8 | \nReceived iron supplementation | 45.8 | 42.7 | 47.9 | 55.7 | \u2225\nFacility-based delivery |  |  |  |  | \nNumber of women | 2644 | 1759 | 2371 | 1732 | \nDelivered at medical facilities | 12.5 | 16.5 | 12.1 | 20.2 | \u2225\nImmediate newborn care among all home-born live births |  |  |  |  | \nNumber of home-born live births | 2006 | 1322 | 1805 | 1231 | \nSterile cord cut\u2021 | 59.2 | 66.9 | 63.3 | 95.1 | \u2225\nNot applying anything on the newly-cut cord | 95.1 | 86.0 | 94.8 | 94.3 | \u2225\nDrying/wiping the baby before delivery of placenta | 2.1 | 3.0 | 2.2 | 14.4 | \u2225\nWrapping the baby before delivery of placenta | 2.4 | 2.7 | 2.9 | 13.5 | \u2225\nDelaying bath to the 3rd day or later | 1.5 | 13.4 | 1.6 | 77.8 | \u2225\nBreastfeeding initiation within 1 hour after birth | 41.2 | 55.0 | 40.9 | 80.0 | \u2225\nBreastfeeding prior to any food/liquid | 28.9 | 50.5 | 29.3 | 87.3 | \u2225\nCare seeking among neonates with complications |  |  |  |  | \nNumber of neonates with 1 or more of the 10 complications\u00a7 | 812 | 400 | 733 | 355 | \nReceived any treatment | 93.7 | 95.9 | 92.9 | 97.3 | \nReceived treatment from a qualified provider\u2020 | 27.4 | 34.6 | 30.7 | 55.7 | \u2225\nKNOWLEDGE (mean danger sign knowledge scores) [range] |  |  |  |  | \nNumber of women | 2644 | 1759 | 2371 | 1732 | \nMaternal danger sign knowledge score: antenatal [0\u201310] | 1.0 | 2.2 | 1.1 | 2.9 | \u2225\nMaternal danger sign knowledge score: labor/delivery [0\u201311] | 1.1 | 1.9 | 1.2 | 2.4 | \u2225\nMaternal danger sign knowledge score: postpartum [0\u20139] | 1.0 | 2.0 | 1.0 | 2.5 | \u2225\nNeonatal danger sign knowledge score [0\u201315] | 2.3 | 2.4 | 2.3 | 2.8 | \u2225\n\n*Adjusted for mother's age at birth (<20 years, 20\u201329 years (reference group), or \u226530 years), maternal educational attainment (<primary school completion vs. \u2265 primary school completion), and the household wealth (the lowest quintile, middle 3 quintiles (reference), or the highest quintile of the wealth index), using linear probability regression models.\nIncludes check-up/treatment by a qualified provider (e.g., doctor, nurse, family welfare visitor, or medical assistant) or at a medical facility (e.g., satellite clinic, Family Welfare Centre, Upazila health complex, qualified doctor's chamber, clinic, or hospital).\nBlade from a clean delivery kit or other instrument that was boiled before use.\nSee Table 2 for the list of complications.\nSignificant differential change over time by study arm (i.e., significant interaction) (p-value <0.05).\n\nLevels and causes of neonatal mortality by time and study arm.\n |  | Baseline |  |  |  | Endline |  | \n |  | Comparison |  | Intervention |  | Comparison |  | Intervention\nLEVELS |  |  |  |  |  |  |  | \nPeriod (calendar years) |  | 2001\u20132002 |  | 2001\u20132002 |  | 2004\u20132005 |  | 2004\u20132005\nNumber of live births | 5292 |  | 4951 |  | 5241 |  | 4616 | \nNumber of deaths | 131 |  | 125 |  | 146 |  | 111 | \nNeonatal mortality rate (per 1000 live births) [95% CI] | 24.8 | [20.7\u201329.4] | 25.2 | [21.0\u201330.1] | 27.9 | [23.5\u201332.8] | 24.0 | [19.8\u201329.0]\nAdjusted hazard ratio* [95% CI] | 1 | (reference) | 1.02 | [0.80\u20131.30] | 1 | (reference) | 0.87 | [0.68\u20131.12]\nCAUSES, based on a hierarchical algorithm\u2020 |  |  |  |  |  |  |  | \nPeriod (calendar years) |  | 2002 |  | 2002 |  | 2004\u20132005 |  | 2004\u20132005\nNumber of live births | 2636 |  | 2463 |  | 5241 |  | 4616 | \nNumber of deaths | 72 |  | 51 |  | 146 |  | 111 | \nNumber of deaths with complete verbal autopsy data (% of total neonatal deaths) | 67 | (93.1) | 42 | (82.4) | 129 | (88.4) | 93 | (83.8)\nNumber of deaths by cause (% of neonatal deaths with complete verbal autopsy data [95% CI]) |  |  |  |  |  |  |  | \nCongenital malformations | 6 | (9.0 [3.4\u201318.5]) | 4 | (9.5 [2.7\u201322.6]) | 10 | (7.8 [3.8\u201313.8]) | 0 | (0.0 [0.0\u20133.9]) \u2021\nTetanus | 3 | (4.5 [0.9\u201312.5]) | 2 | (4.8 [0.6\u201316.2]) | 3 | (2.3 [0.5\u20136.6]) | 1 | (1.1 [0.0\u20135.8])\nPreterm | 16 | (23.9 [14.3\u201335.9]) | 15 | (35.7 [21.6\u201352.0]) | 33 | (25.6 [18.3\u201334.0]) | 25 | (26.9 [18.2\u201337.1])\nBirth asphyxia | 20 | (29.9 [19.3\u201342.3]) | 11 | (26.2 [13.9\u201342.0]) | 63 | (48.8 [39.9\u201357.8]) | 46 | (49.5 [38.9\u201360.0])\nBirth injury | 0 | (0.0 [0.0\u20135.4]) \u2021 | 0 | (0.0 [0.0\u20138.4]) \u2021 | 0 | (0.0 [0.0\u20132.8]) \u2021 | 3 | (3.2 [0.7\u20139.1])\nInfection | 11 | (16.4 [8.5\u201327.5]) | 8 | (19.0 [8.6\u201334.1]) | 15 | (11.6 [6.7\u201318.5]) | 11 | (11.8 [6.1\u201320.2])\nDiarrhea | 0 | (0.0 [0.0\u20135.4]) \u2021 | 0 | (0.0 [0.0\u20138.4]) \u2021 | 1 | (0.8 [0.0\u20134.2]) | 1 | (1.1 [0.0\u20135.8])\nCause not assigned | 11 | (16.4 [8.5\u201327.5]) | 2 | (4.8 [0.6\u201316.2]) | 4 | (3.1 [0.9\u20137.7]) | 6 | (6.5 [2.4\u201313.5])\nCause-specific neonatal mortality rate (per 1000 live births) [95% CI] |  |  |  |  |  |  |  | \nCongenital malformations | 2.3 | [0.8\u20135.0] | 1.6 | [0.4\u20134.2] | 1.9 | [0.9\u20133.5] | 0.0 | [0.0\u20130.8] \u2021\nTetanus | 1.1 | [0.2\u20133.3] | 0.8 | [0.1\u20132.9] | 0.6 | [0.1\u20131.7] | 0.2 | [0.0\u20131.2]\nPreterm | 6.1 | [3.5\u20139.9] | 6.1 | [3.4\u201310.0] | 6.3 | [4.3\u20138.8] | 5.4 | [3.5\u20138.0]\nBirth asphyxia | 7.6 | [4.6\u201311.7] | 4.5 | [2.2\u20138.0] | 12.0 | [9.2\u201315.4] | 10.0 | [7.3\u201313.3]\nBirth injury | 0.0 | [0.0\u20131.4] \u2021 | 0.0 | [0.0\u20131.5] \u2021 | 0.0 | [0.0\u20130.7] \u2021 | 0.6 | [0.1\u20131.9]\nInfection | 4.2 | [2.1\u20137.5] | 3.2 | [1.4\u20136.4] | 2.9 | [1.6\u20134.7] | 2.4 | [1.2\u20134.3]\nDiarrhea | 0.0 | [0.0\u20131.4] \u2021 | 0.0 | [0.0\u20131.5] \u2021 | 0.2 | [0.0\u20131.1] | 0.2 | [0.0\u20131.2]\nCause not assigned | 4.2 | [2.1\u20137.5] | 0.8 | [0.1\u20132.9] | 0.8 | [0.2\u20132.0] | 1.3 | [0.5\u20132.8]\n\n*Analysis of baseline and endline combined, stratified by study arm (n\u200a=\u200a10553, intervention arm, and n\u200a=\u200a9567, intervention arm). Mortality risk at baseline (2001\u20132002) was the reference group in each arm. Estimates were controlled for: sex, mother's age at birth [<20 years, 20\u201329 years [reference group]), or \u226530 years]), maternal educational attainment [<primary school completion vs. \u2265 primary school completion]), and the household wealth [the lowest quintile, middle 3 quintiles [reference]), or the highest quintile of the wealth index]), using a survival-time model with a Weibull survival distribution.\nCauses are presented in the order of the hierarchy. Case definitions for the seven causes are presented elsewhere in detail.\n\nOne-sided 97.5% confidence interval.",
        "label": "unclear",
        "id": "task4_RLD_test_465"
    },
    {
        "paper_doi": "10.2471/blt.14.142489",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: cRCTUnit of randomization: villageNumber of clusters: 79Data collection: baseline, every month for 9 months and endline surveys with 25 randomly-selected households in each target village. At endline, test for malaria by RDT if consenting.Length of follow-up: 13 monthsAdjustment for clustering: yes\n\n\nParticipants: Target treatment group: not described; all agesSample size: 79 villagesExclusion criteria: any person who tested positive who had taken an antimalarial in previous 4 weeks was referred for microscopy\n\n\nInterventions: Staff who received training: drug shop vendorsDuration of training: 2 daysContent of training: using mRDTs, and recommending purchase of first-line ACT (no algorithms)Supervision: monthly visits to track stock and usage of mRDTs and compliance with protocols for testing using a 17-point checklist. Unused kits sent for lot testing every 3 monthsAntimalarials free to participants: no, but subsidizedmRDTs free to participants: no\n\n\nOutcomes: Number receiving an antimalarial\n\n\nNotes: Control: chemical sellers dispensing medicines without test results (community-based treatment of suspected malaria by clinical diagnosis)Country: UgandaSetting: unclearMalaria endemicity: annual transmission rates > 100 bites/personStudy dates: March 2011-April 2012Study sponsor: the Clinton Health Access Initiative and the Bill & Melinda Gates Foundatio\n\n",
        "objective": "To evaluate community\u2010based management strategies for treating malaria or fever that incorporate both a definitive diagnosis with an mRDT and appropriate antimalarial treatment.",
        "full_paper": "Introducing rapid diagnostic tests for malaria to drug shops in Uganda: a cluster-randomized controlled trial\nIntroducing rapid diagnostic tests for malaria to drug shops in Uganda: a cluster-randomized controlled trial\nBulletin of the World Health Organization\nBull. World Health Organ.\nIntroducing rapid diagnostic tests for malaria to drug shops in Uganda: a cluster-randomized controlled trial.\"\nBulletin of the World Health Organization 93 (3): 142-151.\nIntroduction\nIn areas where malaria is endemic, the appropriate management of febrile illness and the effective use of resources for malaria control rely on the availability and use of diagnostic tests. 1\nIn the absence of diagnostic tests, antimalarial drugs are often taken for illnesses that have similar symptoms to those of malaria. [2][3][4][5]\nFailure to diagnose malaria can lead to poor case management, a waste of scarce health resources and increased risk of antimalarial resistance. 1,6\nThe non-treatment or delayed treatment of malaria contribute substantially to malaria-attributable child mortality.\n7,8 In Uganda, only a minority of febrile illnesses are treated with artemisinin combination therapy -i.e. the recommended first-line treatment for malaria -and many of such episodes go untreated.\n9 Similar observations have been made in Kenya, the United Republic of Tanzania and other African countries. [10][11][12][13]\nThe World Health Organization (WHO) recommends parasitological confirmation of malaria before antimalarial drug use.\n14 Although the current Global malaria action plan of the Roll Back Malaria Initiative calls for universal access to malaria testing, 15 such access remains a distant goal in most countries with endemic malaria.\nA study in six African countries found that only 4-31% of children with febrile illnesses were tested for malaria. 9\nIn many countries, patients and caregivers rely heavily on a loosely regulated private sector for malaria treatment. 16,17\nIn consequence, the engagement of the private sector has become an increasingly common strategy in malaria control programmes -as reflected, for example, in the pilot Affordable Medicines Facility-malaria (AMFm). 18\nThe development of inexpensive and simple rapid diagnostic tests for malaria has opened the possibility of widespread access to malaria diagnosis.\nThese antigen detection tests have been shown to be as effective as routine microscopy in malaria diagnosis 19 and can be safely performed by individuals with only basic training.\n20 Although research from Cambodia, 21 Somalia 22 and Uganda 23 has shown that the distribution of rapid diagnostic tests by the private sector is feasible, we know very little of the impact of this approach on population-level rates of malaria diagnosis and purchase of antimalarial drugs.\nWe therefore conducted a trial in eastern Uganda to investigate the impact -on malaria diagnosis and the purchase of antimalarial drugs -of training the vendors from licensed drug shops to test patients with a rapid diagnostic test for malaria.\nThe trained vendors were also encouraged to buy the test, at a subsidized price, from local wholesale providers.\nThe study took place in Uganda's eastern region, where the annual transmission rates for malaria exceed 100 infective bites per person 24 and presumptive symptom-based treatment remains commonespecially when, as commonly occurs, treatment is sought outside the higher level public-health facilities. [25][26][27]\nMalaria is responsible for 30-50% of outpatient visits and 9-14% of inpatient deaths in Uganda. 28\nMethods\nThe study was designed as a clusterrandomized controlled trial, with extensive monitoring of the health behaviour of households before and after a rapid diagnostic test was made available in local drug shops.\nSince the study was designed to explore both cross-sectional and pre-post differences, 67 (85%) of the study villages were randomly selected to receive the intervention while the remaining 12 (15%) were used as a control group.\nVillages\nThe sampling frame for the study included all of the villages in the districts of Budaka, Bukedea, Kibuku, Kumi, Ngora and Pallisa that had at least one drug shop licensed and registered with the ministry of health.\nAlthough the district health office initially identified 92 such villages, only 79 still had at least one licensed drug shop at the time the main study was launched.\nThese 79 villages were included in the cluster randomization (Fig. 1).\nHouseholds\nThe households in all 92 target villages were listed before the launch of the study.\nSubsequently, 25 households in each target village were randomly selected, visited for a baseline survey to record basic demographic characteristics and health behaviours, and re-visited every month for 9 months (Fig. 1) to monitor their health problems and treatment seeking.\nDuring the tenth and final follow-up survey -i.e. the so-called endline survey -all consenting adults and -if their caregivers gave consent -the children in the study households were tested for malaria using the rapid diagnostic test.\nDrug shops\nEach licensed drug shop listed by the ministry of health in the 79 villages and included in the randomization was visited for a baseline survey in April 2011 and an endline survey in March 2012.\nIntervention\nVendors from 108 drug shops in the 67 intervention villages were invited for 2 days of training in the use of a rapid diagnostic test for malaria -i.e. the CareStart Malaria HRP2 (Pf) test (Access Bio, Somerset, USA).\nThe training, described previously, 23 was facilitated by trainers approved by the national ministry of health.\nThese trainers reviewed the signs and symptoms of uncomplicated malaria and severe illness and instructed the trainees on how to perform the diagnostic test and on the recommended first-line treatment for malaria.\nThe first day of training was classroom-based while the second day involved practical experience in a health facility.\nTrainees were not given specific instructions on when to recommend testing and -apart from the promotion of artemisinin combination therapy -were not given treatment algorithms.\nNo retail price for the test was suggested.\nOur main objective was to observe whether and how the trainees chose to integrate the test into their normal practice.\nUpon successful completion of the training, each trainee was given a free box of 40 test kits, gloves, an instruction leaflet, results slips and a sharps disposal box.\nAny new staff employed by the drug shops in the intervention villages after the initial training were invited to a similar training course that was run in October 2011.\nTrained drug shop vendors were visited monthly to track their stocking and usage of the rapid diagnostic test and compliance with the recommended protocols for testing.\nTest kit storage, administration and post-use disposal were monitored using a 17-point checklist.\nInformation on the shop's stock of the test kits and price of the test for patients was recorded.\nAt the same time, any questions the vendors had about the test's administration were answered.\nEvery 3 months, four unused test kits were collected from each shop holding the kits and sent for lot testing at the Foundation for Innovative Diagnostics Laboratory at the Pasteur Institute of Cambodia, in Phnom Penh.\nEvery kit investigated in this way passed lot testing.\nTest kits\nAccording to WHO, the test investigated has a panel detection score of 98.7%, a false-negative rate of < 1% and a total false-positive rate of 2.4%.\n29 Because this test, like other tests based on the detection of histidine-rich protein II, can remain positive for up to 5 weeks after a cured infection, a higher false-positive rate may occur in settings where malaria  is highly endemic.\n30 Trainees were told that any patient who tested positive who had also taken antimalarial drugs in the previous 4 weeks should be referred to a facility where they could be checked for malarial infection by microscopy.\nThe test kits were purchased and imported, at a cost of 0.70 United States dollar (US$) per kit, by the study team.\nThey were then sold to a prominent pharmaceutical wholesaler in Kampala at a subsidized price of US$ 0.12 per kit.\nThe wholesaler's regional pharmacy in the city of Mbale subsequently sold the kits -exclusively to our trainees -at a price of US$ 0.19 per kit. 23\nData entry and analysis\nData were entered using the CSPro 4.0 package (United States Census Bureau, Suitland, USA) and mainly analysed, using Stata version 11.0, in a multivariate linear probability difference-in-differences model.\nA P-value of 0.05 or less was considered statistically significant.\nThe primary outcome was whether a household member with febrile illness was tested for malaria.\nSecondary outcomes included the medication taken to treat febrile illness -if any -and where treatment -if any -was sought.\nThe main independent variable was whether the illness occurred in a village in our intervention or control arm.\nAs vendors from 15% of the drug shops targeted for training did not complete their training, intention-to-treat effects were estimated.\nTo control for seasonal effects and for village characteristics, a full set of village and monthly fixed effects were included in the fully adjusted model.\nEstimated robust standard errors were clustered at the village level.\n31,32 Although adjustments were also made to remove the potential effects of a behaviour change campaign that was rolled out in the later stages of our intervention, the roll-out of the campaign was orthogonal to the main treatment and did not affect our main results.\nOur study was powered to detect an increase in the fraction of fever cases seeking health care at drug shops using the rapid diagnostic kit from 10% to 20%.\nAssuming an incidence of one episode of febrile illness per person-year, a mean household size of five individuals, 40% of households seeking treatment at a drug shop and an intra-class correlation of 0.05, the study was powered to detect the targeted 10% difference in the 9-month intervention period with a probability of 0.92.\nEthics\nEthical approval for this study was given by the Harvard School of Public Health (protocol # P19371-106) and the Uganda National Council for Science and Technology (protocol # HS805).\nTrial registry\nThe trial was registered as clinical trial NCT01652365 at clinicaltrials.gov.\nResults\nDrug shops\nOut of the 108 registered drug shops, 92 shops completed training.\nFourteen vendors declined to be trained, one shop was relocated outside our study area and the vendor from another shop was considered to have failed the training.\nEach of the 92 shops had at least one staff member who successfully completed training (Fig. 1).\nOver the monitoring period from July 2011 to March 2012, shops run by successful trainees bought a mean of 146 (median: 40) diagnostic kits from the local wholesaler (Table 1).\nOverall, 13 440 test kits were bought and the shops investigated 10 412 patients using the test kits that they had been given or bought.\nHowever, 37 such shops (40%) did not purchase any of the test kits and most of the others bought only small numbers of the kits (Table 2).\nTogether, just three shops accounted for 32% (3346) of all of the rapid diagnostic tests performed on patients.\nThe mean price paid by a patient assessed with the rapid diagnostic kit was US$ 0.43 -representing a 125% markup on the kit's wholesale price.\nAccording to the data collected in our household surveys, the median prices paid for artemisinin combination therapy for individuals younger and older than 5 years were US$ 0.57 and US$ 0.76, respectively.\nThe corresponding costs of quinine -which accounted for 72% of purchases of antimalarial drugs other than artemisinin combination therapies -were US$ 0.57 and US$ 0.38, respectively.\n12 Households Fig. 2 illustrates the households investigated and illnesses captured over the study period.\nIn total, 25 758 episodes of illness were reported by study households across the 10 survey rounds.\nRespondents reported the presence of fever in 22 697 (88.1%) of these episodes and we focused on these episodes of febrile illness in our analysis.\nFor 4364 of the reported episodes of illness -including 3908 episodes of febrile illness -treatment was sought at one of the study drug shops.\nThe study households in the intervention and control villages appeared similar in terms of their baseline demographic characteristics (Table 3).\nThe estimated prevalence of malaria at the time of the endline survey was 43% in  both arms of the study.\nTable 4 presents the characteristics of the episodes of febrile illness recorded pre-intervention, in monthly morbidity surveys.\nMost individuals (4448) with febrile illness sought treatment in the public sector, at a private-sector drug shop or pharmacy, or at a private-sector clinic or hospital.\nHowever, 21.5% (1218/5666) of such individuals sought no care.\nRoughly half (1039/2322) of the reported visits to drug shops for the treatment of febrile illness were to a shop that was involved in our study.\nDuring the pre-intervention period, malaria testing and antimalarial drug use among cases of febrile illness were significantly less likely in the intervention villages than in the control villages -25.9% (1258/4861) versus 38% (306/805; P = 0.002) and 54.3% (2641/4861) versus 68.2% (549/805; P = 0.003), respectively.\nThis difference appears to be a reflection of a relatively larger fraction of patients from the control villages seeking treatment at a public facility.\nIn the pre-intervention period, the percentage of febrile patients seeking care in the private sector who were tested for malaria was similar across the two study arms (40% [54/135] versus 33.6% [235/700]; P = 0.565).\nTable 5 shows the population-level crude estimates and the corresponding -and, generally very similar -adjusted estimates of the intervention's impact on testing, medication choice and treatment seeking.\nAmong all cases of febrile illness and among cases of febrile illness in children younger than 5 years, according to the crude model, the in-    tervention significantly increased the probabilities of being tested for malaria, by 6.0 (P = 0.015) and 7.6 percentage points (P = 0.015), respectively.\nAccording to the same model, the intervention increased the likelihood of taking any antimalarial drug or artemisinin combination therapy by 4.8 and 1.6 percentage points and reduced antibiotic usage by 0.3 of a percentage point -but none of these differences reached statistical significance.\nAppendix A (available at: https://cdn1.sph.harvard.edu/wpcontent/uploads/sites/358/2012/08/\nAppendixA.pdf ) demonstrates the robustness of these main results, which remained similar if we (i) used patientreported episodes of suspected malaria rather than all fever episodes, (ii) confined our analysis to the episodes of illnesses that occurred before the roll-out of the behaviour change campaign, or (iii) used logistic regression instead of linear probability models.\nFig. 3 shows the fractions of patients visiting the drug shops in intervention and control villages who were tested for malaria before and after roll-out of the intervention.\nPrior to the first training course about the rapid diagnostic test, the fraction of patients tested for malaria in a study drug shop was similar in the intervention and control villages, 8.9% (70/786) and 10.6% (12/113), respectively.\nAfter the first training course, the fraction of patients tested for malaria in a study drug shop in the control arm remained almost unchanged (33/334) but the corresponding value in the intervention arm almost doubled (390/3011; P < 0.001).\nNearly 90% (3112/3458) of patients investigated using the rapid diagnostic test gave a positive result.\nThe reliability of the test results is discussed in Appendix A.\nFig. 4 summarizes our data on the patients who, during the intervention period, visited and were tested at the study drug shops in the intervention villages.\nAlthough over 80% (285/342) of such patients who tested positive for malaria purchased an antimalarial drug of some kind, nearly 45% (21/48) of those who tested negative did the same.\nTest-positive patients were more than twice as likely to purchase artemisinin combination therapy as test-negative patients (40.9% [140/342] versus 16.7% [8/48]; P < 0.001).\nHowever, less than half of the test-positive patients -and 31.5% (590/1871) of the untested pa-\nAny antimalarial drug\nMedication purchased\nACT: artemisinin combination therapy.\nNotes: Medication is shown separately for the patients who were tested for malaria with a rapid diagnostic kit and found positive, the patients who were similarly tested and found negative and the patients who were not tested.\nResearch\nMalaria tests in drug shops, Uganda Jessica Cohen et al.\ntients -purchased artemisinin combination therapy.\nWe found no differences in antibiotic purchases according to testing status or test result.\nDiscussion\nIn countries where malaria is highly endemic, many episodes of febrile illnesses are treated at drug shops.\nOur results indicate that, by offering training and access to subsidized rapid diagnostic tests to private drug shops, it is possible to increase malaria testing rates significantly.\nThree of the most commonly voiced concerns regarding use of rapid tests for malaria diagnosis by the private sector are poor adherence to protocols, the potential crowding out of public-sector treatment and increased antibiotic purchases by patients who have a negative result.\nWe found no evidence to support any of these concerns.\nIn addition, as we previously reported, our monthly monitoring of the drug shops indicated generally high levels of compliance with recommended treatment, storage, waste management and test-kit administration protocols.\n23 From a policy perspective, a major challenge for any health initiative in the private sector is the achievement of adequate uptake and coverage.\nIn our study setting, the potential impact of the intervention was limited by the need to work only with drug shops that were licensed by the Ugandan Ministry of Health.\nInclusion of unlicensed outlets would have substantially increased the reach of our programme but may also have complicated monitoring and quality control.\nEven among the licensed shops with trained staff, however, uptake was limited.\nOnly about 60% (55/92) of such shops chose to stock the test kits and fewer than 20% (390/2261) of the febrile patients who visited a drug shop that had a trainee were actually tested for malaria.\nWhile no detailed information on the reasons for the relatively weak uptake of the rapid tests was collected, higher rates of uptake could probably be achieved by combining behavioural change efforts with stronger financial incentives.\nThe price subsidies for the test kits could be increased, drug shops could be paid incentives to perform tests and provide appropriate treatment to the test-positive patients, and the level of any subsidy could be correlated with the quality of the testing provided by each shop.\nGreater impact could potentially also be achieved if the test-related training and subsidy could be combined with training on the appropriate treatment of non-malarial illnesses -although this would require policy-makers to support the movement of non-medical professionals further into formal case management.\nA second main challenge, from a policy perspective, is the remarkably low uptake of artemisinin combination therapies that we observed.\nOnly 40.9% of the patients who tested positive for malaria actually purchased such therapies.\nThis modest amount of uptake does not appear to have been driven by lack of access, given that 84.2% (96/114) of our study drug shops reported having artemisinin combination therapies in stock at the time of our endline survey. 12\nIt is possible that patients perceived the prices of such therapies to be too high -compared with those of other antimalarial treatments -even though, in Uganda at the time of our study, the prices of such therapies had been substantially lowered by subsidies from the Affordable Medicines Facility-malaria programme.\n12,18 Furthermore, use of rapid tests for malaria diagnosis in the private sector appears to be a feasible and potentially effective way to increase testing rates and improve overall case management.\nAs discussed in greater detail elsewhere, 33 the subsidizing of such tests for use in the private sector is likely to yield highest returns in settings where malaria prevalence is low and treatment seeking in the private sector is common.\nThe accurate diagnosis of malaria could eliminate the wasteful use of antimalarial drugs for non-malarial illness and improve the management of malaria and other febrile illnesses -particularly if the private sector is properly incentivized and equipped to treat the true causes of non-malarial illness appropriately. \u25a0\nResearch\nResumen\nFig. 1. Study design: introduction and use of rapid diagnostic tests for malaria in drug shops, Uganda 2011-2012\nFig. 2. Monitored households and observed morbidity, Uganda, March 2011-April 2012\nFig. 3. Testing rates for malaria and results for patients visiting study drug shops, Uganda, March 2011-April 2012\n1. Study design: introduction and use of rapid diagnostic tests for malaria in drug shops, Uganda 2011-2012 \nResearch |  | \nMalaria tests in drug shops, Uganda |  | Jessica Cohen et al.\n92 villages |  | \nPre-intervention household |  | \nsurvey and drug shop census |  | \n | No eligible shop\n | (13 villages) | \n7 follow-up surveys plus |  | \nendline survey |  | \nRandomization |  | \n | Vendors invited to training (108 shops in 67 villages) | Vendors not interested (14 shops)\n |  | Moved\n |  | (1 shop)\n |  | Vendor failed training\nControl group | Vendors trained successfully | (1 shop)\n(23 shops in 12 villages) | (92 shops in 59 villages) | \n7 follow-up surveys plus |  | \nendline survey |  | \nTable 1 . Drug shop purchases and use of rapid diagnostic tests for malaria, Uganda, July 2011-March 2012 At the time of the study, 10 000 Ugandan shillings were equivalent to 3.82 United States dollars.\nRapid diagnostic test | Mean | Median | SD | Min Max\nNo. purchased | 146.09 | 40.00 | 261.71 | 0 1320\nNo. performed | 113.17 | 10.00 | 234.92 | 0 1220\nRetail price (Ugandan shillings) a,b | 1125.00 | 1000.00 | 293.52 | 500 2000\nMax: maximum; Min: minimum; SD: standard deviation. |  |  | \na b Values are for 87 of the 92 shops.\nTable 2 . Frequency distributions for drug shop purchases and use of rapid diagnostic tests for malaria, Uganda, July 2011-March 2012 \nNo. of | Purchase of tests | Performance of tests\ntests/shop | No. of shops | % of purchases | No. of shops | % of tests performed\n0 | 37 | 0.0 | 42 | 0.0\n1-100 | 24 | 10.1 | 26 | 8.4\n101-500 | 24 | 41.5 | 18 | 41.7\n501-1000 | 4 | 21.9 | 3 | 17.8\n> 1000 | 3 | 26.5 | 3 | 32.1\nTable 3 . Baseline characteristics of the surveyed households, Uganda, March-April 2011 \nCharacteristic | Households in | Households in | Difference, | P b\n | control villages | intervention villages | percentage | \n | (n = 326) | (n = 1867) | point a | \nMean no. of individuals in household (SD) |  |  |  | \nAll ages | 6.62 (3.32) | 6.34 (3.36) | -0.28 | 0.312\nAged < 5 years | 1.16 (1.07) | 1.06 (1.05) | -0.01 | 0.193\nFraction of household members sleeping under bednets, % | 58.7 | 60.6 | 1.9 | 0.579\nNo. of households own land (%) | 237 (72.7) | 1385 (74.2) | 1.5 | 0.864\nNo. of households treat drinking water (%) | 70 (21.5) | 452 (24.2) | 2.7 | 0.478\nNo. of heads read English (%) | 112 (34.4) | 564 (30.2) | -4.1 | 0.476\nNo. of households have at least one mobile phone (%) | 207 (63.5) | 1099 (58.9) | -4.6 | 0.396\nNo. of households have at least one bicycle (%) | 182 (55.8) | 1088 (58.2) | 2.4 | 0.534\nNo. of households have electricity (%) | 26 (8.0) | 185 (9.9) | 1.9 | 0.595\nSD: standard deviation. a\nFor mean number of individuals, the values represent numbers.\nb Adjusted for clustering at village level.\nResearch Malaria tests in drug shops, Uganda Jessica Cohen et al.\nTable 4 . Pre-intervention treatment-seeking for febrile illness, Uganda, April-June 2011 Characteristic No. of episodes of febrile illness in households (%) Difference, percentage point P a \nIn control villages | In intervention villages\n(n = 805) | (n = 4861)\na Adjusted for clustering at village level.\nb Denominator represents the number of people that visited the facility.\nc Representing 52.6% of those who took any antimalarial drug.\nd Representing 53.7% of those who took any antimalarial drug.\nTable 5 . Impact of intervention on treatment seeking, malaria testing and drug use in response to episodes of febrile illness, Uganda, July 2011-March 2012 Variable No. of episodes included in model Results from unadjusted model Results from adjusted model a \n\u03b2 | 95% CI | % | \u03b2 | 95% CI | %\n |  | change b |  |  | change b\nCI: confidence interval.\na Adjusted for month and village fixed effects.\nb Percentage change in outcome: (beta/pre-intervention mean in intervention areas) x 100.\ndiagnostiques rapides du paludisme dans les pharmacies en Ouganda: un essai contr\u00f4l\u00e9 randomis\u00e9 par grappe Objectif \u00c9valuer l'impact (sur le diagnostic et le traitement du paludisme) de l'introduction de tests diagnostiques rapides dans les pharmacies dans l' est de l'Ouganda. M\u00e9thodes Au total, 2 193 m\u00e9nages vivant dans 79 villages de l' \u00e9tude disposant d'au moins 1 pharmacie homologu\u00e9e ont \u00e9t\u00e9 recrut\u00e9s et suivis pendant 12 mois. Apr\u00e8s 3 mois de suivi, les pharmaciens de 67 villages, choisis al\u00e9atoirement pour l'intervention, se sont vus proposer une formation leur permettant d'utiliser des tests diagnostiques rapides du paludisme et, une fois form\u00e9s, un acc\u00e8s \u00e0 ces tests \u00e0 un prix subventionn\u00e9. Les 12 villages restants de l' \u00e9tude ont servi de villages t\u00e9moins. Un mod\u00e8le de r\u00e9gression des doubles diff\u00e9rences a \u00e9t\u00e9 utilis\u00e9 pour estimer l'impact de l'intervention. R\u00e9sultats Les pharmaciens de 92 pharmacies ont termin\u00e9 leur formation avec succ\u00e8s et 50 d' entre eux ont stock\u00e9 et r\u00e9alis\u00e9 activement les tests rapides. En 9 mois, les pharmaciens form\u00e9s ont r\u00e9alis\u00e9 une moyenne de 146 tests par officine. Les m\u00e9nages ont signal\u00e9 22 697 \u00e9pisodes de maladie f\u00e9brile. La disponibilit\u00e9 des tests rapides dans les pharmacies locales a augment\u00e9 significativement la probabilit\u00e9 qu'une maladie f\u00e9brile soit test\u00e9e pour le paludisme de 23,15% (P = 0,015) et trait\u00e9e par un m\u00e9dicament antipalud\u00e9en de 8,84% (P = 0,056). La probabilit\u00e9 que la polyth\u00e9rapie \u00e0 base d'art\u00e9misinine soit achet\u00e9e, a augment\u00e9, de mani\u00e8re statistiquement non significative, de 5,48% (P = 0,574). Conclusion Dans la zone de notre \u00e9tude, le d\u00e9pistage du paludisme a \u00e9t\u00e9 augment\u00e9 en formant les pharmaciens \u00e0 l'utilisation de tests rapides et en leur fournissant l'acc\u00e8s \u00e0 ces tests \u00e0 un prix subventionn\u00e9. Des interventions suppl\u00e9mentaires peuvent \u00eatre n\u00e9cessaires pour atteindre une couverture plus \u00e9lev\u00e9e du d\u00e9pistage et un taux plus \u00e9lev\u00e9 de r\u00e9ponses appropri\u00e9es aux r\u00e9sultats de tests.\nJessica Cohen et al. | Malaria tests in drug shops, Uganda\n\u6458\u8981 | \n\u4e4c\u5e72\u8fbe\u836f\u5e97\u5f15\u5165\u759f\u75be\u5feb\u901f\u8bca\u65ad\u6d4b\u8bd5 \uff1a\u96c6\u7fa4\u968f\u673a\u5bf9\u7167\u8bd5\u9a8c | \n\u76ee\u7684 \u8bc4\u4f30\u5728\u4e4c\u5e72\u8fbe\u4e1c\u90e8\u836f\u5e97\u5f15\u5165\u5feb\u901f\u8bca\u65ad\u6d4b\u8bd5\u5bf9\u759f\u75be | \u4f9b\u5e94\u5546\u5b8c\u6210\u4e86\u5e73\u5747\u6bcf\u95f4\u5546\u5e97 146 \u6b21\u7684\u6d4b\u8bd5\u3002\u5bb6\u5ead\u62a5\u544a\n\u8bca\u65ad\u548c\u6cbb\u7597\u7684\u5f71\u54cd\u3002 | 22697 \u8d77\u53d1\u70ed\u6027\u75be\u75c5\u3002\u5f53\u5730\u836f\u5e97\u5feb\u901f\u6d4b\u8bd5\u7684\u53ef\u7528\u6027\u8ba9\n\u65b9\u6cd5 \u6574\u4f53\u800c\u8a00\uff0c\u5728 79 \u4e2a\u63a5\u53d7\u7814\u7a76\u7684\u6751\u5e84(\u81f3\u5c11\u6709\u4e00 | \u4efb\u4f55\u53d1\u70ed\u6027\u75be\u75c5\u63a5\u53d7\u759f\u75be\u68c0\u6d4b\u7684\u6982\u7387\u663e\u8457\u63d0\u9ad8 23.15%\n\u5bb6\u6267\u7167\u836f\u5e97)\u6709 2193 \u6237\u5bb6\u5ead\u7eb3\u5165\u8c03\u67e5\uff0c\u5e76\u63a5\u53d7 12 \u4e2a | (P = 0.015) \uff0c\u63a5\u53d7\u6297\u759f\u836f\u7269\u6cbb\u7597\u7684\u6982\u7387\u4e5f\u63d0\u9ad8\u4e86 8.84%\n\u6708\u7684\u89c2\u5bdf\u3002\u7ecf\u8fc7 3 \u4e2a\u6708\u7684\u89c2\u5bdf\uff0c\u5411 67 \u4e2a\u4e61\u6751\u4e2d\u968f\u673a\u9009 | (P = 0.056) \u3002\u9752\u84bf\u7d20\u8054\u5408\u7597\u6cd5\u7684\u6982\u7387\u6709\u7edf\u8ba1\u4e0a\u4e0d\u663e\u8457\u7684\n\u62e9\u8fdb\u884c\u5e72\u9884\u7684\u836f\u5e97\u4f9b\u5e94\u5546\u63d0\u4f9b\u4f7f\u7528\u759f\u75be\u5feb\u901f\u8bca\u65ad\u6d4b\u8bd5 | 5.48% \u7684\u63d0\u9ad8(P = 0.574) \u3002\n\u7684\u57f9\u8bad\uff0c\u5982\u679c\u5df2\u7ecf\u57f9\u8bad\u8fc7\uff0c\u5219\u4ee5\u8865\u8d34\u4ef7\u683c\u4e3a\u5176\u63d0\u4f9b\u6b64 | \u7ed3\u8bba \u5728\u6211\u4eec\u7684\u7814\u7a76\u533a\u57df\uff0c\u57f9\u8bad\u836f\u5e97\u4f9b\u5e94\u5546\u4f7f\u7528\u5feb\u901f\u68c0\n\u7c7b\u6d4b\u8bd5\u7684\u4f7f\u7528\u3002\u5269\u4f59\u7684 12 \u4e2a\u7814\u7a76\u6751\u5e84\u4f5c\u4e3a\u5bf9\u7167\u3002\u4f7f\u7528 | \u6d4b\u4ee5\u53ca\u4ee5\u8865\u8d34\u4ef7\u683c\u4e3a\u5176\u63d0\u4f9b\u6b64\u7c7b\u6d4b\u8bd5\u7684\u4f7f\u7528\u63d0\u9ad8\u4e86\u759f\n\u53cc\u91cd\u5dee\u5206\u56de\u5f52\u6a21\u578b\u8bc4\u4f30\u5e72\u9884\u7684\u5f71\u54cd\u3002 | \u75be\u6d4b\u8bd5\u7684\u6982\u7387\u3002\u53ef\u80fd\u9700\u8981\u989d\u5916\u7684\u5e72\u9884\u6765\u5b9e\u73b0\u6d4b\u8bd5\u7684\u66f4\n\u7ed3\u679c \u6765\u81ea 92 \u5bb6\u836f\u5e97\u7684\u4f9b\u5e94\u5546\u6210\u529f\u5b8c\u6210\u57f9\u8bad\uff0c50 \u5bb6\u79ef | \u9ad8\u8986\u76d6\u7387\u548c\u5bf9\u6d4b\u8bd5\u7ed3\u679c\u66f4\u9ad8\u6bd4\u4f8b\u7684\u6b63\u786e\u53cd\u5e94\u3002\n\u6781\u50a8\u5907\u548c\u6267\u884c\u5feb\u901f\u6d4b\u8bd5\u3002\u5728 9 \u4e2a\u591a\u6708\u4e2d\u7ecf\u8fc7\u57f9\u8bad\u7684 | \nR\u00e9sum\u00e9 | \nIntroduction de tests \u0420\u0435\u0437\u044e\u043c\u0435 | \n\u0412\u043d\u0435\u0434\u0440\u0435\u043d\u0438\u0435 \u0434\u0438\u0430\u0433\u043d\u043e\u0441\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u044d\u043a\u0441\u043f\u0440\u0435\u0441\u0441-\u0442\u0435\u0441\u0442\u043e\u0432 \u043d\u0430 \u043c\u0430\u043b\u044f\u0440\u0438\u044e \u0432 \u0430\u043f\u0442\u0435\u043a\u0438 \u0423\u0433\u0430\u043d\u0434\u044b: \u043a\u043b\u0430\u0441\u0442\u0435\u0440\u043d\u043e\u0435 \u0440\u0430\u043d\u0434\u043e\u043c\u0438\u0437\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u043e\u0435 \u043a\u043e\u043d\u0442\u0440\u043e\u043b\u0438\u0440\u0443\u0435\u043c\u043e\u0435 \u0438\u0441\u0441\u043b\u0435\u0434\u043e\u0432\u0430\u043d\u0438\u0435 \n\u0426\u0435\u043b\u044c \u041e\u0446\u0435\u043d\u0438\u0442\u044c \u0432\u043b\u0438\u044f\u043d\u0438\u0435 \u043d\u0430 \u0434\u0438\u0430\u0433\u043d\u043e\u0441\u0442\u0438\u043a\u0443 \u0438 \u043b\u0435\u0447\u0435\u043d\u0438\u0435 \u043c\u0430\u043b\u044f\u0440\u0438\u0438 | \u0414\u043e\u043c\u043e\u0445\u043e\u0437\u044f\u0439\u0441\u0442\u0432\u0430 \u0441\u043e\u043e\u0431\u0449\u0430\u043b\u0438 \u043e 22 697 \u0441\u043b\u0443\u0447\u0430\u044f\u0445 \u043b\u0438\u0445\u043e\u0440\u0430\u0434\u043e\u0447\u043d\u043e\u0433\u043e\n\u0432\u043d\u0435\u0434\u0440\u0435\u043d\u0438\u044f \u0432 \u0430\u043f\u0442\u0435\u043a\u0438 \u0432\u043e\u0441\u0442\u043e\u0447\u043d\u043e\u0439 \u0423\u0433\u0430\u043d\u0434\u044b \u0434\u0438\u0430\u0433\u043d\u043e\u0441\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u044d\u043a\u0441\u043f\u0440\u0435\u0441\u0441- | \u0437\u0430\u0431\u043e\u043b\u0435\u0432\u0430\u043d\u0438\u044f. \u041d\u0430\u043b\u0438\u0447\u0438\u0435 \u044d\u043a\u0441\u043f\u0440\u0435\u0441\u0441-\u0442\u0435\u0441\u0442\u043e\u0432 \u0432 \u043c\u0435\u0441\u0442\u043d\u044b\u0445 \u0430\u043f\u0442\u0435\u043a\u0430\u0445\n\u0442\u0435\u0441\u0442\u043e\u0432. | \u0437\u043d\u0430\u0447\u0438\u0442\u0435\u043b\u044c\u043d\u043e \u0443\u0432\u0435\u043b\u0438\u0447\u0438\u043b\u043e \u0432\u0435\u0440\u043e\u044f\u0442\u043d\u043e\u0441\u0442\u044c \u043f\u0440\u043e\u0432\u0435\u0440\u043a\u0438 \u043b\u044e\u0431\u043e\u0433\u043e\n\u041c\u0435\u0442\u043e\u0434\u044b \u0412 \u043e\u0431\u0449\u0435\u0439 \u0441\u043b\u043e\u0436\u043d\u043e\u0441\u0442\u0438 \u0432 \u0438\u0441\u0441\u043b\u0435\u0434\u043e\u0432\u0430\u043d\u0438\u0435 \u0431\u044b\u043b\u0438 \u0432\u043a\u043b\u044e\u0447\u0435\u043d\u044b \u0438 | \u043b\u0438\u0445\u043e\u0440\u0430\u0434\u043e\u0447\u043d\u043e\u0433\u043e \u0437\u0430\u0431\u043e\u043b\u0435\u0432\u0430\u043d\u0438\u044f \u043d\u0430 \u043c\u0430\u043b\u044f\u0440\u0438\u044e \u043d\u0430 23,15% (P = 0,015) \u0438\n\u043f\u0440\u043e\u0445\u043e\u0434\u0438\u043b\u0438 \u043c\u043e\u043d\u0438\u0442\u043e\u0440\u0438\u043d\u0433 \u0432 \u0442\u0435\u0447\u0435\u043d\u0438\u0435 12 \u043c\u0435\u0441\u044f\u0446\u0435\u0432 2193 \u0434\u043e\u043c\u043e\u0445\u043e\u0437\u044f\u0439\u0441\u0442\u0432\u0430 | \u0435\u0433\u043e \u043b\u0435\u0447\u0435\u043d\u0438\u044f \u043f\u0440\u043e\u0442\u0438\u0432\u043e\u043c\u0430\u043b\u044f\u0440\u0438\u0439\u043d\u044b\u043c\u0438 \u043f\u0440\u0435\u043f\u0430\u0440\u0430\u0442\u0430\u043c\u0438 \u043d\u0430 8,84% (P =\n\u0432 79 \u0438\u0441\u0441\u043b\u0435\u0434\u0443\u0435\u043c\u044b\u0445 \u0434\u0435\u0440\u0435\u0432\u043d\u044f\u0445 \u043f\u043e \u043a\u0440\u0430\u0439\u043d\u0435\u0439 \u043c\u0435\u0440\u0435 \u0441 \u043e\u0434\u043d\u043e\u0439 | 0,056). \u0412\u0435\u0440\u043e\u044f\u0442\u043d\u043e\u0441\u0442\u044c \u043f\u0440\u0438\u043e\u0431\u0440\u0435\u0442\u0435\u043d\u0438\u044f \u043a\u043e\u043c\u0431\u0438\u043d\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u044b\u0445 \u043f\u0440\u0435\u043f\u0430\u0440\u0430\u0442\u043e\u0432\n\u043b\u0438\u0446\u0435\u043d\u0437\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u043e\u0439 \u0430\u043f\u0442\u0435\u043a\u043e\u0439. \u041f\u043e\u0441\u043b\u0435 3 \u043c\u0435\u0441\u044f\u0446\u0435\u0432 \u043c\u043e\u043d\u0438\u0442\u043e\u0440\u0438\u043d\u0433\u0430 | \u043d\u0430 \u043e\u0441\u043d\u043e\u0432\u0435 \u0430\u0440\u0442\u0435\u043c\u0438\u0437\u0438\u043d\u0438\u043d\u0430 \u0443\u0432\u0435\u043b\u0438\u0447\u0438\u043b\u0430\u0441\u044c \u043d\u0430 \u0441\u0442\u0430\u0442\u0438\u0441\u0442\u0438\u0447\u0435\u0441\u043a\u0438\n\u0444\u0430\u0440\u043c\u0430\u0446\u0435\u0432\u0442\u0430\u043c \u0430\u043f\u0442\u0435\u043a \u0432 67 \u0434\u0435\u0440\u0435\u0432\u043d\u044f\u0445, \u0441\u043b\u0443\u0447\u0430\u0439\u043d\u044b\u043c \u043e\u0431\u0440\u0430\u0437\u043e\u043c | \u043d\u0435\u0437\u043d\u0430\u0447\u0438\u0442\u0435\u043b\u044c\u043d\u044b\u0439 \u043f\u043e\u043a\u0430\u0437\u0430\u0442\u0435\u043b\u044c 5,48% (P = 0,574).\n\u043e\u0442\u043e\u0431\u0440\u0430\u043d\u043d\u044b\u0445 \u0434\u043b\u044f \u0432\u043d\u0435\u0434\u0440\u0435\u043d\u0438\u044f \u044d\u043a\u0441\u043f\u0440\u0435\u0441\u0441-\u0442\u0435\u0441\u0442\u043e\u0432, \u0431\u044b\u043b\u043e \u043f\u0440\u0435\u0434\u043b\u043e\u0436\u0435\u043d\u043e | \u0412\u044b\u0432\u043e\u0434 \u0412 \u043d\u0430\u0448\u0435\u0439 \u043e\u0431\u043b\u0430\u0441\u0442\u0438 \u0438\u0441\u0441\u043b\u0435\u0434\u043e\u0432\u0430\u043d\u0438\u044f \u0447\u0430\u0441\u0442\u043e\u0442\u0430 \u0442\u0435\u0441\u0442\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u043d\u0430\n\u043e\u0431\u0443\u0447\u0435\u043d\u0438\u0435 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043d\u0438\u044e \u0434\u0438\u0430\u0433\u043d\u043e\u0441\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u044d\u043a\u0441\u043f\u0440\u0435\u0441\u0441-\u0442\u0435\u0441\u0442\u043e\u0432 \u043d\u0430 | \u043c\u0430\u043b\u044f\u0440\u0438\u044e \u0431\u044b\u043b\u0430 \u0443\u0432\u0435\u043b\u0438\u0447\u0435\u043d\u0430 \u043f\u0443\u0442\u0435\u043c \u043e\u0431\u0443\u0447\u0435\u043d\u0438\u044f \u0444\u0430\u0440\u043c\u0430\u0446\u0435\u0432\u0442\u043e\u0432 \u0430\u043f\u0442\u0435\u043a\n\u043c\u0430\u043b\u044f\u0440\u0438\u044e \u0438, \u043f\u0440\u0438 \u0443\u0441\u043b\u043e\u0432\u0438\u0438 \u0443\u0441\u043f\u0435\u0448\u043d\u043e\u0433\u043e \u043e\u0431\u0443\u0447\u0435\u043d\u0438\u044f, \u0434\u043e\u0441\u0442\u0443\u043f \u043a \u0442\u0430\u043a\u0438\u043c | \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043d\u0438\u044e \u044d\u043a\u0441\u043f\u0440\u0435\u0441\u0441-\u0442\u0435\u0441\u0442\u043e\u0432 \u0438 \u043f\u0440\u0435\u0434\u043e\u0441\u0442\u0430\u0432\u043b\u0435\u043d\u0438\u044f \u0438\u043c \u0434\u043e\u0441\u0442\u0443\u043f\u0430\n\u0442\u0435\u0441\u0442\u0430\u043c \u043f\u043e \u0441\u0443\u0431\u0441\u0438\u0434\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u043e\u0439 \u0446\u0435\u043d\u0435. \u041e\u0441\u0442\u0430\u043b\u044c\u043d\u044b\u0435 12 \u0438\u0441\u0441\u043b\u0435\u0434\u0443\u0435\u043c\u044b\u0445 | \u043a \u043f\u043e\u0434\u043e\u0431\u043d\u044b\u043c \u0442\u0435\u0441\u0442\u0430\u043c \u043f\u043e \u0441\u0443\u0431\u0441\u0438\u0434\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u043e\u0439 \u0446\u0435\u043d\u0435. \u0414\u043b\u044f \u0434\u043e\u0441\u0442\u0438\u0436\u0435\u043d\u0438\u044f\n\u0434\u0435\u0440\u0435\u0432\u0435\u043d\u044c \u0441\u043b\u0443\u0436\u0438\u043b\u0438 \u0432 \u043a\u0430\u0447\u0435\u0441\u0442\u0432\u0435 \u043a\u043e\u043d\u0442\u0440\u043e\u043b\u044c\u043d\u044b\u0445. \u0414\u043b\u044f \u043e\u0446\u0435\u043d\u043a\u0438 | \u0431\u043e\u043b\u0435\u0435 \u0448\u0438\u0440\u043e\u0432\u043e\u0433\u043e \u043e\u0445\u0432\u0430\u0442\u0430 \u0442\u0435\u0441\u0442\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0435\u043c \u0438 \u0431\u043e\u043b\u0435\u0435 \u0432\u044b\u0441\u043e\u043a\u043e\u0433\u043e\n\u0432\u043d\u0435\u0434\u0440\u0435\u043d\u0438\u044f \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043b\u0430\u0441\u044c \u0440\u0435\u0433\u0440\u0435\u0441\u0441\u0438\u043e\u043d\u043d\u0430\u044f \u043c\u043e\u0434\u0435\u043b\u044c \u00ab\u0440\u0430\u0437\u043d\u043e\u0441\u0442\u044c | \u043f\u0440\u043e\u0446\u0435\u043d\u0442\u0430 \u043f\u0440\u0438\u043d\u044f\u0442\u0438\u044f \u0441\u043e\u043e\u0442\u0432\u0435\u0442\u0441\u0442\u0432\u0443\u044e\u0449\u0438\u0445 \u043c\u0435\u0440 \u043f\u043e \u0440\u0435\u0437\u0443\u043b\u044c\u0442\u0430\u0442\u0430\u043c \u0442\u0435\u0441\u0442\u043e\u0432\n\u0440\u0430\u0437\u043d\u043e\u0441\u0442\u0435\u0439\u00bb. | \u043c\u043e\u0433\u0443\u0442 \u043f\u043e\u0442\u0440\u0435\u0431\u043e\u0432\u0430\u0442\u044c\u0441\u044f \u0434\u043e\u043f\u043e\u043b\u043d\u0438\u0442\u0435\u043b\u044c\u043d\u044b\u0435 \u043c\u0435\u0440\u043e\u043f\u0440\u0438\u044f\u0442\u0438\u044f.\n\u0420\u0435\u0437\u0443\u043b\u044c\u0442\u0430\u0442\u044b \u0424\u0430\u0440\u043c\u0430\u0446\u0435\u0432\u0442\u044b \u0438\u0437 92 \u0430\u043f\u0442\u0435\u043a \u0443\u0441\u043f\u0435\u0448\u043d\u043e \u043f\u0440\u043e\u0448\u043b\u0438 \u043e\u0431\u0443\u0447\u0435\u043d\u0438\u0435 | \n\u0438 50 \u0430\u043a\u0442\u0438\u0432\u043d\u043e \u0444\u043e\u0440\u043c\u0438\u0440\u043e\u0432\u0430\u043b\u0438 \u0437\u0430\u043f\u0430\u0441\u044b \u0438 \u0432\u044b\u043f\u043e\u043b\u043d\u044f\u043b\u0438 \u044d\u043a\u0441\u043f\u0440\u0435\u0441\u0441-\u0442\u0435\u0441\u0442\u044b. | \n\u0417\u0430 \u0434\u0435\u0432\u044f\u0442\u0438\u043c\u0435\u0441\u044f\u0447\u043d\u044b\u0439 \u043f\u0435\u0440\u0438\u043e\u0434 \u043d\u0430 \u043e\u0434\u043d\u0443 \u0430\u043f\u0442\u0435\u043a\u0443 \u043f\u0440\u0438\u0445\u043e\u0434\u0438\u043b\u043e\u0441\u044c \u0432 | \n\u0441\u0440\u0435\u0434\u043d\u0435\u043c 146 \u0442\u0435\u0441\u0442\u043e\u0432, \u0432\u044b\u043f\u043e\u043b\u043d\u0435\u043d\u043d\u044b\u0445 \u043e\u0431\u0443\u0447\u0435\u043d\u043d\u044b\u043c\u0438 \u0444\u0430\u0440\u043c\u0430\u0446\u0435\u0432\u0442\u0430\u043c\u0438. | \nFunding:\nThe study was funded by the Clinton Health Access Initiative and the Bill & Melinda Gates Foundation.\nCompeting interests: During the study period, JC received research support from the Clinton Health Access Initiative, which also employed KM, KB and FA.\n\u202b\u0645\u0644\u062e\u0635\u202c \u202b\u0644\u0644\u0645\u0631\u0627\u0642\u0628\u0629\u202c \u202b\u062e\u0627\u0636\u0639\u0629\u202c \u202b\u0639\u0646\u0642\u0648\u062f\u064a\u0629\u202c \u202b\u0639\u0634\u0648\u0627\u0626\u064a\u0629\u202c \u202b\u062c\u062a\u0631\u0628\u0629\u202c \u202b\u0623\u0648\u063a\u0646\u062f\u0627:\u202c \u202b\u064a\u0641\u202c \u202b\u0627\u0623\u0644\u062f\u0648\u064a\u0629\u202c \u202b\u062e\u0645\u0627\u0632\u0646\u202c \u202b\u0625\u0649\u0644\u202c \u202b\u0644\u0644\u0645\u0627\u0644\u0631\u064a\u0627\u202c \u202b\u0627\u0644\u0631\u0633\u064a\u0639\u0629\u202c \u202b\u0627\u0644\u062a\u0634\u062e\u064a\u0635\u064a\u0629\u202c \u202b\u0627\u0627\u0644\u062e\u062a\u0628\u0627\u0631\u0627\u062a\u202c \u202b\u062a\u0642\u062f\u064a\u0645\u202c",
        "label": "low",
        "id": "task4_RLD_test_991"
    },
    {
        "paper_doi": "10.1111/iwj.12836",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallelEthics and informed consent: yesSample size estimate: noFollow-up period: 30 days postoperativelyITT analysis: yes, number randomised: 129 groin incisions (100 participants), number analysed: 129 incisionsFunding: \"funded by our own department, without any financial or scientific involvement or support from KCI, ACELITY Company\"Preregistration: no\n\n\nParticipants: Location: Germany\nIntervention group: n = 58 incisions,control group: n = 71 incisionsMean age: intervention group = 71 (range 54 to 89),control group = 66.5 (range 41 to 86)\nInclusion criteria: vascular procedures with access to the common femoral artery with at least 1 of the known main risk factors of wound healing: age > 50 years, diabetes mellitus, renal insufficiency, malnutrition, obesity, and chronic obstructive pulmonary disease.\nExclusion criteria: not stated\n\n\nInterventions: Aim/s: to investigate the effectiveness of ciNPT compared with conventional therapy with regard to the incidence of groin WHC on postoperative days 5 to 7 and 30 and the incidence of surgery revisions 30 days postoperatively after various vascular surgeries.\n Group 1 (NPWT) intervention: ciNPT applied for postoperative days 5 to 7Group 2 (control) intervention: a conventional adhesive plaster that was changed daily\nStudy date/s: 1 February to 30 October 2015\n\n\nOutcomes: Wound complications including SSIValidity of measure/s: Szilagyi classificationTime points: the first evaluation took place on postoperative days 5 to 7 during the hospital stay, while the second evaluation was conducted on postoperative day 30 in the outpatient clinic.\n\n\nNotes: \n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "Abstract\nGroin wound infections in patients undergoing vascular procedures often cause a lengthy process of wound healing.\nSeveral clinical studies and case reports show a reduction of surgical site infections (SSIs) in various wound types after using closed incision negative pressure therapy (ciNPT).\nThe aim of this prospective, randomised, single\u2010institution study was to investigate the effectiveness of ciNPT (PREVENA\u2122 Therapy) compared to conventional therapy on groin incisions after vascular surgery.\nFrom 1 February to 30 October 2015, 100 patients with 129 groin incisions were analysed.\nPatients were randomised and treated with either ciNPT (n = 58 groins) or the control dressing (n = 71 groins).\nciNPT was applied intraoperatively and removed on days 5\u20137 postoperatively.\nThe control group received a conventional adhesive plaster.\nWound evaluation based on the Szilagyi classification took place postoperatively on days 5\u20137 and 30.\nCompared to the control group, the ciNPT group showed a significant reduction in wound complications (P < 0\u00b70005) after both wound evaluation periods and in revision surgeries (P = 0\u00b7022) until 30 days postoperatively.\nSubgroup analysis revealed that ciNPT had a significant effect on almost all examined risk factors for wound healing.\nciNPT significantly reduced the incidence of incision complications and revision procedures after vascular surgery.\nIntroduction\nThe healing process of postoperative groin wounds, often rich in complications, frequently leads to a long course with high treatment costs1.\nDue to the anatomical proximity to the lymph nodes and urogenital organs, as well as its function as a leading access in vascular procedures, the groin is prone to infections.\nAdditional wound\u2010healing complications (WHC) in the groin include wound dehiscence, lymphatic leaks with lymphatic fistula and lymphocele, seroma, haematoma, skin necrosis and delayed healing 1, 2, 3, 4, 5, 6, 7.\nSurgical site infections (SSIs) are present in 2\u201322% of all surgical procedures and contribute more than 20% of the costs of all complicated wounds 8, 9.\nIn accordance with international data, the incidence of SSIs after vascular surgery in the groin are 3\u201344%, and deep groin infections with prosthetic material involvement are described in up to 6% of cases 4, 10.\nThe relationship between SSIs and morbidity correlates with extended hospital stay, severe limb ischaemia, extremity loss, massive haemorrhage, systemic sepsis and septic embolisation 1, 4, 5.\nDe spite increasing knowledge of systemic wound\u2010healing factors and many surgical techniques (e.g., sloping groin cut, implantation of obturator and lateral femoral bypasses, use of antibiotic\u2010coated prosthesis, rotation flaps, and fibrin glue), only systemic antibiotic therapy has yielded acceptable results 3, 4, 5, 6, 11.\nAdditionally, negative pressure wound therapy (NPWT; V.A.C.\u00ae\nTherapy, KCI, an ACELITY Company, San Antonio, TX) has been used to manage groin incisions.\nSince the development of NPWT by Morykwas and Argenta in the United States and Fleischmann in Germany in the second half of the 1990s, this method has been used as a supporting therapy for wound healing 12, 13.\nIn subsequent years, several case reports and clinical studies described the effectiveness of NPWT in the management of the following wounds: complex open wounds intended for secondary closure, infected wounds as a supplement to surgical debridement and antibiotic therapy, degloving injuries, sternal wound dehiscence, wounds after open traumatic injuries and high\u2010energy trauma wounds.\nNPWT has also been used as a method to bolster transplanted skin grafts 7, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25.\nThis successful use led to the idea of applying the NPWT dressing on primarily closed wounds to facilitate incision healing.\nUnder the designation closed incision negative pressure therapy (ciNPT), this new technique has resulted in many significant clinical results 11, 13, 25, 26.\nSince 2010, multiple studies and case reports comparing standard\u2010of\u2010care dressings to ciNPT have reported a decrease in SSIs in a wide spectrum of traumatic, orthopaedic, abdominal, sternal and plastic surgery incisions 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37.\nThe reason for this success may be due to the reported mechanisms of action of the ciNPT, which protects the incision from external wound contamination, strengthens the cohesiveness of the wound edges, removes fluids and infectious materials from the wound, decreases the lateral tension around the incision and facilitates oxygen saturation and blood microcirculation within the incision area 11, 38, 39.\nciNPT as delivered by the PREVENA\u2122 Incision Management Therapy System (KCI, an ACELITY Company) consists of a vacuum unit with a battery and a preset negative pressure of \u2212125 mmHg.\nIts integrated individual components includes a mobile therapy unit with a replaceable exudate collection canister, a polyester fabric interface layer with 0\u00b7019% silver for the control of bioburden within the dressing, a polyurethane foam bolster and a polyurethane film with acrylic adhesive.\nA polyurethane shell encapsulates the foam bolster and interface layer, providing a closed system.\nUntil now, four clinical studies have reported on the use of the ciNPT in the groin after vascular surgeries 4, 9, 10, 40; however, these studies lacked a prospective, randomised study design and a subgroup analysis of risk factors and perioperative parameters.\nThe aim of this study was to investigate the effectiveness of ciNPT compared to conventional therapy with regards to the incidence of groin WHC on postoperative days 5\u20137 and 30 and the incidence of surgery revisions 30 days postoperatively after various vascular surgeries.\nAdditionally, subgroup analyses of the main wound\u2010healing risk factors and perioperative risk factors were evaluated to assess the effect of ciNPT on specific patients at risk of postoperative WHC in the groin.\nFurthermore, a logistic regression and a receiver operating characteristic (ROC) analysis were conducted to forecast the risk of postoperative WHC within the main patient risk factors and perioperative risk factors.\nMethods\nThis prospective, randomised, monocentric study design was approved by the ethics committee of Justus Liebig University Giessen.\nThe study was conducted independently and was fully funded by our own department, without any financial or scientific involvement or support from KCI, ACELITY Company.\nFrom 1 February to 30 October 2015, 100 patients with 129 groin incisions were evaluated.\nInclusion criteria were as follows: vascular procedures with access to the common femoral artery with at least one of the known main risk factors of wound healing: age >50 years, diabetes mellitus, renal insufficiency, malnutrition, obesity and chronic obstructive pulmonary disease (COPD).\nAll patients received at least a 5\u2010cm longitudinal incision in the groin.\nThe total number of groins was divided into either the ciNPT group or the control group.\nAll patients received perioperative antibiotic prophylaxis and a preoperative hair shave and sterile skin disinfection with the antiseptic kodan\u00aeTinktur forte (Sch\u00fclke & Mayr GmbH 22840 Norderstedt, Germany) in the surgery area.\nAfter placing a drain, subcutaneous tissue was re\u2010approximated with Vicryl 2.0 sutures (Johnson & Johnson Medical GmbH, Ethicon, Norderstedt, Germany), and the skin was secured with a skin\u2010clamping device (WECK Visistat 35 W. Teleflex Medical GmbH. Kernen, Germany).\nAfterwards, ciNPT (utilizing PREVENA\u2122 Therapy) was applied on the incision (Figure 1).\nOn postoperative days 5\u20137, ciNPT was removed, and a conventional adhesive plaster (Cosmopor E Steril Hartmannn, Heidenheim, Germany) was used.\nThe control group received a conventional adhesive plaster that was changed daily.\nWound evaluations were determined at two\u2010time points.\nThe first evaluation took place on postoperative days 5\u20137 during the hospital stay, and the second evaluation was conducted on postoperative day 30 in the outpatient clinic.\nGroin incisions were graded using the Szilagyi classification 41.\nGrade I describes superficial infections that remain restricted on the skin.\nGrade II contains an infiltration of the subcutaneous layer without participation of the arterial graft.\nGrade III describes an infection involving the arterial graft.\nAlthough this classification system mainly describes only tissue and prosthetic infections, it does consider the anatomical tissue layers for the evaluation of all kinds of groin wound complications.\nIn this study, patients with cutaneous wound dehiscence, skin necrosis and single local infection signs were classified as grade I.\nWound dehiscence in the subcutaneous layer, haematoma, lymphatic fistula, lymphocele, seroma, single local infection signs and systemic infection parameters [leukocytes >13 109/dl, C\u2010reactive protein (CRP) > 100 mg/l] were classified as grade II.\nAll classical local infection signs (pain, swelling, redness and hyperaemia, warmth, dysfunction), systemic infection parameters and arterial graft infections were classified as grade III.\nSubgroup analysis included the main wound\u2010healing risk factors and perioperative risk factors.\nAll risk factors were examined with regards to the incidence of groin incision complications on postoperative days 5\u20137 and 30 and surgery revisions until postoperative day 30.\nThe main risk factors were defined as: age > 50 years, diabetes mellitus with haemoglobin A1c (HbA1c) > 6\u00b75% and 48 mmol/mol glucose, renal insufficiency with glomerular filtration rate < 89 ml/min (stage 2) and creatinine >1\u00b73 mg/dl, overweight with BMI > 25 kg/m2, malnutrition with albumin <35 g/l, protein <65 g/l and transferrin <2 g/l and COPD with the Global Initiative For Chronic Obstructive Lung Disease (GOLD) grade 1 FEV1 \u2265 80%.\nThe perioperative risk factors were defined as wound length > 8 cm, hospital stay >8 days, operative time > 142 minutes, perioperative blood transfusion with haemoglobin <8 mg/dl and previous vascular interventions (Digital Subtraction Angiography or Percutaneous Transluminal Angioplasty).\nStatistical analysis was performed with the Student's test, Levene's test and Fisher's exact test.\nFor the subgroup analyses, the following tests were performed: Fisher's exact test and Pearson Chi Square test.\nFurther analytical methods within the study included logistic regression, ROC analysis and calculation of the correlation coefficients.\nStatistical significance was determined by a P\u2010value <0\u00b705.\nResults\nThe study included 100 patients with 129 groin wounds.\nThe patients included 28 females and 72 males, with a median age of 68\u00b75 \u00b1 9\u00b76.\nOf the 129 groin wounds, 29 were a result of bilateral surgery.\nThe most frequently reported comorbidities were peripheral artery disease (62%) and abdominal aortic aneurysm (21%) (Table 1).\nPrior surgeries included femoral endarterectomy (29%), endovascular aneurysm repair (EVAR) (24%) and femoral popliteal bypass (21%) (Table 2).\nFor the main analysis, there were 35 (27\u00b71%) groin WHCs, with 5 (8\u00b76%) in the ciNPT group (n = 58) and 30 (42\u00b73%) in the control group (n = 71).\nThe first postoperative wound examination on postoperative days 5\u20137 in the ciNPT group showed no WHC in Szilagyi grades I\u2013III wounds, while the control group had 5 (7%) in Szilagyi grade I and 10 (14\u00b71%) in Szilagyi grade II.\nDuring the second examination on postoperative day 30 in the ciNPT group, four (6\u00b79%) WHCs were noted in Szilagyi grade I and one (1\u00b77%) in Szilagyi grade II.\nThe control group showed 3 (4\u00b72%) WHCs in Szilagyi grade I, 10 (14\u00b71%) in grade II and 2 (2\u00b78%) in grade III (Table 3; Figures 2 and 3).\nBoth WHCs in Szilagyi grade III appeared after implantation of a femoro\u2010femoral cross\u2010over bypass.\nBecause of the infection, the prosthesis was removed immediately.\nThe overall incidence of postoperative wound complications (P < 0\u00b70005) and the incidence on postoperative days 5\u20137 (P < 0\u00b70005) and 30 (P = 0\u00b7023) were statistically significant and favoured the ciNPT group (Table 3).\nWhen comparing the incidence of revision surgeries, there was only 1 (1\u00b77%) case in the ciNPT group versus 10 (14\u00b71%) cases in the control group.\nThe comparison of both groups showed a significant advantage for ciNPT (P = 0\u00b7022) (Table 3).\nThe most frequently occurring WHC in the ciNPT group was superficial wound dehiscence.\nIn the control group, haematoma and local infection were the leading WHCs (Table 4).\nFor both groups, local infections with and without revision surgery were treated with antibiotics.\nOne patient died on day 1 postoperatively in the ciNPT group.\nThe cause of death was unrelated to ciNPT.\nFurther subgroup analyses were based on the perioperative risk factors of wound length > 8 cm, hospital stay >8 days, operating time > 42 minutes, previous interventions and perioperative blood transfusions with regards to the incidence of groin WHCs.\nIn patients who had a wound length > 8 cm, the ciNPT group had significantly fewer total WHCs as compared to the control group (P = 0\u00b7003).\nThese results were similar for ciNPT patients in the following subgroups: operation time > 142 minutes (P = 0\u00b70005), hospital stay >8 days (P = 0\u00b7001) and perioperative blood transfusion (P = 0\u00b7004).\nThe significant effect of ciNPT with regards to groin WHCs was also observed on postoperative days 5\u20137 in the following subgroups: wound length > 8 cm (P = 0\u00b7015), operation time > 142 minutes (P = 0\u00b7002), hospital stay >8 days (P = 0\u00b7001) and perioperative blood transfusion (P = 0\u00b7023).\nHowever, only patients in the hospital stay >8 days (P = 0\u00b7014) or operation time > 142 minutes (P = 0\u00b7020) subgroups showed a positive effect of ciNPT on day 30.\nWith respect to revision surgeries, only ciNPT patients who had a hospital stay >8 days had significantly fewer revision surgeries compared to control patients [1 (2\u00b77%) versus 10 (20\u00b78%), respectively; P = 0\u00b7012] (Table 5).\nIn the logistic regression, all single risk factors were examined against the aim variable postoperative WHCs.\nThe biggest predictor for the development of postoperative WHCs could be shown only in the following perioperative risk factors: wound length (P = 0\u00b7003, OR = 4\u00b7800) and operation time (P = 0\u00b7046, OR = 2\u00b7571).\nWhen the all risk factors were investigated, the wound length (P = 0\u00b7015, OR = 7\u00b7503) showed the greatest potential for the prediction of postoperative WHCs.\nIn the ROC analysis, accurate forecasting of a postoperative WHC could be indicated but only for the perioperative risk factors with an area under the curve (AUC) of 0\u00b7662 (P = 0\u00b7016) (Figure 4).\nA closer consideration of wound length and operation time as the biggest predictors demonstrated that the classification achievement of the index perioperative risk factors in the logistic regression in a new ROC analysis is due primarily to the potential of wound length (AUC 0\u00b7664, P = 0\u00b7007) and operation time (AUC 0\u00b7690, P = 0\u00b7005) (Figure 4).\nA complementary calculation of the correlation coefficients of all risk factors demonstrated that wound length with operation time (0\u00b7386) and overweight status (\u22120\u00b7200) had the best correlation.\nOther correlating risk factors included renal insufficiency with age (0\u00b7342) and diabetes mellitus with excessive weight (0\u00b7326).\nThe performed statistical analyses of the subgroups of patients were based on the main wound\u2010healing risk factors of age (>50 years), diabetes mellitus, renal insufficiency, malnutrition, overweight and COPD with regards to the incidence of groin WHCs.\nIn patients whose age was >50 years, the ciNPT group had significantly fewer total WHCs as compared to the control group (P < 0\u00b70005).\nThese results were similar for the ciNPT patients in the following subgroups: diabetes mellitus (P < 0\u00b70005), renal insufficiency (P < 0\u00b70005), malnutrition (P = 0\u00b7043) and overweight status (P < 0\u00b70005).\nOn postoperative days 5\u20137, patients in all wound\u2010healing risk factor subgroups, except malnutrition (P = 0\u00b7081) and COPD (P = 0\u00b7206), showed a significant result for ciNPT.\nOn postoperative day 30, significance was found only for ciNPT patients in the age subgroup (P = 0\u00b7040).\nWith regards to the incidence of revision surgeries, only ciNPT patients in the age (>50 years) subgroup had fewer revision surgeries compared to the control patients [1 (3\u00b72%) versus 6 (23\u00b71%), respectively; P = 0\u00b7029] (Table 5).\nDiscussion\nAgainst the background of extended hospital stays and higher treatment costs, preventive measures to reduce the incidence of postoperative complications play an important role.\nIn order to optimise the advancement of preventive procedures, the effectiveness of ciNPT on groin wounds after vascular surgeries was evaluated.\nThe effect of ciNPT has been demonstrated in clinical studies and case reports in a variety wound types 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37.\nIn spite of many publications, clinical studies involving groin wounds are rare.\nTo date, there exist only four clinical studies in which ciNPT were examined on groin wounds after vascular surgeries 4, 9, 10, 40.\nOur results point out that ciNPT had a significant effect on the reduction of the incidence of wound complications and revision surgeries as well as the effect on almost all mean and perioperative risk factors.\nThe results of this study are comparable to the Matatov et al. study 4 given that both studies are similar in design and sample size.\nIn our study and the Matatov et al. evaluation, there were significant reductions in the overall incidence of WHCs (P = 0\u00b70011 and P < 0\u00b70005, respectively).\nAll three wound infections (6%) in the Matatov study were classified as Szilagyi grade I, while in this study, there were four incision complications classified as Szilagyi grade I and only one grade II.\nSimilarly, the results of this study were consistent with the results of Weir 9 (with only two incision complications) and Karl and Woeste 10 (with no complications) after application of ciNPT.\nThis observation supports the reduction of incision complications in deep tissue layers with possible revision surgeries when ciNPT is utilised.\nThis observation was confirmed in this study by the significant decrease of revision surgeries within the first 30 days postoperatively.\nThe statistical significance (P < 0\u00b70005) in the Szilagyi grade II complications after two evaluation periods with a wound proportion of 1:20 brings out the effectiveness of ciNPT (Table 3).\nThe detailed view of the separate types of wound complications in this study shows a particular reduction of subcutaneous haematoma with a clear statistical significance (P = 0\u00b7020) and clarifies the positive impact of the ciNPT suction effect (Table 4).\nAn 8\u00b76% difference with no WHCs on days 5\u20137 postoperatively and five WHCs on day 30 postoperatively could suggest that ciNPT loses its effectiveness after being removed.\nThe removal of ciNPT may compromise the sterile wound conditions and can lead to potential wound contamination during subsequent wound dressing changes, which may be interpreted as a casual explanation for higher wound complication rates on day 30 postoperatively.\nThis view is based on evidence on the number of local wound infections in the control group (Table 4).\nciNPT functions as a barrier against potential wound contamination and serves as an effective method for the removal of wound exudate.\nTo prolong the beneficial wound management effects of ciNPT over days 5\u20137 postoperatively, a longer application time of ciNPT should be considered.\nAlternative options to this could be a strictly sterile execution of the routine changes of the wound dressing and covering the incision wound with a dry gauze with an adhesive dressing attached on it as a contamination barrier.\nCertainly, the question about the application period of these measures should be discussed prior to removing surgical staples in order for the procedure to serve as a guideline.\nDespite significant results in the subgroup analysis, a loss of effectiveness was shown in almost all risk factors after days 5\u20137 postoperatively.\nA possible reason for this may be found in the loss of effectiveness of ciNPT over a longer time period as described above.\nIf special pathological influences on the individual risk factors exist and therefore limit the effectiveness of the ciNPT, then they must be cleared by secondary examinations.\nIn the study by Matatov et al., 4 subgroup analysis and isolated view on two evaluation periods regarding the incidence of incision complications were unfortunately not carried out.\nIn this subgroup analysis, ciNPT had a significant effect on clinical use with regards to all main and perioperative risk factors (Table 5).\nThis is strengthened by the results of our logistic regression analysis, the ROC analysis and the correlation calculation.\nIn particular, the perioperative risk factors, wound length and operation time (in the logistic regression analysis and in the ROC analysis) were the strongest predictors for postoperative WHCs (Figure 4).\nThe calculation of the correlations especially emphasised a stronger relationship between the wound length and the operation time.\nTherefore, the wound length and operation time could be determined in all three investigations as the strongest predictors with regards to the cause for postoperative WHCs in the groin.\nBy these data, the option of a specific indication for the application of the ciNPT is given and an arbitrary execution of this therapy can be prevented.\nAs a practical assistance in the specific use of the ciNPT, a scoring system to justify its indication appears to be useful.\nTherefore, we constructed a scoring system based on the significant study data in which the risk factors with the highest significance were assigned a value of 2 points and the risk factors with a lower significance were assigned 1 point (Table 6).\nThe limit for the indication of ciNPT was based on the average score of the point values of the scoring system.\nThe average score for all patients treated with the ciNPT was 7\u00b75 points.\nThe average score for patients with the ciNPT without WHCs was 8\u00b74 points.\nIn accordance with these average scores, the lower limit for the indication of ciNPT was set at 8 points.\nWith the use of this scoring system, 12% of the patients in the ciNPT group showed a WHC compared to 88% without a WHC.\nAccording to these data, it becomes evident that by using this scoring system for the ciNPT, postoperative groin WHCs may be prevented.\nThrough a firm consideration of the significant risk factors in this study within the preoperative phase, ciNPT can be specifically used as an efficient incision management measure to prevent postoperative inguinal WHCs.\nA similar conclusion was presented by Willy et al. in the international multidisciplinary consensus recommendations, where the experts examined 100 publications and recommended the consideration of ciNPT use for patients with risk factors and high\u2010risk procedures 42.\nGiven the fact that the study data regarding the application of ciNPT in groin wounds after vascular surgery are based solely on the clinical examinations with the PREVENA\u2122\nIncision Management Therapy System, the question arises whether other ciNPT systems can achieve a similar positive effect on groin wounds.\nAn almost equivalent ciNPT system is the PICO\u2122 Single Use Negative Pressure Wound Therapy System (Smith & Nephew, London, UK), which in comparison with the PREVENA\u2122\nIncision Management Therapy System eliminates the wound secretions by evaporation and uses less negative pressure (\u221280 mmHg).\nTo clarify whether both ciNPT systems show related positive clinical results in groin wounds, further studies are needed.\nConclusions\nIn comparison to conventional adhesive plaster, the use of ciNPT demonstrates a statistically significant reduction of postoperative WHCs in the groin on postoperative days 5\u20137 and 30 and revision surgeries until day 30 postoperatively in patients after several vascular surgeries.\nThe results of the subgroup analysis show a significant effect of ciNPT on almost all examined risk factors, through which a specific preventative use may be possible for patients with a corresponding risk profile.\n(A) Components of ciNPT; B) ciNPT after aortobifemoral bypass.\nWound complications of study patients based on Szilagyi classification. (A) Szilagyi I: Skin necrosis, superficial wound dehiscence and local infection; (B) Szilagyi II: Deep wound dehiscence and fat necrosis; (C) Szilagyi III: Prosthetic graft infection.\nWound results after removing ciNPT on (A) 5\u20137 days and (B) 30 days postoperatively.\nROC curve of (A) all perioperative risk factors, (B) perioperative risk factor operation time and (C) perioperative risk factor wound length.\n\nPatient characteristics\n | ciNPT group | Control group | P\u2010value\nNumber of patients | 43 | 57 | \nNumber of groin incisions | 58 | 71 | \nGender |  |  | \nMale | 29 (67%) | 43 (75%) | 0\u00b75\nFemale | 14 (33%) | 14 (25%) | 0\u00b75\nMean age [years] | 71 (range 54\u201389) | 66\u00b75 (range 41\u201386) | 0\u00b7020\nMean BMI [kg/m2] | 26\u00b77 (range 19\u00b71\u201337\u00b73) | 27\u00b78 (range 18\u00b74\u201337\u00b72) | 0\u00b7205\nHypertension | 38 (88%) | 53 (93%) | 0\u00b7325\nCoronary artery disease | 22 (51%) | 13 (23%) | 0\u00b7003\nDiabetes mellitus | 22 (51%) | 29 (51%) | 1\nRenal insufficiency | 27 (63%) | 30 (53%) | 0\u00b7415\nDialysis | 0 (0%) | 2 (35%) | 0\u00b7322\nMalnutrition | 13 (30%) | 22 (39%) | 0\u00b7406\nCOPD | 9 (21%) | 8 (14%) | 0\u00b7791\nSmoker | 23 (53%) | 22 (39%) | 0\u00b7159\nPreoperative anaemia | 19 (44%) | 30 (53%) | 0\u00b7426\nPostoperative anaemia | 19 (44%) | 27 (51%) | 0\u00b7840\nPostoperative leucocytosis | 22 (51%) | 33 (58%) | 0\u00b7547\nPeripheral artery disease |  |  | \nFontaine classification grade II | 13 (30%) | 26 (46%) | 0\u00b7149\nFontaine classification grade III | 5 (12%) | 2 (4%) | 0\u00b7234\nFontaine classification grade IV | 6 (14%) | 10 (18%) | 0\u00b7785\nInfrarenal abdominal aortic aneurysm | 14 (33%) | 7 (12%) | 0\u00b7024\nThoracic aortic aneurysm | 1 (2%) | 4 (7%) | 0\u00b7387\nThoracic abdominal aortic aneurysm | 3 (7%) | 5 (9%) | 1\nInfrarenal aortic stenosis | 0 (0%) | 1(2%) | 1\nArtery occlusion (thrombosis/embolism) | 0 (0%) | 3 (5%) | 0\u00b7257\nVisceral artery aneurysm | 0 (0%) | 1 (2%) | 1\nLeriche syndrome | 1 (2%) | 1 (2%) | 1\n\nBMI, body mass index; COPD, chronic obstructive pulmonary disease.\n\nPerioperative characteristics\n | ciNPT group | Control group | P\u2010value\nMean operative time [minutes] | 140 (range 40\u2013436) | 146 (range 32\u2013402) | 0\u00b7706\nMean hospital stay [days] | 12\u00b78 (range 5\u201343) | 13\u00b70 (range 5\u201344) | 0\u00b7909\nMean wound length [cm] | 7\u00b77 (range 5\u201315) | 8\u00b76 (range 5\u201315) | 0\u00b7017\nPerioperative blood transfusion | 9 (21%) | 13 (23%) | 1\nProcedure types |  |  | \nEVAR/TEVAR | 19 (44\u00b72%) | 17 (30%) | 0\u00b7148\nRevascularisation | 26 (61%) | 41 (72%) | 0\u00b7284\nBilateral procedures | 19 (44\u00b72%) | 14 (26%) | 0\u00b7053\nProsthetic material used |  |  | \nPTFE | 4 (9\u00b73%) | 6 (10\u00b75%) | 1\nDacron | 2 (4\u00b77%) | 4 (7%) | 0\u00b7697\nDacron patch | 10 (23\u00b73%) | 18 (31\u00b76%) | 0\u00b7380\nVein | 6 (14%) | 7 (12\u00b73%) | 1\n\nEVAR, endovascular aortic repair; PTFE, polytetrafluoroethylene; TEVAR, thoracic endovascular aortic repair.\n\nIncidence of wound\u2010healing disturbances with reference to the total number of groin incisions, wound evaluation on 5\u20137 and 30 day postoperatively and revision surgery on 30 day postoperatively based on Szilagyi classification\n | Total number | 5\u20137 day postoperatively | 30 day postoperatively | Revision surgery on 30 day postoperatively\nSzilagyi classification | ciNPT group n = 58 | Control group n = 71 | P\u2010value | ciNPT group n = 58 | Control group n = 71 | P\u2010value | ciNPT group n = 58 | Control group n = 71 | P\u2010value | ciNPT group n = 58 | Control group n = 71 | P\u2010value\nSzilagyi grade I | 4 (6\u00b79%) | 8 (11\u00b73%) | 0\u00b7545 | 0 (0%) | 5 (7%) | 0\u00b7064 | 4 (6\u00b79%) | 3 (4\u00b72%) | 0\u00b7070 | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501\nSzilagyi grade II | 1 (1\u00b77%) | 20 (28\u00b72%) | <0\u00b70005 | 0 (0%) | 10 (14\u00b71%) | 0\u00b7005 | 1 (1\u00b77%) | 10 (14\u00b71%) | 0\u00b7022 | 1 (1\u00b77%) | 6 (8\u00b75%) | 0\u00b7128\nSzilagyi grade III | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501 | 0 (0%) | 0 (0%) | 1 | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501 | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501\nTotal number | 5 (8\u00b76%) | 30 (42\u00b73%) | <0\u00b70005 | 0 (0%) | 15 (21\u00b71%) | <0\u00b70005 | 5 (8\u00b76%) | 15 (21\u00b71%) | 0\u00b7023 | 1 (1\u00b77%) | 10 (14\u00b71%) | 0\u00b7022\n\n\nTypes of wound complications within the three grades of Szilagyi classification\n | ciNPT group | Control group | P\u2010value\nSuperficial wound dehiscence | 3 (7%) | 4 (7%) | 1\nSkin necrosis | 1 (2\u00b73%) | 3 (5%) | 0\u00b7632\nDeep wound dehiscence with fat necrosis | 1 (2\u00b73%) | 4 (7%) | 0\u00b7387\nHaematoma | 0 (0%) | 8 (14%) | 0\u00b7020\nSeroma | 0 (0%) | 1 (1\u00b78%) | 1\nLymphatic fistula | 1 (2\u00b73%) | 3 (5\u00b73%) | 0\u00b7632\nArterial graft infection | 0 (0%) | 2 (4%) | 0\u00b7322\nLocal infection | 1 (2\u00b73%) | 10 (17\u00b75%) | 0\u00b7022\n\n\nAnalyses on subgroups of patients based on main wound\u2010healing risk factors and perioperative risk factors with regards to WHCs and revision surgeries\n | Analysis Intervals | \n | Total number of WHCs | Number of WHCs at postoperative days 5\u20137 | Number of WHCs at postoperative day 30 | Patients requiring revision surgery on 30 day postoperatively\nPatient subgroups | ciNPT group | Control group | P\u2010value | ciNPT group | Control group | P\u2010value | ciNPT group | Control group | P\u2010value | ciNPT group | Control group | P\u2010value\nAge (>50 years) | n = 31, 2 (6\u00b75%) | n = 26, 18 (69\u00b72%) | <0\u00b70005 | n = 31, 0 (0%) | n = 26, 11 (42\u00b73%) | <0\u00b70005 | n = 31, 2 (6\u00b75%) | n = 26, 7 (26\u00b79%) | 0\u00b7040 | n = 31, 1 (3\u00b72%) | n = 26, 6 (23\u00b71%) | 0\u00b7029\nDiabetes mellitus | n = 22, 2 (9\u00b71%) | n = 29, 17 (58\u00b76%) | <0\u00b70005 | n = 22, 0 (0%) | n = 29, 8 (27\u00b76%) | 0\u00b7007 | n = 22, 2 (9\u00b71%) | n = 29, 9 (31%) | 0\u00b7059 | n = 22, 1 (4\u00b75%) | n = 29, 7 (24\u00b71%) | 0\u00b7061\nRenal insufficiency | n = 27, 2 (7\u00b74%) | n = 30, 15 (50%) | <0\u00b70005 | n = 27, 0 (0%) | n = 30, 9 (30%) | 0\u00b7002 | n = 27, 2 (7\u00b74%) | n = 30, 6 (20%) | 0\u00b7163 | n = 27, 1 (3\u00b77%) | n = 30, 3 (10%) | 0\u00b7347\nMalnutrition | n = 13, 2 (15\u00b74%) | n = 22, 11 (50%) | 0\u00b7043 | n = 13, 0 (0%) | n = 22, 5 (22\u00b77%) | 0\u00b7081 | n = 13, 2 (15\u00b74%) | n = 22, 6 (27\u00b73%) | 0\u00b7355 | n = 13, 0 (0%) | n = 22, 4 (18\u00b72%) | 0\u00b7140\nOverweight | n = 32, 5 (15\u00b76%) | n = 41, 23 (56\u00b71%) | <0\u00b70005 | n = 32, 0 (0%) | n = 41, 14 (34\u00b71%) | <0\u00b70005 | n = 32, 5 (15\u00b76%) | n = 41, 9 (22%) | 0\u00b7354 | n = 32, 1 (3\u00b71%) | n = 41, 5 (12\u00b72%) | 0\u00b7167\nCOPD | n = 9, 1 (11\u00b71%) | n = 8, 4 (50%) | 0\u00b7111 | n = 9, 0 (0%) | n = 8, 2 (25%) | 0\u00b7206 | n = 9, 1 (11\u00b71%) | n = 8, 2 (25%) | 0\u00b7453 | n = 9, 1 (11\u00b71%) | n = 8, 1 (12\u00b75%) | 0\u00b7735\nWound length (>8 centimetre) | n = 25, 4 (16%) | n = 49, 25 (51%) | 0\u00b7003 | n = 25, 0 (0%) | n = 49, 13 (26\u00b75%) | 0\u00b7015 | n = 25, 4 (16%) | n = 49, 12 (24\u00b75%) | 0\u00b7197 | n = 25, 1 (4%) | n = 49, 9 (18\u00b74%) | 0\u00b7083\nHospital stay (> 8 days) | n = 37, 3 (8,1%) | n = 48, 28 (58\u00b73%) | 0\u00b7001 | n = 37, 0 (0%) | n = 48, 14 (29\u00b72%) | 0\u00b7001 | n = 37, 3 (8\u00b71%) | n = 48, 14 (29\u00b72%) | 0\u00b7014 | n = 37, 1 (2\u00b77%) | n = 48, 10 (20\u00b78%) | 0\u00b7012\nOperation time (> 142 minutes) | n = 21, 2 (9\u00b75%) | n = 28, 21 (75%) | <0\u00b70005 | n = 21, 0 (0%) | n = 28, 10 (35\u00b77%) | 0\u00b7002 | n = 21, 2 (9\u00b75%) | n = 28, 11 (39\u00b73%) | 0\u00b7020 | n = 21, 1 (4\u00b77%) | n = 28, 7 (25%) | 0\u00b7062\nPreviousinterventions | n = 9, 1 (11\u00b71%) | n = 18, 7 (38\u00b79%) | 0\u00b7149 | n = 9, 0 (0%) | n = 18, 4 (22\u00b72%) | 0\u00b7174 | n = 9, 1 (11\u00b71%) | n = 18, 3 (16\u00b77%) | 0\u00b7593 | n = 9, 0 (0%) | n = 18, 3 (16\u00b77%) | 0\u00b7279\nPerioperative blood transfusion | n = 9, 1 (11\u00b71%) | n = 13, 10 (77%) | 0\u00b7004 | n = 9, 0 (0%) | n = 13, 6 (46\u00b71%) | 0\u00b7023 | n = 9, 1 (11\u00b71%) | n = 13, 4 (31%) | 0\u00b7230 | n = 9, 1 (11\u00b71%) | n = 13, 2 (15\u00b74%) | 0\u00b7642\n\nCOPD, chronic obstructive pulmonary disease.\n\nScoring system for ciNPT based on the significant risk factors for groin WHCs\nRisk factors | Points\nPatient age (P < 0\u00b70005) | 2\nDiabetes mellitus (P < 0\u00b70005) | 2\nRenal insufficiency (P < 0\u00b70005) | 2\nOverweight (P < 0\u00b70005) | 2\nOperation time (P < 0\u00b70005) | 2\nMalnutrition (P = 0\u00b7043) | 1\nWound length (P = 0\u00b7003) | 1\nPerioperative blood transfusion (P = 0\u00b7004) | 1\n",
        "label": "unclear",
        "id": "task4_RLD_test_714"
    },
    {
        "paper_doi": "10.1186/1471-2393-12-11",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Parallel-arm individual-randomised RCT conducted at 3 primary care trusts in Birmingham, UK between Jul 2010 and Oct 2011. Trial name: ELSIPS.\n\n\nParticipants: Sample size: 1324 women.Inclusion criteria: nulliparous women < 28 weeks' gestation assessed by a midwife as having specific social risk. (Risk factors included housing problems, lack of social support, smoking, low maternal weight or obesity, teenage, late booking for ANC.)Exclusion criteria: women under 16 years of age, or teenage mothers recruited to another national trial of additional support during pregnancy.\n\n\nInterventions: Target: health system (re-organisation of health services: home visits).Arm 1 (662 women): POW provided support, including home visits, in addition to standard ANC and PNC. The POW organised antenatal visits and advised on lifestyle changes. In addition to emotional and health-related support, the POW helped with financial, legal or benefits problems and with housing. The POW also provided support with care of the newborn, including breastfeeding.Arm 2 (662 women): women in the control group received standard ANC and PNC.\n\n\nOutcomes: Trial primary outcome: Edinburgh Postnatal Depression Scale1 (EPDS) 8-12 weeks postpartum and antenatal visits attended.Review outcomes reported:Primary: ANC coverage (at least 10 contacts).Secondary: preterm birth (< 34 weeks), low birthweight infants, perinatal mortality.Other: depression scores.\n\nFollow-up: intervention involved individual support during pregnancy and follow-up to 8-12 weeks postpartum.We did not include data for preterm birth < 34 weeks because our review's definition of preterm birth < 37 weeks.Outcome data from unpublished paper obtained from author: SL Kenyon, s.kenyon@bham.ac.uk.\n\n\nNotes: Funders: this work was funded by the National Institute for Health Research (NIHR) through the Collaborations for Leadership in Applied Health Research and Care for Birmingham and Black Country (CLAHRC-BBC) programme. The views expressed in this publication are not necessarily those of the NIHR, the Department of Health\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Background\nMaternal, neonatal and child health outcomes are worse in families from black and ethnic minority groups and disadvantaged backgrounds.\nThere is little evidence on whether lay support improves maternal and infant outcomes among women with complex social needs within a disadvantaged multi-ethnic population in the United Kingdom (UK).\nMethod/Design\nThe aim of this study is to evaluate a lay Pregnancy Outreach Worker (POW) service for nulliparous women identified as having social risk within a maternity service that is systematically assessing social risks alongside the usual obstetric and medical risks.\nThe study design is a randomised controlled trial (RCT) in nulliparous women assessed as having social risk comparing standard maternity care with the addition of referral to the POW support service.\nThe POWs work alongside community midwifery teams and offer individualised support to women to encourage engagement with services (health and social care) from randomisation (before 28 weeks gestation) until 6 weeks after birth.\nThe primary outcomes have been chosen on the basis that they are linked to maternal and infant health.\nThe two primary outcomes are engagement with antenatal care, assessed by the number of antenatal visits; and maternal depression, assessed using the Edinburgh Postnatal Depression Scale at 8-12 weeks after birth.\nSecondary outcomes include maternal and neonatal morbidity and mortality, routine child health assessments, including immunisation uptake and breastfeeding at 6 weeks.\nOther psychological outcomes (self efficacy) and mother-to-infant bonding will also be collected using validated tools.\nA sample size of 1316 will provide 90% power (at the 5% significance level) to detect increased engagement with antenatal services of 1.5 visits and a reduction of 1.5 in the average EPDS score for women with two or more social risk factors, with power in excess of this for women with any social risk factor.\nAnalysis will be by intention to treat.\nQualitative research will explore the POWs' daily work in context.\nThis will complement the findings of the RCT through a triangulation of quantitative and qualitative data on the process of the intervention, and identify other contextual factors that affect the implementation of the intervention.\nDiscussion\nThe trial will provide high quality evidence as to whether or not lay support (POW) offered to women identified with social risk factors improves engagement with maternity services and reduces numbers of women with depression.\nMREC number\n10/H1207/23\nTrial registration number\nISRCTN: ISRCTN35027323\nBackground\nMaternal, neonatal and child health outcomes are worse in women from black and ethnic minority and disadvantaged groups and there are a range of factors likely to be contributing, one of these being inclusivity and engagement with services.\nThis was an important focus of the National Service Framework (NSF) on Maternity and Maternity Matters, both of which emphasise choice, access and continuity of care in a safe service.\nThe last two Confidential Enquiries into Maternal Deaths and Saving Mothers' Lives made recommendations about care for vulnerable women with socially complex lives.\nSocial disadvantage, living in a poor community and being from a minority ethnic group, including asylum seekers and newly arrived refugees, were all major risk factors.\nBlack African women, including asylum seekers and newly arrived refugees had a mortality rate nearly six times higher than White women..\nClearly maternal death is rare but it is well documented that women from these vulnerable groups book for antenatal care later, make fewer visits, experience greater pregnancy morbidity and have a higher risk of adverse fetal and child health outcomes.\nA recent UK national cohort study found severe maternal morbidities were significantly more common among women from black African and Caribbean and Pakistani ethnic groups than in White women.\nThe authors suggested that these differences may be due to pre-existing medical factors or factors related to care during pregnancy, labour or birth but they are unlikely to be due to differences in age, socioeconomic or smoking status, body mass index or parity.\nThis study further highlighted the importance of tailored maternity services and improving access to care for women of ethnic minorities.\nBased on the assumption that increased engagement with antenatal services will result in improved maternal and perinatal health outcomes, the maternity NSF recommended that services are proactive in engaging all women, particularly those from disadvantaged groups.\nThis includes contact early in their pregnancy and maintenance of contact before and after birth.\nIt notes that some women in these groups may require more support and access to social or other services, for example housing and benefits advice.\nIn informing the NSF, evidence was sought on how services may be organised and delivered to improve outcomes for disadvantaged groups, but little good evidence was found.\nThe National Institute for Health and Clinical Excellence (NICE) has recently published a Guideline for Models of Service Provision for pregnant women with complex social factors and found little high quality evidence.\nOne of the research recommendations was to answer the question 'Is intervention and/or family support provided by statutory and 3rd sector agencies effective in improving outcomes for women and their babies?'\nAdditional social support during pregnancy for vulnerable groups might, on the face of it, be of possible benefit.\nHowever, a recently updated Cochrane review of 'Support during pregnancy for women at increased risk of low birth weight' (LBW) found 18 RCTs and concluded that programmes offering additional social support were not associated with improvements in any perinatal outcomes.\nIn most of the trials, however, participants were selected because they had obstetric rather than social risks for LBW and almost all support interventions were delivered by trained professionals, which may not be the most likely person to improve outcomes.\nThe review did find an overall reduction in Caesarean section (RR 0.87, 95%CI 0.78 to 0.97), and noted that some trials found improvements in maternal psychosocial outcomes.\nIt is well documented that a reduction in maternal depression, in addition to improving maternal wellbeing, will have a beneficial effect on short and long term child outcomes.\nThere is some evidence on benefits of lay support in other areas of maternity care from the Cochrane review on 'Continuous Support for Women during Childbirth', which suggested that the beneficial effects associated with continuous support were greater when the provider was not a member of the hospital staff.\nThe review showed that when the providers of continuous support were members of staff, spontaneous vaginal birth (SVB) was increased.\nWhen the providers of support were not staff members SVB appeared to be further increased.\nThis therefore adds support to the hypothesis that care may be better received when provided by lay people rather than health professionals.\nEvidence from three much quoted trials by Olds in the United States, of nurse home visitation from pregnancy up to 2 years for vulnerable groups (one mainly teen, single mothers and the second, young deprived African-Americans) found improvements in some pregnancy-related outcomes, including greater engagement with maternity and related services, and went on to find beneficial effects in several long-term maternal and child outcomes.\nOnly the first trial, however, found any effect on preterm delivery and birth weight and only in the small sub-groups of those aged 14-16 and of smokers.\nThe third trial by Olds also examined the effects of home visitation by lay workers but found no benefit from this type of worker.\nEvaluation of the Old's support model of nurse home visitation is currently being undertaken throughout the UK (Family Nurse Partnership (FNP)).\nAt present therefore, there is little evidence on whether lay support improves maternal and infant outcomes among women with complex social needs within a disadvantaged multi-ethnic population in the UK.\nA recent meta-synthesis into barriers to antenatal care for marginalised women in high income countries has suggested that a non-judgemental, contextually tailored antenatal service that pays attention to the specific circumstances of disadvantaged women may increase sustained access to care.\nSo at least in theory, care that provides individual case management including home visiting, as provided by a POW service, could be of benefit.\nThe aim of this study is to evaluate, by a randomised controlled trial, a POW service for nulliparous women identified as having social risk within a maternity service that is systematically assessing social risks alongside the usual obstetric and medical risks.\nMethods/Design\nDesign\nThe study design is an individually randomised controlled trial involving three primary care trusts (PCTs) in Birmingham, with nulliparous women assessed as having social risk, randomised to standard maternity care or the addition of referral to the POW support service (See Figure 1).\nSetting and population\nThe POW service and its evaluation will run across the whole of Birmingham, which currently comprises three PCTs: Heart of Birmingham (HoB PCT); South Birmingham (SB PCT) and Birmingham East and North (BEN PCT).\nAlthough varying in proportion, these PCTs all include a population that has high levels of deprivation and with a variety of ethnic groups, including many recently arrived mothers, refugees and asylum seekers.\nFor example, in HoB PCT there are 5500-6000 births each year; almost 90% of which are to women in deprived wards (Index of Multiple Deprivation (IMD) quintile 5).\nOnly about 15% are of European ethnicity, 54% are within the South Asian group (31% Pakistani, 11% Bangladeshi and 9% Indian), 13% are African and 10% African-Caribbean.\nOne in four mothers has themselves been born outside the UK and many are recent immigrants who have little English language skills.\nIn SB PCT, about 50% of births occur to women in the most deprived IMD quintile of deprivation and 60% of mothers are white.\nBEN PCT has a population somewhere between these in terms of the proportion of women in the ethnic minority groups and levels of deprivation.\nIn summary, women from black and ethnic minority groups with complex social risk and needs accounts for a large and increasing proportion of the maternity population cared for by the three maternity units (Birmingham Women's National Health Service (NHS) Foundation Trust, Sandwell & West Birmingham NHS Trust, and Heart of England NHS Foundation Trust) in these PCTs.\nUsual care (control)\nTo maximise access and engagement with maternity services the local PCTs have commissioned a new service model based on social risk assessment, alongside assessing and managing obstetric and medical risk in the usual way.\nTo assess social risk a set of items have recently been included in the standardised maternity notes used universally in Birmingham.\nThese are completed at the booking visit by the midwife and identify whether a woman has any of the following factors:\n\u2022 UK resident for under a year.\n\u2022 Difficulty with the English language, both spoken and written.\n\u2022 Housing problems, such as rent arrears, temporary accommodation, registered with National Asylum Support Service (NASS) or of No Fixed Abode (NFA).\n\u2022 No support from either partner or family or friend\n\u2022 Woman/household member in receipt of social services support, including child protection.\n\u2022 Identified benefit problem.\n\u2022 Smoking.\n\u2022 Drug misuse, including other's in the household.\n\u2022 Alcohol misuse.\n\u2022 Clinical diagnosis of past or present mental illness.\n\u2022 Teen parent (under 20 years old).\n\u2022 Domestic abuse.\n\u2022 Body Mass Index less than or equal to 18 OR more than or equal to 35.\n\u2022 Late booking (defined as booking after 18 weeks gestation).\n\u2022 Did Not Attend 2 or more antenatal appointments (under 28 weeks gestation).\nThe intention is that systematic social risk assessment will maximise the likelihood that social risk is identified and needs are met.\nMidwives identifying women with social risk factors currently either signpost women to services that may be beneficial or refer them to specialised agencies or personnel.\nThis may mean they signpost to support agencies (for example, housing or benefit offices) or refer to other agencies (for example, social services), or refer onto the specialist midwives in their Trust.\nThese specialist midwives act as a contact point and provide specific advice and support for women experiencing problems such as domestic abuse, mental health issues or who are teenagers when they are pregnant.\nIntervention\nThe PCTs have also decided to provide a POW service to complement these pathways, with the intention of further increasing full engagement with care during pregnancy and postpartum, and to improve the women's social conditions.\nThe ultimate aim of this, theoretically at least, is to improve the health of both mother and baby by increasing engagement with antenatal services, which should reduce perinatal mortality and morbidity.\nIt is also hoped that additional support of this nature would improve women's psychological health, which in turn would have a positive impact on the child.\nThe POW service is in addition to standard maternity care and will not be available to nulliparous women other than within the trial.\nWomen assessed as having social risk and randomised to the intervention group will be referred to a POW who will provide individual case management including home visiting.\nThe purpose of the POW service is to ensure that women attend antenatal appointments and engage with required care, such as taking prescribed medication, attending scan appointments, and including making lifestyle changes, such as smoking cessation.\nAdditionally in the postnatal period the POWs are providing breast feeding support (World Health Organization Baby Friendly Initiative) and advice about feeding and caring for the baby.\nThe POWs also provide social support on such issues as ensuring that available benefits are obtained, housing difficulties are dealt with, mental health problems managed and overall well-being is maximised.\nThe philosophy underlying POW support is an attempt to help women to become more able to manage problems that arise in life, that is, to enhance their general self-efficacy.\nThe POWs receive appropriate training to National Vocational Qualification (NVQ) level 3 which is provided by 'Gateway Family Services' and have access to supervision from experts with specific skills and knowledge.\nPostpartum POW contact will continue until 6 weeks after birth when transfer to the Family Support Worker (FSW) would take place for those who require it.\nEligibility\nInclusion criteria\n\u2022 Nulliparous women < 28 weeks gestation.\n\u2022 Assessed by the midwife as having specified social risk through routine systematic assessment.\nNulliparous is defined as never having given birth to a child; this will include women who have had a miscarriage/s or termination/s of pregnancy.\nWe have chosen under 28 weeks as an inclusion criterion to give adequate time for the POW service to impact on the outcomes.\nExclusion criteria\nOne of the PCTs (SB PCT) participating in this trial is also involved in a national trial of additional support to pregnant teenagers, called the Family Nurse Partnership (FNP).\nThe FNP intervention is health professionals providing intensive support throughout pregnancy up to 2 years after birth.\nWe will exclude teenagers recruited to FNP, but do not expect this to greatly affect recruitment to ELSIPS.\nWe will also exclude those under 16 years of age due to the complexity of gaining informed consent from this group.\nRecruitment and randomisation\nAs part of the booking visit, all women in Birmingham have a systematic social risk assessment undertaken by the midwife.\nOnce social risk has been identified the midwife will offer the support routinely available (as part of standard care) and will discuss the additional support of the POW only available through the trial.\nIf the woman is potentially interested in taking part, her details will be passed to the ELSIPS midwife within the community midwifery team.\nRandomisation is by random permuted blocked design stratified by Trust.\nThe randomisation lists were generated by the trial statistician (KH) and then forwarded to the University of Birmingham Primary Care Clinical Research and Trials Unit who provided a telephone randomisation service thus ensuring concealment of allocation.\nThe ELSIPS midwives\nAll the midwives within their individual teams refer eligible women to the ELSIPS midwife.\nThe ELSIPS midwife is responsible for obtaining informed consent and randomising the women.\nThey are also responsible within their teams for promoting the trial and training other midwives in trial processes.\nThe ELSIPS midwives work for a varying number of hours (3-7) per week depending on the size of their team and the number of births.\nThey each have an agreed target for the numbers of women we expect to be recruited each month.\nThey are trained and supported by the University of Birmingham team.\nOutcome measures\nThe primary outcomes have been chosen on the basis that they are linked to maternal and infant health.\nThe two primary outcomes are engagement with antenatal care, assessed based on number of antenatal visits, and maternal depression, assessed using the Edinburgh Postnatal Depression Scale (EPDS) at 8-12 weeks after birth.\nSecondary outcomes\nMaternal outcomes will include\n\u2022 length of labour (first, second and third stages),\n\u2022 mode of birth (spontaneous vaginal birth, instrumental birth or caesarean section),\n\u2022 perineal trauma (episiotomy, degree of laceration),\n\u2022 incidence of maternal morbidity (e.g., postpartum haemorrhage, shoulder dystocia, chorionamnioitis),\n\u2022 length of stay in hospital,\n\u2022 engagement with other services, as required (e.g., smoking cessation service).\nBaby outcomes are mainly markers of poor perinatal outcome\n\u2022 composite outcome of adverse perinatal outcome comprising:\n- perinatal mortality;\n- preterm birth before 34 weeks;\n- birth weight 10th centile or below;\n- admission to neonatal unit\n\u2022 Apgar score at 5 minutes,\n\u2022 arterial cord blood gases, if taken\n\u2022 breastfeeding initiation rate,\n\u2022 length of stay in hospital,\n\u2022 oxygen at 36 weeks post conceptual age, if applicable,\n\u2022 retinopathy of prematurity, if applicable,\n\u2022 abnormal cerebral ultrasound prior to discharge (e.g., intraparenchymal cerebral bleed, hydrocephalus, parenchymal cysts), if applicable,\n\u2022 necrotising enterocolitis (Bells Stage I, II or III), if applicable,\n\u2022 culture positive sepsis requiring greater than 5 days antibiotic treatment, if applicable.\nLonger term infant outcomes\n\u2022 Routine child health assessments, including immunisation uptake and breastfeeding continuation at 6 weeks.\nPsychological outcomes\n\u2022 Self efficacy (using Pearlin and Schooler Mastery Scale).\n\u2022 Mother-to-infant bonding tool.\nThe detrimental impact of maternal bonding difficulties on both the emotional and cognitive development of the child and the quality of the mother-infant relationship has been well documented in the literature.\nWe have therefore chosen to evaluate mother-to-infant bonding.\nSample size\nCurrently, NICE recommend, in the Antenatal Care Guideline, that the schedule of appointments should be determined by the function of the appointments and for nulliparous women with an uncomplicated pregnancy, they recommend that a schedule of ten visits should be adequate.\nThe actual number of appointments attended was the subject of a survey of women's experiences of maternity care carried out in 2006.\nThis was a national survey which used a random sample of 4800 women and achieved a response rate of 63%.\nWomen were sent a postal questionnaire three months after birth.\nNulliparous women reported they attended a mean of 10.9 (standard deviation [SD] 6) antenatal appointments (Table 1).\nThere is some evidence relating to number of antenatal visits and perinatal outcomes.\nA systematic review found that in settings where the number of visits is already low, reduced visits programmes of antenatal care for low risk women are associated with an increase in perinatal mortality compared to standard care, although admission to neonatal intensive care may be reduced.\nAn observational study explored the relationship between the number of antenatal visits made by 17,765 British women and adverse perinatal outcomes.\nNo consistent relationship between admission to neonatal unit or perinatal mortality and number of antenatal visits was found.\nA significant positive relationship was found between number of antenatal visits and Caesarean section, and low birth weight (less than 2500 g) was positively associated with number of visits for nulliparous women but not for parous women.\nMore recently, a cohort study from Finland found under-attending free antenatal care was associated with adverse pregnancy outcomes.\nLogistic regression analyses found there were significantly more low birthweight infants in the under and non-attenders, with more fetal and neonatal death.\nEstimates of the baseline Edinburgh postnatal depression score are taken from the controls of a Cochrane systematic review of Psychosocial and psychological interventions for preventing postpartum depression.\nIt is well documented that a reduction in maternal depression, in addition to improving maternal wellbeing, will have a beneficial effect on short and long term child outcomes and it is plausible that social support provided by the POWs could reduce the numbers of women becoming depressed.\nStudies have shown that depressed mothers are more likely to demonstrate impaired maternal-infant interactions and negative perceptions of infant behaviour.\nChildren of depressed mothers are more likely to suffer a range of adverse outcomes, including insecure attachment, behavioural problems, cognitive developmental deficits and difficulties in emotional functioning, some of these continuing into adolescence.\nA sample size of 421 per arm would provide 90% power (at the 5% significance level) to detect a reduction of 1.5 in the average EPDS score from say 7 to 5.5, and would provide greater than 90% power to detect increased engagement with antenatal services of either 1.5 or 2 visits (Table 2).\nThis calculation has also allowed for 20% drop-out or loss to follow-up.\nFollowing a successful six month pilot in which 475 women were recruited a revision to the sample size was agreed.\nPrior to the pilot there was no data on the extent of the social risk factors amongst women and data from the pilot showed that 36% of the women recruited had one social risk factor.\nIt was agreed to power the study to detect the pre-specified differences in the primary outcomes in the sub group of women with two or more social risk factors, which lead to an increase in the sample size to 658 women per arm.\nIt is anticipated that this sample size of 1316 will be obtained by 31st December 2011.\nStatistical analysis\nWe will calculate the mean number of visits and mean EPDS for each arm and compare differences using the t-test or other appropriate non-parametric test.\nVariations will be explored by pre-specified sub-group comparisons.\nVariations in participant baseline characteristics between intervention and control groups will also be explored, and if appropriate, variations in participant characteristics adjusted for using generalised linear models with appropriate consideration of strata of randomisation.\nFor secondary continuous outcomes, variations between control and intervention groups will be investigated using the t-test and, if necessary, covariate adjustment made using generalised linear models (to include strata affects); for secondary binary outcomes, differences will be compared using chi-squared tests and, if necessary, logistic regression.\nFor all analyses, assumptions of various tests and models will be explored and non-parametric tests used if required.\nAll analyses will be by intention to treat and 95% confidence intervals will be quoted throughout.\nMissing covariate and outcome data will be examined.\nReasons for withdrawal, lack of participation and any reasons for non-compliance will be documented and explored.\nComplete case and available case analyses will be completed in the first instance.\nIf the amount of missing data is not insignificant, then multiple imputation will be used to evaluate sensitivity to the missing completely at random assumption, and inferences compared to those under the lesser missing at random assumption.\nPre-specified sub-group comparisons will be according to number of social risks (1 social risk or 2 or more social risks) identified and gestation at recruitment (< 12 weeks, 12-19 + 6 weeks, 20-27 + 6 weeks).\nThese pre-specified subgroup comparisons will be for the primary outcomes and the more clinically important secondary outcomes (perinatal composite outcome, self-efficacy and Mother-to-infant bonding).\nThe level of significance will at 0.05.\nThe remaining secondary outcomes will have the level of significance at 0.01.\nQualitative component to the ELSIPS study\nWe will undertake additional qualitative research in order to understand the nature of the work of the POW in more depth.\nThis will complement the findings of the RCT, in which process information is already being collected by the POWs about the components of their work, including the frequency, venue, duration, support offered, additional social risk disclosure and referrals to other agencies.\nThere are two aims of this work:\n1. It will allow us to ensure that the components and process of the intervention itself (the relationship and support offered by the POW) are fully understood, and help future policy makers determine whether the intervention might need to be adapted to reflect their population and health system.\n2. It may help to identify aspects of the intervention that are particularly successful and areas that would benefit from future redesign or efficiency/quality improvement work.\nQualitative approach: A grounded theory approach will be taken to this part of the study, in order to understand the nature of the POWs work from their perspective.\nWe have chosen to avoid formal interviews of the POWs about their work as this tends to reproduce the 'theory' about the role, idealised accounts and retrospective explanations about action and is therefore less likely to be able to fully uncover any disjunction between the process information collected about their work and the complexity of daily practice.\nWe have selected shadowing, a form of focused ethnography, to understand the nature and content of the daily work of the POWs in a naturalistic environment.\nThis enables observations of action-incontext that can be triangulated with informal reflective discussions with POWs on aspects of their work that have been observed.\nData collection method: Two researchers (one social scientist, one clinical researcher) will singly observe POWs undertaking their daily work, including meetings with clients, until theme saturation is achieved.\nIt is anticipated that this will take approximately 100 hours.\nSampling: Initial sampling will be purposive to include two POWs from each of the three localities covered by the service.\nThe rationale behind this was that each locality serves a population with very different socio-demographic characteristics.\nWe will coordinate with the POW managers to ensure that we are able to observe interactions with women at different stages in the POW service.\nAny further sampling will be theoretical (see analysis below).\nRecruitment: Researchers will attend team meetings to inform POWs about the study.\nWritten consent will be obtained from the POWs for the shadowing.\nIf POWs are selected to participate, all their current clients will receive a letter informing them that their POW may be accompanied by a researcher and that they can opt out without it affecting the service they receive.\nEach POW will confirm orally that the client has received and understood the letter on a case-by-case basis before any meetings are observed.\nAnalysis: Field notes will be taken during the observation and detailed reflective accounts of the observation written up immediately following the observation.\nAfter each day of shadowing, the two researchers will meet to debrief, discuss the data and emerging themes, and to identify additional data required to elaborate the properties of emerging themes and test them.\nEmergent themes will be used to interrogate existing related theory in the literature and extend it.\nData anonymization and storage: Full field notes will only be available to the core qualitative analysis group (NG, LI, SK).\nBefore dissemination, all data will be anonymized and any potentially identifying features of the clients the POWs worked with will be removed.\nTrial oversight\nA Steering Committee has been formed from all those involved to monitor progress and oversight is provided by the Birmingham and Black Country Collaboration in Leadership in Applied Research and Care (CLAHRC) Steering Committee (Chair Dr Rashmi Shukla).\nThis trial comprises of part of the work undertaken by Theme 5.\nEthical approval has been obtained from South Birmingham Ethics Committee (10/H1207/23)\nApproval at each site has been obtained from the local Research and Development Directorates.\nParticipants receive an information leaflet (full and summary) and sign a consent form which are available on the website, http://www.bham.ac.uk/elsips\nDiscussion\nThe original intention of the ELSIPS trial was to include all pregnant women.\nHowever, our detailed investigations of current social support within the PCTs have found that many multiparous women with high social need will already have been allocated a FSW through their local Children's Centre.\nThis is because multiparous women access the Children's Centres for services for their children under 5 years old and so come into contact with the FSW on a regular basis.\nThese FSWs are provided through local education, as well as health services funding, and many provide very similar support to a POW.\nThe main distinction between FSWs and POWs is the FSWs do not specifically engage with nulliparous women in the antenatal period.\nOn this basis if multiparae were included, the trial comparison may fail to find a real difference because of the dilution effect of FSWs.\nIn addition, contamination may arise if systematically more multiparae randomised to standard care who did not already have a FSW were subsequently provided with one because of not having been allocated to a POW.\nWe explored powering the study using a composite primary outcome of perinatal morbidity and mortality but the substantial sample size required to show even a large difference, together with funding constraints for the POW service whilst under evaluation, have meant we have opted for the smaller sample size required for the primary outcomes of engagement with services and EPDS at 8-12 weeks after birth.\nAlthough the scientific basis for antenatal care does not appear to be as robust as it might, it is based on the assumption that engagement with services results in improved maternal and perinatal health outcomes.\nSo, the number of visits (both consultant and midwife) attended should act as a surrogate for improved maternal and neonatal outcomes.\nWe have chosen to evaluate self-efficacy as one of the psychological outcomes rather than self-esteem as it is more closely related to the changes the POWs are intended to facilitate.\nSelf-esteem is believed to reflect evaluation of one's overall self-worth, while self-efficacy is specifically concerned with evaluation of one's performance.\nWe have chosen to evaluate differences in general self-efficacy rather than parenting specific efficacy because it is this broad construct that we believe will be promoted by the POWs.\nThe aim of the ELSIPS trial is to provide much needed high quality evidence of the effect of individualised support provided by a lay worker (in this instance a POW), working alongside community midwifery teams, to encourage engagement of nulliparous women with identified social risk factors and the effect on health related outcomes for both mother and baby.\nIt is anticipated that results will be available in the summer of 2013.\nFlow diagram to summarise allocation and contact throughout the trial.\n\nEstimated frequency of the events which make up the primary outcomes\nOutcome | Average | Source of information\nNumber of antenatal visits by nullips (consultant and midwife) | Average number of visits 10.9 (SD 6) | Recorded Delivery \n\nEdinburgh Postnatal Depression Scale (EPDS) | Average score5-7 (SD 6) | Psychosocial and psychological interventions for preventing postpartum depression (Cochrane review) \n\n\nEstimated sample size calculations for number of antenatal visits\nAverage number of antenatal visits | Average number in intervention arm | Difference | Power | Sample sizeper arm\n8.9 | 10.9 | 2 | 90% | 190\n\n9.4 | 10.9 | 1.5 | 90% | 337\n",
        "label": "low",
        "id": "task4_RLD_test_473"
    },
    {
        "paper_doi": "10.1186/1475-2875-7-237",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: An open label RCTFollow-up: Malaria film and Hb level on days 0, 1, 2, 3, 7, 14, and 28, plus QT-NASBA for detection of sub-microscopic gametocytaemiaAdverse event monitoring: Adverse events were recorded at each visit in the case record form. An adverse event defined as any unfavourable and unintended sign.\n\n\nParticipants: Number of participants: 146Inclusion criteria: Age 6 months to 12 years, axillary temp > 37.5 degC or history of fever, P. falciparum mono-infection 1000 to 200,000/uL, informed consentExclusion criteria: Severe malaria, any other underlying illness\n\n\nInterventions: 1. DHA-P, fixed dose combination, 20 mg/160 mg tablets (Sigma-Tau)4 to 7 kg  1/2 tablet once daily for 3 days7 to 13 kg 1 tablet once daily for 3 days13 to 24 kg 2 tablets once daily for 3 days24 to 35 kg 4 tablets once daily for 3 days2. Artemether-lumefantrine, fixed dose combination, 20/120 mg tablets (Novartis)5 to 14 kg 1 tablet twice daily for 3 days15 to 24 kg 2 tablets twice daily for 3 days25 to 34 kg 3 tablets twice daily for 3 daysAll doses supervised and given with a glass of milk.\n\n\nOutcomes: Recurrent parasitaemia at day 28, PCR-adjusted and PCR-unadjustedGametocyte prevalence during follow-upMean Hb at day 28Adverse eventsNot included in this review:Fever clearanceParasite clearance\n\n\nNotes: Country: KenyaSetting: Health centreTransmission: High transmissionResistance: Not reportedDates: Apr 2007 to Jul 2007Funding: The Knowledge and Innovation Fund, Koninklijk Instituut voor de Tropen/Royal Tropical Institute. DHA-P provided free of charge by Sigma-Tau\n\n",
        "objective": "To evaluate the effectiveness and safety of DHA\u2010P compared to other ACTs for treating uncomplicated P. falciparum malaria in adults and children.",
        "full_paper": "Background\nMany countries have implemented artemisinin-based combination therapy (ACT) for the first-line treatment of malaria.\nAlthough many studies have been performed on efficacy and tolerability of the combination arthemeter-lumefantrine (AL) or dihydroartemisinin-piperaquine (DP), less is known of the effect of these drugs on gametocyte development, which is an important issue in malaria control.\nMethods and results\nIn this two-arm randomized controlled trial, 146 children were treated with either AL or DP.\nBoth groups received directly observed therapy and were followed for 28 days after treatment.\nBlood samples were analysed with microscopy and NASBA.\nIn comparison with microscopy NASBA detected much more gametocyte positive individuals.\nMoreover, NASBA showed a significant difference in gametocyte clearance in favour of AL compared to DP.\nThe decline of parasitaemia was slower and persistence or development of gametocytes was significantly higher and longer at day 3, 7 and 14 in the DP group but after 28 days no difference could be observed between both treatment arms.\nConclusion\nAlthough practical considerations could favour the use of one drug over another, the effect on gametocytogenesis should also be taken into account and studied further using molecular tools like NASBA.\nThis also applies when a new drug is introduced.\nTrial registration\nCurrent controlled trials ISRCTN36463274\nBackground\nIn response to widespread resistance of Plasmodium falciparum parasites to the commonly used drugs chloroquine (CQ) and sulphadoxine-pyrimethamine (SP), many African countries recently adopted artemisinin-based combination therapy (ACT) as first-line treatment for uncomplicated malaria.\nThe combination artemether-lumefantrine (AL) proved to be highly effective and well-tolerated in several studies in Africa.\nDisadvantages of this drug combination are the twice-daily dosing and the fact that it should be administered with a fat-rich meal or at least a cup of soya milk.\nIn Uganda, in an area of intense malaria transmission, recurrence of parasitaemia within 28 days occurred in 29% of AL treated patients, in 8.9% adjusted by genotyping, indicating recrudescence.\nAnother ACT, dihydroartemisinin combined with piperaquine (DP), which was originally developed in China, is increasingly used in Southeast Asia.\nPiperaquine is an orally active bisquinoline with a half-life elimination time of 2.5\u20134 weeks.\nThe drug is structurally related to CQ, but still active against highly CQ-resistant P. falciparum strains.\nThis relatively inexpensive drug was well tolerated and highly effective in Southeast Asia as was the case in two studies in Africa.\nConsequently, both AL and DP are considered to be amongst the most promising artimisinin-based drugs.\nArtemisinin-based drugs also act on gametocytes and thus on transmission, at least in low transmission areas.\nIn high transmission areas of Africa, not much information is yet available on gametocytaemia after ACT treatments and on possible influence on transmission.\nIn a comparative study of AL and DP in Uganda, the appearance of gametocytes in those who did not have gametocytes at the start of treatment was lower from day 15 to day 42 of follow up in those treated with DP than in those treated with AL.\nA limitation of this study was the fact that gametocytaemia was assessed by microscopic examination only.\nIt has recently been shown that sub-microscopic gametocyte densities may significantly contribute to the transmission of malaria.\nAdequate assessment of gametocytaemia is important.\nQuantitative nucleic sequence based amplification technique (QT-NASBA) has been shown to be much more sensitive for the detection of gametocytes than microscopy.\nIn this study, the post-treatment prevalence of gametocytes in children in Mbita, western Kenya was assessed, after treatment with AL and DP.\nMicroscopy and QT-NASBA for the quantification of gametocytaemia were compared and effectiveness of both drugs regarding clinical symptoms, clearance of parasites and tolerability was assessed.\nMaterials and methods\nStudy site and population\nThe study was conducted in Mbita, western Kenya, at the shores of Lake Victoria during the high malaria transmission season of April-July 2007.\nMbita, is an area with highly variable transmission that depends on the local environmental circumstances that can support mosquito conditions.\nThe EIR is calculated to be 6 infectious bites per person per month.\nChildren (6 months-12 years of age) visiting the out-patient clinic of the health centre and diagnosed with uncomplicated malaria were included after informed consent from parents or guardians.\nInclusion criteria were: uncomplicated P. falciparum malaria with initial parasitaemia between 1,000 and < 200,000 parasites/\u03bcl blood, axillary temperature \u2265 37.5\u00b0C (measured with a digital thermometer) or a history of fever.\nChildren with severe malaria, mixed infection or other underlying illness were excluded from the study.\nIn total 146 children were recruited for the present study.\nEthical approval for this study was obtained from appropriate local authorities and The Kenya Medical Research Institute (KEMRI, Nairobi, Kenya) Ethical Steering Committee (SSC protocol No 948).\nThe trial was registered as an International Standard Randomized Controlled Trial at current controlled trials (ISRCTN36463274).\nStudy design and treatment\nFollowing diagnosis (at day 0), the patients were randomly allocated to one of the two treatment groups following a computer generated randomization list.\nOne group was assigned DP (Sigma-Tau, Italy) once per day for three days.\nOne tablet of the study drug contained 20 mg of dihydroartemisinin and 160 mg of piperaquine (paediatric formulation).\nTreatment was according to body weight as follows: children between 4\u20137 kg received half a tablet per dose, those between 7\u201313 kg 1 tablet, 13\u201324 kg 2 tablets per dose and children between 24\u201335 kg 4 tablets.\nThe other group was assigned to AL (Novartis Pharma, Switzerland).\nEach tablet contained 20 mg artemether and 120 mg lumefantrine.\nPatients received treatment according to bodyweight; i.e. children between 5\u201314 kg received one tablet per dose, those between 15\u201324 kg two tablets and those between 25\u201334 kg received three tablets per dose.\nDoses were given twice daily.\nAll treatments were given with a glass of milk under direct supervision at the clinic or, for the 2nd dose of AL, at home.\nOutcomes: efficacy\nEfficacy was assessed using the WHO in vivo test with a follow-up period of 28 days.\nAt enrollment (day 0) a full clinical examination was performed; information was recorded on a case record form.\nAt initial diagnosis (day 0) and during follow-up (day 1, 2, 3, 7, 14, and 28), finger prick blood samples were collected for microscopy, measurement of haemoglobin level and molecular analysis.\nHaemoglobin was measured with Hemocue 201+ analyser and cuvettes (HemoCue diagnostics B.V. Waarle, The Netherlands).\nResponse to treatment was measured and defined according to WHO guidelines.\nPatients showing complications or treatment failure were treated with appropriate supportive therapy.\nChildren developing danger signs or severe malaria on day 1 or 2 of the study were withdrawn from the study, referred to the hospital, and given alternative treatment.\nAdverse events were recorded on the case record forms.\nAn AE was defined as an unfavourable and unintended symptom, sign or disease.\nA serious adverse event (SAE) was defined as a symptom or sign that is temporally associated with the drugs administered to the patient that is life threatening or results in hospitalization, permanent and significant disability or death.\nSAE's were immediately reported to the ethical committee of KEMRI and the drug safety department of Sigma-Tau, Italy.\nOutcomes: parasite clearance and gametocyte dynamics\nParasite clearance and gametocyte dynamics were assessed microscopically as well as with quantitative nucleic acid sequence based amplification assay, QT-NASBA.\nMicroscopy\nGiemsa-stained thick and thin smears were prepared according to WHO guidelines.\nTwo independent experienced microscopists, who were blinded to the treatment and clinical status of the patient, examined the smears for the presence of parasites and identified the observed parasite species.\nParasitaemia was determined by counting the number of parasites against 200 leukocytes for the asexual stages (assuming that there are 8,000 leukocytes/\u03bcl blood).\nThe presence of gametocytes was examined against 500 leukocytes.\nQT-NASBA\nFinger prick blood (50 ul) for NASBA analysis was collected on Whatman 903 filter paper (Whatman international Ltd. Maidston, United Kingdom) and air-dried at room temperature.\nNucleic acid extraction was performed as previously described by Boom et al .\nReal-time 18S rRNA QT-NASBA was applied to study asexual parasite clearance below microscopical threshold.\nIn order to quantify the number of parasites in blood, a 10 fold serial dilution of 106 to 10 in vitro cultured parasites/ml was used as reference and processed and analysed with NASBA.\nFurthermore, to assess prevalence of gametocytes below the detection limit of microscopy, QT-NASBA targeting Pfs25 mRNA as described by Schneider et al  was used on blot spots collected during follow-up.\nGenotyping\nIn order to discriminate between re-infection (RI) and recrudescence (RE), merozoite surface protein 1 and 2 (msp1 and msp2) and glutamate rich protein (GLURP) genotyping was performed as described by Snounou on blood spots obtained at primary (day 0) and secondary infection (time point of re-occurrence).\nBlood spots were collected on Whatman 903 filter paper (Whatman international Ltd. Maidston, United Kingdom) and air-dried at room temperature for PCR analysis.\nDNA was isolated as described by Boom et al .\nMolecular analysis was performed at Royal Tropical Institute, Amsterdam and was done blinded from the treatment that was given to the patients.\nSample size and statistical analysis\nThe aim of the study was to compare gametocytaemia after AL and DP and to compare assessment of gametocytaemia by microscopical examination versus QT-NASBA All data were entered in excel and analysed with SPSS for windows (version 12.0).\nParasite densities were analysed after natural log-transformation.\nWhere appropriate, proportions were compared with the \u03c72-test and means were compared with the one-way ANOVA or Student t-test.\nA simplified trapezoid area under the curve (AUC) analysis using gametocyte data from days 0, 3, 7, 14 and 28, as a surrogate for the infectiousness of the participants in the different treatment groups, was performed.\nResults\nPatient recruitment\nIn total 1882 cases suspected of uncomplicated malaria were screened for eligibility into the study during an 8-week recruitment period in April and May 2007.\n1,736 children were excluded because they did not meet the inclusion criteria (Figure 1).\n146 patients fulfilling the inclusion criteria entered the study; 73 were randomly allocated to the DP arm and 73 to the AL arm.\nBoth study groups were comparable at baseline for their demographical and clinical characteristics and parasite densities (Table 1).\nOn completion of follow up (day 28) data of 134 patients (92%) were available for analysis.\nTwelve patients did not reach the study endpoint.\nSeven patients were lost during follow up, one was unable to take oral medication, one developed severe anaemia, one did not receive the proper drugs, one withdrew from the study and one patient died.\nTreatment outcome\nThere were no early treatment failures during the first three days of follow up.\nOnly one patient in the AL arm had a recurrent parasitaemia (43,880 parasites/\u03bcl) at day 28 of follow up.\nGenotyping analysis revealed that this patient had a reinfection with P. falciparum.\nAll other 133 patients who completed follow-up had an adequate clinical and parasitological response.\nAfter one day of treatment, over 90% of the patients had no microscopically detectable asexual parasites.\nIn the AL group no parasites could be detected with microscopy in any of the patients at day two.\nOne patient was still microscopically positive at day two in the DP group with 40 parasites/\u03bcl, but this patient was also microscopically negative at day 3.\nThe parasite reduction ratios at 48 hours reproduction cycle (parasite count on admission/parasite count at 48 hours) was 8.96 * 105 at 48 hours for the AL treatment and 2.06 * 104 at 48 hours for the DP treatment.\nNASBA was also applied to monitor parasite dynamics below sub-microscopical level.\nHumidity in some of the filter papers degraded the RNA in the blood spots of some of the samples.\nThis led to several extraction failures.\nIn order to have a clear picture of parasite dynamics only those series with a full range of follow-up samples, i.e. 56 DP and 54 AL treated patients, were analysed.\nBoth treatment arms showed a steep decline in parasitaemia from the day of enrollment (day 0) to day 1; 62% reduction after DP treatment and 89% reduction in the AL arm.\nAt day 2, the level of parasitaemia was reduced to 1.2% in the AL group and 2.75% in the DP group.\nHb convalescence, fever clearance and adverse events\nAt baseline Hb levels in both treatment groups were comparable (Table 1).\nAt day 28 all groups had a significant increase of Hb however no significant difference between the treatments on the Hb convalescence was found.\nFinal mean Hb levels were 7.15 mmol/l \u00b1 1.07 for the DP treatment group and 6.79 mmol/l \u00b1 1.24 for the AL group.\nA possible influence of anaemia on gametocyte carriage at enrollment was not observed in the present study (p > 0.05).\nFever clearance was defined as the time from receiving the assigned treatment to the time a normal body temperature was recorded (\u2264 37.5\u00b0C), in study cases who presented with fever.\nFever clearance was rapid in both study groups.\nOn day 1, 10 cases (13.7%) in the DP group presented with fever and six cases (8.2%) were observed in the AL group.\nIn the DP group fever was observed on day 2 in four cases (5.5%) and three cases (4.1%) on day 3.\nIn the AL group cases with fever also presented on day 2 (12 study subjects, 16.4%) and on day 3 (two individuals, 2.7%).\nFever was not observed during follow-up after day 7, with the exception of the child that presented with a P. falciparum reinfection on day 28 in the AL group and the child that developed broncho-pneumonia (case presented below).\nFurthermore the presence of fever at recruitment was no predictor for gametocyte carriage (p > 0.05)\nMost adverse events were mild, self limiting and consistent with symptoms of malaria.\nThere was no significant difference between the two study groups (Table 2).\nOne patient died.\nThe child (63 months) had been ill for two weeks prior to presentation at the clinic.\nPlasmodium falciparum infection with parasitaemia of 20,120 parasites/\u03bcl was diagnosed.\nThere was a fever (38.6\u00b0C), but there were no other complaints and no signs of severe anaemia (Hb: 6.4 mmol/L).\nOn day 3, there were no signs of illness.\nOn day 7, the child presented with fever (38.6\u00b0C), cough and complaints of anorexia.\nThere was no history of significant illness or allergies.\nAfter examination (microscopy was negative for malaria parasites), broncho-pneumonia was diagnosed and the child was treated with oral phenoxymethylpenicillin for five days and Paracetamol syrup.\nOn day 14, the child did not attend the follow-up visit, the parents reported that the child died a day before in a local health post.\nThe event was assessed as unrelated to the study drug.\nAutopsy was not performed.\nGametocyte dynamics\nThe presence of gametocytes in clinical samples was assessed by microscopy and NASBA and is presented in Table 3.\nAt the start of the study, three patients in the DP arm (4.5%) and six patients in the AL arm (9.0%) carried microscopically detectable gametocytes.\nMicroscopical follow-up of the presence of gametocytes during the whole study period revealed that in total 39 samples (distributed over 13 patients) in the DP arm and 18 samples (distributed over nine patients) in the AL arm carried gametocytes (not significantly different).\nIt was observed that the microscopical detection of gametocytes in blood slides during the study was subjected to fluctuations (for example a case positive on day 0, negative on day 1 and subsequently again positive at day 3), which is probably due to the fact that gametocytes circulate at low levels.\nHowever, on day 7, three patients in the DP group and 1 in the AL group showed gametocyte positive slides for the first time.\nOn day 28, in none of the cases gametocytes were observed by microscopy.\nThere was no difference between children older than 60 months and younger as regards carriage of gametocytes and density.\nNASBA analysis on 56 DP treated subjects and 54 AL treated subjects detected strikingly more gametocyte carriers at the start of the study compared to microscopy; i.e. 22 study subjects in the DP arm (39.3%) and 21 in the AL (38.9%) were harbouring gametocytes before.\nThe pfs25 NASBA revealed that 34 cases (60.7%) were gametocyte positive in the DP group on day 3, of which 13 cases were newly identified compared to day 0.\nIn contrast, a significantly lower number of patients (20; 37%) were gametocyte positive in the AL treatment group on day 3, of which six new cases.\nThis trend was also observed on day 7: the DP arm had 33 (58.9%) gametocyte positive samples, whereas the AL treated group had a significantly lower number of gametocyte positive samples (11, 20.3%).\nHowever on day 14 (AL: 12 positive [22.2%], six new; DP: 17 [30.4%], five new) and day 28 (AL: 5 positive [9.3%], no new cases; DP: 8 [14.3%], no new cases) of follow-up no significant difference in gametocyte carriage was observed between both treatment groups.\nThe AUC (day 0\u201328) of the two treatment groups was calculated to be 20.0 infectious persons/day for the DP treatment arm and 10.5 infectious persons/day for the AL treatment arm.\nStratifying the data for age under and above 60 months showed no difference in either of the groups.\nDiscussion\nSeveral studies have analysed the efficacy and tolerability of AL and DP and all show very good results.\nIn the present study, the two drugs showed to be similar with respect to effectiveness and tolerability compared to other studies.\nHowever, most of these studies have a follow up of 42 days which makes a direct comparison of the results difficult.\nNo adverse events other than those related to malaria itself were observed in the current study, which is in line with other reports.\nIn the present study, all children experienced haemoglobin convalescence without difference between the two treatment arms, in contrast to the study of Kamya et al, who found a greater increase in the DP treated patients.\nThe difference in follow-up time and the numbers of patients included in both studies may be responsible for this difference.\nFurther studies with comparable study length should be done to give an answer to these discrepancies.\nThe effects on gametocytaemia and possibly malaria transmission deserve further study.\nWhereas asexual parasites were cleared in three days after the initiation of the two treatment schedules, gametocytaemia appeared different when assessed by microscopy as well as with NASBA.\nGametocytes were present in low numbers throughout follow-up in both study groups.\nArtemisinin derivatives have in general a negative effect on gametocyte development and survival and thus influence malaria transmission, at least in low transmission areas.\nIn this study, the actual infectiousness of the remaining gametocyte populations in both treatment arms was not assessed; the presence of gametocytes does not necessarily mean that they actually contribute to transmission.\nSeveral studies have shown that gametocytes persist in a large population of previously infected and treated children.\nA large proportion of these carriers has a parasite load below microscopical detection limit, a load that can be detected with molecular assays like NASBA.\nPatients with submicroscopic parasite densities may still be infectious to mosquitoes and may contribute to transmission, as confirmed with membrane feeding experiments.\nStudies that include reduction of transmission as a component of efficacy of drugs, thus need to incorporate highly sensitive molecular assays to reliably assess gametocyte densities.\nThe present study showed a limited effect of DP on gametocyte development in comparison with AL when a sensitive tool like NASBA is used for gametocyte detection, which could limit the usefulness of DP to areas with low transmission but this finding should be further investigated in larger studies in different study sites with different transmission intensities.\nIt is not clear if plasma concentrations of dihydroartemisinin in the blood could play a role.\nDihyrodartemisinin is the major and the active metabolite of artemether.\nSo far, no studies have been performed that compare the plasma levels of dihydroartemisinin when given as such or after administration of artemether and subsequent metabolisation.\nThis should be further investigated together with effect on gametocytogenesis, which should incorporate a sensitive detection tool for gametocytes such as NASBA.\nThe effect that drugs can have on gametocyte clearance as measured with NASBA could have some implications for the introduction drugs and especially the introduction of new drugs.\nThis study showed that with sensitive detection tools a difference in parasite clearance can be observed but these results should be confirmed in larger studies and in other study areas with different malaria transmission intensities.\nTransmission intensity varies significantly in the different African countries and within a country high and low transmission areas can often be identified.\nMalaria endemic countries generally have a national malaria drug policy for the whole country.\nAlthough this is logical from a practical and logistical point of view, it may not be the best approach for effective malaria control.\nIt could, therefore, be more effective if a country develops specific drug policies to suit regional instead of national requirements.\nSchematic representation and flowchart of the study.\n\nBaseline characteristics of patients included in the study at the time of enrollment in the study\nCharacteristic | DHA-PQP (n = 73) | ALN (n = 73)\nSex ratio male:female | 33:40 | 40:33\nAge (in months), median (IQR) | 60 (44) | 52 (44)\nBody weight, mean kg (range) | 17.62 (6\u201337) | 17.32 (6\u201342)\nTemperature, mean \u00b0C \u00b1 SD | 38.1 \u00b1 0.99 | 37.8 \u00b1 0.73\nHaemoglobin mmol/L \u00b1 SD | 6.33 \u00b1 1.29 | 6.28 \u00b1 1.27\nParasites/\u03bcl geometric mean (range) as determined by microscopy | 12145 (1000\u201372640) | 13379 (1080\u201372000)\n\n\nSummary of adverse events recorded during the study\nAdverse event | DHA-PQP | ALN | P-value\nHeadache | 43 (58.9%) | 37 (50.7%) | 0.318\nAbdominal pain | 25 (32.4%) | 26 (35.6%) | 0.862\nWeakness | 19 (26.0%) | 30 (41.1%) | 0.035\nAnorexia | 8 (10.9%) | 10 (13.7%) | 0.439\nDiarrhea | 9 (12.3%) | 7 (9.6%) | 0.785\nCough | 16 (21.9%) | 17 (23.3%) | 0.843\nVomiting | 11 (15.1%) | 9 (12.3%) | 0.806\nPruritis | 4 (5.5%) | 3 (4.1%) | 0.698\n\n\nOccurrence of gametocytes as detected by microscopy or NASBA in the different study groups at the start of the study and during subsequent follow-up.\nGametocyte positive samples | Day 0 | Day 1 | Day 2 | Day 3 | Day 7 | Day 14 | Day 28\nDP group microscopy (n = 67) |  |  |  |  |  |  | \nTotal number of positive cases | 3 | 7 | 7 | 10 | 7 | 5 | 0\nNumber of new cases observed |  | 5 | 2 | 0 | 3 | 0 | 0\n\nDP group Nasba (n = 56) |  |  |  |  |  |  | \nTotal number of positive cases | 22 |  |  | 34a | 33a | 17 | 8\nNumber of new cases observed |  |  |  | 13 | 8 | 5 | 0\n\nAL group microscopy (n = 67) |  |  |  |  |  |  | \nTotal number of positive cases | 6 | 3 | 3 | 3 | 2 | 1 | 0\nNumber of new cases observed |  | 1 | 1 | 0 | 1 | 0 | 0\n\nAL group Nasba (n = 54) |  |  |  |  |  |  | \nTotal number of positive cases | 21 |  |  | 20 | 11 | 12 | 5\nNumber of new cases observed |  |  |  | 6 | 4 | 6 | 0\n\na Number of gametocyte carriers detected with NASBA is significantly higher in the DP treated group compared to the AL treated group.",
        "label": "low",
        "id": "task4_RLD_test_554"
    },
    {
        "paper_doi": "10.1111/iwj.12436",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallelEthics and informed consent: yesSample size estimate: noFollow-up period: unknownITT analysis: yes, number randomised: 20, number analysed: unclearFunding: unclear. MHB gave scientific presentations for KCI.Preregistration: no\n\n\nParticipants: Location: Nuremberg, Germany\nIntervention group: n = 10,control group: n = 10Mean age: intervention group = 52.3 (16.3),control group = 57.8 (15.2)\nInclusion criteria: patients with spinal fractures who were scheduled for internal fixation\nExclusion criteria: not reported\n\n\nInterventions: Aim/s: to evaluate the different aspects of wound healing in spinal fractures treated with internal fixationGroup 1 (NPWT) intervention: the iNPWT group was treated with a PICO system (Smith & Nephew, UK). The PICO system was left on the wound for 5 days including the day of surgery. In addition to daily clinical examination, all wounds/seroma were analysed by ultrasonography on day 5 and day 10 after surgery.Group 2 (control) intervention: standard department wound dressing consisting of dry wound coverage (compresses attached to the skin) was used.\nStudy date/s: not reported\n\n\nOutcomes: SeromaValidity of measure/s: ultrasound was used as a standardised imaging modality to detect seromas in the wound area.Time points: day 5 and day 10 after surgery\n\n\nNotes: Investigator contacted for additional details\n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "Abstract\nTo evaluate the clinical use and economic aspects of negative pressure wound therapy (NPWT) after dorsal stabilisation of spinal fractures.\nThis study is a prospective randomised evaluation of NPWT in patients with large surgical wounds after surgical stabilisation of spinal fractures by internal fixation.\nPatients were randomised to either standard wound dressing treatment (group A) or NPWT (group B).\nThe wound area was examined by ultrasound to measure seroma volumes in both groups on the 5th and 10th day after surgery.\nFurthermore, data on economic aspects such as nursing time for wound care and material used for wound dressing were evaluated.\nA total of 20 patients (10 in each group) were enrolled.\nThroughout the whole study, mean seroma volume was significantly higher in group A than that in group B (day 5: 1\u00b79 ml versus 0 ml; P = 0\u00b70007; day 10: 1\u00b76 ml versus 0\u00b75 ml; P <0\u00b7024).\nFurthermore, patients of group A required more wound care time (group A: 31 \u00b1 10 minutes; group B 13\u00b78 \u00b1 6 minutes; P = 0\u00b70005) and more number of compresses (total number; group A 35 \u00b1 15; group B 11 \u00b1 3; P = 0\u00b70376).\nNPWT reduced the development of postoperative seroma, reduced nursing time and reduced material required for wound care.\nIntroduction\nIn recent years, negative pressure wound therapy (NPWT) became a widely used therapy for many different indications in the treatment of wounds 1, 2, 3, 4.\nRecently, NPWT has also been used in the treatment of closed surgical wounds.\nThe incisional NPWT (iNPWT) has shown beneficial effects when administered after severe trauma 5, 6, 7, 8.\nThe indications and the evidence for the efficacy of iNPWT have increased recently 9, 10, 11.\nHowever, only a few prospective randomised studies have been published on those indications mostly dealing with hip arthroplasty 11, 12.\nThe mode of action of iNPWT is still not completely understood.\nThis is thought to be mediated by increased oxygen delivery to the tissue as result of enhanced tissue perfusion and promotion of angiogenesis 5, 9, 13.\nThe purpose of this study was to evaluate the different aspects of wound healing in spinal fractures treated by internal fixation.\nWe compared a standard wound dressing group with an iNPWT group in the context of postoperative seroma formation in the wound area, the total time of secretion, the total time needed for the wound care (dressing changes) and the material needed for dressing changes.\nMaterials and methods\nA total of 20 patients with spinal fractures were scheduled for internal fixation.\nThey were randomised into two groups.\nGroup A (10 patients) received the standard wound dressing of our department, consisting of a dry wound coverage (compresses attached to the skin).\nGroup B (10 patients) was treated with iNPWT over the sutured wound area.\nThe surgical intervention was identical in both the groups.\nAn open reduction technique with internal fixation system was performed for all the patients, which was obtained from the same manufacturer (Synthes, West Chester, PA).\nAll patients received two Redon\u00ae drains, one on each side of the spinal column subcutaneously.\nPostoperative physiotherapy and mobilisation were also identical for both the groups.\nThe iNPWT group (group B) was treated with a PICO\u2122 system (Smith & Nephew plc, London, UK).\nThe PICO\u2122 system was left on the wound for 5 days including the day of surgery.\nIn addition to daily clinical examination, all wounds/seroma were analysed by ultrasonography on the 5th and 10th day after surgery.\nBefore surgery, plasmatic coagulation was assessed in all patients using the Quick prothrombin time test.\nPostoperatively, the immediate amount of wound secretion in the Redon drain canisters was quantified.\nIn addition, the length of the incision was measured.\nThe duration of secretion from the wounds was monitored and the total number of dressing changes and the time required to perform the dressing changes were also assessed.\nThe material used for dressing changes was also quantified (compresses and gloves).\nStatistical significance was calculated with the Prism v6.0 GraphPad Software, Inc. (La Jolla, CA).\nFor Gaussian distributed data, the Student's t\u2010test was used.\nFor non\u2010Gaussian distributed data, the Mann\u2013Whitney test was used.\nInformed consent was obtained from each patient.\nThe study was approved by the local ethics committee (Re\u2010No.139_12 B) and conforms to the principles of the Declaration of Helsinki.\nResults\nIn this study, 10 patients (mean age 57\u00b780 \u00b1 15\u00b724 years) were randomised to group A and 10 patients (mean 52\u00b730 \u00b1 16\u00b732 years) to group B. Both groups displayed normal coagulation times according to the Quick prothrombin time test (group A: 94\u00b720 \u00b1 9\u00b752%; group B: 96\u00b700 \u00b1 7\u00b780%; P = 0\u00b773).\nThere was no significant difference in the postoperative wound size between both groups (group A: 17\u00b725 \u00b1 5\u00b770 cm; group B: 14\u00b760 \u00b1 4\u00b738 cm; P = 0\u00b726).\nFurthermore, both groups displayed almost equal volumes of wound secretion in the Redon\u00ae drain canisters after 2 days (group A: 621\u00b75 \u00b1 286\u00b75 ml; group B: 454\u00b70 \u00b1 229\u00b76 ml; P = 0\u00b716).\nThe seroma volume underneath the surgical wound was significantly lower at day 5 and day 10 in the iNPWT group (day 5: group A: 1\u00b79 \u00b1 2\u00b77 ml versus group B: 0 \u00b1 0 ml (P = 0\u00b70007); day 10: group A: 1\u00b76 \u00b1 2\u00b76 ml versus group B: 0\u00b75 \u00b1 1\u00b70 ml; P = 0\u00b7024).\nThe patients treated with iNPWT required fewer dressing changes: 48 dressing changes in group B patients, equating to 4\u00b78 per patient and 79 dressing changes in group A patients equating to 7\u00b79 per patient (Figure 1).\nGroup B patients also had lesser number of days of wound secretion (Figure 2) and required lesser time for wound care (Figure 3) and lesser material for dressing changes (Figures 4 and 5).\nDiscussion\nSince the development of NPWT, the indications for the use of NPWT have been mainly acute and chronic wounds 1, 2, 3, 14, 15, 16, 17, 18; however, the indications have increased over time 2, 19.\nNPWT exerts a positive effect on wound healing resulting in a reduction in wound healing complications.\nFurthermore, NPWT treatment excels because of its ease of application and a low risk of side effects 20, 21.\nRecent studies using NPWT showed a reduction of seromas in wounds after hip surgery.\nThe beneficial effect of NPWT was found after elective total hip arthoplasty and after arthroplasty of femoral neck fractures 11, 12.\nA recently published review confirmed the reduction of wound complications in high\u2010risk wounds 9.\nIn our study, we evaluated for the first time the possible effects of iNPWT in spinal fractures treated with open reduction and internal fixation.\nBesides the anatomical area of application, it is the first prospective randomised study in orthopaedics using the PICO\u2122 system (Figure 6) (Smith & Nephew plc).\nThis study provides compelling evidence that iNPWT may be useful to treat large surgical incision wounds after surgical treatment of spinal fractures.\nTo the best of our knowledge, significant reduction of wound complications in fractures of the spine treated with open reduction and internal fixation by using iNPWT has not been previously reported.\nIn addition, iNPWT treatment reduced the time and dressing material needed for postoperative wound care.\nThe present study used ultrasound as a standardised imaging modality to detect seromas in the wound area.\nThe high sensitivity of this imaging modality allowed for the detection of seroma volumes directly underneath the surgical incision.\nThis method was previously described for the evaluation of iNPWT after surgical interventions of the hip and femoral neck fracture 11, 12.\nIn line with these studies, we found a significant reduction of wound draining days in the iNPWT group compared with that in the control group.\nIn addition to reduced wound secretion, Stannard et al. showed that iNPTW treatment reduced haematoma size in patients suffering from high\u2010energy trauma injuries 5, 22.\nGiven the fact that haemotomas might favour as nutrient\u2010rich environments for bacterial replication and that persisting wound drainage might facilitate the entrance of bacteria into the wounds, the reduction of wound secretion and the haematoma size might reduce the risk of a wound infection.\nHaematomas are thought to serve as rich nutrient sources for infection 23.\nIn our opinion, a prolonged secretion is an important risk factor of early postoperative infections as well.\nHowever, to the best of our knowledge, we are not aware of a study that has addressed this issue.\nHence, further investigations into these mechanisms are warranted.\nFurthermore, it is not fully understood how iNPWT leads to a reduced seroma and haematoma formation in the wounded tissue.\nHorch et al. suggested that NPWT results in a significantly increased tissue perfusion and oxygenation 13.\nIn addition to altering tissue perfusion, treatment with iNPWT might reduce wound edge tension and thereby promote healing 24.\nOur results in this study are consistent with the findings of other studies regarding the reduction of wound healing complications after the treatment with iNPWT.\nWe and others have demonstrated that wound treatment with iNPWT after elective total hip arthroplasty for the treatment of osteoarthritis of the hip and the treatment with iNPWT after surgical treatment of femoral neck fractures reduced the wound healing complications 11, 12.\nIn addition, patients who suffered from severe soft tissue damage after trauma displayed a faster recovery when the iNPWT device was placed on the wounded tissue at early time points 13, 25.\nIn our study, we attached the iNPWT device immediately after surgical wound closure to the skin over the wound.\nIn a previous study, we observed that the time dedicated to the wound care and the consumption of wound care material were significantly shorter than those in the control group.\nIn agreement with our previous study 11, we found that iNPWT treatment reduced the time needed for wound care and the consumption of wound care material.\nThe 48 dressing changes in the iNPWT group also included the removal of the redon drains and the removal of the device.\nTo avoid a bias, these dressing changes were also documented; the number would have been even smaller if this study\u2010related dressing changes were not taken into account.\nLimitations of the present study are the relatively small number of enrolled patients and the use of an iNPWT device from a single manufacturer.\nThe difference between the different manufacturers seems to be relatively small and the principles of NPWT seem to be transferable between the different manufacturers.\nHowever, this has to be confirmed by comparative studies.\nConclusion\nIn summary, our results support the use of iNPWT after spinal surgery.\nApart from its economic benefits, iNPWT promotes wound healing and might prevent wound infections, which are the dreaded complications of spine surgery.\nDistribution of number of dressing changes (P < 0\u00b70001).\nDistribution of number of days of wound secretion.\nDistribution of wound care time.\nDistribution of number of used gloves for dressing changes.\nDistribution of number of used compresses for dressing changes.\nApplication of a PICO system to the wound.",
        "label": "unclear",
        "id": "task4_RLD_test_709"
    },
    {
        "paper_doi": "10.1371/journal.pone.0119772",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Cluster-RCT in Bulgan, Mongolia.\n\n\nParticipants: Sample size: 501 women randomised.Clusters: the unit of randomisation was the Soum and bag, small geographic areas in Mongolia. Each Soum has a healthcare facility where women must register their newborn. 18 geographic areas were randomised, after selection for administrative convenience and to avoid contamination.Individuals: pregnant women living in Bulgan, Mongolia.\n\n\nInterventions: Target: community.Arm 1: distribution of maternal and child health handbooks during pregnancy. The MCH handbook logged maternal health and personal information, pregnancy, delivery and postpartum health and weight, dental health, parenting classes, child developmental milestones from 0-6 years, immunisation records and height and weight charts for children.Arm 2: women received standard care.\n\n\nOutcomes: Trial primary outcome: number of antenatal visits; proportion of women attending 6 or more antenatal visits. (The national standard for ANC in Mongolia is 6 visits.)Review outcomes reported:Primary: ANC coverage of at least 4 visits, maternal mortalitySecondary: maternal outcomes: morbidity during pregnancy, mode of delivery, breastfeeding initiation, maternal depression and health (EPDS and GHQ). Infant outcomes: birthweight, Apgar score, NICU admission, neonatal mortality at discharge. Maternal healthy behaviours.Follow-up: data collection at 1 month postpartum.\n\n\nNotes: Funders: this study was funded by the National Center for Global Health and Medicine, Tokyo, Japan.Significant group differences noted for distances travelled to nearest health centre (greater in the intervention group) and for wealth index (the control group was poorer). The authors report that travel time did not function as an effect modifier; however, women from a higher socioeconomic background attended more ANC visits.Trial authors provided unpublished outcome data upon request. The trial statistician (HN) calculated ORs and 95% confidence intervals using the generalised estimating equations (GEE) method to adjust for cluster design and baseline differences, including wealth\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Objective\nTo assess the effectiveness of the Maternal and Child Health (MCH) handbook in Mongolia to increase antenatal clinic attendance, and to enhance health-seeking behaviors and other health outcomes.\nMethods\nA cluster randomized trial was conducted using the translated MCH handbook in Bulgan, Mongolia to assess its effectiveness in promoting antenatal care attendance.\nPregnant women were recruited from 18 randomly allocated districts using shuffled, sealed envelopes.\nThe handbook was implemented immediately for women at their first antenatal visit in the intervention group, and nine months later in the control group.\nThe primary outcome was the number of antenatal care visits of all women residing in the selected districts.\nCluster effects were adjusted for using generalized estimation equation.\nMasking was not possible among care providers, pregnant women and assessors.\nFindings\nNine districts were allocated to the intervention group and the remainder to the control group.\nThe intervention group (253 women) attended antenatal clinics on average 6\u20229 times, while the control group (248 women) attended 6\u20222 times.\nSocioeconomic status affected the frequency of clinic attendance: women of higher socioeconomic status visited antenatal clinics more often.\nPregnancy complications were more likely to be detected among women using the handbook.\nConclusion\nThe MCH handbook promotes continuous care and showed an increase in antenatal visits among the intervention group.\nThe intervention will help to identify maternal morbidities during pregnancy and promote health-seeking behaviors.\nTrial Registration\nUMIN Clinical Trial Registry UMIN000001748\nIntroduction\nMaternal and child health continues to present a significant public health challenge in Mongolia.\nDespite a marked improvement in the maternal and neonatal mortality ratios over the past 20 years, with 89\u00b76 per 100,000 births in 2007 and 14 per 1,000 births during 2001\u20132003, respectively, as well as a decline in the mortality of older children, the quality of antenatal care is still low and complications during pregnancy remain a significant hurdle for improving maternal health in Mongolia.\nEffective interventions to enhance maternal and child health outcomes are crucial to address these challenges and to maintain the achievement of health-related Millennium Developmental Goals (MDGs) 4 and 5.\nAn ongoing challenge for researchers and health professionals is how to deliver effective interventions to reduce maternal and neonatal mortality in resource-limited settings.\nEffective maternal health interventions should aim to encourage health-seeking behaviors among pregnant women and increase their maternal knowledge.\nAs the role of health workers is to promote healthcare-seeking behaviors and initiate preventive action, a health record book such as Japan\u2019s Maternal and Child Health (MCH) handbook could be used as an effective tool by community healthcare workers and professional hospital staff to enhance client\u2013provider communication about health, raise health awareness, and identify complications earlier in the pregnancy.\nThe purpose of introducing the handbook to Mongolia, which was proposed by the Mongolian Ministry of Health, was to increase antenatal visits and enhance client-provider communication during pregnancy to improve long-term health outcomes for mother and child.\nThe handbook was first considered by the Mongolian government as a key intervention in maternal and child health in 2007, and our study initiated the national adoption of the MCH handbook in Mongolia in 2010.\nRegarded as Japan\u2019s flagship intervention in the context of health aid, the handbook has been adopted in other countries, such as Indonesia and Bangladesh, and previous studies have evaluated its impact on perinatal health.\nHowever, a high-quality study that assesses the effectiveness of the handbook to facilitate long-term information-sharing has not previously been undertaken.\nThe World Health Organization (WHO) emphasises the importance of effective interventions that focus on delivering a continuum of care.\nThe MCH handbook facilitates continuum of care throughout pregnancy, delivery and postpartum as well as the child\u2019s infancy using the handbook\u2019s continuous record of basic educational information that encompasses antenatal care and the milestones of child development from the ages of 0\u20136 years.\nWomen use the handbook by filling out relevant sections with their personal, maternal and child health information, and bringing the handbook with them to all antenatal and postnatal appointments.\nAt the appointment, the midwife and/or doctor then check the relevant section of the handbook pertaining to the woman\u2019s stage of pregnancy or the child\u2019s stage of development, and record in it results of tests, such as protein in the urine during pregnancy, or other notes.\nThe handbook also contains information on MCH care and serves as a valuable communication and educational tool between pregnant women and healthcare professionals, through which women can raise specific health concerns and healthcare professionals can convey important health messages and guidance at point of care.][\nIn this study, we aim to measure improved health-seeking behavior by increased antenatal clinic attendance in the Mongolian province of Bulgan.\nThe effectiveness of the intervention will be investigated through a cluster randomized control trial evaluating antenatal attendance, maternal physical and mental health, neonatal health and healthy behaviors.\nImplementation of the MCH handbook\u2014a communication tool between women and healthcare professionals\u2014can only be conducted at cluster level, and therefore a cluster-randomized trial was employed.\nMethods\nStudy design\nA cluster-randomized controlled study was conducted from 1 May 2009 until 1 September 2010 among pregnant women and their infants who lived in Bulgan, Mongolia.\nThe allocation ratio was 9/9 = 1.00 and the unit of randomisation in this study was the soum\u2014a small administrative unit in Mongolia\u2014and the bag, which is a subdivision of a soum.\nParticipants/ Study population\nEligible participants included pregnant women living in the Bulgan province of Mongolia.\nHealth centres in Bulgan are located in each soum and all women must register their newborn infants at their local health centre, regardless of the infant\u2019s birthplace.\nData confidentiality was strictly maintained throughout all steps of this study.\nRandomisation and masking\nSoums and bags were selected for administrative convenience and to avoid contamination.\nBulgan province is comprised of 17 soums and 4 bags and they differ in size, health outcomes, and available healthcare facilities.\nOf the combined soums and bags (21), 18 units (16 soums and 2 bags) were selected for inclusion in this study, and randomized in equal number between intervention and control group.\nThree units were excluded because one soum was the subject of a pilot study, and two bags were included in another health promotion project.\nRandomisation was conducted using shuffled, sealed envelopes, and an envelope was selected by each soum representative.\nSince the unit of randomisation is a soum and the intervention is visible, the intervention and outcomes could not be masked.\nWritten informed consent was sought from all women for permission to use the collected data in the study.\nInterventions\nThe Mongolian edition of the handbook was translated into Mongolian from the original Japanese version.\nThe MCH handbook contained a log for recording information on maternal health and personal information, course of pregnancy, delivery and postpartum health, weight during and after pregnancy, dental health, parenting classes, child development milestones from the ages of 0\u20136 years, immunization and illnesses, and height and weight charts for children.\nThe handbook was used as the intervention at both the cluster and individual participant level.\nThe handbooks were implemented at the beginning of the study observational period, and after a delay of seven months in the control group.\nOutcomes\nThe primary outcome was the number of antenatal care visits and the proportion of women who made six antenatal care visits or more.\nIn Mongolia, the national standard of antenatal care visits is a minimum of six.\nHealthcare professionals working in each cluster recorded each antenatal visit for their soum.\nData was collected for all participants at one month postpartum.\nSecondary outcomes included clinical outcomes (mortality and morbidities) of women and their infants, as well as healthy behaviors of women and their families.\nCharacteristics and other outcomes for mothers and their infants were collected 28 days after childbirth via self-reported questionnaires and interviews conducted by trained data collectors.\nThe data collectors visited the family clinic or regional hospital as well as the household to undertake routine check-ups of mothers and infants using a questionnaire.\nAll mortality and morbidity ratios are derived from routinely collected national statistics using the ICD-10 classification system.\nStatistical analyses\nThe primary analyses followed the intention-to-treat principle and compared the proportion of women who visited health centres for antenatal check-ups and their number of visits between the intervention and control groups.\nIn the analysis of a cluster-randomized trial, correlations of the outcomes of participants in the same cluster should be adequately adjusted.\nTo do this, the generalized estimating equations (GEE) method was adopted to estimate mean difference, risk ratio and risk difference as a measure of the effect, and to calculate their 95% confidence intervals (CI).\nMultivariate GEE analyses was performed to adjust for possible effects of baseline variables.\nTo quantify household wealth status, principle component analysis was used according to the procedure outlined in the Demographic and Health Survey (DHS) guidelines.\nThe whole population sample in the Bulgan province of Mongolia was used to create the wealth index.\nThe sample size was determined to detect one mean difference in antenatal care visits between the two groups with a two-sided alpha level of 0.05 and 80% power.\nAssuming 0.01 intra-cluster correlation, it was estimated that approximately 500 women were required in total.\nAll statistical analyses were conducted with SAS version 9.2 (SAS Institute, Cary, NC, USA).\nThis clinical trial is registered at the UMIN Clinical Trial Registry (UMIN000001748).\nBoth the protocol and CONSORT checklist of the present trial are presented as S1 and S2 Documents.\nRole of the funding source\nThis study was supported by the National Center for Global Health and Medicine, Tokyo, Japan.\nThe funding source did not affect the conduct, analyses or results of the study in any way.\nResults\nBaseline characteristics of the soums\nThis study had nine clusters in the intervention group and the total intervention population was 253 women with an average of 28\u00b70 people in a cluster.\nThe intervention was implemented between May 2009 and January 2010, and data was collected between February 2010 and August 2010.\nOf the whole intervention group, only 210 participants received the intervention.\nThere were nine clusters in the control group and the total number of participants for this group was 248 women.\nThere was no difference in the size of the cluster between the two groups.\nFig. 1 shows the selection process in detail.\nBaseline characteristics of women and infants\nAll baseline characteristics were similar in both the intervention and control groups, presented in Table 1.\nIn this study, 32\u00b74% of participants from the intervention group and 31\u00b70% of the control group were experiencing their first pregnancy; 94\u00b71% of the intervention group and 95\u00b72% of the control group were married and the mean age of both groups was 27 years of age.\nFrom the intervention group, 9\u00b75% of participants were educated to elementary school level compared with 10\u00b75% of the control group.\nStatistically significant differences in travel time were observed between women\u2019s homes and the nearest health centres, (p = 0\u00b7008), and in the wealth index (p = 0\u00b7001) of the intervention group compared to the control group.\nPrimary outcome\nWomen in the intervention group attended antenatal clinics an average of 6\u00b79 times, while women from the control group attended 6\u00b72 times, as shown in Table 2.\nIn the primary GEE analysis, there is no significant difference between the two groups in the number of antenatal care visits and the proportion of women who have visited more than 6 times.\nThe travel time to antenatal clinics did not significantly affect the association between the intervention and the primary outcome.\nHowever, socioeconomic status was found to influence the frequency of clinic attendance: women of a higher socioeconomic status visited antenatal clinics more often than women from a lower socioeconomic background.\nSocioeconomic status acted as a statistically significant effect modification on outcomes by the multivariate GEE analyses.\nTherefore the analysis of primary outcomes was stratified by socioeconomic-status quintile.\nResults of the GEE analysis of primary outcomes stratified by wealth index are presented in Figs. 2, 3 and 4.\nWomen\u2019s health\nComplications in maternal health were more likely to be identified, with maternal morbidity during pregnancy at 12\u00b73% in the intervention group compared with 5\u00b77% in the control group.\nThis difference was statistically significant (p-value 0\u00b701).\nNo evidence of difference was observed in women who scored both higher than 12 points in EPDS (RR 0.99 [0.94\u20131.04], p = 0.56) and higher than 4 points in GHQ (RR 1.01 [0.99\u20131.03], p = 0.41).\nInfant health\nA higher rate of early breastfeeding initiation amongst the intervention group was a significant neonatal health outcome.\nIn the intervention group, 94\u00b71% of infants initiated breastfeeding within one hour of childbirth compared to 87\u00b75% of infants in the control group.\nThis difference, though notable, was not statistically significant.\nHealthy behaviors\nAn increase in healthy behaviors was observed amongst the intervention group.\nThe majority of women stopped drinking alcohol during pregnancy: only 7\u00b79% of women from the intervention group continued to drink alcohol when pregnant compared with 14\u00b71% in the control group.\nIn the intervention group, a statistically significant reduction in smoking was found among women\u2019s family members living in the same household, with 51\u00b70% of family members continuing to smoke during women\u2019s pregnancies compared with 60\u00b79% living with control group participants.\nDiscussion\nOur findings show that pregnant women who used the MCH handbook increased their number of antenatal visits from the national requirement of six visits to a mean of 6.9 visits, compared to a mean of 6\u00b72 visits in the control group.\nSocioeconomic background was also found to play a significant role in clinic attendance for both groups.\nAfter adjusting for confounders in the GEE model, the intervention effect was statistically significant, but only among the wealthy.\nParticipants in the wealthiest two quintiles were more likely to attend antenatal clinics more than six times.\nComplications in maternal health were more likely to be detected among pregnant women who used the handbook.\nHealthy behaviors were also adopted by partners and other family members of pregnant women in the intervention group: the majority of women did not drink alcohol (7\u00b79% in the intervention group compared with 14\u00b71% in the control group), and approximately half of family members stopped smoking at home, thereby reducing the harm of passive smoking for expectant mothers.\nTo our knowledge, this study is the first to assess the MCH handbook using a cluster-randomized design in Mongolia.\nThis is also the first cluster-randomized controlled trial to assess a Japanese health aid intervention.\nThe MCH handbook provided pregnant women with a useful educational aid that promotes healthcare-seeking behaviors, fosters continuity of care and enhances communication between pregnant women and their healthcare providers.\nThe intervention raised women\u2019s awareness of maternal and child health concerns and prompted them to seek out healthcare, as illustrated by an increase in antenatal visits.\nNot only does the handbook instigate the delivery of key health messages from healthcare providers to pregnant women during antenatal visits, but also from pregnant women to their families at home.\nThis study used a randomized cluster design and population-based data collection so its results may be more representative of the effectiveness of the MCH handbook in the community; however, several limitations are present.\nMasking was not possible among care providers, pregnant women and assessors.\nRecall bias likely exists in the analysis, because data collection was performed at one month after birth.\nThe unbalanced distribution of socioeconomic status acts as an effect modifier.\nThe per-protocol analysis showed a significant increase in women\u2019s clinic attendance, and therefore it is likely that women of a lower socioeconomic status did not receive the handbook.\nA systematic review of a similar intervention highlights the potential benefits of giving women their own health record to use during pregnancy.\nThe review included three randomized trials.\nThough none of the trials assessed the rate of antenatal care visits as an outcome, the intervention resulted in favourable outcomes such as enhancing a mother\u2019s control over her health, and satisfaction with the care provided.\nA cross-sectional study conducted by Osaki et al showed the MCH handbook increased utilisation of health services and deliveries with trained personnel.\nAlthough the rate of antenatal visits is not reported in Osaki et al\u2019s study, the findings of our study are compatible with this and other studies.\nA significant outcome of this study was an increase in the proportion of women who attended antenatal care visits among pregnant women who used the MCH handbook.\nTravel time did not function as an effect modifier; however, socioeconomic background was particularly relevant, with women from a higher socioeconomic background visiting antenatal clinics more often than those of a lower socioeconomic background.\nThe handbook also facilitated the identification of maternal morbidities during pregnancy and minimized passive smoking in the households of intervention group participants.\nIn response to the study\u2019s main findings, the MCH handbook was implemented as part of the national health policy in Mongolia in 2010 soon after the trial was finished, and the results support the policy.\nHowever, policies to reach women of lower socioeconomic status are yet to be developed and more research is required to address this issue.\nOur study showed the effectiveness of the MCH handbook to promote long-term information sharing through an increase in antenatal clinic attendance among women who used the handbook.\nThe intervention promotes better communication between women and healthcare specialists and acts as a reference point for women to raise particular concerns and questions about their own health at antenatal clinics, while at the same time giving healthcare workers the opportunity to deliver important health messages.\nThe handbook\u2019s role in enhancing long-term information sharing can make an important contribution to maintaining MDGs 4 and 5.\nFurther interventions are also necessary to specifically target pregnant women from a lower socioeconomic background in outreach efforts that aim to increase antenatal clinic attendance.\nFuture research should also focus on the effectiveness of the handbook in other provinces within Mongolia as well as other low- to middle-income countries, where the handbook can be used as an effective tool in maternal health education to further promote maternal health awareness and healthy behaviors, enable early interventions, reduce adverse birth outcomes in developing settings and sustain the achievement of MDGs 4 and 5.\nUse of the latest information technology, such as a smartphone application of the MCH handbook to facilitate use of the intervention, may also provide a valuable opportunity to enhance accessibility of the handbook, and would benefit from further research.\nFlow diagram of the study population.\nPrimary outcome: Mean difference by wealth index.The y-axis shows the mean difference with confidence intervals of the number of antenatal care visits between the intervention and control groups. The x-axis shows the wealth index quintile.\nPrimary outcome: Risk ratio by wealth index.The y-axis shows the risk ratio with confidence intervals of the number of women who made six antenatal care visits during their pregnancy in the intervention and control groups. The x-axis shows the wealth index quintile.\nPrimary outcome: Risk difference by wealth index.The y-axis shows the risk differences with confidence intervals of the number of women who made six antenatal care visits during their pregnancy in the intervention and control groups. The x-axis shows the wealth index quintile. Note: 1st quintile represents the highest wealth index, and the 5th represents the lowest.\n\nBaseline characteristics of women and infants.\n |  | Intervention | Control | p-value\nN = 253 | N = 248\nFirst pregnancy | N (%) | 82 (32.41) | 77 (31.05) | 0.743\nNumber of pregnancies | Mean (SD) | 2.49 (1.37) | 2.32 (1.24) | 0.154\nmissing | 0 | 1\nOutcome of previous pregnancies |  |  |  | \nLive birth | Mean (SD) | 1.42 (1.35) | 1.29 (1.19) | 0.229\nAbortion | 0.11 (0.41) | 0.09 (0.43) | 0.556\nMiscarriage | 0.11 (0.39) | 0.07 (0.32) | 0.233\nAdoption | 0.00 (0.00) | 0.02 (0.16) | 0.099\nPre-pregnancy weight | Mean (SD) | 61.10 (9.02) | 60.15 (8.76) | 0.237\nmissing | 2 | 2\nWeight at first antenatal care visit | Mean (SD) | 63.13 (9.20) | 61.88 (9.19) | 0.132\nmissing | 1 | 6\nTravel time from home to antenatal care clinic | Median | 40 | 40 | 0.008\n(25\u201375%) | (20\u201399) | (20\u201360)\n(min., max.) | (4, 1440) | (2, 180)\nMarital status Married | N (%) | 238 (94.1) | 236 (95.2) | 0.590\nMean maternal age (SD) | Mean (SD) | 27.3 (6.13) | 27.7 (5.67) | 0.390\nmissing | 1 | 3\nMaternal educational attainment (up to elementary level education) | N (%) | 24 (9.49) | 26 (10.48) | 0.947\nNumber of family members in the household | Mean (SD) | 4.332 (1.23) | 4.185 (1.196) | 0.177\nWealth index | Mean (SD) | 0.448 (2.194) | -0.225 (2.356) | 0.001\n\n\nPrimary outcome and outcomes for mothers, infants and healthy behaviors.\n |  | Intervention N = 253 | Control N = 248 | Effect of measure [MD: Mean difference, RR: Risk ratio, RD: Risk difference] (95%CI), p: p-value, *GEE analysis\nPrimary outcome\nAntenatal care visits | Mean (SD) | 6.615 (1.525) | 6.407 (1.765) | [MD] 0.208 (\u20130.710\u20131.125) (p = 0.66)*\nAntenatal care visits | \u2265 6 N(%) | 206 (81.7%) | 175 (70.6%) | [RR] 1.158 (0.876\u20131.532), p = 0.30*, [RD] 11.2% (-9.9%-32.3%), p = 0.30*\nWomen\u2019s outcomes\nComplications identified during pregnancy | N (%) | 31 (12.25) | 14 (5.65) | P = 0.012\nmissing | 1 | 1\nMultiple pregnancies | N (%) | 6 (2.37) | 4 (1.61) | \nGestational age | Mean (SD) | 38.95 (1.25) | 39.06 (1.18) | \nMedian (25\u201375%) | 39 (38\u201340) | 39 (39\u201340)\nmissing | 7 | 14\nCephalic fetal presentation | N (%) | 246 (97.23) | 236 (95.16) | \nSpontaneous vaginal deliveries | N (%) | 202 (79.84) | 212 (85.48) | \nEPDS: Postnatal depression Over cut-off 12 points | N(%) | 15 (5.93) | 11 (4.44) | RR 0.99 (0.94\u20131.04), p = 0.560, RD\u20140.014 (\u20130.062\u20130.034), p = 0.561\nGHQ: General Health Questionnaire Over cut-off 4 | N (%) | 3 (1.2%) | 5 (2.0%) | RR 1.01 (0.99\u20131.03), p = 0.412, RD 0.0085 (\u20130.012\u20130.029), p = 0.411\nInfant outcomes\nApgar score 5 minutes | Mean(SD) | 7.55 (0.89) | 7.34 (1.25) | MD 0.210 (\u20130.212\u20130.632), p = 0.330\nMedian (25\u201375%) | 8 (7\u20138) | 7 (7\u20138)\nmissing | 7 | 6\nBirthweight | Mean(SD) | 3388.61(449.00) | 3429.11(486.40) | MD\u201340.50 (\u2013141.53\u201360.53), p = 0.432\nmissing | 2 | 1\nFemale | N (%) | 123 (48.6) | 120 (48.39) | \nAny congenital malformation | N (%) | 6 (2.37) | 3 (1.21) | \nAdmission of newborn to the Intensive Care Unit | N (%) | 6 (2.37) | 5 (2.02) | \nWhen did breastfeeding start? | N (%) |  |  | RR 1.07 (0.97\u20131.18), p = 0.186, RD 0.062 (\u20130.028\u20130.153), p = 0.176\n1) Within one hour after birth | 238 (94.07) | 217 (87.50)\n2) Between one hour and 24 hours after birth | 10 (3.95) | 25 (10.08)\n3) After 24 hours | 3 (1.19) | 2 (0.81)\n4) Breastfeeding not initiated before discharge/ after birth | 1 (0.40) | 2 (0.81)\nNeonatal status at discharge Death | N (%) | 1 (0.40) | 2 (0.81) | RR 1.00 (0.99\u20131.02), p = 0.512, RD 0.0041 (\u20130.0082\u20130.016), p = 0.512\nHealthy behaviors\nDrinking during pregnancy | N (%) | 20 (7.91) | 35 (14.11) | RR 1.07 (0.97\u20131.18), p = 0.166, RD 0.061 (\u20130.024\u20130.15), p = 0.161\nmissing | 2 | 0\nMaternal smoking | N (%) | 5 (1.98) | 7 (2.82) | RR 1.01 (0.98\u20131.04), p = 0.572, RD 0.0086 (\u20130.021\u20130.038), p = 0.571\nmissing | 0 | 1\nSmoking among other members of the household during pregnancy | N (%) | 129 (50.98) | 151 (60.89) | RR 0.841 (0.71\u20130.99), p = 0.039, RD \u22120.097 (\u22120.194\u2013\u22120.001), p = 0.048\nmissing | 1 | 1\n",
        "label": "unclear",
        "id": "task4_RLD_test_488"
    },
    {
        "paper_doi": "10.1371/journal.pone.0000542",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Trial design: randomized, open\nTime period/duration of trial: 5 June 2005-8 September 2005\nDuration of follow-up: 6 months after enrolment\n\n\nParticipants: Setting: presenting to the outpatient or emergency department of Patan Hospital\nLocation: Lalitpur, Nepal\nAge: 2-65 years; median age 17 (range: 2-64 years)\nGender: male = 247, female = 135\nHealth status of participants: not recorded\nInclusion criteria:clinically diagnosed enteric feverresidence within 2.5 km of the hospitalability to take oral medicationsno pregnancy/lactationno history of seizuresability to stay in the city for the duration of the treatmentno contraindications to quinolones or cephalosporinsability to give written informed consentExclusion criteria:signs of complicated typhoid defined as the presence of jaundice, gastrointestinal bleeding, peritonism, shock, encephalopathy, convulsions, myocarditis or arrhythmia at the time of enrolmentreceipt of a third-generation cephalosporin, fluoroquinolone or macrolide in the week prior to presentation at the clinic\n\n\nInterventions: Cefixime: oral, 20 mg/kg/day in 2 divided doses for 7 daysGatifloxacin: oral, 10 mg/kg/day as a single dose for 7 days\n\n\nOutcomes: Primary outcomeFever clearance time - time to first drop in oral temperature <= 37.5 degC and remaining as such for 48 hSecondary outcomesAcute treatment failure: severe complications, persistence of fever > 38 degC, persistence of symptoms for > 7 days after start of therapy, requirement for additional or rescue treatmentOverall treatment failure: acute treatment failure + relapse + deathRelapse: fever with a positive blood culture within 1 month of completing treatment in a patient who had been successfully treated. Excluded individuals given prolonged or rescue treatmentCultures: faecal samples taken at end of 1st, 3rd and 6th monthAdverse events: not prespecified, listed in the results\n\n\nOrganism type and antimicrobial susceptibility: No breakdown of results by organism/susceptibility\n\n\nNotes: 482 clinical cases of enteric fever, with 390 enrolled and randomized.92 were ineligible, with the commonest reason being receipt of antibiotics in the week before trial entry (n = 49).187 participants assigned to receive cefixime, and 77 of these were culture-positive; 203 were assigned to gatifloxacin and 92 of these were culture-positiveBoth ITT (all 390 randomized participants) and positive pre-treatment culture analyses (defined as the per protocol participants, with positive pre-treatment culture results).We requested and received data from the trial authors to calculate the mean time to defervescence\n\n",
        "objective": "To evaluate the effectiveness of cephalosporins for treating enteric fever in children and adults compared to other antimicrobials.",
        "full_paper": "Objective\nTo assess the efficacy of gatifloxacin versus cefixime in the treatment of uncomplicated culture positive enteric fever.\nDesign\nA randomized, open-label, active control trial with two parallel arms.\nSetting\nEmergency Room and Outpatient Clinics in Patan Hospital, Lagankhel, Lalitpur, Nepal.\nParticipants\nPatients with clinically diagnosed uncomplicated enteric fever meeting the inclusion criteria.\nInterventions\nPatients were allocated to receive one of two drugs, Gatifloxacin or Cefixime.\nThe dosages used were Gatifloxacin 10 mg/kg, given once daily for 7 days, or Cefixime 20 mg/kg/day given in two divided doses for 7 days.\nOutcome Measures\nThe primary outcome measure was fever clearance time.\nThe secondary outcome measure was overall treatment failure (acute treatment failure and relapse).\nResults\nRandomization was carried out in 390 patients before enrollment was suspended on the advice of the independent data safety monitoring board due to significant differences in both primary and secondary outcome measures in the two arms and the attainment of a priori defined endpoints.\nMedian (95% confidence interval) fever clearance times were 92 hours (84\u2013114 hours) for gatifloxacin recipients and 138 hours (105\u2013164 hours) for cefixime-treated patients (Hazard Ratio[95%CI]\u200a=\u200a2.171 [1.545\u20133.051], p<0.0001). 19 out of 70 (27%) patients who completed the 7 day trial had acute clinical failure in the cefixime group as compared to 1 out of 88 patients (1%) in gatifloxacin group(Odds Ratio [95%CI]\u200a=\u200a0.031 [0.004 \u2013 0.237], p<0.001).\nOverall treatment failure patients (relapsed patients plus acute treatment failure patients plus death) numbered 29.\nThey were determined to be (95% confidence interval) 37.6 % (27.14%\u201350.2%) in the cefixime group and 3.5% (2.2%\u201311.5%) in the gatifloxacin group (HR[95%CI]\u200a=\u200a0.084 [0.025\u20130.280], p<0.0001).\nThere was one death in the cefixime group.\nConclusions\nBased on this study, gatifloxacin is a better treatment for uncomplicated enteric fever as compared to cefixime.\nTrial Registration\nCurrent Controlled Trials ISRCTN75784880\nIntroduction\nEnteric fever (Typhoid and Paratyphoid fever) is a systemic infection caused by the bacterium Salmonella enterica serovar Typhi (S.typhi) or Salmonella enterica serovar Paratyphi (S. paratyphi) which in humans is transmitted through the fecal-oral route.\nToday the vast burden of disease is encountered in the developing world where sanitary conditions remain poor.\nThe best global estimates are of at least 22 million cases of typhoid fever each year with 200,000 deaths.\nCrucially these are almost exclusively confined to resource poor countries.\nA recent Cochrane review on typhoid treatments underscored the need for large sample size drug interventional trials, especially in children in whom this disease predominates.\nIn 1948 the introduction of chloramphenicol revolutionized the treatment of typhoid fever.\nUnfortunately the emergence of resistance to the \u201cfirst line\u201d antimicrobials (for example, ciprofloxacin) has been a major setback and has given rise to the possibility of untreatable enteric fever.\nGatifloxacin, a relatively inexpensive fluoroquinolone antibiotic in South Asia with once daily oral administration, is a new broad spectrum synthetic 8-methoxyfluoroquinolone which has the lowest minimum inhibitory concentration (MIC) against S. typhi from Nepal.\nThis in vitro activity needs to be verified clinically before gatifloxacin can be recommended for widespread use.\nCefixime, an orally administered third generation cephalosporin, is a commonly used drug in South Asia for the treatment of enteric fever.\nAlthough cefixime is recommended as a drug of choice by the World Health Organization (WHO) for the treatment of resistant typhoid fever it is relatively expensive in South Asia and has to be administered for a longer duration than the currently used fluoroquinolones.\nClearly there is an urgent need for a treatment that combines ease of oral administration, with speed of clinical response, reduction in secondary transmission and inexpensiveness.\nIn this open randomized trial, we aimed to compare clinical outcomes for the treatment of uncomplicated enteric fever with gatifloxacin or cefixime in an outpatient setting.\nMethods\nParticipants\nThe study was approved by Nepal Health Research Council and Oxford Tropical Research Ethics Committee.\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nWe enrolled patients who presented to the outpatient or emergency department of Patan Hospital, Lalitpur, Nepal from June 5, 2005 to September 8, 2005.\nPatan Hospital is a 318 \u2013bed hospital located in the Lalitpur district in Kathmandu Valley.\nPatients were eligible to enter the study if they had clinically diagnosed enteric fever and their residence was within approximately 2.5 km radius from the hospital.\nOther inclusion criteria were that patients must be aged between 2 and 65 years, able to take oral medications, non-pregnant and non-lactating, without a history of seizures, able to stay in the city for the duration of the treatment, not known to have contraindications to either cephalosporins or fluoroquinolones and willing to give informed written consent to take part in the study.\nFor children enrolled into the study, written informed consent was taken from a parent.\nPatients were excluded from the study if they had any signs of complicated typhoid defined as the presence of jaundice, gastrointestinal bleeding, peritonism, shock, encephalopathy, convulsions, myocarditis or arrythmia at the time of enrollment.\nPatients who had received a third generation cephalosporin, fluoroquinolone or macrolide in the week prior to presentation to our clinic were also excluded.\nInterventions\nOn presentation to Patan Hospital all patients with fever without an obvious focus were referred to the enteric fever study clinic, where they were seen by the study physician.\nPatients who fulfilled the inclusion criteria were randomly assigned to receive Gatifloxacin (Broadband\u2122, Novartis AG Basel, Switzerland) 10 mg/kg/day, in a single dose orally for 7 days or Cefixime (Cifex\u2122, Aegis, Nicosia, Cyprus) 20 mg/kg/day in two divided doses orally for 7 days.\nBoth drugs were administered in tablet form, cut and weighed in a sensitive scale to ensure that underdosing did not occur.\nTo children who were apprehensive of swallowing the tablet, the drug was embedded in a banana and given.\nAll patients were asked to swallow the study drug under direct observation during each visit.\nEach patient had haematocrit, total leucocyte count with differential, serum creatinine, total bilirubin, alanine aminotranferase(ALT), and aspartate aminotransferase(AST) measured, and blood and stool cultures were also performed before the start of the study intervention.\nThe exact location of the patient's home was recorded and the first dose of drug administered at the clinic.\nWe employed six Community Medical Auxiliaries (CMA) who had all received at least 15 months of prior formal primary health care worker training and been registered in a government recognized institution.\nThe CMAs visited patients twice daily at their homes to perform a simple clinical assessment, measure the oral temperature and give directly observed therapy with the study drugs.\nThe CMA visited the patient's home every 12 hours, morning and evening, until day 10 following enrollment or complete resolution of illness, whichever came later.\nThe oral temperature of the patient was recorded twice every day by the CMA and a note was made of the timing and dosages of acetaminophen intake.\nThe quality of patient-visits was ensured by regular unplanned supervisory checks in which the study doctor accompanied the CMA during the visits to patients' homes\nCMAs were asked to send patients immediately to the hospital on encountering any severe symptoms, and the patients also were asked to attend clinic if they had any severe symptoms at any other time.\nA symptom questionnaire was used daily during each visit to monitor any adverse events.\nAny patient with any severe symptom was seen by the study physician.\nThe CMAs and study physicians held daily case conferences at which all the study patients were discussed.\nAll patients regardless of the culture results were seen at hospital on Day 10 following enrollment.\nBlood and stool cultures were repeated on Day 10 in all culture positive patients and thereafter if the patient again became ill with probable enteric fever.\nAll culture positive patients were followed up until six months after enrollment, and stool cultures were performed at the end of the first, third and sixth month.\nMicrobiological Procedures\nBlood culture was performed on media containing tryptone soya broth and sodium polyethanol sulphonate, incubated at 37 C and examined daily for growth over 7 days.\nSalmonella enterica serotype Typhi or Paratyphi A, B or C isolated in culture were identified using standard biochemical tests and specific antisera (Murex Biotech, Dartford, England).\nAntibiotic susceptibilities were determined during isolation using the Kirby-Bauer disc diffusion method involving antibiotic discs containing Nalidixic acid, Ofloxacin, Ciprofloxacin, Chloramphenicol, Ampicillin, Cotrimoxazole, Cefixime and Cefotaxime (HiMedia Laboratories, Mumbai, India).\nMinimum Inhibitory Concentrations (MICs) were determined later for organisms stored in glycerol (bacterial preserver) at -70C.\nThe MICs were determined by Chloramphenicol, Nalidixic acid, Gatifloxacin, Cefixime, Ceftriaxone and Gemifloxacin E-tests\u2122 (AB Biodisk, Solna, Sweden), according to the manufacturer's instructions.\nThe sensitivity tests were interpreted using Clinical and Laboratory Standards Institute criteria for Enterobacteriaceae.\nObjectives\nThe objective of the study was to compare the efficacy of Gatifloxacin and Cefixime in the treatment of uncomplicated culture positive enteric fever.\nOutcomes\nThe primary outcome was the fever clearance time (FCT).\nFCT was defined as time to first drop in oral temperature \u2264 37.5\u00b0C, remaining \u2264 37.5\u00b0C for 48 hours.\nThe secondary outcomes included acute treatment failure.\nAcute treatment failure was defined as including any severe complication; the persistence of fever (> 38 C); the persistence of symptoms for more than 7 days after the start of treatment , requiring additional or rescue treatment.\nIf a patient had a temperature above 37.5 and below 38 for more than 7 days, but did not need additional or rescue therapy, and subsequently their fever cleared by day 10, that patient would not qualify as an acute treatment failure.\nPatients who failed the study treatment were given rescue treatment.\nThe rescue drug was Ofloxacin 20 mg/kg/day orally in two divided doses for 14 days for the Cefixime group, and Ceftriaxone 40 mg/kg/day IV in a single daily dose for 14 days for the Gatifloxacin group.\nFor the Cefixime group alone, if on day 8 of treatment the patient still had a fever of >\u200a=\u200a38\u00b0C, the study drug was continued for 10 days and the patient categorized as acute treatment failure.\nIf the temperature on Day 10 was >37.5, rescue treatment was given.\nA relapse was defined as fever with a positive blood culture within a month of completing treatment.\nAll the relapses were patients that were initially categorized as successfully treated.\nAny patient given rescue treatment or prolonged treatment was precluded from the \u201crelapse\u201d group.\nPatients categorized as \u201coverall treatment failures\u201d included patients experiencing acute treatment failure, plus those falling into the relapsed category, plus all deaths within the trial follow up period.\nSample size\nThe sample size was calculated to detect a FCT difference of approximately 48 hours between gatifloxacin (assumed median FCT 156 hrs) and cefixime (assumed median FCT 204 hrs) with p\u200a=\u200a0.05 and power\u200a=\u200a80%.\nThe accrual time for recruitment was assumed to last 70 days, and that the last patient would be followed up until 8 days after recruitment.\nTherefore, we estimated the minimum sample size at 235 participants.\nAssuming a loss to follow-up of 5%, the sample size was calculated as 125 blood culture positive patients in each arm.\nBefore the recommended sample size had been reached, once 169 blood culture positive patients had been enrolled, the independent data safety monitoring committee (DSMC) advised the Principal Investigators to stop recruitment to the trial based on a priori defined difference (p<0.01) between the two treatment arms in the primary endpoints of the study.\nRandomization\u2014Sequence generation\nPatients were randomized in blocks of 100 from a computer generated randomization list, by an investigator not involved in patient recruitment or assessment.\nRandomization\u2014Allocation concealment\nThe randomization sequence and block size was concealed from the physicians allocating treatment and managing the patients, prior to patient enrollment.\nTreatment allocations were kept in sealed opaque envelopes, which were opened only on enrollment of the patient to the study after all inclusion and exclusion criteria had been checked.\nRandomization\u2014Implementation\nParticipants were enrolled by the study physician in the same order in which they presented to the study clinic.\nThe sealed envelopes were opened in strict numeric sequence.\nBlinding\nBlinding was not feasible in this trial due to logistical reasons.\nStatistical methods\nAll data were entered into an electronic database (Microsoft Office Access Version 2003, Wash., USA), and analyses was performed using Stata 9 (Stats Corp LP; Texas, USA). ).\nContinuous covariates were compared between groups of patients using the Mann-Whitney test, and categorical covariates were compared using the chi-square test or Fisher's exact test when appropriate.\nFever clearance times and time to relapse were analyzed using Kaplan Meier survival curves and compared between the two groups using the logrank test.\nBinary outcomes (clinical failures) were compared between the two treatment groups using Fisher's exact test.\nAnalysis was done in all randomized patients (intention to treat, ITT) and separately in patients with positive pretreatment culture (per protocol, PP) and negative pretreatment culture.\nResults\nParticipant flow\nOf the 482 patients from the study area who were clinically diagnosed with enteric fever, 390 patients were enrolled into the study and randomized.\n92 patients were ineligible, the main reason (49 patients) being a history of already having taken antibiotics (fluoroquinolone, macrolide, or third generation cephalosporin) within one week prior to study entry (Figure 1).\nAmong all randomized patients, 187 patients were assigned to receive cefixime and 203 to gatifloxacin.\n77 patients assigned to receive cefixime were blood culture positive for enteric fever whilst 92 of those assigned to receive gatifloxacin were culture positive .\nThere were unequal number of positive patients in each of the study arms.\nOne possible reason for the difference in number of culture positive patients between study arms is that cultures were drawn and culture results obtained after randomization had been done.\nRecruitment\nWe enrolled patients who presented to the outpatient or emergency department of Patan Hospital, Lalitpur, Nepal from June 5, 2005 to September 8, 2005.\nAll enrolled patients were followed up for at least 10 days after recruitment.\nPatients with a positive pretreatment blood culture were followed up for 6 months after enrollment.\nAt the point that the DSMC asked to examine the trial data for the primary outcome measure in positive pre-treatment patients, the median fever clearance time was 92 hours\u00a0(95% CI, 84\u2013114 hours) for the gatifloxacin treated patients and 138 hours\u00a0(95% CI, 105\u2013164 hours) for cefixime treated patients.\nThe difference between the two treatment arms was 46 hours (p<0.0001).\nBaseline data\nAdmission characteristics are shown for all ITT patients (Table 1) and for all PP patients (Table 2).\nThe median age of patients enrolled into the trial was 17 with a range of 2\u201364 years.\nThere were no baseline differences in the culture positive and culture negative groups, other than temperature at presentation, AST and ALT which were higher and platelets and total WBC which were lower in the culture positive patients as compared to the culture-negative patients.\nAmong all PP patients, there were no differences in the baseline characteristics between the two treatment groups.\nThere were 40 patients, 15 in the gatifloxacin arm and 25 in the cefixime arm, who had taken amoxycillin up to the week before study entry.\nOf these 4 and 7 were culture-positive respectively.\nNumbers analyzed\nAnalysis was done in all 390 randomised patients (ITT) and separately in 169 patients with positive pre-treatment culture (PP).\nAll endpoints were analysed in the ITT and PP populations, apart from relapse which was only analysed in the PP population.\nOutcomes and estimation\nPrimary outcome\nIn all ITT patients, median (95% confidence interval) fever clearance time was 102 (90\u2013117) hours for the cefixime group and 72 (62\u201380) hours for the gatifloxacin group, logrank test p<0.0001, Hazard Ratio[95%Confidence Interval]\u200a=\u200a1.821 [1.466\u20132.263].\nThe proportion of all patients failing through time to clear fever is shown in Figure 2.\nAt day 7 fever clearance rate was 73.9% (67.0% \u2013 80.3 %) in cefixime group and 94.2% (90.2% \u2013 96.9%) in gatifloxacin group.\nIn the PP group, median (95% CI) fever clearance time was 92 hours (84\u2013114 hours) for gatifloxacin recipients and 138 hours (105\u2013164 hours) for cefixime-treated patients (HR[95%CI]\u200a=\u200a2.171 [1.545\u20133.051], p<0.0001).\nThe proportion failing to clear fever for each study drug through time after treatment is shown (Figure 3).\nAt day 7 the fever clearance rate was 62.7% (95 % CI\u200a=\u200a51.5%\u201373.8%) in the cefixime group and 91.8% (95 % CI\u200a=\u200a84.8%\u201396.4%) in the gatifloxacin group.\nIn the group with negative blood culture but clinically diagnosed enteric fever (Fig 1), the FCT was 82 hours (95% CI\u200a=\u200a44\u201394 hours) for the cefixime group and 39 hours (95%CI\u200a=\u200a28\u201354 hours) for the gatifloxacin group (HR[95%CI]\u200a=\u200a1.740 [1.309\u20132.312], p<0.0001 logrank test).\nSecondary Outcomes\nIn the ITT group, overall, 30 out of 167 (18%) in the cefixime group and 2 out of 190 (1%) in the gatifloxacin group were acute clinical failures, OR[95%CI]\u200a=\u200a0.049 [0.011\u20130.207], p<0.001, Fisher's exact test.\nIn the PP group, 19 out of 70 (27%) patients who completed the 7-day trial had acute clinical failure in the cefixime recipients as compared to 1 out of 88 (1%) in the gatifloxacin recipients (Odds Ratio [95%CI]\u200a=\u200a0.031 [0.004 \u2013 0.237], p<0.001).\nConsidering all patients to be failures who dropped out of the study before completion of the seven day treatment course, 26 out of 77 (34%) failed in the cefixime group as compared to 5 out of 92 (5%) in the gatifloxacin group (OR[95%CI]\u200a=\u200a0.112 [0.041 \u2013 0.312], p<0.001).\n138 patients were evaluable for relapse; 20 had acute treatment failure and 11 withdrew from the study before day 7.\nIn total, eight relapses (Figure 1) were observed.\nRelapse rates were 12.4% (6/51) in the cefixime group and 3.4% (2/87) in gatifloxacin group (HR[95%CI]\u200a=\u200a0.185 [0.037\u20130.915], p\u200a=\u200a 0.0199).\nThe Kaplan-Meier plots for the time of relapse are shown in Figure 4.\nOverall failures (acute treatment failure plus relapse plus death) were 29 in number (Figure 1).\nOverall failure rate at 1 month was estimated as 37.6% (95% CI\u200a=\u200a27.14% \u2013 50.2%) in the cefixime group and 3.5% ( 95% CI \u200a=\u200a2.2%\u201311.5%) in the Gatifloxacin group (HR[95%CI]\u200a=\u200a0.084[ 0.025\u20130.280], p<0.0001) ( Figure 5).\nFrom patients with negative cultures, 11 had acute clinical failures, 10 (out of 97, 10%) in Cefixime group and 1 (out of 103, 1%) in the Gatifloxacin group, OR[95%CI]\u200a=\u200a0.086 [0.011\u20130.686], p\u200a=\u200a0.004, Fisher's exact test.\nSimilarly, treating drop-out as treatment failures, we had 50 out of 187 (27%) in the Cefixime group and 15 out of 203 (7%) in the Gatifloxacin group acute treatment failures, OR[95%CI]\u200a=\u200a0.219 [0.118\u20130.405], p<0.001, Fisher's exact test.\nAncillary analyses\nAmong all culture positive patients in the cefixime group, one patient (1/70, 1%) had S. Paratyphi A cultured from her blood on day 10,but there were no (0/88, 0%) positive blood culture growths in the gatifloxacin group on day 10.\nNo patient was found to be a persistent carrier of S. Typhi or Paratyphi A in their stool.\nA positive stool culture for S. Typhi was seen for one patient on day 10 and for another on day 30.\nSubsequent cultures were negative for both patients.\nWe were able to obtain stool cultures from 147 (87%), 141 (83%), and 130 (77%) pretreatment blood culture positive patients at one, three, and six months respectively.\nMicrobiology\nAntibiotic sensitivity testing revealed that all strains were sensitive to gatifloxacin, cefixime, ceftriaxone or gemifloxacin.\nOne strain was resistant to chloramphenicol, and 136 (83%) of the pretreatment isolates were nalidixic acid resistant strains (NARST).\nMinimum inhibitory concentration (MIC) was determined for 161 of the pretreatment blood culture isolates.\nThe median (range) MICs for each antibiotic were as follows: gatifloxacin 0.125 (0.006\u20130.5) \u00b5g/mL, cefixime 0.380 (0.016\u20132.0) \u00b5g/mL, nalidixic acid >256 (1.5->256) \u00b5g/mL, chloramphenicol 8.0 (1.5->256) \u00b5g/mL, ceftriaxone 0.125 (0.047\u20130.5) \u00b5g/mL and gemifloxacin 0.125 (0.004\u20130.5) \u00b5g/mL.\nAdverse events\nAmong all patients who received cefixime, there was one death, which might have been due to the development of disease-related complications during treatment.\nThis patient was enrolled on the fourteenth day of his illness.\nOn day 6 of treatment the patient complained of reddish stool and petechiae and was immediately admitted to hospital where he developed severe thrombocytopenia and gastrointestinal bleeding.\nHe developed acute respiratory distress syndrome and was mechanically ventilated.\nHe developed disseminated intravascular coagulation and succumbed to his illness on day 21 of entry into the trial.\nHis pretreatment blood culture grew S. Paratyphi A which was sensitive to cefixime with an MIC of 0.38 \u00b5g/mL.\nOne patient developed erythematous skin rash which needed two doses of oral antihistamine.\nAmong all patients who received gatifloxacin there were 2 patients with excessive vomiting, which required intravenous antiemetics and fluids and observation in the hospital emergency room for upto 6 hours.\nThere were an additional 23 patients who complained of excessive nausea and occasional vomiting after ingestion of the drug.\nOf these, two needed oral antiemetics; in the remaining 21 patients no intervention was required.\nDiscussion\nInterpretation\nIn this study examining fever clearance time, acute treatment failure and relapse as indicators of treatment efficacy, that the results raise doubts on the usefulness of cefixime and suggest that gatifloxacin is a potent choice for the treatment of uncomplicated enteric fever.\nFebrile illness is one of the most common reasons for presentation to hospitals in many developing countries.\nIn patients with fever, a very common clinical diagnosis is enteric fever, and S. enterica serotype Typhi or Paratyphi A are the two most commonly isolated pathogens from the blood in febrile patients in our hospital.\nBefore the advent of multi-drug-resistant (MDR) S. Typhi, chloramphenicol, ampicillin or cotrimoxazole were successfully used as the first line drug in the treatment of enteric fever.\nAfter the emergence of MDR strains, fluoroquinolones and third-generation cephalosporins have been suggested and used as alternative antimicrobials.\nHowever the emergence and spread of point mutations in the gyrA gene of the bacterial genome has conferred resistance to nalidixic acid and reduced susceptibility to the commonly used fluoroquinolones such as ofloxacin, leading to a poorer clinical response.\nA recent study in Viet Nam (CM Parry, unpublished) showed ofloxacin at the dose of 20 mg/kg/day was able to achieve a cure rate in only 64% of patients.\nIn our context of high nalidixic acid resistance, gatifloxacin is the most effective and appropriate choice for treatment of enteric fever.\nGatifloxacin (Sandoz, India) is relatively inexpensive (US$1.2 for a 7 day treatment course) and needs to be administered just once a day; both of these features are attractive in this setting.\nGatifloxacin has a different binding motif than some other fluroquinolones, and this characteristic enables it to retain activity against Salmonella enterica serovar Typhi or Salmonella enterica serovar Paratyphi A even in the presence of marked reduction in sensitivity to the older fluoroquinolones.\nCefixime, a third generation cephalosporin, is widely trusted to be effective for enteric fever as first line treatment, and is also used as second line therapy when initial treatment with a fluoroquinolone in a patient suspected to be enteric fever fails.\nThe fact that we saw a high overall failure rate associated with cefixime despite all of the strains being fully sensitive in vitro to the drug shows that the mechanism of action of cefixime may not be suited to the eradication of S. Typhi or Paratyphi A from the body or blood, and the poor intracellular penetration into macrophages and reticulo endothelial tissues where the typhoid organisms colonize may be the cause of high failure rates.\nThis study was unique in that we used CMAs to simulate a hospital setup in the community.\nCMA's directly observed patients taking the therapies, monitored fever and identified complications early; these characteristics have not been used in the past for typhoid trials although enteric fever in endemic areas is treated on an outpatient basis.\nA major advantage of follow-up using CMAs was that the health workers knew the exact house location of the patients, and therefore follow up even after the successful completion of the initial seven-day drug trial was possible.\nIn developing countries follow up of patients can be very difficult because of a lack of a proper address and relative unavailability of other means of communication, for example,a telephone.\nIn addition to its relevance to culture confirmed enteric fever, another major strength of this large randomized study is that gatifloxacin proved to be more efficacious than cefixime with respect to fever clearance time and failure rates, even in the subgroup of patients who were clinically presumed to have enteric fever but who had a negative blood culture.\nAntibiotic treatment for typhoid in highly endemic areas is usually started based on the presence of a \u201csyndromic\u201d illness (acute fever for a few days and constitutional symptoms with no known source of infection) before culture results are known.\nEnteric fever, which continues to be a neglected disease, is an important cause of morbidity and mortality, and facilities for blood culture or other reliable methods of diagnosis rarely exist in this setting.\nGeneralizability\nDespite widespread resistance to Nalidixic acid in Kathmandu, and rising MICs to the older fluoroquinolones, ciprofloxacin and ofloxacin, gatifloxacin has proven to be a potent drug for the treatment of enteric fever.\nOur study has relevance to South Asia, as resistance to nalidixic acid is widely prevalent there.\nInevitably there will be emergence of resistance to gatifloxacin in areas with both MDR and NARST; and in this situation alternative antibiotics like azithromycin may need to be used.\nOf interest, in keeping with anecdotal reports from elsewhere in South Asia, only one strain was resistant to chloramphenicol in the present study.\nIn areas of the world where chloramphenicol susceptibility has reemerged there may be an argument for reassessing chloramphenicol.\nIn the present study Gatifloxacin was associated with nausea in 12% of patients and it may be important to forewarn patients of this possible side effect.\nThere have been sporadic reports of dysglycemia caused by gatifloxacin, and a recent population-based, case controlled study examining gatifloxacin usage amongst elderly individuals in Canada (mean age 77 years) who developed dysglycemia also raises possible concerns.\nWe did not do any blood sugar testing to look for dysglycemia.\nHowever in a study involving a younger age group where blood sugar testing was done the results revealed no dysglycemia: 887 children were treated with gatifloxacin (10 mg/kg) for otitis media and were followed for a year with no signs of alteration of glucose homeostasis either acutely or otherwise.\nClearly, it would be prudent to treat diabetics and elderly people suffering from enteric fever with an alternative antibiotic such as azithromycin and avoid the potential problems in this specific population with gatifloxacin.\nLimitations of the study\nThe DSMC advised the Principal Investigators in this study to stop recruitment to the trial based on a priori defined difference (p<0.01) between the two treatment arms in the primary endpoints of the study.\nIt is possible that if the trial had been continued with a larger sample size, other important information could have been gathered.\nIn addition if patients and/or investigators had been blinded to treatment assignments, the study would have been further strengthened; however as in most typhoid trials, it was not possible to do this due to the difference in dosing schedule for the two drugs being compared.\nAnother limitation of this study was that temperature was only measured every 12 hours.\nHowever, to address this limitation, and to avoid missing increases in temperature, we checked temperatures for 10 days after enrollment, or for 48 hours after resolution of fever, whichever came later, in all patients.\nFinally, a telephone or internet based system of randomization would be ideal, but such a system does not exist here.\nOverall evidence\nWe have compared the outcomes from our trial with those of other comparable studies, identified from a recent Cochrane review, WHO typhoid guidelines, and a search of Medline using these terms: cefixime, typhoid trials.\nThe findings in our study are consistent with those of a 1995 study done in Viet Nam which showed that cefixime (20 mg/kg/day) for 7 days was inferior to ofloxacin (10 mg/kg/day) for 5 days in the treatment of MDR typhoid fever in children.\nHowever other studies have suggested cefixime can be successful in the treatment of enteric fever.\nOverall these studies, both descriptive and randomized, have examining the use of cefixime in confirmed enteric fever (total of 292 patients) and with treatment durations of mostly 14 days, have found failure rates ranging from 4% to 23%.\nBesides the general undesirability of a longer course with cefixime with increased morbidity and possibly complications, this drug is also more expensive (a 7-day course costs US $7 (Blue Cross Laboratories, India)).\nThe present study is the largest randomized controlled trial ever conducted with cefixime in enteric fever and clearly shows, even in a setting with fully sensitive strains, that cefixime is a poor drug for this disease.\nThese findings are contrary to the recommendation by many sources including the World Health Organization that cefixime can be used as first or second line therapy in the treatment of enteric fever.\nBased on the present study, we believe gatifloxacin to be an optimal choice in the treatment of uncomplicated enteric fever.\nProfile of the Trial.The consort flow diagram showing the flow of participants through the trial.\nProportion of all patients still febrile.Kaplan-Meier survival curve showing the proportion of all patients(ITT) still febrile through time.\nProportion of culture positive patients still febrile.Kaplan-Meier survival curve showing the proportion of culture positive(PP) patients still febrile through time.\nProportion of relapse free patients.Kaplan-Meier survival curve showing the proportion of relapse free patients in the culture positive population.\nProportion of overall failure free patients.Kaplan-Meier survival curve showing the proportion of overall failure free patients in the culture positive population.\n\nBaseline characteristics all patients.\nPATIENT CHARACTERISTICS | Culture negative (213) | Culture positive (169)\nNo of males/No of females | 136/77 | 111/58\nAge (yrs) | 18 (2\u201364) | 17 (2.75\u201350)\nNumber Aged <14 years (%) | 79(37.1) | 60 (35.5)\nWeight (Kg) | 44 (10\u201380) | 46 (10\u201373)\nDuration of fever before treatment (days) | 5 (0\u201321) | 5 (2\u201323)\nMedian oral temperature at presentation(95% CI, range) (in degrees C) | 38.7 (38.6\u201339; 36.5\u201340.7) | 39(38.8\u201339.2; 36.8\u201341)\nHeadache, Number with (%) (median duration [days]) | 204 (95.7) (4) | 164 (97.0) (5.)\nAnorexia, Number with (%) (median duration [days]) | 160 (75.1)  | 129 (76.3) \nAbdominal Pain, Number with (%) (median duration [days]) | 88 (41.3)  | 80 (47.3) \nCough, Number with (%) (median duration [days]) | 83 (39.0)  | 59 (34.9) \nDiarrhoea, Number with (%) (median duration [days]) | 45 (21.1)  | 41 (24.3) \nVomiting, Number with (%) (median duration [days]) | 30 (14.1) | 27 (16.0) \nAbdominal tenderness ( n [%]) ) | 32 [15.1] | 23 [13.6]\nSplenomegaly ( n [%]) | 18 [8.5] | 18 [10.6]\nHepatomegaly ( n [%]) | 12 [5.76] | 9[5.3]\nHematocrit (in%) | 40 (27\u201353) | 40 (29\u201350)\nWhite Cell Count (in \u00d71000 per microlitre) | 7.2 (2.3\u201324.2) | 6.7 (3.0\u201320.0)\nPlatelet Count (in \u00d71000 per microlitre) | 192 (66\u2013546) | 180 (65\u2013380)\n* ALT ( in U/L ) | 30(11\u2013240) | 37 (12\u2013200)\n**AST ( in U/L ) | 43 (20\u2013354) | 52 (21\u2013169)\nTotal Bilirubin ( in mg/dL ) | 0.8 (0.17\u20133.6) | 0.89 (0.18\u20133.2)\n\nBaseline epidemiological, clinical and laboratory features at presentation of all intention to treat patients showing a comparison between culture positive and culture negative groups.\nALT (serum alanine aminotransferase) normal range 5\u201334 U/L\nAST (serum aspartate aminotransferase) normal range 5\u201334 U/L\nAll data presented as median (range) unless specified.\n\nBaseline characteristics at presentation of culture positive patients.\nPATIENT CHARACTERISTICS | GATIFLOXACIN (n\u200a=\u200a92) | CEFIXIME (n\u200a=\u200a77)\nNo of males/No of females | 67/25 | 44/33\nAge (yrs) | 18 (2.75\u201345) | 15 (3\u201350)\nNumber Aged <14 years (%) | 27 (29%) | 33 (43%)\nWeight (Kg) | 49 (10\u201370) | 42 (11\u201373)\nDuration of fever before treatment (days) | 5.2 | 5.4\nMedian oral temperature at presentation(95% CI, range) (in degrees C) | 39 (38.9\u201339.2; 37.5\u201341.0) | 39 (38.8\u201339.2; 36.8\u201340.5)\nHeadache, Number with (%) (median duration [days]) | 88 (95.7%) (5) | 76 (98.7%) (4.5)\nAnorexia, Number with (%) (median duration [days]) | 73 (79.3%) (4) | 56 (73%) (4)\nAbdominal Pain, Number with (%) (median duration [days]) | 43 (46.7%) (4) | 40 (52%) (4)\nCough, Number with (%) (median duration [days]) | 37 (40.2%) (3) | 22 (29%) (3)\nDiarrhoea, Number with (%) (median duration [days]) | 21 (22.8%) (3) | 20 (26%) (3)\nVomiting, Number with (%) (median duration [days]) | 17 (18.5%) (2) | 10 (13%) (1.5)\nAbdominal tenderness ( n [%]) ) | 14 (15.2%) | 8 (10.4%)\nSplenomegaly ( n [%]) | 10 (10.9%) | 8 (10.4%)\nHepatomegaly ( n [%]) | 5 (5.4%) | 4 (5%)\nHematocrit (in%) | 41 (30\u201350) | 40 (29\u201350)\nWhite Cell Count (in \u00d71000 per microlitre) | 6.8(3.0\u201318) | 6.7 (3.1\u201320)\nPlatelet Count (in \u00d71000 per microlitre) | 180(65\u2013367) | 186 (120\u2013380)\n* ALT ( in U/L ) | 36 (12\u2013180) | 39(18\u2013200)\n**AST ( in U/L ) | 53 (24\u2013155) | 49 (21\u2013169)\nTotal Bilirubin ( in mg/dL ) | 0.85 (0.18\u20133.2) | 0.9 (0.35\u20132.3)\nPositive pretreatment fecal cultures ( n [%]) | 9 (9.8%) | 3 (3.8%)\n\nBaseline epidemiological, clinical and laboratory features at presentation of all blood culture positive patients showing a comparison between the gatifloxacin and cefixime arms.\nALT (serum alanine aminotransferase) normal range 5\u201334 U/L\nAST (serum aspartate aminotransferase) normal range 5\u201334 U/L\nAll data presented as median ( range) unless specified.",
        "label": "high",
        "id": "task4_RLD_test_200"
    },
    {
        "paper_doi": "10.1186/s12936-019-2775-1",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: 2 x 2 factorial, cluster-randomized, controlled trial. Villages were randomized to 1 of 4 arms: 1. LLIN + IRS; 2. LLIN alone; 3. IRS alone; or 4. control. For this review, the relevant comparison was 1. LLIN + IRS vs 2. LLIN alone.Unit of allocation: villageNumber of units: 176 total (44 in each arm)Outcome assessment/surveillance type: active and passive case detection. Through weekly household visits, study participants with a fever or history of fever were encouraged to present to the nearest health posts for testing and treatment. Health centres were regularly visited to find malaria cases not reported to field workers.Length of follow-up: 2.5 years (121 weeks)Adjustment: the incidence of malaria was calculated using methods for cluster randomized trials that take into account the intracluster correlation coefficient.\n\n\nParticipants: Number of participants:  9068 in IRS and ITNs arm, 8521 in LLIN-only arm - approximately 196 per clusterPopulation characteristics: predominantly rural. Residents primarily depend on farming, livestock rearing, and to a lesser extent, fishingInclusion criteria for participants: all consenting residents of households in all clusters were recruited for the study.\n\n\nInterventions: IRSActive ingredient and dosage:  propoxur: 2 g/m2Formulation: WPFrequency of spraying: annuallyTime of spraying: prior to transmission season (September 2014, July 2015, July 2016)Spraying conducted by: locally recruited spray personnel and supervisors.Coverage: 95-96%LLIN Active ingredient and dosage: deltamethrin 55 mg/m2 (SD 25%)Time of implementation: at the beginning of study, all households in the IRS + LLIN and LLIN-alone arms received new LLINs free of charge (procured June 2014, first follow-up October 2014). 1 net was given for a family of 1-2 people, 2 nets for a family of 3-5 people, 3 nets for a family of 6-7 people, and 4 nets for a family of >= 8 peopleFormulation: PermaNet 2.0Coverage and compliance both decreased significantly during the study period.Coverage measure:  household ownership of >= 1 LLINCoverage in IRS arm: 100% (at baseline)Coverage in control arm: 100% (at baseline)Coverage after 110 weeks (both arms): 8% Compliance measure: whether any household members used an LLIN the night before the day of the interview Compliance: in IRS arm: 47% (weeks 1-26), 26% (weeks 26-53), 8% (weeks 53-79), 1% (weeks 79-121)\nCompliance in control arm: 49% (weeks 1-26), 27% (weeks 26-53), 6% (weeks 53-79), 1% (weeks 79-121)Cointerventions: none described\n\n\nOutcomes: Malaria case incidence in all ages (determined by the detection of P falciparum or P vivax by RDTs in participants with a fever or history of fever within the previous 48 hours upon arrival to health posts)Malaria prevalence in all age groups at week 57Anaemia prevalence (Hb < 11 g/dL) in children aged 6-59 months\n\n\nLocation profile: Study location: Ethiopia (Adami Tullu, Adami Tullu-Jiddo-Kombolcha district, East Shewa Zone, Oromia Regional State)Malaria endemicity: seasonal, with the peak malaria transmission season from September to DecemberPlasmodium species: P falciparum and P vivax\n\n\nVector profile: Primary (and secondary) vector species: An arabiensis (primary), An pharoensis (auxiliary)Phenotypic resistance profile:An arabiensis was susceptible to propoxur (a carbamate), but resistant to the pyrethroid insecticides. An pharoensis was susceptible to all pyrethroids and carbamates testedMethod of mosquito collection: malaria vectors were collected in randomly selected houses using light trap catches (LTC), pyrethrum spray catches (PSC), and artificial outdoor pit shelters (PIT). LTC and PIT were placed in 1 house per cluster. PSC was performed in 4 houses per cluster. LTC, PSC, and PIT were used to monitor the impact of the interventions on An arabiensis host-seeking density, indoor resting density, and outdoor resting density, respectively. In addition, human landing catch was performed indoors and outdoors in 1 house in 1 cluster per study arm to monitor the impact the interventions on An arabiensis human biting rates.\n\n\nNotes: Intracluster correlation coefficient: 0.01 (obtained from study authors. This was used to calculate an adjusted Incidence RR by review authors)\n\n",
        "objective": "To summarize the effect on malaria of additionally implementing IRS, using non\u2010pyrethroid\u2010like or pyrethroid\u2010like insecticides, in communities currently using ITNs.",
        "full_paper": "Background\nConflicting results exist on the added benefit of combining long-lasting insecticidal nets (LLINs) with indoor residual spraying (IRS) to control malaria infection.\nThe main study objective was to evaluate whether the combined use of LLINs and IRS with propoxur provides additional protection against Plasmodium falciparum and/or Plasmodium vivax among all age groups compared to LLINs or IRS alone.\nMethods\nThis cluster-randomized, controlled trial was conducted in the Rift Valley area of Ethiopia from September 2014 to January 2017 (121\u00a0weeks); 44 villages were allocated to each of four study arms: LLIN\u2009+\u2009IRS, IRS, LLIN, and control.\nEach week, 6071 households with 34,548 persons were surveyed by active and passive case detection for clinical malaria.\nPrimary endpoints were the incidence of clinical malaria and anaemia prevalence.\nResults\nDuring the study, 1183 malaria episodes were identified, of which 55.1% were P. falciparum and 25.3% were P. vivax, and 19.6% were mixed infections of P. falciparum and P. vivax.\nThe overall malaria incidence was 16.5 per 1000 person-years of observation time (PYO), and similar in the four arms with 17.2 per 1000 PYO in the LLIN\u2009+\u2009IRS arm, 16.1 in LLIN, 17.0 in IRS, and 15.6 in the control arm.\nThere was no significant difference in risk of anaemia among the trial arms.\nConclusions\nThe clinical malaria incidence and anaemia prevalence were similar in the four study groups.\nIn areas with low malaria incidence, using LLINs and IRS in combination or alone may not eliminate malaria.\nComplementary interventions that reduce residual malaria transmission should be explored in addition to LLINs and IRS to further reduce malaria transmission in such settings.\nTrial registration PACTR201411000882128 (08 September 2014)\nElectronic supplementary material\nThe online version of this article (10.1186/s12936-019-2775-1) contains supplementary material, which is available to authorized users.\nBackground\nDespite remarkable achievements in the fight against malaria over the last decade, the World Health Organization (WHO) recommends universal coverage of populations at risk with long-lasting insecticidal nets (LLINs) and targeted indoor residual spraying (IRS) with an insecticide for the control of malaria.\nBoth LLINs and IRS have been shown to be effective in reducing malaria transmission when applied independently.\nIn an effort to accelerate the control and ultimate elimination of malaria, IRS in combination with LLINs has been deployed in some countries, and the available evidence from large surveys, cohort studies, and a randomized trial suggests that the joint intervention of LLINs and IRS should be scaled up and that the combined effect of these interventions should be further evaluated.\nReviews by Pluess et al. in 2009 and WHO in early 2014 documented that historical and programme documentation had clearly established the impact of IRS.\nHowever, the number of high-quality trials was too few to quantify the size of effect in different transmission settings.\nEvidence from randomized comparisons of IRS vs no IRS had confirmed that IRS reduced malaria incidence in unstable malaria settings.\nSome limited data suggested that LLIN gives better protection than IRS in unstable areas, and these reviews together with modelling efforts, recommended that more trials were needed to compare the effects of LLINs with IRS, as well as to quantify their combined effects.\nDespite an increasing interest in the simultaneous use of both interventions, no clear guidelines existed at the start of this study on how these interventions should be combined.\nAt the same time, there is a paucity of evidence concerning whether their combined use is more effective in reducing the incidence of malaria than using either intervention alone.\nSome non-randomized observational studies and mathematical modelling exercises indicate modest effectiveness or conflicting results when combining interventions for malaria reduction compared to either intervention alone.\nConsequently, it is difficult to draw any conclusions regarding whether the combination of IRS and LLINs is beneficial against malaria compared to one of the interventions alone.\nRecent reviews indicated that only a few of the published randomized controlled trials showed additional protection against fighting malaria when the use of LLINs was combined with IRS, compared to either method alone.\nA multi-intervention trial in Benin reported no reduction in clinical malaria in children under the age of 5\u00a0years from houses sprayed with bendiocarb in combination with LLINs, compared to children in houses with LLINs alone.\nSimilarly, in The Gambia, a combination of IRS using DDT and universal coverage of LLINs showed no added protection against malaria among children 6\u00a0months to 14\u00a0years old compared to universal coverage of LLINs alone.\nBy contrast, a recent cluster-randomized controlled trial in Tanzania, where the usage of LLINs was less than 50%, found some evidence of added protection against malaria infection in children 6\u00a0months to 14\u00a0years from the combination of LLINs and IRS with bendiocarb compared to LLINs alone.\nThe specific objectives of this intervention study were: (1) to determine whether the combined use of LLINs and IRS with propoxur provides additional protection against malaria (P. falciparum and/or P. vivax) among all age groups in the study area compared to LLINs or IRS alone; and, (2) determine whether LLINs\u2009+\u2009IRS improves haemoglobin (Hb) concentration and reduces anaemia among children under 5\u00a0years of age compared with children in LLINs or IRS arms alone.\nMethods\nThis study was conducted to evaluate the effect of LLINs and IRS to prevent malaria in southern Ethiopia, and followed the recommendations of Lines and Kleinschmidt.\nThis report includes a comprehensive assessment of the trial results.\nIn parallel with this study, monitoring of LLIN ownership and use, entomological studies and monitoring of insecticide resistance are published in separate reports.\nThe study protocol has been published previously.\nStudy setting\nThis study was carried out in the Adami Tullu part of the Adami Tullu-Jiddo-Kombolcha woreda (district) in the East Shewa Zone of the Oromia Regional State in Ethiopia.\nThe woreda is a local administrative unit in the country, which consists of several kebeles (the lowest government administrative unit; kebele is further divided into gares, or villages).\nAdministratively, the district comprises 48 kebeles, each with a population size of approximately 1000 to 5000 people.\nThe projected population size of the district for 2014 was approximately 173,000 people.\nThe main ethnic group is the Oromo, and the predominant religion is Islam.\nThe majority of the population lives in rural areas in houses made of mud or cement walls and thatched or iron roofs.\nLocal residents primarily depend on farming, livestock rearing, and to a lesser extent, fishing in Lake Zeway, for their subsistence.\nIn 2014, there was one public and one non-governmental organization hospital, 9 public health centres, and 43 health posts in the district.\nEach kebele has at least one health post staffed by two health extension workers (HEWs) reporting to the health centre.\nThe peak malaria transmission season in the study area is from September to December, following the rainy season during June to August.\nPlasmodium falciparum and P. vivax are the main causes of malaria infection in the area.\nAnopheles arabiensis is the major malaria vector in the district and An. pharoensis is considered to have an auxiliary role.\nA study performed prior to the start of this trial demonstrated that An. arabiensis was susceptible to propoxur (a carbamate), but resistant to the pyrethroid insecticides.\nHowever, An. pharoensis was susceptible to all pyrethroids and carbamates tested.\nIn Ethiopia, LLINs and IRS are applied either simultaneously or separately depending on the local setting.\nDesign\nThis 2\u2009\u00d7\u20092 factorial, cluster-randomized, controlled trial was carried out for 121\u00a0weeks from late September 2014 to January 2017.\nThe village (or cluster) constituted the unit of randomization and an equal number of villages were randomized to one of the following four arms: (1) LLIN\u2009+\u2009IRS; (2) LLIN alone; (3) IRS alone; or, (4) control.\nThe control arm received the routine standard practice of malaria prevention of the Ethiopian Malaria Control and Elimination Programme.\nThe control households would receive new LLINs and IRS spraying when the district health office found it appropriate, but during the study period, no communities in the woreda received such additional interventions.\nAll people living in the area were offered malaria diagnosis and treatment, if needed, when presenting at a health institution.\nThe trial was performed as described in the previous protocol.\nAlthough the study was planned for 104\u00a0weeks follow-up, this was extended to 121\u00a0weeks to add one additional malaria transmission season.\nParticipants\nThis trial was conducted in the rural communities of the district.\nPrior to implementing the intervention and randomizing villages to arms, a baseline survey, mapping and a pilot study were carried out to estimate an optimum sample size.\nA population survey in the study households was repeated at the start of each year.\nVillage inclusion and exclusion criteria\nVillages located within 5\u00a0km from Lake Zeway or the Bulbula River were included in the study, as preliminary findings indicated that the incidence of malaria was highest in this part of the area.\nParticipant inclusion and exclusion criteria\nAll consenting residents of households in all clusters were recruited for the study.\nResidents and household heads who did not provide informed consent were excluded.\nRandomization and masking\nFrom a total of 48 rural kebeles in the Adami Tullu district, 13 kebeles adjacent to Lake Zeway and Bulbula River were included in the study.\nFrom the total list of the clusters in 13 kebeles, 207 were included in the sampling frame, of which 176 were randomly selected (see flowchart, Fig.\u00a01).\nRandomization was carried out in Bergen, Norway, to prevent selection bias by concealing the allocation sequence from the field researchers assigning villages to intervention groups until the moment of assignment.\nThus, a researcher not involved in the study randomly allocated a random number that was used as the seed for the computer-generated list of villages using SPSS software.\nBecause of many clusters in each arm, stratification of clusters or restricted randomization was not done.\nThe baseline data collections were carried out before the start of the study in 2014 showed that the study groups were comparable, except for house design (Table\u00a01).\nDue to the nature of the interventions, blinding of the study participants was not possible.\nForty-four clusters were assigned to each of the intervention groups.\nFigure\u00a02 provides information about the interventions in each of the groups, follow-up information, and participants included in the analysis.\nThe LLIN use coverage before the intervention was 11%, and no household had received IRS spraying the year prior to the study.\nInterventions\nLong-lasting insecticidal nets\nThe LLINs distributed for this trial were PermaNet 2.0 rectangular, 100 deniers, light blue, large size (160\u00a0cm width\u2009\u00d7\u2009180\u00a0cm length\u2009\u00d7\u2009150\u00a0cm height) purchased in June 2014 from the Vestergaard Frandsen Group SA (Vestergaard Frandsen, Lausanne, Switzerland).\nAll households in the IRS\u2009+\u2009LLIN and LLIN alone arms received new LLINs free of charge at the beginning of the intervention regardless of previous ownership, with householders maintaining their existing nets at the time of distribution.\nThe number of new LLINs distributed to each household was based on the household size recommended by national malaria guidelines, i.e., one net for a family of 1\u20132, two nets for a family of 3\u20135, three nets for a family of 6\u20137, and four nets for a family of \u2265\u20098 persons.\nIn advance of the LLIN distribution, all village residents were informed about the distribution of the nets through house-to-house visits, village leaders and community elders.\nHouseholds not receiving nets during the first distribution, received nets later.\nEducation about and a demonstration of how to use LLINs were given to the recipients by trained field staff and selected village residents (Fig.\u00a02).\nAll study participants were followed on a weekly basis for 121\u00a0weeks, from October 2014 to January 2017.\nAll study participants were followed until the end of the study or until they were lost to follow-up.\nNewcomers (individuals who joined a cohort as new household members) and newborns during the study period were included in the study (Fig.\u00a02).\nA cross-sectional survey was carried out at the 110th week post-distribution to assess LLIN ownership among all households that received LLINs at baseline and to validate the results of LLIN use.\nWeekly home visits were carried out to record the LLIN use of the study participants.\nEach week, the heads of households or family members aged more than 18\u00a0years were asked whether any household members used an LLIN the night before the day of the interview.\nThe names and codes of the individuals who used the LLIN were recorded.\nIf the visited houses were closed, or if heads of households or family members aged more than 18\u00a0years were absent, the data collectors visited the house at least three more times within the same week.\nIf one or more or all of the household members had left the study area during the study period, the individuals were considered lost to follow-up.\nIn subsequent follow-ups, the households were visited on the same day of the week to maintain a seven-day gap between visits.\nThe visits were carried out early in the morning to observe whether the LLINs were hung in the sleeping space.\nFor the LLIN ownership survey, respondents were asked if they had useable LLINs in their household.\nThe presence of usable LLINs was verified in the visited household by observation.\nIf the LLINs were lost, the reasons for the loss were asked.\nIndoor residual spraying (IRS)\nIndoor residual spraying with propoxur was carried out three times (September 2014, July 2015, July 2016) during the study period in the LLIN\u2009+\u2009IRS and IRS alone arms.\nSpraying was done once a year prior to the peak transmission season, following the national spraying operation guidelines.\nA 6-day spray operation training was given for locally recruited spray personnel and supervisors.\nThe spraying teams were organized into teams of four spray personnel and a porter, and supervised by a squad leader.\nApproximately 12 houses were sprayed by each spray operator per day using an 8-l Hudson X-pert (HD Hudson Manufacturing Co, Chicago, IL, USA).\nPrior to spraying, community sensitization was performed to inform residents about the safety, purpose and time of spraying.\nIRS operation was performed using propoxur (isopropoxy-phenyl methylcarbamate) purchased from the state-owned Adami Tullu Pesticide Processing Share Company located in the study district.\nStudy endpoints\nThe primary outcome measure was malaria incidence determined by the detection of P. falciparum or P. vivax by rapid diagnosis tests (RDTs; CareStart Malaria Pf/Pv combo test; Access Bio Inc, NJ, USA) in patients with a fever or having a history of fever within the previous 48\u00a0h upon arrival to health posts by active and passive case detection (see \u201cData collection methods\u201d for details).\nThe other main outcome was anaemia and Hb concentration in children under the age of 5\u00a0years which was measured using a portable photometer (Hb310 analyser, HemoCue\u00ae AB, Angelhom, Sweden) at the end of each transmission season.\nSample size\nMalaria incidence and anaemia prevalence\nThe sample size was calculated based on earlier model studies, and assuming that the two interventions would provide protection independently of each other by about 40\u201350%, assuming an additional effect of IRS and LLIN combination of 25%.\nThe sample size calculations were based on epidemiological data collected in a baseline pilot study in villages adjacent to Lake Zeway during September to December, 2013.\nThe sample size for the primary endpoint, i.e., the incidence of malaria, was calculated using methods for cluster randomized trials that take into account the intracluster correlation coefficient (ICC), incidence rate, expected effect, and power of the study.\nA baseline malaria incidence rate of 7.85 per 10,000 person-weeks, or 40.8 per 1000 person years (PY) was used, and the coefficient of variation between clusters within each group of k\u2009=\u20090.27 was used for the sample size estimation.\nIn the study, 176 of 207 clusters living within a distance of 5\u00a0km from Lake Zeway were randomly selected.\nThese selected villages had 6071 households with approximately 196 people per cluster were followed for 121\u00a0weeks, achieving a 90% power to detect a 25% reduction in the malaria incidence rate in the IRS\u2009+\u2009LLIN arm compared to LLINs alone or the IRS-only arm, using a two-sided 5% significance level.\nSome 6071 households with 34,548 people were included the trial (Fig.\u00a01).\nThe proposed sample size had the power to detect a mean difference between the study arms of 0.5\u00a0mg/ml Hb concentration in children under the age of 5\u00a0years.\nData collection methods\nEpidemiological data collection\nActive and passive case detection was done to diagnose malaria cases at the health posts throughout the trial using RDTs.\nThrough weekly household visits, study participants with a fever or having a history of fever within the previous 48\u00a0h were given numbered identification cards and encouraged to present to the nearest health posts for testing and treatment.\nAll persons with possible malaria were checked if they had actually visited the health post.\nIn addition, the health centres and the hospital were regularly visited to find malaria cases that could have visited these health facilities without reporting to the field workers.\nVery few such cases were found, but any person from the study villages treated for malaria at a health centre or hospital in the district was included in the study.\nIndividuals who were found to be positive for P. falciparum by RDT were given artemether\u2013lumefantrine [(AL) Coartem\u00ae, Novartis, Basel, Switzerland] two times a day for 3\u00a0days based on body weight, according to national guidelines.\nAL is a fixed dose combination of 20\u00a0mg of artemether plus 120\u00a0mg of lumefantrine.\nPersons with P. vivax infections were treated with chloroquine, 25\u00a0mg/kg for 3\u00a0days (10\u00a0mg base per kg on days 1 and 2, and 5\u00a0mg base per kg on day 3).\nTreatment of other conditions was performed in accordance with national guidelines, or referred to higher-level health facilities.\nPatients with severe illness at the time of visit (from malaria or other causes) were referred to the nearest health facility.\nThe Hb concentration was measured in children 6\u201359\u00a0months old in the study households at three time points during the study (December 2014\u20132016, at end of each year\u2019s main malaria transmission season) to assess the prevalence of anaemia.\nThrough house-to-house visits, a single finger-prick sample was taken from each child, and height and weight were measured.\nChildren with Hb values less than 11\u00a0g/dl were defined as anaemic.\nValidation study\nTo validate the weekly incidence data, a community-based malaria prevalence survey was done on a randomly selected sample of households taking part from each arm of the trial during the main transmission season in November 2015 among all age groups.\nAll household members were eligible and volunteered to be included in the study.\nThe heads of the households were interviewed using a pre-tested structured questionnaire, and tested individuals for malaria parasites using RDT.\nIn this survey, 4450 persons were included, and 0.46% (21 persons) of those who volunteered to contribute had malaria as assessed by RDT.\nThere was no significant difference among the intervention arms, and the prevalence at week 57 was similar to the number of cases reported during the same week by the malaria incidence assessment.\nData management\nThe data collection was done using standardized paper-based forms and questionnaires according to standardized operating procedures.\nData were entered by trained data entry clerks and verified by range and consistency checks, and data cleaning was performed weekly.\nAny discrepancies were corrected by cross-checking against the corresponding original forms and subsequently amended in the final dataset.\nTo minimize any loss to follow-up, all residents were followed and recorded if they moved out of the trial area or moved from one cluster to another cluster with a different intervention.\nFor residents or respondents who were present at the time of the visit by project staff, basic information about dates and reasons for absence were recorded from other community members, such as friends or neighbours.\nAnalysis\nThe primary health outcome measure was malaria incidence determined by the detection of P. falciparum or P. vivax using RDTs.\nAll analyses were done on an intention-to-treat basis, regardless of whether the individual household members used LLINs, IRS or neither.\nAll analyses were conducted using Stata version 13 (Stata Corp LP, College Station, TX, USA).\nOutcomes were compared between study arms.\nTo control for potential confounding factors, the clustering effect of villages and the effect of repeated measurement in the same individual and individual-level covariates (such as age, gender, LLIN use) were taken into consideration during the analysis.\nBuilding materials (roof type) is another potential confounding factor, and was also adjusted for in the regression analysis, which was estimated by a proportional hazards model.\nFor ease of analysis, corrugated iron was merged with cement/concrete roof type.\nThough the main analysis plan was intention-to-treat, considering known protective effect of LLINs, there was a need to see if the rate of malaria infection for an average LLIN user was different from non-user.\nThus, to determine if LLINs provided individual-level protection against malaria, a generalized estimating equation with Poisson log linear model was used to adjust for within-cluster correlation of measurements.\nPrincipal component analysis was used to construct a wealth index, as has been described before.\nSatScan v9.1.1 (http://www.satscan.org/) software was used for spatial and space\u2013time statistical analysis, to identify statistically significant retrospective space\u2013time malaria clusters.\nEthical approval\nThe study was approved by the Institutional Review Board (IRB) of the College of Health Sciences at Addis Ababa University, the Ministry of Science and Technology, Ethiopia (ref: 3.10/446/06) and the Regional Committee for Medical and Health Research Ethics, Western Norway (ref: 2013/986/REK Vest).\nThis study contains a control group, which did not receive any additional interventions except for the routine malaria work carried out by the district health office.\nThe three ethical review boards accepted that such a group was included provided that the malaria incidence was followed closely, and if malaria incidence was not higher than expected.\nThe study regularly monitored the malaria incidence in all four groups throughout the study, and the research did not observe higher incidences in the control group, nor any epidemics.\nCommunity consultation and sensitization\nPrior to the implementation of interventions, a consultative workshop and several meetings were held to explain the objectives, kebele selection and randomization, implementation procedures, and expected outcomes of the trial to the communities, with representatives from the Oromia Regional Health Bureau, the East Shewa Zone Health Department, the Adami Tullu District Health Office and the District Administration.\nStudy communities were sensitized prior to randomization through meetings and discussions with community leaders, kebeles, village leaders, and community elders.\nInformation and informed consent\nVerbal informed consent to participate in the study was obtained from the study participants and from parents or guardians for children under 18\u00a0years old using the local Afan Oromo language.\nVerbal consent was used because many of the participants could not read and write.\nThis consenting procedure was approved by the three ethical committees.\nInformation sheets were provided about the purpose of the study, and the participants were informed that involvement in the study was voluntary and that they had the right to withdraw at any time regardless of reason.\nAt each data collection, verbal consent was obtained from all study participants, and verbal assent was obtained from parents or guardians for children using the local language.\nAssurance was given that a refusal to participate in this study would not affect their access to services at the health posts in the study villages in the community.\nAfter completion of the study, households in the IRS and control groups received LLINs according to the national guidelines for bed net distribution.\nTimelines of activities\nEthical approval and pilot study were conducted in 2013, followed by trial registration, actual intervention and outcome measurement (see Additional file 1: Figure S1).\nResults\nIntervention coverage\nBaseline data collections done before the start of the study in 2014 showed that the study groups were comparable, except for house design (Table\u00a01).\nMore households in the control group (57.2 vs 42.8%; \u03c72\u2009=\u200969.4, P\u2009<\u20090.001) had corrugated iron roofs.\nThe study population consisted of 34,548 people (70,356 PYs of observation) with an average of 196 people per cluster (Table\u00a01).\nOf these, 6488 (19%) were children under 5\u00a0years of age, 11,136 (32%) were between 5 and 14\u00a0years, and 16,924 (49%) were older than 15\u00a0years of age.\nTable\u00a02 provides information about the follow-up of the four intervention arms.\nWith an average of 2.57 nets per household, a total of 3006 households (1618 households in LLIN\u2009+\u2009IRS arm and 1388 households in LLIN arm) in both arms of the trial received 7740 LLINs (4157 nets in LLINs\u2009+\u2009IRS and 3583 nets in LLINs only).\nIncidence of malaria by study arm\nDuring the 121\u00a0weeks from September 2014 to January 2017, there were 1183 malaria episodes, of which 652 (55.1%) were P. falciparum infections, 299 (25.3%) were P. vivax infections, and 232 (19.6%) were mixed P. falciparum and P. vivax infections (Table\u00a03); 124 repeated episodes of malaria were diagnosed (Table\u00a03), and the repeated episodes occurred more than 4\u00a0weeks after their first episode.\nThe overall malaria incidence was 16.5 per 1000 PYs of observation time (PYO).\nIncidence rates were similar in the four arms with 17.2 per 1000 PYO in the LLIN\u2009+\u2009IRS arm, 16.1 in the LLIN arm, 17.0 in the IRS arm, and 15.6 in the control arm (Table\u00a03).\nThe incidence of P falciparum infection was 9.1 per 1000 PYO (95% CI 8.4\u20139.8), for P. vivax 4.2 (3.7\u20134.6), and for mixed P. falciparum and P. vivax infection 3.3 per 1000 PYO (95% CI 2.8\u20133.6).\nThere was no difference in malaria incidence among the four arms adjusting for roof type.\nThe hazard rate of malaria infection for those living in thatched roofs was 18% higher than in households with corrugated iron roofs (Table\u00a04).\nThe generalized estimating equation (GEE) showed that LLINs did not provide individual protection against malaria infection in the study setting (P\u2009=\u20090.53).\nOf the 1059 households with first episodes of malaria, 484 episodes occurred in areas with malaria clustering.\nThe incidence of malaria in the clustered areas was 38.2 per 1000 PYO, and higher than 9.2 per 1000 PYO in the non-clustered areas; the incidence risk ratio (IRR) was 3.93 (95% CI 3.48\u20134.38).\nHowever, the IRR between the intervention groups in the clustered areas were similar.\nMalaria and anaemia\nThe prevalence of anaemia was 28.2% (95% CI 26.6\u201329.8) in 2014 and increased to 36.8% (95% CI 35.1\u201338.5) in 2015, and fell to 29.8% (95% CI 28.2\u201331.4) at the end of the study.\nThere was no significant difference in risk of anaemia among the trial arms (Table\u00a05).\nDiscussion\nThe main finding in this study was that LLINs and IRS, alone or in combination, did not reduce malaria incidence to levels feasible for malaria elimination.\nThe average malaria incidence across study arms was 16.5 episodes per 1000 PYO and there were no significant differences between study arms.\nThe potential reasons for these results are discussed below.\nIn this study, incidence of malaria was low, and the trial did not demonstrate any reduction in malaria incidence in the intervention groups.\nThe study did not document any additional benefit in using the combination of LLINs plus IRS compared with single interventions with a low malaria incidence.\nHowever, the entomological results from the study indicate that combining IRS with LLINs reduced An. arabiensis densities compared to LLINs alone and to the control group.\nDespite the fact that the population is representative of the rural population living in similar ecological settings in Ethiopia, the generalizability of the study findings might be affected by the context of the study period.\nIn the years 2015 and 2016, the study area was affected by an unexpected severe drought and food shortages.\nThis may partially also explain the low LLINs use, as the use of bed nets is associated with lower perceived risk of malaria infection.\nAlthough the interventions resulted in lower mosquito densities in houses using IRS compared with LLINs and the control arm, the current study did not find a similar effect on malaria incidence.\nThe study shows that IRS or LLINs [even at varying degrees of coverage (Table\u00a03)] may not be able to reduce malaria incidence further in areas with a low malaria incidence.\nThe study suggests that using LLINs and IRS alone in areas with low malaria incidence may not be able to substantially reduce malaria incidence or eliminate malaria, as has also been suggested in a recent review and modelling studies.\nA study performed in Benin could not document any effect of LLINs or the combination of IRS and IRS on the incidence of malaria and neither did a study from The Gambia.\nA cohort study with a higher incidence carried out in southwest Ethiopia demonstrated that the use of LLINs, although providing individual protection, did not have an effect on incidence, while IRS spraying showed a reduction in malaria incidence.\nThe study results should not be interpreted to indicate that areas with higher incidences would not benefit from such interventions.\nLLINs and IRS did not have an observable impact in this study which was conducted in an area of low transmission.\nHowever, both interventions reduced malaria incidence in southwest Ethiopia and Sudan in areas of higher transmission.\nThe occurrence of malaria is different from other African countries and is characterized by a mixture of P. falciparum and P. vivax infections, and where An. arabiensis is the principal vector.\nThe impact of residual malaria transmission was mainly driven by outdoor biting and early indoor biting behaviour of An. arabiensis.\nComplementary interventions that reduce the risk of acquiring malaria infections both outdoors and before sleeping hours, such as toxic sugar bait and interventions that reduce the density of mosquitoes that feed on cattle, e.g., ivermectin, should be explored in addition to LLINs and IRS to further reduce malaria transmission in such settings.\nA strength of this study is that it was followed by regular monitoring of insecticidal susceptibility of An. arabiensis.\nDuring the trial period, there was no change either in susceptibility to the carbamates or to the pyrethroids of An. arabiensis.\nAnopheles arabiensis was resistant to deltamethrin, while An. pharoensis remained susceptible to all insecticides.\nThe bio-efficacy of LLINs was acceptable for at least 24\u00a0months.\nNevertheless, IRS use remained high during the malaria transmission seasons, and An. arabiensis was sensitive to propoxur, assessed by doing monthly cone bioassays for at least the malaria transmission periods.\nThe frequency of pyrethroid resistance to An. arabiensis remained high (over 90%) and stable throughout the study.\nThe bio-efficacy of nets to insecticide-susceptible insectary colony of An. arabiensis was high.\nSusceptibility to deltamethrin was restored after exposure of An. arabiensis to piperonyl butoxide (PBO), implicating the role of mixed function oxidases in the resistance of this insecticide.\nRecent trials using LLINs with permethrin (a pyrethroid) and pyriproxyfen had increased efficacy compared with LLINs treated with permethrin alone, and introducing this new LLIN could be explored in areas where there exists An. arabiensis.\nIn another study, the PBO long-lasting insecticidal net and non-pyrethroid indoor residual spraying interventions showed improved control of malaria transmission compared with standard long-lasting insecticidal nets where pyrethroid resistance is prevalent.\nAs a result, WHO has since recommended to increase coverage of PBO long-lasting insecticidal nets.\nThe study showed a low LLIN ownership after 2\u00a0years, and a low LLIN use, despite 100% net coverage at baseline.\nThe use of LLINs was closely monitored through weekly home visits, and this rigorous monitoring gives a more realistic assessment than some cross-sectional surveys.\nAnother study from the same trial population found behavioural, socio-cultural, economic, and ecological conditions, weak education, communication and social support structures were important in understanding and accounting for why a low level of intended use and a widespread misuse and repurposed use.\nThe study highlights the need to design strategies to increase LLIN ownership and use in low malaria transmission setting.\nA strength of this study is that it was based on a random selection of villages: typical rural communities in Ethiopia.\nMoreover, the study included a large sample with high power and an adequate follow-up period.\nHowever, as the incidence rate of malaria was lower than expected at the start of the study, this could affect the statistical power of the study.\nThe research was based on a hypothesised effect size of 30\u201350% reduction in malaria incidence.\nUsing an effect size of 30% between the LLIN\u2009+\u2009IRS and the Control arm, a sensitivity analysis showed that the statistical power is 82%.\nIn the study, except for housing type and the interventions, the baseline characteristics of the study arms were balanced (Table\u00a01).\nIn addition, quality-checks on the reported malaria cases took place at health posts, health centres and hospitals in the area.\nIt is unlikely that the study would have missed many malaria cases.\nThe method used to find malaria cases was based on active and passive case finding using trained staff and appropriate RDTs to diagnose malaria.\nAs an additional quality check, a prevalence survey at one point demonstrated that the incidence and the prevalence survey provided similar results, which were also comparable to results from the Malaria Indicator Survey in the same period.\nAnother limitation of the study could be the potential spill-over effect between clusters, and such an effect could have diluted any difference in the outcome measure.\nHowever, the IRS\u2009+\u2009LLINs were as effective as IRS alone in reducing densities and human biting rates of An. arabiensis, and the effectiveness of the two interventions combined was better than LLINs alone in reducing densities and human biting rates of the vector.\nAdded impact of the combination intervention against malaria infectivity rates of An. arabiensis compared to either intervention alone remains unknown and warrants further research and action.\nThe study shows that malaria infection is a risk factor for anaemia, but the prevalence of anaemia was similar in the trial arms (Table\u00a05).\nDespite the malaria prevention efforts, an unexpected increase in anaemia prevalence was observed during the 1st year of this study, most probably because of increasing rates of stunting during this period with food shortages.\nThe risk of anaemia was high among children with malaria, children from poor families, stunted children, and children under 36\u00a0months old.\nConducting malaria prevention trials in drought-prone areas may bring challenges, and a broader assessment of causes of anaemia than used may be appropriate in settings similar to those in this trial.\nConclusions\nThe clinical malaria incidence and anaemia prevalence were similar in the four study groups.\nIn areas with low malaria incidence, using LLINs and IRS in combination or alone may not eliminate malaria.\nComplementary interventions that reduce residual malaria transmission should be explored in addition to LLINs and IRS to further reduce malaria transmission in such settings.\nAdditional file\nFlow diagram illustrating trial profile\nFlow diagram illustrating follow up of trial participants\n\nBaseline characteristics of study clusters at the beginning of the transmission in 2014\n | Intervention arms\nIRS\u2009+\u2009LLINs | % | LLINs | % | IRS | % | Control | % | Total | %\nNumber of clusters | 44 |  | 44 |  | 44 |  | 44 |  | 176 | \nNumber of households | 1618 |  | 1388 |  | 1527 |  | 1538 |  | 6071 | \nPopulation | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \nPopulation per cluster | 207 |  | 183 |  | 195 |  | 201 |  | 196 | \nAge group (in years)\n\u00a0<\u20095\u00a0years | 1673 | 18.4 | 1528 | 19.0 | 1576 | 18.4 | 1711 | 19.4 | 6488 | 18.8\n\u00a05\u201314\u00a0years | 2840 | 31.2 | 2706 | 33.7 | 2832 | 33.1 | 2758 | 31.2 | 11,136 | 32.2\n\u00a015\u00a0years and older | 4591 | 50.4 | 3804 | 47.3 | 4159 | 48.5 | 4370 | 49.4 | 16,924 | 49.0\nTotal | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \nGender\n\u00a0Male | 4612 | 50.7 | 4006 | 49.8 | 4312 | 50.3 | 4397 | 49.7 | 17,327 | 50.2\n\u00a0Female | 4492 | 49.3 | 4032 | 50.2 | 4255 | 49.7 | 4442 | 50.3 | 17,221 | 49.8\n\u00a0Total | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \nEthnicity\n\u00a0Oromo | 8819 | 96.9 | 7550 | 93.9 | 7627 | 89.0 | 7821 | 88.5 | 31,817 | 92.1\n\u00a0Amhara | 15 | 0.2 | 34 | 0.4 | 34 | 0.4 | 126 | 1.4 | 209 | 0.6\n\u00a0Gurage | 8 | 0.1 | 21 | 0.3 | 164 | 1.9 | 65 | 0.7 | 258 | 0.7\n\u00a0Others | 262 | 2.9 | 433 | 5.4 | 742 | 8.7 | 827 | 9.4 | 2264 | 6.6\n\u00a0Total | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \nMain roof material\n\u00a0Thatch/leaf | 4353 | 47.8 | 3863 | 48.1 | 4035 | 47.1 | 3774 | 42.7 | 16,025 | 46.4\n\u00a0Corrugated iron | 4726 | 51.9 | 4133 | 51.4 | 4484 | 52.3 | 5055 | 57.2 | 18,398 | 53.3\n\u00a0Cement/concrete | 25 | 0.3 | 42 | 0.5 | 48 | 0.6 | 10 | 0.1 | 125 | 0.4\n\u00a0Total | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \nReligion\n\u00a0Orthodox Christian | 717 | 7.9 | 918 | 11.4 | 921 | 10.8 | 937 | 10.6 | 3493 | 10.1\n\u00a0Muslim | 8275 | 90.9 | 6923 | 86.1 | 7437 | 86.8 | 7547 | 85.4 | 30,182 | 87.4\n\u00a0Protestant Christian | 102 | 1.1 | 182 | 2.3 | 199 | 2.3 | 322 | 3.6 | 805 | 2.3\n\u00a0Other | 10 | 0.1 | 15 | 0.2 | 10 | 0.1 | 33 | 0.4 | 68 | 0.2\n\u00a0Total | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \nEducation status\n\u00a0Illiterate | 5187 | 57.0 | 4625 | 57.5 | 5043 | 58.9 | 4897 | 55.4 | 19,752 | 57.2\n\u00a0Read and write only | 866 | 9.5 | 918 | 11.4 | 1139 | 13.3 | 815 | 9.2 | 3738 | 10.8\n\u00a0Primary | 2206 | 24.2 | 1888 | 23.5 | 1820 | 21.2 | 2241 | 25.4 | 8155 | 23.6\n\u00a0Secondary and above | 845 | 9.3 | 607 | 7.6 | 565 | 6.6 | 886 | 10.0 | 2903 | 8.4\n\u00a0Total | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \nSocio-economic status\n\u00a0Lower class | 2887 | 31.7 | 3171 | 39.5 | 2987 | 34.9 | 2414 | 27.3 | 11,459 | 33.2\n\u00a0Middle class | 3084 | 33.9 | 2587 | 32.2 | 2754 | 32.1 | 3153 | 35.7 | 11,578 | 33.5\n\u00a0Upper class | 3133 | 34.4 | 2280 | 28.4 | 2826 | 33.0 | 3272 | 37.0 | 11,511 | 33.3\n\u00a0Total | 9104 |  | 8038 |  | 8567 |  | 8839 |  | 34,548 | \n\n\nCoverage of the interventions of long-lasting insecticidal nets (LLIN) and indoor residual spraying (IRS) in the study arms at different time periods\n | Intervention arms: coverage of interventions\nIRS\u2009+\u2009LLINsN\u2009=\u20091618 | LLINsN\u2009=\u20091388 | IRSN\u2009=\u20091527 | ControlN\u2009=\u20091538 | TotalN\u2009=\u200960,781\nLLIN ownership (at baseline) | 100 | 100 |  |  | 100\nMean LLIN use during the specified period (%)\n\u00a0Weeks 1\u201326 | 47 | 49 |  |  | 48\n\u00a0Weeks 26\u201352 | 26 | 27 |  |  | 27\n\u00a0Weeks 53\u201379 | 8 | 6 |  |  | 7\n\u00a0Weeks 79\u2013121 | 1 | 1 |  |  | 1\nMean IRS coverage during specified period (%)\n\u00a0Weeks 1\u201352 | 96 |  | 97 |  | 97\n\u00a0Weeks 53\u2013104 | 95 |  | 92 |  | 94\n\u00a0Weeks 105\u2013121 | 95 |  | 94 |  | 95\n\nLLIN long-lasting insecticidal nets, IRS indoor residual spraying; N number of households\n\nMalaria incidence rates for the intervention arms for some background variables\n | Intervention arm\nIRS\u2009+\u2009LLIN | LLIN | IRS\nMalaria episodes | Person years | Incidence (95% CI) | Malaria episodes | Person years | Incidence (95% CI) | Malaria episodes | Person years | Incidence (95% CI)\nMalaria episodes\n\u00a0First episode (cases) | 287 | 18,376 | 15.6 (13.9\u201317.5) | 254 | 16,972 | 15.0 (13.2\u201316.8) | 261 | 16,869 | 15.5 (13.7\u201317.4)\n\u00a0All malaria episodes | 321 | 18,713 | 17.2 (15.3\u201319.1) | 278 | 17,244 | 16.1 (14.3\u201318.1) | 291 | 17,153 | 17.0 (15.1\u201319.0)\n\u00a0P. falciparum | 180 | 18,713 | 9.6 (8.2\u201311.0) | 173 | 17,243 | 10.0 (8.5\u201311.5) | 153 | 17,154 | 8.9 (7.5\u201310.3)\n\u00a0P. vivax | 86 | 18,713 | 4.6 (3.6\u20135.6) | 69 | 17,243 | 4.0 (3.1\u20134.9) | 68 | 17,154 | 4.0 (3.0\u20134.9)\n\u00a0Mixed (Pf\u2009+\u2009Pv) | 57 | 18,713 | 3.0 (2.3\u20133.8) | 36 | 17,243 | 2.1 (1.4\u20132.8) | 68 | 17,154 | 4.0 (3.0\u20134.9)\nAge\n\u00a00\u20135\u00a0years | 79 | 3252 | 24.3 (18.9\u201329.6) | 83 | 3105 | 26.7 (21.4\u201333.0) | 54 | 2951 | 18.3 (13.9\u201323.7)\n\u00a06\u201315\u00a0years | 103 | 5975 | 17.2 (14.1\u201320.8) | 91 | 5933 | 15.3 (12.2\u201318.5) | 83 | 5840 | 14.2 (11.2\u201317.3)\n\u00a0Older than 16\u00a0years | 139 | 9481 | 14.7 (12.4\u201317.3) | 104 | 8206 | 12.7 (10.4\u201315.3) | 154 | 8360 | 18.4 (15.5\u201321.3)\nTime period\n\u00a0Week 1\u201326 | 83 | 4071 | 21.1 (17.0\u201326.0) | 56 | 3738 | 15.0 (11.1\u201319.0) | 75 | 3839 | 19.5 (15.1\u201324.0)\n\u00a0Week 26\u201352 | 76 | 4114 | 18.5 (14.3\u201322.6) | 61 | 3765 | 16.2 (12.1\u201320.3) | 54 | 3789 | 14.3 (10.5\u201318.1)\n\u00a0Week 53\u201379 | 62 | 4091 | 15.2 (11.4\u201319.0) | 59 | 3786 | 15.6 (11.6\u201320.0) | 64 | 3682 | 17.4 (13.1\u201321.6)\n\u00a0Weeks 79\u2013121 | 100 | 6437 | 15.5 (12.5\u201318.6) | 102 | 5954 | 17.1 (13.8\u201320.5) | 98 | 5845 | 16.8 (13.5\u201320.1)\nClustering\n\u00a0Cluster area | 104 | 2720 | 38.2 (31.4\u201346.1) | 126 | 2735 | 46.1 (38.5\u201354.7) | 169 | 3513 | 48.1 (41.2\u201355.8)\n\u00a0Non-cluster area | 217 | 15,993 | 13.6 (11.9\u201315.5) | 152 | 14,509 | 10.5 (8.9\u201312.2) | 122 | 13,639 | 8.9 (7.5\u201310.6)\nLLIN use\n\u00a0Over 50% | 13 | 551 | 23.6 (13.1\u201339.3) | 13 | 449 | 29 (16.1\u201348.3) | 0 | 9 | 0.0\n\u00a025\u201349% | 80 | 5618 | 14.2 (11.3\u201317.6) | 72 | 5224 | 13.8 (10.9\u201317.3) | 7 | 135 | 52 (22.7\u2013102.6)\n\u00a00\u201324% | 228 | 12,543 | 18.2 (15.9\u201320.7) | 193 | 11,571 | 16.7 (14.5\u201319.2) | 284 | 17,010 | 16.7 (14.8\u201318.7)\n\n | Control | Total\nMalaria episodes | Person years | Incidence (95% CI) | Malaria episodes | Person years | Incidence (95% CI)\nMalaria episodes\n\u00a0First episode (cases) | 257 | 18,441 | 13.9 (12.3\u201315.7) | 1059 | 70,658 | 15.0 (14.1\u201315.9)\n\u00a0All malaria episodes | 293 | 18,752 | 15.6 (13.9\u201317.5) | 1183 | 71862 | 16.5 (15.5\u201317.4)\n\u00a0P. falciparum | 146 | 18,752 | 7.8 (6.5\u20139.0) | 652 | 71,862 | 9.1 (8.4\u20139.8)\n\u00a0P. vivax | 76 | 18,752 | 4.1 (3.1\u20135.0) | 299 | 71,862 | 4.2 (3.7\u20134.6)\n\u00a0Mixed (Pf\u2009+\u2009Pv) | 71 | 18,752 | 3.8 (2.9\u20134.7) | 232 | 71,862 | 3.3 (2.8\u20133.6)\nAge\n\u00a00\u20135\u00a0years | 54 | 3429 | 15.7 (11.6\u201320.0) | 270 | 12,742 | 21.2 (18.6\u201323.7)\n\u00a06\u201315\u00a0years | 98 | 5979 | 16.4 (13.2\u201320.0) | 375 | 23,727 | 15.8 (14.2\u201317.4)\n\u00a0Older than 16\u00a0years | 141 | 9344 | 15.1 (12.6\u201317.6) | 538 | 35,393 | 15.2 (13.9\u201316.5)\nTime period\n\u00a0Week 1\u201326 | 78 | 4107 | 19.0 (14.8\u201323.2) | 292 | 15,755 | 18.5 (16.4\u201320.7)\n\u00a0Week 26\u201352 | 60 | 4112 | 15.0 (11.0\u201318.3) | 251 | 15,780 | 16.0 (14.0\u201318.0)\n\u00a0Week 53\u201379 | 71 | 4 087 | 17.4 (13.3\u201321.4) | 256 | 15,646 | 16.4 (14.4\u201318.4)\n\u00a0Weeks 79\u2013121 | 84 | 6445 | 13.0 (10.3\u201316.0) | 384 | 24,681 | 15.6 (14.0\u201317.1)\nClustering\n\u00a0Cluster area | 156 | 3,704 | 42.1 (35.9\u201349.1) | 484 | 12,672 | 38.2 (34.9\u201341.7)\n\u00a0Non-cluster area | 137 | 15,049 | 9.1 (7.6\u201310.7) | 575 | 59,190 | 9.7 (8.9\u201310.5)\nLLIN use\n\u00a0Over 50% | 0 | 0 | 0.0 | 26 | 1009 | 25.8 (17.2\u201337.2)\n\u00a025\u201349% | 3 | 94 | 32 (8.1\u201386.9) | 162 | 11,071 | 14.6 (12.5\u201317.0)\n\u00a00\u201324% | 290 | 18,658 | 15.5 (13.8\u201317.4) | 995 | 59,782 | 16.6 (15.6\u201317.7)\n\nLLIN long-lasting insecticidal nets; IRS indoor residual spraying\n\nProportional hazards model comparing incidence of malaria among the arms, adjusted for roof type\n | HR (95% CI)\nArms\n\u00a0LLIN\u2009+\u2009IRS | 1\n\u00a0LLIN only | 0.97 (0.82\u20131.15)\n\u00a0IRS only | 1.01 (0.85\u20131.19)\n\u00a0Routine | 0.92 (0.78\u20131.08)\nType of house roof\n\u00a0Thatched/leaf | 1.18 (1.04\u20131.33)**\n\u00a0Corrugated iron and cement/concretea | 1\n\nLLIN long-lasting insecticidal nets; IRS indoor residual spraying\n**\u2009<\u20090.01\naThere were only 125 individuals living under a cement/concrete roof\n\nPrevalence of anaemia during the three surveys\nSurveys | Number of anaemia cases (haemoglobin\u2009<\u200911\u00a0g/dl | Mean haemoglobin (g/dl) | Anaemia prevalence | OR (95% CI)\n(95% CI) | Percent (95% CI)\nSurvey 1: December 2014\n\u00a0LLINs\u2009+\u2009IRS | 199 | 11.74 (11.63\u201311.85) | 26.8 (23.8\u201330.1) | 1\n\u00a0LLINs only | 220 | 11.52 (11.39\u201311.64) | 28.2 (25.5\u201331.9) | 1.09 (0.87\u20131.37)\n\u00a0IRS only | 199 | 11.54 (11.42\u201311.68) | 29.1 (25.8\u201332.6) | 1.12 (0.89\u20131.41)\n\u00a0Control arm | 223 | 11.54 (11.42\u201311.65) | 28.3 (25.2\u201331.6) | 1.08 (0.86\u20131.35)\nAll | 841 | 11.59 (11.53\u201311.65) | 28.2 (26.6\u201329.8) | \nSurvey 2: December 2015\n\u00a0LLINs\u2009+\u2009IRS | 310 | 11.13 (11.01\u201311.24) | 38.1 (34.8\u201341.5) | 1\n\u00a0LLINs only | 282 | 11.22 (11.10\u201311.34) | 35.0 (31.8\u201338.4) | 0.88 (0.72\u20131.07)\n\u00a0IRS only | 272 | 11.20 (11.07\u201311.32) | 38.5 (35.0\u201342.2) | 1.02 (0.83\u20131.25)\n\u00a0Control arm | 287 | 11.38 (11.27\u201311.50) | 35.8 (32.5\u201339.2) | 0.91 (0.74\u20131.11)\nAll | 1151 | 11.23 (11.17\u20131129) | 36.8 (35.1\u201338.5) | \nSurvey 3: December 2016\n\u00a0LLINs\u2009+\u2009IRS | 240 | 11.58 (11.48\u201311.69) | 29.5 (26.5\u201332.8) | 1\n\u00a0LLINs only | 227 | 11.55 (11.45\u201311.66) | 31.1 (27.8\u201334.6) | 1.08 (0.87\u20131.34)\n\u00a0IRS only | 192 | 11.62 (11.52\u201311.73) | 28.8 (25.4\u201332.2) | 0.96 (0.77\u20131.20)\n\u00a0Control | 236 | 11.57 (11.47\u201311.67) | 29.7 (26.7\u201333.0) | 1.01 (0.81\u20131.25)\nAll | 895 | 11.58 (11.53\u201311.63) | 29.8 (28.2\u201331.4) | \n",
        "label": "low",
        "id": "task4_RLD_test_237"
    },
    {
        "paper_doi": "10.1186/1475-2875-13-208",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Cluster-RCTUnit of randomization: householdICC is not reported.Trial duration: 1 month baseline and 6 months' intervention from April to October 2007.\n\n\nParticipants: Adults or children living in an endemic regionParticipants were screened forP. vivax and parasites were cleared at start.\n\n\nInterventions: Mosquito coils (0.03% transfluthrin) and no treatment.Co-interventions: LLINsTreatment arms:- Control (no treatments) arm - 513 households- 0.03% transfluthrin coils arm - 512 households- LLINs arm - 513 households- LLINs + 0.03% transfluthrin coils arm - 514 households\n\n\nOutcomes: - Participants with malaria parasitaemia confirmed through mRDTs (P. falciparum or P. vivax) and verified by external microscopist through thick film;- Adherence to regular usage of the intervention measured through village leaders' reports and self-reporting; and- Reduction in indoor density of mosquitoes measured through collections using CDC light traps indoor households from the four treatment arms (monthly arithmetic mean of mosquito densities).\n\n\nNotes: Conducted in rural areas of China in the Ruili County, Yunnan Province, close to the Myanmar border.Trial registration number: NCT00442442Funded by SC Johnso\n\n",
        "objective": "To assess the impact of topical repellents, insecticide\u2010treated clothing, and spatial repellents on malaria transmission.",
        "full_paper": "Background\nMosquito coils are the most commonly used household insecticidal product in the world with sales exceeding 50 billion coils, used by two billion people worldwide annually.\nDespite strong evidence that coils prevent mosquito bites a systematic review concluded that there is no evidence that burning mosquito coils prevents malaria acquisition.\nTherefore, the current trial was designed to measure and compare prevention of malaria infection by mosquito coils or long-lasting insecticidal net (LLIN) or a combination of the two in Yunnan, China in the Greater Mekong sub-region.\nMethods\nA four-arm single blind household-randomized design was chosen as coils emanate insecticide throughout the household.\nHouseholds enrolled at baseline were randomly allocated by the lottery method to one of the four intervention arms: (i) nothing, (ii) 0.03% transfluthrin coils alone, (iii) deltamethrin long-lasting insecticide treated nets, (LLINs) alone or (iv) a combination of transfluthrin coils and deltamethrin LLINs.\nAll household members were recruited to the study, with only those households excluded with pregnant or breastfeeding mothers, members with chest complaints or allergies or members that regularly slept away from home.\nThe main outcome of interest was Plasmodium falciparum malaria prevalence detected by rapid diagnostic tests (RDTs) during six repeated monthly cross-sectional surveys.\nThe secondary outcome of interest was the effect on Plasmodium vivax prevalence detected in the same way.\nResults\nA total of 2,052 households were recruited into the study, comprising 7,341 individuals\nThe odds ratios of testing positive by RDT with P. falciparum or P. vivax were >75% lower for all intervention arms compared with the control arm.\nCoils alone provided 77% protection (95% CI: 50%-89%), LLINs provided 91% protection (95% CI: 72%-97%) and the combination of coils and LLINs provided 94% protection (95% CI: 77%-99%) against P. falciparum compared with the control arm.\nThere was no statistically significant difference between the protective efficacies of the different interventions.\nConclusions\nThis is the first robust clinical evaluation of transfluthrin mosquito coils as a means to reduce malaria and the high degree of infection prevented would indicate they represent a potentially highly effective tool, which could be integrated into larger vector control programmes.\nTrial registration\nClinicalTrials.gov Identifier: NCT00442442, March 2007.\nBackground\nMosquito coils are the most commonly used household insecticidal product in the world with sales exceeding 45 to 50 billion coils used by 2 billion people worldwide each year .\nThe popularity of coils is due to their low cost at $0.02 each , their ability to be used without electricity or equipment and cultural acceptance, because smoke is used in many cultures to drive away mosquitoes .\nThese products present a great opportunity for public health, because such products could provide a means of disease control that is already proven highly acceptable to end-users and has undergone stringent safety testing.\nDespite strong evidence that coils prevent mosquito bites, a systematic review concluded that there is no evidence that burning mosquito coils prevents malaria acquisition .\nTherefore, the current trial was designed to measure and compare prevention of malaria infection by mosquito coils or long lasting insecticidal net (LLIN) or a combination of the two.\nThe trial was undertaken in Yunnan, a region of China in the Greater Mekong sub-region where malaria remains endemic despite substantial investment in LLIN programmes  and artemisinin resistant malaria threatens worldwide initiatives to control and eliminate malaria .\nThe product selected for evaluation was a 0.03% transfluthrin mosquito coil (RAID\u00ae, SC Johnson, USA).\nTransfluthrin is a highly effective fast-acting pyrethroid insecticide used extensively in household and hygiene products, mainly against flying insects, such as mosquitoes and flies.\nThe WHO have carried out an evaluation of the extensive toxicity literature available on transfluthrin and concluded that transfluthrin is: \u201cunlikely to present acute hazard in normal use\u201d .\nCentrally organized vector-based malaria control programmes have never incorporated the use of coils as a methodology, presumably as there is a lack of clinical evidence of efficacy, yet there is little doubt over their effects on preventing mosquitoes from biting .\nA large proportion of global malaria transmission occurs before sleeping hours and outdoors due to outdoor and early biting malaria vectors , which are of particular importance in Southeast Asia.\nGiven that mosquito coils are known to be effective at reducing biting, they are relatively cheap, universally available, purchased and used by so many households, there is clearly a need to examine more closely their potential for use in integrated vector control programmes.\nThe study tested the hypothesis that using transfluthrin mosquito coils, LLINs or a combination of both interventions would reduce 1) the incidence of Plasmodium vivax and Plasmodium falciparum malaria and 2) numbers of malaria vector mosquitoes attempting to feed on humans, among those households using the interventions relative to those households not using any of the interventions other than their normal malaria prevention practices.\nMethods\nStudy design\nA household-randomised design was chosen as coils emanate insecticide throughout the household, and LLINs are also know to have a protective effect at the household level.\nField workers and participants were not blinded to treatment allocation, as this was impossible in practice.\nHowever, the field staff collecting monthly RDT data were not aware of the intervention which individuals had been using thus achieving single blinding (investigator) of the study.\nFurthermore, microscopists at Yunnan Institute of Parasitic diseases that verified positive RDTs by miscoscopy and the statistician was blind to the allocation.\nThe intervention of interest was the use of 0.03% transfluthrin coils (SC Johnson), deltamethrin long-lasting insecticide-treated nets (LLINs) (TianJin-Yorkool Ltd, Tianjian, PR China, and Lantrade Global Supplies Ltd, Gerrards Cross, UK), or a combination of transfluthrin coils and deltamethrin LLINs compared with a non-intervention control arm.\nThe study was designed to investigate the benefits of using mosquito coils with or without LLINs on malaria prevention.\nThe main outcome of interest was P. falciparum malaria prevalence detected by rapid diagnostic tests (RDTs) during six repeated monthly cross-sectional surveys.\nThe secondary outcome of interest was the effect on P. vivax prevalence detected in the same way.\nParticipants\nThe study was conducted in Ruili County, Yunnan Province, P.R. China, close to the Myanmar border between April and October 2007.\nYunnan is one of only two provinces in China that remains malaria endemic and Ruili County has a high number of cases due to proximity of heavily forested areas, a high proportion of migrant populations moving over the border between countries and the remote location of minority group habitations which are difficult to cover with centralized vector control and public health programmes.\nAll communities enrolled were in rural areas.\nThe border area stretches 170\u00a0km with a population of 111,449 people.\nThe average elevation was 780 metres with mean temperature of 20\u00b0C and annual average rainfall 1,394.8\u00a0mm.\nDuring 2005 to 2006, 1,125 malaria cases were recorded, of which 31% was P. falciparum.\nThe main malaria vectors are Anopheles minimus and Anopheles sinensis.\nVillages selected were Leiying, Dengga, Leilong, Huyiu, Banling, Hulan, Mendian, Najingli, Ruili farm, due to their relatively high annual malaria incidence.\nThe total population is 4832 households, with 789 malaria cases (approximate annual malaria incidence of 0.02) with 254 cases of P. falciparum (32%) during 2005 .\nPrior to the clinical study there has never been any organized vector control in the area but residents burn fires and mosquito coils and some use untreated bednets for protection against bites .\nRecruitment and allocation\nThe recruitment process began with a meeting between the village head/elders and the local malaria control staff who explained the study to the villagers.\nAs there is a possibility that mosquitoes could be repelled from entering a house using coils or an LLIN and thus being actively diverted to a neighbouring house nearby only those households that were more than 20\u00a0m apart were enrolled, and no two adjacent houses were used.\nIn addition, to ensure potential diversion was kept to an absolute minimum and did not lead to bias a maximum of 20% of the houses in any village were recruited into the study.\nIn this way only 15% of houses in any village had a study intervention, so any diverted mosquitoes would not be measured in study households, but would be absorbed by the large number of houses not participating in the study.\nGiven the protocol inclusion/exclusion criteria of a maximum of 20% of households in any village being available for the study (to negate diversion bias), the number of households/village was calculated.\nFrom this list, those villages that had reported malaria cases in the previous 48\u00a0months were selected.\nOnce houses eligible and happy to participate in the study were identified the study coordinator randomly assigned them to treatment groups by drawing lots.\nMembers of the field team then recruited all household inhabitants to the study on written informed consent.\nHouseholds were excluded if household members included 1) children under six years, 2) pregnant or breastfeeding mothers, 3) those with chest complaints or allergies, and 4) those that regularly slept away from home.\nThis was done to ensure that participants were healthy and not likely to contract malaria outside of the household.\nHowever, in practice children under six years of age were included in the study with their parent or guardian\u2019s consent.\nIntervention\nThose people allocated to LLIN were provided with an new LLIN free of charge for each sleeping space, those allocated to coils were given two coils to burn each evening, those allocated to LLIN\u2009+\u2009coils were given both products and the untreated control group continued to use their own personal protection methods.\nIt would be unethical to ask anyone not to do this but a record was kept of such ad-hoc coil use in the negative control group and those reporting the use of one box or more (10 coils/5 nights) were excluded from the analysis for that round.\nAt each monthly visit, the households in the coil and the coil+LLIN groups were given sufficient coils to last for the next month, and the empty boxes of the previous month were collected as a compliance check.\nHouseholds were permitted to use up to two coils/night as it is common practice for the family to sit in a living area in the early evening and then retire to a separate bedroom area at night.\nWhere houses had separate living and sleeping areas it was thus possible to light one coil early in the evening in the living area, then light another in the sleeping area before the family retired to bed, overcoming problems of moving coils once alight.\nCompliance\nEvery night in the village, the village leader was requested to check whether his village used the coils or nets; and every month a questionnaire was delivered on compliance to household heads enrolled in the study.\nIndividuals were asked about the number of night\u2019s use of the intervention to which they had been allocated (coils, LLINs or both), or use of interventions to which they had not been allocated (coil-use in the control group).\nThere were no LLINs in the control arm.\nAs a further check of compliance the empty boxes of the previous month were collected to confirm they had been used.\nData collection\nA baseline survey was undertaken among household heads, along with the first cross-sectional malaria survey after randomization in April 2007 to collect data on socio-demographic factors including age, sex, level of education, occupation and history of malaria in the previous 12\u00a0months.\nPresence of malaria infection among all household members was detected using CareStart pLDH Malaria G0121 (AccessBio Inc., Monmouth, USA) rapid diagnostic tests that distinguished P. falciparum and P. vivax parasites during six cross-sectional surveys carried-out monthly between May and October 2007.\nFor those RDTs that showed a positive result a thick film blood slide was taken for verification by Microscopy at Yunnan Institute of Parasitic Diseases.\nData were also collected on the reported history of fever in the previous month and use of treatment including anti-malarial drugs.\nAll positive cases were referred to the local health centre to receive appropriate anti-malaria treatment.\nTreatment for P. vivax was chloroquine (1.5\u00a0g)\u2009+ primaquine (180\u00a0mg), while treatment for P. falciparum was artesunate (600\u00a0mg)\u2009+\u2009primaquine (45\u00a0mg).\nThe protocol stated that patients should be excluded if they contracted malaria to prevent bias from household clustering of malaria infections.\nHowever, on analysis, data showed that there was a protocol deviation and no patients that contracted malaria were excluded in practice.\nAll malaria cases were treated with primaquine, which clears hypnozoites from the liver, excluding the possibility of recrudescence biasing data from the trial on malaria incidence.\nEntomology\nAs an additional outcome measure a series of entomological monitoring studies were included within the main clinical trial.\nOne of the main villages, Dennggar, included in the study was chosen as a sentinel site where entomological collections were made over a 4-month period mid study (July \u2013 October).\nStandard CDC light traps (Bioquip Inc., USA) were hung indoors close to an occupied bed each night in four houses from each of the four treatment groups (n\u2009=\u200916).\nTraps were set in the evening before dusk and left running all night.\nIn the morning, all mosquitoes were collected from each trap, knocked down by chilling, then identified by local experts.\nThe number of mosquitoes indoors was recorded in each study group to determine whether any treatment was significantly reducing mosquito entry into houses.\nData were analysed as a proportion of mosquitoes, per house, per night.\nIt should be noted that for the control arm, houses with untreated bednets were selected as the controls to ensure that the CDC light trap catch was optimised and could be compared to the protective efficacy of each the interventions to the user sleeping next to the light trap.\nSample size\nA sample size of 400 households per intervention arm, each with an average of five individuals, was calculated to detect a 50% reduction in P. falciparum incidence in the households using transfluthrin coils compared with those not allocated to an intervention with 90% power and 95% significance .\nA total of 2,052 households were recruited into the study at baseline (target number was 1,600) comprising 7,341 individuals (target number was 8,000).\nHouseholds enrolled at baseline were randomly allocated by the lottery method to one of the four intervention arms (i) nothing, (ii) coils alone, (iii) LLINs alone or (iv) coils and LLINs.\nAs the study was clustered at the household level the increase in numbers of households recruited was a positive occurrence.\nThe slightly lower total number of individuals was presumably due the smaller family size per house of four rather than five, as used in the sample size estimate.\nAs there is a possibility that mosquitoes could be repelled from entering a house using coils or an LLIN and thus being actively diverted to a neighbouring house nearby only those households that were more than 20\u00a0m apart were enrolled, and no two adjacent houses were used.\nIn addition, to ensure potential diversion was kept to an absolute minimum and did not lead to bias we only recruited a maximum of 20% of the houses in any village into the study.\nIn this way only 15% of houses in any village had a study intervention, so any diverted mosquitoes would not be measured in study households, but would be absorbed by the large number of houses not participating in the study.\nData analyses\nThe data were analysed using Stata 11.0 (StataCorp. 2009.\nStata Statistical Software: Release 11. College Station, TX: StataCorp LP.).\nBaseline socio-demographic variables (age, sex, education level of adults) were compared between the intervention arms.\nThe main analysis was an intention-to-treat analysis according to the groups to which individuals had been randomised.\nNo stratification by village was used.\nA secondary per-protocol analysis restricted data to those with\u2009>\u200990% reported compliance for the month prior to each cross-sectional survey, equivalent to non-compliance of three days or less per month.\nA random-effects logistic regression model adjusting for repeated sampling of the same individual was used to identify potential confounding factors such as age group and level of education among adults in the control group.\nA multilevel, mixed-effects logistic regression model was used to estimate odds ratios (OR) of each intervention (fixed effect) compared with the control arm, using nested random effects to adjust for the non-independence of observations from individuals in the same household, and for repeated sampling of the same individual.\nThe protective efficacy was estimated as (1 \u2013 OR) \u00d7 100%.\nEthics\nLSHTM University of London Ethics Committee and the Yunnan Bureau of Health approved the study.\nAt the end of the study, all households that had not been provided with an LLIN (the coil only group and the negative control group) were given LLINs as required by the ethics committee to ensure parity among all participants.\nResults\nBaseline data\nOver 75% of the study participants were over 16\u00a0years of age, 75% of these had primary level education or above and the main occupation of the head of household was as a farmer (83%).\nAge, gender, level of education among adults and reported history of malaria in the previous 12\u00a0months was evenly distributed between the intervention arms (Table\u00a0 1).\nReported history of malaria in all the arms corresponded to that at County level (0.02 \u2013 0.03 cases per year).\nLoss to follow up was less than 2% in each of the four treatment arms, with lowest loss to follow up in the LLIN only arm (1.5%), loss to follow up of 2% in the coil only arm and 1.9% in the coil & LLIN arm although these were not significantly different (Figure\u00a0 1).\nMalaria infection and potential confounders\nThe prevalence of P. falciparum infection was generally low, ranging from 0.11 to 0.66% in the control arm, while P. vivax infection prevalence varied from 0 to 0.72% in the control arm (Table\u00a0 2).\nThere were no mixed species infections.\nNo individuals were positive for P. falciparum more than once, while four individuals tested positive for P. vivax twice and one individual tested positive three times.\nAn investigation of potential risk factors for malaria in the control group identified a weak association between age and P. falciparum infection with those aged \u226515\u00a0years having an odds ratio of 4.31 compared with those aged 2\u201314 years (95% CI: 0.93, 19.98; p\u2009=\u20090.062).\nThere was also some weak evidence of an association between level of education among adults and P. falciparum infection in the control group.\nThose with primary education had an odds ratio of 0.40 compared with those with no or limited education (95% CI: 0.15, 1.04; p\u2009=\u20090.061), but this effect was not significant for those with secondary or higher education (p\u2009=\u20090.148).\nNeither covariable showed a significant association with P. vivax prevalence in the control arm.\nEfficacy of the intervention\nThe unadjusted odds ratios of testing positive by RDT with P. falciparum or P. vivax were considerably lower for all intervention arms compared with the control arm (see Table\u00a0 3).\nThe confidence intervals overlapped between all arms, suggesting no significant difference between the protective efficacies of the different interventions, although the magnitude of effect was greatest for the combined intervention of LLINs plus coils.\nAs there were no repeat positives or evidence of household clustering for P. falciparum, the mixed-effects, multi-level model gave almost identical results to a random effects model adjusting only for repeated observations of the same individual.\nAge and level of education among adults (\u226515\u00a0years old) were tested separately for inclusion in a multivariable random effects model, and education was no longer significant (p\u2009>\u20090.1).\nAdults were nearly five times more likely to be infected with P. falciparum than children, with an odds ratio of 4.86 (95% CI: 1.18, 20.05; p\u2009=\u20090.029).\nThe odds ratios for the interventions, adjusted for the effect of age, were almost identical to the unadjusted odds ratios (Table\u00a0 3).\nAs there was no evidence of an association of either age or education with P. vivax infection among controls, these factors was not tested for inclusion in a multivariable model of P. vivax infection.\nThe level of compliance with the allocated interventions was high: > 94% of individuals used the coils and/or LLINs for\u2009>\u200990% of the month prior to the surveys.\nConversely, those in the control arm were less likely to follow the request of the study directors to not use any intervention, with 13-19% using local coils for 3 or more days in the month prior to the survey.\nA per-protocol analysis including only those with\u2009>\u200990% compliance gave almost identical results to the intention-to-treat analysis.\nEntomology\nMosquito densities were lower in all treatment groups with a >80% reduction from the use of LLINs or mosquito coils compared to an untreated bed net (Table\u00a0 4).\nDiscussion\nCompared to the untreated control arm, both P. falciparum and P. vivax malaria cases in the group using mosquito coils were reduced by more than 75%, a remarkably high reduction in infection.\nThe level of malaria prevention afforded by coils was very similar to that of the LLIN in the same study.\nWhen the two methods were employed in combination, the reduction in malaria infection was increased further to more than 90%.\nThis may indicate that malaria in this area is likely to be transmitted by more than one vector with a mixture of biting behaviour over a prolonged period of the evening and night.\nIndeed, entomological data collected from the area at the same time as the trial (July 2007) indicated that An. sinensis was the most abundant malaria vector comprising 50% of Anophelines collected, in addition to smaller numbers of Anopheles kochi, Anopheles splendidus, Anopheles barbirostris, Anopheles vagus, Anopheles jeyporiensis, Anopheles annularis, Anopheles philippinsis, Anopheles minimus, Anopheles tessallatus, Anopheles maculatus, Anopheles barbumbrosus, Anopheles dirus and Anopheles culicifacies.\nOne might speculate that the coils are having a greater effect in the evening before people retire to bed, and that the LLIN is further preventing infection later into the night, but this would need additional investigation.\nGiven that coils are universally sold and used in most malaria risk areas globally, they are readily accepted by the local population (as they reduce nuisance biting) and are relatively inexpensive at around 2 cents per coil, it would seem that the high reduction in infections demonstrated here could make transfluthrin mosquito coils suitable for inclusion in some vector-borne disease control programmes.\nAs only a single type of coil (Raid\u00ae) was evaluated, which is known to be of high quality, the results obtained should not automatically be extrapolated to other products available in the market.\nSimilarly, the study was conducted in rural areas of S.E. Asia and other locations are likely to have different habitats and vector species , which are likely to influence the success of coils to some degree.\nThe current study investigated entomological and malaria infection outcomes and monitored use and acceptance of the various interventions.\nHowever, there was no attempt to measure clinical malaria or to record other physical or physiological effects of either the LLIN or the smoke from the coils  and these aspects may need to be taken into consideration if coil use were to be promoted more widely.\nConclusions\nThis is the first robust clinical evaluation of coils as a means to reduce malaria and the high reduction in infections achieved would indicate that they represent a potentially highly effective tool, which could be implemented at the household or community level, or integrated into larger vector control programmes.\nStudy flowchart. Numbers in parentheses are for per-protocol analysis i.e. >90% compliance in previous month with intervention allocated.\n\nBaseline socio-demographic characteristics of individuals enrolled by intervention arm\n\u00a0 | Control | Coils | LLINs | Coils\u2009+\u2009LLINs\nIndividuals enrolled | 1841 | 1843 | 1828 | 1901\nFemale (%)\u00a7 | 919 (49.9) | 911 (49.4) | 918 (50.2) | 924 (48.6)\nMean age (years) [s.d.] | 34.2 [17.9] | 33.9 [18.2] | 34.3 [17.7] | 34.5 [18.2]\n\u226515\u00a0years old (%)\u00a7 | 1464 (79.7) | 1428 (77.5) | 1451 (79.3) | 1498 (78.9)\nLevel of education if \u226515\u00a0years old (%)\u00a7 | n\u2009=\u20091456 | n\u2009=\u20091425 | n\u2009=\u20091434 | n\u2009=\u20091497\nNone/limited | 384 (26.4) | 338 (23.7) | 358 (25.0) | 378 (25.2)\nPrimary | 444 (30.5) | 471 (33.1) | 424 (29.6) | 495 (33.1)\nSecondary | 628 (43.1) | 616 (43.2) | 652 (45.5) | 624 (41.7)\nHistory of malaria in previous 12\u00a0months (%)\u00a7 | 61 (3.3) | 43 (2.4) | 43 (2.4) | 63 (3.3)\n\n\u00a7Discrepancies in percentages are due to missing data for some characteristics.\n\nNumber (%) positive over results available by rapid diagnostic test from of six repeated monthly cross-sectional surveys\n\u00a0 | Control | Coils | LLINs | Coils\u2009+\u2009LLINs\n\u00a0 | P. falciparum | P. vivax | P. falciparum | P. vivax | P. falciparum | P. vivax | P. falciparum | P. vivax\nSurvey 1 | 5/1819 (0.27) | 0/1819 (0) | 1/1831 (0.05) | 0/1831 (0) | 1/1819 (0.05) | 0/1819 (0) | 1/1878 (0.05) | 0/1878 (0)\nSurvey 2 | 3/1795 (0.17) | 4/1795 (0.22) | 2/1812 (0.11) | 1/1812 (0.06) | 0/1799 (0) | 3/1799 (0.17) | 0/1862 (0) | 1/1862 (0.05)\nSurvey 3 | 6/1812 (0.33) | 13/1812 (0.72) | 0/1830 (0) | 2/1830 (0.11) | 1/1811 (0.06) | 0/1811 (0) | 0/1887 (0) | 0/1887 (0)\nSurvey 4 | 12/1824 (0.66) | 3/1824 (0.16) | 3/1829 (0.16) | 4/1829 (0.22) | 1/1814 (0.06) | 2/1814 (0.11) | 0/1880 (0) | 1/1880 (0.05)\nSurvey 5 | 7/1806 (0.39) | 9/1806 (0.50) | 1/1814 (0.06) | 1/1814 (0.06) | 0/1808 (0) | 3/1808 (0.17) | 1/1875 (0.05) | 1/1875 (0.05)\nSurvey 6 | 2/1801 (0.11) | 9/1801 (0.50) | 1/1821 (0.05) | 0/1821 (0) | 0/1803 (0) | 1/1803 (0.6) | 0/1870 (0) | 0/1870 (0)\n\n\nOdds ratios and protective efficacy of interventions using an intention to treat analysis, adjusting for household clustering and repeated observation of individuals with mixed-effects logistic regression\n\u00a0 | Control | Coils | LLINs | Coils\u2009+\u2009LLINs\nOdds Ratio of being P. falciparum positive | 1 | 0.23 | 0.09 | 0.05\n(95% Confidence Interval [CI]) | - | (0.10,0.49) | (0.03, 0.28) | (0.01, 0.23)\nAge-adjusted OR | - | 0.23 | 0.09 | 0.06\n(95% CI) | - | (0.11, 0.50) | (0.03, 0.28) | (0.01, 0.23)\np-value\u00a7 | \u00a0 | 0.0002 | <0.0001 | <0.0001\nProtective efficacy | \u00a0 | 77% | 91% | 94%\n(95% CI) | \u00a0 | (50, 89) | (72, 97) | (77, 99)\nOdds Ratio of being P. vivax positive | 1 | 0.20 | 0.21 | 0.07\n(95% Confidence Interval [CI]) | - | (0.09, 0.44) | (0.10, 0.47) | (0.02, 0.24)\np-value | - | <0.0001 | 0.0001 | <0.0001\nProtective efficacy | \u00a0 | 80% | 79% | 93%\n(95% CI) | \u00a0 | (56, 91) | (53, 90) | (76, 98)\n\n\u00a7P-values for unadjusted and age-adjusted odds ratios were identical.\n\nArithmetic mean monthly indoor mosquito catch/house/night\n\u00a0 | Control | LLIN | Coils | Coils\u2009+\u2009LLIN\nJuly | 22 | 6 | 5 | 3\nAugust | 38 | 8 | 2 | 4\nSeptember | 15 | 1 | 2 | 0\nOctober | 12 | 1 | 1 | 0\nMean | 21.75 | 4 | 2.5 | 1.75\nStd. Dev. | 11.6 | 3.6 | 1.7 | 2.1\n% reduction | - | 82% | 88% | 92%\n",
        "label": "low",
        "id": "task4_RLD_test_289"
    },
    {
        "paper_doi": "10.1007/bf01992163",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Trial design: randomized\nTime period/duration of trial: unclear, before 1993\nDuration of follow-up: 2 months\n\n\nParticipants: Setting: unclear\nLocation: unclear\nAge: 42 adult participants > 16 years (mean age: 28.2 years in ceftriaxone group, 26.8 years in ciprofloxacin group)\nGender: no details given\nHealth status of participants: not recorded\nInclusion criteria:blood culture positiveacute S typhi infectionExclusion criteria:inability to take oral medicationpossible or proven pregnancylack of fever at time of admission\n\n\nInterventions: Ceftriaxone: IV, 3 g once daily for 7 daysCiprofloxacin: oral, 500 mg twice daily for 7 days\n\n\nOutcomes: Clinical cure: clinical failure defined as fever > 38 degC after 7 days of therapy or who deteriorated clinically after 5 full days of therapy; cure if patient afebrile and asymptomatic on or before day 7 and did not require additional therapy during 2 months of follow-upRelapse: readmission for typhoid within 2 months of discharge with a positive blood or stool culture for S typhi with the same antibiogram as previousConvalescent faecal carriage: not defined, but stool culture positivity assessed at 28 days post-enrolment\n\n\nOrganism type and antimicrobial susceptibility: S typhi = 42 (MDR S typhi = 22), S paratyphi = 0\n\n\nNotes: The trial was terminated when the clinicians involved in the trial felt that it was no longer ethical to randomize patients to receive ceftriaxone, given the higher cost, need for IV access and perceived lower efficacy of this regime\n\n",
        "objective": "To evaluate the effectiveness of cephalosporins for treating enteric fever in children and adults compared to other antimicrobials.",
        "full_paper": "iprofloxacin versus Ceftriaxone in the Treatment of ultlresistant Typhoid Fever\niprofloxacin versus Ceftriaxone in the Treatment of ultlresistant Typhoid Fever\nA randomized trial comparing ceftriaxone (3 g given parenterally per day for 7 days) to ciprofloxacin (500 mg given orally twice a day for 7 days) in the treatment of blood culture positive typhoid fever was conducted.\nTwenty patients were openly randomized to receive ciprofloxacin and 22 to receive ceftriaxone.\nThe outcome was classified as clinical failure in 6 patients (27 %) in the ceftriaxone group, but in none in the ciprofloxacin group (p = 0.01).\nThe mean duration of fever was four days in the ciprofloxacin group and about five days in the ceftriaxone group (p = 0.04).\nIn the six patients in the ceftriaxone group who experienced failure, therapy was switched to ciprofloxacin and the patients became afebrile and asymptomatic within 48 hours.\nPatients with resistant strains of Salmonella typhi and patients with sensitive strains responded equally well to ciprofloxacin therapy.\nAnalysis of a subset of 12 of the multiresistant strains revealed that resistance was encoded for by a transferable 180 kilobase plasmid.\nCiprofloxacin represents a useful treatment option in areas where multiresistant strains are likely to be encountered.\nTyphoid fever is traditionally treated with a two-Week course of either chloramphenicol, cotrirnoxazole or ampicillin/amoxiciilin.\nDespite minor differences in toxicity, duration of fever, carriage and relapse rate, these agents are roughly equal in clinical efficacy (1).\nResistance to these agents has occurred sporadically over the past two de-Cades in a variety of locations (1,2), but beginning in 1989, Salmonella typhi strains resistant to all three standard antimicrobial agents have been reported with alarming frequency from locations as diverse as the Indian subcontinent, the Arabian (Persian) Gulf, the UK and China (1)(2)(3)(4).\nThese multiresistant strains are fully pathogenic, often Causing illness more severe than that due to sensitive strains (5).\nTreatment of typhoid caused by multiresistant Salmonella typhi strains is not standardized.\nBoth third-generation cephalosporins (particularly ceftriaxone) and the new fluoroquinolones have been used with some success (1,2), but no study has yet been published which makes a direct comparison of these two classes of antibiotics in the therapy of typhoid fever.\nIn response to the rapid dissemination of multiresistant Salmonella typhi in Bahrain (4), we initiated a trial comparing oral ciprofloxacin with parenteral ceftriaxone for the treatment of typhoid fever.\nWe also investigated the nature of antibiotic resistance in selected multiresistant Salmonella typhi recently introduced into the Arabian Gulf area.\nPatients and Methods\nPatients.\nResults\nPatients and Bacterial Strains: A total of 43 patients met the study entry criteria.\nOne patient was subsequently excluded when he proved to have active tuberculosis in addition to Salmonella typhi bacteremia and failed to become afebrile after more than two weeks of therapy with both ciprofloxacin and ceftriaxone.\nOf 42 evaluable patients, 20 were randomized to receive ciprofloxacin and 22 to receive ceftriaxone.\nThere were no significant differences when the treatment groups were compared by age, duration of fever prior to admission, prevalence of multiresistant strains, white blood cell count, serum sodium level, hematocrit, or proportion of patients with diarrhea, constipation, splenomegaly or occult blood in the stool (Table 1).\nAll Salmonella typhi isolates were sensitive to both ceftriaxone and ciprofloxacin when tested by the Kirby-Bauer technique.\nTriresistant strains were defined as those resistant to cotrimoxazole, ampicitlin and chloramphenicol, whereas sensitive strains were sensitive to all three antibiotics.\nNo isolate had an intermediate pattern of sensitivity.\nOutcome of Therapy.\nThere were no cases of clinical failure in the ciprofloxacin group, whereas there were six cases of failure in the ceftriaxone group (p = 0.01).\nAll ciprofloxacin patients were asymptomatic and afebrite by day 6 (mean: 4 days of fever).\nThe ceftriaxone group required a significantly longer time for resolution of fever (mean: about 5 days, p = 0.04).\nFive of the six cases of failure in the ceftriaxone group were thus classified because of persistent fever after a full seven days of therapy; the sixth patient had fever and persistent severe neuropsychiatric symptoms on day 6 of ceftriaxone therapy and was deemed a case of clinical failure by the investigators (Table 2).\nAll six cases of failure in the ceftriaxone group\nWere subsequently allocated to receive ciprofloxacin therapy.\nThese six patients were afebrile and asymptomatic within 48 hours, and all patients had an uneventful recovery while completing a one-week course of ciprofloxacin.\nBlood cultures were done on day 3 of initial therapy in all 42 study patients and on day 8 in the six patients receiving ciprofloxacin after failure of ceftriaxone; all cultures were negative.\nOne patient in the ceftriaxone group experienced relapse four weeks after therapy, both blood and Stool cultures being positive for a SalmoneUa typhi strain with the same antibiogram as the initial isolate.\nOne patient in the ciprofloxacin group was readmitted with fever eight weeks after discharge.\nThis patient's stool grew a sensitive Salmonella typhi; however, her prior isolate was triresistant and infection was attributed to reinfection rather than relapse.\nAll other patients had negative stool cultures four weeks after therapy and did not relapse within a two-month follow-up period.\nThe study was terminated when the clinicians in-Volved in the study felt that it was no longer ethical to randomize patients to receive ceftriaxone, given the higher cost, need for intravenous access and lower efficacy of this regimen.\nPhage Types and Drug Resistance.\nSeven of the 12 randomly selected strains belonged to Vi-phage type El, three to type M1, one to type A, and one to type 51.\nAll these strains were resistant to chloramphenicol, ampicillin and trimethoprim, but were sensitive to ceftriaxone, nalidixic acid and ciprofloxacin.\nThe complete spectrum of resistance was transferable at 28 \u00b0C but not at 37 \u00b0C.\nIn all cases, the spectrum of resistance was en-\nDiscussion\nIn this study, ciprofloxacin given orally produced more rapid and reliable resolution of fever than parenteral ceftriaxone.\nProlonged fever in ceftriaxone treated typhoid patients has been observed in other studies (10,11) and may reflect the relatively poor intracellular penetration of cephalosporins.\nCiprofloxacin, with its excellent intracellular penetration, has been almost uniformly successful in the treatment of typhoid caused by both sensitive and resistant Salmonella typhi isolates (1,12,13).\nThe short course (7 days) of ciprofloxacin used in our study was efficacious and not associated with a high rate of stool carriage or relapse, thus having significant advantages compared to the longer two-week course of traditional agents.\nBecause of these advantages, ciprofloxacin has recently been advocated in the UK as the drug of choice for treatment of typhoid in patients with a high pretreatment likelihood of infection with strains resistant to traditional agents (14).\nDespite its advantages, ciprofloxacin does have appreciable drawbacks.\nIts use in children and pregnancy is controversial due to concern about possible cartilage injury.\nWhile much less expensive than ceftriaxone, ciprofloxacin is still more expensive than oral drugs of choice used in the past.\nAlthough touted as a drug for treatment of typhoid carriers (15), its failure to reliably eliminate stool carriage in a recent outbreak of Salmonella java is also disquieting (16).\nMost disconcerting of all is a report from India of decreasing susceptibility of Salmonella typhi to ciprofloxacin and the need for higher doses (1.5 g/day) to achieve a cure (17).\nIn spite of these potential problems, on the basis of the findings of this study, oral ciprofloxacin (500 mg orally b.i.d.) can be recommended for the initial therapy of typhoid in areas where resistant strains are responsible for a sizeable proportion of cases of typhoid fever.\nIt is an effective oral drug which can cure typhoid in a one-week course of therapy.\nThe rapid spread of multiresistant typhoid fever over large geographic areas presents multiple challenges, especially in less developed countries where access to newer and more expensive antimicrobial agents may be limited.\nFurther research efforts must continue to focus on oral agents with good intracellular penetration which can be used for short courses of therapy with the chances of a high cure rate.\nTable 1 : Comparison of the two treatment groups at study entry.\nCipro- | Ceftri-\nfloxacin | axone\n(n = 20) | (n = 22)\n,...._.\nTable 2 : Outcome of therapy in the two treatment groups.\n | Cipro-Ceftriaxone P value\n | floxacin | (n = 22) | \n | (n=20) |  | \nClinical failure | 0/20 | 6/22 | 0.01\nRelapse | 0 | 1 | NS\nDays to resolution |  |  | \nof fever (mean) | 4.0 | 5.2 | 0.04\nNS: not significant. |  |  | \nAcknowledgements\nThe authors kindly thankDr. E.C, Oldfield Ili for his critical review.\nWe also thank Ms. C. Cesefia for her help in manuscript preparation and the entire staff of Salmaniya Medical Centre for their cooperation in this study.\nSpecial thanks to Dr. A. L. Bourgeois for his help with the bacterial isolates.\nThe Chief, Navy Bureau of Medicine and Surgery, Washington, DC, Clinical Investigation Program, sponsored this study No. 84-16-1968-366, Operation Desert Storm Scholarly Work, as required by HSETCINST 6000.41.\nThis work was also supported by the Naval Medical Research and Development Command, Naval Medical Command, National Capital Region, Bethesda, MD (work unit No. 3M162770AR122), the Ministry of Health of Bahrain, and Naval Medical Research Unit No. 3, Cairo, Egypt (work unit No. 3M162787A870AN121).\nThe opinions and assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the views of the United States Department of the Navy, the Department of Defense or the United States Government.",
        "label": "high",
        "id": "task4_RLD_test_207"
    },
    {
        "paper_doi": "10.7189/jogh.09.020402",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: DesigncRCTAllocation of clusters50 schools randomized to intervention, 50 to control\n\n\nParticipants: 9258 primary school-aged children\n\n\nInterventions: Broad multiple\n\n\nOutcomes: Any STH\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Background\nWater, sanitation, and hygiene (WASH) in schools is promoted by development agencies as a modality to improve school attendance by reducing illness.\nDespite biological plausibility, the few rigorous studies that have assessed the effect of WASH in schools (WinS) interventions on pupil health and school attendance have reported mixed impacts.\nWe evaluated the impact of the Laos Basic Education, Water, Sanitation and Hygiene Programme \u2013 a comprehensive WinS project implemented by UNICEF Lao People\u2019s Democratic Republic (Lao PDR) in 492 primary schools nationwide between 2013 and 2017 \u2013 on pupil education and health.\nMethods\nFrom 2014-2017, we conducted a cluster-randomized trial among 100 randomly selected primary schools lacking functional WASH facilities in Saravane Province, Lao PDR.\nSchools were randomly assigned to either the intervention (n\u2009=\u200950) or comparison (n\u2009=\u200950) arm.\nIntervention schools received a school water supply, sanitation facilities, handwashing facilities, drinking water filters, and behavior change education and promotion.\nComparison schools received the intervention after research activities ended.\nAt unannounced visits every six to eight weeks, enumerators recorded pupils\u2019 roll-call absence, enrollment, attrition, progression to the next grade, and reported illness (diarrhea, respiratory infection, conjunctivitis), and conducted structured observations to measure intervention fidelity and adherence.\nStool samples were collected annually prior to de-worming and analyzed for soil-transmitted helminth (STH) infection.\nIn addition to our primary intention-to-treat analysis, we conducted secondary analyses to quantify the role of intervention fidelity and adherence on project impacts.\nResults\nWe found no impact of the WinS intervention on any primary (pupil absence) or secondary (enrollment, dropout, grade progression, diarrhea, respiratory infection, conjunctivitis, STH infection) impacts.\nEven among schools with the highest levels of fidelity and adherence, impact of the intervention on absence and health was minimal.\nConclusions\nWhile WinS may create an important enabling environment, WinS interventions alone and as currently delivered may not be sufficient to independently impact pupil education and health.\nOur results are consistent with other recent evaluations of WinS projects showing limited or mixed effects of WinS.\nSchool-aged children in low-income settings are at substantial risk for water, sanitation, and hygiene (WASH)-related infections such as pathogens causing diarrheal diseases, soil-transmitted helminths (STH), and trachoma.\nCrowded, unsanitary conditions may facilitate the spread of pathogens, and increase pupils\u2019 risk for disease.\nImproved access to WASH facilities combined with sufficient behavior change may not only prevent the spread of pathogens within the school domain but also lead to beneficial WASH habits at home and throughout the life course.\nThe limited data available indicate that only 69% of schools worldwide have access to sanitation facilities, while only 66% have access to water.\nWASH in schools (WinS) targets and indicators have been included in the Sustainable Development Goals.\nDespite the biological plausibility of WinS interventions to reduce illness and subsequently school absence, evidence of impact has been mixed.\nSome WinS efficacy studies, such as those assessing intensive handwashing programs in China and Egypt, reported reductions in absence and absence due to illness.\nHowever, with only 6- and 3-month follow up periods, respectively, and with soap being continuously supplied by the intervention or school administration, respectively, the long-term sustainability of handwashing behaviors linked to these impacts is unknown.\nEffectiveness trials of WinS projects have not replicated this success.\nA matched-control evaluation of a comprehensive WinS program in Mali revealed reductions in pupil-reported diarrhea, symptoms of respiratory infection, and absence due to diarrhea, but higher odds of absence overall among pupils enrolled in beneficiary schools.\nHowever, there were imbalances between the beneficiary and comparison groups at baseline, and the study was further limited by inconsistent fidelity to the intervention by implementing partners and participating schools.\nA randomized controlled trial (RCT) of a WinS program in Kenya reported a 44% reduction in odds of Ascaris lumbricoides reinfection, but no overall impact on absence or diarrhea.\nProgram impact differed by intervention arm (as individual and combined WASH interventions were employed) and subsets of the sample population.\nAbsence among girls in the hygiene promotion and water treatment arm reduced by 58%.\nIn water-scarce schools that received a comprehensive WASH intervention, including water supply improvements, risk of diarrhea among pupils reduced by 61%, while diarrhea among pupils\u2019 siblings under 5 years old reduced by 56%.\nHowever, program impact may have been affected by incomplete and inconsistent intervention delivery (fidelity) and uptake and use by the target population (adherence).\nA WinS intervention in Lao People\u2019s Democratic Republic (Lao PDR), Cambodia, and Indonesia had no impact on STH infection or being underweight, but reported evidence of improvement in dental cavities.\nAgain, this evaluation was potentially limited by incomplete fidelity and adherence to the program, as well as a non-randomized design and contamination from concurrent programming in control schools.\nHere, we present results from the Water, Sanitation, and Hygiene for Health and Education in Laotian Primary Schools (WASH HELPS) study, a cluster-RCT designed to measure the impact of a comprehensive WinS project \u2013 water supply, sanitation, handwashing, and behavior change - in Lao PDR on pupil absence, diarrhea, respiratory infection, and STH infection.\nGiven past challenges in program fidelity and adherence to project outputs and behaviors, we also apply two analyses that have previously been used to evaluate the role of intervention fidelity and adherence on WinS project impacts.\nMETHODS\nStudy setting and intervention\nThe Laos Basic Education, Water, Sanitation and Hygiene Programme was implemented by UNICEF in 492 primary schools across thirteen provinces between 2013 and 2017.\nThe WASH HELPS Study, a research component of the intervention, was conducted between September 2014 and May 2017 in Saravane Province, which was selected because it was the only province in which intervention activities had not yet occurred, thus allowing a randomized intervention trial.\nThe study setting, baseline results, intervention components, intervention outputs and outcomes, and their fidelity and adherence have been described in detail elsewhere.\nKey outputs and outcomes of the project are listed in Table 1.\nBriefly, the comprehensive WinS project included provision of a school water supply, sanitation facilities, handwashing facilities (individual and group), drinking water filters, and behavior change education and promotion.\nThe project was implemented in two phases; lessons learned from Group 1 schools (n\u2009=\u200952; intervention started in 2014) were applied to improve the project for Group 2 schools (n\u2009=\u200948; intervention started in 2015), leading to different levels of achievement at output and outcome levels between groups, as well as different durations of follow-up.\nStudy design, sampling, and data collection\nWe conducted a cluster-randomized, controlled trial among 100 primary schools (50 intervention, 50 comparison).\nStudy design, sampling, and data collection methods have been previously published.\nWe used stratified random sampling to help ensure equal representation of control and intervention schools in each district, and that the number of schools selected in each district was proportional to the number of eligible schools in each district.\nWe selected up to 40 pupils from grades 3-5 in each school using systematic stratified sampling, with grade and sex as the stratification variables.\nPupils selected at baseline were followed throughout the entire study period; pupils who left the school due to abandonment or transfer were replaced at the beginning of the following academic year, maintaining equal grade and sex ratios when possible.\nPupils who progressed from fifth to the sixth grade were replaced with pupils from grade three the following academic year.\nA total of 3993 pupils were enrolled throughout the study period.\nData were collected over three or two school years (Group 1 and 2 schools, respectively) to measure uptake and sustainability of facilities and behavior change.\nTo account for variabilities across time and season, data were collected throughout the school year, which consists of 33 weeks across two semesters (September-January and February-May), with five to six hours of instruction per day.\nTrained enumerators visited study schools every six to eight weeks during the school year through March 2017, for a total of 11 (Group 1) or 7 (Group 2) visits per school.\nAll visits were unannounced and during school hours.\nAt each visit, enumerators conducted a roll call of all students enrolled in the school using sex- and grade-specific ledgers; interviewed the school directors; interviewed sampled pupils in grades 3\u20135; observed conditions and functionality of WinS hardware; and observed individual and group handwashing practices.\nEach year, stool samples were collected from up to 50 pupils per school prior to distribution of preventative chemotherapy as part of the National School Deworming Programme.\nStool samples were tested for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus) using the Kato Katz technique.\nMeasures\nOur primary impact of interest was pupil absence measured by school-wide roll-call at each visit.\nAt the beginning of each data collection visit, enumerators visited each classroom with a roster of all students enrolled in the school, stratified by grade and sex.\nAt each visit, enumerators confirmed with the head teacher whether there were any new students since the last visit or if any students had left the school.\nNew students were added to the roster.\nDropout was recorded for students who had dropped out since the last visit.\nAbsences that were followed by a designation of dropout or transfer were removed from roll-call analysis.\nSecondary educational impacts included enrollment, dropout, and progression.\nEnrollment was calculated at each visit by summing the count of pupils on the roll-call roster and subtracting those who had dropped out or transferred.\nIn addition to student-level dropout recorded in the roll-call register, an aggregate school-level count of dropout was reported by the school at the end of each school year.\nPupils who transferred to another school were not considered to have dropped out.\nProgression was school-reported at the end of each academic year as the count of students who passed the national exam and progressed to the next grade level.\nAll secondary educational impacts were stratified by grade and sex.\nSecondary health impacts included diarrhea, symptoms of respiratory infection, and conjunctivitis/non-vision related eye illness and were collected through pupil interviews.\nAll health impacts were binary and self-reported with a one week recall period.\nPupils were asked if they had had diarrhea using local terminology and were also asked how many times they had defecated each day; a pupil was considered to have had diarrhea if he or she had reported having diarrhea and had defecated three or more times in a 24-hour period.\nPupils were considered to have symptoms of respiratory infection if they reported cough, runny nose, stuffy nose, or sore throat.\nDuring the last visit we included negative-control questions about self-reported cuts/scrapes and toothache.\nThese questions served as a measure of respondent bias, as there is no biological plausibility of an association between a WinS program and cuts/scrapes or toothache.\nData on STH infection were collected yearly.\nAny sample testing positive for the hookworms, A. lumbricoides, or T. trichuria considered positive for STH infection.\nIntervention fidelity and adherence for this study has been described previously.\nTo measure fidelity- defined as how the intervention was delivered per the stated design- we created an index score in which one point was given for each of the 20 output criteria fulfilled (Table 1).\nFor each visit, the minimum intervention fidelity score was zero and the maximum score was 20.\nTo measure adherence \u2013 defined as achievement of behavioral outcomes promoted by the intervention \u2013 a similar index score was created.\nAlthough there were five behavioral outcomes of interest (Table 1), we excluded group compound cleaning from the index given that reported participation in group compound cleaning was nearly universal among both intervention and comparison schools at baseline (97.9%), therefore the adherence score ranged from 0-4.\nA behavior was considered to be achieved when >75% of pupils reported or were observed to complete the behavior except for group handwashing, which was binary (either the school performed group handwashing or did not).\nAnalysis\nData were analyzed using Stata Statistical Software: Release 13 (StataCorp LP, College Station, TX, USA) and SAS version 9.4 (SAS Institute, Cary, NC, USA).\nIntention to treat analysis\nOur primary analysis was an intention-to-treat (ITT) analysis, which was used on all primary and secondary impacts.\nFor binary impacts (roll-call absence, diarrhea, symptoms of respiratory infection, conjunctivitis/non-vision related eye illness, STH infection, toothache, cuts/scrapes), we estimated relative risk using a \u201cmodified Poisson\u201d approach.\nThis is a validated method to produce relative risk ratios for binary data using a multi-level mixed Poisson model with robust error variances, and was chosen for this analysis because Stata does not support the use of log-linear binomial regression when using mixed effects generalized linear models.\nOdds ratios were obtained when the modified Poisson model did not converge for a specific impact (eg, toothache).\nRandom intercepts at the school and pupil levels were included to account for clustering of pupils within schools and for repeated measures of pupils over time, respectively.\nFor count impacts (enrollment, abandonment, progression), we estimated relative risk using Poisson regression models.\nAs these data were aggregated at the school-level, we included a random intercept at the school level only.\nAll ITT models compared intervention schools to comparison schools as they were randomly allocated to intervention and comparison groups, without regard to project fidelity or adherence.\nIntervention and comparison schools were balanced on key indicators at baseline, therefore intervention schools were included in the analysis once UNICEF documented that full intervention implementation (eg, both hardware and behavior change components) was complete.\nSince full implementation generally occurred at the same time in each district, comparison schools were also included once implementation occurred in their respective districts.\nModels included several design variables, including the district and visit number, and controlled for the following fixed effects, determined a priori based on biological plausibility of affecting impacts: pupil sex, pupil grade, school enrollment size, season (rainy or dry).\nThe rice crop calendar (planting, growing, harvesting) was included as a fixed effect in the absence model because rice agriculture is the predominant economic activity in the province and the need to stay home and support the family was the leading cause of pupil-reported absence.\nFully adjusted models were used to produce adjusted risk ratios (aRR) for each of the associations of interest.\nThese fixed effects, as well as whether the school was concurrently receiving aid from the World Food Program (WFP) school feeding program, were also assed for effect modification.\nCovariates were determined to be effect modifiers if an interaction term between the covariate and intervention group was significant in the full model.\nIntervention fidelity and adherence are important considerations when evaluating the impact of WASH programs.\nAssessing these factors along the causal \u2018theory of change\u2019 allows us to understand not only if but why and how that intervention succeeded or not in that context (ie, was there theory failure?).\nFurther, assessment of the process can determine if the intervention followed the intervention protocol to activate that theory of change (ie, was there intervention failure?).\nIn contexts where fidelity and adherence to the intervention is imperfect, ITT results may underestimate the causal effect for the potential impact of changes to outputs or outcomes, resulting in null or mixed effects.\nGiven the suboptimal fidelity and adherence of the intervention based on our monitoring data, we conducted a secondary analysis to quantify the impact of the project as implemented by UNICEF and adhered to by schools and pupils on the primary impact (roll-call absence) and select secondary impacts (diarrhea, respiratory infection, and STH prevalence).\nWe explore two modeling frameworks that have been previously used to evaluate the role of fidelity and adherence to a school WASH intervention on project impacts: As-treated (AT) analysis and Structural Nested Models (SNMs).\nEach framework operates under different assumptions and differ in robustness and efficiency; as such, comparing estimates lends a more informed picture of project impact.\nWe conducted a sensitivity analysis to identify a meaningful threshold of fidelity and adherence.\nThe scale of 20 outputs (fidelity) were categorized with cut-points at each 10th percentile and the scale of four outcomes (adherence) were unadjusted.\nWe observed lower risk of absence among schools with 70%-80% intervention fidelity and higher, but there was no clear evidence of a threshold for any other association (Figure 1).\nWe thus selected a threshold of 75%, which is consistent with previous research on fidelity to WinS projects.\nOnly the SNM requires specifying a threshold of fidelity/adherence, however, we also applied the 75% threshold to the AT models for comparability between the two approaches.\nAs-treated analysis\nThe AT analysis groups subjects according to the treatment received and does not consider the treatment intended (as is the case with ITT analysis).\nAdvantages to the AT approach are that it is analytically straightforward and easily supports our clustered and longitudinal study design.\nDisadvantages are that characteristics of schools with good fidelity or students who adhere to behaviors may be fundamentally different from those with poor fidelity/adherence, which can lead to confounding.\nThis confounding may be remedied by controlling for the prognostic factors that led participants to choose to adhere, but only if those prognostic factors are known, which is often not the case.\nFor the AT analysis, we ran two separate models that were structurally identical to the ITT models.\nHowever, instead of using intervention status as the primary predictor, as in the ITT analysis, schools were grouped according to intervention fidelity (ie, fulfilling \u226575% of outputs or not) and adherence (ie, fulfilling \u226575% of outcomes or not), respectively.\nAT models included the same covariates as the ITT models, with a priori identified fixed effects and random intercepts at the school and pupil levels.\nOnly data collected after the implementor reported intervention delivery was complete were included.\nAT models were stratified by effect modifiers identified in the ITT analysis.\nStructural nested model analysis\nSecond, we assessed the role of fidelity and adherence using SNMs, an instrumental variable approach.\nSNMs resolve the potential confounding issue presented by AT models because they do not break the randomization of intervention status.\nInstead, SNMs create a counterfactual for each study participant in order to compare the risk of an impact among adherers against the risk of the impact had the same individual not adhered.\nUnlike the ITT and AT models, to control for relevant covariates, a weighted distribution of population data are produced in order to remove the association between population-level confounders and randomization.\nWhile SNMs are advantageous because they account for unknown or unmeasured confounders, drawbacks are that they are more computationally intensive and rely on strong assumptions.\nSNM assumptions are described in detail elsewhere; briefly, they are as follows: (1) Exclusion restriction \u2013 randomization has no direct effect on the outcome; (2) Consistency \u2013 observed outcomes are possible under the fidelity/adherence level actually observed; (3)\nThe potential outcomes used to estimate the SNM effects are independent of randomization; (4) No interaction \u2013 the model\u2019s causal effect is consistent across randomization groups.\nOur code was derived from Garn et al and adapted for a 2-arm trial.\nBecause the SNM methodology we used does not accommodate repeated measures, we averaged time-varying pupil-level data (eg, grade, absence, reported diarrhea, reported symptoms of respiratory infection) and school-level data (output index, behavior index) across the data collection period.\nAs such, binary variables such as absence, reported diarrhea, and reported symptoms of respiratory infection became a continuous variable between zero and one, in which zero indicated never being absent, reporting diarrhea, or reporting symptoms of respiratory infection, whereas one indicated always being absent, reporting diarrhea, or reporting symptoms of respiratory infection.\nSimilar to the ITT and AT models described above, observations were included only after full implementation had been achieved.\nModels were adjusted using the same covariate variables as we used in the ITT and AT models.\nAs with the AT models, achievement of \u226575% of outputs and \u226575% of outcomes were considered achieving fidelity and adherence, respectively.\nSNMs were stratified by effect modifiers identified in the ITT analysis.\nFor all analyses, results were considered statistically significant if the P-value was <0.05.\nEthics\nThe WASH HELPS Study was approved by Emory University\u2019s Institutional Review Board (IRB0076404) and the Lao Ministry of Health\u2019s National Institute of Public Health National Ethics Committee (No. 043 NIOPH/NECHR).\nBoth Institutional Review Boards approved consent in loco parentis (in the place of the parent) signed by the school director.\nPupils who were selected for the pupil interview and/or stool collection provided informed verbal assent prior to any data collection.\nAll consent/assent procedures occurred after randomization.\nThe intervention was delivered to comparison schools in April 2017, after research activities ended.\nThe study is registered in ClinicalTrials.gov (NCT02342860).\nRESULTS\nBaseline results and intervention fidelity and adherence\nA total of 100 schools (n\u2009=\u200950 intervention, n\u2009=\u200950 comparison) were randomized, received the intervention, and included in the analysis (Figure 2).\nThere were no significant differences in key pupil-level or school-level indicators between intervention and comparison groups at baseline, indicating that the cluster-randomization was successful in creating balanced groups.\nFollowing full intervention implementation, intervention fidelity was 30.9% across all schools and visits and intervention adherence was 29.4%.\nData on fidelity to specific project outputs and adherence to specific project behaviors across the evaluation period have been previously published.\nIntention-to-treat analysis\nWe found no impact of the intervention on the primary impacts (roll-call absence) or secondary impacts (enrollment, progression, pupil-reported diarrhea, pupil-reported symptoms of respiratory infection, pupil-reported conjunctivitis, STH infection; Table 2).\nThere was some evidence of effect modification.\nRisk of diarrhea was higher in the rainy season compared to the dry season; when stratified by season, there was no significant impact of the intervention on diarrhea in either season (Dry season adjusted risk ratio (aRR)\u2009=\u20090.69, 95% confidence interval (CI)\u2009=\u20090.44, 1.10; Rainy season aRR\u2009=\u20091.14, 95% CI\u2009=\u20090.65, 1.99).\nPupil sex, pupil grade, school enrollment size, receiving support from the WFP school feeding program, and the rice crop calendar (absence model only) did not modify the effect of any primary or secondary impacts.\nWe found no difference in reported prevalence of toothache or cuts/scrapes (the negative control questions) among pupils attending intervention vs comparison schools (toothache adjusted odds ratio (aOR\u2009=\u20090.64, 95% CI\u2009=\u20090.23, 1.84; cuts/scrapes aRR\u2009=\u20091.06, 95% CI\u2009=\u20090.66, 1.72), indicating that any respondent bias that may have been present occurred equally between groups.\nAs-treated analysis\nAT results are presented in Table 3.\nIntervention fidelity \u2013 meeting \u226575% of output indicators associated with water supply, toilets, handwashing facilities, promotion of group hygiene activities, group handwashing facilities, and filtered drinking water\u2013 was associated with roll call absence and prevalence of STH.\nCompared to students attending schools without intervention fidelity, students attending schools with intervention fidelity had a 23% lower risk of absence (aRR\u2009=\u20090.76, 95% CI\u2009=\u20090.64, 0.91) and a 20% higher risk of STH prevalence (aRR\u2009=\u20091.20, 95% CI\u2009=\u20091.01, 1.43).\nDiarrhea was significantly higher during the rainy season, but when stratified there was no significant difference by fidelity status (Dry season aRR\u2009=\u20090.84, 95% CI\u2009=\u20090.48, 1.49; Rainy season aRR\u2009=\u20091.65, 95% CI\u2009=\u20090.82, 3.33).\nIntervention adherence \u2013 meeting outcome indicators associated with toilet use, handwashing with soap after toilet use, daily group handwashing, and daily group toilet cleaning \u2013 was not significantly associated with any impacts.\nStructural nested model analysis\nResults from the SNMs are presented in Table 3.\nDiarrhea was the only impact associated with fidelity or adherence.\nWhen stratified by season, diarrhea was lower in the dry season among students attending schools with intervention fidelity (aRR\u2009=\u20090.45, 95% CI\u2009=\u20090.24, 0.85) and adherence (aRR\u2009=\u20090.42, 95% CI\u2009=\u20090.21, 0.87); there was no significant difference in diarrhea between groups during the rainy season.\nDISCUSSION\nIn the primary analysis, we found no evidence that the intervention had an effect on absence, school enrollment, dropout, grade progression, pupil-reported diarrhea, pupil-reported symptoms of respiratory infection, pupil-reported conjunctivitis, or prevalence of STH.\nThese results contribute to the growing body of research showing limited or mixed impacts of WinS effectiveness trials on pupil health and education.\nSince 2010, access to WASH has been a fundamental human right recognized by the United Nations General Assembly.\nAs such, regardless of its potential education and health impacts, WinS access is an important objective, evidenced by its inclusion in the Sustainable Development Goals.\nHowever, if improvements in education and health indicators are to be achieved, results from this and other rigorously evaluated WinS programs suggest that WinS interventions alone, and as currently delivered in many contexts, may be insufficient to achieve anticipated education and health impacts.\nThe theory of change for WinS programs posits that improved WASH access leads to reductions in pathogen exposure at the school level and the habitualization of hygiene behaviors that can be practiced both at school at and home, which in-turn leads to reduced illness and thus reduced school absence.\nNumerous factors influence school absence, such as household wealth, distance to school, and number of siblings.\nLao PDR is a least-developed country, with over 65% of the population working in agriculture.\nIn Saravane Province, where over half of the population lives in poverty, the school calendar largely coincides with rice planting and harvesting seasons, and children are often kept home from school to assist in the fields and with other household chores.\nIndeed, in the current study, the leading pupil-reported cause of school absence was the need to stay home to support the family in economic activities (9.4% of pupils in intervention group and 8.7% of pupils in comparison group across all visits), not illness (5.1% of pupils in intervention group and 5.8% of pupils in comparison group across all visits), which may explain the lack of an impact of the intervention on absence.\nThus, the role of school WASH in supporting an enabling environment may be critical, but ultimately not sufficient to reduce absence when other factors like household economic needs or food security is the main driver of truancy from school.\nComplementary approaches to WinS may be necessary to achieve improvements in absence and other educational impacts.\nFor example, WinS may be successful in combination with school feeding programs or conditional cash transfers, both of which have been associated with reduced absence and increased enrollment in other low- and middle-income contexts.\nAlthough our results did not reveal a significant interaction between the WFP school feeding program and absence or enrollment, our study was not designed or adequately powered to detect a difference.\nAlthough there are potential mechanisms by which improved WASH may impact illness independently of measurable impacts on absence, we found no overall impact of the WinS intervention on pupil illness.\nThese results contrast to previous WinS research that reported overall reductions in diarrhea, respiratory infection, and absence due to illness, but are consistent with results from a WinS intervention in Lao PDR, Cambodia, and Indonesia that found no impact of the intervention on STH or underweight.\nOne explanation for the lack of an effect of the WinS intervention on pupil illness is low household WASH access; in this study context, the health benefits linked to improvements in school WASH conditions and behaviors provided by this intervention were likely not sufficient to overcome other potential transmission pathways at home or elsewhere in the community.\nEnvironmental improvements in both the domestic and public domains may be required for successful control of infections targeted by environmental improvements, such as diarrhea.\nAs such, WinS alone may not achieve significant health gains without concurrent community and household WASH improvements.\nFidelity and adherence are fundamental antecedents to achieving intervention effects.\nIt is possible that the lack of an effect of the intervention could be due, in part, to sub-optimal or unsustained fidelity and adherence.\nHowever, our secondary analyses yielded limited evidence of an effect of the intervention, even at high levels of intervention fidelity and adherence.\nAdditionally, our sensitivity analysis showed no clear trend in impacts across the fidelity/adherence continuum.\nWith two exceptions \u2013 the association between fidelity and lower absence (AT analysis) and the association between fidelity and adherence and lower diarrhea during the dry season (SNM analysis) \u2013 we did not find that fidelity and adherence led to improved education or health.\nThese results support the above conclusion that factors other than WinS \u2013 such as low household WASH access or household economics \u2013 may supersede health and education benefits of a WinS intervention in low-income contexts.\nHowever, the AT evidence should be should be interpreted cautiously due to the limited potential for causal inference resulting from breaking the randomization assignment in the AT analysis.\nThe two fidelity and adherence analyses results were inconsistent and sometimes yielded estimates of effect in opposite directions (eg, associations between adherence and diarrhea, respiratory infection, and STH), which is likely due to unaccounted for confounding in the AT analysis.\nIV analyses are known to yield estimators with high variance, especially when compliance is low, which may also partially explain differences between the AT and SNM results.\nThe choice of which method to use depends on numerous factors, including study design, plausibility of meeting analysis assumptions, and available analytical resources; our conflicting estimates highlight the importance of testing the sensitivity of multiple fidelity analysis options.\nStrengths and limitations\nThe design, methods and approach of the WASH HELPS study were robust.\nRandomized controlled trials offer the greatest potential for causal inference.\nThe longitudinal design allowed us to collect data across three full school years of in Group 1 schools and two full school years of in Group 2 schools, allowing us to capture inter-seasonal and inter-year variations in the outputs, outcomes, and impacts.\nAll data were collected during unannounced school visits so that schools could not prepare for the visit and bias observations.\nOur primary measure of impact \u2013 roll-call absence - is an objective measure of school absence.\nThis impact evaluation was conducted by external researchers, to foster an unbiased assessment of the project impact.\nOur field team was composed of experienced Laotian enumerators to ensure the tools were designed and delivered with cultural and contextual appropriateness.\nThis robust study design lends strong internal validity, and results may be generalized to the larger, nationwide WinS project.\nThis was an effectiveness trial evaluating an intervention as conducted in a real-world setting.\nThe lessons from this project, taken with other recent WinS trials, reveal heterogeneity of findings that can inform programming across contexts.\nLastly, in addition to comparing two methods to analyze the effect of intervention fidelity on WinS impacts, our fidelity analysis also examines adherence to intervention behaviors, which has not been previously included in WinS fidelity analyses.\nThere are a number of limitations to this evaluation.\nFirst, the secondary health impact measures (diarrhea, symptoms of respiratory infection, conjunctivitis) were based on self-report by pupils, which may be subject to bias, and this evaluation was not blinded for either the beneficiaries or data collectors.\nMore objective and robust measures of pupil health, such as molecular methods to detect enteric infection in stool samples, would improve our confidence in the reported impacts, though these measures can be costly, time consuming, and require specialized equipment and laboratory staff.\nAs a way to measure potential reporting bias, we included a negative control question about symptoms of illness unrelated to WASH access (cuts/scrapes and toothache) at the last survey visit.\nDifferences in reported symptoms of these illnesses between intervention and comparison groups would indicate a potential reporting bias, but we found no evidence to suggest that any bias may have existed to a greater degree among either the intervention group or the comparison group.\nAdditionally, schools in the comparison group did not have functional WASH facilities, so it is unlikely that the null results could be explained by a change in behaviors among the comparison group.\nSecond, the intervention was delivered across two different school years, so Group 1 schools had one more year of surveillance than Group 2 schools.\nFollowing a single cohort of schools over the same time period would have provided a more accurate measure of WinS hardware and software performance, sustainability, and impact.\nThird, implementation was delayed in many Group 1 schools.\nThe intervention was fully implemented in Group 1 schools at visit 4, with the exception of Samoui district, in which the intervention was fully implemented at visit 9.\nOur analysis excludes visits prior to full intervention implementation, thus power may have been limited by dropping observations under incomplete intervention delivery.\nLast, we were unable to account for the quality of intervention design or dose of the intervention received, which are important components of fidelity and adherence.\nCONCLUSIONS\nOur findings and those of other rigorous WinS trials suggest that WinS programs \u2013 as currently designed and delivered \u2013 do not have a population-level benefit on education and health.\nIn this context, the WinS improvements alone were not sufficient to address the other powerful causes of absenteeism, enrollment, and dropout that are not related to \u2013 but possibly more influential than \u2013 school WASH.\nWe believe this likely holds in many similar settings.\nSimilarly, WinS improvements, though potentially critical for the enabling environment, may not be sufficient to overcome disease transmission in areas where community and household WASH coverage is poor.\nWinS, independent of its stated purpose of improving education and health, is an important objective for dignity, inclusivity, and development.\nHowever, if intended impacts are to be achieved, improving intervention fidelity and adherence and including other complementary approaches for WASH may be required.\nTo better understand how to improve intervention fidelity and adherence, evaluations of WinS interventions need to better understand and adapt to contextual drivers of key impacts and outcomes, further develop and test theories of change, and conduct rigorous process evaluations to understand where along the causal pathways interventions are falling short.\nAssociation between intervention fidelity and adherence continuum and intervention impacts.\nFlow diagram of school and pupil selection.\n\nIntervention outputs and behavioral outcomes and their measurement indicators\nOutput | Indicator and criteria\nWater supply | \u2022 Improved* water point on school compound\n\u00a0\u00a0\u00a0\u00a0- Water point functional in the previous year (director reported)\n\u2022 Water tank to supply toilet and handwashing stations\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water observed in tank\nToilets | \u2022 At least one improved* toilet compartment\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is sex separated (by designation)\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is unlocked\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is clean\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet has water available inside compartment for flushing\nHandwashing facilities | \u2022 At least one individual handwashing station available to pupils\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water available at individual handwashing station\n\u00a0\u00a0\u00a0\u00a0\u00a0- Soap available at individual handwashing station\nPromotion of daily group hygiene activities | \u2022 Daily group handwashing schedule posted in at least one classroom or near toilet\n\u2022 Daily group compound cleaning schedule posted in at least one classroom or near toilet\n\u2022 Daily group toilet cleaning schedule posted in at least one classroom or near toilet\nGroup handwashing | \u2022 Group handwashing facility available to pupils\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water available at group handwashing facility\n\u00a0\u00a0\u00a0\u00a0\u00a0- Soap available at group handwashing facility\nWater filters | \u2022 At least one drinking water filter available in a classroom for pupil use\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water in filter\nOutcome | Indicator\nToilet use | \u2022 Percentage of students using toilet for defecation during school hours (pupil-reported)\nHandwashing (individual) | \u2022 Percentage of students washing hands with soap and water upon exiting toilet (observation)\nDaily group handwashing | \u2022 School conducted daily group handwashing the day of visit (observation)\nDaily group toilet cleaning | \u2022 Percentage of students participating in daily group toilet cleaning within the previous five school days (pupil-reported)\nDaily group compound cleaning | \u2022 Percentage of students participating in daily group compound cleaning within the previous five school days (pupil-reported)\n\n*Defined according to Joint Monitoring Programme (JMP) standards.\n\nAssociation between WinS intervention and health and educational impacts, Saravane Province, Lao People\u2019s Democratic Republic, 2014-2017 (n\u2009=\u2009100 schools)\nImpact | Comparison* | Intervention* | Adjusted risk ratio | 95% confidence interval\nRoll-call absence\u2020 | 6024 (32.2%) | 7147 (29.9%) | 1.01 | 0.84, 1.20\nEnrollment\u2021 | 68.2 (49.7) | 71.6 (50.0) | 1.07 | 0.84, 1.35\nDropout\u2021 | 0.8 (2.6) | 0.4 (1.0) | 0.56 | 0.25, 1.24\nGrade progression\u2021 | 64.4 (48.6) | 67.3 (48.6) | 1.07 | 0.91, 1.25\nDiarrhea\u2020,\u00a7 | 1032 (21.1%) | 947 (14.7%) | 0.80 | 0.51, 1.26\nSymptoms of respiratory infection\u2020,\u2016 | 1414 (28.9%) | 2064 (32.1%) | 1.08 | 0.95, 1.23\nConjunctivitis\u2020,\u00a7 | 41 (0.8%) | 48 (0.8%) | 0.89 | 0.53, 1.52\nPrevalence of any STH\u2020,\u00b6 | 1833 (39.8%) | 1935 (41.6%) | 1.00 | 0.85, 1.17\n\n*Data are n (%) for impacts at the pupil level (roll-call absence, diarrhea, symptoms of respiratory infection, conjunctivitis, and prevalence of STH) and mean (SD) for impacts at the school-level (enrollment, dropout, grade progression) across study period.\n\u2020Risk ratios were calculated using a Poisson model with robust error variances and random intercepts at the school and pupil level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size, and season (rainy or dry). Absence model additionally controlled for and rice crop calendar (planting, growing, harvesting).\n\u2021Risk ratios were calculated using a Poisson model with random intercepts at the school level. All models adjusted for district and visit number.\n\u00a7Pupil-reported in previous week.\n\u2016Pupil-reported cough, runny nose, stuffy nose, or sore throat in previous week.\n\u00b6Samples testing positive for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus).\n\nAssociation between WinS intervention fidelity and adherence and absence, diarrhea, respiratory infection, and soil-transmitted helminth infection (STH), Saravane Province, Lao PDR, 2014-2017 (n\u2009=\u2009100 schools)\n | As-treated analysis | Structural nested model analysis\n | Adjusted risk ratio* | 95% confidence interval | Adjusted risk ratio\u2020 | 95% confidence interval\nRoll-call absence:\nFidelity\u2021 | 0.76 | 0.64, 0.91 | 0.97 | 0.33, 2.81\nAdherence\u2021 | 0.91 | 0.79, 1.05 | 0.96 | 0.19, 4.97\nDiarrhea:\u00a7\nFidelity, dry season\u2021 | 0.84 | 0.48, 1.49 | 0.45 | 0.24, 0.85\nAdherence, dry season\u2021 | 1.00 | 0.70, 1.44 | 0.42 | 0.21, 0.87\nFidelity, rainy season\u2021 | 1.65 | 0.82, 3.33 | 1.03 | 0.42, 2.51\nAdherence, rainy season\u2021 | 1.41 | 0.61, 3.26 | 0.50 | 0.19, 1.30\nSymptoms of respiratory infection\u2016:\nFidelity\u2021 | 1.00 | 0.89, 1.14 | 1.41 | 0.93, 2.13\nAdherence\u2021 | 0.97 | 0.84, 1.11 | 2.30 | 0.54, 8.87\nPrevalence of any STH:\u00b6\nFidelity\u2021 | 1.20 | 1.01, 1.43 | 1.10 | 0.57, 2.13\nAdherence\u2021 | 0.93 | 0.77, 1.12 | 1.18 | 0.37, 3.73\n\nSTH \u2013 soil-transmitted helminth\n*Risk ratios calculated using a Poisson model with robust error variances and random intercepts at the school and pupil level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size, season (rainy or dry). Absence models additionally controlled for rice crop calendar (planting, growing, harvesting).\n\u2020Risk ratios calculated using a Structural Nested Model with random intercepts at the school level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size.\n\u2021Fulfilling \u226575% of intervention outputs was considered fidelity. Fulfilling \u226575% of intervention outcomes was considered adherence.\n\u00a7Pupil-reported in previous week.\n\u2016Pupil-reported cough, runny nose, stuffy nose, or sore throat in previous week.\n\u00b6Samples testing positive for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus).",
        "label": "high",
        "id": "task4_RLD_test_781"
    },
    {
        "paper_doi": "10.4314/ahs.v13i2.44",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: DesigncRCTAllocation of clusters10 villages randomized to intervention, 9 to control\n\n\nParticipants: 558 children younger than age 5\n\n\nInterventions: Primarily education\n\n\nOutcomes: Any STH\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Design and implementation of participatory hygiene and sanitation transformation (PHAST) as a strategy to control soil-transmitted helminth infections in Luweero, Uganda\nDesign and implementation of participatory hygiene and sanitation transformation (PHAST) as a strategy to control soil-transmitted helminth infections in Luweero, Uganda\nAfrican Health Sciences\nAfr H. Sci.\nBackground:\nThe study is a continuation of a research carried out in Luweero district in Uganda 1 .\nIt investigated whether PHAST was a suitable tool for reducing transmission of soil transmitted helminths.\nPHAST means Participatory Hygiene and Sanitation Transformation; a participatory approach that uses visual tools to stimulate the participation of people in promotion of improved hygiene and sanitation.\nObjective: To assess the effect of PHAST on intestinal helminth transmission in children under five years.\nMethods: Three phases namely; (1) Baseline survey (2) PHAST intervention (3) Follow up were conducted.\nDuring Phase 1, the subjects' stool samples were examined for presence of helminthic ova and questionnaires administered.\nIn Phase 2, PHAST was conducted only in experimental villages.\nAll subjects in the experimental and control villages were treated thrice with Albendazole.\nDuring Phase 3, all steps of Phase 1 were repeated.\nResults: There was an overall reduction in the prevalence of children infected with helminths after PHAST intervention.\nAlso, comparison of pre-intervention and post-intervention multivariate results indicates that the likelihood of children getting infected with helminths reduced in most of the experimented variables.\nConclusion: Health stakeholders should utilize PHAST approach to sensitize communities on the importance of hygiene to curb soil-transmitted helminth infections.\nIntroduction\nIntestinal helminth infections have been reported in many parts of Uganda including Luweero district [1][2][3][4] .\nThe district has had inadequate investigations on the problem, yet it has had the infections for quite a long time 3 ; it was therefore necessary to investigate a way of controlling its transmission.\nThus, this study was conducted to assess the effect of PHAST intervention on intestinal helminth infections in children less than 5 years.\nPHAST is a participatory approach, which was developed to encourage people to analyze their own situation and identify key problems, decide what things need to be improved, plan how they are going to do it and then act.\nFor many years, conventional messages on hygiene and sanitation had been known and largely understood by people.\nHowever, these messages had not translated into significant improvement in good hygiene practices.\nIn 1993, WHO and the Regional Water and Sanitation Group for East and Southern Africa (RWSG-ESA) initiated the PHAST strategy to address this concern 5 .\nAmong the PHAST tools that were developed, 4 were purposely selected as most appropriate for the current study to cover 5 of the essential steps to community planning namely; Sanitation ladder, Three-Pile Sorting, Faecal-Oral disease Transmission Routes and Barriers (FTRB), and Tippy Tap 6 .\nSanitation ladder (set of pictures showing various methods of excreta disposal) was selected for problem identification.\nParticipants arranged pictures in form of ladder, from the worst practice/latrine to the best; identifying their own situation and looking at advantages of moving up the ladder.\nThree-Pile Sorting was for analyzing problems and selecting options; participants sorted out 30 pictures of hygiene/sanitation related situations depending on whether they considered them \"good\", \"bad\" or \"in-between\" giving reasons in each case.\nFTRB was for planning for solutions; participants organized a set of pictures basing on what they knew about faecal-oral transmission routes and then worked out how to block the routes using common barrier pictures.\nTippy tap was for planning for new facilities and behaviour change; this was a hand washing facility demonstrated to participants; they were encouraged to make and use it in their homes.\nMonitoring was done after every training session and family members assessed their progress using pictorial monitoring forms.\nDespite the fact that PHAST had been employed in some programmes such as RUWASA, effects of the use of participatory methods had not been systematically monitored or documented.\nRecorded effects were merely anecdotal and it lacked baseline surveys to prove its effectiveness 5,7 .\nThis study therefore tested the approach as no studies had been carried out to assess its impact.\nMethods\nThe study was a randomized community intervention trial with pre-and post-intervention phases.\nThe study was implemented in 3 phases.\nPhase 1 was a baseline cross-sectional descriptive survey that investigated the prevailing helminth status described in already published paper 1 .\nTwo sub-counties, selected by simple random sampling, consisted of 4 parishes from which 19 study villages were studied.\nStool samples from 727 eligible children were examined for presence of different types of helminth ova using Kato-Katz 8 technique.\nSemi-structured questionnaires were also administered to parents/ guardians and inspection of households conducted to assess their hygiene status.\nDuring Phase 2, the four parishes were randomly assigned 10 experimental and 9 control villages with 357 and 370 children respectively.\nPHAST health education was carried out thrice among the experimental group (parents/guardians) only.\nAfter each training session, the respondents' households were visited to reinforce what had been discussed during the training.\nAt each visit, household members freely discussed and identified their household sanitation and hygiene status they had attained plus the ones they were aiming at using pictorial monitoring forms.\nIn addition, all the children were treated with a single oral dose of Albendazole depending on age once every 3 months; those below 2 years were given one 200-mg tablet whereas those between 2 and 5 years two 200-mg tablets.\nIt was a directly observed therapy.\nDuring Phase 3, all steps and procedures in Phase 1 were repeated.\nThe results were analyzed using univariate and bivariate analyses.\nChi-square test was used and the level of significance set to 95% level.\nThe relationship between the variables was established as statistically significant when found to be equal to or less than 0.05 9 .\nThe odds ratios (OR) of the children were determined using the odds ratio statistic in a 2x2 analysis.\nMultivariate analysis was further used to investigate how helminth infection was related to more than one variable at a time while controlling for confounders.\nA binary logistic regression model was used to obtain the adjusted OR.\nResults\nTable 1 indicates that the prevalence rate of children infected with helminth ova which was 27.6% (201/ 727) at baseline reduced to 16.5% after PHAST intervention.\nComparison of Phase 1 and 3 preintervention (pr) and post-intervention (po) multivariate results for different variables also indicate that the odds of children getting infected with helminth ova reduced after PHAST intervention in all the variables except three; these either had an increase or a constant Odds Ratio (tables 2a and 2b).\nIf OR after PHAST is lower than OR before PHAST, there was a reduction in helminth infections.\nThe variables that indicated the most significant reduction were; condition of latrines, respondents' hand washing after handling children's faeces and keeping of pigs.\nThese variables were OR -Odds ratio CI -Cornfield 95% confidence limits for OR # -Respondents did not wash hands * -Most significant OR reductions after PHAST intervention \u00a7 -Children pushing against the ground as a way of cleaning themselves after visiting significantly associated with helminth infection before the intervention but were not associated after the intervention.\nFor instance, there was nearly a threefold reduction in OR noted among helminth-infected children who were living in homes with poorly maintained latrines compared to those in homes with fairly maintained latrines (OR pr = 1.90; 95% CI = 1.17 -3.10 Vs OR po = 0.74; 95% CI = 0.30 -1.80).\nFurthermore, of the children infected with helminths, the likelihood of those who lived in homes with pigs getting infected also decreased by a half after the intervention compared to those in homes without pigs, (OR pr = 1.73; 95% CI = 1.17 -2.58 Vs OR po = 0.82; 95% CI = 0.48 -1.39).\nTable 3 indicates that generally, there was a statistically significant decline in prevalence in the control group (chi square =16.90, p= 0.00).\nVariable\nPre-intervention -\nOR -Odds ratio CI -Cornfield 95% confidence limits for OR\n# -Respondents did not wash hands * -Most significant OR reductions after PHAST intervention \u00a7 -Children pushing against the ground as a way of cleaning themselves after visiting\nThere was also a statistically significant decline in the prevalence in experimental group (Chi-square = 6.41, p= 0.01).\nA difference in the decline in prevalence for the control and experimental groups was noted but not statistically significant (The control group declined 6 percent points more than the experimental group (Chi-square = 1.83, p > 0.05).\nDiscussion\nEffect of PHAST intervention on helminthic infections\nThe overall prevalence rate of helminth infections reduced after the intervention in Phase 3.\nThe likelihood of children getting infected with helminths also reduced in almost all the variables.\nThe observed reduction in these aspects is likely to be attributed to PHAST intervention.\nAs stipulated by Simpson-Herbert et al., 1997 6 , tools and techniques that were used during PHAST participatory health education in the experimental group could have stimulated participation even among participants who did not know how to read.\nBy taking them through the various steps and activities, it was possible to help them understand that poor hygiene and sanitation behaviours and practices are the principal causes of many preventable diseases, such as helminthiasis.\nWhereas conventional health education, which is normally delivered by didactic teaching, is aimed at reducing transmission and re-infection by encouraging healthy behaviour, it is not normally presented as a user friendly package involving participants.\nUNDP recommends that participatory methods create a non-threatening environment in which all participants regardless of class, age and sex can express their views freely 10 .\nThis freedom was observed during PHAST sessions where almost every participant was able to identify their own problems and plan for solutions in a relaxed atmosphere.\nEvidently, PHAST approach was an eye opener to the respondents as noted by a reduction in odds ratio after the intervention in most aspects.\nReduction, for instance, was noted in the risk of children acquiring infections through respondents' hand washing with soap after visiting the latrine and after handling children's faeces.\nReduction in the likelihood of children getting infected with worms in homes with mud floors indicates that, despite the fact that floors remained uncemented, participants cleaned them better after PHAST exposure.\nDecrease in the risk of children acquiring infections for those living in homes with soiled compounds could be attributed to respondents' becoming aware that indiscriminate dumping of faeces was harmful and therefore ensured that it was properly disposed of.\nThe improvement in the type of latrine, latrine floor and condition was linked with the newly constructed or improved latrines observed during Phase 2 monitoring.\nThis in turn resulted in the reduction of risk of children acquiring worm infections.\nThe decline in odds of children acquiring worm infection in homes keeping pigs could be attributed to respondents who could have restricted the roaming of pigs after the intervention, thus reducing on indiscriminate defaecation.\nAccording to Scott 11 , the association with pig ownership is intriguing giving the continuing interest in zoonotic potential of transmission of Ascaris from pigs to humans.\nThe results of respondents who had difficulties in accessing water show that, despite the problem, they could have started using it sparingly for improving hygiene practices.\nHowever, an increase in the risk of children acquiring worm infections for those who cleaned their anal region by pushing themselves against the ground (sliding) after visiting the latrine, could be attributed to the fact that it was not easy for the parents/guardians to teach their kids the proper way of cleaning themselves as they were either busy with domestic chores or away from home whenever they defaecated.\nThe fact that the risk of children getting infected for those living in homes with refuse dumps increased whereas that of children in homes without hand washing facilities (tippy taps) remained the same after intervention could imply that more home visits were still required to sensitize respondents to clean up compounds and to produce more of such facilities.\nIn view of the findings discussed above that there were significant changes in hygiene related behaviour after the intervention, it is possible that within the short run, chemotherapy has a greater effect on worm prevalence generally than the PHAST intervention as indicated in Table 3.\nThe increase in hygiene related behaviour in various variables, however, suggests that PHAST intervention would have a greater effect on worm prevalence but in a long run.\nIn this respect, it should be noted that least decline was recorded in Kasala parish (2.4%) where PHAST sessions were least attended and highest decline in Kisimula (15%) where there was good attendance yet the parish had had highest prevalence (38.8%) prior intervention.\nConclusion\nIt is possible that higher reductions would have been registered if PHAST activities were conducted for a longer period.\nSwitzerland et al., 2010 12 reports that PHAST is a process that takes time.\nThe study findings, therefore, raise questions on how long and how frequently PHAST has to be implemented in order to be fully effective.\nThis requires further research.\nAccording to Loevinsohn 13 , successful health education depends on using a few messages, of proven benefit, repeatedly, and in many fora.\nTable 1 : Comparison of Phases 1 and 3 helminth infections -Phase 3 percentages are based on a total of 92 infected cases, 101 includes double infectionsThere was a high drop rate basically due to migration from study area; hence the difference in the two study populations during Phase 1 & 3 (727 and 558 respectively). Phase 3 data contains both experimental and control groups.\nType of worm | No. of cases before intervention No. of cases after intervention\n | -Phase 1 (n = 727) |  | -Phase 3 (n = 558) | \n | No. of infections | (%) | No. of infections \u00de (%)\nAncylostoma duodenale /Necator americans | 165 | (82.1) | 64 | (71.1)\nAscaris lumbricoides | 38 | (18.9) | 10 | (11.1)\nTrichuris trichiura | 14 | (7.0) | 22 | (24.4)\nHymenolepis nana | 1 | (1.0) | 0 | 0\nEnterobius vermicularis | 2 | (0.5) | 5 | (5.6)\nTotal | 220 | 27.6 | 101 | 16.5\n\u00de\nTable 2a : Comparison of pre-intervention and post-intervention multivariate results\nVariable | Pre-intervention -Phase 1 | Post-intervention -Phase 3\n | Adjusted OR (95% CI) | Adjusted OR (95% CI)\nLevel of Education |  | \nNone | 1.06 (0.70 -1.63) | 0.63 (0.34 -1.17)\nAt least primary education |  | \nRespondents' hand washing after latrine | \nNothing # /Water only | 1.03 (0.55 -1.94) | 0.48 (0.21 -1.08)\nWashing with soap |  | \nDo after handling children's faeces |  | \nNothing # /Water only | 1.79 (1.03 -3.11) | 0.87* (0.46 -1.63)\nWashing with soap |  | \nCleaning of child after latrine |  | \nSliding  \u00a7 | 0.46 (0.30 -0.70) | 1.85 (1.05 -3.26)\nLeaves/Bathing |  | \nType of floor of house |  | \nMud | 1.14 (0.62 -2.09) | 0.60 (0.24 -1.51)\nCemented |  | \nSoiling the compound |  | \nYes | 1.05 (0.68 -1.64) | 0.50 (0.14 -1.78)\nNo |  | \nPresence of refuse dumps |  | \nYes | 1.12 (0.73 -1.71) | 1.71 (0.61 -4.81)\nNo |  | \nStatus of compound |  | \nPoorly maintained | 1.23 (0.79 -1.90) | 0.80 (0.37 -1.71)\nFairly maintained |  | \nEasy |  | \nTable 2b : Comparison of pre-intervention and post-intervention multivariate results\n | Phase 1 | Post-intervention -Phase 3\n | Adjusted OR (95% CI) | Adjusted OR (95% CI)\nKeep pigs |  | \nYes | 1.73 (1.17 -2.58) | 0.82* (0.48 -1.39)\nNo |  | \nType of latrine |  | \nTNR  \u2021 | 1.16 (0.76 -1.80) | 0.73 (0.41 -1.29)\nTIL \u00b1 |  | \nLatrine floor |  | \nLogs/Mud | 1.32 (0.76 -2.32) | 1.09 (0.47 -2.53)\nConcrete |  | \nCondition of latrine |  | \nPoor | 1.90 (1.17 -3.10) | 0.74* (0.30 -1.80)\nFair |  | \nHand washing facility |  | \nNo | 0.78 (0.38 -1.61) | 0.79 (0.40 -1.59)\nYes |  | \nAccessibility to water |  | \nDifficult | 1.28 (0.82 -1.99) | 0.85 (0.47 -1.55)\nEasy |  | \nTable 3 : Comparison of prevalence of helminth infections of the study groups for Phase 1 \u00a5 and Phase 3 \u00a9\nStudy groups Prevalence (%) | Prevalence (%) | Percent Decline | \n | Per Parish | Per study | Per | Per study | Per Parish | Per study | Chi-Square \u20ac\n |  | Arm | Parish | Arm |  | Arm | (P-value)\nControl | (24.6) | 29.7 | (10.2) | 15.9 | 14.4 | 14.5 | 16.90\n |  |  |  |  |  |  | (0.00)\n | (34.2) |  | (20.7) |  | 13.5 |  | \nExperimental (12.3) | 23.5 | (9.9) | 17.1 | 2.4 | 8.5 | 6.41\n | (38.8) |  | (23.8) |  | 15.0 |  | (0.01)\n |  | 27.6 |  | 16.5 |  | 11.1 | 1.83 0.18\n\u00a5 Phases 1 -Pre-intervention \u00a9 Phases 3 -Post-intervention \u2022 Mantel -Haenszel's Chi-square used",
        "label": "unclear",
        "id": "task4_RLD_test_798"
    },
    {
        "paper_doi": "10.1186/s12984-021-00813-7",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Design: feasibility RCT\nCountry: USA\nSense(s) addressed: tactile\nStudy recruitment and setting details: see Table 7 \n\n\nParticipants: Inclusion criteriaHistory of stroke > 1 year priorImpaired touch sensation in the hand (Semmes Weinstein Monofilament Exam score of >= 0.2 grams on 3 of 20 measured locations on the hand)Passive range of motion allows user to don a gloveEnglish speaker, age 18 or olderExclusion criteriaIntact sensation in the hand (determined by Semmes Weinstein Monofilament Exam)Active range of motion within normal limits for all joints of the fingersCognitive deficits, dementia or aphasia (MMSE score of < 22) that prevent informed consentOther neurological condition that may affect motor response (e.g. Parkinson's, amyotrophic lateral sclerosis (ALS), or multiple sclerosis (MS))Pain in the limb that substantially interferes with ADLs or prior arm injuryEnrolment in a conflicting study, Botox treatment, or other upper extremity rehabilitation programme during the study periodStudy population (number randomised): 16\nDropout details given in Table 8\nParticipant details given in Table 9\n\n\nInterventions: Comparison: active intervention vs controlActive intervention \nName: vibrotactile stimulation glove\nClassification of intervention: rehabilitation (restitution)\nMaterials: a wearable computing glove providing vibrotactile stimulation. A vibration motor was attached to each dorsal phalanx, allowing a designated actuator for each finger while stimulating a region where vibrations can reach the glabrous skin of the palm and the finger extensor tendons. A circuit board and micro-controller activates motors in a pre-programmed sequence when the switch is turned \"on.\" Small, coin-shaped vibration motors from Precision Microdrives (ERM-type, Model #310-113) provide the stimulation\nProcedures: stimulation transmitted at a frequency range of 10 Hz to 400 Hz (ideally 250 Hz). Stimulation pattern and timing was designed to be intensive but not uncomfortable by using many vibration pulses with a changing location across the fingers. Vibration motors were driven at a voltage of 3.3 V for an approximate amplitude of 1.5 g and 210 Hz vibration frequency (measured in a laboratory setting for validation at 1.3 g and 175 Hz when attached to the glove). Two stimulation sequences were used, each based on the finger pattern for a piano song which provided a framework for pseudo-random stimulation. The protocol includes no required exercises.\nWho delivered: self-delivery \nMode: not reported\nWhere: participant's home\nSession: 56 sessions (daily for 8 weeks)\nDuration: 3 hours per day for 8 weeks (21 hours per week)\nTailoring: not reported\nModification: not reported\nDoes normal therapy continue? Participants continued their standard of careControl\nName: sham\nMaterials: a wearable computing glove\nProcedures: participants in the sham control condition receive a glove with vibration disabled. They were instructed to wear the glove on their affected hand, switched on, for three hours daily while awake.\nWho delivered: self-delivery \nMode: not reported\nWhere: participant's home\nSession: 56 sessions (daily for 8 weeks)\nDuration: 3 hours per day for 8 weeks (21 hours per week)\nTailoring: not reported\nModification: not reported\n\n\nOutcomes: Motor: voluntary angular range of motion \nSensory: Semmes Weinstein Monofilament Exam\nOther: Modified Ashworth Scale\nTiming: immediately after interventionFor overview of included outcome measures, see Table 10.\n\n\nFunding statement: Funding statement: this research was supported, in part, by the National Science Foundation (NSF) Graduate Research Fellowship program, a grant from the Georgia Tech Graphics, Visualization, and Usability (GVU) consortium, and a Microsoft Research PhD Fellowship\nConflict of interest statement: the authors declare that they have no competing interests\n\n\nNotes: Trial registration details: as a feasibility study, the trial was not listed with clinicaltrials.gov \nPublished protocol: no\nPPI: none reportedNo statement on pilot/feasibility design; no power calculation reporte\n\n",
        "objective": "To assess the effectiveness of interventions aimed at perceptual disorders after stroke compared to no intervention or control (placebo, standard care, attention control), on measures of performance in activities of daily living.",
        "full_paper": "Objective\nEvaluate the feasibility and potential impacts on hand function using a wearable stimulation device (the VTS Glove) which provides mechanical, vibratory input to the affected limb of chronic stroke survivors.\nMethods\nA double-blind, randomized, controlled feasibility study including sixteen chronic stroke survivors (mean age: 54; 1-13 years post-stroke) with diminished movement and tactile perception in their affected hand.\nParticipants were given a wearable device to take home and asked to wear it for three hours daily over eight weeks.\nThe device intervention was either (1) the VTS Glove, which provided vibrotactile stimulation to the hand, or (2) an identical glove with vibration disabled.\nParticipants were randomly assigned to each condition.\nHand and arm function were measured weekly at home and in local physical therapy clinics.\nResults\nParticipants using the VTS Glove showed significantly improved Semmes-Weinstein monofilament exam results, reduction in Modified Ashworth measures in the fingers, and some increased voluntary finger flexion, elbow and shoulder range of motion.\nConclusions\nVibrotactile stimulation applied to the disabled limb may impact tactile perception, tone and spasticity, and voluntary range of motion.\nWearable devices allow extended application and study of stimulation methods outside of a clinical setting.\nBackground\nOver 15 million people have a stroke each year, making it one of the leading causes of disability in the United States and worldwide.\nUpper limb disability occurs in about 50% of cases and diminished tactile perception in about 35-55%.\nCurrent methods of therapy for upper limb dysfunction after stroke focus on activities which use the limb; however, these forms of rehabilitation are not accessible to survivors with very limited function.\nSomatosensory stimulation may be an effective and accessible modality for rehabilitation.\nMost fundamentally, somatosensory input is known to drive cortical organization and skill acquisition.\nSomatosensory input has also been associated with sensorimotor recovery after CNS injury in animal as well as human studies.\nAfferent input is also integral to limb use.\nTactile perception and proprioception are factors in motor performance and are thought to co-activate with motorcortical circuits.\nAfferent electrical stimulation has been studied as a means for providing sensory input to the disabled extremity of stroke survivors, and preliminary evidence shows changes in tactile perception, motor function and brain activity.\nMechanical, vibratory stimulation can be applied without the placement, electrodes or gel of electrical stimulation.\nAfferent electrical stimulation most often targets cutaneous sensory receptors in the skin; while vibrotactile stimulation can activate both muscle afferent fibers and cutaneous sensory receptors without inducing movement.\nVibrotactile stimulation has been coupled with other methods such as robotic manipulation or music practice exercises for rehabilitation, and applied to the arm for in-situ dexterity improvement.\nImproved spasticity and significant neuromuscular changes have been found in laboratory studies of whole-body vibration (WBV) and focal muscle/tendon vibration.\nDespite encouraging data, vibrotactile stimulation is not widely used outside the clinic because there are no mobile devices that can deliver and study this form of mechanical stimulation for prolonged periods of time.\nHere we designed a lightweight, wireless, wearable device to apply vibrotactile stimulation to the hand.\nWearable devices are closely coupled with the body, and thus allow stimulation for extended periods of time and in the background of daily life.\nThe intervention is mobile and simple to apply without access to a clinic.\nUsers simply wear the device, requiring little exertion and time, which may facilitate adherence.\nThe device was deployed in a controlled feasibility trial of chronic stroke survivors with upper limb sensorimotor deficits.\nIf wearable stimulation proves to be effective it could directly impact healthcare delivery, because it may provide a mobile, affordable rehabilitation option for patients who otherwise would not have access to high intensity stroke rehabilitation.\nMethods\nThe study was a double-blind, randomized controlled study performed in Atlanta, Georgia.\nEligible participants were randomly assigned to the vibrotactile stimulation glove (VTS) or sham control glove (control) condition.\nAll were asked to wear the device on their affected hand for three hours each day for eight weeks.\nAs a feasibility study, the trial was not listed with clinicaltrials.gov but was approved and overseen by the Office of Research Integrity\u2019s IRB board of Georgia Institute of Technology.\nAll participants were screened using the Mini Mental State Exam (MMSE) and provided written consent before beginning the study.\nParticipants\nThe study included 16 chronic stroke survivors with upper extremity deficits (ages 28-68; 1-13 years post stroke (Mean=3.7, SD=3.3); 8 VTS condition/8 control condition).\nParticipants were recruited through stroke support groups in the Atlanta metropolitan area.\nFigure 2 shows a breakdown of participant demographics.\nIndividuals with various levels of arm function could participate.\nThe protocol requires no exercises and thus is accessible to patients with very limited movement.\nBecause this investigation is preliminary, no prior data are available for optimal sample size calculation.\nInclusion criteria\nHistory of stroke > 1 year prior\nImpaired touch sensation in the hand (Semmes-Weinstein monofilament exam score of  0.2 grams on 3 of 20 measured locations on the hand)\nPassive range of motion allows user to don a glove\nEnglish speaker, age 18+\nExclusion criteria\nIntact sensation in the hand (determined by Semmes-Weinstein monofilament exam)\nActive Range of Motion within normal limits for all joints of the fingers\nCognitive deficits, dementia or aphasia (MMSE score of <22) that prevent informed consent\nOther neurological condition that may affect motor response (e.g. Parkinson\u2019s, ALS, MS)\nPain in the limb that substantially interferes with ADLs or prior arm injury\nEnrollment in a conflicting study, Botox treatment, or other upper extremity rehabilitation program during the study period\nStudy design\nThe study consisted of eight weeks using the stimulation or sham device during daily life.\nParticipants wore the glove daily and met with blinded study administrators for weekly visits to measure sensorimotor function.\nAt the first visit, all participants received a device, cord and safety manual to take with them.\nParticipants were instructed to wear the device, turned on, for three hours every day while awake.\nUsers were notified that an onboard measurement unit would track usage time, and that 21 h of weekly use is required.\nAll participants were advised to charge the glove each night using the cord provided, just as one might do with a cell phone.\nThen, participants wear the device on-the-go or at home during their normal routine.\nWearing did not need to be continuous each day, but had to total three hours.\nThe dosage was chosen to be intensive, while not requiring too much daily commitment for participants.\nApparatus\nA wearable computing glove was designed to provide vibrotactile stimulation to participants throughout their daily life (Fig. 1).\nIt can be taken home and used outside of the clinic environment.\nAdditionally, the glove is worn while users conduct their daily life - making the rehabilitation low-effort and \u201cpassive.\u201d\nWearable device\nThe wearable device (\u201cVTS Glove\u201d) is designed to be low-cost, lightweight, and mobile.\nThe device is a fingerless glove with a vibration motor attached to each dorsal phalanx.\nThis design allows a designated actuator for each finger, while stimulating a region where vibrations can reach the glabrous skin of the palm and the finger extensor tendons.\nThe heart of the device is a circuit board and microcontroller, which activates these motors in a pre-programmed sequence when the switch is turned \u201con.\u201d\nThe onboard gyroscope logs movement data along with usage data onto a microSD card which is checked by proctors for protocol adherence each week.\nThe glove is rechargeable and has a battery life that allows wireless stimulation for four hours between charges.\nDesign and implementation of the device is reported in detail in a companion manuscript.\nStimulus design\nFor this experiment, stimulation characteristics were designed to target cutaneous mechanoreceptors \u2013 specifically the Pacinian corpuscles \u2013 which respond to direct vibration and vibration transmitted through the body at a frequency range of 10-400 Hz (preferentially responding around 250 Hz).\nStimulation pattern and timing was designed to be intensive but not uncomfortable by using many vibration pulses with a changing location across the fingers.\nSmall, coin-shaped vibration motors from Precision Microdrives (ERM-type, Model #310-113) provide the stimulation for this experiment.\nThese motors were driven at a voltage of 3.3V for an approximate amplitude of 1.5 g and 210 Hz vibration frequency (measured in a laboratory setting for validation at 1.3 g and 175 Hz when attached to the glove).\nStimulation was the same for all participants and could be perceived by the investigators.\nTwo stimulation sequences were used, each based on the finger pattern for a piano song.\nSong patterns provided a framework for pseudo-random stimulation and the option to later combine stimulation with music practice exercises for a lighthearted therapy routine.\nEach song pattern (Ode to Joy and Happy Birthday) was extended with a short sequence to balance stimulation evenly across all fingers.\nDuring each repetition the pattern played once quickly (250 ms vibrations, 100 ms pause between each stimulus) and once slowly (700 ms vibrations, 100 ms pauses).\nThese songs were chosen for their recognizable, one-handed melodies with 5-7 notes which could be played on the keyboard with little-to-no hand shifting.\nThe stimulation pattern was switched weekly, alternating between the two \u201csongs.\u201d\nConditions\nParticipants continued their standard of care, and none were enrolled in concurrent upper limb rehabilitation programs.\nIntervention condition\nParticipants in the vibrotactile stimulation (VTS) condition received a glove with vibration enabled.\nThe protocol includes no required exercises.\nParticipants were asked to wear their glove, switched on (so the indicator light appears), for three hours daily while awake.\nUsers should also charge the battery each night and as needed.\nControl condition\nParticipants in the sham control condition receive a glove with vibration disabled.\nThe appearance of the device was the same as the experimental condition.\nAll indicator lights on the computer board activate in the same fashion.\nInstructions and language also matched those in the VTS condition: wear the glove on their affected hand, switched on, for three hours daily while awake, and charge the battery each night.\nThe control condition was assigned a sham device (rather than no intervention) to examine the tolerance of the wearable device with and without stimulation, evaluate if the vibrotactile stimulation itself may have an impact on measures, and provide some data on mechanisms underlying this technique by comparing the conditions.\nOutcome measures\nBaseline demographic information collected was sex, age, date of stroke, type of stroke, and side affected.\nMeasurements are taken during weekly visits throughout the study.\nVisits occur at the patient\u2019s home or a midway meeting spot.\nAll measures were performed by trained proctors not involved in the intervention or data analysis.\nFor all participants, key measurements were taken by a blinded occupational therapist.\nThose measures were taken at the beginning (day 0), middle (4 weeks), and end (8 weeks) of the study.\nThe therapist and study proctor for each participant was consistent to minimize inter-rater variability.\nThe intent of this study was to examine the initial feasibility in this device and technique.\nThus, data on engineering, design, comfort and usability was collected through weekly surveys and observations.\nEngineering, comfort, and usability data are presented in another manuscript along with subsequent design work.\nHere we provide data on measures of arm function.\nAdherence for users in both conditions was measured each week using self-reported usage times matched with data from the glove\u2019s inertial measurement unit.\nIf usage time had not been within three hours of the required weekly time (21 hours) for two consecutive weeks, the participant would have been released from the study.\nNo such occurrences happened during the trial.\nPrimary outcome measures\nThe Semmes\u2013Weinstein Monofilament Exam (SWME) is used to assess cutaneous tactile perception in the affected hand.\nA 5-piece monofilament hand kit was used.\nLocations on the dorsal and volar side of the hand are assessed: each fingertip, each dorsal proximal phalanx, index and pinkey volar proximal phalanx, six points on the palm, and two points on the dorsal hand.\nThe SWME is a standardized measure with moderate reliability, greater than the static two-point discrimination test in some conditions.\nThis assessment was done weekly using the same brand of filaments, same evaluator, and filaments were replaced if damaged or bent.\nSecondary outcome measures\nThe Modified Ashworth Scale (MAS) is used to assess resistance to passive motion from involuntary muscle tone and spasticity .\nIn this study, MAS was measured for flexion and extension of the PIP/MCP finger joints, thumb, wrist, elbow and shoulder of the participant\u2019s affected upper limb.\nMAS ratings are reported here on a scale of 0-5.\nConfounding factors for this measure were controlled whenever possible including: time of day, time after medication dosage, arm position, and rater.\nVoluntary angular range of motion (Active Range of Motion or AROM) is used to assess motor impairment.\nThis measure can capture changes in function when participant dexterity is too low to perform tests like the Jebsen-Taylor Hand Function Test.\nHere, these measures were made for flexion and extension of the fingers, wrist, elbow and shoulder of the participant\u2019s affected upper limb.\nMeasures are taken in the neutral gravity plane whenever possible.\nCompensation from other muscles and synergy with spasticity are not included as voluntary range.\nFinger and elbow extension is measured from a flexed position, not from neutral, so as to report voluntary extension that may be used for activities such as releasing objects from grasp.\nA trained occupational therapist performed all movement and spasticity measures in a clinical setting at the beginning, middle and end of the study.\nEach week, participants are also given a worksheet to report what they did while wearing the device, observations, and comments about the device.\nData analysis\nUsing an intention-to-treat analysis, we processed data for all participants including two who had to withdraw prematurely due to unrelated circumstances.\nThe last measured values were used for the determination of any missing values in the case of dropouts or a missed visit, conservatively assuming that no changes occurred since the last measure.\nPaired observations were compared using the Wilcoxon signed-ranks test, and measures between groups were compared using the Mann-Whitney U test.\nRepeated measures were compared using a Friedman test and the Conover post-hoc test.\nA p-value < 0.05 was considered statistically significant.\nThe Semmes-Weinstein monofilament exam was taken at 20 points on the hand, yielding one minimum perceivable force value per location.\nThe minimum perceivable force level at each location was summed across all locations tested.\nA minimum sum of 1.4 grams (20 points  0.07 grams per point) corresponds to \u201cnormal\u201d sensation at all points and a maximum sum of 6000 g (20 points x 300 grams per point) corresponds to only \u201cresidual deep pressure sensation\u201d at all points.\nSmaller perceived forces equate to better tactile perception.\nAngular range of motion is reported, for clarity, at four body areas: the shoulder, elbow, wrist, and fingers.\nThe reported AROM value for each of these areas is the sum of the joint\u2019s movement (i.e., voluntary angular motion for the shoulder is the sum of shoulder flexion, extension, and abduction).\nFinger AROM is measured at the MCP and PIP joints, and those values are summed.\nThus, \u201cFinger Flex.\u201d and \u201cFinger Ext.\u201d include change in both the average MCP and PIP ranges.\nHealthy ranges for these measures would be as follows: shoulder , elbow (flexion + extension from flexed position) , wrist (flexion + extension + R/U deviation) , finger flex.\n(PIP + MCP flexion ROM per finger) , finger ext. (PIP + MCP extension ROM from flexed position, per finger) .\nResults\nSemmes\u2013Weinstein monofilament exam (SWME)\nStarting means (M=832.4 grams, SD=1206 for VTS; M=501.6 grams, SD=949.7 for control) were compared between groups using Mann-Whitney U test (U=18; z=\u22121.10; p=0.271).\nStarting ranges were 2705\u20131.79 for VTS, and 2448-1.66 for control.\nOne participant in the VTS condition is not included in this range and these calculations because their starting measures prevent representation on the graphs.\nThis user initially presented as insensate at all points but could accurately report deep pressure sensation at three points later in the study.\nBaseline measures of the VTS experimental group were compared to measures at eight weeks (M=9.701 grams, SD=14.25) and results suggest that there is a significant difference (t-test: t(6)=\u22123.50; p=0.006; signed-ranks: Z=\u22121.89; p=0.039).\nAs Fig. 3 shows, the VTS condition is able to sense smaller forces than the control condition at eight weeks (M=91.15 grams, SD=224.1).\nThe sham control condition also showed a change in SWME measures, but this change was not statistically significant (t-test: t(7)=1.190; p=0.254; signed-ranks: Z=\u22121.40; p=0.098).\nWe also performed a Friedman\u2019s test that found a significant difference in the repeated measures for the VTS group (X2F(8)=24.04, p=0.002), and no significant difference between measures for the control group (X2F(8)=17.29, p=0.27).\nA signed-ranks test suggests that the difference from baseline in the VTS group becomes significant at week 4.\nThe Conover post-hoc test adjusted by the Benjaminyi-Hochberg FDR method suggests a significant difference from baseline begins at week 5.\nFigure 4 shows the trends in these values throughout the entire study.\nModified Ashworth Scale (MAS)\nModified Ashworth Scale (MAS) was measured in a clinical setting for flexion and extension of MCP/PIP finger joints, thumb, wrist, elbow, and shoulder.\nHere, results are reported for the fingers (average of PIP and MCP joints) which showed the most change in values.\nStarting means (M=3.11, SD=1.08 for VTS; M=2.25, SD=0.78 for control) were compared using a Mann-Whitney U test (U=16; z=1.63; p=0.10) and no significant difference was found.\nEach participant\u2019s mean MAS can be found in Fig.\n5. Differences in experimental group MAS were found to be significant using a Wilcoxon signed-ranks test comparing starting measures to measures at 8 weeks (Z=-2.31; p=0.01).\nMAS difference at 8 weeks for the VTS condition was an average of -1.44 points on the Ashworth scale for each of the two measured joints on the affected limb (MCP and PIP).\nDifferences in control group MAS at 8 weeks were also compared using the Wilcoxon signed-ranks test (Z=1.06; p=0.15) but the average difference (M=\u22120.27) was not considered significant.\nChange from baseline was compared between conditions and found to be significantly different (Mann-Whitney: U=7.5; z=\u22122.48; p=0.007 ).\nParticipant 7 had severe spasticity before the study, which led to a Baclofen pump and wrist fusion surgery.\nThese interventions were failing to stop the progression of tone and spasticity in their hand; however, their tone was reduced after participation in the study.\nTwo users (5 and 7) agreed to follow-up six months post-study.\nThere was no significant relapse in values at follow-up vs. study end.\nActive range of motion (AROM)\nStarting means for arm motion and finger flexion had a significant difference between conditions.\nThe control group included fewer members with low to moderate starting function.\nBaseline function may be a factor in the AROM results for the control group, but further study is needed to examine its influence.\nThe control group showed no significant difference in shoulder (Starting Mean, , Avg. Change at Week ), elbow (Starting , SD=115.7, ), wrist (Starting , SD=38.9, ), finger flexion (Starting , SD=71.9, Avg. ) or finger extension range (, , ).\nThe experimental VTS condition showed improvements in sum of shoulder (Starting Mean=63.5, SD=66.7, Avg. Change at Week Eight=44.6), elbow (Starting Mean=54.1, SD=52.5, Avg.Change =69.5), and finger flexion (Starting Mean=25.0, SD=31.3, Avg. Change=50.9) range of motion.\nA Wilcoxon signed-ranks test found these changes to be significant (Z=-2.03, -2.17, -2.10, p=0.02, 0.01, 0.02), and this finding was consistent with a paired t-test (t(7)=2.59, 2.98, 2.16, p=0.02, 0.01, 0.03).\nChange in range of motion for the wrist (Starting Mean=9.5, SD=12.5, Avg. Change=16.5) and finger extension (Starting Mean=10.9, SD=18.1, Avg.Change=57.2) was not found to be statistically significant.\nChanges in voluntary ROM are shown in Fig. 6.\nDiscussion\nParticipants who received vibrotactile stimulation showed significant change in measures whereas those in the control group did not.\nThe wearable devices successfully delivered mobile stimulation throughout the duration of the study, and all participants were able to adhere to the daily wearing protocol.\nChanges in SWME measures suggest that participants showed improved tactile perception.\nFigure 4 suggests that the trend in improvement was gradual.\nOne participant reported the return of protective sensation in cases of joint hyper-extension, and one reported being able to feel the vibrations when they could not initially.\nSome participants in the VTS condition provided observations that the affected hand was more open and flexible.\nThese observations were consistent with changes in Modified Ashworth Scale measures.\nFigure 5 shows each person\u2019s starting and ending measures.\nAll but one person in the experimental condition showed a reduction in MAS values.\nParticipants in both conditions must frequently stretch open their affected hand to don the device, and stretching may be associated with changes in MAS.\nHowever, participants in the control group (who also stretched to don the device daily) did not show a significant change in MAS values, which suggests that stimulation rather than stretching is associated with these changes.\nTone and spasticity lack effective or lasting treatment options, yet 40-50% of stroke survivors with upper extremity disability may be affected .\nMore study is needed, but this promising preliminary evidence along with that in prior work suggests that afferent stimulation may be used to address spasticity and tone.\nThe VTS Glove allows extended stimulation and further study of this technique.\nFuture work can adjust stimulation characteristics to target different sensory receptors and examine optimal settings.\nSome participants with flexed fingers struggled to don the glove device, so the design was subsequently revised for accessibility.\nChanges in voluntary range of motion may be due in part to reduction in involuntary tone.\nSome participants showed large increases in range, with near-normal finger extension and flexion at week eight.\nOther participants showed no change in voluntary range of motion.\nFurther study can provide details on what markers, such as initial motor ability, predict outcomes using this device.\nImproved range of motion in proximal joints of the arm, such as the elbow, may be attributable to increased limb use and attention or to afferent input that reaches these proximal regions.\nVibration can be widely conducted throughout the human body via bones and other tissues.\nFuture work should examine if these results are maintained, but the informal follow-ups that were accepted by the two participants suggest that improvements may be lasting.\nSome participants had their stroke many years ago, and demonstrated change in measures.\nParticipants used the device for over 140 h each, an intensity enabled by the wearable form factor and the passive stimulation method.\nIn line with this result, a body of research has previously associated rehabilitation intensity (practice time) with improved outcomes.\nPrior studies applying WBV and focal muscle vibration also found reductions in spasticity and disabled limb function; however, most prior work is limited to stimulation for periods of 5-30 minutes in a laboratory setting because existing apparatus are large and immobile.\nParticipants in the experimental condition reported new capabilities on the weekly worksheet that included helping to cook, cleaning their hobby equipment, donning winter gloves and holding their partner\u2019s hand.\nThey also reported new tactile perception from the hand including sensing the vibrations, hyperextension during stretching, and the spray of water.\nThree participants reported a greater sense of embodiment or ownership of the limb.\nParticipants took advantage of the mobile nature of the device: reporting wearing the device to events such as church, lunch, and the movies.\nStudy limitations\nThis investigation intends to establish the feasibility of wearable vibrotactile stimulation to improve diminished limb function.\nParticipants include various levels of disability, which provides initial data on who may be suited for this stimulation.\nThe Modified Ashworth Scale is a standard measure of tone and spasticity, but there are confounding factors for this measure.\nThese factors were controlled whenever possible, including arm position, time of day, and rater.\nEffects of the control condition\nSome change in measures may be expected when using the sham device.\nThe sham device provided cutaneous sensory stimulation via the fabric of the glove; while the VTS experimental device provided additional cutaneous, and proprioceptive, stimulation via vibration.\nFurthermore, both conditions encouraged attention and engagement with the limb.\nHowever, in contrast to the experimental group, the control group did not show significant changes.\nPossible mechanisms behind changes in limb function\nA wearable device can facilitate engagement with the disabled limb, which may help discourage maladaptive plastic changes from sensory deprivation and learned non-use.\nLearned non-use is thought to be one of the reasons behind limited functional improvement of limbs after stroke: survivors learn to compensate and do not force themselves to re-learn the use of their limb.\nIn addition, participants stretched open their affected hand to don and doff the device several times per day.\nThis stretching was expected to impact Modified Ashworth measures.\nLesion location was not recorded in the study, but this information would provide interesting additional data if recorded in future work.\nThe control condition allowed us to examine the impacts of these mechanisms.\nAll participants interacted with a wearable device, but the experimental VTS group showed significantly different clinical measures after eight weeks.\nThis difference suggests that engagement and stretching may not be the only mechanisms to influence the participants.\nChanges in tactile perception may be due central mechanisms.\nAfferent input, transmitted by intact peripheral nervous pathways, may activate central nervous system regions.\nThis sensory input could impact central organization as is found in constraint-induced movement therapy after brain injury, or during normal sensorimotor skill acquisition.\nVibration may help regulate electrophysiology associated with spasticity via afferent feedback.\nReduced threshold of the stretch reflex has been implicated as one of the mechanisms behind symptoms of spasticity.\nSupraspinal control usually regulates this reflex, but can be disrupted in events such as spinal cord injury or stroke.\nThese reflexes are also mediated by afferent feedback produced during limb movement.\nVibration provides similar feedback \u2013 like many small muscle stretches \u2013 activating cutaneous mechanoreceptors and proprioceptive afferents.\nAfferent feedback then may induce reflex suppression and involuntary muscle contraction \u2013 which may impact spasticity and are found during whole body vibration (WBV) and focal muscle/tendon vibration.\nPresynaptic inhibition from afferent discharge is cited as a possible mechanism underlying reflex suppression during vibration.\nContinuous passive motion is another treatment for spasticity, but removal of proprioceptive afferents was shown to prevent normalization suggesting that sensory feedback may underlie this method.\nInvestigation of these factors is beyond the scope of this work, but the promising results warrant further study.\nImproved voluntary range of motion may be unlocked when spasticity and tone decreases.\nAnother hypothesis for changes in voluntary motion is that such sensory stimulation provides excitatory feedback and co-activation of motor systems, and helps restore somatosensation useful in motor function.\nThis hypothesis is supported by work in sensory stimulation for motor learning and performance, and motor rehabilitation.\nConclusions\nA controlled, randomized trial of 16 participants evaluated the feasibility of a wearable vibrotactile stimulation method to reduce upper limb disability in chronic stroke.\nAll users were assigned to wear a computerized glove on their affected hand for three hours per day.\nUsers in the sham control group received no stimulation and those in the experimental condition received vibrotactile stimulation from the glove.\nThe wireless, wearable device was used during daily life, not in a clinical setting.\nParticipants who received vibrotactile stimulation demonstrated a significant change in measures of tactile perception, voluntary motion, and spasticity after eight weeks.\nSome participants reported increase in protective sensation, sense of embodiment, and return to activities of daily living such as cleaning, cooking and writing using their disabled hand.\nThe computerized glove that provides vibrotactile stimulation for this study\nDemographics and notes for participants in the study. The experimental VTS group includes participants 1\u20138, and the sham control group includes participants 9\u201316. These participant numbers were assigned only to present data in this manuscript\nSemmes\u2013Weinstein Monofilament Exam results by group at baseline and eight weeks. This graph shows the group\u2019s average sum of perceived forces across 20 locations on the hand. Smaller perceived forces equate to greater tactile perception. Logarithmic scale used to render all force levels. Error bars indicate standard deviation plotted on a linear scale where each tick mark indicates 500 grams\nTrajectory of Semmes\u2013Weinstein Monofilament Exam results over eight weeks for both conditions. This graph shows the group\u2019s average sum of perceived forces across 20 locations on the hand. Smaller perceived force values equate to greater tactile perception. Logarithmic scale used to render all force levels. Shaded regions indicate the standard deviation over time (linear scale)\nModified Ashworth values for the fingers (average of PIP and MCP joint) at baseline and after eight weeks. The MAS rating scale is reported here as a scale of 0\u20135. Lower scores are better\nIncrease in angular degrees of voluntary movement for four upper body locations between baseline and study end. Shoulder, elbow and wrist values include both flexion and extension (from flexed) ranges. Finger flexion and extension is shown separately to provide greater detail, and these values include both MCP and PIP ranges. Zero values most often occurred when the participant had no voluntary movement in the joint at baseline and 8 weeks",
        "label": "unclear",
        "id": "task4_RLD_test_167"
    },
    {
        "paper_doi": "10.3390/tropicalmed4040141",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Design\ncNON-RCT*\nAllocation of clusters1 village randomized* to intervention, 1 to control\n\n\nParticipants: 527 individuals ages 3 to 70\n\n\nInterventions: Broad multiple\n\n\nOutcomes: Any STH\n\n\nNotes: *The study may have used a random mechanism to allocate the intervention, but there was only 1 intervention area compared to 1 control area, so randomization in this case not likely to have reduced confounding or imbalances\n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Many latrine campaigns in developing countries fail to be sustained because the introduced latrine is not appropriate to local socio-economic, cultural and environmental conditions, and there is an inadequate community health education component.\nWe tested a low-cost, locally designed and constructed all-weather latrine (the \u201cBALatrine\u201d), together with community education promoting appropriate hygiene-related behaviour, to determine whether this integrated intervention effectively controlled soil-transmitted helminth (STH) infections.\nWe undertook a pilot intervention study in two villages in Central Java, Indonesia.\nThe villages were randomly allocated to either control or intervention with the intervention village receiving the BALatrine program and the control village receiving no program.\nSTH-infection status was measured using the faecal flotation diagnostic method, before and eight months after the intervention.\nOver 8 months, the cumulative incidence of STH infection was significantly lower in the intervention village than in the control village: 13.4% vs. 27.5% (67/244 vs. 38/283, p < 0.001).\nThe intervention was particularly effective among children: cumulative incidence 3.8% (2/53) for the intervention vs. 24.1% (13/54) for the control village (p < 0.001).\nThe integrated BALatrine intervention was associated with a reduced incidence of STH infection.\nFollowing on from this pilot study, a large cluster-randomised controlled trial was commenced (ACTRN12613000523707).\n1. Introduction\nThe global prevalence of infection with soil-transmitted helminths (STH) remains high, with 1.5 billion people infected worldwide, many of them children.\nOver two thirds of STH infections are in Asia, mostly in Southeast Asia.\nThe prevalence of STH infection in Indonesia is high at 45\u201365%, with areas having poor sanitation reaching 80% prevalence.\nIn Central Java, research into STH infection among elementary school children by Laksono and later the Health Department, found an infection prevalence of 84\u201396%.\nMore recently, a cross-sectional survey in Semarang, Central Java, found a prevalence of 34% among a cohort of 6466 people aged between two and 93 years.\nAnthelmintic drugs aimed at reducing morbidity are effective, but only temporarily, with a cure often followed by subsequent reinfection.\nIn rural areas, open defecation coupled with poorly constructed or inadequate latrines allows STH eggs to spread infection.\nTherefore, for long-term prevention, improved sanitation and community education are essential.\nA recent systematic review and meta-analysis concluded that \u201cintegrated control approaches emphasizing health education and environmental sanitation are needed to interrupt transmission of STH\u201d.\nIn particular, interventions that improve the hygienic disposal of faeces to reduce soil and/or water contamination have been identified as a key strategy to control transmission and prevent related diseases.\nIn Indonesia, open defecation is common, with 55% of the poorest and 18% of the richest households practicing open defecation.\nIn 2010, less than 40% of the people in rural areas had improved latrines, defined as facilities that hygienically separate human excreta from human contact.\nThe country did not reach its Millennium Development Goal of 75% sanitation coverage by 2015.\nIn rural areas, which include 118 million people, or 46.3% of the country\u2019s population, it has been estimated that 47% of the population have improved latrines, 12% shared latrines, 12% other unimproved latrines and 29% no latrines (i.e., practice open defecation).\nCompounding the problems caused by the lack of improved latrines is inappropriate hygiene-related behaviour, particularly related to hand washing, with 2007 National baseline data indicating that less than a quarter (23.2%) of the population had appropriate hand-washing behaviour.\nThe aim of this study was to develop and test an integrated approach to the prevention of STH infection and reduction of both transmission and reinfection.\nOur intervention included anthelmintic drugs, the construction and adoption of improved latrines, and effective education regarding hygienic and sanitary behaviour.\n2. Methods\n2.1. Study Design\nThis study was conducted in two villages, Palemon and Cepoko, in the Gunungpati sub-district of the city of Semarang, Central Java, Indonesia (see Figure 1), from July 2011 to June 2012.\nA random selection was made from these villages regarding which one should receive the integrated intervention and which one should be the control, by researchers who had no prior knowledge or contact with the villagers or village officials.\nThe two villages though similar in size and local topography, were not in close proximity to each other.\nThe study area is wooded and hilly and most of the houses, often made of local brick, are constructed by the householders themselves.\nMore than half of the households in the study villages did not have their own latrines.\nA randomly selected cohort (control: n = 244; intervention: n = 283) was followed over the eight-month duration of the study.\nA questionnaire was administered at baseline and follow-up to all village residents, with the eligibility criterion of being over two years of age.\nParticipants also provided two stool samples for parasitological examination.\nFollowing the baseline survey, all residents (regardless of infection status) were treated with anthelmintic medication and the incidence of STH infection was assessed at follow-up.\nFor ethical reasons, participants who were found to be STH positive at follow-up were re-treated.\n2.2. Ethics \nEthical approval was given by the Semarang City authorities (ref. 070/613/IV/2011), and from the Human Research Ethics Committee at Griffith University (ref. PBH/17/11/HREC).\n2.3. Procedure\nOur study procedure reflected the integrated model previously described (see also Figure 2), comprising chemotherapy, a locally constructed latrine (the \u201cBALatrine\u201d) and community health education.\nFollowing WHO Guidelines, a single oral dose of Albendazole (400mg) was administered immediately after the baseline survey.\nThe BALatrine is designed for resource poor rural communities and emergency situations, to be made by local people using local materials.\nTesting for cultural acceptance was conducted in the field through pilot studies in Pekalongan in Central Java and BALatrines were also used in an emergency refugee camp during the 2010 eruption of the Mt Merapi volcano, where they were proven appropriate for the level of technology available in the village context.\nThe BALatrine is a relatively simple squat latrine (Figure 3) that can be constructed by village residents using inexpensive materials.\nWhen water for flushing is available, a U-bend (\u2018goose-neck\u2019) water closet can be attached.\nWhen water is not available, such as during the dry season, the latrine can be used in a dry-pit configuration (with removal of the U-bend attachment), with a lid to isolate it from insects and to prevent odours from escaping.\nThus, it can function despite seasonal changes in water supply.\nBesides being inexpensive (cost at the time of the study was $50 USD for the latrine and local building materials; equating to approximately $80\u201390 at the time of publication), for people with limited financial and educational resources it is easy to copy.\nIts construction and use reflect critical resource, environmental and technical issues and due to local village input it also overcomes some major disincentives embodied in conventional latrines by being culturally familiar, simple and easy to use.\nThe community health education programme is an essential adjunct to latrine construction.\nIn the intervention village, all residents were given health education regarding hygiene, sanitation, and prevention of STH infections.\nThis health education component was delivered via community meetings in each village.\nAll village residents were invited and meetings were held in the village meeting hall.\nThe health education/health promotion component of the intervention was implemented through a two-hour village-wide mobilization meeting, which formed the project launch and was designed to mobilize households by consciousness-raising and provision of information about parasite infection and burden of STH.\nSubsequently, a series of small group workshops took place with the villagers in order to describe the BALatrine construction in detail and how to plan, construct, use, and maintain their latrines, as well as to discuss STH disease pathways.\nThe content of the health education programme comprised information about the dangers of STH infections and, using illustrated leaflets, how the transmission of STH infections can be prevented by the construction of latrines and with appropriate hygiene-related behaviours.\n2.4. Measurements and Analyses\nThe primary outcome measures of the integrated intervention were STH infection status at baseline and at follow-up.\nSTH infection status was measured through laboratory analysis of stool samples collected from each participant at baseline and at follow-up eight months after the BALatrines were constructed.\nThe samples were analysed microscopically for the presence of helminth eggs, according to the Willis-Mollay Flotation technique.\nAfter preparing the samples, a cover slip was placed over each sample tube and left for 10 minutes.\nAfter 10 minutes, a drop of eosin solution (2%) was added to a glass slide onto which the cover slip was then placed and observed using a light microscope at 10 \u00d7 magnification.\nA positive sample was where at least one STH egg was identified.\nA face-to-face questionnaire (the Helminth Education and Latrine Project (HELP) questionnaire) was also administered at baseline and follow-up and this provided information about villagers\u2019 demographic attributes.\nWe also assessed local village environmental contamination with faeces.\nThese findings have been published elsewhere.\nData were analysed using SPSS Version 22 (IBM, New York, United States), Microsoft Excel and the tools at Open Source Epidemiologic Statistics for Public Health.\nDifferences between participants in control and intervention villages were analysed using the unpaired t-test and the Pearson\u2019s chi-squared (\u03c72) test.\nLogistic regression analysis was performed and odds ratios calculated.\nTo compare the intervention village with the control village, we report both crude odds ratios and adjusted for age and sex.\n3. Results \n3.1. Characteristics of the Participants\nThere were 527 participants in the study at baseline, 244 in the control village and 283 in the intervention village.\nTheir ages ranged from three to 70 years, with a mean \u00b1 SD of 29.4 \u00b1 16.1 years in the control village and 32.2 \u00b1 17.9 years in the intervention village (Table 1; for the age difference between villages, p = 0.06).\nIn both villages, similar proportions of the participants had completed elementary education (control, 211/244, 86.5%; intervention, 253/283, 89.4%).\nIn the control village, 98.8% of the residents had a monthly income below 1 M Indonesian Rupiah (IDR) or about US$70, whereas in the intervention village the comparable percentage was 97.2% (p = 0.017).\nIn the control village, 38.5% of the residents lived in a home in which all floor spaces were dry, whereas in the intervention village the comparable percentage was 54.1% (p < 0.001).\nIn the intervention village greater than 90% of households adopted the BALatrine, as measured at the follow up.\n3.2. STH-Infection Status\nAt baseline, the prevalence of STH infection was almost the same in the two villages (Table 1).\nAt follow-up, the cumulative incidence of infection was much lower in the intervention village than in the control village (Table 2, 13.4% vs. 27.5%, p < 0.001).\nAfter adjustments for age and gender, the benefit of the intervention was still clear (adjusted odds ratio = 0.38, 95% CI 0.25\u20130.60, Table 2).\nThe intervention was particularly effective in children (adjusted odds ratio = 0.12, 95% CI 0.03\u20130.56, Table 2).\n4. Discussion\nImproving access to adequate sanitation is a critical step toward the sustainable interruption of STH transmission.\nYet this is an ongoing challenge in low resource settings where public sewerage system infrastructure rarely exists, on-site sanitation systems are either improperly designed or poorly functioning, and open defecation is seen as culturally acceptable, especially in rural areas.\nOften people must rely on shared sanitation facilities, which have been shown to increase the risk of adverse health outcomes compared to individual household latrines.\nCompounding the issue is limited access to materials and a lack of technical expertise to build or improve latrines.\nThe BALatrine was specifically designed to overcome many of these challenges whilst also taking into consideration cultural appropriateness and acceptability.\nUsing simple technology and locally sourced, inexpensive materials, the BALatrine is cheap, easy to build and maintain and adaptable for wet and dry conditions.\nBuilding the latrines through community mobilisation also helps keep the costs down and enables the householders themselves to take ownership over the latrines and their maintenance and consequently, benefit from improved health, helping alleviate the cycle of poverty that is often associated with STH infections.\nIn the present study, we evaluated the effectiveness of an integrated BALatrine intervention at reducing human worm burden through a pilot study in two villages in Central Java.\nPeople in the intervention village were 2.6 times less likely to be infected following the BALatrine-based intervention than those in the control village indicating that the BALatrine is associated with a reduced worm burden.\nHowever, it is important to note that we did not resurvey participants after baseline deworming to determine the efficacy of the Albendazole treatment nor did we differentiate between STH species, which are known to respond differently to treatment.\nWe have based our interpretation of the results on the assumption that treatment was effective at temporarily reducing infections to zero.\nConsequently, it is possible that the effect seen could be a result of differential cure rates between villages based on their STH profile at baseline.\nHowever, a prevalence study we conducted the following year (manuscript under review) across 16 villages in two neighbouring subdistricts of Semarang, including Gunung Pati, revealed Ascaris lumbricoides as the predominant species (mean prevalence of 26% vs. 7.9% and 1.8% for hookworm and Trichuris trichuria, respectively).\nWe therefore believe that it is not unrealistic to assume that our study villages also had similar burdens of each of the STH species at baseline and that treatment would be similarly effective across the two sites.\nOur study also found that nearly all households adopted the BALatrine suggesting a strong willingness among villagers to improve sanitation and a desire to improve the health of their families.\nAccess to improved latrines does not guarantee their use, however, particularly over time when old habits or cultural preferences can be difficult to overcome.\nIntegrating health education and promotion programmes to improve peoples\u2019 understanding and knowledge of the link between open defecation and ill-health and WASH-related behaviours, is therefore extremely important.\nPeople also need to understand the importance of properly cleaning and maintaining their latrines, particularly as this is associated with higher latrine use.\nIn the present study, a community health education programme was administered prior to the construction and installation of the latrines and aimed to raise awareness, improve hygiene behaviours and motivate the villagers to build and continue to use their new latrines, which may have led to the high uptake observed in this study.\nHowever, we did not assess the impact of this program on participants\u2019 knowledge and behaviour change or assess latrine use over time, which are key limitations of this study.\nIt is well established that chemotherapy alone will not break the transmission cycle.\nRecent studies have also shown that programmes solely focusing on WASH have limited effect on STH incidence and may provide no additional impact compared to mass drug administration programmes alone.\nIn contrast, health education and promotion programmes can be highly effective at reducing the incidence of STH infections if designed appropriately such as the highly successful \u201cMagic Glasses\u201d study, which resulted in a 50% reduction in STH infections.\nHowever, sustained reinforcement of health messages is required in order to increase their effectiveness over the longer term.\nUltimately, eliminating STH will be best achieved through integrated control programmes.\nThe current study adds to the growing body of research into the impact of integrated control programs on soil-transmitted helminthiases and demonstrates that sanitation interventions can be effective at reducing worm burden when designed appropriately for the local context and combined with health education and promotion.\nIn conclusion, our findings provide \u201cproof of principle\u201d that the BALatrine-based intervention is effective in preventing STH infection.\nWe will now undertake a full-scale randomized controlled trial and contribute much needed evidence based on WASH and STH.\nMap of Gunung Pati subdistrict in Semarang, Central Java (source: Wikipedia Indonesia, 2017).\nFlowchart of the study.\nSchematic presentation of the BALatrine.\n\nBaseline characteristics of participants.\nVillage Status | Control | Intervention\nVillage  | Cepoko | Palemon\nSample Size | 244 | 283\nMean Age (years) | 29.4 | 32.2\nPrevalence of STH infection: % (95% CI) | 21.7% (16.5\u201326.9) | 25.8% (20.7\u201330.9)\nSex Ratio (F/M) | 141/103 | 151/132\nPrevalence of STH infection by Sex (F/M)  | 22.0%/21.4% | 20.5%/31.8%\n\n\nInfection rates in the control and intervention villages.\nVariable | Control | Intervention | Odds Ratio | Odds Ratio\nCrude | p Value | Adjusted * | p Value\nAll participants | n = 244 | n = 283 |  |  |  | \nPrevalence of infection at baseline: % (95% CI) | 21.7 (16.5\u201326.9) | 25.8 (20.7\u201330.9) | - | - | - | -\nCumulative incidence of infection at follow-up: % (95% CI) | 27.5 (21.9\u201333.1) | 13.4 (9.5\u201317.4) | 0.41 (0.26\u20130.64) | <0.001 | 0.38 (0.25\u20130.60) | <0.001\nChildren | n = 54 | n = 53 |  |  |  | \nPrevalence of infection at baseline: % (95% CI) | 18.8 (8.2\u201328.9) | 18.9 (8.3\u201329.4) | - | - | - | -\nCumulative incidence of infection at follow-up: % (95% CI) | 24.1 (12.7\u201335.5) | 3.8 (0.0\u20138.9) | 0.12 (0.03\u20130.58) | 0.01 | 0.12 (0.03\u20130.56) | 0.01\n\n* The model for all participants was adjusted for age and sex. The model for children (<14 years) was adjusted for gender only.",
        "label": "unclear",
        "id": "task4_RLD_test_816"
    },
    {
        "paper_doi": "10.1371/journal.pone.0000542",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Trial design: randomized, open\nTime period/duration of trial: 5 June 2005-8 September 2005\nDuration of follow-up: 6 months after enrolment\n\n\nParticipants: Setting: presenting to the outpatient or emergency department of Patan Hospital\nLocation: Lalitpur, Nepal\nAge: 2-65 years; median age 17 (range: 2-64 years)\nGender: male = 247, female = 135\nHealth status of participants: not recorded\nInclusion criteria:clinically diagnosed enteric feverresidence within 2.5 km of the hospitalability to take oral medicationsno pregnancy/lactationno history of seizuresability to stay in the city for the duration of the treatmentno contraindications to quinolones or cephalosporinsability to give written informed consentExclusion criteria:signs of complicated typhoid defined as the presence of jaundice, gastrointestinal bleeding, peritonism, shock, encephalopathy, convulsions, myocarditis or arrhythmia at the time of enrolmentreceipt of a third-generation cephalosporin, fluoroquinolone or macrolide in the week prior to presentation at the clinic\n\n\nInterventions: Cefixime: oral, 20 mg/kg/day in 2 divided doses for 7 daysGatifloxacin: oral, 10 mg/kg/day as a single dose for 7 days\n\n\nOutcomes: Primary outcomeFever clearance time - time to first drop in oral temperature <= 37.5 degC and remaining as such for 48 hSecondary outcomesAcute treatment failure: severe complications, persistence of fever > 38 degC, persistence of symptoms for > 7 days after start of therapy, requirement for additional or rescue treatmentOverall treatment failure: acute treatment failure + relapse + deathRelapse: fever with a positive blood culture within 1 month of completing treatment in a patient who had been successfully treated. Excluded individuals given prolonged or rescue treatmentCultures: faecal samples taken at end of 1st, 3rd and 6th monthAdverse events: not prespecified, listed in the results\n\n\nOrganism type and antimicrobial susceptibility: No breakdown of results by organism/susceptibility\n\n\nNotes: 482 clinical cases of enteric fever, with 390 enrolled and randomized.92 were ineligible, with the commonest reason being receipt of antibiotics in the week before trial entry (n = 49).187 participants assigned to receive cefixime, and 77 of these were culture-positive; 203 were assigned to gatifloxacin and 92 of these were culture-positiveBoth ITT (all 390 randomized participants) and positive pre-treatment culture analyses (defined as the per protocol participants, with positive pre-treatment culture results).We requested and received data from the trial authors to calculate the mean time to defervescence\n\n",
        "objective": "To evaluate the effectiveness of cephalosporins for treating enteric fever in children and adults compared to other antimicrobials.",
        "full_paper": "Objective\nTo assess the efficacy of gatifloxacin versus cefixime in the treatment of uncomplicated culture positive enteric fever.\nDesign\nA randomized, open-label, active control trial with two parallel arms.\nSetting\nEmergency Room and Outpatient Clinics in Patan Hospital, Lagankhel, Lalitpur, Nepal.\nParticipants\nPatients with clinically diagnosed uncomplicated enteric fever meeting the inclusion criteria.\nInterventions\nPatients were allocated to receive one of two drugs, Gatifloxacin or Cefixime.\nThe dosages used were Gatifloxacin 10 mg/kg, given once daily for 7 days, or Cefixime 20 mg/kg/day given in two divided doses for 7 days.\nOutcome Measures\nThe primary outcome measure was fever clearance time.\nThe secondary outcome measure was overall treatment failure (acute treatment failure and relapse).\nResults\nRandomization was carried out in 390 patients before enrollment was suspended on the advice of the independent data safety monitoring board due to significant differences in both primary and secondary outcome measures in the two arms and the attainment of a priori defined endpoints.\nMedian (95% confidence interval) fever clearance times were 92 hours (84\u2013114 hours) for gatifloxacin recipients and 138 hours (105\u2013164 hours) for cefixime-treated patients (Hazard Ratio[95%CI]\u200a=\u200a2.171 [1.545\u20133.051], p<0.0001). 19 out of 70 (27%) patients who completed the 7 day trial had acute clinical failure in the cefixime group as compared to 1 out of 88 patients (1%) in gatifloxacin group(Odds Ratio [95%CI]\u200a=\u200a0.031 [0.004 \u2013 0.237], p<0.001).\nOverall treatment failure patients (relapsed patients plus acute treatment failure patients plus death) numbered 29.\nThey were determined to be (95% confidence interval) 37.6 % (27.14%\u201350.2%) in the cefixime group and 3.5% (2.2%\u201311.5%) in the gatifloxacin group (HR[95%CI]\u200a=\u200a0.084 [0.025\u20130.280], p<0.0001).\nThere was one death in the cefixime group.\nConclusions\nBased on this study, gatifloxacin is a better treatment for uncomplicated enteric fever as compared to cefixime.\nTrial Registration\nCurrent Controlled Trials ISRCTN75784880\nIntroduction\nEnteric fever (Typhoid and Paratyphoid fever) is a systemic infection caused by the bacterium Salmonella enterica serovar Typhi (S.typhi) or Salmonella enterica serovar Paratyphi (S. paratyphi) which in humans is transmitted through the fecal-oral route.\nToday the vast burden of disease is encountered in the developing world where sanitary conditions remain poor.\nThe best global estimates are of at least 22 million cases of typhoid fever each year with 200,000 deaths.\nCrucially these are almost exclusively confined to resource poor countries.\nA recent Cochrane review on typhoid treatments underscored the need for large sample size drug interventional trials, especially in children in whom this disease predominates.\nIn 1948 the introduction of chloramphenicol revolutionized the treatment of typhoid fever.\nUnfortunately the emergence of resistance to the \u201cfirst line\u201d antimicrobials (for example, ciprofloxacin) has been a major setback and has given rise to the possibility of untreatable enteric fever.\nGatifloxacin, a relatively inexpensive fluoroquinolone antibiotic in South Asia with once daily oral administration, is a new broad spectrum synthetic 8-methoxyfluoroquinolone which has the lowest minimum inhibitory concentration (MIC) against S. typhi from Nepal.\nThis in vitro activity needs to be verified clinically before gatifloxacin can be recommended for widespread use.\nCefixime, an orally administered third generation cephalosporin, is a commonly used drug in South Asia for the treatment of enteric fever.\nAlthough cefixime is recommended as a drug of choice by the World Health Organization (WHO) for the treatment of resistant typhoid fever it is relatively expensive in South Asia and has to be administered for a longer duration than the currently used fluoroquinolones.\nClearly there is an urgent need for a treatment that combines ease of oral administration, with speed of clinical response, reduction in secondary transmission and inexpensiveness.\nIn this open randomized trial, we aimed to compare clinical outcomes for the treatment of uncomplicated enteric fever with gatifloxacin or cefixime in an outpatient setting.\nMethods\nParticipants\nThe study was approved by Nepal Health Research Council and Oxford Tropical Research Ethics Committee.\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nWe enrolled patients who presented to the outpatient or emergency department of Patan Hospital, Lalitpur, Nepal from June 5, 2005 to September 8, 2005.\nPatan Hospital is a 318 \u2013bed hospital located in the Lalitpur district in Kathmandu Valley.\nPatients were eligible to enter the study if they had clinically diagnosed enteric fever and their residence was within approximately 2.5 km radius from the hospital.\nOther inclusion criteria were that patients must be aged between 2 and 65 years, able to take oral medications, non-pregnant and non-lactating, without a history of seizures, able to stay in the city for the duration of the treatment, not known to have contraindications to either cephalosporins or fluoroquinolones and willing to give informed written consent to take part in the study.\nFor children enrolled into the study, written informed consent was taken from a parent.\nPatients were excluded from the study if they had any signs of complicated typhoid defined as the presence of jaundice, gastrointestinal bleeding, peritonism, shock, encephalopathy, convulsions, myocarditis or arrythmia at the time of enrollment.\nPatients who had received a third generation cephalosporin, fluoroquinolone or macrolide in the week prior to presentation to our clinic were also excluded.\nInterventions\nOn presentation to Patan Hospital all patients with fever without an obvious focus were referred to the enteric fever study clinic, where they were seen by the study physician.\nPatients who fulfilled the inclusion criteria were randomly assigned to receive Gatifloxacin (Broadband\u2122, Novartis AG Basel, Switzerland) 10 mg/kg/day, in a single dose orally for 7 days or Cefixime (Cifex\u2122, Aegis, Nicosia, Cyprus) 20 mg/kg/day in two divided doses orally for 7 days.\nBoth drugs were administered in tablet form, cut and weighed in a sensitive scale to ensure that underdosing did not occur.\nTo children who were apprehensive of swallowing the tablet, the drug was embedded in a banana and given.\nAll patients were asked to swallow the study drug under direct observation during each visit.\nEach patient had haematocrit, total leucocyte count with differential, serum creatinine, total bilirubin, alanine aminotranferase(ALT), and aspartate aminotransferase(AST) measured, and blood and stool cultures were also performed before the start of the study intervention.\nThe exact location of the patient's home was recorded and the first dose of drug administered at the clinic.\nWe employed six Community Medical Auxiliaries (CMA) who had all received at least 15 months of prior formal primary health care worker training and been registered in a government recognized institution.\nThe CMAs visited patients twice daily at their homes to perform a simple clinical assessment, measure the oral temperature and give directly observed therapy with the study drugs.\nThe CMA visited the patient's home every 12 hours, morning and evening, until day 10 following enrollment or complete resolution of illness, whichever came later.\nThe oral temperature of the patient was recorded twice every day by the CMA and a note was made of the timing and dosages of acetaminophen intake.\nThe quality of patient-visits was ensured by regular unplanned supervisory checks in which the study doctor accompanied the CMA during the visits to patients' homes\nCMAs were asked to send patients immediately to the hospital on encountering any severe symptoms, and the patients also were asked to attend clinic if they had any severe symptoms at any other time.\nA symptom questionnaire was used daily during each visit to monitor any adverse events.\nAny patient with any severe symptom was seen by the study physician.\nThe CMAs and study physicians held daily case conferences at which all the study patients were discussed.\nAll patients regardless of the culture results were seen at hospital on Day 10 following enrollment.\nBlood and stool cultures were repeated on Day 10 in all culture positive patients and thereafter if the patient again became ill with probable enteric fever.\nAll culture positive patients were followed up until six months after enrollment, and stool cultures were performed at the end of the first, third and sixth month.\nMicrobiological Procedures\nBlood culture was performed on media containing tryptone soya broth and sodium polyethanol sulphonate, incubated at 37 C and examined daily for growth over 7 days.\nSalmonella enterica serotype Typhi or Paratyphi A, B or C isolated in culture were identified using standard biochemical tests and specific antisera (Murex Biotech, Dartford, England).\nAntibiotic susceptibilities were determined during isolation using the Kirby-Bauer disc diffusion method involving antibiotic discs containing Nalidixic acid, Ofloxacin, Ciprofloxacin, Chloramphenicol, Ampicillin, Cotrimoxazole, Cefixime and Cefotaxime (HiMedia Laboratories, Mumbai, India).\nMinimum Inhibitory Concentrations (MICs) were determined later for organisms stored in glycerol (bacterial preserver) at -70C.\nThe MICs were determined by Chloramphenicol, Nalidixic acid, Gatifloxacin, Cefixime, Ceftriaxone and Gemifloxacin E-tests\u2122 (AB Biodisk, Solna, Sweden), according to the manufacturer's instructions.\nThe sensitivity tests were interpreted using Clinical and Laboratory Standards Institute criteria for Enterobacteriaceae.\nObjectives\nThe objective of the study was to compare the efficacy of Gatifloxacin and Cefixime in the treatment of uncomplicated culture positive enteric fever.\nOutcomes\nThe primary outcome was the fever clearance time (FCT).\nFCT was defined as time to first drop in oral temperature \u2264 37.5\u00b0C, remaining \u2264 37.5\u00b0C for 48 hours.\nThe secondary outcomes included acute treatment failure.\nAcute treatment failure was defined as including any severe complication; the persistence of fever (> 38 C); the persistence of symptoms for more than 7 days after the start of treatment , requiring additional or rescue treatment.\nIf a patient had a temperature above 37.5 and below 38 for more than 7 days, but did not need additional or rescue therapy, and subsequently their fever cleared by day 10, that patient would not qualify as an acute treatment failure.\nPatients who failed the study treatment were given rescue treatment.\nThe rescue drug was Ofloxacin 20 mg/kg/day orally in two divided doses for 14 days for the Cefixime group, and Ceftriaxone 40 mg/kg/day IV in a single daily dose for 14 days for the Gatifloxacin group.\nFor the Cefixime group alone, if on day 8 of treatment the patient still had a fever of >\u200a=\u200a38\u00b0C, the study drug was continued for 10 days and the patient categorized as acute treatment failure.\nIf the temperature on Day 10 was >37.5, rescue treatment was given.\nA relapse was defined as fever with a positive blood culture within a month of completing treatment.\nAll the relapses were patients that were initially categorized as successfully treated.\nAny patient given rescue treatment or prolonged treatment was precluded from the \u201crelapse\u201d group.\nPatients categorized as \u201coverall treatment failures\u201d included patients experiencing acute treatment failure, plus those falling into the relapsed category, plus all deaths within the trial follow up period.\nSample size\nThe sample size was calculated to detect a FCT difference of approximately 48 hours between gatifloxacin (assumed median FCT 156 hrs) and cefixime (assumed median FCT 204 hrs) with p\u200a=\u200a0.05 and power\u200a=\u200a80%.\nThe accrual time for recruitment was assumed to last 70 days, and that the last patient would be followed up until 8 days after recruitment.\nTherefore, we estimated the minimum sample size at 235 participants.\nAssuming a loss to follow-up of 5%, the sample size was calculated as 125 blood culture positive patients in each arm.\nBefore the recommended sample size had been reached, once 169 blood culture positive patients had been enrolled, the independent data safety monitoring committee (DSMC) advised the Principal Investigators to stop recruitment to the trial based on a priori defined difference (p<0.01) between the two treatment arms in the primary endpoints of the study.\nRandomization\u2014Sequence generation\nPatients were randomized in blocks of 100 from a computer generated randomization list, by an investigator not involved in patient recruitment or assessment.\nRandomization\u2014Allocation concealment\nThe randomization sequence and block size was concealed from the physicians allocating treatment and managing the patients, prior to patient enrollment.\nTreatment allocations were kept in sealed opaque envelopes, which were opened only on enrollment of the patient to the study after all inclusion and exclusion criteria had been checked.\nRandomization\u2014Implementation\nParticipants were enrolled by the study physician in the same order in which they presented to the study clinic.\nThe sealed envelopes were opened in strict numeric sequence.\nBlinding\nBlinding was not feasible in this trial due to logistical reasons.\nStatistical methods\nAll data were entered into an electronic database (Microsoft Office Access Version 2003, Wash., USA), and analyses was performed using Stata 9 (Stats Corp LP; Texas, USA). ).\nContinuous covariates were compared between groups of patients using the Mann-Whitney test, and categorical covariates were compared using the chi-square test or Fisher's exact test when appropriate.\nFever clearance times and time to relapse were analyzed using Kaplan Meier survival curves and compared between the two groups using the logrank test.\nBinary outcomes (clinical failures) were compared between the two treatment groups using Fisher's exact test.\nAnalysis was done in all randomized patients (intention to treat, ITT) and separately in patients with positive pretreatment culture (per protocol, PP) and negative pretreatment culture.\nResults\nParticipant flow\nOf the 482 patients from the study area who were clinically diagnosed with enteric fever, 390 patients were enrolled into the study and randomized.\n92 patients were ineligible, the main reason (49 patients) being a history of already having taken antibiotics (fluoroquinolone, macrolide, or third generation cephalosporin) within one week prior to study entry (Figure 1).\nAmong all randomized patients, 187 patients were assigned to receive cefixime and 203 to gatifloxacin.\n77 patients assigned to receive cefixime were blood culture positive for enteric fever whilst 92 of those assigned to receive gatifloxacin were culture positive .\nThere were unequal number of positive patients in each of the study arms.\nOne possible reason for the difference in number of culture positive patients between study arms is that cultures were drawn and culture results obtained after randomization had been done.\nRecruitment\nWe enrolled patients who presented to the outpatient or emergency department of Patan Hospital, Lalitpur, Nepal from June 5, 2005 to September 8, 2005.\nAll enrolled patients were followed up for at least 10 days after recruitment.\nPatients with a positive pretreatment blood culture were followed up for 6 months after enrollment.\nAt the point that the DSMC asked to examine the trial data for the primary outcome measure in positive pre-treatment patients, the median fever clearance time was 92 hours\u00a0(95% CI, 84\u2013114 hours) for the gatifloxacin treated patients and 138 hours\u00a0(95% CI, 105\u2013164 hours) for cefixime treated patients.\nThe difference between the two treatment arms was 46 hours (p<0.0001).\nBaseline data\nAdmission characteristics are shown for all ITT patients (Table 1) and for all PP patients (Table 2).\nThe median age of patients enrolled into the trial was 17 with a range of 2\u201364 years.\nThere were no baseline differences in the culture positive and culture negative groups, other than temperature at presentation, AST and ALT which were higher and platelets and total WBC which were lower in the culture positive patients as compared to the culture-negative patients.\nAmong all PP patients, there were no differences in the baseline characteristics between the two treatment groups.\nThere were 40 patients, 15 in the gatifloxacin arm and 25 in the cefixime arm, who had taken amoxycillin up to the week before study entry.\nOf these 4 and 7 were culture-positive respectively.\nNumbers analyzed\nAnalysis was done in all 390 randomised patients (ITT) and separately in 169 patients with positive pre-treatment culture (PP).\nAll endpoints were analysed in the ITT and PP populations, apart from relapse which was only analysed in the PP population.\nOutcomes and estimation\nPrimary outcome\nIn all ITT patients, median (95% confidence interval) fever clearance time was 102 (90\u2013117) hours for the cefixime group and 72 (62\u201380) hours for the gatifloxacin group, logrank test p<0.0001, Hazard Ratio[95%Confidence Interval]\u200a=\u200a1.821 [1.466\u20132.263].\nThe proportion of all patients failing through time to clear fever is shown in Figure 2.\nAt day 7 fever clearance rate was 73.9% (67.0% \u2013 80.3 %) in cefixime group and 94.2% (90.2% \u2013 96.9%) in gatifloxacin group.\nIn the PP group, median (95% CI) fever clearance time was 92 hours (84\u2013114 hours) for gatifloxacin recipients and 138 hours (105\u2013164 hours) for cefixime-treated patients (HR[95%CI]\u200a=\u200a2.171 [1.545\u20133.051], p<0.0001).\nThe proportion failing to clear fever for each study drug through time after treatment is shown (Figure 3).\nAt day 7 the fever clearance rate was 62.7% (95 % CI\u200a=\u200a51.5%\u201373.8%) in the cefixime group and 91.8% (95 % CI\u200a=\u200a84.8%\u201396.4%) in the gatifloxacin group.\nIn the group with negative blood culture but clinically diagnosed enteric fever (Fig 1), the FCT was 82 hours (95% CI\u200a=\u200a44\u201394 hours) for the cefixime group and 39 hours (95%CI\u200a=\u200a28\u201354 hours) for the gatifloxacin group (HR[95%CI]\u200a=\u200a1.740 [1.309\u20132.312], p<0.0001 logrank test).\nSecondary Outcomes\nIn the ITT group, overall, 30 out of 167 (18%) in the cefixime group and 2 out of 190 (1%) in the gatifloxacin group were acute clinical failures, OR[95%CI]\u200a=\u200a0.049 [0.011\u20130.207], p<0.001, Fisher's exact test.\nIn the PP group, 19 out of 70 (27%) patients who completed the 7-day trial had acute clinical failure in the cefixime recipients as compared to 1 out of 88 (1%) in the gatifloxacin recipients (Odds Ratio [95%CI]\u200a=\u200a0.031 [0.004 \u2013 0.237], p<0.001).\nConsidering all patients to be failures who dropped out of the study before completion of the seven day treatment course, 26 out of 77 (34%) failed in the cefixime group as compared to 5 out of 92 (5%) in the gatifloxacin group (OR[95%CI]\u200a=\u200a0.112 [0.041 \u2013 0.312], p<0.001).\n138 patients were evaluable for relapse; 20 had acute treatment failure and 11 withdrew from the study before day 7.\nIn total, eight relapses (Figure 1) were observed.\nRelapse rates were 12.4% (6/51) in the cefixime group and 3.4% (2/87) in gatifloxacin group (HR[95%CI]\u200a=\u200a0.185 [0.037\u20130.915], p\u200a=\u200a 0.0199).\nThe Kaplan-Meier plots for the time of relapse are shown in Figure 4.\nOverall failures (acute treatment failure plus relapse plus death) were 29 in number (Figure 1).\nOverall failure rate at 1 month was estimated as 37.6% (95% CI\u200a=\u200a27.14% \u2013 50.2%) in the cefixime group and 3.5% ( 95% CI \u200a=\u200a2.2%\u201311.5%) in the Gatifloxacin group (HR[95%CI]\u200a=\u200a0.084[ 0.025\u20130.280], p<0.0001) ( Figure 5).\nFrom patients with negative cultures, 11 had acute clinical failures, 10 (out of 97, 10%) in Cefixime group and 1 (out of 103, 1%) in the Gatifloxacin group, OR[95%CI]\u200a=\u200a0.086 [0.011\u20130.686], p\u200a=\u200a0.004, Fisher's exact test.\nSimilarly, treating drop-out as treatment failures, we had 50 out of 187 (27%) in the Cefixime group and 15 out of 203 (7%) in the Gatifloxacin group acute treatment failures, OR[95%CI]\u200a=\u200a0.219 [0.118\u20130.405], p<0.001, Fisher's exact test.\nAncillary analyses\nAmong all culture positive patients in the cefixime group, one patient (1/70, 1%) had S. Paratyphi A cultured from her blood on day 10,but there were no (0/88, 0%) positive blood culture growths in the gatifloxacin group on day 10.\nNo patient was found to be a persistent carrier of S. Typhi or Paratyphi A in their stool.\nA positive stool culture for S. Typhi was seen for one patient on day 10 and for another on day 30.\nSubsequent cultures were negative for both patients.\nWe were able to obtain stool cultures from 147 (87%), 141 (83%), and 130 (77%) pretreatment blood culture positive patients at one, three, and six months respectively.\nMicrobiology\nAntibiotic sensitivity testing revealed that all strains were sensitive to gatifloxacin, cefixime, ceftriaxone or gemifloxacin.\nOne strain was resistant to chloramphenicol, and 136 (83%) of the pretreatment isolates were nalidixic acid resistant strains (NARST).\nMinimum inhibitory concentration (MIC) was determined for 161 of the pretreatment blood culture isolates.\nThe median (range) MICs for each antibiotic were as follows: gatifloxacin 0.125 (0.006\u20130.5) \u00b5g/mL, cefixime 0.380 (0.016\u20132.0) \u00b5g/mL, nalidixic acid >256 (1.5->256) \u00b5g/mL, chloramphenicol 8.0 (1.5->256) \u00b5g/mL, ceftriaxone 0.125 (0.047\u20130.5) \u00b5g/mL and gemifloxacin 0.125 (0.004\u20130.5) \u00b5g/mL.\nAdverse events\nAmong all patients who received cefixime, there was one death, which might have been due to the development of disease-related complications during treatment.\nThis patient was enrolled on the fourteenth day of his illness.\nOn day 6 of treatment the patient complained of reddish stool and petechiae and was immediately admitted to hospital where he developed severe thrombocytopenia and gastrointestinal bleeding.\nHe developed acute respiratory distress syndrome and was mechanically ventilated.\nHe developed disseminated intravascular coagulation and succumbed to his illness on day 21 of entry into the trial.\nHis pretreatment blood culture grew S. Paratyphi A which was sensitive to cefixime with an MIC of 0.38 \u00b5g/mL.\nOne patient developed erythematous skin rash which needed two doses of oral antihistamine.\nAmong all patients who received gatifloxacin there were 2 patients with excessive vomiting, which required intravenous antiemetics and fluids and observation in the hospital emergency room for upto 6 hours.\nThere were an additional 23 patients who complained of excessive nausea and occasional vomiting after ingestion of the drug.\nOf these, two needed oral antiemetics; in the remaining 21 patients no intervention was required.\nDiscussion\nInterpretation\nIn this study examining fever clearance time, acute treatment failure and relapse as indicators of treatment efficacy, that the results raise doubts on the usefulness of cefixime and suggest that gatifloxacin is a potent choice for the treatment of uncomplicated enteric fever.\nFebrile illness is one of the most common reasons for presentation to hospitals in many developing countries.\nIn patients with fever, a very common clinical diagnosis is enteric fever, and S. enterica serotype Typhi or Paratyphi A are the two most commonly isolated pathogens from the blood in febrile patients in our hospital.\nBefore the advent of multi-drug-resistant (MDR) S. Typhi, chloramphenicol, ampicillin or cotrimoxazole were successfully used as the first line drug in the treatment of enteric fever.\nAfter the emergence of MDR strains, fluoroquinolones and third-generation cephalosporins have been suggested and used as alternative antimicrobials.\nHowever the emergence and spread of point mutations in the gyrA gene of the bacterial genome has conferred resistance to nalidixic acid and reduced susceptibility to the commonly used fluoroquinolones such as ofloxacin, leading to a poorer clinical response.\nA recent study in Viet Nam (CM Parry, unpublished) showed ofloxacin at the dose of 20 mg/kg/day was able to achieve a cure rate in only 64% of patients.\nIn our context of high nalidixic acid resistance, gatifloxacin is the most effective and appropriate choice for treatment of enteric fever.\nGatifloxacin (Sandoz, India) is relatively inexpensive (US$1.2 for a 7 day treatment course) and needs to be administered just once a day; both of these features are attractive in this setting.\nGatifloxacin has a different binding motif than some other fluroquinolones, and this characteristic enables it to retain activity against Salmonella enterica serovar Typhi or Salmonella enterica serovar Paratyphi A even in the presence of marked reduction in sensitivity to the older fluoroquinolones.\nCefixime, a third generation cephalosporin, is widely trusted to be effective for enteric fever as first line treatment, and is also used as second line therapy when initial treatment with a fluoroquinolone in a patient suspected to be enteric fever fails.\nThe fact that we saw a high overall failure rate associated with cefixime despite all of the strains being fully sensitive in vitro to the drug shows that the mechanism of action of cefixime may not be suited to the eradication of S. Typhi or Paratyphi A from the body or blood, and the poor intracellular penetration into macrophages and reticulo endothelial tissues where the typhoid organisms colonize may be the cause of high failure rates.\nThis study was unique in that we used CMAs to simulate a hospital setup in the community.\nCMA's directly observed patients taking the therapies, monitored fever and identified complications early; these characteristics have not been used in the past for typhoid trials although enteric fever in endemic areas is treated on an outpatient basis.\nA major advantage of follow-up using CMAs was that the health workers knew the exact house location of the patients, and therefore follow up even after the successful completion of the initial seven-day drug trial was possible.\nIn developing countries follow up of patients can be very difficult because of a lack of a proper address and relative unavailability of other means of communication, for example,a telephone.\nIn addition to its relevance to culture confirmed enteric fever, another major strength of this large randomized study is that gatifloxacin proved to be more efficacious than cefixime with respect to fever clearance time and failure rates, even in the subgroup of patients who were clinically presumed to have enteric fever but who had a negative blood culture.\nAntibiotic treatment for typhoid in highly endemic areas is usually started based on the presence of a \u201csyndromic\u201d illness (acute fever for a few days and constitutional symptoms with no known source of infection) before culture results are known.\nEnteric fever, which continues to be a neglected disease, is an important cause of morbidity and mortality, and facilities for blood culture or other reliable methods of diagnosis rarely exist in this setting.\nGeneralizability\nDespite widespread resistance to Nalidixic acid in Kathmandu, and rising MICs to the older fluoroquinolones, ciprofloxacin and ofloxacin, gatifloxacin has proven to be a potent drug for the treatment of enteric fever.\nOur study has relevance to South Asia, as resistance to nalidixic acid is widely prevalent there.\nInevitably there will be emergence of resistance to gatifloxacin in areas with both MDR and NARST; and in this situation alternative antibiotics like azithromycin may need to be used.\nOf interest, in keeping with anecdotal reports from elsewhere in South Asia, only one strain was resistant to chloramphenicol in the present study.\nIn areas of the world where chloramphenicol susceptibility has reemerged there may be an argument for reassessing chloramphenicol.\nIn the present study Gatifloxacin was associated with nausea in 12% of patients and it may be important to forewarn patients of this possible side effect.\nThere have been sporadic reports of dysglycemia caused by gatifloxacin, and a recent population-based, case controlled study examining gatifloxacin usage amongst elderly individuals in Canada (mean age 77 years) who developed dysglycemia also raises possible concerns.\nWe did not do any blood sugar testing to look for dysglycemia.\nHowever in a study involving a younger age group where blood sugar testing was done the results revealed no dysglycemia: 887 children were treated with gatifloxacin (10 mg/kg) for otitis media and were followed for a year with no signs of alteration of glucose homeostasis either acutely or otherwise.\nClearly, it would be prudent to treat diabetics and elderly people suffering from enteric fever with an alternative antibiotic such as azithromycin and avoid the potential problems in this specific population with gatifloxacin.\nLimitations of the study\nThe DSMC advised the Principal Investigators in this study to stop recruitment to the trial based on a priori defined difference (p<0.01) between the two treatment arms in the primary endpoints of the study.\nIt is possible that if the trial had been continued with a larger sample size, other important information could have been gathered.\nIn addition if patients and/or investigators had been blinded to treatment assignments, the study would have been further strengthened; however as in most typhoid trials, it was not possible to do this due to the difference in dosing schedule for the two drugs being compared.\nAnother limitation of this study was that temperature was only measured every 12 hours.\nHowever, to address this limitation, and to avoid missing increases in temperature, we checked temperatures for 10 days after enrollment, or for 48 hours after resolution of fever, whichever came later, in all patients.\nFinally, a telephone or internet based system of randomization would be ideal, but such a system does not exist here.\nOverall evidence\nWe have compared the outcomes from our trial with those of other comparable studies, identified from a recent Cochrane review, WHO typhoid guidelines, and a search of Medline using these terms: cefixime, typhoid trials.\nThe findings in our study are consistent with those of a 1995 study done in Viet Nam which showed that cefixime (20 mg/kg/day) for 7 days was inferior to ofloxacin (10 mg/kg/day) for 5 days in the treatment of MDR typhoid fever in children.\nHowever other studies have suggested cefixime can be successful in the treatment of enteric fever.\nOverall these studies, both descriptive and randomized, have examining the use of cefixime in confirmed enteric fever (total of 292 patients) and with treatment durations of mostly 14 days, have found failure rates ranging from 4% to 23%.\nBesides the general undesirability of a longer course with cefixime with increased morbidity and possibly complications, this drug is also more expensive (a 7-day course costs US $7 (Blue Cross Laboratories, India)).\nThe present study is the largest randomized controlled trial ever conducted with cefixime in enteric fever and clearly shows, even in a setting with fully sensitive strains, that cefixime is a poor drug for this disease.\nThese findings are contrary to the recommendation by many sources including the World Health Organization that cefixime can be used as first or second line therapy in the treatment of enteric fever.\nBased on the present study, we believe gatifloxacin to be an optimal choice in the treatment of uncomplicated enteric fever.\nProfile of the Trial.The consort flow diagram showing the flow of participants through the trial.\nProportion of all patients still febrile.Kaplan-Meier survival curve showing the proportion of all patients(ITT) still febrile through time.\nProportion of culture positive patients still febrile.Kaplan-Meier survival curve showing the proportion of culture positive(PP) patients still febrile through time.\nProportion of relapse free patients.Kaplan-Meier survival curve showing the proportion of relapse free patients in the culture positive population.\nProportion of overall failure free patients.Kaplan-Meier survival curve showing the proportion of overall failure free patients in the culture positive population.\n\nBaseline characteristics all patients.\nPATIENT CHARACTERISTICS | Culture negative (213) | Culture positive (169)\nNo of males/No of females | 136/77 | 111/58\nAge (yrs) | 18 (2\u201364) | 17 (2.75\u201350)\nNumber Aged <14 years (%) | 79(37.1) | 60 (35.5)\nWeight (Kg) | 44 (10\u201380) | 46 (10\u201373)\nDuration of fever before treatment (days) | 5 (0\u201321) | 5 (2\u201323)\nMedian oral temperature at presentation(95% CI, range) (in degrees C) | 38.7 (38.6\u201339; 36.5\u201340.7) | 39(38.8\u201339.2; 36.8\u201341)\nHeadache, Number with (%) (median duration [days]) | 204 (95.7) (4) | 164 (97.0) (5.)\nAnorexia, Number with (%) (median duration [days]) | 160 (75.1)  | 129 (76.3) \nAbdominal Pain, Number with (%) (median duration [days]) | 88 (41.3)  | 80 (47.3) \nCough, Number with (%) (median duration [days]) | 83 (39.0)  | 59 (34.9) \nDiarrhoea, Number with (%) (median duration [days]) | 45 (21.1)  | 41 (24.3) \nVomiting, Number with (%) (median duration [days]) | 30 (14.1) | 27 (16.0) \nAbdominal tenderness ( n [%]) ) | 32 [15.1] | 23 [13.6]\nSplenomegaly ( n [%]) | 18 [8.5] | 18 [10.6]\nHepatomegaly ( n [%]) | 12 [5.76] | 9[5.3]\nHematocrit (in%) | 40 (27\u201353) | 40 (29\u201350)\nWhite Cell Count (in \u00d71000 per microlitre) | 7.2 (2.3\u201324.2) | 6.7 (3.0\u201320.0)\nPlatelet Count (in \u00d71000 per microlitre) | 192 (66\u2013546) | 180 (65\u2013380)\n* ALT ( in U/L ) | 30(11\u2013240) | 37 (12\u2013200)\n**AST ( in U/L ) | 43 (20\u2013354) | 52 (21\u2013169)\nTotal Bilirubin ( in mg/dL ) | 0.8 (0.17\u20133.6) | 0.89 (0.18\u20133.2)\n\nBaseline epidemiological, clinical and laboratory features at presentation of all intention to treat patients showing a comparison between culture positive and culture negative groups.\nALT (serum alanine aminotransferase) normal range 5\u201334 U/L\nAST (serum aspartate aminotransferase) normal range 5\u201334 U/L\nAll data presented as median (range) unless specified.\n\nBaseline characteristics at presentation of culture positive patients.\nPATIENT CHARACTERISTICS | GATIFLOXACIN (n\u200a=\u200a92) | CEFIXIME (n\u200a=\u200a77)\nNo of males/No of females | 67/25 | 44/33\nAge (yrs) | 18 (2.75\u201345) | 15 (3\u201350)\nNumber Aged <14 years (%) | 27 (29%) | 33 (43%)\nWeight (Kg) | 49 (10\u201370) | 42 (11\u201373)\nDuration of fever before treatment (days) | 5.2 | 5.4\nMedian oral temperature at presentation(95% CI, range) (in degrees C) | 39 (38.9\u201339.2; 37.5\u201341.0) | 39 (38.8\u201339.2; 36.8\u201340.5)\nHeadache, Number with (%) (median duration [days]) | 88 (95.7%) (5) | 76 (98.7%) (4.5)\nAnorexia, Number with (%) (median duration [days]) | 73 (79.3%) (4) | 56 (73%) (4)\nAbdominal Pain, Number with (%) (median duration [days]) | 43 (46.7%) (4) | 40 (52%) (4)\nCough, Number with (%) (median duration [days]) | 37 (40.2%) (3) | 22 (29%) (3)\nDiarrhoea, Number with (%) (median duration [days]) | 21 (22.8%) (3) | 20 (26%) (3)\nVomiting, Number with (%) (median duration [days]) | 17 (18.5%) (2) | 10 (13%) (1.5)\nAbdominal tenderness ( n [%]) ) | 14 (15.2%) | 8 (10.4%)\nSplenomegaly ( n [%]) | 10 (10.9%) | 8 (10.4%)\nHepatomegaly ( n [%]) | 5 (5.4%) | 4 (5%)\nHematocrit (in%) | 41 (30\u201350) | 40 (29\u201350)\nWhite Cell Count (in \u00d71000 per microlitre) | 6.8(3.0\u201318) | 6.7 (3.1\u201320)\nPlatelet Count (in \u00d71000 per microlitre) | 180(65\u2013367) | 186 (120\u2013380)\n* ALT ( in U/L ) | 36 (12\u2013180) | 39(18\u2013200)\n**AST ( in U/L ) | 53 (24\u2013155) | 49 (21\u2013169)\nTotal Bilirubin ( in mg/dL ) | 0.85 (0.18\u20133.2) | 0.9 (0.35\u20132.3)\nPositive pretreatment fecal cultures ( n [%]) | 9 (9.8%) | 3 (3.8%)\n\nBaseline epidemiological, clinical and laboratory features at presentation of all blood culture positive patients showing a comparison between the gatifloxacin and cefixime arms.\nALT (serum alanine aminotransferase) normal range 5\u201334 U/L\nAST (serum aspartate aminotransferase) normal range 5\u201334 U/L\nAll data presented as median ( range) unless specified.",
        "label": "low",
        "id": "task4_RLD_test_198"
    },
    {
        "paper_doi": "10.1136/bmj.n893",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: randomised control trialStudy grouping: parallel designEthics and informed consent: informed consent obtained; approved by the ethics committees of the Royal Brisbane and Women's Hospital and Griffith University\nFollow-up period: 30 daysSample size estimate: calculated the sample size based on the proportion of women who developed a SSI within 30 days of CS. Conservatively estimated that 15% of women in the control group were likely to develop an SSI; determined that an absolute reduction in rate of SSI of 5 percentage points would be clinically important. The sample size required to detect a reduction in the cumulative incidence of SSI at 30 days from 15% to 10% was 950 per group (90% power and 5% significance level; inflated the sample size by 10% to allow for loss to follow up (n = 1045/group; total sample size 2090).ITT analysis: yes; number randomised: 2035, number analysed: 2035Funding: Australian National Health and Medical Research CouncilPreregistration: ANZCTR identifier 12615000286549\n\n\nParticipants: Location: Australia (4 tertiary hospitals)Intervention group: 1017,control group: 1018Mean age: 31 (5.5) vs 31 (5.4)Inclusion criteria: women booked for elective (category 4) or for semi-urgent (categories 2-3) caesarean section; who recorded a pre-pregnancy BMI of >= 30 kg/m2 and were able to give informed consent.Exclusion criteria: women who needed an urgent CS (category 1), had an infection in hospital including during labour or immediately prior to CS, had participated in the trial in a previous pregnancy, or were unable to speak or understand English with no interpreter present.\n\n\nInterventions: Aims: to determine the effectiveness of closed incision negative pressure wound therapy (NPWT) compared with standard dressings in preventing surgical site infection (SSI) in obese women undergoing caesarean section.Group 1 (NPWT) intervention: women assigned to the NPWT group received a PICO(tm) dressing (Smith & Nephew, Hull, UK), which was left intact for approximately 5-7 days as recommended by the manufacturer.Group 2 (control) intervention: women received the standard hospital dressing. The choice of standard dressings was based on the treating obstetrician's usual choice of dressing, (e.g. hydrocolloid or transparent) applied according to the manufacturer's recommendations. The standard dressing was left intact for 5-7 days.\n\n\nOutcomes: PrimarySSI (including superficial, deep or organ/body space)DehiscenceMortality (included in the serious adverse events outcome)SecondaryHaematomaSeromaBlisteringReadmissionReoperationPainValidity of measure/s: CDC definition for SSITime points: 30 days\n\n\nNotes: Full cost-effectiveness analysis planned to be reported separately (protocol)\n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "Closed incision negative pressure wound therapy versus standard dressings in obese women undergoing caesarean section: multicentre parallel group randomised controlled trial\nClosed incision negative pressure wound therapy versus standard dressings in obese women undergoing caesarean section: multicentre parallel group randomised controlled trial\nBMJ\nBMJ\nObjective\nTo determine the effectiveness of closed incision negative pressure wound therapy (NPWT) compared with standard dressings in preventing surgical site infection (SSI) in obese women undergoing caesarean section.\nDesign\nMulticentre, pragmatic, randomised, controlled, parallel group, superiority trial.\nsetting Four Australian tertiary hospitals between October 2015 and November 2019.\nParticiPants\nEligible women had a pre-pregnancy body mass index of 30 or greater and gave birth by elective or semiurgent caesarean section.\ninterventiOn 2035 consenting women were randomised before the caesarean procedure to closed incision NPWT (n=1017) or standard dressing (n=1018).\nAllocation was concealed until skin closure.\nMain OutcOMe Measures\nThe primary outcome was cumulative incidence of SSI.\nSecondary outcomes included depth of SSI (superficial, deep, or organ/body space), rates of wound complications (dehiscence, haematoma, seroma, bleeding, bruising), length of stay in hospital, and rates of dressing related adverse events.\nWomen and clinicians were not masked, but the outcome assessors and statistician were blinded to treatment allocation.\nThe pre-specified primary intention to treat analysis was based on a conservative assumption of no SSI for a minority of women (n=28) with missing outcome data.\nPost hoc sensitivity analyses included best case analysis and complete case analysis.\nresults\nIn the primary intention to treat analysis, SSI occurred in 75 (7.4%) women treated with closed incision NPWT and in 99 (9.7%) women with a standard dressing (risk ratio 0.76, 95% confidence interval 0.57 to 1.01; P=0.06).\nPost hoc sensitivity analyses to explore the effect of missing data found the same direction of effect (closed incision NPWT reducing SSI), with statistical significance.\nBlistering occurred in 40/996 (4.0%) women who received closed incision NPWT and in 23/983 (2.3%) who received the standard dressing (risk ratio 1.72, 1.04 to 2.85; P=0.03).\ncOnclusiOn Prophylactic closed incision NPWT for obese women after caesarean section resulted in a 24% reduction in the risk of SSI (3% reduction in absolute risk) compared with standard dressings.\nThis difference was close to statistical significance, but it likely underestimates the effectiveness of closed incision NPWT in this population.\nThe results of the conservative primary analysis, multivariable adjusted model, and post hoc sensitivity analysis need to be considered alongside the growing body of evidence of the benefit of closed incision NPWT and given the number of obese women undergoing caesarean section globally.\nThe decision to use closed incision NPWT must also be weighed against the increases in skin blistering and economic considerations and should be based on shared decision making with patients.\nIntroduction\nThe use of caesarean section in birthing women varies widely, with Nordic countries reporting low rates and other Western countries such as Australia, Canada, the UK, and the US reporting higher rates (15-17% v 25-32%).\n1 Compared with vaginal birth, caesarean section is associated with increased morbidity and mortality.\n2 The World Health Organization defines people as obese doi: 10.1136/bmj.n893 | BMJ\n2021;373:n893 | the bmj if their body mass index is greater than or equal to 30.0.\n1 Obesity in pregnancy is increasingly common; in Australia, more than 50% of women are overweight or obese on entering pregnancy.\n1 Postoperative wound complications such as surgical site infection (SSI), dehiscence (splitting open of a surgically closed wound), and formation of haematoma and seroma are common complications of surgical procedures, 3 particularly among women with obesity, diabetes, or both.\n4 SSI is an important global concern that can contribute to re-intervention and treatment, increased length of stay in hospital, delayed wound healing, and, in some cases, death.\n5 6 Maternal obesity increases the woman's risk of developing SSI and other wound complications threefold, which delays recovery, increases discomfort, and reduces quality of life.\n4 7 Over the past decade, the use of single use closed incision negative pressure wound therapy (NPWT) dressings in high risk surgical incisions has been increasing, with the aim of reducing the risk of SSI and other associated wound complications.\n8 Closed incision NPWT is a sealed non-invasive system that applies suction (negative pressure) on the wound site that has been closed, for example, by sutures, staples, or glue.\nThe surgical incision is covered with semiocclusive adhesive dressing connected by tubing to a suction pump. 9\nThe suction pump exerts negative pressure to the closed incision and removes wound fluid with recommended pressures usually between -50 mm Hg and -125 mm Hg, 10 depending on the manufacturer's instructions. 11\nThe mechanism of action is unclear but is purported to include reduced bacterial entry into the wound while removing blood and exudate and stimulating granulation.\nIn 2010-11, two simplified NPWT devices became commercially available (Prevena (KCI) and PICO (Smith & Nephew)).\nA Cochrane review published before we started this trial and its subsequent update found only low quality evidence in any population, with most studies sponsored by industry.\n8 12 Meta-analytic results of the updated Cochrane review reported inconclusive evidence of the effectiveness of closed incision NPWT specifically for obese women undergoing caesarean section (seven studies). 8\nAt the time we began our research, all other trials in this population were small, single site, and industry funded.\nIn this study, we aimed to compare the effectiveness and safety of prophylactic closed incision NPWT and standard surgical dressings on the cumulative incidence of SSI in obese women undergoing elective and semi-urgent caesarean section.\nMethods study design and participants\nWe conducted a pragmatic, randomised, controlled, parallel group, superiority trial in four large public hospitals in southeast Queensland, Australia.\nWe made no changes to the methods after the start of the trial.\nWe identified potentially eligible women at their routine 36 week antenatal visit.\nResearch nurses at each site screened women in antenatal clinics, antenatal wards, and birthing suites.\nWomen were eligible if they were booked for elective (category 4) or semi-urgent (categories 2-3) caesarean section, 13 recorded a prepregnancy body mass index of 30 or higher, and were able to provide written informed consent.\nWe excluded women who needed an urgent caesarean section (category 1), had an infection in hospital including during labour or immediately before caesarean section, had participated in the trial in a previous pregnancy, or were unable to speak or understand English with no interpreter present.\nWritten informed consent was obtained from all participants.\nThe protocol has been published. 14\nrandomisation and masking\nWe used a web based central randomisation service to randomly assigned eligible, consenting women (1:1) just before the caesarean procedure to receive either a closed incision NPWT dressing or the standard hospital dressing.\nTo ensure that equal numbers of participants were assigned to each group, we used random block sizes of four, six, and eight, stratified by hospital.\nAllocation was concealed until after skin closure.\nThe nature of the intervention meant that women, clinical staff, and research staff were not blinded to treatment after allocation.\nData were reviewed by two independent, blinded outcome assessors to determine primary and secondary wound endpoints, and discrepancies were adjudicated by a third blinded assessor.\nPrincipal investigators, including the trial statistician, were also blinded to group allocation.\nThe clinical trial coordinator trained and supervised research nurses and audited the quality of data and compliance of randomisation.\nProcedures\nAll women received standard care, according to local hospital and national health department guidelines.\n15 Before the skin incision, the woman's abdomen was prepared with either alcoholic or aqueous chlorhexidine or betadine.\nAll women received a lower transverse suprapubic skin incision, and two obstetricians, usually a trainee registrar supervised by a consultant, carried out the operation.\nThe method of skin closure (suture or staples) was based on the obstetrician's preference.\nThe operating obstetrician (or delegate) applied the closed incision NPWT and standard dressings under sterile conditions in the operating room immediately after skin closure.\nWomen assigned to the closed incision NPWT group received a PICO dressing (Smith & Nephew, Hull, UK), which was left intact for approximately five to seven days as recommended by the manufacturer.\nThis particular NPWT product was used in two earlier pilot studies.\n16 17 The PICO product (size 10\u00d730 cm or 10\u00d740 cm) has a small discrete pump powered by two AA lithium batteries with an absorbent polyurethane foam dressing that holds wound exudate away from the skin.\nA tube is inserted into the foam, and a continuous negative pressure of 80 mm Hg is applied after application of the dressing.\non 7 May 2024 at University of California Med Ctr Medical Center Library.\nProtected by http://www.bmj.com/\nBMJ: first published as 10.1136/bmj.n893 on 5 May 2021.\nDownloaded from each of the four edges with four pieces of adhesive tape included in the dressing kit, as per the manufacturer's instructions.\nAll clinical staff providing care received ongoing training and support in the correct application and use of the PICO dressings, as well as monitoring dressing changes and completing documentation daily for assessment of protocol fidelity.\nThe control group comprised women allocated to the standard hospital dressing.\nThe choice of standard dressings was based on the treating obstetrician's usual choice of dressing (for example, hydrocolloid or transparent), applied according to the manufacturer's recommendations after skin closure in the operating room.\nAcross all hospital sites, the standard dressing was left intact for five to seven days.\nWe collected clinical data from several sources, including electronic records, direct observation, and self-reporting by women during hospital admission and after discharge.\nDemographic data (pre-pregnancy body mass index, parity/gravidity, comorbidities, measurement of health status (Health Related Quality of Life Short Form Survey SF-12 v-2) were obtained on enrolment; surgical data (American Society of Anaesthesiologists category, type of anaesthetic, antibiotic administration, hair removal method, surgical approach, wound closure layers, suture materials, length of operation) were obtained on the day of the caesarean section.\nResearch nurses visited women on postoperative day 2 and collected vital signs, SSI related data using a structured tool based on the Centres of Healthcare Related Infection Surveillance and Prevention guidelines identifying signs and symptoms of SSI (that is, redness, swelling, pain/tenderness, watery or purulent discharge), 18 pain associated with the dressing, and women's satisfaction with the NPWT dressing.\nAfter discharge from hospital, research nurses conducted telephone interviews with all women weekly (from the day of their surgery) until 28 days after discharge.\nThey asked women a series of questions about SSI symptoms, SF-12 v2, and related resource use including health professional visits.\nOn day 30, research nurses audited all participants' hospital electronic health records to check for documented evidence of SSI and wound complications (chart data documented wound complications, reoperations and hospital readmission due to wound complications, use of antibiotics for wound complications, type of SSI, signs and symptoms of SSI).\nOutcome assessors were blinded to group allocation, the intervention and its comparator, and study hypotheses.\nThese assessors were experienced registered nurses and performed outcome assessment of primary (SSI) and secondary wound related outcomes (SSI type, wound complications) for all women enrolled in the study.\nEach outcome assessor independently ascertained wound outcomes, and regular inter-rater consistency checks were undertaken throughout the trial.\nWhere discrepancies in assessment of signs and symptoms existed, a third outcome assessor (nurse practitioner in wound care) adjudicated decisions.\nWe defined loss to follow-up as lacking both 30 day medical record data and follow-up phone interview data over the four weekly time points on the primary outcome (SSI).\nThus, a woman might be missing up to three interviews but would not be considered lost to follow-up unless her 30 day chart was also missing.\nEach week, nurses attempted to contact women or their contact person up to three times.\nTherefore, for all women who were not lost to follow-up and did not withdraw their participation after randomisation, we had data on primary outcome, SSI type (where SSI occurred), and wound complications.\nAll data were entered directly into secure portable tablets using a purpose built research data capture (REDCap) database and form based interface.\nResearch nurses had access to the data at their hospital site only, and clinical staff did not have access to research data.\nThe clinical trial coordinator audited the quality and completeness of data and adherence to the protocol, as well as visiting sites for training and monitoring.\nOutcomes\nThe primary outcome was the cumulative incidence of SSI at 30 days after surgery, as defined by Centers for Disease Control and Prevention (CDC) guidelines.\nSecondary clinical outcomes included type of SSI (superficial, deep, or organ/body space), 18 any type of wound complication (dehiscence, haematoma, seroma, bleeding), type/number of individual wound complications, length of stay in hospital, and number of wound related hospital readmissions in the 30 days after surgery.\nDefinitions and measures used for primary and secondary outcomes are included in supplementary table A.\nOther secondary outcomes, including dressing related adverse events, such as rash, itchiness, and blistering, were assessed by research nurses.\nSerious adverse events (maternal death, admission to intensive care unit, life threatening condition) were monitored and reported to the human research ethics committee at each site.\nAn independent Data Safety Monitoring Committee was established to assess the safety of the intervention.\nThis committee, comprising an obstetrician, a statistician, and an infection control nurse specialist, oversaw the trial and reviewed interim analyses, undertaken twice during the life of the trial.\nThe trial would not be stopped unless the committee deemed that significant safety problems were present during safety monitoring of the trial intervention.\nstatistical analysis\nWe calculated the sample size on the basis of the proportion of women who developed an SSI within 30 days of caesarean section.\nOn the basis of previous work in this area, 19 we conservatively estimated that 15% of women in the control group were likely to develop an SSI.\nFollowing discussions with infectious disease experts and obstetricians, we determined that an absolute reduction in the rate of SSI of 5 percentage points would be clinically important.\nPower Analysis & Sample Size system (PASS, V.12), NCSS).\nWe inflated the sample size by 10% to allow for loss to follow-up (n=1045 per group; total sample size 2090 women).\nWe summarised baseline characteristics comprising binary data by using counts and proportions and continuous data as mean and standard deviation or median and interquartile range, depending on the distribution.\nWe used Cohen's \u03ba to calculate inter-rater consistency between outcome assessors.\nThe pre-specified primary outcome analysis was by intention to treat.\nFor women lost to follow-up and withdrawn from the study post-randomisation who were missing the primary outcome, we conservatively (favoured standard treatment, as it had higher levels of missing data) assumed that they did not develop an SSI (worst case analysis).\nAs per the protocol, we explored differences in prognostic variables between groups.\nThe prognostic factors assessed were identified in the literature 20 21 and based on expert opinion (body mass index, age, diabetes, smoking, rupture of membranes, parity, caesarean section elective/semiurgent, and length of procedure).\nWe found differences between groups relative to body mass index and group allocation.\nThus, following the protocol, we analysed the primary outcome by using a logistic regression model, adjusting for these.\nWe used a planned per protocol analysis of treatment for device related and serious adverse events.\nWe compared binary outcomes (that is, SSI, wound complications, adverse/serious adverse events) by using a \u03c7 2 test or Fisher's exact test and risk ratios with 95% confidence intervals.\nWe reported continuous variables with non-normal distribution (that is, length of surgery, length of stay in hospital) by using medians and interquartile ranges and compared them by using a Mann-Whitney U test.\nFor all inferential tests, we considered a P value below 0.05 to be statistically significant.\nPost hoc analyses\nTo check for the robustness of conclusions to the effect of assumptions around missing primary outcome data, we repeated the intention to treat analysis as described above assuming that all women missing the primary outcome did have an SSI (favouring closed incision NPWT; that is, best case analysis) and excluding women missing the primary outcome (complete case analysis).\nAdditionally, we did a per protocol analysis excluding women lost to follow-up, women withdrawn after randomisation, and women treated against their randomised allocation (for example, treated with closed incision NPWT when in the standard dressing arm).\nSecondary outcomes (type of SSI, wound complications, length of stay in hospital, readmissions, pain, reoperations) were analysed by complete case analysis (excluding women without primary outcome) and by per protocol analysis (as with the primary outcome: excluding women lost to follow-up, women withdrawn post-randomisation, and women treated against their randomised allocation).\nPatient and public involvement\nPatients were not involved in defining the research question or outcome measures or in the interpretation or writing up of results of this study.\nThis study was conceived in 2013, when the patients as co-researchers movement had not widely been adopted in Australia.\nresults\nBetween 26 October 2015 and 1 November 2019, 8558 of the 12 077 women screened were excluded, leaving 3519 women who were eligible.\nHowever, 338 (9.6%) could not be recruited as their caesarean section occurred after hours, and 1072 women were not enrolled for various reasons, including refusals, vaginal delivery, or delivery at another facility; 2109 (60%) women were enrolled.\nWe randomly assigned 2035 women to receive NPWT (n=1017) or standard surgical wound dressings (n=1018) (fig 1).\nFollowup concluded on 1 December 2019.\nIntention to treat analysis of primary and secondary outcomes (SSI, type of SSI, wound complications, length of stay in hospital, readmissions, pain, reoperations) included the 2035 women randomly assigned to the intervention and control groups.\nBaseline demographic and obstetric characteristics were similar between groups (table 1).\nThe average age of participants was 31 (SD 5.5; range 16-54) years.\nHalf of all women (1012; 50%) had a pre-pregnancy body mass index of 35 or higher (range 30-72), and most (1472; 72%) had an elective caesarean section.\nOne third of women (657; 32%) across the sample had either gestational diabetes or diabetes mellitus.\nAt the time of caesarean section, most women (1729; 85%) had intact membranes.\nMost women (1942; 95%) had subcutaneous layer closure in addition to subcuticular (skin) closure; staples were rarely used (27; 1%).\nInterrater reliability between outcome assessors for the primary outcome SSI and the secondary outcome type of SSI yielded \u03ba=0.764 (95% confidence interval 0.72 to 0.81), and \u03ba=0.712 (0.66 to 0.76), respectively.\nIn the primary analysis, our \"worst case\" intention to treat analysis assumed that women whose primary outcome was missing did not develop SSI (table 2).\nThe SSI rate across the entire sample was 8.6% (n=174).\nWe observed a 3 percentage point reduction in the absolute risk of SSI in women treated with NPWT compared with standard dressings; this difference was not statistically significant (7.4% v 9.7%; risk ratio 0.76, 95% confidence interval 0.57 to 1.01; P=0.06) (table 2).\nIn terms of SSI type, only 1 (<1%) woman in the NPWT group developed an organ/ space SSI (table 2).\nThe rates of all types of wound complications in the intervention and control groups were comparable.\nWound dehiscence was the most common complication in both groups ( No differences between the groups in the distribution of prognostic factors were apparent (table 1).\nMultivariable logistic regression analysis, including all a priori identified prognostic factors, showed that body mass index 40-49.9 (P=0.02) was a statistically significant model covariate (model likelihood \u03c7 2 =26.16, df 11; P<0.05; Nagelkerke R 2 =2.9).\nThe full results are shown in supplementary table B.\nWe did a planned per protocol analysis for dressing related adverse events and serious adverse events (table 3).\nDressing related adverse events reported included skin blistering, itchiness, and rash.\nWe observed a 2 percentage point increase in the absolute risk of skin blistering among women in the closed incision NPWT group, which was statistically significant (4.0% (40) v 2.3% (23); risk reduction 1.72, 1.04 to 2.85; P=0.03).\nOverall, 17 serious adverse events occurred, including three neonatal deaths.\nRates of serious adverse events were low and did not differ between intervention and control groups (intensive care unit admission, life threatening condition: 1.2% v 0.5%; risk ratio 2.57, 0.92 to 7.17; P=0.06).\nMost of the admissions to intensive care related to the lack of available high dependency unit beds.\nOne woman developed a pulmonary embolism.\nAll serious adverse events were reported to the ethics board, and none was deemed related to the intervention.\nPost hoc sensitivity analyses\nPost hoc sensitivity analyses of the cumulative incidence of all types of SSI favoured closed incision NPWT therapy compared with our main crude analysis (reported above).\nThe \"best case\" intention to treat analysis assumed that women with missing outcome data developed SSI (supplementary table C).\nThe SSI incidence across the entire sample was 9.9% (n=202).\nWe observed a 4 percentage point reduction in the absolute risk of SSI in women treated with closed\nWe also did a per protocol analysis of primary and secondary outcomes based on 1979 women (supplementary table E).\nIn this analysis, we excluded the 56 women; 29 (1.4%) did not receive the allocated treatment, 16 (<1%) withdrew consent after randomisation (this included one woman who did not receive the allocated treatment), and 12 (<1%) women were lost to follow-up.\nThe exclusion of 56 (2.7%) women (with missing data) in the per protocol analysis yielded results consistent with the intention to treat analysis for the SSI incidence (7.4% (74) v 10% (98); risk reduction 0.75, 0.56 to 1.0; P=0.05).\nAssessed for eligibility\nExcluded\ndiscussion\nOn balance, the results of the four analytic scenarios suggest that closed incision NPWT may be effective in reducing SSI in obese women undergoing caesarean section.\nOur pre-specified primary analysis indicated that 9% of women in this trial developed an SSI of any type-7% in the closed incision NPWT group and 10% in the control group.\nThis difference was close to statistical significance.\nThe results of the best case, complete case, per protocol sensitivity, and multivariable analyses were consistent, favouring the closed incision NPWT intervention.\nThe primary analysis was based on a conservative assumption that women lost to follow-up did not develop an SSI; this result showed a significant relative reduction of 29% in the cumulative incidence of SSI in the closed incision NPWT group.\nIt is therefore possible that our primary\ncomparison with other studies\nOur results across all analytic scenarios were consistent, showing no significant differences in the incidence of superficial and deep SSI by trial arm.\nThe results of other studies using closed incision NPWT in this population have yielded mixed results. [22][23][24]\nVariations in SSI rates as reported in other studies in this population are likely related to the different definitions used to classify and detect SSI, 20 smaller samples, 16 25 and use of pilot and cohort designs, 16 22 26 which carry a high risk of bias and uncertainty in the results.\nThe results of several smaller trials in this population, some of which were non-blinded and industry funded, showed significant reductions of up to 50% in superficial SSI rates.\n23 26 27 A recently updated Cochrane review of use of closed incision NPWT in primary wounds included a subgroup analysis of seven studies involving 1886 obese women undergoing caesarean section. 8\nThe results of that subgroup analysis indicated a 27% reduction, albeit non-significant, in superficial SSI incidence.\nThe results of our trial, the largest in this field, suggest that closed incision NPWT may reduce superficial SSI incidence in this patient population.\nGiven that approximately 29.7 million births occur through caesarean section globally, 2 this result is clinically important.\nHowever, the decision to use closed incision NPWT in this population needs to be considered alongside any economic benefit.\nWe found no statistically significant differences in organ/space SSI.\nNotably, this study was not powered to detect potential differences.\nOur results are similar to previous research in this population.\n23 26 28\nWe also found no significant group differences in wound complications in relation to bleeding, dehiscence, haematoma, or seroma.\nimplications of findings\nThe finding of a 72% relative increase in blistering associated with closed incision NPWT may have implications for healthcare decision making.\nThe recently updated Cochrane review highlighted very low certainty evidence around blistering when comparing closed incision NPWT and standard dressings.\n8 Whether blistering (under the adhesive dressing and tape) occurred because of the dressing itself or the adhesive tape that was applied (per manufacturer's instructions) around the dressing to reinforce the dressing and help to maintain suction is not clear.\nResults of several previous trials in this population reported adverse skin reactions including blistering, erythema, and bruising.\n16 22 28\nThe occurrence of a minor treatable adverse event such as blistering that we found in this trial needs to be balanced with probable reductions in the incidence SSI.\nThus, informing women about the potential risks of closed incision NPWT, and providing targeted training to clinicians in its application, may reduce the potential for blistering.\nImportantly, patients should be partners in the decision to use closed incision NPWT as an alternative wound management therapy.\nThe generalisability of our results needs to be considered relative to the inclusion criteria applied and the low rates of SSI in our study.\nWe excluded women undergoing emergency caesarean section because they are a different population and their risk factors for SSI are not similar to those women undergoing elective and semi-urgent caesarean section.\n21 Also, emergency caesarean section as a surgical procedure is much less \"standardised\" than other more \"routine\" caesarean procedures.\nGiven the greater heterogeneity of women undergoing emergency caesarean section and of emergency caesarean section procedures, and wanting to increase internal validity to more precisely detect the potential impact of closed incision NPWT, we had to control for potential confounding variables as much as possible.\nTherefore, excluding these women meant that the caesarean procedure was more consistent in its technique and associated processes such as skin preparation and antibiotic use.\nIn terms of SSI event rates, the baseline infection rate we found was much lower than we had assumed in our sample size calculation.\nOur trial was underpowered given the low event rate and thus may not be generalisable to other clinical settings.\nThe women in this trial probably received a high standard of clinical care, based on clinical practice guidelines.\nHowever, the \"true\" rate of SSI is often underestimated using routinely collected surveillance data.\n29 With the body of evidence for the effectiveness of closed incision NPWT growing, our findings may be useful for physicians' and women's decision making regarding dressing type irrespective of the centre's SSI rates.\nstrengths and limitations of study\nStrengths of this study include its sample size, the rigorous randomisation, and prospective data collection, including weekly follow-up by dedicated research staff.\nThe results of a per protocol analysis were consistent with the intention to treat analysis, indicating minimal effect of missing data and loss to follow-up and the robustness of our results.\nAcross both intervention and control groups, the time that dressings were left in situ was consistent, with both being intact for five days.\nFurthermore, SSI and wound complication outcomes were based on the definitions in the CDC's guideline. 18\nThe pragmatic nature of this trial and the characteristics of the dressings precluded blinding of participants, clinical staff, or data collec tors.\nHowever, outcome assessors were blinded to group allocation and the intervention/comparator.\nThe process of outcome ascertainment was rigorous: two blinded outcome assessors independently ascertained SSI and wound complication data, and a third outcome assessor adjudicated any discrepancies.\non 7 May 2024 at University of California Med Ctr Medical Center Library.\nProtected by http://www.bmj.com/\nBMJ: first published as 10.1136/bmj.n893 on 5 May 2021.\nDownloaded from among outcome assessors was moderate.\nA Data Safety Monitoring Board provided oversight in terms of safety checks.\nThis trial is also one of the few in this area that was not funded by industry, thus reducing potential biases relative to its conduct and reporting.\nHowever, we note several limitations.\nFirstly, women undergoing urgent (category 1) caesarean section were excluded, despite this population having an even higher risk of developing a SSI. 21\nWe excluded these women because of ethical concerns related to trying to obtain valid consent.\nIn most instances, these women would not have enough time to consider participation in this trial.\nSecondly, 60% of eligible women were enrolled and, of these, 73% had an elective caesarean section, affecting generalisability.\nGeneralisability was maximised by recruiting women from four large public hospitals, who underwent both elective and semi-urgent caesarean section.\nThe proportions of women undergoing elective versus semi-urgent caesarean section in Queensland public hospitals typically reflects the proportions recruited in this trial.\nThirdly, over the four week data collection period, we were able to collect 30 day follow-up outcome data for all women except for the 16 women who withdrew their consent after randomisation.\nFourthly, the potential exists for false positive or false negative outcome assessments of SSI and wound complications; however, blinding and use of two outcome assessors adjudicated by a third minimised this risk.\nFifthly, we followed up the women with telephone interviews.\nThe decision to use telephone interviews was pragmatic; to bring women in weekly to assess SSI would have created an increased burden on participants and likely resulted in huge loss to followup.\nWe used this approach in preference to having missing data.\nAlso, we know that data from routine surveillance are inferior in quality to those from dedicated follow-up.\nThe survey tool we used was a previously validated patient reported tool to assess for SSI. 30\nIt had a series of questions about signs and symptoms of SSI such as redness, pain/tenderness at the incision, and discharge, as well as questions related to involvement of health professionals in the management of the wound and antibiotics prescribed for the wound.\nTo ensure the quality and consistency of the data, research nurses used an interview script based on SSI symptoms and related resource use.\nAdditionally, to minimise loss to follow-up, the research nurses contacted women on three separate occasions for each week if the women did not answer.\nOther research has shown that self-report of wound related complications is accurate when validated tools are used. 17 30\nWe cannot rule out the possibility that trial participants may have incorrectly reported their wound characteristics, but we have no reason to think that this was likely to occur.\nSixthly, we did not do time analysis because reporting time to SSI using weekly data provides limited information.\nSeventhly, despite the large sample size, the cumulative incidence of SSI was lower than expected; thus, given the wider 95% confidence intervals (less precision) for the primary outcome, a false negative result (type II error) is still possible.\nFurthermore, underestimation of the incidence of SSI is possible, given the way that missing data have been treated in the analysis of primary analysis.\nFinally, we did not have access to general practice data or information as to whether women went back to different hospitals or on any use of antibiotics for wound infection.\nTherefore, some wound complications and infections may have been missed, leading to an underestimation of SSI incidence.\nNevertheless, the women in this study were able to accurately self-report any wound related complications and treatments (for example, antibiotics).\nconclusions\nOn the basis of our primary intention to treat analysis, assigning no SSI to missing data, prophylactic closed incision NPWT for obese women after caesarean section resulted in a 24% reduction in the relative risk of SSI compared with standard dressings (3% reduction in absolute risk).\nThis difference, although close to statistical significance, possibly underestimates the effectiveness of closed incision NPWT in this population.\nOn balance, the results of the conservative primary, multivariable adjusted model, and post hoc sensitivity analyses should be considered alongside the growing body of evidence of the benefits of closed incision NPWT and given the number of obese women undergoing caesarean section globally.\nHowever, the decision to use closed incision NPWT needs to be weighed against the increase in skin blistering and economic considerations and based on shared decision making.\nASA=American Society of Anesthesiologists; Hb=haemoglobin; IQR=interquartile range; NPWT=negative pressure wound therapy.\n*Percentages might not add up because of rounding.\n\u2020Missing data for \u22645 women.\n\u2021Weight in kilograms divided by square of height in meters.\ntable 2 | clinical outcomes for intention to treat population with missing data on primary outcome (28 women) assumed to be no surgical site infection (ssi), conservatively favouring standard care*. values are numbers (percentages) unless stated otherwise IQR=interquartile range; NPWT=negative pressure wound therapy. *Worst case analysis based on effect estimate; 28 women missing primary outcome data (12 lost to follow-up; 16 withdrawn) assumed not to have SSI (favouring standard dressing as this arm has higher levels of missing data). \u2020Using \u03c7 2 test, Fisher's exact test, or Mann-Whitney U test. \u2021Data not available for randomised patients withdrawn from study. \u00a7Pain associated with surgical wound requiring readmission measured as binary variable (yes/no). \u00b65 participants had reoperations for wound complications before hospital discharge.\n |  | closed incision | standard dressing | relative risk | \nclinical outcomes | all (n=2035) | nPWt (n=1017) | (n=1018) | (95% ci) | P value \u2020\nAll SSI types | 174 (8.6) | 75 (7.4) | 99 (9.7) | 0.76 (0.57 to 1.01) | 0.06\nSuperficial | 163/174 (94) | 70/75 (93) | 93/99 (94) | 0.75 (0.56 to 1.02) | 0.72\nDeep incision | 10/174 (5.7) | 4/75 (5) | 6/99 (6) | 0.67 (0.19 to 2.36) | 0.72\nOrgan/space | 1/174 (0.6) | 1/75 (1) | 0/99 (0) | - | 0.50\nComplications | 247 (12.1) | 123 (12.1) | 124 (12.2) | 0.99 (0.79 to 1.25) | 0.95\nBleeding | 30 (1.5) | 14 (1.4) | 16 (1.6) | 0.88 (0.43 to 1.79) | 0.72\nDehiscence | 211 (10.4) | 108 (10.6) | 103 (10.1) | 1.05 (0.81 to 1.36) | 0.71\nHaematoma | 17 (0.8) | 11 (1.1) | 6 (0.6) | 1.84 (0.68 to 4.94) | 0.22\nSeroma | 53 (2.6) | 27 (2.7) | 26 (2.6) | 1.04 (0.61 to 1.77) | 0.89\nMedian (IQR) HLOS, days (n=2019) \u2021 | 3.0 (2.0-4.0) | 3 (2.0-4.0) | 3 (2.0-4.0) | - | 0.32\nReadmissions \u2021 | 36 (1.8) | 23 (2.3) | 13 (1.3) | 1.76 (0.90 to 3.46) | 0.09\nPain \u2021 \u00a7 | 32 (1.6) | 21 (2.1) | 11 (1.1) | 1.90 (0.92 to 3.93) | 0.07\nReoperations \u2021 \u00b6 | 9 (0.4) | 4 (0.4) | 5 (0.5) | 0.80 (0.22 to 2.96) | 0.75\nHLOS=hospital length of stay; |  |  |  |  | \nICU=intensive care unit; NPWT=negative pressure wound therapy; SSI=surgical site infection.\n*Per protocol population excludes participants (n=56) who were lost to follow-up (12 (<1%) participants), did not receive treatment to which they were originally allocated (28 (1%) participants), or subsequently withdrew from study (16 (<1%) participants).\n1 participant withdrawn from study did not receive treatment to which she was originally allocated.\n\u2020Using \u03c7 2 test or Fisher's exact test.\nthe bmj | BMJ 2021;373:n893 | doi: 10.1136/bmj.n893\nFunding: The trial was funded by a competitive peer reviewed grant (APP1081026) from the Australian National Health and Medical Research Council.\nThe funders had no role in considering the study design or in the collection, analysis, or interpretation of data, the writing of the report, or the decision to submit the article for publication.\nThe views expressed are those of the authors and not necessarily those of the NHS, the National Institute for Health Research (NIHR), or the Department of Health and Social Care.\nRESEARCH\nNo commercial reuse: See rights and reprints http://www.bmj.com/permissions\nSubscribe: http://www.bmj.com/subscribe protocol, critically revised the manuscript for important intellectual content, and approved the final manuscript.\nThe corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.\nBMG is the guarantor.\nDissemination to participants and related patient and public communities: The results have been and will be presented at national and international conferences.\nDissemination plans to inform the patient community of this study's results include social media, press release, and the hospital's newsletter.\nStudy results will be disseminated to the trial participants by email or letter upon their request.\nProvenance and peer review: Not commissioned; externally peer reviewed.",
        "label": "low",
        "id": "task4_RLD_test_700"
    },
    {
        "paper_doi": "10.1136/bmj.n256",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Design: RCTUnit: 20 clusters (slums)\n\n\nParticipants: Location/setting: study was carried out in Mumbai, India. Health workers who had passed 10th grade education were trained to conduct CBE within 4 weeks to perform on participants in the study.Sample size: 151,538 women aged 35 to 64Sex: only females were included.Mean age: mean age for enrolment for all trial participants in the screening and control arm was 44.84 and 44.92 years respectively.Inclusion criteria: women aged 35 to 64 years who did not have a history of breast cancer.Exclusion criteria: women outside of ages 35 to 64 years, and all women with a history of breast cancer.\n\n\nInterventions: Intervention (screening) group: (n = 75,360) of 10 clusters must undergo 9 rounds of biennial monitoring for breast cancer occurrence and mortality, 4 rounds of screening by CBE and cancer awareness education and 5 rounds of active surveillance.Control group: (n = 76,178) of 10 clusters must undergo 9 rounds of biennial monitoring for breast cancer occurrence and mortality, 1 round cancer awareness education, 8 rounds of active surveillance.Trial duration was 20 years.\n\n\nOutcomes: Primary outcomes: the down-staging of breast cancer at diagnosis and reduction in mortality from breast cancer.Secondary outcomes: CBE coverage, sensitivity, and specificity.Timing of outcome assessments: total trial duration was 20 years, but database was locked in March 2019 for analysis.\n\n\nNotes: Study start date: May 1998Study end date: March 2019Funding source: US National Institutes of Health (grant number: RO1CA074801)Conflicts of interest: author conflicts not declare\n\n",
        "objective": "To assess whether training in CBE affects the ability of health workers in LMICs to detect early breast cancer.",
        "full_paper": "Abstract\nObjective\nTo test the efficacy of screening by clinical breast examination in downstaging breast cancer at diagnosis and in reducing mortality from the disease, when compared with no screening.\nDesign\nProspective, cluster randomised controlled trial.\nSetting\n20 geographically distinct clusters located in Mumbai, India, randomly allocated to 10 screening and 10 control clusters; total trial duration was 20 years (recruitment began in May 1998; database locked in March 2019 for analysis).\nParticipants\n151\u2009538 women aged 35-64 with no history of breast cancer.\nInterventions\nWomen in the screening arm (n=75\u2009360) received four screening rounds of clinical breast examination (conducted by trained female primary health workers) and cancer awareness every two years, followed by five rounds of active surveillance every two years.\nWomen in the control arm (n=76\u2009178) received one round of cancer awareness followed by eight rounds of active surveillance every two years.\nMain outcome measures\nDownstaging of breast cancer at diagnosis and reduction in mortality from breast cancer.\nResults\nBreast cancer was detected at an earlier age in the screening group than in the control group (age 55.18 (standard deviation 9.10) v 56.50 (9.10); P=0.01), with a significant reduction in the proportion of women with stage III or IV disease (37% (n=220) v 47% (n=271), P=0.001).\nA non-significant 15% reduction in breast cancer mortality was observed in the screening arm versus control arm in the overall study population (age 35-64; 20.82 deaths per 100\u2009000 person years (95% confidence interval 18.25 to 23.97) v 24.62 (21.71 to 28.04); rate ratio 0.85 (95% confidence interval 0.71 to 1.01); P=0.07).\nHowever, a post hoc subset analysis showed nearly 30% relative reduction in breast cancer mortality in women aged 50 and older (24.62 (20.62 to 29.76) v 34.68 (27.54 to 44.37); 0.71 (0.54 to 0.94); P=0.02), but no significant reduction in women younger than 50 (19.53 (17.24 to 22.29) v 21.03 (18.97 to 23.44); 0.93 (0.79 to 1.09); P=0.37).\nA 5% reduction in all cause mortality was seen in the screening arm versus the control arm, but it was not statistically significant (rate ratio 0.95 (95% confidence interval 0.81 to 1.10); P=0.49).\nConclusions\nThese results indicate that clinical breast examination conducted every two years by primary health workers significantly downstaged breast cancer at diagnosis and led to a non-significant 15% reduction in breast cancer mortality overall (but a significant reduction of nearly 30%in mortality in women aged \u226550).\nNo significant reduction in mortality was seen in women younger than 50 years.\nClinical breast examination should be considered for breast cancer screening in low and middle income countries.\nTrial registration\nClinical Trials Registry of India CTRI/2010/091/001205; ClinicalTrials.gov NCT00632047.\nIntroduction\nThe incidence of breast cancer is rising in all countries of the world, but particularly so in low and middle income countries.\nFor example, in Mumbai, India, the incidence of breast cancer has risen by nearly 40% between 1992 and 2016, and breast cancer is the leading cause of death from cancer in women in most states of India.\nBreast cancers in low and middle income countries are frequently detected in advanced stages, and consequently, more than half the global deaths from breast cancer occur in these countries.\nWhile mammography is the established screening tool in developed countries, the screening modality that is appropriate for India and other low and middle income countries remains undetermined.\nBreast self-examination might not be useful as a general strategy,  largely because it is not feasible to ensure women perform it well.\nHowever, a case-control study based on data from the Canadian National Breast Screening Study showed that in a controlled setting, where the quality of breast self-examination was carefully evaluated, women who conducted the procedure benefitted well.\nMammography, which is widely practiced in Western countries, might not be an appropriate approach in low and middle income countries because of its cost and complexity.\nFurthermore, most women in low and middle income countries are younger than 50, and mammography is less effective in this age group.\nClinical breast examination (CBE) is an alternative screening method, and was one of the components of screening in two important randomised trials.\nThe Health Insurance Plan Study was conducted in greater New York, USA, in the 1960s during which 62\u2009000 women aged 40-64 were randomised to receive yearly CBE plus mammography or no screening.\nDuring the 1960s, mammography was in its early stages of development, and a disproportionately large number of breast cancers were detected by CBE.\nAn estimated two thirds of the reduction in breast cancer mortality in the Health Insurance Plan study could be attributed to CBE.\nTo determine the relative contributions of mammography and CBE in the reduction of breast cancer mortality, the Canadian National Breast Screening Study was initiated in the early 1980s.\nIn one of two parts of the study, women aged 50-59 were randomly allocated to receive either yearly CBE plus mammography or yearly CBE alone.\nThe trial had the potential to determine whether mammography provided any added benefit in terms of mortality reduction in addition to that provided by CBE.\nAfter 13 years of follow-up and five rounds of screening, deaths from breast cancer in the two arms were almost identical.\nThese results remained unchanged after 25 years of follow-up.\nThe findings of the Health Insurance Plan Study and Canadian National Breast Screening Study provided a strong argument for a randomised trial to compare CBE with no screening,  and formed the basis for the Mumbai study.\nThis study aimed to determine whether CBE plus provision of cancer awareness would downstage breast cancer at diagnosis and reduce mortality from the disease, compared with no screening.\nMethods\nThe Mumbai study had two components: screening for cervix cancer by visual inspection and screening for breast cancer by CBE.\nThe results of the cervical cancer component have been reported, as well as details of methodology to include design, mechanisms of community outreach, recruitment and informed consent, training of primary health workers and medical social workers, sample size estimation, adherence to screening (after three rounds), and mechanism of referral and treatment.\nThe above methodological aspects are summarised in this paper.\nDefinition of a cluster\nA cluster comprised of many closely situated dwellings in congested slum areas, defined by geographical boundaries such as railway lines, water pipelines, highways, roads, public parks, and canals.\nEach cluster had 9000 to 10\u2009000 dwellings with a population of 50\u2009000-65\u2009000, of which about 7500 women were aged 35-64.\nTransfer between control and intervention clusters was unlikely because the clusters were geographically separate, and because virtually none of the participants underwent breast screening outside the trial.\nThe standard of care in our study population was no screening.\nRandomisation method\nRandomisation was by cluster, where groups rather than individuals were chosen as units of randomisation.\nTwenty independent clusters were numbered 1-20 and randomly allocated to screening or control groups by a draw of lots.\nWith this procedure, 10 clusters were assigned as screening clusters and 10 as control clusters.\nTrial participants and intervention\nThe current study, a cluster randomised controlled trial, recruited 151\u2009538 women aged 35-64 from 20 clusters in Mumbai.\nWomen in the screening arm (n=75\u2009360) received four rounds of CBE conducted by trained female primary health workers and cancer awareness information every two years, followed by five rounds of active surveillance by way of home visits every two years.\nWomen in the control arm (n=76\u2009178) received one round of cancer awareness followed by eight rounds of active surveillance every two years.\nParticipants in both arms were eligible for free diagnostic evaluation and treatment at the Tata Memorial Hospital; women in both groups were provided with identical identity cards to obtain free treatment at the hospital.\nRecruitment started in May 1998 and was completed in April 2002.\nFour rounds of CBE were concluded in December 2007 and follow-up continued until May 2018.\nThe database was locked in March 2019 for analysis.\nSample size considerations\nWe based sample size calculations primarily on expected incidence and mortality data from breast and cervical cancer over the long duration of the study.\nIntracluster correlation was estimated using age, education status, and religion of women in the study.\nThe computation was done using MLWin Software.\nFor estimation of sample size, we considered two primary outcomes\u2014breast and cervical cancer mortality.\nSample size derived was 150\u2009000 women, which was calculated to detect 25% reduction in mortality from breast cancer with 80% power and 5% type I error, after adjusting for intracluster correlation and design effect (0.00013758 and 2.0408, respectively).\nWith these considerations, 230 deaths from breast cancer in the control group were required for mortality analysis to be recommended.\nThe smaller design effect observed in the study indicated that the sample size was adequate to estimate reduction in mortality with anticipated power.\nThree way data linkage\nTo capture information on death from any cause, the trial had a three way data linkage system.\nPrimary collection of data was done by trained medical social workers by home visits.\nData were matched with those of Mumbai Municipal Death Records and with the Mumbai Cancer Registry.\nMore information about the linkage systems and process has been provided in the supplementary material.\nBreast cancer deaths\nBreast cancer as the cause of death among women who were diagnosed with breast cancer was blindly ascertained by two independent experts.\nIf there was a discrepancy between the two experts, the records were blindly reviewed by a third independent reviewer.\nCause of death was assigned to breast cancer when at least two of the three reviewers concurred.\nCause of death could not be ascertained in 40 women.\nStatistical analysis\nWe calculated incidence rates in both arms by taking into account the number of person years determined from the date of entry into the trial to the date of diagnosis.\nThe number of person years for calculating mortality rates was determined from the date of entry in the trial to the date of death.\nData were censored during analysis for women who had migrated or were lost to follow-up, or who had died from other causes.\nAll deaths in both arms were included for all cause mortality estimates.\nWe used a Poisson regression model to estimate incidence and mortality rate ratios and their 95% confidence intervals.\nAdjustments were made for design effect.\nAll statistical tests were two sided, and P<0.05 was considered to be statistically significant.\nThe data were analysed on the basis of intention to screen (all women, irrespective of compliance), and when the predefined number of events (230 deaths) were documented in the control arm.\nAll analyses were carried out in Stata software version 12 (Stata, College Station, TX).\nThe study underwent several protocol amendments during its long course, particularly in the initial years.\nThe amendments were suggested by consultants or the data safety monitoring committee from time to time and were duly approved by the institutional review board.\nThese amendments were also approved by the funding agency (US National Cancer Institute).\nAll interpretations in the manuscript are aligned with the finally amended protocol.\nPatient and public involvement\nPatients and public were not involved in setting the research question, outcome measures, design, interpretation, or writing of the results.\nHowever, involvement of local community leaders was sought during recruitment of study participants and study implementation.\nResults\nThe CONSORT flow diagram depicting the overall trial schema is presented in figure 1.\nDemographic and breast cancer risk factors were well balanced between the two arms indicating that randomisation was without bias (supplementary table 1).\nCompliance, quality assurance, and breast cancer detection\nThe mean adherence to screening after four rounds was 67.07%, and mean adherence to hospital referral for confirmation of diagnosis was 76.21% (supplementary table 2); overall, 94.82% (n=71\u2009456) of the participants were screened at least once.\nThe average screen positivity rate was 1.28% in the four screening rounds (supplementary table 2).\nAfter four rounds of screening, 199 women with breast cancer were identified (supplementary table 3).\nBreast cancers included 114 screen detected cancers, 77 interval cancers, and eight cancers among women who did not adhere to screening in the preceding round (supplementary table 3).\nAs a quality assurance measure, a random sample of 5% of women (n=10\u2009021) was also examined by a qualified medical officer.\nThe \u0138 value for concordance was found to be 0.76, (95% confidence interval 0.72 to 0.81), indicating that the quality of CBE conducted by primary health workers met quality assurance requirements.\nAverage adherence to rounds 5-9 of active surveillance after CBE screening was 77.57%, which was similar to the average adherence to rounds 5-9 received by the control arm (77.57% v 76.22%, P=0.99; supplementary tables 4 and 5).\nOf 641 cancers detected in the screening arm overall, 199 (31%) were detected during screening rounds 1-4 and 442 (69%) were detected during the active surveillance rounds 5-9 after CBE screening (supplementary tables 2 and 4).\nAdherence to treatment and to evidence based guidelines was similar in both arms (supplementary table 6); mean adherence of these women to treatment was 98.88%.\nAdherence in the control arm to the first and the only round of cancer awareness was 90.88% (n=69\u2009231).\nAverage adherence to the subsequent eight rounds of active surveillance was 78.14% (supplementary table 5).\nAfter nine rounds of active surveillance, 655 breast cancer cases were recorded in the control arm (supplementary table 5).\nProgressively more breast cancers were detected in each round as the women aged.\nMean adherence of these women to treatment was 97.63%.\nAge at enrolment and age at diagnosis of breast cancer\nMean age at diagnosis of breast cancer in women in the screening arm was 55.18 (standard deviation 9.10 (95% confidence interval 54.47 to 55.88)).\nMean age at diagnosis in the control arm was 56.50 (9.10 (55.80 to 57.20)).\nThis difference indicated that screening had brought forward breast cancer diagnosis by 16 months (P=0.01; table 1).\nAt the time of recruitment, over 70% women in both the screening and control arms were younger than 50, whereas at the time of breast cancer diagnosis, this proportion was reversed with nearly 75% of women aged 50 and older in both arms (table 1).\nThese data implied that breast cancer was diagnosed predominantly in older women, or in younger women after they reached age 50.\nThis finding formed the basis for us to analyse the subsequent data relating to breast cancer downstaging and mortality by using age 50 as the cutoff threshold, although this threshold was not prespecified in the protocol and should be considered a post hoc analysis.\nDownstaging of breast cancer\nBiennial CBE led to significant downstaging of breast cancer in all women (P=0.001; table 2), as well as in women younger than 50 (P=0.005) and in those aged 50 and older (P=0.05).\nStaging information was unavailable in 41 women in the screening arm and 73 women in the control arm.\nHowever, we saw no difference when comparing the survival of these women with missing information (supplementary figure 1).\nBreast cancer incidence and absence of overdiagnosis\nAt the end of screening, we found 198 women with breast cancer in the screening arm and 151 in the control arm, which translated into a crude incidence rate of 60.57 and 45.30 per 100\u2009000 women years, respectively (rate ratio 1.34 (95% confidence interval 1.05 to 1.71); P=0.02; table 3).\nWe saw an excess of 47 diagnoses of breast cancer in the screening arm compared with the control arm (table 3).\nAfter a median follow-up of 18 years, the screening and control arms had 640 and 655 cases of breast cancer, respectively, which translated into a crude incidence rate of 62.76 and 64.43 per 100\u2009000 women years, respectively (0.97 (0.87 to 1.09), P=0.66; table 3).\nSupplementary table 7 shows that although, as expected, a higher incidence of breast cancer was seen in the screening group than in the control group up to study year 10 (that is, until the end of screening round 4), this difference reduced gradually from study year 12 onwards (starting surveillance round 1) and disappeared completely by study year 18 (surveillance round 5).\nBreast cancer mortality\nWe recorded 213 breast cancer deaths in the screening arm and 251 deaths in the control arm (rate ratio 0.85 (95% confidence interval 0.71 to 1.01), P=0.07; table 3).\nThus, overall, a 15% non-significant reduction in mortality was seen when women of all ages were considered.\nAmong women younger than 50, 149 breast cancer deaths were recorded in the screening arm and 158 deaths in the control arm (0.93 (0.79 to 1.09), P=0.37).\nAmong women aged 50 and older, 64 breast cancer deaths were recorded in the screening arm and 93 deaths in the control arm (0.71 (0.54 to 0.94), P=0.02; table 3).\nThis subset analysis based on the age 50 threshold was not stipulated in the protocol and was a post hoc analysis.\nThe cumulative breast cancer mortality in the screening and control arms over 20 years is shown in figure 2.\nAn excess of breast cancer deaths in the screened population was seen in both age subgroups (age <50 and \u226550) in the early years after randomisation (fig 2), which lasted for about 14 years in women younger than 50 and about six years in those aged 50 and older.\nWhen breast cancer mortality data were analysed on the basis of attendance to the number of CBE screening rounds, we found that even women younger than 50 who attended all four rounds of screening benefitted significantly in terms of mortality reduction (rate ratio 0.66 (95% confidence interval 0.53 to 0.83), P<0.001).\nBut this benefit did not exist if these women attended only three rounds (0.88 (0.60 to 1.27), P=0.48).\nWomen aged 50 and older, however, benefitted from attending both three as well as four rounds of screening (attendance to all four rounds (0.64 (0.45 to 0.93), P=0.02); attendance to three rounds (0.66 (0.44 to 1.00), P=0.05); supplementary table 8).\nAll cause mortality\nWhen we considered all cause mortality during the 20 year period, we saw a non-significant reduction of 5% in the screening arm.\nAll cause mortality rates were 1100.59 and 1162.25 per 100\u2009000 women years in the screened and controls arms, respectively (rate ratio 0.95 (95% confidence interval 0.81 to 1.10); P=0.49).\nThe subdivision of all cause mortality by age (<50 and \u226550) is also represented (table 3).\nBreast cancer deaths comprise less than 3% of deaths from all causes in women in India; and hence a reduction in all cause mortality was not expected.\nThe cumulative all cause mortality in the screening and control arms over 20 years is shown in supplementary figure 2.\nDiscussion\nStatement of principal findings\nWe report here results of our randomised trial that compared CBE screening with no screening.\nWe showed that biennial CBE performed by trained female primary health workers significantly advanced breast cancer diagnosis by 16 months, and also downstaged the disease with fewer stage III or IV cancers in screened women.\nOverall, CBE led to a non-significant 15% reduction in breast cancer mortality; however, a significant reduction of nearly 30%was observed in women aged 50 and older.\nIn women younger than 50, despite successful downstaging, no mortality reduction was observed.\nLack of mortality reduction in younger women is consistent with data reported in some mammography trials, and could point to undetermined biological factors.\nParticipant attendance to the number of screening rounds also appeared to be important in breast cancer mortality reduction for women younger than 50.\nWe found a 34% mortality reduction in this age group if the women attended all four rounds of screening (P<0.001).\nThis benefit, however, disappeared if they attended only three rounds (mortality reduction 13%, P=0.48).\nFor women aged 50 and older, however, we observed mortality reduction after attendance to three and four rounds of screening (34%, P=0.05 and 36%, P=0.02, respectively; supplementary table 8).\nStrengths and weaknesses in relation to other studies\nTwo other randomised trials have compared CBE screening with no screening.\nA cluster randomised controlled trial was initiated in Kerala, India, in 2006 where three rounds of CBE every three years was planned to evaluate whether CBE can reduce incidence of advanced breast cancers and mortality from the disease.\nEarly results have shown a higher proportion of early stage breast cancers in the intervention arm than in the control arm.\nAnother trial comparing CBE screening with no screening in the Philippines could not be satisfactorily concluded because of unacceptably low levels of adherence, possibly because of external investigators not fully anticipating cultural and psychosocial barriers.\nIn our study, an excess mortality from breast cancer was seen in the screening arm during the first few years of screening for the total study population as well as when stratified by age groups.\nSuch an excess mortality was also seen in the cervical cancer component of this trial.\nA meta-analysis of seven breast cancer screening trials suggested an excess breast cancer mortality up to the fifth year of screening in women younger than 50 and in the first year in older women.\nThis excess was, however, not apparent in a combined analysis of Swedish trials.\nThe possible finding of early excess cancer mortality needs exploring.\nThe theory of biological predeterminism (pre-existing micrometastases before diagnosis and surgery) fails to explain this excess mortality but could point towards an impact of events at the time of diagnosis and surgery on mortality.\nStrengths and weaknesses of this study\nOne crucial element of our study that led to its successful completion was that it was entirely indigenous.\nThe trial was conceived, designed and implemented by a team based in Mumbai and had full understanding of the psychosocial, geopolitical, and geographical ground realities that influence the conduct of complex, public health randomised trials in low and middle income countries.\nOur study was conducted in slum areas largely inhabited by socioeconomically disadvantaged women who often moved residence requiring our medical social workers to trace their new abodes, sometimes in far flung parts of the city.\nOwing to our medical social workers making innumerable home visits to a population that was often mobile, we were able to achieve a satisfactory compliance at all levels of screening.\nThe quality of CBE performed by our primary health workers was also of high standard, which was confirmed by comparing the screening findings with a specialist breast clinician.\nWe were also able to capture death records of a high proportion of cases because of the three way data linkage system.\nFinally, our study included near perfect randomisation for a cluster randomised controlled trial; all demographic and breast risk factors were equally distributed in the screening and control arms.\nProvision of timely treatment could have helped to improve quality of life in screened women by preventing advanced stage disease, including local recurrence.\nOur study also had some limitations.\nCancer staging data were unavailable from 41 women in the screening arm and 73 women in the control arm.\nThis limitation probably did not affect the study results because the survival curves of patients with missing staging information were similar in the screening and control arms (supplementary figure 1).\nHowever, a sensitivity analysis of patients with missing staging information, in which all 41 women from the screening arm were assigned cancer stages III or IV and all 73 women from the control arm were assigned to cancer stages I or II, led to loss of statistical significance in the downstaging effect of screening.\nAnother study limitation was that cause of death information was not available through death certificates and the available documents for some women.\nTo overcome this limitation, three independent experts reviewed the records of all women with breast cancer who had died.\nBreast cancer was assigned as a cause of death only when at least two reviewers concurred (213 (83%) of 258 in the screening arm and 251 (90%) of 278 in the control arm).\nOur blinded review process for assigning cause of death was based on similar mechanisms used in other screening trials.\nHowever, the possibility of some residual uncertainty cannot be excluded; some degree of variability is inevitable in screening trials when death certificates are often modestly accurate and medical records often incomplete.\nWe did not observe a significant reduction in all cause mortality.\nBut because breast cancer deaths comprise less than 3% of all deaths in women in India, we did not expect a reduction in all cause mortality in our study.\nMeaning of the study\u2014possible explanations and implications for clinicians and policymakers\nOur study validates CBE as an alternative modality of breast screening.\nIt demonstrates that CBE screening is effective in reducing breast cancer mortality in Indian women aged 50 and older without any overdiagnosis.\nIn our trial, we were able to use a vertical programme with dedicated staff that was centrally controlled.\nFurthermore, women in India and in many other low and middle income countries are relatively lean and have smaller breasts than women in Western countries.\nThe health workers who screened women with CBE in this trial had passed 10th grade education and could be trained to perform CBE within a minimal training period (about four weeks).\nWe believe that CBE screening by primary health workers is replicable in the general population, and CBE has already been implemented in other parts of India as pilot schemes.\nOur study suggests that implementation of population screening by CBE in low and middle income countries is feasible, provided that adequate training of screening providers, careful monitoring, and quality of performance are assured.\nWhether the use of CBE in low and middle income countries at the community level can lead to a reduction in breast cancer mortality is still unknown.\nIts success can only be ascertained several years after CBE has been implemented as public health programme.\nWhat is already known on this topic\nBreast cancer screening by mammography reduces mortality in women aged 50 and older, but its effectiveness in women younger than 50 is questionable\nBreast self-examination has not been proven to be an effective method for early detection of breast cancer\nWhether screening by clinical breast examination can reduce mortality from breast cancer is not known\nWhat this study adds\nIn a 20 year study, clinical breast examination conducted by trained female health workers in Mumbai led to a downstaging of breast cancer at diagnosis and reduced mortality from the disease by nearly 30% in women aged 50 and older, but with no mortality reduction seen in women younger than 50\nA 5% reduction in all cause mortality was seen in the screening arm compared with the control arm, but was not statistically significant\nClinical breast examination should be considered for breast cancer screening in low and middle income countries\nExtra material supplied by authors\nTrial flow diagram\nCumulative breast cancer mortality during 20 years of study\n\nAge at enrolment of all women and age at diagnosis of breast cancer\nArm | Age at enrolment (for all trial participants) |  | Age at diagnosis (for participants with breast cancer only)\nTotal No | No of women aged <50 (%) | No of women aged \u226550(%) | P value | Mean age (SD (95% CI)) | Difference (95% CI) | Total No | No of women aged <50 (%) | No of women aged \u226550(%) | P value | Mean age (SD (95% CI)) | Difference (95% CI)\nScreening | 75\u2009177* | 54\u2009212 (72.11) | 20965 (27.89) | 0.06 | 44.84 (7.90 (44.78 to 44.90)) | 0.078 (\u22120.002 to 0.158) |  | 640\u2020 | 161 (25.16) | 479 (74.84) | 0.01 | 55.18 (9.10 (54.47 to 55.88)) | 1.321(0.330 to 2.312)\nControl | 76\u2009097* | 54\u2009188 (71.21) | 21909 (28.79) | 44.92 (8.00 (44.86 to 44.97)) |  | 655 | 147 (22.44) | 508 (77.56) | 56.50 (9.10 (55.80 to 57.20))\n\nSD=standard deviation.\nInformation on age was not available for 183 women in the screening arm and 81women in the control arm among the total women enrolled.\nOf the 641 women with breast cancer in the screening arm, one had bilateral breast cancer, who was considered only once.\n\nStaging of breast cancer at diagnosis \nAge group | Randomised group | Stages I or II (No (%)) | Stages III or IV (No (%)) | Total No | Pearson x2 | Difference (%) in stages III+IV between screening and control arms (95% CI)\nAll women* | Screening arm | 379 (63) | 220 (37) | 599 | 11.757 (P=0.001) | 9.83 (4.208 to 15.368)\nControl arm | 311 (53) | 271 (47) | 582\n<50\u2020 | Screening arm | 271 (63) | 161 (37) | 432 | 8.034 (P=0.005) | 9.77 (3.008 to 16.423)\nControl arm | 206 (53) | 183 (47) | 389\n\u226550\u2021 | Screening arm | 108 (65) | 59 (35) | 167 | 3.906 (P=0.05) | 10.27 (0.094 to 20.092)\nControl arm | 105 (54) | 88 (46) | 193\n\nStaging information unavailable from 41 women in the screening arm and 73 women in the control arm.\nStaging information unavailable from six women in the screening arm and 12 women in the control arm.\nStaging information unavailable from 35 women in the screening arm and 61 women in the control arm.\n\nBreast cancer incidence, breast cancer mortality, and all cause mortality after 20 years since commencement of study\n | Screening arm |  | Control arm | Rate ratio(95% CI)\u2020 | P value\nTotal No of women | No of diagnoses or deaths | No of person years | Crude rate per 100\u2009000 person years (95% CI) | Total No of women | No of diagnoses or deaths | No of person years | Crude rate per 100\u2009000 person year (95% CI)\nBreast cancer incidence\nCompletion of active screening | 75\u2009360 | 198 | 326\u2009891.2 | 60.57 (49.87 to 74.62) |  | 76\u2009178 | 151 | 333\u2009346.7 | 45.30 (38.51 to 53.64) | 1.34 (1.05 to 1.71) | 0.02\nCompletion of 20 years of study | 75\u2009360 | 640 | 1\u2009019\u2009761 | 62.76 (57.02 to 69.35) |  | 76\u2009178 | 655 | 1\u2009016\u2009616 | 64.43 (60.43 to 68.90) | 0.97 (0.87 to 1.09) | 0.66\nBreast cancer mortality\nAll ages* | 75\u2009360 | 213 | 1\u2009023\u2009097 | 20.82 (18.25 to 23.97) |  | 76\u2009178 | 251 | 1\u2009019\u2009500 | 24.62 (21.71 to 28.04) | 0.85 (0.71 to 1.01) | 0.07\nAge <50 | 54\u2009212 | 149 | 763\u2009141.8 | 19.53 (17.24 to 22.29) |  | 54\u2009188 | 158 | 751\u2009367.0 | 21.03 (18.97 to 23.44) | 0.93 (0.79 to 1.09) | 0.37\nAge \u226550 | 20\u2009965 | 64 | 259\u2009955.2 | 24.62 (20.62 to 29.76) |  | 21\u2009909 | 93 | 268\u2009133.1 | 34.68 (27.54 to 44.37) | 0.71 (0.54 to 0.94) | 0.02\nAll cause mortality\nAll ages* | 75\u2009360 | 11\u2009261 | 1\u2009023\u2009180 | 1100.59 (989.98 to 1224.58) |  | 76\u2009178 | 11\u2009853 | 101\u20099831 | 1162.25 (1037.16 to 1303.45) | 0.95 (0.81 to 1.10) | 0.49\nAge <50 | 54\u2009212 | 4450 | 763\u2009177.7 | 583.09 (539.66 to 629.69) |  | 54\u2009188 | 4708 | 751\u2009508.2 | 626.47 (572.73 to 684.32) | 0.931 (0.829 to 1.045) | 0.23\nAge \u226550 | 20\u2009965 | 6811 | 260\u2009001.8 | 2619.6 (2456.3 to 2796.9) |  | 21\u2009909 | 7145 | 268\u2009323.2 | 2662.8 (2498.2 to 2835.8) | 0.984 (0.902 to 1.073) | 0.71\n\nInformation on age not available for 183 women in the screening arm and 81 women in the control arm among study participants of all ages.\nRate ratio calculated by Poisson regression model after adjusting for cluster design. ",
        "label": "high",
        "id": "task4_RLD_test_906"
    },
    {
        "paper_doi": "10.1371/journal.pmed.1001445",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: A parallel arm cluster-RCT conducted in 90 sites in Vietnam, Quang Ninh Province between Jul 2008 and Jun 2011.\n\n\nParticipants: Sample size: 90 clusters (22,561 births; 1243 mother-newborn pairs randomly selected for secondary outcomes).Clusters : eligible districts had NMR >= 15/1000 in 2005.Individuals: all mother-newborn pairs within the study area with births from July 2008 to June 2011 were eligible. There were 22,561 births registered in the study area during the study period.\n\n\nInterventions: Target: health system (policy/practice change).Arm 1 (44 clusters, 11,906 births): the intervention consisted of facilitated work with stakeholder groups (primary care staff, local politicians and women's union representatives) on the commune level and included the identification of local perinatal health problems and use of a problem-solving cycle. The 44 communes in the intervention group had a total of 1508 maternal and newborn health groups (MNHG) meetings, lasting approximately 2 hours each. The problem-solving process identified 15-27 unique problems which resulted in 19-27 unique actions applied 297-649 times per year.Arm 2 (46 clusters, 10,655 births): standard health care in control communities. A 6% random sample of all registered live births, surviving the neonatal period, was continuously selected each month in order to represent the entire birth cohort for secondary outcome data. Home visits were performed for families of a deceased newborn in another random sample.\n\n\nOutcomes: Trial primary outcome: neonatal mortality.Review outcomes reported:Primary: maternal mortality.Secondary: ANC coverage (at least 1 visit), health facility deliveries, tetanus protection, perinatal mortality, neonatal mortality.Other: postnatal home visit.\n\nFollow-up: data collection took place monthly for 3 years.\n\n\nNotes: Funders: Swedish International Development Cooperation Agency (Sida), Swedish Research Council, and Uppsala University\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Lars \u00c5ke Persson and colleagues conduct a cluster randomised control in northern Vietnam to analyze the effect of the activity of local community-based maternal-and-newborn stakeholder groups on neonatal mortality.\nPlease see later in the article for the Editors' Summary\nBackground\nFacilitation of local women's groups may reportedly reduce neonatal mortality.\nIt is not known whether facilitation of groups composed of local health care staff and politicians can improve perinatal outcomes.\nWe hypothesised that facilitation of local stakeholder groups would reduce neonatal mortality (primary outcome) and improve maternal, delivery, and newborn care indicators (secondary outcomes) in Quang Ninh province, Vietnam.\nMethods and Findings\nIn a cluster-randomized design 44 communes were allocated to intervention and 46 to control.\nLaywomen facilitated monthly meetings during 3 years in groups composed of health care staff and key persons in the communes.\nA problem-solving approach was employed.\nBirths and neonatal deaths were monitored, and interviews were performed in households of neonatal deaths and of randomly selected surviving infants.\nA latent period before effect is expected in this type of intervention, but this timeframe was not pre-specified.\nNeonatal mortality rate (NMR) from July 2008 to June 2011 was 16.5/1,000 (195 deaths per 11,818 live births) in the intervention communes and 18.4/1,000 (194 per 10,559 live births) in control communes (adjusted odds ratio [OR] 0.96 [95% CI 0.73\u20131.25]).\nThere was a significant downward time trend of NMR in intervention communes (p\u200a=\u200a0.003) but not in control communes (p\u200a=\u200a0.184).\nNo significant difference in NMR was observed during the first two years (July 2008 to June 2010) while the third year (July 2010 to June 2011) had significantly lower NMR in intervention arm: adjusted OR 0.51 (95% CI 0.30\u20130.89).\nWomen in intervention communes more frequently attended antenatal care (adjusted OR 2.27 [95% CI 1.07\u20134.8]).\nConclusions\nA randomized facilitation intervention with local stakeholder groups composed of primary care staff and local politicians working for three years with a perinatal problem-solving approach resulted in increased attendance to antenatal care and reduced neonatal mortality after a latent period.\nTrial registration\nCurrent Controlled Trials ISRCTN44599712\nPlease see later in the article for the Editors' Summary\nEditors' Summary\nBackground\nOver the past few years, there has been enormous international effort to meet the target set by Millennium Development Goal 4 to reduce the under-five child mortality rate by two-thirds and to reduce the number of maternal deaths by three-quarters, respectively, from the 1990 level by 2015.\nThere has been some encouraging progress and according to the latest figures from the World Health Organization, in 2011, just under 7 million children aged under 5 years died, a fall of almost 3 million from a decade ago.\nHowever, currently, 41% of all deaths among children under the age of 5 years occur around birth and the first 28 days of life (perinatal and neonatal mortality).\nSimple interventions can substantially reduce neonatal deaths and there have been several international, national, and local efforts to implement effective care packages to help reduce the number of neonatal deaths.\nWhy Was This Study Done?\nIn order for these interventions to be most effective, it is important that the local community becomes involved.\nCommunity mobilization, especially through local women's groups, can empower women to prioritize specific interventions to help improve their own health and that of their baby.\nAn alternative strategy might be to mobilize people who already have responsibility to promote health and welfare in society, such as primary care staff, village health workers, and elected political representatives.\nHowever, it is unclear if the activities of such stakeholder groups result in improved neonatal survival.\nSo in this study from northern Vietnam, the researchers analyzed the effect of the activity of local maternal-and-newborn stakeholder groups on neonatal mortality.\nWhat Did the Researchers Do and Find?\nBetween 2008 and 2011, the researchers conducted a cluster-randomized controlled trial in 90 communes within the Quang Ninh province of northeast of Vietnam: 44 communes were allocated to intervention and 46 to the control.\nThe local women's union facilitated recruitment to the intervention, local stakeholder groups (Maternal and Newborn Health Groups), which comprised primary care staff, village health workers, women's union representatives, and the person with responsibility for health in the commune.\nThe groups' role was to identify and prioritize local perinatal health problems and implement actions to help overcome these problems.\nOver the three-year period, the Maternal and Newborn Health Groups in the 44 intervention communes had 1,508 meetings.\nEvery year 15\u201327 unique problems were identified and addressed 94\u2013151 times.\nThe problem-solving processes resulted in an annual number of 19\u201327 unique actions that were applied 297\u2013649 times per year.\nThe top priority problems and actions identified by these groups dealt with antenatal care attendance, post-natal visits, nutrition and rest during pregnancy, home deliveries, and breast feeding.\nNeonatal mortality in the intervention group did not change over the first two years but showed a significant improvement in the third year.\nThe three leading causes of death were prematurity/low birth-weight (36%), intrapartum-related neonatal deaths (30%), and infections (15%).\nStillbirth rates were 7.4 per 1,000 births in the intervention arm and 9.0 per 1,000 births in the control arm.\nThere was one maternal death in the intervention communes and four in the control communes and there was a significant improvement in antenatal care attendance in the intervention arm.\nHowever, there were no significant differences between the intervention and control groups of other outcomes, including tetanus immunization, delivery preparedness, institutional delivery, temperature control at delivery, early initiation of breastfeeding, or home visit of a midwife during the first week after delivery.\nWhat Do These Findings Mean?\nThese findings suggest that local stakeholder groups comprised of primary care staff and local politicians using a problem-solving approach may help to reduce the neonatal mortality rate after three years of implementation (although the time period for an expected reduction in neonatal mortality was not specified before the trial started) and may also increase the rate of antenatal care attendance.\nHowever, the intervention had no effect on other important outcomes such as the rate of institutional delivery and breast feeding.\nThis study used a novel approach of community-based activity that was implemented into the public sector system at low cost.\nA further reduction in neonatal deaths around delivery might be achieved by neonatal resuscitation training and home visits to the mother and her baby.\nAdditional Information\nPlease access these Web sites via the online version of this summary at http://dx.doi.org/10.1371/journal.pmed.1001445.\nThe World Health Organization provides comprehensive statistics on neonatal mortality\nThe Healthy Newborn Network has information on community interventions to help reduce neonatal mortality from around the world\nIntroduction\nPerinatal and late neonatal mortality remains a challenge in low- and middle-income countries in contrast to the progress in reducing post-neonatal and child mortality.\nA number of single interventions and packages has proven effective in reducing the number of stillbirths and/or neonatal deaths.\nCoverage of such interventions is currently being monitored across the continuum of care from pregnancy to early childhood.\nScaling-up of efficacious single or package interventions may be difficult and result in low effectiveness.\nSocial and geographic variation in reach and quality of services may result in inequity in neonatal survival and obstacles in reaching child mortality goals.\nThese dilemmas have prompted the development of different social or systems interventions, e.g., through community mobilization by facilitation of local women's groups.\nSuch groups usually follow a problem-solving cycle from identification of their own prioritized perinatal problem to evaluation of the actions taken.\nTwo trials in Nepal and India, in contexts with a relatively high initial level of neonatal mortality, have reported a reduction by 30%\u201345% that has been judged to be cost-effective.\nA participatory community-based intervention in Uganda that was informed by local health data was also effective in reducing infant mortality.\nFacilitation of community groups in Mumbai slums did not result in effects on health care or mortality.\nAn effort to scale-up a community-based intervention in Bangladesh did not result in lower neonatal mortality.\nStudies of women's group interventions and perinatal outcomes are ongoing in African countries but so far it is not known to what extent these approaches may be effective in contexts outside South Asia and in settings with a medium-level neonatal mortality rate (NMR), i.e., in the range 15\u201330/1,000.\nThe care providers have not been directly represented in these community mobilization approaches, which may preclude actions related to provision of services and quality of care.\nOne may also question the sustainability of an approach with women's groups that are specifically established for this purpose.\nAn alternative strategy may be to mobilize people, who already have responsibility to promote health and welfare in society, e.g., primary care staff, village health workers (VHWs), and elected representatives of local political or non-governmental organizations.\nIt is not known whether facilitation of such local stakeholder groups may result in improved perinatal survival.\nThus, the aim of this trial was to analyse the effect of facilitation of local maternal-and-newborn stakeholder groups on neonatal mortality in a province in northern Vietnam.\nMethods\nThe study protocol of this trial has been published (Figure 1).\nStudy Location and Population\nThe trial was performed in Quang Ninh province, located in the northeast of Vietnam, which has a long coastline and borders China.\nThere are more than 1 million inhabitants with 89% being the Kinh majority and 11% belonging to ethnic minority groups, mainly in the mountains in the western part.\nVietnam, now classified as a middle-income country, had reportedly a crude birth rate in 2010 of 17/1,000, mortality before the age of 5 y of 19 per 1,000 live births, a NMR of 12 per 1,000 live births, and a maternal mortality ratio of 56 per 100,000 live births (adjusted, year 2008).\nA survey was performed in 2006, covering all live births and neonatal deaths in Quang Ninh province in 2005.\nThe main aim was to map perinatal health services and analyse levels of neonatal mortality and its geographical variation and plan for this trial.\nCommune health centres (CHC) provided antenatal care, which was attended by three-quarters of the mothers.\nDelivery care was offered by CHCs, or by hospitals at district, province, and regional levels.\nNeonatal mortality was 16/1,000 live births.\nA quarter of the mothers who had lost a newborn in the neonatal period had not had any contact with the health system at time of death\u2014and this situation was more likely to be the case among mothers of ethnic minorities.\nThere was a huge variation in neonatal mortality between the districts, ranging from 10 in the urban districts along the coast to 44/1,000 live births in the mountainous inland areas.\nProcedures\nThe study area of this trial were 90/187 communes in the province located in districts with NMR\u226515/1,000 in 2005 (Figure 2).\nThe number of inhabitants in the different communes ranged from 1,000 to 18,000; amounting to a total of 350,000 within the study area.\nThe rationale for excluding districts with lower mortality was the assumption, that a community-based participatory intervention potentially should be less effective in areas with a low NMR.\nThe included districts had 6,306 births and an average NMR of 24/1,000 in 2005.\nThe trial had a cluster-randomized design, with the geopolitical unit commune constituting the cluster.\nAn individual randomization was not possible, due to the intervention on the commune level.\nEach commune has a political leadership, provides services to the population including a CHC, and is a well-recognised entity in Vietnamese society.\nThe sampling strategy was one-stage cluster sampling with probability proportional to size (PPS) of the clusters.\nThe PPS, in this case number of births per year, was chosen to obtain similar distribution of sizes of clusters across intervention and control communes.\nSampling was neither blocked nor paired.\nA sampling frame was established with a cumulative list of number of births in each of the communes in the 2005 survey.\nA random number list was used to subsequently allocate \u201cintervention\u201d or \u201ccontrol\u201d to the list of communes, and 44 out of the 90 communes were allocated to intervention and 46 to control.\nThe randomization was performed by one of the involved researchers at Uppsala University.\nThe sequence was concealed until the intervention was assigned; otherwise the allocation was not masked.\nAll mother\u2013newborn pairs within the study area with births from July 2008 to June 2011 were eligible to be included in the trial.\nThe intervention consisted of facilitated work of maternal-and-child stakeholder groups on the commune level that included identification of local perinatal health problems followed by a problem-solving cycle.\nWe hypothesised that this would improve the quality and coverage of perinatal services and after some latency lead to improved neonatal survival in comparison with control communes.\nPrimary outcome was neonatal mortality.\nA latent period before effect on the primary outcome is expected in this type of intervention, but this timeframe was not pre-specified.\nSecondary outcomes were: (a) care-seeking behaviour (in the analyses represented by attendance to antenatal care, tetanus immunisation during antenatal care, reported material and financial preparedness for delivery as part of antenatal care, and institutional delivery); (b) exclusive breast-feeding, represented by initiation of breast-feeding within 1 h; (c) temperature control at delivery (defined as the newborn placed naked and dried on mother's chest immediately after delivery in combination with the not being bathed within the first hour of life); (d) home visit by midwife during first week after delivery; and (e) perinatal health knowledge of primary health staff.\nThe last secondary outcome is not being reported here; it will be presented in a separate publication.\nStatistical power calculations were based on estimates of neonatal mortality, which was the primary outcome; NMR 24/1,000 and 6,251 live births in the study area in 2005 (on average 69 in each commune).\nWe arbitrarily estimated the design effect to be 1.5, considering the relatively high number of clusters and the low average cluster size.\nA 3-y sample would allow demonstrating a significant reduction of 7/1,000 in NMR, i.e., to 17/1,000 or less, with 80% power at a 0.05 significance level.\nThis effect size is comparable to those reported from the trials in Nepal and India.\nA steering board for the intervention was established, chaired by the hospital director of the regional Uong Bi hospital, and with members from the Women's Union (WU), the Provincial Health Bureau, the Ministry of Health, and the collaborating research institutions.\nThe board monitored the progress of the trial and tried to motivate the local health institutions to support the efforts of the trial.\nThe results of an interim analysis in 2010 and of all 3 y completed in 2011 were reported to the steering board.\nNo stopping rules were applied, since we did not anticipate any negative effect of the intervention on the cluster or individual level.\nWe recruited lay women from the WU in the province to act as facilitators in supporting CHC staff and key commune stakeholders in improving perinatal health care practices.\nThe WU is an organization with high national coverage working with various issues related to the situation of women in Vietnam.\nThere is a long tradition of involvement at local and regional levels of WU in issues of welfare, particularly surrounding health care.\nIn a pilot study we trained two facilitators and evaluated their ability to work in this role that included participatory communication techniques and found that the facilitation strategy was feasible.\nEight individuals from local WU organizations were recruited for the trial and were trained by research team members for 2 wk.\nRecruitment criteria included being an experienced WU member, having completed secondary school, and having children.\nAdditionally three facilitators were recruited during the trial to replace facilitators leaving because of childbirth or new employment.\nThe 11 facilitators (nine from the ethnic majority Kinh and two from the Tay ethnic minority group) had a mean age of 32 y at recruitment.\nThey were paid by the project on full-time basis during the entire 3 y of intervention.\nThe training program of the facilitators included theoretical sessions, group discussions, role-plays, and field practice.\nIt covered topics such as group dynamics, quality improvement methods (e.g., brainstorming and the plan-do-study-act cycle) and basic evidence-based perinatal care.\nA facilitation manual and a specific diary were developed to guide the work of the facilitators.\nTwo research team members coordinated the facilitation process and acted as supervisors of the facilitators; i.e., field supervision and performing 2-d meetings with all facilitators once a month during the entire trial period.\nMaternal and Newborn Health Groups (MNHG) were constituted in each intervention commune (by directives from the Provincial Health Bureau).\nThese groups consisted of eight members: three CHC staff (physician, midwife, nurse); one of the VHWs of the commune; one population collaborator, the chairperson, or vice chairperson of the commune (having responsibility for health in the commune); and two WU representatives (from village and commune levels).\nThe facilitators primarily used the plan-do-study-act cycle in mobilizing the groups in identifying and priorizing local perinatal health problems and accomplishing improvement cycles that included concrete actions on prioritized problems (Figure 3).\nSuch improvement cycles on different problems were performed continously over the intervention period in all MNHGs.\nWe strived to keep each facilitator performing monthly meetings with the same MNHG over the 3-y intervention period.\nEach facilitator was responsible for five to eight MNHGs.\nTwenty MNHGs kept the same facilitator the whole period, 22 changed facilitator once, and two MNHGs changed facilitator twice.\nWhen appropriate, the facilitators were recommended to highlight recommendations provided by the Vietnamese National Standard and Guidelines on Reproductive Health Care Services from 2003.\nMNHG members were not paid and did not receive any allowances for participating in the NeoKIP project as the improvement work was assumed to be an integrated part of their normal duties.\nAll members of the MNHGs had in their different professions important roles in the commune's health and welfare.\nThe facilitation process and the work of the MNHGs were monitored.\nMeetings, issues dealt with, and actions taken were recorded and are briefly reported in the result section.\nFacilitators kept diaries, and focus groups were held with facilitators and MNHG members (unpublished data).\nData collectors, who had no contact with the facilitation process, were responsible for collection of data on all births and neonatal deaths in the study area.\nThey attended monthly meetings at CHCs, where VHWs regularly report vital events to CHC staff.\nThey also performed monthly visits to all district hospitals and to the two provincial and regional level hospitals in the area to collect information on all births, stillbirths, maternal deaths, and neonatal deaths in the area.\nThey kept an updated list of pregnant women, which was used to ascertain that valid information on all pregnancy outcomes was obtained.\nTriangulation was systematically performed of birth and neonatal death data from all included sources (records from district, provincial, and regional hospitals, records and reports from CHCs, reports provided by VHWs).\nData of births and birth outcomes were carefully crosschecked between the different sources of information in order to ascertain that all births and their outcome were registered, and that no duplication of information occurred.\nThis methodology was developed in the 2005 baseline survey and judged to provide valid information.\nStillbirth was defined as birth of a dead foetus after an estimated 28 wk of gestation.\nLive birth was defined as birth of a foetus with any sign of viability.\nNeonatal death was defined as death of a live birth during the first 28 d of life.\nWhen a probable neonatal death was identified, a data collector performed a home visit to ascertain the case, and perform a verbal autopsy-based classification of cause of death.\nA questionnaire, modified from the World Health Organization (WHO) generic verbal autopsy instrument was used in the interview that also included the mother's or other family member's story of the events prior to the neonatal death.\nIn the training of the data collectors the importance to adhere to the pregnancy outcome definitions was stressed as well as the ability to differentiate between stillbirths and early neonatal deaths.\nThree paediatricians independently scrutinized the information provided and assigned a cause of death according to the WHO ICD-10 classification.\nIn case of disagreement the cause was established in a consensus process.\nA questionnaire-based interview on socio-economic information and health care utilisation was also performed.\nDetails of the cause-of-death results have been presented elsewhere.\nCoordinates of households of all neonatal deaths were also registered with a global positioning system tool (GPS).\nA 6% random sample of all registered live births, surviving the neonatal period, was continuously selected (each month) in order to represent the entire birth cohort.\nAs for the families of the deceased newborn cases home visits were performed to families in this random sample, and interviews on socio-economic information, perinatal health care utilisation, and newborn care were performed.\nGPS coordinates of the households of the sample of newborns were collected.\nThis random subsample of the population of live births will be referred to as \u201creferents\u201d and the newborn deaths will be referred to as \u201ccases.\u201d\nAnalyses\nSupervisors checked the data collected in the field.\nA second data quality control was done before computerisation, and the databases were carefully checked for completeness and consistency.\nSpatial information was used to produce a map of intervention and control communes and all CHCs in the area.\nProblems identified in the facilitated group meetings and actions taken were quantified and described.\nInformation from first-year randomly selected birth interviews was used to describe socio-economic characteristics and basic perinatal health services utilisation in intervention and control communes.\nInformation on births and pregnancy outcomes in the 2005 baseline survey was also analysed for intervention and control communes.\nThe random sample of referents was used to analyse the secondary outcomes.\nThe cases and the sample of referents were also used in a supportive analysis of effect of the intervention on neonatal mortality outcome adjusting for baseline covariates (nested case-referent analysis).\nFrequencies of births, stillbirths, live births, neonatal deaths, and maternal deaths were analysed for the entire trial period (July 2008 to June 2011) as well as for each of the three 12-mo periods (July 2008 to June 2009, July 2009 to June 2010, and July 2010 to June 2011, respectively).\nStillbirth rates (per 1,000 births), perinatal mortality rates (per 1,000 births), and neonatal mortality rates (per 1,000 live births) were calculated with 95% CI.\nIn order to assess the homogeneity within the communes (clusters) the intraclass (or intracluster) correlation coefficient (ICC) for the binomial variable neonatal death was calculated by use of the R package \u201caod\u201d.\nIn \u201caod\u201d (that stands for analysis of overdispersed data) point estimates of ICC for clustered binomial data are estimated using a one-way random effect model.\nThe effect of intervention (intervention versus control) on the primary outcome neonatal mortality (binary; deceased versus not deceased) for all live births in the study population was analysed by means of generalized linear mixed models (GLMM) using the R package \u201clme4\u201d.\nIn the model intervention was included as a fixed factor, nested within the random factor commune (i.e., the clusters).\nResults are presented as odds ratios (OR) and 95% CIs.\nAnalyses were performed for the entire trial period of 3 y as well as for each year separately (calendar periods defined above).\nTime trends with regard to neonatal mortality were analysed by including the fixed factor year (1, 2, and 3, respectively) as well as the interaction term year*intervention in the main model described above.\nFurther, we estimated intervention effects on neonatal mortality outcome by including adjustment for socio-economic covariates (presented in Table 1) in a nested case-referent analysis (since baseline characteristics were not available for the entire cohort).\nIntervention was included as a fixed factor, nested within the random factor commune (i.e., the clusters).\nThe baseline socio-economic characteristics were included as fixed factors in the model.\nThe effect of intervention on the binary secondary outcomes (i.e., antenatal care, tetanus immunisation, delivery preparedness, institutional delivery, temperature control, early breastfeeding, home visits) in the previously described subsample of live births was analysed in the same manner as the analyses of the primary outcome, and time trend analyses were also performed accordingly.\nAlso, baseline socio-economic factors (Table 1) were controlled for.\nEthics Statement\nInformed consent was sought from parents of neonates in the random sample of all live births as well as from parents of the deceased neonates at home-based interviews.\nThe project was ethically reviewed and approved by the Ministry of Health, Hanoi, and the Regional Research Ethics Committee, Uppsala University.\nBefore start of study the project was presented and approved at the provincial health authority level, and thereafter presented and informally approved in a series of meetings at district and commune levels.\nResults\nThe 90 communes in districts with NMR \u226515/1,000 in 2005 were randomly allocated to 44 intervention communes and 46 control communes (Figure 1).\nNo communes were lost to follow-up.\nOne intervention commune stopped the facilitation meetings after 21 mo, i.e., 15 mo before the end of trial, while all others completed the intervention.\nEthnicity, economic situation, education, and utilisation of health services were similar among delivering women in randomized intervention and control communes (Table 1).\nPregnancy outcomes and neonatal mortality rates in interventions and control communes had been similar in the 2005 baseline (Table 2).\nIn the 44 communes allocated to the intervention the facilitated groups had 1,508 meetings (out of a possible total of 1,584 monthly meetings).\nOn average a meeting lasted for 2 h.\nEvery year 15\u201327 unique problems were identified and addressed 94\u2013151 times.\nThe problem-solving processes resulted in an annual number of 19\u201327 unique actions that were applied 297\u2013649 times per year.\nThe problems most frequently identified, and the most frequent actions taken are listed in Box 1.\nMost problems were related to the utilisation of health services and the mother's own care of herself and her newborn.\nThe actions most frequently dealt with communication of messages from maternal and child health service providers and counselling of mothers.\nBox 1. Most Frequently Identified Problems and Actions Taken (n of Times That Problem/Action Was Identified and Implemented)\nProblems\nLow frequency of antenatal visits at the right time (42)\nLow frequency of post-natal home visits (33)\nLow awareness among pregnant women of appropriate diet, work, and rest (23)\nHigh frequency of home deliveries (16)\nLow awareness among pregnant women about appropriate breast feeding practices (14)\nActions\nCommunication activities (623)\nPrepare education material and train VHWs (154)\nPost-natal home visits (63)\nCreate lists of pregnant and newly delivered women (28)\nDistribute leaflets (25)\nThere were 22,561 births registered in the study area from July 2008 to June 2011, whereof 184 resulted in stillbirths (8.2/1,000 births).\nOf the 22,377 live births 389 died in the neonatal period (17/1,000 live births), Table 2.\nIntraclass correlation coefficient (ICC) (neonatal death) was 0.0073 for the entire trial period and 0.0091 for year 3 (July 2010 to June 2011).\nNMR from July 2008 to June 2011 was 16.5/1,000 (195 deaths per 11,818 live births) in the intervention communes compared with 18.4/1,000 (194 per 10,559 live births) in control communes (adjusted OR 0.96 [95% CI 0.73\u20131.25]).\nThere was a significant monotonic downward time trend of NMR in intervention communes but not in control communes (time trend analysis of intervention arm: year 2 versus year 1 p\u200a=\u200a0.856; year 3 versus year 1 p\u200a=\u200a0.007; year 3 versus year 1 and 2 p\u200a=\u200a0.003.\nControl arm: year 2 versus year 1 p\u200a=\u200a0.324; year 3 versus year 1 p\u200a=\u200a0.561; year 3 versus year 1 and 2 p\u200a=\u200a0.184).\nWhen evaluating the time trend in the entire material with an interaction term intervention:year the third year was also significantly different (p\u200a=\u200a0.0128).\nNo significant difference in NMR was observed during the first 2 y of the trial (July 2008 to June 2009 and July 2009 to June 2010, respectively) while the third year (July 2010 to June 2011) showed a significantly lower NMR in intervention arm; adjusted OR 0.51 (95% CI 0.30\u20130.89), Table 2.\nAn additional analysis was performed with a nested case-referent approach adjusting the neonatal mortality outcome for baseline characteristics (ethnicity, maternal education level, maternal age, and household poverty).\nThe effect estimates showed a similar pattern as in the cohort analysis (Table S1).\nSecondary outcomes represented various aspects of antenatal care, delivery care, and newborn care (Table 3).\nAntenatal care was significantly more common among women in intervention communes (adjusted OR 2.27, 95% CI 1.07\u20134.80) with a significant time trend in the intervention arm.\nThere were no significant differences in the following secondary outcomes: tetanus immunisation as part of the antenatal services (OR 1.64, 95% CI 0.83\u20133.24), the presence of delivery preparedness (OR 1.33, 95% CI 0.67\u20132.64), institutional delivery (OR 1.88, 95% CI 0.60\u20135.87), temperature control at delivery (OR 1.28, 95% CI 0.50\u20133.25), early initiation of breastfeeding (OR 0.93, 95% CI 0.6\u20131.37), or home visit of a midwife during the first week after delivery (OR 1.06, 95% CI 0.51\u20132.20) (Table 3).\nThe three leading causes of death were prematurity/low birth-weight (36%), intrapartum-related neonatal deaths (30%), and infections (15%) (Table 4).\nDuring the third year of the trial a reduction in the number of deaths in intervention communes was seen in low birth weight/prematurity, intrapartum-related, as well as infectious diseases deaths.\nStillbirth rates were 7.4 per 1,000 births (95% CI 5.9\u20138.9) in intervention arm and 9.0 per 1,000 births (95% CI 7.2\u201310.8) in control arm.\nThere was one maternal death in the intervention communes and four in the control communes.\nDiscussion\nA randomized facilitation intervention in a Vietnamese province with local maternal and newborn stakeholder groups composed of primary care staff and local politicians, who used a problem-solving approach through monthly meetings for 3 y, resulted in reduced neonatal mortality after a latent period.\nThe top priority problems and actions identified by these groups dealt with antenatal care attendance, post-natal visits, nutrition and rest during pregnancy, home deliveries, and breast feeding.\nThe intervention also increased antenatal care attendance, while there were no effects on secondary outcomes around delivery and newborn care.\nBaseline characteristics of mothers were similar in intervention and control communes, with the exception poor households that were slightly more common in control communes.\nAn additional analysis with adjustment for baseline covariates using a nested case-referent approach did not change the level of effect estimates on neonatal mortality.\nFurther, neonatal mortality rate did not differ between intervention and control communes in the 2005 baseline survey.\nNo deviations from the protocol were observed except from one out of 44 MNHGs that ended the facilitation intervention two-thirds of the way into the trial.\nThe facilitated intervention with MNHGs maintained a high activity with a large number of problems identified and actions taken, in spite of no extra financial benefits to the group members.\nThe birth and neonatal outcome data collection system was separated from the facilitation intervention activities, and it included triangulation of different data sources, careful cross-checking of data, as well as systematic control of pregnancy outcomes of the cumulative lists of pregnant women in the commune that the data collectors kept.\nOne of the VHWs in the commune (around one out of ten) was involved in the MNHG while all VHWs were informants in the data collection system.\nThe information on births and/or neonatal deaths provided by the VHWs who were involved in the intervention was most likely not biased, since the updated lists of pregnant women enabled a systematic enquiry on a montly basis on pregnancy outcomes, as well as a triangulation of information sources and cross-checking of data.\nThe system for collection of outcome data was developed for the 2005 survey, and was judged to provide the best possible data on births and neonatal deaths in this setting.\nFor the study design and the facilitation intervention we were inspired by the Promoting Action on Research Implementation in Health Services model.\nThis middle-range theory highlights three major ingredients for being successful in implementing research into practice: (1) the nature of the evidence being used, (2) the quality of the context in terms of coping with change, and (3) the type of facilitation needed to ensure a successful change process.\nImplementation is conceived as a multifaceted intervention, rather than a more straightforward, linear process of translating knowledge from experts to the local level.\nIn the trial we analysed the effect of facilitation of local stakeholder groups focusing maternal and neonatal health problems and actions.\nExcellent evidence is available for a series of effective preventive and curative activities that have been translated to national guidelines for the Vietnamese health system context.\nThe intervention did not impose the guidelines on the local stakeholder groups.\nThey were free to decide which problems to focus on and what actions to take in order to address those problems.\nChallenging existing practices and supporting new ways of doing things facilitated this bottom-up approach to change and development.\nNeoKIP was a complex social intervention.\nIt was context-specific and continuously subject to negotiation and interpretation among the involved local stakeholders.\nTherefore, it could not be expected that this type of intervention should result in immediate effects on neonatal mortality that might be achieved by a single or package intervention\u2014a delay must be anticipated before any mortality reduction can be shown.\nIn the protocol of this trial a latent period was considered, but the duration of this was not pre-specified.\nThis is a limitation for the results we present, but when considering this point, it should also be noted that there was a significant downward trend in NMR in the intervention arm resulting in a reduction of neonatal mortality of public health importance in the third 12-mo period of the study (OR 0.51, 95% CI 0.30\u20130.89), while the control arm had no significant trend.\nThe intervention was strengthened by the selection of members in the MNHGs, who had professional responsibilities that related to maternal and newborn health and welfare; as midwives at health centres, as members of the powerful Women's Union, as VHWs or as local politicians.\nHowever, the NeoKIP intervention was a new approach for local stakeholders, who were not used to collaborate in this kind of group activity.\nThis type of approach requires active and diciplined stakeholders who assess, discuss, and find ways to overcome contextual barriers that may impede the process of implementation.\nThus, to succeed with this undertaking time, commitment, and perseverance were needed.\nMNHG members gradually identified a number of antenatal, delivery, and neonatal problems and decided on actions directed towards pregnant women and their households, the health services, or the general public.\nThey used loudspeakers (common in Vietnamese villages) to motivate the public for antenatal care and delivery at hospitals, they trained VHWs to support an optimal utilisation of perinatal services, they produced and distributed leaflets regarding perinatal health issues, and mobilized the local community.\nThere are several features in the causal inference framework that support the effect on neonatal mortality; the reports from the MNHGs reflect intense activities for improved perinatal health, the information on utilisation of services suggests a process of change in intervention communes that could result in improved pregnancy outcome and the change in cause-specific mortality.\nMost of the problems identified by the MNHGs dealt with the demand side and less with problems of the health care providers.\nChanges reportedly occurred in one of the most frequent problems identified, i.e., low antenatal care attendance, where health care utilisation of women in intervention communes differed favourably from those in control communes.\nA major part of actions taken dealt with communication and counselling of mothers.\nIn the Makwanpur trial in Nepal of participatory women's groups, where maternal and newborn mortality was reduced, a process evaluation indicated that women in intervention clusters attended antenatal care to a larger extent, and moved towards behaviours favourable for perinatal health.\nIn the NeoKIP trial we did not see any overall sigificant increase in institutional deliveries or care of the newborn, such as temperature control, initiation of breastfeeding, and home-visits by health personnel in the early neonatal period.\nA participatory approach with groups of women is maybe more likely to influence the immediate newborn care of the birthing women.\nAlthough some reduction was noted intrapartum-related neonatal deaths or asphyxia still remained a major problem, especially in home deliveries and at district hospitals.\nA participatory approach with local stakeholders may potentially influence quality of perinatal care, but the records of the MNHGs reveal that the main focus was on the demand side.\nA balanced intervention with a community participatory approach combined with efforts to improve quality of perinatal care could potentially reduce mortality even further.\nA further reduction of intrapartum-related neonatal deaths could be achieved by neonatal resuscitation training.\nHome visits to the mother and her newborn child (part of Ministry of Health guidelines) and improved management of neonatal infections are also needed.\nAs part of the 2005 survey of perinatal health services and outcomes the level of knowledge of neonatal among primary care level in the study area was assessed to be low.\nFurther, lack of resources, low frequency of deliveries, lack of formal training in perinatal care, and poorly paid staff were observed barriers to keeping skills at an adequate level in the health care context.\nThis novel approach is an example of a community-based activity that was implemented into the public sector system,, where so far knowledge on effectiveness has been missing.\nThe Vietnamese health system struggles to develop and meet changing needs in a society in rapid economic growth and transition.\nIn this province the private or non-governmental sector plays a limited role in relation to maternal and neonatal health services.\nIn other provinces private care providers play a more important role, although private delivery services are limited.\nThe reports from the MNHGs tell that efforts were made to improve communication between public sector care providers and mothers, and maybe with other household members, which may have motivated more mothers to attend antenatal care.\nThe study province in northern Vietnam has a medium-level neonatal mortality and a comprehensive health system for maternal and neonatal services.\nOverall there were relatively few home deliveries but a great geographic and social inequity in coverage of perinatal services and level of neonatal survival.\nThe facilitation intervention with local stakeholder groups composed of primary care staff and local politicians working for 3 y with a perinatal problem-solving approach reduced neonatal mortality after a latent period.\nIncremental costs for this type of intervention are judged to be low, and mainly related to cost for the laywoman facilitator and the indirect costs of MNHGs monthly meetings (total incremental cost US$77,000 or US$6.5 per birthing woman).\nQuang Ninh province is relatively typical for many Vietnamese provinces, with the dominating ethnic majority in towns and most rural areas, and ethnic minority groups in the remote, mountainous areas.\nWe have shown that community-based participatory approaches to reduce neonatal mortality is not only effective in South Asian societies but also in a South-East Asian society in rapid transition.\nStudy flow.\nQuang Ninh province, Vietnam with study area, and randomized intervention and control communes.\nAction cycle in Maternal and Newborn Health Groups.\n\nBaseline characteristics of mothers in intervention and control communes. Data from a random sample (398/7,033) of live births, first year of trial.\nCharacteristic | Intervention Communes | Control Communes\nEthnic minority household | 33 (71/213) | 38 (70/185)\nPoor household | 19 (41/213) | 27 (50/185)\nMother farmer | 42 (89/213) | 51 (95/185)\nMother lack formal education | 15 (32/213) | 21 (38/185)\nMother<20 y old | 8.9 (19/213) | 9.2 (17/185)\nFirst-born child | 39 (84/213) | 38 (71/185)\n\nData are percent (numerator/denominator).\n\nNeonatal mortality outcome.\nOutcome | Baseline (2005) | Year 1 | Year 2 | Year 3 | Year 1\u20133\n | Intervention | Control | Intervention | Control | Intervention | Control | Intervention | Control | Intervention | Control\nBirths | 3,264 | 3,042 | 3,783 | 3,303 | 4,038 | 3,625 | 4,085 | 3,727 | 11,906 | 10,655\nStillbirths | 26 | 29 | 23 | 29 | 37 | 35 | 28 | 32 | 88 | 96\nLive births | 3,238 | 3,013 | 3,760 | 3,274 | 4,001 | 3,590 | 4,057 | 3,695 | 11,818 | 10,559\nNeonatal deaths | 80 | 70 | 72 | 59 | 76 | 57 | 47 | 78 | 195 | 194\nEarly (0\u20136 d) | 68 | 57 | 56 | 50 | 61 | 41 | 37 | 55 | 154 | 146\nLate (7\u201328 d) | 12 | 13 | 16 | 9 | 15 | 16 | 10 | 23 | 41 | 48\nMaternal deaths | 2 | 4 | 0 | 2 | 0 | 1 | 1 | 1 | 1 | 4\nStillbirths/1,000 births | 8.0 | 9.5 | 6.1 | 8.8 | 9.2 | 9.7 | 6.9 | 8.6 | 7.4 | 9.0\nPerinatal deaths/1,000 births | 28.8 | 28.3 | 20.9 | 23.9 | 24.3 | 21.0 | 15.9 | 23.3 | 20.3 | 22.7\nNeonatal deaths/1,000 live births (95% CI) | 24.8 (20.0\u201330.7) | 23.2 (18.4\u201329.3) | 19.1 (15.2\u201324.0) | 18.0 (14.0\u201323.2) | 19.0 (15.2\u201323.7) | 15.9 (12.3\u201320.5) | 11.6 (8.7\u201315.4) | 21.1 (17.0\u201326.3) | 16.5 (14.4\u201319.0) | 18.4 (16.0\u201321.1)\nAdjusted ORa | \u2014 | \u2014 | 1.08 (0.66\u20131.77) | 1.0 | 1.23 (0.75\u20132.01) | 1.0 | 0.51 (0.30\u20130.89) | 1.0 | 0.96 (0.73\u20131.25) | 1.0\n\nAdjusted for cluster-randomization, generalized linear mixed models (GLMM; binomial).\n\nSecondary outcomes, time period July 2008 to June 2011.\nOutcome | Intervention | Control | ORa | 95% CI\nAntenatal careb | 91 (596/656) | 82 (482/587) | 2.27 | 1.07\u20134.80\nTetanus immunisation | 89 (575/646) | 84 (474/566) | 1.64 | 0.83\u20133.24\nDelivery preparedness | 83 (543/656) | 78 (460/587) | 1.33 | 0.67\u20132.64\nInstitutional delivery | 91 (594/656) | 87 (510/587) | 1.88 | 0.60\u20135.87\nTemperature control | 6.2 (41/656) | 5.1 (30/587) | 1.28 | 0.50\u20133.25\nEarly breast feeding | 57 (372/656) | 57 (334/587) | 0.93 | 0.64\u20131.37\nHome visit | 9.0 (59/656) | 7.8 (46/587) | 1.06 | 0.51\u20132.20\n\nAnalysis performed on random sample of mothers with live births (n\u200a=\u200a1,243). Data are percent (numerator/denominator), adjusted OR with 95% CI.\nGeneralized linear mixed models, adjusted for cluster design and socio-economic covariates (ethnic minority, lack of formal education, mother<20 y of age and poor household).\nTime trend analysis antenatal care attendance in the intervention arm with year 1 as reference, year 2 p\u200a=\u200a0.066, and year 3 p\u200a=\u200a0.021.\n\nCauses of neonatal death.\nCause of death | ICD-10 | Year 1 | Year 2 | Year 3 | Year 1\u20133\n |  | Intervention | Control | Intervention | Control | Intervention | Control | Intervention | Control\nLow birth weight, prematurity | P07.0-4 | 35 | 17 | 27 | 17 | 17 | 28 | 79 | 62\nIntrapartum-related neonatal death | P21 | 25 | 20 | 25 | 19 | 11 | 15 | 62 | 54\nInfections |  | 5 | 5 | 15 | 11 | 7 | 16 | 27 | 32\nSepsis | P36.9 | 5 | 2 | 10 | 8 | 7 | 15 | 22 | 25\nPneumonia | P23 | 0 | 1 | 3 | 3 | 0 | 1 | 3 | 5\nTetanus | A33 | 0 | 2 | 2 | 0 | 0 | 0 | 2 | 2\nMalformation | Q04\u2013Q89 | 2 | 8 | 3 | 3 | 1 | 3 | 6 | 14\nOther causes |  | 1 | 6 | 0 | 2 | 5 | 9 | 6 | 17\nHypoglycaemia | E15 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0\nRespiratory distress syndrome | P22 | 1 | 0 | 0 | 1 | 0 | 0 | 1 | 1\nUmbilical haemorrhage | P51 | 0 | 2 | 0 | 0 | 0 | 3 | 0 | 5\nNeonatal jaundice | P58 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1\nIntoxication | P93 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0\nNeglect | P96.8 | 0 | 3 | 0 | 1 | 3 | 6 | 3 | 10\nUnknown cause | P96.9 | 5 | 4 | 5 | 4 | 6 | 7 | 15 | 15\nTotal number of deaths |  | 73 | 60 | 75 | 56 | 47 | 78 | 195 | 194\n",
        "label": "unclear",
        "id": "task4_RLD_test_503"
    },
    {
        "paper_doi": "10.21147/j.issn.1000-9604.2020.04.06",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Accrual: June 2011 to December 2015Single centrePhase of trial: 3Study design: open-label RCTCountry or countries where the trial was conducted: ChinaMedian follow-up: 57.3 months\n\n\nParticipants: Age: median 49, range 22-64 yearsNodal status of breast cancer: 37% node positive, 63% node negativeAdjuvant or neoadjuvant: adjuvantNotable exclusion criteria: none\n\n\nInterventions: Arm 1: paclitaxel 150 mg/m2 + carboplatin AUC3 every 2 weeks for 8 cyclesArm 2: epirubicin 80 mg/m2 and cyclophosphamide 600 mg/m2 every 2 weeks for 4 cycles followed by paclitaxel 175 mg/m2 every 2 weeks for 4 cycles\n\n\nOutcomes: Primary3-year DFS rate, defined as the date of randomisation to the date of the first local/distant recurrence (in the absence of other primary malignancies)SecondaryOS, defined as the time from randomisation to death due to any causeToxicity, according to NCI-CTCAE, version 3.0\n\n\nNotes: Trial registration record: NCT01378533All randomised participants were included in analysis.Study did not report assessing the proportional hazards assumption.Funding considerations: funded by the National Key Research and Development Program of China and the Chinese Academy of Medical Science Initiative for Innovative Medicine\n\n",
        "objective": "To evaluate the benefits and harms of platinum\u2010based chemotherapy as adjuvant and neoadjuvant treatment in people with early triple\u2010negative breast cancer.",
        "full_paper": "Objective\nThe objective of this open-label, randomized study was to compare dose-dense paclitaxel plus carboplatin (PCdd) with dose-dense epirubicin and cyclophosphamide followed by paclitaxel (ECdd-P) as an adjuvant chemotherapy for early triple-negative breast cancer (TNBC).\nMethods\nWe included Chinese patients with high recurrence risk TNBC who underwent primary breast cancer surgery.\nThey were randomly assigned to receive PCdd [paclitaxel 150 mg/m2 on d 1 and carboplatin, the area under the curve, (AUC)=3 on d 2] or ECdd-P (epirubicin 80 mg/m2 divided in 2 d and cyclophosphamide 600 mg/m2 on d 1 for 4 cycles followed by paclitaxel 175 mg/m2 on d 1 for 4 cycles) every 2 weeks with granulocyte colony-stimulating factor (G-CSF) support.\nThe primary endpoint was 3-year disease-free survival (DFS); the secondary endpoints were overall survival (OS) and safety.\nResults\nThe intent-to-treat population included 143 patients (70 in the PCdd arm and 73 in the ECdd-P arm).\nCompared with the ECdd-P arm, the PCdd arm had significantly higher 3-year DFS [93.9% vs. 79.1%; hazard ratio (HR)=0.310; 95% confidence interval (95% CI), 0.137\u22120.704; log-rank, P=0.005] and OS (98.5% vs. 92.9%; HR=0.142; 95% CI, 0.060\u22120.825; log-rank, P=0.028).\nWorse neutropenia (grade 3/4) was found in the ECdd-P than the PCdd arm (47.9% vs. 21.4%, P=0.001).\nConclusions\nPCdd was superior to ECdd-P as an adjuvant chemotherapy for early TNBC with respect to improving the 3-year DFS and OS.\nPCdd also yielded lower hematological toxicity.\nThus, PCdd might be a preferred regimen for early TNBC patients with a high recurrence risk.\nIntroduction\nTriple-negative breast cancer (TNBC) is characterized by the lack of expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor 2 (HER2).\nTNBCs, which account for approximately 12%\u221220% of all invasive breast cancers, are resistant to endocrine and HER2-targeted therapy; their aggressive behavior and poor prognosis make them one of the most challenging cancers to treat.\nPostoperative adjuvant therapy for early breast cancer, which is an important part of comprehensive treatment, can reduce the risk of recurrence and metastasis.\nAt present, since the use of polygenic prognostic detection in domestic hospitals is low, the decision of adjuvant treatment for early breast cancer patients is relatively conservative.\nThe clinical application of an anthracycline sequential taxane regimen and aromatase inhibitors has also reached an expert consensus.\nSystemic chemotherapy is generally recommended by guidelines and is, thus, currently considered as a mainstay of TNBC management.\nHowever, the proposed chemotherapy regimens remain controversial.\nIn routine clinical practice, anthracycline and taxane-containing regimens are the most commonly used systemic cytotoxic regimens for TNBC patients.\nAdding platinum to neoadjuvant chemotherapy regimens not only substantially increases the pathological complete response (pCR) rate but may also improve the event-free survival (EFS) or overall survival (OS) of TNBC patients according to previous trials.\nPlatinum-based neoadjuvant\u00a0chemotherapy\u00a0may be recommended as an option in\u00a0TNBC\u00a0patients with the cost of higher hematological toxicity incidence.\nHowever, there is limited direct evidence regarding an appropriate platinum-based adjuvant chemotherapy.\nFurthermore, determination of the optimal regimen balancing well-tolerated adverse toxicity with high efficacy is difficult.\nUnderlying genetic conditions appear to play an important role in TNBC.\nBRCA1-positive tumors show distinct clinic pathological characteristics.\nSeventy percent of all BRCA1-positive breast cancers and up to 23% of BRCA2 carriers have a TNBC phenotype.\nTNBC tumors with germline BRCA (gBRCA) mutation are associated with a better response to DNA-damaging systemic regimens such as the platinum agents.\nDose-dense chemotherapy (i.e., a chemotherapy regimen in which each cycle has a shortened treatment interval) is associated with significant improvements in survival and has been considered for use in the adjuvant setting for TNBC.\nWith granulocyte colony-stimulating factor (G-CSF) support, dose-dense chemotherapy regimens at the optimal dose have been permitted at two-week intervals rather than the conventional three-week cycle in early breast cancer regimens.\nData supporting platinum-based adjuvant regimens for TNBC are scarce and are based mostly on retrospective research.\nGiven the lack of well-established prospective or randomized studies, we conducted this study to compare the efficacy and safety of dose-dense paclitaxel plus carboplatin (PCdd) with those of the commonly used dose-dense epirubicin and cyclophosphamide followed by paclitaxel (ECdd-P) as adjuvant chemotherapy treatment in Chinese TNBC patients with high recurrence risk.\nMaterials and methods\nStudy design\nThis was a randomized, open-label, single-center study conducted in Chinese females with TNBC at high recurrence risk.\nThe study was approved by the Independent Ethics Committee of the National Cancer Center/Cancer Hospital (No. CH-BC-012).\nAll interventions were performed in accordance with the Declaration of Helsinki, guidelines of the International Conference for Harmonization/Good Clinical Practice.\nThe study was registered with the ClinicalTrials.gov (No. NCT01378533).\nParticipants\nAll participating patients provided written informed consent.\nFemale patients aged 18\u221265 years who had undergone primary breast surgery for confirmed ER-negative, PR-negative, and HER2-negative breast cancer were eligible.\nER, PR and HER2 status were determined by immunohistochemistry (IHC) on patients\u2019 tumor sections.\nThe IHC cutoff for ER-negative and PR-negative status was 1% or less positive tumor cells with nuclear staining.\nHER2-negative status was determined by IHC by giving a score of 0 or 1 or by the absence ofHER2 amplification (HER2/CEP17 ratio <2.0 and HER2 copies <4.0) upon fluorescence in situ hybridization (FISH) analysis.\nER, PR and HER2 analyses were performed centrally in a single laboratory of National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences.\nPatients were selected with positive axillary lymph or with other high-risk factors for recurrence (e.g., age <35 years, grade III disease, and intravascular cancer embolus).\nFurther details regarding this study protocol are available in the Supplementary Table S1.\nRandomization and masking\nThe patients were randomly assigned to receive either the PCdd or the ECdd-P regimen.\nSimple randomization was conducted with no stratification factors and was carried out by using\u00a0random allocation sequence.\nThe patients, medical staff, and investigators were aware of treatment allocation and assessing outcomes.\nProcedures\nPatients in both study arms received treatment in two-week cycles.\nPatients assigned to the PCdd arm received paclitaxel 150 mg/m2 on d 1 plus carboplatin AUC=3 on d 2 for 8 cycles.\nPatients assigned to the ECdd-P arm received epirubicin 80 mg/m2 divided in 2 d and cyclophosphamide 600 mg/m2 on d 1 for 4 cycles followed by paclitaxel 175 mg/m2 on d 1 for 4 cycles.\nProphylactic G-CSF 3 \u00b5g/kg was administered during each cycle according to European Society for Medical Oncology (ESMO) and American Society of Clinical Oncology (ASCO) guidelines in the dose-dense setting.\nToxicities were managed through dose delays of up to 3 weeks, and dose reductions were permitted in the following events: grade 4 hematological, grade 3 or 4 non-hematological, or other protocol-specified toxic effects.\nSafety was monitored with adverse events (AEs) reports, physical examinations, regular laboratory tests and electrocardiogram assessments at the end of each cycle until the 30th day of the last follow-up cycle.\nOutcomes\nThe primary efficacy endpoint was the 3-year disease-free survival (DFS) rate, which was calculated from the date of randomization to the date of the first local/distant recurrence (in the absence of other primary malignancies).\nSecondary objectives included OS and safety.\nOS was defined as the time from randomization to death due to any cause.\nWe analyzed the DFS and OS in patients who received at least one dose of the study treatment (intention-to-treat population, ITT).\nIn the safety analysis, we evaluated the numbers and proportions of patients in each treatment arm who had any AEs, delay of chemotherapy, and dose reduction.\nAEs were graded using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 (NCI-CTCAE, version 3.0).\nIn addition, we conducted exploratory subgroup analyses according to age (\u226440 vs. >40 years), Ki-67 index (\u226430 vs. >30), tumor size (<2 cm vs. \u22652 cm), nodal status (negative vs. positive), and surgery-chemotherapy interval (<30 dvs. \u226530 d) to investigate whether the treatment effect varied by subgroup.\nSample size computation\nThe sample size was calculated based on the primary endpoint, i.e., 3-year DFS rate.\nAssuming an approximate higher proportion of 0.10 as a primary outcome in PCdd regimen (results of our preliminary clinical research demonstrated the proportion achieving 3-year DFS in the ECdd-P regimen was 80.0%), an overall sample size of 133 participants (66 in the ECdd-P arm and 67 in the PCdd arm) was calculated to achieve 80.0% power with an alpha level at 0.05, with a 5% dropout rate in each control/treatment arm.\nSince the censoring proportion during the course of the study might be higher than expected; therefore, the sample size was increased to 143 patients to ensure the target number of events would be reached in a reasonable time frame.\nStatistical analysis\nAll statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) software (Version 22.0; IBM Corp., New York, USA).\nData were presented as numbers (%) or as the mean standard deviation.\nFrequency tables were analyzed by using the\u00a0\u03c72 test.\nThe survival analysis was estimated using the Kaplan-Meier product-limit method in the ITT population.\nThe hazard ratio (HR) and 95% confidence interval (95% CI) were estimated using the Cox proportional hazard model.\nPatients not showing progression were censored at the study cutoff date.\nThe multivariable Cox model was used for subgroup analysis to explore the influence of clinical characteristics on the 3-year DFS.\nThe safety analysis set included all randomized patients who received at least one dose of the study treatment and underwent at least one post-baseline safety assessment.\nA P value of <0.05 was considered statistically significant.\nResults\nPatients\nFrom June 2011 to December 2015, 143 patients were randomly enrolled in the PCdd arm (n=70) or the ECdd-P arm (n=73).\nAfter excluding 11 patients [treatment discontinuation due to tumor progression (n=1), withdrawal after chemotherapy (n=3), and lost to follow-up (n=7)], 132 patients who completed the planned eight cycles of chemotherapy were included in the per-protocol analysis (Figure 1).\nThe data cutoff for the primary analysis was November 30th, 2018.\nBaseline characteristics were balanced between arms (Table 1).\nAll enrolled patients had an Eastern\u00a0Cooperative\u00a0Oncology\u00a0Group (ECOG) performance status score of 0\u22121, and 62.2% were postmenopausal.\nThe median age was 49 (range, 22\u221264) years.\nIn total, 111 patients (77.6%) were aged older than 40 years.\nMost patients had stage II or III disease (n=92, 64.3%), and 90.9% of patients had invasive ductal carcinoma.\nMore than 42% had T2\u2212T4 tumors, and 53 patients (37.1%) were clinically node positive.\nMore than 75% of patients had a Ki-67 proliferation index >30%.\nSurvival outcomes\nAs of cutoff date, the median duration of follow-up was 57.3 (range, 1.2\u221298.6) months, with 58.1 months in the PCdd arm and 56.1 months in the ECdd-P arm.\nIn total, 98 patients (74.2%) were followed up over 4 years.\nDuring the study period, 23 relapse events were recorded, 5 in the PCdd arm and 18 in the ECdd-P arm.\nMost events (96.2%) were observed during the first 3 years after first diagnosis.\nIn the full analysis of ITT population, patients had significantly fewer DFS events in the PCdd arm than in the ECdd-P arm (5 vs. 18; HR=0.310; 95% CI, 0.137\u22120.704; log-rank, P=0.005).\nThe 3-year DFS was 93.9% (95% CI, 88.2%\u221299.6%) in the PCdd arm and 79.1% (95% CI, 69.7\u221288.5%) in the ECdd-P arm.\nThe Kaplan-Meier curves for DFS remained separated for the rest of the 3-year follow-up (Figure 2).\nData on OS were immature.\nEight patients died in the ECdd-P arm, whereas only one died in the PCdd arm; all deaths were cancer related.\nPreliminary data showed a potential trend on a higher 3-year OS rate in the PCdd arm (98.5% vs. 92.9%; HR=0.142; 95% CI, 0.060\u22120.825; log-rank, P=0.028) (Figure 3).\nSubgroup analyses showed a consistent DFS benefit in the PCdd arm, with the difference reaching statistical significance in the following subgroups: age >40 years (HR=4.31; 95% CI, 1.42\u221213.11; P=0.010), Ki-67 index >30% (HR=3.80; 95% CI, 1.08\u221213.36; P=0.038), and clinically evaluated lymph nodes (HR=5.73; 95% CI, 1.28\u221225.65; P=0.022) ( Figure 4).\nAEs\nOverall, both regimens were well tolerated with manageable AEs.\nThere were more patients who experienced chemotherapy delay [25 (35.7%) vs. 23 (31.5%), P=0.361] and dose reduction [16 (22.9%) vs. 14 (19.2%), P=0.369] in the PCdd arm than in the ECdd-P arm, but the difference was not significant (Table 2).\nThe most frequent AEs were neutropenia, nausea and emesis.\nThe incidence of grade 3 or 4 neutropenia was significantly higher in the ECdd-P arm than that in the PCdd arm [35 (47.9%) vs. 15 (21.4%), P=0.001], while the incidence of other grade 3 and 4 AEs was similar between the two arms.\nThere was also no significant difference in the incidence of peripheral neuropathy between the two arms (Table 3).\nNo death or life-threatening event was recorded during the study or within 30 days after the last cycle of treatment.\nDiscussion\nThis open-label, randomized study achieved its primary endpoint, with a statistically significant difference in the 3-year DFS rate in patients randomized to receive PCdd as adjuvant chemotherapy for high-risk early TNBC vs. ECdd-P (93.9% vs. 79.1%; HR=0.310; 95% CI, 0.137\u22120.704; log-rank P=0.005).\nFurther, PCdd was better tolerated than ECdd-P, with fewer hematological toxicities (grade 3/4) (21.4% and 47.9%, respectively).\nCollectively, these results indicate that PCdd might be an appropriate regimen for TNBC.\nPCdd not only is superior to ECdd-P as adjuvant chemotherapy with respect to improving the 3-year DFS and OS rates but also yields lower chemotherapy-related toxicities in early TNBC patients regardless of theBRCA mutation status.\nThus, PCdd might be a beneficial standard adjuvant regimen for early TNBC patients at a high recurrence risk, as indicated herein by the clinically meaningful improvement in survival and safety.\nTo our knowledge, this is an innovative randomized clinical study to evaluate the efficacy of a dose-dense carboplatin-based regimen in the adjuvant setting for TNBC with high recurrence risk.\nTNBC may be more sensitive to platinum-based regimens.\nCarboplatin increased the pCR rate from 41% to 54% in the CALGB40603 trial and from 36.9% to 53.2% in the GeparSixto trial.\nIn the GeparSixto study, the improved pCR rate significantly increased the 3-year DFS rate from 76.1% to 85.8% (HR=0.56; 95% CI, 0.33\u22120.96; P=0.024).\nHowever, in the CALGB40603 study, the 5-year distant recurrence-free interval was 76.3% with no significant difference.\nThe randomized phase III clinical trial EA 1131 (NCT02445391) has also been designed to prove the efficacy of adjuvant cisplatin or carboplatin following neoadjuvant chemotherapy in patients with residual TNBC.\nThe BrighTNess study has also confirmed that carboplatin-containing regimen appears to have a favorable risk-to-benefit profile for patients with high-risk TNBC in the neoadjuvant setting.\nHowever, the clinical benefit of adjuvant carboplatin in TNBC has not been well-investigated.\nFor an adjuvant scenario, a retrospective, single-center study in a Swiss breast cancer center reported a 5-year relapse-free survival (RFS) of 90% in patients treated with carboplatin.\nIn the present study, the PCdd regimen achieved significantly better survival benefit (3-year DFS and OS rates) for TNBC patients in the adjuvant setting compared with historical data from standard chemotherapy regimens (60%\u221280% with taxane-based regimens, 65%\u221285% with anthracycline- and taxane-based therapy, and 83.7% with anthracycline-based chemotherapy plus bevacizumab).\nA dose-dense regimen has been hypothesized to minimize residual tumor burden compared to dose escalation and serve as a more effective method for high-risk breast cancer.\nIn the CALGB9741 trial, the 4-year DFS rate was 82% in the dose-dense group.\nA previous study from our institution also compared the epirubicin and cyclophosphamide followed by paclitaxel (EC-P) or epirubicin plus paclitaxel (EP) dose-dense group and the EP regular group regarding postoperative adjuvant treatment for high-risk breast cancer.\nThe dose-dense group had higher 3-year RFS rates (84.1% vs. 80.0%, P=0.501) and OS rates (95.6% vs. 90.0%, P=0.153).\nOur trial is a novel prospective study showing significant improvements in the 3-year DFS and OS rates by using a dose-dense anthracycline-free platinum-based adjuvant chemotherapy regimen for TNBC regardless of the BRCA mutation status.\nThe 3-year DFS (93.9%) and OS (98.5%) rates in the PCdd arm were also superior to those of a dose-dense regimen reported by the Early Breast Cancer Trialists\u2019 Collaborative Group (EBCTCG).\nAlthough the survival data in our study are immature at present, a relatively long follow-up time will allow us to report a beneficial trend in OS.\nIn addition, these data are comparable to previous data on anthracycline- and taxane-based dose-dense regimens.\nBecause the TNBC phenotype is closely associated with hereditary breast cancer, the administration of platinum-based regimens has received a new impetus.\nHowever, in the Chinese population, BRCA1/2 mutations are prevalent in only 10.5% of TNBC patients younger than 50 years.\nThe benefit of adjuvant carboplatin in TNBC with BRCA1/2 mutation(s) is still controversial.\nThe GeparSixto trial showed that carboplatin is more effective in TNBC patients; however, a secondary analysis of the GeparSixto demonstrated that TNBC patients without BRCA1 and BRCA2 germline mutations would also benefit from the addition of carboplatin, which increased the DFS rate (85.3% in the carboplatin group and 73.5% in the non-carboplatin group; HR=0.53; 95% CI, 0.29\u22120.96; P=0.04).\nThe BRCA1/2 mutation status plays an important role for tumor identification in TNBC patients with higher response rate of platinum-based neoadjuvant therapy.\nHowever, other studies have shown that the clinical use of the homologous recombination deficiency (HRD) test may also have the potential to identify patients with TNBC that may respond to the treatment of DNA damage, in excess of those currently identified by gBRCA1/2 mutational screening.\nIt has been suggested that tumors carrying gBRCA mutations may be sensitive to DNA-damaging chemotherapeutic drugs, including platinum.\nIn the present study, we found that for early TNBC patients, the addition of carboplatin to paclitaxel was superior to epirubicin plus paclitaxel with respect to the 3-year DFS among BRCA1/2 unselected patients.\nTo analyze the trends in adjuvant regimens for TNBC and to explore the factors influencing efficacy, we demonstrated that patients aged >40 years, with Ki-67 index >30%, and clinically evaluated lymph nodes were found to have a survival advantage from the PCdd regimen.\nFuture refinement of platinum-sensitive subgroups for targeting specific tumor biomarkers in TNBC is warranted ().\nWith respect to tolerance, previous trials showed a high incidence of AEs and an increasing discontinuation rate for dose-dense chemotherapy of TNBC.\nThe PCdd regimen, which yields fewer adverse toxicities, may be considered a better alternative for the high-risk group of patients in our study, particularly for older patients.\nThe toxicity profile in our study was as anticipated: gastrointestinal toxic effects were more common in the PCdd arm, while grade 3/4 hematological toxicity was more common in the ECdd-P arm.\nAll gastrointestinal toxic effects were manageable and self-limiting.\nThese findings indicate that the PCdd regimen can be recommended to reduce unnecessary toxicities.\nOur study has some limitations, including its small sample size and the potential investigator bias from a single-center institutional experience.\nFurther, we had limited statistical power to show a significant OS benefit.\nA longer follow-up time is necessary, and the median OS should be further evaluated.\nIn addition, given the financial and technical limitations during the study period, the BRCA mutation status was not analyzed to identify whether the gBRCA subgroup will benefit from the PCdd regimen.\nFurther prospective trials to evaluate other platinum-based regimens in the adjuvant setting for TNBC are warranted, particularly to define a sensitive population.\nAn ongoing phase III trial in National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences (NCT03876886 at http://ClinicalTrials.gov) might provide further insight to evaluate the incorporation of platinum in the adjuvant setting, to detect HRD, and to identify specific TNBC patients who might benefit from carboplatin-based therapy.\nConclusions\nPCdd not only is superior to ECdd-P as adjuvant chemotherapy with respect to improving 3-year DFS and OS rates but also yields lower chemotherapy-related toxicities in early TNBC patients regardless of the BRCA mutation status.\nThus, PCdd might be a beneficial standard adjuvant regimen for early TNBC patients at a high recurrence risk, with clinically meaningful improvement in survival and safety data.\nFlow diagram of study design. ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nKaplan-Meier plot of disease-free survival (DFS). Cross marks indicate censored observations. Data for the intention-to-treat population. Hazard ratio (HR), 0.310, 95% confidence interval (95% CI), 0.137\u22120.704; Log-rank P=0.005; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nKaplan-Meier plot of overall survival (OS). Cross marks indicate censored observations. Data for the intention-to-treat population. Hazard ratio (HR), 0.142, 95% confidence interval (95% CI), 0.060\u22120.825, Log-rank P=0.028; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nSubgroup analyses of disease-free survival (DFS). The analyses of two arm patients were stratified for modified intention-to-treat population in clinically relevant subgroups. ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin; SCI, surgery-chemotherapy interval; HR, hazard ratio; 95% CI, 95% confidence interval.\n\nSynopsis of study protocol\nItem | Description\nER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor 2; ANC, absolute neutrophil count; AST, aspartate aminotransferase; ALT, alanine aminotransferase; AUC, area under the curve; 5-HT, 5-hydroxytryptamine; G-CSF, granulocyte colony-stimulating factor.\nStudy ID | CH-BC-012\nStudy title | Randomized phase III trial comparing dose-dense epirubicin and cyclophosphamide followed by paclitaxel with paclitaxel plus carboplatin as adjuvant therapy for triple-negative\u00a0breast cancer\nProtocol date | 4/20/2011\nTrial stage principal | Phase III\nInvestigator | Binghe Xu, M.D. & PhD. National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College. Email: xubinghe@medmail.com.cn;\nQing Li, B.S.Med. National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College. Email: cheryliqing@126.com\n\nParticipating study left | National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing 100021, China\nObjectives | To compare the efficacy and safety of dose-dense epirubicin and cyclophosphamide (ECdd) followed by paclitaxel (P) with dose-dense paclitaxel plus carboplatin (PCdd) as adjuvant therapy for patients with triple-negative breast cancer (TNBC) at high risk of recurrence\nPrimary objective:\n\u2022 Compare 3-year disease-free survival (DFS) of early TNBC patients at high risk treated with PCdd to those treated with ECdd-P regimens\nSecondary objectives:\n\u2022 Compare 3-year overall survival (OS) in the same population\n\u2022 Compare the toxicity of the PCdd to the ECdd-P in patients with TNBC at high risk of recurrence\n\nStudy population | Patients with early TNBC at high risk of recurrence\nStudy design | This is a single-left, open label, randomized, comparative phase III trial. The trial includes two groups: ECdd-P and PCdd.\nEligible participants will be randomly assigned in a 1:1 ratio to the PCdd group or the ECdd-P group. Randomization was conducted with no stratification factors. Eligible patients will be continually enrolled into the study until the total number of patients reached the planned sample size. The patients, medical staff and investigators were aware of treatment allocation. Sample size was determined based on a superiority test of 3-year DFS rate. To detect a difference of an approximate higher proportion of 0.10 between the two regimens (result of our preliminary clinical research demonstrated the proportion surviving in the ECdd-P regimen was 80.0%), an overall sample size of 133 subjects (66 in the ECdd-P arm and 67 in the PCdd arm) was calculated to achieve 80.0% power at a one-sided 0.050 significance level, with a 10% dropout rate (5% in each control/treatment arm). The accrual pattern across time periods was uniform (all periods equal). Primary and secondary efficacy analyses include the intent-to-treat (ITT) population of all randomly assigned patients. The safety analysis population includes all patients who received at least one dose of treatment.\n\nEligibility | Inclusion criteria: 1) Patient must accept the primary breast surgery; 2) Patients with histologically confirmed ER (\u2212), PR (\u2212) and HER2 (\u2212),i.e., <1% positive tumor cells with nuclear staining in IHC and no HER2 overexpression; 3) Positive axillary lymph nodes; negative axillary lymph node with age <35 years or III grade or intravascular cancer embolus; 4) Age between 18 years to 65 years; 5) Able to give informed consent; 6) Patients with an Eastern Cooperative Oncology Group (ECOG) performance score of 0 or 1; 7) Not pregnant, and on appropriate birth control if of child-bearing potential; 8) Adequate bone marrow reserve with ANC >1.5\u00d710 9/L and platelets >100\u00d710 9/L; 9) Adequate renal function with serum creatinine <2.0\u00d7 the upper limit of normal; 10) Adequate hepatic reserve with serum bilirubin <2.0\u00d7 the upper limit of normal, AST/ALT <2\u00d7 the upper limit of normal, and alkaline phosphatase < 5\u00d7 the upper limit of normal. Serum bilirubin >2.0 is acceptable in the setting of known Gilbert\u2019s syndrome; and 11) No active major medical or psychosocial problems that could be complicated by study participation.\nExclusion criteria: 1) Received neo-adjuvant therapy; 2) cardiac dysfunction documented by an ejection fraction less than the lower limit of the facility normal by multi-gated acquisition (MUGA) scan, or 45% by echocardiogram; 3) uncontrolled medical problems; 4) evidence of active acute or chronic infection; 5) pregnant or breast feeding; or 6) hepatic, renal or bone marrow dysfunction as detailed above.\n\nSample size calculation | The target sample size was calculated based on the primary endpoint, i.e., 3-year DFS rate. To detect a difference of 0.13 between the two regimens (result of our preliminary clinical research demonstrated the proportion surviving in the ECdd-P regimen was 80.0%), an overall sample size of 133 subjects (66 in the ECdd-P arm and 67 in the PCdd arm) was calculated to achieve 80.0% power at a one-sided 0.050 significance level. The accrual pattern across time periods was uniform (all periods equal). The proportion of drop out in the control and treatment group was 0.1000 (each 0.05).\nRandomization | Upon meeting the eligibility criteria, patients will be randomised under concealment, by the study lead investigator (Cancer Hospital, Chinese Academy of Medical Sciences), according to prespecified randomisation number lists to receive ECdd-P or PCdd.\nTreatment | Administration: Patients in both study groups received treatment in 14-day cycles. Patients assigned to the PCdd arm received paclitaxel 150 mg/m2 on d 1 plus carboplatin AUC=3 on d 2 for 8 cycles. Patients assigned to the ECdd-P arm received epirubicin 80 mg/m2 divided in 2 d and cyclophosphamide 600 mg/m2 on d 1 for 4 cycles followed by paclitaxel 175 mg/m2 on d 1 for 4 cycles. Prophylactic antiemetic measures, including 5-HT3 receptor antagonists, and dexamethasone, were allowed. Premedication with dexamethasone and histamine antagonists was administered before paclitaxel to prevent hypersensitivity reactions. Prophylactic G-CSF 3 \u00b5g/kg in d 5\u22129 was given for each chemotherapy cycle.\n\nSafety assessments and dose modifications | Safety assessments included 12-lead electrocardiograms, vital sign taking and clinical laboratory evaluations every cycle. Adverse events (AEs) were recorded at each treatment cycle until 28 follow-up d after the end of study visit. Toxicity was graded by using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 (NCI-CTCAE, version 3.0). Febrile neutropenia was managed according to institutional treatment guidelines in China. Toxicities were managed through dose delays of up to 3 weeks, and dose reductions were permitted in the following events: grade 4 hematological, grade 3 or 4 non-hematological, or other protocol-specified toxic effects.\nStudy drugs | Drug: epirubicin, cyclophosphamide, paclitaxel, carboplatin, G-CSF epirubicin 80 mg/m2 iv divide in 2 d cyclophosphamide 600 mg/m2 iv d 1 G-CSF 3 \u00b5g/kg in d 5\u22129 q14d \u00d74 cycles paclitaxel 175 mg/m2 iv d 1 G-CSF 3 \u00b5g/kg in d 5\u22129 q14d \u00d74 cycles paclitaxel 150 mg/m2 iv d 1 carboplatin AUC=3 iv d 2 G-CSF 3 \u00b5g/kg in d 5\u22129 q14d \u00d78 cycles.\n\nConcomitant medications | 1. Antiemetics can be prescribed to patients who are vomiting due to administration of treatment drug(s);\n2. Patients experiencing peripheral neuropathy can be treated with neurotropic supplements such as duloxetine, vitamin B, etc.;\n3. Analgesics can be used for patients who have pain affecting quality of life;\n4. Patients with constipation, diarrhea, or other conditions can be treated using appropriate medication for their respective condition;\n5. Prophylactic antiemetic measures, including 5-HT3 receptor antagonists, and dexamethasone, were allowed.\n6. Premedication with dexamethasone and histamine antagonists was administered before paclitaxel to prevent hypersensitivity reactions.\n\nOutcome measures | Primary outcome measure:\nThe primary endpoint is 3-year DFS rate. DFS was calculated from the date of randomization to the date of the first local/distant recurrence (without second primary malignancies), according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.\nSecondary outcome measures:\nSecondary endpoints include 3-year OS (defined as the time from randomization to death due to any cause) and safety of the treatment. Toxicity was graded by using the NCI- CTCAE, version 3.0.\n\nSafety parameters | AEs, vital signs and clinical laboratory tests\nStatistical analysis | All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) software (Version 22.0; IBM Corp., New York, USA). Data on clinical characteristics, chemotherapy, recurrence, and survival were analyzed. Data were presented as the number (%) or the mean standard deviation. Continuous variables were compared using the Student\u2019s t test, while categorical variables were compared using the \u03c72 or Fisher\u2019s exact test.\nThe proportion of patients remaining event-free over time will be displayed using the Kaplan-Meier method and analyzed using a two-sided log-rank test. All statistical tests were two-sided, and a P value of <0.05 was considered statistically significant.\nThe safety population will include all patients who received at least one dose of treatment. For safety analysis, AEs will be coded using the Medical Dictionary for Regulatory Activities (MedDRA). Analysis of AEs will be based on treatment-emergent adverse events (TEAEs). TEAEs are AEs not present prior to medical treatment, or are already present and worsen either in intensity or frequency following treatment. The incidence rate of TEAEs will be described according to system organ class (SOC) and preferred term (PT). Meanwhile, serious AEs (SAEs) and AEs leading to study discontinuation will be similarly summarized and tabulated. Laboratory tests will be analyzed using descriptive statistical analysis.\n\nFollow-up | All treated patients will be followed-up with once every 3 months to collect survival information for DFS and OS. Patients who discontinue treatment due to any causes will be followed-up with once every 3 months until disease recurrence or death. After disease recurrence, patient follow up can be conducted by phone or as general clinical visits until death.\n\n\nBaseline characteristics of patients with triple-negative breast cancer\nVariable | ECdd-P arm (N=73) [n (%)] | PCdd arm (N=70) [n (%)] | P\nECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; MRM, modified radical mastectomy; BCS, breast conservative surgery; SLN, simple mastectomy and sentinel lymph node biopsy; SCI, surgery chemotherapy interval.\nAge [mean (range)] (year) | 46 (26\u221264) | 49 (22\u221263) | 0.216\n\u3000\u226440 | 20 (27.4) | 12 (17.1) | 0.163\n\u3000>40 | 53 (72.6) | 58 (82.9) | \nMenopause at diagnosis |  |  | \n\u3000Post-menopause | 50 (68.5) | 39 (55.7) | 0.124\n\u3000Pre-menopause | 23 (31.5) | 31 (44.3) | \nPathology |  |  | 0.114\n\u3000IDC | 63 (86.3) | 67 (95.7) | \n\u3000ILC | 2 (2.7) | 0 (0) | \n\u3000Other type | 8 (11.0) | 3 (4.3) | \nTumor size (cm) |  |  | 0.179\n\u3000<2 | 27 (37.0) | 34 (48.6) | \n\u3000\u22652 | 46 (63.0) | 36 (51.4) | \nLymph node metastasis |  |  | 0.604\n\u3000Yes | 29 (39.7) | 24 (34.3) | \n\u3000No | 44 (60.3) | 46 (65.7) | \nIntravascular cancer embolus |  |  | 0.167\n\u3000Yes | 16 (21.9) | 10 (14.3) | \n\u3000No | 57 (78.1) | 60 (85.7) | \nNuclear grade |  |  | 0.999\n\u3000Grade 1, 2 | 23 (31.5) | 22 (31.4) | \n\u3000Grade 3 | 50 (68.5) | 48 (68.6) | \nKi-67 |  |  | 0.108\n\u3000\u226430 | 12 (16.4) | 20 (28.6) | \n\u3000>30 | 61 (83.6) | 50 (71.4) | \nTNM stage |  |  | 0.104\n\u3000I | 24 (32.9) | 27 (38.6) | \n\u3000II/III | 49 (67.1) | 43 (61.4) | \nType of surgery |  |  | 0.309\n\u3000MRM | 57 (78.1) | 54 (77.1) | \n\u3000BCS | 13 (17.8) | 9 (12.9) | \n\u3000SLN | 3 (4.1) | 7 (10.0) | \nRadiotherapy |  |  | 0.141\n\u3000Yes | 42 (57.5) | 33 (47.1) | \n\u3000No | 31 (42.5) | 37 (52.9) | \nSCI (d) |  |  | 0.609\n\u3000<30 | 47 (64.4) | 42 (60.0) | \n\u3000\u226530 | 26 (35.6) | 28 (40.0) | \n\n\nTreatment exposure in TNBC patients treated with ECdd-P/PCdd chemotherapy\nVariables | n (%) | P\nECdd-P Arm (N=73) | PCdd Arm (N=70)\nTNBC, triple-negative breast cancer; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin.\nFollow-up time [Median (range)] (month) | 56.1 (2.8\u221298.6) | 58.1 (1.2\u221276.6) | 0.320\nNumber of chemotherapy cycles |  |  | \n\u3000Total | 573 | 552 | \n\u3000Median | 8 (3\u22128) | 8 (2\u22128) | 0.783\nDelay of chemotherapy |  |  | 0.361\n\u3000Yes | 23 (31.5) | 25 (35.7) | \n\u3000No | 50 (68.5) | 45 (64.3) | \nDose reduction |  |  | 0.369\n\u3000Yes | 14 (19.2) | 16 (22.9) | \n\u3000No | 59 (80.8) | 54 (77.1) | \n\n\nCommon adverse events in TNBC patients treated with ECdd-P/PCdd chemotherapy\nAdverse events | n (%) | P*\nECdd-P arm (n=73) | PCdd arm (n=70)\nA patient could have experienced more than one specific toxicity. TNBC, triple-negative breast cancer; ECdd-P, dose-dense epirubicin and cyclophosphamide followed by paclitaxel; PCdd, dose-dense paclitaxel plus carboplatin; AST, aspartate aminotransferase; ALT, alanine aminotransferase; TBIL, total bilirubin; CRE, creatinine; *, P values for differences in two arms are tested by \u03c72 test or Fisher exact test.\n\nHematologic toxicities | Grade 1/2 | Grade 3/4 | Grade 1/2 | Grade 3/4 | Grade 3/4\n\u3000Anemia | 28 (38.4) | 0 (0) | 14 (20.0) | 0 (0) | \u2212\n\u3000Leukopenia | 39 (53.4) | 26 (35.6) | 39 (55.7) | 12 (17.1) | 0.010\n\u3000Neutropenia | 30 (41.1) | 35 (47.9) | 31 (44.3) | 15 (21.4) | 0.001\n\u3000Thrombocytopenia | 8 (11.0) | 0 (0) | 9 (12.9) | 2 (2.9) | 0.238\nNon-hematologic toxicities | Grade 1/2 | Grade 3/4 | Grade 1/2 | Grade 3/4 | Grade 3/4\n\u3000Alopecia | 36 (49.3) | 8 (11.0) | 32 (45.7) | 4 (5.7) | 0.204\n\u3000Stomatitis | 38 (52.1) | 0 (0) | 29 (41.4) | 0 (0) | \u2212\n\u3000Nausea emesis | 65 (89.0) | 0 (0) | 56 (80.0) | 1 (1.4) | 0.490\n\u3000Diarrhea | 5 (6.8) | 1 (1.4) | 1 (1.4) | 0 (0) | 0.490\n\u3000Mucositis/cutaneous | 3 (4.1) | 1 (1.4) | 1 (1.4) | 0 (0) | 0.490\n\u3000Peripheral neuropathy | 28 (38.4) | 1 (1.4) | 31 (44.3) | 4 (5.7) | 0.170\n\u3000Foot and hand syndrome | 6 (8.2) | 0 (0) | 1 (1.4) | 0 (0) | \u2212\n\u3000Myalgia/arthralgia | 12 (16.4) | 1 (1.4) | 11 (15.7) | 0 (0) | 0.490\n\u3000Asthenia | 8 (11.0) | 1 (1.4) | 6 (8.6) | 0 (0) | 0.490\n\u3000Allergic | 1 (1.4) | 0 (0) | 3 (4.3) | 0 (0) | \u2212\n\u3000Cardiac toxicity | 3 (4.1) | 0 (0) | 2 (2.9) | 0 (0) | \u2212\n\u3000ALT elevation | 25 (34.2) | 3 (4.1) | 19 (27.1) | 1 (1.4) | 0.326\n\u3000AST elevation | 30 (41.1) | 0 (0) | 26 (37.1) | 0 (0) | \u2212\n\u3000TBIL elevation | 29 (39.7) | 0 (0) | 26 (37.1) | 0 (0) | \u2212\n\u3000CRE elevation | 3 (4.1) | 0 (0) | 7 (10.0) | 0 (0) | \u2212\n",
        "label": "unclear",
        "id": "task4_RLD_test_9"
    },
    {
        "paper_doi": "10.1186/1475-2875-12-363",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Study design: cRCT\n\nUnit of allocation: village (paired on the basis of geographical location)\n\nNumber of units: 10:10\n\nLength of follow-up: 1 year\n\nOutcome assessment: passive case detection at local health centre. Children were followed up in 2 postintervention cross-sectional surveys: after 5 months and at the end of the study (10 months).\n\nAdjustment: cluster adjustment using intraclass correlation coefficient of 0.048, between-cluster variation 0.006, and within-cluster variation 0.12\n\n\nParticipants: Number of participants: 8395\n\nInclusion criteria: children under 10 years of age\n\n\nInterventions: Intervention: bed net (n = 4066)\n\nInsecticide and dosage: deltamethrin (25 mg/m2)Retreatment: not stated\n\nUsage: not stated\n\nControl: no net (n = 4109)\n\n\nOutcomes: Outcomes measured:Malaria prevalenceMalaria incidenceNumber of deathsNumber of cases of severe malaria\n\n\nNotes: Study location: Rakhine State, Western Myanmar, in 2 areas: Dabhine and MyothugyiEIR: not statedMalaria transmission: predominantly low with pockets of intense transmissionMain vectors: not stated% P vivax cases: 52%, and 2% P falciparum/P vivax mixe\n\n",
        "objective": "The primary objective of this review was to assess the impact of ITNs on mortality and malaria morbidity, incorporating any evidence published since the previous update into new and existing analyses, and assessing the certainty of the resulting evidence using GRADE.",
        "full_paper": "Background\nInsecticide-treated bed nets (ITN) reduce malaria morbidity and mortality consistently in Africa, but their benefits have been less consistent in Asia.\nThis study\u2019s objective was to evaluate the malaria protective efficacy of village-wide usage of ITN in Western Myanmar and estimate the cost-effectiveness of ITN compared with extending early diagnosis and treatment services.\nMethods\nA cluster-randomized controlled trial was conducted in Rakhine State to assess the efficacy of ITNs in preventing malaria and anaemia in children and their secondary effects on nutrition and development.\nThe data were aggregated for each village to obtain cluster-level infection rates.\nIn total 8,175 children under 10\u00a0years of age were followed up for 10\u00a0months, which included the main malaria transmission period.\nThe incidence and prevalence of Plasmodium falciparum and Plasmodium vivax infections, and the biting behaviour of Anopheles mosquitoes in the area were studied concurrently.\nThe trial data along with costs for current recommended treatment practices were modelled to estimate the cost-effectiveness of ITNs compared with, or in addition to extending the coverage of early diagnosis and treatment services.\nResults\nIn aggregate, malaria infections, spleen rates, haemoglobin concentrations, and weight for height, did not differ significantly during the study period between villages with and without ITNs, with a weighted mean difference of \u22122.6 P. falciparum episodes per 1,000\u00a0weeks at risk (95% Confidence Interval \u22127 to 1.8).\nIn areas with a higher incidence of malaria there was some evidence ITN protective efficacy.\nThe economic analysis indicated that, despite the uncertainty and variability in their protective efficacy in the different study sites, ITN could still be cost-effective, but not if they displaced funding for early diagnosis and effective treatment which is substantially more cost-effective.\nConclusion\nIn Western Myanmar deployment of ITNs did not provide consistent protection against malaria in children living in malaria endemic villages.\nEarly diagnosis and effective treatment is a more cost effective malaria control strategy than deployment of ITNs in this area where the main vector bites early in the evening, often before people are protected by an ITN.\nBackground\nMalaria is a major cause of morbidity and mortality in Myanmar.\nMalaria control activities in this country have been concentrated on early, mostly clinical, diagnosis and treatment.\nLimited availability of curative services in remote areas, the difficulties in accessing malarious areas, and the relatively high costs of effective treatment of multi-drug resistant malaria, have compromized malaria control efforts.\nEffective malaria control activities are needed, and there is a recent substantial increase in donor support for these activities.\nLevels of chloroquine and sulphadoxine-pyrimethamine resistance in Plasmodium falciparum have been high in this region for decades.\nSubsidies for highly effective artemisinin combination treatment (ACT) have considerably increased their availability in recent years.\nUnfortunately artemisinin monotherapies are still widely available providing continued selective pressure on emerging artemisinin resistance in the east of the country.\nRegular and systematic indoor spraying of residual insecticides was stopped in 1993 and is now only used for special situations (outbreaks, new settlements).\nLargely ignored by the outside world until 2007, since then there has been an over fifty-fold increase in external donor funding for malaria control in Myanmar, the majority of which has been spent on insecticide-treated mosquito nets (ITN).\nITNs are an appealing approach to the control of malaria.\nThey repel and kill malaria vectors, and they prevent sporozoite-bearing anopheline mosquitoes from biting the occupants lying under or near the bed net.\nITNs have been shown to exert a mass effect reducing vector populations in villages with extensive use and high transmission.\nITNs are usually greatly appreciated by their occupants, because they prevent all kinds of insect bites, and they are very safe.\nOverall use of ITNs prevented approximately one in five childhood deaths in sub-Saharan Africa.\nThese substantial benefits have rightly led to the inclusion of ITNs as a central component of many malaria control initiatives throughout the world.\nIn endemic areas of South and South-East Asia, where malaria transmission is lower and unstable, mobile young adult males are particularly affected, and vector behaviour is different, the results of ITN efficacy studies are mixed, and much less clear than in Africa.\nThis is not because of insecticide resistance, although some vector species do exhibit low-grade pyrethroid resistance, but because of the behaviour of the vectors which often bite early in the evening or morning outdoors or away from dwellings, and the corresponding behaviour of humans.\nITN always provide some protection against malaria \u2013 the critical question is how much?\nAll individuals in malaria endemic areas should ideally have ITN, even if they provide only modest benefit, but whilst there are financial constraints, choices need to be made.\nIt is necessary therefore to assess the efficacy of ITN in the South and South-East Asian region at a local level to assist decisions on their wider-scale deployment in relation to alternative proven highly effective malaria control measures (i.e. early diagnosis and effective treatment: EDAET).\nThe effectiveness of ITN in reducing malaria morbidity and mortality depends on several factors, including the level of malaria endemicity and the behaviour and immunity of the population, the climate, the acceptance and usage of the ITN by the population and, crucially, the biting behaviour of the main anopheline vectors.\nA cluster-randomized controlled trial was carried out in Western Myanmar, to evaluate the protective efficacy of village-wide usage of insecticide-impregnated ITNs on malaria incidence and prevalence, anaemia, and the development of children.\nThe anopheline vector abundance and biting behaviour were also studied, and data were collected on population sleeping habits, to assess the potential of ITNs to reduce man-mosquito contact in this region; these findings are described in the accompanying paper.\nWhile effective diagnosis and treatment facilities were an ethical and research necessity in the study setting, these are often not present in routine settings.\nBudgetary limitations may result in ITN programmes competing with EDAET services for limited resources.\nA cost-effectiveness analysis was therefore carried out to assess the costs and benefits of ITN when compared with those of EDAET when neither has been implemented, and to assess the incremental cost-effectiveness of introducing one of the interventions when the other is already in place.\nMethods\nStudy area and population\nThe study was conducted in Rakhine State, Western Myanmar.\nThe monsoon season is from May to October and yearly rainfall is very high (+/\u2212 5000\u00a0mm/year).\nMalaria transmission occurs throughout the year and peaks during the post monsoon (November-January), and sometimes in the early monsoon (June-July) periods (Figure\u00a01).\nThe transmission intensity varies from predominantly low to pockets of intense transmission, depending on the location, and there are considerable variations over short geographic distances and also between years and between seasons.\nSymptomatic malaria occurs at all ages but is seen predominantly in children.\nAccess to effective treatment has historically been very poor.\nOutbreaks of falciparum malaria can spread over several townships, or may be limited to a single village at other times.\nIn 1994, M\u00e9d\u00e9cins Sans Fronti\u00e8res - Holland (MSF-H) began a malaria control programme in Rakhine State in cooperation with the Myanmar VBDC (Vector Borne Disease Control) department.\nThe programme focussed on early diagnosis and treatment, through support of 30 fixed field clinics with laboratories and riverboat \u2018mobile malaria clinics\u2019.\nPlasmodium falciparum was responsible for approximately 80% of the malaria infections in patients who presented to these clinics.\nDrug resistance in P. falciparum in the area was studied in 1995; high levels of resistance were found with treatment failure rates of 82% to chloroquine and 67% to sulphadoxine-pyrimethamine, whereas mefloquine was very effective (93% cure rate).\nSince 1996 all patients with falciparum malaria have been treated with a combination of mefloquine and artesunate.\nBetween 1997 and 2007 over one million patients (approximately 43% of whom were children under 10\u00a0years of age) received treatment for slide confirmed malaria, supported by this programme.\nIt was unclear whether ITN should be deployed as a priority so the effectiveness of ITN was studied between May 1998 and February 1999, covering the peak malaria transmission season, in villages around two Rural Health Clinics (RHC), in Dabhine and Myothugyi, two village-tracts located in the townships Sittwe and Maungdaw, which are approximately 90\u00a0km apart.\nThe study villages in the Dabhine area are all within 5 miles from the Bay of Bengal.\nThis is a coastal plain area without hills or forest, where rice and other crops are cultivated.\nCommon breeding sites for Anophelines are paddy fields, brackish water in streams and tide pools and saline ponds used for prawn culture which are close to villages, swamps, and small ponds used for growing watercress.\nThe area was classified generally as meso- to hyperendemic for malaria.\nA malaria blood-slide survey, performed in September 1995, found a prevalence of 34% for P. falciparum and 11% for Plasmodium vivax, and a spleen prevalence of 36% among primary school children.\nMyothugyi is a densely populated area situated 5\u20137 miles from the Bay of Bengal.\nThe area is characterized by rice-fields and partly forested hills.\nPotential breeding sites for Anophelines are freshwater creeks, ponds and stagnant water in rice fields, as well as brackish water in tide pools, prawn-breeding ponds and small pools and streams at foothills.\nIt has been classified generally as a low meso-endemic malaria area.\nA malaria blood-slide survey in January 1996, 10 kilometres south of Myothugyi, found a prevalence of 13% for P. falciparum and 3% for P. vivax and none of the primary schoolchildren examined had palpable splenomegaly.\nThese surveys, and the substantial burden of malaria in children, prompted discussions with the community and authorities as to what control measures could be applied.\nIn view of the uncertainties over ITN efficacy and their limited availability at the time, and after consultation with the authorities and village leaders, it was decided jointly that an evaluation was needed to set priorities before consideration of wide-scale deployment.\nIt was agreed that if there was any evidence of benefit, then after the evaluation, ITNs would be given to all villages included in the evaluation.\nStudy design, randomization and sample size calculation\nIdeally, a study population is randomized on an individual or household basis.\nHowever, the use of insecticide-impregnated bed nets (ITN) has an impact on the vector populations in the nearby environment and, therefore, on malaria transmission and risk, not only in the household itself, but also among people living nearby, including those who do not have ITN (the \u201cmass\u201d effect).\nTherefore, the unit chosen for randomization in this study was the village, and this was agreed with the village leaders in consultation with the community.\nThe primary outcome measure was the incidence and prevalence of malaria in children.\nIn December 1997 and January 1998 a population survey was done in the study areas.\nOutreach Workers (ORW) were trained to register all families, number houses, and to collect demographic data on family size and number of children under 10\u00a0years of age.\nDuring this period, a sample of the population of children under 10 was checked for parasitaemia and splenomegaly.\nThe results of the pre-intervention survey were used to determine the number of clusters and the sample sizes needed for the larger study.\nThe pre-intervention malaria survey was performed on 1,088 children (Table\u00a01).\nThe intra-class correlation coefficient (ICC) calculated on the basis of the pre-intervention data was 0.048, with a between-cluster variation of 0.006 and within-cluster variation of 0.12.\nThe prevalence of P. falciparum parasitaemia by microscopy averaged 17% (range 0 \u2013 53%).\nMicroscopy was quality assured throughout the study with independent cross-checking.\nThe number of village-clusters and their size needed for a valid comparative study was calculated using the methods described by Thompson.\nIn order to detect a 50% reduction in the incidence of falciparum malaria, with 80% power at a 5% significance level, a total of 10 village pairs were required with an average of 420 children per village, amounting to a total sample size of 8,400 children.\nInitially, 22 villages were informed about the study-procedures and were invited to participate, but after further discussions two villages declined to join.\nFinally 20 villages (clusters) were paired, and for each pair one was selected randomly (using a computer generated random number) to receive impregnated bed nets (ITNs), while the other acted as the control village.\nMatching was done according to geographical location so that the intervention and control villages were nearby (villages were 1\u20132 miles apart).\nThis minimized differences in the environment in and around each village pair (i.e. proximity to foothills, forest fringe, prawn-ponds, rice-fields, waterways), which could have had a strong influence on vector populations and consequently on the prevalence and incidence of malaria.\nThe travel-time to the health clinic and its influence on the number of patient visits, and therefore on malaria incidence, were comparable for intervention and control villages, as the distances were matched equally.\nITN distribution\nBetween the 26th and 29th of April 1998, all households in the intervention villages received a number of ITNs, according to the number of family members and were instructed explicitly about the correct use of the nets (Figure\u00a02).\nMore than 5000 ITN were distributed.\nThe ITN (green colour, polyester, size 130 \u00d7 180 \u00d7 150 cm (11.6\u00a0m2) or 190 \u00d7 180 \u00d7 150 cm (14.5\u00a0m2), Siam-Dutch Co, Thailand) were already impregnated with deltamethrin (25\u00a0mg/m2) by the manufacturer.\nAll households of the control villages received ITNs after the study was completed in May 1999.\nIncidence surveillance\nMalaria morbidity surveillance consisted of passive case detection.\nEach family from both study groups received a specific registration card at the beginning of the study and anyone with complaints of fever was urged to visit the RHC and bring the registration card.\nIn both study areas a RHC operated by staff from the Department of Health, was supported by doctors, microscopists, and out-reach workers from MSF-H. Microscopy was quality checked on a regular basis.\nThe RHCs were open seven days a week.\nOut-reach workers regularly visited the villages and informed the population about using the clinics in case of illness.\nIn the RHC blood-smears of patients with fever or a history of fever were examined for malaria parasites.\nData-files were kept for all children under 10\u00a0years coming from the villages under study.\nPatients with falciparum malaria were treated with a single dose of mefloquine 15\u00a0mg base/kg and artesunate 4\u00a0mg/kg (then the current treatment).\nVivax malaria was treated with chloroquine (10\u00a0mg base/kg on day 0 and 1, 5\u00a0mg/kg on day 2), followed by primaquine (0.25\u00a0mg base/kg/day for 14\u00a0days).\nPatients who were not part of the study were also provided with medical services.\nNo other significant health services were available in the areas under study.\nCross-sectional surveys\nAfter distribution of ITN, children were followed up for weight, height and haemoglobin at three time-points: at the beginning of the study (May 11 - June 10, 1998), after 5\u00a0months (14 Sept- 9 Oct. 1998) and at the end of the study (25 Jan - 22 Feb 1999) (Figure\u00a02).\nDuring the last survey a blood-smear for the detection of malarial parasitaemia was taken as well.\nDuring the surveys people were asked about the usage of ITN and whether they had washed their ITN.\nEthics statement\nThe study was discussed in detail with community representatives of each village and with officials from the Ministry of Health.\nThere was general agreement on the need for an evaluation but no local ethical review committee or relevant formal organizational ethics review process was available before the study started in 1998.\nThe reasons for the study were discussed and it was explained that individual or community decision to participate or not would not in any way jeopardize anti-malarial screening, treatment, or subsequent deployment of ITN.\nVillages were included in the study only after extensive discussions and full approval of the village representatives.\nMany study participants were illiterate, and for these individuals fully informed witnessed verbal consent was obtained from adults and the parents of children involved in the study.\nThe Myanmar National Health authorities granted approval for this study.\nStatistical analysis\nMalariometric data\nThis study was designed as a cluster randomized trial and, therefore, the data were aggregated for each village (i.e. cluster) to obtain cluster-level infection rates.\nThis approach is appealing because the cluster is both the unit of randomization and the unit of analysis.\nThe villages with bed nets were then compared to the other villages using a weighted paired t test.\nThe results of blood smears, collected at the clinics, were used to calculate the incidence density of P. falciparum, defined as the number of malaria episodes per 1,000 child-weeks at risk.\nThe number of weeks at risk for each child was calculated from the start of the study till the end of the study (the last malaria prevalence survey) or the final date they were seen before they moved or died.\nIf children moved in the first half of the study (before the 2nd prevalence survey), the weeks at risk could be not calculated and these children were, therefore, excluded from the analysis of malaria episodes per 1,000 child weeks at risk.\nIf children moved in the second half of the study, the weeks at risk were taken from the first half of the study.\nChildren who had falciparum malaria were treated with mefloquine and artesunate and were considered not at risk of reinfection because of post-treatment prophylaxis for four weeks following treatment.\nThree categorical variables were also calculated, in which each child was classified to have one or more infections with either P. falciparum, P. vivax or \u201call species of malaria (i.e. any malaria)\u201d between 11 May 1998 and 27 February 1999.\nThis was analysed on an intention-to-treat basis.\nData from the cross-sectional surveys were also aggregated to obtain cluster-level prevalences of malaria, anaemia, and malnutrition.\nCost-effectiveness analysis\nFour scenarios were modelled for the costs and benefits of malaria control using ITN and/or EDAET as compared with a baseline of no intervention.\nThe study data on incidence of P. falciparum cases in the individual villages were used, with Poisson distributions assigned to each of these independently, to capture the uncertainty and variability in these findings.\nIncidence data for vivax cases per village were not available and the proportion of children with at least one case of vivax over the study period was, therefore, adapted as a proxy for incidence (likely to be an under-estimate of actual incidence), and converted to the rate per 1,000 person weeks.\nThe costs of diagnosis, treatment and ITN were adapted to present day (2013) recommended strategies and their current costs (i.e. rapid tests for the diagnosis of malaria prior to treatment, the current recommended dosage of artemether-lumefantrine for falciparum malaria, continued use of chloroquine with primaquine for vivax malaria, and the use of long-lasting insecticide treated nets (LLIN) instead of ITN).\nA cost for village malaria workers was also included, that would be required for the EDAET strategy at $1.05 per child per annum (adapted from a study of VMWs in Cambodia).\nThe costs of treatment are based on the currently recommended dose regimen of 3\u00a0days of artemether-lumefantrine (current first line policy); the provider unit costs in Myanmar in 2013 for a full treatment course for children weighing between 5-14\u00a0kg was $0.47 and for children weighing 15-24\u00a0kg it was $0.94.\nThe cost of one chloroquine tablet was $0.021 and for primaquine it was $0.015.\nThe cost of RDTs was estimated at $0.8.\nAn extra 15% was added to the costs of diagnostics and treatment for shipping and wastage.\nThe cost of LLIN including delivery was assumed to be $10 based on a recent review of LLIN costs which conformed with estimates from donors currently active in Myanmar.\nIt was assumed that each ITN is shared by 1.8 people on average (while ITN can be shared by two children a family of five will receive three nets) and that they have a useful life of three years.\nThese costs were applied for each 1,000 person-weeks.\nClinical cases of falciparum malaria were converted to disability-adjusted life-years (DALYs) assuming a mortality rate of 0.1% in treated cases and 1% in untreated P falciparum cases, and a mortality rate of 0.01% and 0.05% in vivax malaria respectively.\nModelling also assumed a one-week total duration of illness for uncomplicated treated malaria and four weeks for untreated malaria.\nResults were generated by running 10,000 Monte Carlo simulations sampling from the probability distributions and calculating the mean costs and effects for each strategy.\nCost-effectiveness acceptability curves were generated to summarize the parameter uncertainties and identify which strategy was most likely to be cost-effective across a range of willingness to pay thresholds.\nResults\nCharacteristics of study groups\nDuring December 1997 and January 1998 data were collected from 1,088 children (13.3% of the total study population) on the prevalence of malaria and palpable spleens.\nThere was considerable variability in the proportions of children with P. falciparum and/or P. vivax parasitaemia, and palpable spleens among the clusters (Table\u00a01).\nIn May 1998, immediately after the ITN were distributed (but before ITN could have had an impact on malaria related pathology), base-line demographic and clinical data of the study children were gathered.\nA total of 8,175 children were recruited from twenty clusters.\nThe age and sex distributions of children were similar for the intervention (4,066 children) and control villages (4,109 children), as were the variables \u201cweight for height\u201d, and the proportions of children with anaemia (Hgb\u2009<\u200910.0\u00a0g/dL).\nIncidence of malaria\nDuring the study-period (10 May 1998 to 28 February 1999) [42\u00a0weeks], 2,408 children visited the clinics with complaints of fever and were confirmed by microscopy to have malaria; 1,102 P. falciparum (46%), 1,257 P. vivax (52%), and 49 mixed i.e. with both species (2%).\nOf the children infected with P. falciparum who visited the clinics, 15.5% came twice with a P. falciparum infection (32 from ITN clusters and 117 children from NN clusters), 1.8% came three times (two from ITN cluster and 15 from NN clusters), and three children (NN cluster) came four times with a P. falciparum infection during the 10-month follow up period.\nThe number of P. falciparum episodes per 1,000 child-weeks at risk was calculated for each cluster (Table\u00a02, Figure\u00a03).\nThere was considerable heterogeneity in the effects.\nEight out of the ten ITN villages showed some protective efficacy against falciparum malaria compared to their control village while the remaining two village pairs showed the opposite trend.\nThe study design (by cluster) provides a limited number of data points so further investigation by stratifying clusters by transmission and other characteristics was not feasible.\nThe protective effect of ITN appeared to increase in both relative and absolute terms with the incidence of malaria in the control village in each cluster pair (Figure\u00a03).\nHowever the weighted difference between the intervention and control villages for the incidence density of P. falciparum was not statistically significant (weighted mean difference of \u22122.6 episodes per 1,000\u00a0weeks at risk, (95% Confidence Interval; -7 to 1.8); p\u2009=\u20090.22).\nTwelve children (1% of acute falciparum malaria) developed severe falciparum malaria; in Dabhine, six severe patients were reported from the NN-villages and two from the ITN-villages.\nIn Myothugyi, one severe patient was reported from the NN-villages and three from the ITN-villages.\nThus, overall seven children developed severe malaria in NN villages and five in ITN villages.\nSeven out of 10 village pairs showed a protective efficacy of ITN for both one or more falciparum malaria episodes and one or more vivax malaria episodes while the remaining three village pairs showed the opposite trend.\nThe overall proportion of children having had one or more malaria episode was 16.3% in ITN villages and 22.5% in NN villages, a weighted difference of \u22126.4% (95% CI; -21.2% to 8.4%), p\u2009=\u20090.35 (Table\u00a03, Figure\u00a04).\nFor P. falciparum this was 8.4% (341/4066) in ITN villages and 15.2% (618 /4109) in NN control villages respectively, a weighted difference of \u22126.9% (95% CI; \u201318.9% to 5.1%), p\u2009=\u20090.23, and for P. vivax, 9.9% (401/4066) in ITN villages versus 12.3% (505/4109) in NN villages, a weighted difference of \u22122.6% (95% CI; -10.9% to 5.7%), p\u2009=\u20090.25 (Table\u00a03).\nAge-groups\nThere was no significant trend for falciparum malaria-infection by age.\nRestricting the analysis to children under five years of age, similar results were observed compared to the whole study group.\nVivax malaria was more common among younger children.\nRelapse of vivax malaria cannot usually be distinguished reliably from incident infection, although first infections of life are by definition incident.\nAmong children under one year of age, 143 (17,5%) had at least one vivax infection during the 10\u00a0month study period (13.7% in ITN villages and 21.0% in NN villages) and children in their second year of age had 185 (19.4%) infections (18.1% in ITN villages and 20.7% in NN villages).\nAfter the second year the proportion of children with an episode of vivax malaria decreased steadily by age and only 3.7% of 9-year-old children (3.8% in ITN villages and 3.7% in NN villages) had an episode of vivax malaria in the same period.\nCross-sectional surveys\nOf the 8175 children registered, 7764 (95%) were present at all three \u2018prevalence\u2019 surveys.\nOf the other 411 children, 46 children reportedly died during the study-period (overall child mortality 5.62% in the 42\u00a0weeks), (20 from NN villages and 26 from ITN villages; p\u2009>\u20090.2) and 365 had moved out of the area or were absent during one or two of the follow-up surveys (NN 142, ITN 223).\nMalaria prevalence\nThe prevalences of children with malaria and splenomegaly were measured at the end of the study (January \u2013 early February 1999), which coincided with the end of the peak transmission season.\nA total of 7,828 children were present at this survey.\nThere was considerable variation in malaria prevalence, both for P. falciparum and P. vivax, and also in the splenomegaly prevalences among the villages (Table\u00a04, and in further detail in Additional files 1, 2, 3 and 4).\nThe overall prevalence of P. falciparum was 5.7% (225) in NN villages and 4.2% (161) in ITN villages.\nSome protective efficacy from impregnated bed nets was seen in six village pairs, while four village pairs showed the opposite effect; overall difference - 1.9% (95% CI; -5.8% to 2.0%), p\u2009=\u20090.30 (Additional files 5, 6 and 7).\nFor P. vivax the overall prevalence was 14.2% (564) in NN and 11.6% (446) in ITN villages.\nFour village pairs suggested protection by ITN while five did not, and in one village pair, both clusters, had no vivax malaria at all; overall difference\u2009=\u2009\u22122.8% (95% CI; - 8.5%, 2.9%); p\u2009=\u20090.30.\nFor splenomegaly prevalence, six village pairs suggested protection by ITN, while four did not; overall difference\u2009=\u2009\u2212 7.3% (95% CI; - 20.1%, 5.5%), p\u2009=\u20090.23.\nThe overall splenomegaly prevalence was 14% in bed net villages and 21% in non-bed net villages.\nAnaemia and malnutrition\nThe prevalences of anaemia (Hb <10\u00a0g/dL) and malnutrition (Wt/Ht; Z-score < \u22122) at baseline in the ITN and NN villages were similar.\nThe weighted mean difference in the prevalence of anaemia was \u22125.2% (ITN-NN; 95% CI: -13.8%, 3.4%) and the weighted mean difference in the prevalence of malnutrition was 0% (ITN-NN; 95% CI: -4%, +4%).\nThe mean haemoglobin level of children during the first cross-sectional survey was lower in Myothugyi (8.85\u00a0g/dL; village means varying from 7.84 to 9.24\u00a0g/dL) than in Dabhine (9.70\u00a0g/dL; village means varying from 9.15 to 10.02\u00a0g/dL).\nThe proportion of anaemic children (Hb\u2009<\u200910\u00a0g/dL) decreased during the study period in all but two clusters (1 NN and 1 ITN village).\nAt the final cross-sectional survey, the decrease in the proportion of children with anaemia in bed net villages was more pronounced than in control villages in 7 cluster pairs while 3 pairs showed the opposite trend (Table\u00a05).\nThe prevalences of anaemia at the end of the study were not significantly different comparing NN and ITN clusters (weighted mean difference of - 3.3% (95% CI; - 11.3, 4.7), p\u2009=\u20090.38 (Additional file 8).\nThe proportion of children with moderate acute malnutrition (Wt/Ht; Z-score\u2009<\u2009\u22122) decreased over the study period in 14 of the 20 villages; 8 ITN villages and 6 NN villages (Table\u00a05).\nAt the final nutrition survey the decrease of malnutrition in ITN villages was greater than in control villages in seven cluster pairs, while three pairs showed a greater decrease in control villages.\nThe malnutrition prevalences at the end of the study were not significantly different comparing NN and ITN clusters, either for all children under 10\u00a0years (\u2212 1.7% (95% CI; - 3.8, 0.4), p\u2009=\u20090.094), or for the children under five years of age (\u2212 1.8% (95% CI; - 5.9, 2.3), p\u2009=\u20090.35) (Additional file 9).\nCost-effectiveness analysis of ITN and EDAET\nThe average weight of a child under 10\u00a0years of age presenting to the malaria programme in Rakhine State was 12.7\u00a0kg, which puts the current cost of the drugs for the treatment of falciparum malaria for an average person in this study area at $0.62 and for vivax malaria at approximately $0.3.\nThe model output suggests that in the absence of ITN or EDAET, 1.4 DALYs are accumulated per 1,000 person-weeks at no cost to the provider.\nThe use of ITN would add a cost of $35 and reduce the number of DALYs to 0.69, or an incremental cost-effectiveness ratio (ICER) of $51/DALY averted.\nThe use of EDAET alone would cost $23 per 1000 person-weeks and reduce DALYs to 0.19, or an ICER of $19 per DALY averted (i.e. nearly three times less).\nWhile both options are considered cost-effective, the implementation of ITN instead of EDAET would incur higher costs and avert less than a third of the number of DALYs.\nWhere EDAET is already in place the introduction of ITN would avert an additional 0.2 DALYs at a cost of $148 per DALY averted.\nThe cost-effectiveness acceptability curves (Figure\u00a05) illustrate these results, and show that if policy makers are willing to pay over approximately $280 per DALY averted, the combination of both ITN and EDAET is likely to be cost-effective.\nIf resources are more constrained then EDAET alone will always dominate the use of ITN alone.\nThis analysis derives from the data in this study and, therefore, specifically to children in this area of Western Myanmar.\nEDAET is likely to be equally effective in adults, but ITNs are likely to be less effective in the adult population who are often outside dwellings in the early evening, and go to sleep later and leave the home earlier than children.\nThus the comparative economic advantage of EDAET over ITN is likely to be even greater in adults and, therefore, the overall advantage of EDAET over ITN for the population overall is predicted to be greater than reported here.\nThis analysis assumes no impact of EDAET on incidence, although a comparison of the baseline and post-intervention surveys in the control villages suggests that EDAET might itself be reducing transmission.\nThis assumption does not impact on the incremental cost/gain of either of the strategies compared with each other, but it could potentially underestimate the relative advantage of EDAET as compared with a baseline of doing nothing, while overestimating that of ITN.\nDiscussion\nThe results of this study suggest that ITNs provided some protection for children against malaria in most villages in this area of Rakhine State, but they did not in the others, and the overall result was that there was no significant benefit evident from their deployment.\nMalaria episodes were less common overall in villages with ITN than in the control group, because differences in favour of ITN were greater than differences in favour of no nets (Figures\u00a03 and 4), but when these data were aggregated per cluster, systematic use of ITN was not found to reduce either falciparum or vivax malaria significantly.\nImportantly they also did not attenuate the adverse consequences of malaria on anaemia or growth.\nThis study focussed on children, as they are the most vulnerable age group for malaria in this area.\nThis provided the highest estimate of potential protective efficacy and thus benefit from ITN, as children go to bed earlier and they sleep longer than adults (particularly relevant in this area with early evening outdoor biting vectors), and they are less likely to travel far from the house.\nAs all villagers in the ITN clusters were provided with a net, any positive outcome could have been further enhanced through a mass insecticidal effect on the malaria vectors.\nIn these villages malaria affected children more than adults but malaria in the South East Asian region is often a disease particularly affecting young mobile adults who work in the forest and do not tend to use nets, so any benefits demonstrated here are likely to be greater than in the total population at risk in the region.\nAdherence with instructions for correct use of ITN was generally good, but as the benefits were small, incorrect use did not lead to more malaria.\nEven though the ITN and control clusters were selected randomly, the malaria prevalence before the study was higher in the study villages than in the control villages.\nThis might indicate a higher endemicity, which could have influenced the incidence and final prevalence results of the ITN clusters, but transmission intensity varies greatly per year and the differences were small and unlikely to make a material difference to the results.\nThis study was tightly controlled, with a high coverage and user rate compared to the normal context of use.\nDespite this no significant overall benefit from ITN deployment in terms of malaria protection could be demonstrated.\nThe proportion of children with anaemia decreased both in villages provided with ITNs and in the control villages over the duration of the study.\nThis is probably largely because diagnosis and effective treatment of malaria were provided in all villages, and because children with anaemia were treated promptly, which emphasizes the benefit provided by prompt and effective anti-malarial treatment in such areas of low unstable transmission.\nThe most likely explanation for this rather disappointing result from ITN deployment in this malaria endemic area of Western Myanmar is the early evening biting pattern and strong preference for outdoor biting of most malaria vectors in this area, as in many other areas of South-East Asia.\nFocussing on the human bite catches between July 1998 and January 2000 (n\u2009=\u20092,895), the overall peak biting time was between 6\u00a0pm and 7\u00a0pm, and over half of the anophelines (51%) were caught before 8\u00a0pm.\nThis would differentiate any benefits of ITNs between children, who are often in or near an ITN at this time, and adults who often are not.\nAs the behaviour of malaria vectors differs so much, particularly in Asia, entomological information is essential before ITNs are deployed.\nThese somewhat negative results contrast with the widespread positive perception of the uniform value of ITNs in malaria control.\nA systematic review of ITN evaluations, which included 22 studies from Africa, Asia and South America, concluded that ITNs reduce childhood morbidity and mortality and that ITNs should be employed in all malarious areas.\nOver the past decade ITNs have been taken up enthusiastically as a key component of many malaria control programmes, but caution is needed in interpreting the findings from bed net studies with different designs and analytical methods.\nAccording to Lengeler, of 29 identified studies conducted in Asia, only four studies used correct statistical procedures for their trials and were included in the meta-analysis.\nAnother study did use correct statistical procedures, but was not included because it studied only pregnant women.\nOf these five \u201cstatistically correct\u201d studies, three showed reduced morbidity while the other two did not.\nIn other studies, non-comparable control groups were used or ITN were allocated at a village-level, whereas end-points were calculated for individuals.\nThere is no doubt concerning the consistent and large benefits provided by ITNs in Africa which fully justifies current deployment initiatives.\nIn contrast, the epidemiology of malaria in Asia is extremely heterogeneous.\nTransmission is highly seasonal and unstable and intensities vary greatly over short distances, and also show great variation from year to year.\nIt is not surprising that ITNs are relatively ineffective in areas of unstable malaria where the principal malaria vectors bite outdoors early in the evening, before people go in or near their ITN.\nThis and previous studies argue against generalising from ITN studies in Africa to the rest of the malaria affected world.\nWhere the main malaria vectors bite mainly indoors after adults and children have gone to sleep, ITNs should be very effective, but in many places, this is not the case.\nThis is not to say that ITNs provide no benefit in these places\u2013 they do protect against malaria, albeit sometimes to a small extent, and if cost were no obstacle then everyone in malaria endemic areas should certainly be provided with an ITN.\nBut funds for malaria control are usually limited and ITN deployment is often included uncritically in many malaria control efforts in Asia in preference to other control measures.\nEarly diagnosis and effective treatment reduces malaria morbidity and mortality, but in a large proportion of patients in south and Southeast Asia, the diagnosis of malaria is still made clinically and, therefore, incorrectly.\nA lack of resources means that most patients with malaria in Myanmar and in adjacent countries under similar constraints still receive inadequate treatment.\nMany patients take a few tablets of artesunate until they feel better.\nSometimes chloroquine is still used which is inexpensive but ineffective.\nArtesunate-based combination treatment is a proven highly effective treatment, which significantly reduces malaria transmission in low-transmission areas.\nWithout accounting for these transmission blocking properties, the present economic model indicates that effective diagnostic and treatment (EDAET) services are less costly and substantially more effective than ITN, when comparing each of these to a baseline of no intervention, which is still the reality in some areas in Myanmar.\nOnce EDAET is implemented effectively, ITNs can still provide additional benefit.\nDuring the extended interval since this study was carried out there has been a consistent and substantial decline in malaria incidence across the region, including many areas in Myanmar even when allowing for improved reporting rates.\nWhile the Rakhine state still accounts for approximately half the burden in Myanmar, the incidence rates reported in this study are likely to be higher than those that would be documented today in this region.\nThis could further strengthen the argument in favour of EDAET.\nLower incidence of malaria will imply that protective measures which require the same costs applied to the entire population will be relatively less cost-effective than those that target infected individuals such as diagnosis and treatment.\nITNs are very popular with international donor agencies.\nMillions of dollars are spent on ITN programmes in this populous malaria endemic region.\nITNs are both provided free by donor organizations or promoted through social marketing, encouraging the population to purchase their own ITN.\nIn this malaria endemic region of Myanmar, the economic modelling based on the study data and contemporary prices suggests that early-diagnosis and effective treatment is substantially more cost-effective than deployment of LLINs as a malaria control measure.\nThis argues for careful evaluation in each region before large-scale deployment.\nThe findings here provide some evidence of a protective efficacy of ITNs in areas of higher endemicity within this region, suggesting that they can be a cost-effective intervention in the context of improving child health.\nIf, however, donor funding is targeting specifically the control of malaria and ultimately its elimination from the region it seems likely that improving preventive, diagnostic and curative interventions for all ages and particularly in adult, often migrant males, could offer better returns on investment.\nConclusion\nMalaria infections, palpable splenomegaly, haemoglobin concentrations, and weight for height, did not differ significantly during the study period between villages with and without ITNs.\nThe limited efficacy of ITNs may be explained by the biting behaviour (peak biting time between 1800 and 1900\u00a0hours, mainly outdoors) of the most common Anopheles mosquito vectors.\nGiven the lack of significant efficacy and relatively high costs of ITNs, the first priority in implementation of malaria control interventions in this area should be the provision of effective diagnosis and treatment.\nWhere EDAET services are already in place and sufficient budgets are available then the use of ITN can be cost-effective.\nClimate and seasonal malaria incidence in Rakhine State in 1998\u20131999. Upper. Monthly rainfall and minimum/maximum temperature, average of Sittwe and Maungdaw Townships (data obtained from township weather stations). Lower. Monthly number of malaria-patients visiting 6 malaria clinics in the region (not only at the study sites).\nTime frame of the study of insecticide treated mosquito nets and entomological surveys  between 1995 and 2000 in the study areas.\nIncidence rate ratio of P.falciparum for ITN versus control (NN) villages by incidence rate (per 1000\u00a0weeks) in the control villages.\nRisk difference in malaria (ever) in ITN villages by proportion compared with malaria in control (NN) villages.\nCost-effectiveness acceptability curves for the four options, accounting for the uncertainty due to the variability in the different study sites and varying levels of willingness to pay per DALY averted.\n\nPlasmodium falciparum, Plasmodium vivax and splenomegaly prevalences in ITN and control villages during the pre-study malaria survey (December 1997)\n\u00a0 | Intervention villages | Control villages\n\u00a0 | N | M.S. | Pf (%) | Pv (%) | Spleen (%) | N | M.S. | Pf (%) | Pv (%) | Spleen (%)\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 707 | 125 | 37 (30) | 35 (28) | 79 (63) | 796 | 102 | 38 (37) | 33 (32) | 68 (67)\nPair 2 | 551 | 89 | 26 (29) | 33 (37) | 47 (53) | 679 | 48 | 9 (19) | 7 (15) | 16 (34)\nPair 3 | 462 | 37 | 8 (22) | 10 (27) | 16 (42) | 298 | 49 | 11 (22) | 6 (12) | 30 (61)\nPair 4 | 223 | 33 | 7 (21) | 8 (24) | 9 (28) | 184 | 32 | 4 (13) | 16 (50) | 19 (59)\nPair 5 | 74 | 18 | 3 (17) | 4 (22) | 5 (28) | 87 | 15 | 8 (53) | 1 (7) | 6 (40)\nSubtotal | 2017 | 302 | 81 (27) | 90 (30) | 156 (52) | 2044 | 246 | 70 (28) | 63 (26) | 139 (57)\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 78 | 16 | 3 (19) | 1 (6) | 6 (38) | 47 | 15 | 3 (20) | 0 (0) | 1 (7)\nPair 7 | 714 | 96 | 1 (1) | 0 (0) | 15 (16) | 819 | 105 | 0 (0) | 0 (0) | 30 (29)\nPair 8 | 492 | 54 | 0 (0) | 1 (2) | 13 (24) | 602 | 70 | 2 (3) | 1 (1) | 6 (9)\nPair 9 | 659 | 57 | 4 (7) | 4 (7) | 7 (12) | 650 | 82 | 1 (1) | 0 (0) | 1 (1)\nPair 10 | 153 | 25 | 10 (40) | 2 (8) | 8 (32) | 120 | 20 | 7 (35) | 1 (5) | 4 (20)\nSubtotal | 2096 | 248 | 18 (7) | 8 (3) | 49 (19) | 2238 | 292 | 13 (4) | 2 (1) | 42(14)\nTotal | 4113 | 550 | 99 (18) | 98 (18) | 205 (37) | 4282 | 538 | 83 (15) | 65 (12) | 181 (34)\n\nN\u2009=\u2009number of children\u2009<\u200910\u00a0yrs (data from local authorities).\nMS\u2009=\u2009number of children of whom a malaria smear was taken.\nPf\u2009=\u2009P. falciparum prevalence (including mixed infections), Pv\u2009=\u2009P. vivax prevalence (including mixed infections).\n\nFalciparum malaria incidence per 1,000\u00a0weeks exposure in ITN and control villages (from 10 May 1998 to 28 February 1999)\n\u00a0 | ITN villages | Control villages | ITN Protective efficacy\n\u00a0 | N | Pf episodes per 1,000\u00a0weeks exposure | N | Pf episodes per 1,000\u00a0weeks exposure | Rate ratio | Rate difference\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 677 | 4.04 | (97 /24034) | 715 | 15.25 | (376 /24662) | 0.26 | \u221211.21\nPair 2 | 522 | 7.37 | (137 /18592) | 653 | 3.51 | (83 /23625) | 2.09 | +3.86\nPair 3 | 429 | 1.57 | (24 /15304) | 280 | 9.38 | (91 /9697) | 0.17 | \u22127.81\nPair 4 | 208 | 2.83 | (21 /7411) | 171 | 8.03 | (48 /5980) | 0.35 | \u22125.20\nPair 5 | 67 | 2.13 | (5 /2343) | 71 | 13.32 | (33 /2477) | 0.16 | \u221211.19\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 74 | 3.68 | (9 /2447) | 44 | 3.84 | (6 /1562) | 0.96 | \u22120.16\nPair 7 | 737 | 0.23 | (6 /26632) | 786 | 0.46 | (13 /28427) | 0.50 | \u22120.23\nPair 8 | 482 | 0.06 | (1 /17344) | 595 | 0.97 | (21 /21637) | 0.06 | \u22120.91\nPair 9 | 653 | 1.61 | (37 /22973) | 631 | 0.57 | (13 /22852) | 2.82 | +1.04\nPair 10 | 140 | 2.90 | (14 /4824) | 107 | 7.68 | (29 /3775) | 0.38 | \u22124.78\nTotal | 3,989 | 2.47 | (351 /141903) | 4053 | 4.93 | (713 /144694) | weighted paired t-test p\u2009=\u20090.216\n\n\nProportion of children (%) with one or more malaria infections during the study\n\u00a0 | \u00a0 | \u00a0 | P. falciparum- ever | P. vivax - ever | Any malaria \u2013 ever\n\u00a0 | ITN (N) | Control (N) | ITN | NN | Risk ratio | Risk difference | ITN | NN | Risk ratio | Risk difference | ITN | NN | Risk ratio | Risk difference\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 695 | 740 | 89 | 305 | 0.31 | - 28.4 | 151 | 243 | 0.66 | - 11.2 | 214 | 429 | 0.53 | - 27.2\n\u00a0 | \u00a0 | \u00a0 | (12.8) | (41.2) | \u00a0 | \u00a0 | (21.7) | (32.9) | \u00a0 | \u00a0 | (30.8) | (58.0) | \u00a0 | \u00a0\nPair 2 | 527 | 662 | 126 | 80 | 1.98 | + 11.8 | 120 | 72 | 2.09 | + 11.9 | 215 | 139 | 1.94 | + 19.8\n\u00a0 | \u00a0 | \u00a0 | (23.9) | (12.1) | \u00a0 | \u00a0 | (22.8) | (10.9) | \u00a0 | \u00a0 | (40.8) | (21.0) | \u00a0 | \u00a0\nPair 3 | 441 | 285 | 26 | 81 | 0.21 | - 22.5 | 46 | 64 | 0.43 | - 13.8 | 67 | 120 | 0.36 | - 26.9\n\u00a0 | \u00a0 | \u00a0 | (5.9) | (28.4) | \u00a0 | \u00a0 | (10.4) | (22.5) | \u00a0 | \u00a0 | (15.2) | (42.1) | \u00a0 | \u00a0\nPair 4 | 213 | 176 | 24 | 46 | 0.43 | - 14.8 | 32 | 60 | 0.44 | - 19.1 | 53 | 89 | 0.49 | - 25.7\n\u00a0 | \u00a0 | \u00a0 | (11.3) | (26.1) | \u00a0 | \u00a0 | (15.0) | (34.1) | \u00a0 | \u00a0 | (24.9) | (50.6) | \u00a0 | \u00a0\nPair 5 | 67 | 72 | 5 | 29 | 0.19 | - 32.8 | 6 | 33 | 0.20 | - 36.8 | 11 | 49 | 0.24 | - 51.7\n\u00a0 | \u00a0 | \u00a0 | (7.5) | (40.3) | \u00a0 | \u00a0 | (9.0) | (45.8) | \u00a0 | \u00a0 | (16.4) | (68.1) | \u00a0 | \u00a0\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 77 | 44 | 8 | 4 | 1.14 | + 1.3 | 6 | 0 | 3.35 | + 5.4 | 12 | 4 | 1.71 | + 6.5\n\u00a0 | \u00a0 | \u00a0 | (10.4) | (9.1) | \u00a0 | \u00a0 | (7.7) | (0) | \u00a0 | \u00a0 | (15.6) | (9.1) | \u00a0 | \u00a0\nPair 7 | 745 | 792 | 6 | 13 | 0.05 | - 0.8 | 1 | 4 | 0.20 | - 0.4 | 7 | 16 | 0.45 | - 1.1\n\u00a0 | \u00a0 | \u00a0 | (0.8) | (1.6) | \u00a0 | \u00a0 | (0.1) | (0.5) | \u00a0 | \u00a0 | (0.9) | (2.0) | \u00a0 | \u00a0\nPair 8 | 489 | 598 | 1 | 23 | 0.05 | - 3.6 | 6 | 13 | 0.56 | - 1.0 | 7 | 34 | 0.25 | - 4.3\n\u00a0 | \u00a0 | \u00a0 | (0.2) | (3.8) | \u00a0 | \u00a0 | (1.2) | (2.2) | \u00a0 | \u00a0 | (1.4) | (5.7) | \u00a0 | \u00a0\nPair 9 | 664 | 632 | 38 | 13 | 2.71 | + 3.6 | 28 | 6 | 4.20 | + 3.2 | 59 | 17 | 3.30 | + 6.2\n\u00a0 | \u00a0 | \u00a0 | (5.7) | (2.1) | \u00a0 | \u00a0 | (4.2) | (1.0) | \u00a0 | \u00a0 | (8.9) | (2.7) | \u00a0 | \u00a0\nPair 10 | 148 | 108 | 18 | 24 | 0.55 | - 10.0 | 5 | 4 | 0.92 | - 0.3 | 19 | 26 | 0.53 | - 11.3\n\u00a0 | \u00a0 | \u00a0 | (12.2) | (22.2) | \u00a0 | \u00a0 | (3.4) | (3.7) | \u00a0 | \u00a0 | (12.8) | (24.1) | \u00a0 | \u00a0\nTotal | 4066 | 4109 | 341 | 618 | \u00a0 | \u00a0 | 401 | 505 | \u00a0 | \u00a0 | 664 | 923 | \u00a0 | \u00a0\n\u00a0 | \u00a0 | \u00a0 | (8.4) | (15.0) | \u00a0 | \u00a0 | (9.9) | (12.3) | \u00a0 | \u00a0 | (16.3) | (22.5) | \u00a0 | \u00a0\nWeighted paired t-test | \u00a0 | \u00a0 | p\u2009=\u20090.23 | p\u2009=\u20090.25 | p\u2009=\u20090.35\n\nP. falciparum-ever includes P. falciparum\u2009+\u2009mixed, P. vivax-ever includes P. vivax\u2009+\u2009mixed. ITN: insecticide treated nets, NN: no nets.\n\nPlasmodium falciparum, Plasmodium vivax and splenomegaly prevalences (%) in ITN and control clusters during the peak season at the end of the study (January 1999)\n\u00a0 | N | N | P. falciparum | P. vivax | Spleen\n\u00a0 | ITN | NN | ITN | NN | Prevalence difference | ITN | NN | Prevalence difference | ITN | NN | Prevalence difference\nDabhine | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 1 | 659 | 708 | 31 (5) | 107 (15) | - 10.4 | 183 (28) | 280 (40) | - 11.8 | 170 (26) | 432 (61) | - 35.2\nPair 2 | 508 | 644 | 39 (8) | 35 (5) | + 2.3 | 120 (24) | 135 (21) | + 2.7 | 155 (31) | 131 (20) | + 10.2\nPair 3 | 413 | 268 | 17 (4) | 21 (8) | - 3.7 | 57 (14) | 83 (31) | - 17.2 | 63 (15) | 90 (34) | - 18.3\nPair 4 | 203 | 169 | 10 (5) | 6 (4) | + 1.4 | 23 (11) | 25 (15) | - 3.5 | 56 (28) | 74 (44) | - 16.2\nPair 5 | 62 | 71 | 3 (5) | 7 (10) | - 5.0 | 23 (37) | 22 (31) | + 7.1 | 23 (38) | 31 (44) | - 5.9\nMyothugyi | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPair 6 | 65 | 43 | 14 (22) | 3 (7) | + 14.6 | 7 (11) | 1 (2) | + 8.4 | 11 (17) | 4 (9) | + 7.6\nPair 7 | 717 | 763 | 2 (0) | 5 (1) | - 0.4 | 0 (0) | 0 (0) | 0 | 6 (1) | 23 (3) | - 2.2\nPair 8 | 462 | 580 | 0 (0) | 6 (1) | - 1.0 | 1 (0) | 4 (1) | - 0.7 | 1 (0) | 9 (2) | - 1.3\nPair 9 | 627 | 620 | 5 (1) | 3 (0) | + 0.3 | 9 (1) | 2 (0) | + 1.1 | 11 (2) | 3 (0) | + 1.3\nPair 10 | 143 | 103 | 40 (28) | 32 (31) | - 3.1 | 23 (16) | 12 (12) | + 4.4 | 37 (26) | 23 (22) | + 3.5\nTotal | 3859 | 3969 | 161 (4) | 225 (6) | \u00a0 | 446 (12) | 564 (14) | \u00a0 | 533 (14) | 818 (21) | \u00a0\nWeighted Paired t-test | \u00a0 | \u00a0 | Pf : -1.9% (\u22125.8%, 2.0%), | Pv : -2.8% (\u22128.5%, 2.9%), | Spleen : -7.3% (\u221220.1%, 5.5%),\np-value\u2009=\u20090.30 | p-value\u2009=\u20090.30 | p-value\u2009=\u20090.23\n\nP. falciparum% and P. vivax% indicate the percentage of children with P. falciparum or P. vivax parasitaemia, both values include mixed infections. ITN: insecticide treated nets, NN: no nets.\n\nPrevalence of anaemia and malnutrition during the 1st and 3rd cross-sectional surveys\nMalnutrition (Wt/Ht < \u22122Z1)\n\u00a0 | ITN clusters | NN clusters\n1st prevalence | 3rd prevalence | Change | 1st prevalence | 3rd prevalence | Change | Relative change\nPair 1 | 87/692 (13) | 67/651 (10) | \u22122.3% | 98/728 (13) | 76/701 (11) | \u22122.6% | \u22120.3%\nPair 2 | 61/516 (12) | 58/506 (11) | \u22120.4% | 86/651 (13) | 102/634 (16) | +2.9% | 3.3%\nPair 3 | 72/435 (17) | 71/413 (17) | +0.6% | 51/277 (18) | 43/266 (16) | \u22122.5% | \u22123.1%\nPair 4 | 62/210 (30) | 28/201 (14) | \u221215.6% | 38/169 (22) | 32/167 (19) | \u22123.3% | 12.3%\nPair 5 | 17/64 (27) | 6/61 (10) | \u221216.7% | 11/71 (15) | 6/70 (9) | \u22126.9% | 9.8%\nPair 6 | 8/77 (10) | 2/65 (3) | \u22127.3% | 3/42 (7) | 5/42 (12) | +4.8% | 12.1%\nPair 7 | 105/735 (14) | 92/701 (13) | \u22121.2% | 158/780 (20) | 92/760 (12) | \u22128.2% | \u22127.0%\nPair 8 | 60/486 (12) | 56/450 (12) | +0.01% | 67/585 (11) | 81/572 (14) | +2.7% | 2.7%\nPair 9 | 87/652 (13) | 55/620 (9) | \u22124.5% | 64/627 (10) | 78/612 (13) | +2.5% | 7.0%\nPair 10 | 39/148 (26) | 8/140 (6) | \u221220.6% | 10/104 (10) | 7/100 (7) | \u22122.6% | 18.0%\nTotal | 598/4015 (15) | 443/3808 (12) | 586/4034 (15) | 586/4034 (15) | 522/3924 (13) | \u22121.2% | 2.1%\nAnaemia (Hb <10\u00a0g/dL)\n\u00a0 | ITN clusters | NN clusters\n\u00a0 | 1st prevalence | 3rd prevalence | Change | 1st prevalence | 3rd prevalence | Change | Relative change\nPair 1 | 362/695 (52) | 299/659 (45) | \u22126.7% | 440/739 (60) | 413/708 (58) | \u22121.2% | 5.5%\nPair 2 | 303/527 (57) | 235/508 (46) | \u221211.2% | 332/662 (50) | 305/644 (47) | \u22122.8% | 8.4%\nPair 3 | 228/441 (52) | 182/413 (44) | \u22127.6% | 152/285 (53) | 125/268 (47) | \u22126.7% | 0.9%\nPair 4 | 133/213 (62) | 88/203 (43) | \u221219.1% | 110/176 (63) | 59/169 (35) | \u221227.6% | - 8.5%\nPair 5 | 46/67 (69) | 22/62 (35) | \u221233.2% | 29/72 (40) | 33/71 (46) | +6.2% | 39.4%\nPair 6 | 76/77 (99) | 52/65 (80) | \u221218.7% | 42/44 (95) | 40/43 (93) | \u22122.4% | 16.3%\nPair 7 | 485/745 (65) | 602/716 (84) | +19.0% | 657/792 (83) | 558/761 (73) | \u22129.6% | \u221228.6%\nPair 8 | 355/489 (73) | 307/461 (67) | \u22126.0% | 532/597 (89) | 473/579 (82) | \u22127.4% | \u22121.4%\nPair 9 | 579/664 (87) | 443/627 (71) | \u221216.5% | 540/632 (85) | 446/619 (72) | \u221213.4% | 3.1%\nPair 10 | 142/148 (96) | 108/143 (76) | \u221220.4% | 104/108 (96) | 95/103 (92) | \u22124.1% | 16.3%\nTotal | 2709/4066 (67) | 2338/3857 (61) | \u22126.0% | 2938/4107 (72) | 2547/3965 (64) | \u22127.3% | \u22121.3%\n\n1 Weighted paired t test 3rd prevalence Wt/Ht - 2 Z score, p-value\u2009=\u20090.094. Relative change\u2009=\u2009(Prev 3 - Prev 1) NN cluster - (Prev 3 - Prev 1) ITN cluster. ITN: insecticide treated nets, NN: no nets.",
        "label": "low",
        "id": "task4_RLD_test_895"
    },
    {
        "paper_doi": "10.1371/journal.pone.0009696",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Parallel arm cluster-RCT conducted in Mirzapur, Bangladesh, between Dec 2003 and Dec 2006.\n\n\nParticipants: Sample size: 12 clusters (21,140 individuals randomised, 10,700 women with at least 1 pregnancy during 10 preceding months analysed).Clusters: rural unions surrounding an urban central union (excluded from the study) served by a 750 bed private referral-level hospital.Individuals: all married women of reproductive age (i.e. 15-49 years) in the intervention arm were eligible for enrolment. Women in the survey were eligible if they had had a pregnancy outcome in the last 3 years.\n\n\nInterventions: Target: health system (addition of home visits).Arm 1: 2 home visits (12-16 and 32-34 weeks); they were given a labour card for women to present upon arrival at hospital for delivery and 3 postnatal visits on days 2, 5 and 8. CHWs facilitated free-of-charge transfer of ill neonates to hospital.The purpose of the antenatal component of the intervention was to increase uptake of ANC (3 visits taking place at home or at a health centre or satellite clinic - distinct from the 2 antenatal CHW home visits), tetanus toxoid vaccination, general pregnancy and newborn care education, and birth preparedness (including delivery at a health facility).Arm 2: standard ANC.\n\n\nOutcomes: Trial primary outcomes: antenatal and immediate newborn care behaviours, knowledge of danger signs, care seeking for neonatal complications, and neonatal mortality.Review outcomes reported:Primary: not reported.Secondary: ANC coverage (at least 1 visit), health facility delivery, IPT for malaria, neonatal mortality.\n\nFollow-up: data collection at delivery and during pre and postnatal home visits. Endline survey Jan - May 2006, before the end of the trial.\n\n\nNotes: Funders: The Wellcome Trust: Burroughs Wellcome Fund Infectious Disease Initiative 2000 and the Office of Health, Infectious Diseases and Nutrition, Global Health Bureau, United States Agency for International Development (USAID) through the Global Research Activity Cooperative agreement with the Johns Hopkins Bloomberg School of Public Health (award HRN-A-00-96-90006-00). Support for data analysis and manuscript preparation was provided by the Saving Newborn Lives program through a grant by the Bill & Melinda Gates Foundation to Save the Children-US. The study was registered at clinicaltrials.gov, No. NCT00198627\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Background\nTo evaluate a delivery strategy for newborn interventions in rural Bangladesh.\nMethods\nA cluster-randomized controlled trial was conducted in Mirzapur, Bangladesh.\nTwelve unions were randomized to intervention or comparison arm.\nAll women of reproductive age were eligible to participate.\nIn the intervention arm, community health workers identified pregnant women; made two antenatal home visits to promote birth and newborn care preparedness; made four postnatal home visits to negotiate preventive care practices and to assess newborns for illness; and referred sick neonates to a hospital and facilitated compliance.\nPrimary outcome measures were antenatal and immediate newborn care behaviours, knowledge of danger signs, care seeking for neonatal complications, and neonatal mortality.\nFindings\nA total of 4616 and 5241 live births were recorded from 9987 and 11153 participants in the intervention and comparison arm, respectively.\nHigh coverage of antenatal (91% visited twice) and postnatal (69% visited on days 0 or 1) home visitations was achieved.\nIndicators of care practices and knowledge of maternal and neonatal danger signs improved.\nAdjusted mortality hazard ratio in the intervention arm, compared to the comparison arm, was 1.02 (95% CI: 0.80\u20131.30) at baseline and 0.87 (95% CI: 0.68\u20131.12) at endline.\nPrimary causes of death were birth asphyxia (49%) and prematurity (26%).\nNo adverse events associated with interventions were reported.\nConclusion\nLack of evidence for mortality impact despite high program coverage and quality assurance of implementation, and improvements in targeted newborn care practices suggests the intervention did not adequately address risk factors for mortality.\nThe level and cause-structure of neonatal mortality in the local population must be considered in developing interventions.\nPrograms must ensure skilled care during childbirth, including management of birth asphyxia and prematurity, and curative postnatal care during the first two days of life, in addition to essential newborn care and infection prevention and management.\nTrial Registration\nClinicaltrials.gov NCT00198627\nIntroduction\nNeonatal mortality declined by approximately 20% over the last decade in Bangladesh, however, the rate of decline was less than in the postneonatal and 1\u20134 year-old periods.\nNeonatal deaths now account for almost half of under-5 child deaths in Bangladesh and efforts to reduce neonatal mortality are crucial to achieving Millennium Development Goal 4 for child survival.\nSince 90% of births and most neonatal deaths still occur at home, community-level interventions must be introduced while linking with the healthcare system for treatment of life-threatening newborn illness.\nSeveral recent community-based trials of packages of maternal and neonatal interventions in low resource settings in South Asia have shown statistically significant reductions in neonatal mortality, employing a variety of healthcare delivery approaches.\nThe focus of the interventions, however, has been primarily on averting deaths due to serious infections.\nHome-based health education and routine neonatal assessment and antibiotic treatment of serious infections by community health workers (CHWs) decreased mortality in rural India and rural northeastern Bangladesh, although regulatory approval for and availability of CHWs for home-based treatment of illness is lacking in most settings.\nA preventive maternal and neonatal care behavior change management program implemented by CHWs through home visits as well as community mobilization also reported a mortality reduction of 54% in a very high mortality area of Uttar Pradesh, India.\nLady health workers in Hala, Pakistan, promoted essential maternal and newborn care through home visits, community education group sessions, and linkages with local traditional birth attendants (TBAs), resulting in a 28% reduction in mortality.\nStudies without home-based interventions also reported mortality reductions of about 30% through community-based participatory interventions in Nepal and by improving TBAs' clean delivery practices and strengthening their linkages with primary health facilities in Larkana, Pakistan.\nTo provide cost-effective essential preventive and curative services in low-resource settings, strategies must take into account the risk factors for and causes of mortality, the quality and accessibility of the health care system, and community perception and acceptance of the interventions.\nCommunity-based preventive care coupled with basic management of childhood illness and facilitated referral by CHWs is a potentially effective model where access to quality health care at facilities can be ensured.\nWe developed a preventive service delivery strategy in a rural area of central Bangladesh with good access to facility-based care to promote household newborn care practices through home visits by CHWs, and conducted routine, home-based illness surveillance coupled with facilitated referral of sick newborns to health facilities.\nA cluster-randomized controlled trial was conducted to examine its impact on knowledge and practice of newborn care and neonatal mortality.\nMethods\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nStudy Population and Design\nProjahnmo-Mirzapur was a cluster-randomized, controlled intervention trial of a preventive and curative maternal-neonatal healthcare package, in which was nested surveillance for community-acquired neonatal bacteremia..\nThe trial was implemented in Mirzapur, a sub-district of Tangail district, Dhaka division, Bangladesh, located 2 hours by car from the capital city of Dhaka, during January 2004\u2013 December 2006.\nThe neonatal mortality rate (NMR) was estimated at 24 per 1000 live births in 2002.\nThe area was served by Kumudini Hospital \u2013 a 750-bed, private, referral-level hospital, located in a central urban union which was excluded from the study.\nThe remaining population of about 292,000 was divided into 12 rural unions, which were randomly allocated to either comparison or intervention arm using a computer-generated pseudo-random number sequence without stratification or matching (Figure 1).\nBlinding was unachievable given the nature of the intervention.\nNewborns in the comparison arm received the usual health services provided by the government, non-governmental organizations and private providers.\nIn the intervention arm, each union had six CHW areas, each of which consisted of approximately 4000 population served by one CHW.\nThe CHW-to-population ratio was similar to the primary healthcare worker-to-population ratio in the Bangladesh government health system, thus facilitating sustainability and scalability of the healthcare delivery strategy.\nAll married women of reproductive age (i.e., 15\u201349 years) in the intervention arm were eligible for enrolment, and were administered informed verbal consent by the CHW in their area.\nDesign and Implementation of Interventions\nCommunity-level interventions were developed based on findings from formative research on newborn care practices in the study population, conducted during November 2002\u2013 April 2003.\nInformation on pregnancy, delivery, immediate newborn care and care seeking for newborn illness was collected through 26 unstructured interviews with women, husbands, mothers-in-law, and TBAs, and through semi-structured, in-depth interviews with 40 women and/or family members and 54 healthcare providers, including TBAs, health workers of BRAC or village associations, and village doctors.\nFindings from formative research were used to design the communications and negotiation approach to promote safe and clean delivery and preventive, household newborn care practices (Table 1).\nCHWs were trained for 36 days on pregnancy surveillance, counseling and negotiation skills, essential newborn care, neonatal illness surveillance and management of illness based on a clinical algorithm adapted from Integrated Management of Childhood Illness.\nAfter initial training and evaluation, routine monitoring and refresher training were provided each fortnight.\nFurther information on recruitment, characteristics, training and monitoring of CHWs is presented elsewhere.\nIn addition, TBAs serving in the intervention unions (n\u200a=\u200a84) attended a two-day orientation session on the aims and activities of the project, essential newborn care practices, and indications for referral of newborns and mothers.\nTable 1 presents detailed information on interventions provided by CHWs in the intervention arm.\nCHWs identified pregnancies in their population through bimonthly household pregnancy surveillance.\nBirth and newborn care preparedness (BNCP) was promoted by CHWs through two antenatal home visits scheduled at 12\u201316 and 32\u201334 weeks of gestation.\nCHWs gave a labor notification card to each woman with instructions for a family member to seek out and present the card to the CHW when the pregnant woman started into labor.\nCHWs, notified by the card, attended the delivery whenever possible, or visited the mother and newborn infant as early as possible in the postnatal period.\nCHWs conducted three additional postnatal visits on days 2, 5 and 8 to promote preventive newborn care practices and to identify and refer sick neonates to Kumudini Hospital.\nDuring each of the postnatal visits, CHWs completed a standardized newborn assessment form, identified the presence of serious illnesses requiring referral to Kumudini Hospital \u2013 including illness indicative of infections, potentially requiring antibiotic treatment \u2013 and made referral to the hospital according to the clinical algorithm.\nCHWs' classification of neonates with illness had high validity compared to physicians' classification.\nUse of the clinical algorithm by CHWs during routine household surveillance was also validated in identifying severely ill neonates needing urgent referral to the hospital and those who subsequently died.\nTo eliminate potential barriers to care seeking for illness, CHWs facilitated transport, if necessary, for neonates needing referral-level evaluation at Kumudini Hospital, and all care at the hospital was free-of-charge for referred neonates.\nThe mean travel time to the hospital was about one hour, and formative research suggested positive community perception of the quality of care at the hospital.\nIf the family refused to be referred, the CHW continued to encourage referral but managed the neonate in the home according to the algorithm, without use of injectable antibiotics.\nData\nIn order to examine intervention effectiveness, baseline and endline surveys were conducted in both study arms, using comparable questionnaires.\nPrimary outcome measure was neonatal mortality; secondary outcomes included antenatal and immediate newborn care behaviours, knowledge of danger signs, and care seeking for neonatal complications.\nThe surveys included all households with recently delivered women (RDW) (i.e., women who had a pregnancy outcome in the three calendar years before the survey), and collected household wealth and basic demographic information from all household members.\nTo measure the mortality outcome, a hypothesized 40% reduction during the intervention period, a total sample size of 14,872 neonates was required, based on the baseline NMR of 28 per 1000 live births, power of 80% and an estimated design effect of 2.55 derived from the baseline data.\nGiven a crude birth rate of 27 per 1000 population, 7884 live births were expected per year, and the surveys collected life-time pregnancy history from all eligible RDW at both baseline and endline.\nWe anticipated that, after three months of initial intervention scale-up, a two-year period of enrolment during which the implementation of the intervention was stabilized would be sufficient.\nIn addition, for all identified neonatal deaths during a defined period (see below), verbal autopsy data, including signs and symptoms of illness leading to deaths, were collected by separate interviewers who were trained in verbal autopsy data collection for six days.\nTo measure indicators of care practice and knowledge, the surveys also collected knowledge (K) of maternal and newborn care practices, household practice (P) of maternal-newborn preventive and curative care behaviours; and program coverage (C) among RDWs who had a pregnancy outcome in the last 12 months before each survey, hereafter referred to as KPC-RDW.\nAt baseline, all eligible KPC-RDW were interviewed, while a sample of KPC-RDW was interviewed during the endline survey.\nThe endline sample of KPC-RDW was randomly selected within each union, based on a sample size calculation to provide estimates for all KPC indicators assuming 50% prevalence with \u00b16% precision and response rate of 85% for each union.\nBaseline household listing and mapping was conducted during March \u2013 June 2003, and households with at least one RDW who had a pregnancy outcome between 2000 and 2002 were identified.\nThe baseline survey was conducted during April \u2013 July 2003.\nResponse rates were 86.9% (14532/16725) among all RDW and 92.4% among all KPC-RDW (4636/5015).\nVerbal autopsy data were collected during September \u2013 December 2003, on average 14.8 months (SD: 3.6, n\u200a=\u200a109) after the death, for neonatal deaths among those born in 2002 (response rate 88.6%, 109/123).\nThe intervention was introduced and scaled up to the entire study area during December 2003\u2013 February 2004.\nImplementation continued through December 2006.\nEndline enumeration of households with RDW who had a pregnancy outcome between 2003 and 2005 was conducted during December 2005\u2013 April 2006.\nThe endline survey was conducted during January \u2013 May 2006, before the end of the trial, in order to maintain community cooperation and minimize potential end-of-project effect in outcome measurement.\nThe response rate was 87.8% (14731/16771) among all RDW.\nThe KPC-RDW response rate was 94.0% (3519/3744).\nFor all neonatal deaths among those born during the intervention period (2004\u20132005), verbal autopsy information was collected during April \u2013 August 2006 (response rate 86.4%, 222/257).\nThe mean interval between a death and verbal autopsy data collection was 16.5 months (SD: 8.1, n\u200a=\u200a222).\nIn addition, two interim adequacy surveys of knowledge and practice were conducted to monitor the coverage or adequacy of the intervention, and to guide adjustments in the implementation to optimize coverage and quality of the intervention.\nRandom samples of households were selected from the baseline household listing for the two adequacy surveys.\nSample size was calculated to provide estimates for selected KPC indicators, assuming 50% prevalence, \u00b110% precision, and response rate of 85% for each union.\nThe first and second adequacy surveys were conducted during December 2004\u2013 January 2005 and August \u2013 September 2005, respectively.\nIn total, 1141 and 1213 women who had a pregnancy outcome in the 12 months before each survey were enumerated in the first and second adequacy surveys, respectively.\nResponse rates were 82.7% (1141/1380) for the first, and 86.5% (1194/1380) for the second adequacy survey.\nInformed verbal consent was administered by survey interviewers for all participants.\nStatistical Analysis\nWe analyzed the two adequacy surveys and the endline survey to estimate coverage changes in three consecutive 8-month periods in the intervention arm.\nAnalyses were restricted to pregnancies which ended during the following 8-month periods, to avoid overlap between surveys: April \u2013 November 2004 (from adequacy survey 1), December 2004\u2013 July 2005 (from adequacy survey 2), and August 2005\u2013 March 2006 (from the endline survey).\nCoverage of the program was assessed in three areas: antenatal (whether a CHW visited the home at least once during pregnancy), delivery (whether a CHW attended at delivery), and postnatal (whether a CHW assessed a neonate at least once within the first 2, 7, and 28 days of life, respectively, and, among those who received postnatal visits, the mean time of first visit and the mean number of visits).\nThe baseline and endline surveys were analyzed to assess changes in three main outcomes in both comparison and intervention arms: reported maternal and newborn care practices, knowledge of maternal and newborn danger signs of illness, and neonatal mortality, controlled for basic demographic and socioeconomic characteristics.\nWe first estimated means and proportions of RDW with selected background characteristics, by study arm and survey, including mother's age at birth (<20 years, 20\u201329 years, and \u226530 years), mother's educational attainment (< primary school completion vs. \u2265 primary school completion), and household wealth status.\nA household wealth index score, based on the pooled data of baseline and endline surveys, was constructed using principal component analysis of household assets.\nHouseholds in each survey were ranked based on the index score and categorized into quintiles.\nThe lowest and highest quintiles were classified as poor and rich, respectively, relative to the three middle quintiles.\nAntenatal and neonatal care practices were measured among KPC-RDW.\nThe last pregnancy was used as an index pregnancy if there were two or more pregnancies within the 12-month recall period.\nA woman was considered to have received routine antenatal care from a qualified provider (distinct from BNCP home visits by CHWs) if she had received \u22651 antenatal check-up either at a medical facility (i.e., satellite clinic, Union Health and Family Welfare Centre \u2013 a primary health facility serving approximately 20,000 population in the union, Upazila health complex \u2013 a first-level referral public hospital in each sub-district, qualified doctor's chamber, clinic or hospital) or by a qualified provider (i.e., doctor, nurse, Family Welfare Visitor \u2013 health personnel at a Union Health and Family Welfare Centre, or medical assistant).\nAmong all home-born live births, seven selected immediate essential newborn care variables were measured, including sterile cord cut (i.e., the cord was cut by either a blade which was boiled before use or a blade from a clean delivery kit); drying/wiping the baby before delivery of the placenta; wrapping the baby before delivery of the placenta; delaying the first bath to the third day of life or later; initiating breastfeeding within one hour after delivery; breastfeeding prior to giving any food or liquid; and not applying anything to the cord immediately after cutting and tying it.\nCare seeking to a qualified provider (defined above) was measured among all neonates who had signs of complications based on maternal report.\nThe baseline and endline surveys collected information on 11 and 19 complication signs, respectively, and we restricted analyses to neonates with \u22651 of 10 signs collected in both surveys (Table 2).\nKnowledge of maternal and neonatal danger signs was measured among KPC-RDW, using unprompted binary knowledge variables of 10 antenatal, 11 childbirth, 9 postpartum maternal, and 16 neonatal danger signs (Table 3).\nFour composite knowledge score variables were constructed for antenatal (range [0\u201310]), childbirth (0\u201311), postnatal maternal (0\u20139) and neonatal danger signs (0\u201316), by adding un-weighted positive answers for each of the individual signs within the category.\nComposite variables were treated as having missing values if the respondent had not completed all questions in each category.\nTo investigate differential changes in knowledge and practices, we conducted intention-to-treat analyses at the study arm level, using difference-in-difference test with interaction terms for time (baseline vs. endline) and study arm (comparison vs. intervention).\nWe estimated predicted mean of each knowledge or practice indicator by time and study arm and compared the change between baseline and endline by study arm, controlling for maternal and household background characteristics described above.\nLinear probability regression models were used to test the null hypothesis that the difference-in-difference was zero.\nRobust standard errors were adjusted for clustering on each union.\nNeonatal mortality was examined using pregnancy history by all RDW.\nWe assessed mortality data quality by examining distributions of the monthly number of live births, the monthly number of neonatal deaths, and age at deaths by year.\nWe included live births in two-calendar-year-periods, January 2001\u2013 December 2002 from the baseline survey and January 2004\u2013\nDecember 2005 from the endline survey, to control for the potential seasonal effect on mortality and to eliminate the 11-month pre-intervention period (January \u2013 November 2003) included in the 3-year pregnancy history recall period for the endline survey.\nWe estimated NMR and 95% confidence intervals (CI) by time and study arm.\nWe used a survival-time model with a Weibull survival distribution to estimate relative hazard of mortality between the study arms at baseline and at endline, adjusted for child sex and background characteristics described above.\nRobust standard errors adjusted for clustering on each union.\nFinally, verbal autopsy data were analyzed to estimate cause-specific neonatal mortality rate by time and study arm, using neonatal deaths among those born in 2002 (baseline), and in 2004\u20132005 (endline).\nWe applied a standard hierarchical algorithm to assign one primary cause out of the seven major causes of neonatal mortality in the order of: congenital malformation, tetanus, preterm birth, birth asphyxia, birth injury, sepsis or pneumonia, and diarrhea.\nA p-value of 0.05 was considered statistically significant, and all analyses of KPC indicators were adjusted for sampling weight.\nSTATA 9.0 statistical software (Stata Corporation, College Station, TX, USA) was used for all analyses.\nThe study was approved by the Committee on Human Research at the Johns Hopkins Bloomberg School of Public Health, the Ethical Review Committee and the Research Review Committee at ICDDR,B, the Ethical Review Committee at Dhaka Shishu Hospital and the Ethical Review Committee at Oxford University.\nThe study was registered at clinicaltrials.gov, No. NCT00198627.\nResults\nEnrolment\nA total of 9987 women of reproductive age had 5031 pregnancy outcomes in the intervention arm, including 302 miscarriages, 113 stillbirths and 4616 live births (Figure 2).\nIn the comparison arm, 11153 women of reproductive age had 5669 pregnancy outcomes, including 319 miscarriages, 109 stillbirths and 5241 live births.\nThere were no differences in the rates of miscarriage and stillbirth between the two arms.\nEnrolment rates did not vary across unions.\nCoverage of the Intervention\nIn the intervention arm, percent of pregnant women receiving \u22651 BNCP visit reached over 90% during the first survey period (April \u2013 November 2004) and remained comparable in the subsequent two 8-month survey periods (Table 4).\nSimilar rates were seen for receipt of two BNCP visits.\nPercent of home-deliveries attended by CHWs was 12% during April \u2013 November 2004, increased to 20% during December 2004\u2013 July 2005, but remained at 14% during August 2005\u2013 March 2006.\nPercent of home-born newborns assessed by CHWs within the first two and seven days of life improved from 54% to 69% and from 66% to 74%, respectively, over the survey periods.\nAmong those who were assessed by CHWs at least once during the first 28 days of life, the average timing of the initial assessment decreased and the mean number of assessments increased over the periods.\nPractice\nIndicators of maternal and newborn care practice and knowledge were similar in the intervention and comparison arms at baseline.\nAdjusted for significant improvement in background socioeconomic characteristics in each arm (Table 5), proportions of women who received \u22651 routine antenatal check-up (distinct from antenatal BNCP visits by CHWs) from a qualified provider and took antenatal iron supplements increased significantly in the intervention arm (reaching 69% and 56%, respectively), but not in the comparison arm (49% and 43%, respectively) (Table 6).\nThere was no change in the proportions of women who received \u22651 tetatus toxoid immunization during pregnancy in either study arm (approximately 75%), however, the proportion of women who received \u22652 tetanus toxoid immunizations during pregnancy decreased in both study arms (from about 55% to about 40%), likely associated with national shortage of the vaccine.\nPercent of women who delivered at a health facility remained small, but increased significantly more in the intervention (from 12 to 20%) than the comparison (from 13% to 17%) arm (Table 6).\nAmong all home-born neonates, sterile cord cut, delaying the first bath, early breastfeeding initiation and breastfeeding before any food or liquid increased in both arms, but the increases were substantially larger in the intervention arm than in the comparison arm, reaching about 80% or more (Table 6).\nImmediate drying and immediate wrapping of the baby improved only in the intervention arm, reaching about 14%.\nFinally, among neonates who had \u22651 of the 10 selected complication signs, care seeking from a qualified provider increased significantly more in the intervention arm (from 31% to 56%) than in the comparison arm (from 27% to 35%).\nKnowledge\nUnprompted knowledge of maternal and neonatal danger signs increased significantly in both study arms between the baseline and endline, adjusted for improvement in background socioeconomic characteristics (Table 5).\nHowever, the improvements in the intervention arm were significantly larger than those in the comparison arm (Table 6).\nNevertheless, intervention-arm women identified only about three signs among 15 neonatal danger signs at the endline, and recognition improved only in selected individual neonatal signs, including redness around or discharge from the umbilicus, body cold/shivering, skin lesions, and convulsion (results not shown).\nMortality\nNMR estimates did not vary significantly by time or study arm (Table 7).\nNMR was 24.8 (95% CI: 20.7\u201329.4) and 27.9 (95% CI: 23.5\u201332.8) in the comparison arm at baseline and endline, respectively, and was 25.2 (95% CI: 21.0\u201330.1) and 24.0 (95% CI: 19.8\u201329.0) in the intervention arm at baseline and endline, respectively.\nAdjusted mortality hazard ratio in the intervention arm, compared to the comparison arm, was 1.02 (95% CI: 0.80\u20131.30) at baseline and 0.87 (95% CI: 0.68\u20131.12) at endline.\nVerbal autopsy data ascertainment rate did not vary significantly by sex, age at death, or study arm (results not shown).\nCause-specific neonatal mortality rates did not differ by time or study arm.\nThe most common causes of death during the intervention period (2004\u20132005) in the intervention and comparison areas combined were birth asphyxia (109/222, 49%), prematurity (58/222, 26%) and infection (26/222, 12%) (Table 7).\nDiscussion\nThis cluster randomized controlled trial of a package of maternal and newborn healthcare interventions successfully achieved good coverage of antenatal (\u223c90%) and postnatal (\u223c70%) home visits by CHWs, and significantly improved several key newborn care practices and care seeking for newborn complications from qualified providers.\nKnowledge of maternal and newborn danger signs also improved, although to a limited extent.\nHowever, there was no evidence for an impact of the intervention on neonatal mortality.\nThese results are in contrast to several recent trials which decreased neonatal mortality in various settings in South Asia, and also contrasts with a large-scale program evaluation in rural India where lack of mortality impact seemed to stem from inadequate implementation and insufficient coverage of the interventions.\nOur program coverage for both the antenatal and postnatal components was comparable with levels achieved in other effective trials.\nIn addition, we had strict quality assurance of implementation through regular supervision of CHWs and through intensive monitoring of quality of program implementation through household \u201cadequacy\u201d surveys; data from the surveys was used to identify potential areas for improvement in program implementation, and to guide adjustments in intervention delivery to optimize program impact.\nIn Sylhet, Bangladesh, we achieved a 34% reduction in mortality through a similar package of interventions, supervision and monitoring; further analysis of that program revealed that a 64% reduction in mortality was seen among the newborns who were visited within the first two days of life whereas no mortality impact was found among those who were visited only after the two days.\nCoverage of the first visit within the two days, however, was similar in Sylhet (62%) and Mirzapur (69%), and the magnitude of changes in care practices were also similar.\nThus, factors other than reaching families with the intervention must be considered to explain the lack of mortality impact in this study and to guide future strategies to reduce mortality in moderate mortality settings such as Mirzapur.\nLack of evidence of mortality impact can be due to lack of power to test our hypothesis that the intervention would result in a 40% reduction in mortality in the intervention arm \u2013 a level of reduction that had been observed in other efficacy trials and that we thought would be needed to compel policy and program change in Bangladesh.\nGiven the lower number of live births in the study area than anticipated during study design, we did not achieve our enrolment target of 14,872 births.\nWe speculated that a number of factors contributed to this, including declining fertility in rural areas of Bangladesh, an overestimated initial population size, and potential omission of live births in the retrospective pregnancy history.\nIn particular, preliminary results from the Mirzapur Demographic Surveillance Systems since 2007 suggest that the initial population of 292,000 in 2003 was likely overestimated by about 18%, while the annual number of live births during 2004 and 2005 recorded in the retrospective birth history data was about 5% lower than the prospective demographic surveillance results.\nExpanding the study area or extending the intervention period would have been an option to achieve the target enrolment.\nHowever, the catchment area could not be extended in order to ensure access to Kumudini Hospital; and, there was no compelling reason to continue the trial longer than planned, due to the lack of evidence for a downward mortality trend in the intervention arm using program implementation data.\nIn Sylhet, for example, a non-significant downward trend in mortality was observed within 6 month after the intervention started, and a significant program effect on mortality was observed 2 years after the initial intervention introduction.\nIn addition, improvement in care seeking for illness with qualified providers at Kumudini Hospital by families in both study arms, coupled with the provision of quality, life-saving care for any who reached the hospital, likely contributed to the lack of mortality impact of the intervention relative to the comparison area.\nMost importantly, however, our results highlight that local epidemiology, including levels and causes of mortality in the community, must be taken into careful account during intervention design.\nAs NMR decreases, particularly below about 30 per 1000 live births, the cause structure of mortality and, thus, the relative importance of various risk factors for mortality changes.\nIn most other community-based trials, baseline NMR exceeded 45 per 1000 live births, and serious infections, including sepsis, pneumonia, and tetanus, likely accounted for >40% of neonatal deaths.\nAlthough our intervention was designed to address the major causes of mortality in neonates, it was most robust for the prevention and management of infections.\nIn the Mirzapur population, however, nearly 60% of deaths were due to birth asphyxia or prematurity, and the program had limitations in reaching households at the critical times (i.e., during labour, childbirth and immediately after delivery) to address these conditions, and the CHWs lacked the necessary tools and skills to effectively address these conditions.\nCHWs attended <20% of home deliveries, largely due to difficulties in receiving timely notification of labour onset and in travelling to the home to intervene during delivery, given their population catchment area which extended over four villages and, particularly at night, strong discouragement from CHW families for travelling outside the village out of safety concerns.\nTBAs attended most home deliveries (97%) but in spite of brief but focused training in clean delivery, immediate newborn care, and danger sign recognition and referral, they lacked the capabilities to provide skilled care at birth, including resuscitation of birth asphyxiated newborns.\nSome evidence suggests, however, that TBA training in resuscitation is a potentially effective intervention.\nMoreover, recent reviews and meta-analyses suggest that TBAs have some potential for promoting antenatal care, detecting obstetric complications, referring women to skilled obstetric care and positively impacting stillbirths and neonatal outcomes.\nWe found, however, that the numbers and diversity of TBAs in the community made it challenging to train, supervise and manage them to uniform standards of care, and that TBAs and CHWs infrequently encountered a newborn that required bag-and mask resuscitation, which further complicates attempts to train and equip them to provide effective resuscitation in the community.\nCurrent policy in Bangladesh does not promote TBA training programs, however, and implementation of newborn resuscitation outside health facilities is challenging.\nThus, skilled attendance at delivery remains a key policy and program priority for reducing both neonatal and maternal mortality in Bangladesh.\nIn addition to skilled care at delivery, early postnatal care is also critical for reducing mortality in moderate neonatal mortality settings, considering the preponderance of early deaths due to prematurity, birth asphyxia, and, to a lesser extent, vertically acquired sepsis.\nAlthough our overall coverage of postnatal care was good, only 18% and 33% of neonates who died within the first day and the first week of life, respectively, were visited by CHWs prior to the death.\nMoreover, among newborns who were assessed by CHWs and found to be ill, only 54% complied with referral to hospital and compliance with referral was 30% less likely in the first week of life, despite attempts to eliminate major care seeking barriers \u2013 danger sign recognition, access to the hospital and cost.\nThus, emphasis must be placed on community mobilization and empowerment, and on greater understanding of and development of improved approaches to overcome social and financial barriers to referral compliance and care seeking at facilities, especially in the first week of life and in settings where cultural seclusion after birth remains a social norm.\nMoreover, as NMR is reduced below about 30 per 1000, reliance on community-based care is likely to be inadequate to address the needs of extremely preterm infants, who often need additional interventions beyond essential newborn care interventions (e.g., breastfeeding, warmth and hygiene, and emollient therapy), including corticosteroid administration to the mother prior to delivery, surfactant therapy at birth, and assisted ventilation such as continuous positive airway pressure.\nSkilled attendance at facility-based deliveries, along with adaptation of these additional interventions for implementation in first-level facilities in low resource settings, can help to ensure their coverage.\nEmerging evidence suggests that in addition to understanding and overcoming social barriers to care seeking at facilities, programs to address financial barriers may also provide a powerful stimulus to families to access skilled care for delivery and immediate postnatal care at health facilities.\nFinally, for treatment of serious neonatal infections, community-based case management is a viable alternative to facility-based care even where access to quality health care at facilities can be ensured.\nIn Sylhet, Bangladesh, while only 34% of referrals of sick newborns to hospital by CHWs were complied with, another 43% accepted injectable antibiotic treatment at home.\nNeonates in each treatment group had a significantly reduced hazard of mortality, compared to sick neonates who received no treatment or treatment from unqualified providers, indicating that with the addition of home-based treatment, approximately three-fourths of sick neonates received effective curative antibiotic treatment preventing death, a substantial improvement over what was achieved in Mirzapur, where we did not offer home-based treatment with injectable antibiotics.\nIn summary, for optimal survival improvement in low resource populations with moderate NMR, the intervention design must include a clear pathway to survival that links risk factors with causes of mortality, and identifies locally contextualized approaches to risk reduction.\nAs community-based interventions mature and NMR comes down, programs must ensure, in addition to essential newborn care; skilled care during childbirth, including interventions to prevent and manage birth asphyxia and respiratory distress syndrome in preterm infants; and high coverage of curative postnatal care in the first two days of life.\nBarriers to care seeking for illness must also be addressed.\nWhere poor care seeking at referral-level hospitals exists during the early neonatal period, adaptation of interventions for extremely preterm infants for use at community clinic level must be prioritized, and consideration given to inclusion of home-based treatment of serious infections integrated into community case management strategies for childhood infections.\nDistribution of study unions (clusters), Mirzapur sub-district, Tangail district, Bangladesh.Red circle: Union Head Quarter. Star: Kumudini Hospital. Light blue line: River/Beel. Pink shade: Intervention Area, Purple shade: Comparison Area.\nTrial profile for measurement of neonatal mortality.*Participants are women of reproductive age (15\u201349).\n\nAntenatal (birth and newborn care preparedness) and postnatal interventions at home by community health workers.\nPRENATAL: Two home visits scheduled at 12\u201316 weeks and 32\u201334 weeks to:\n1. Promote antenatal care, including:\n(1) Making three antenatal care visits from a health centre or a satellite clinic\n(2) Receiving two doses of tetanus toxoid vaccine\n(3) Procuring adequate iron-folic acid (IFA) supplementation\n(4) Eating extra food\n(5) Care seeking for the following maternal danger signs:\n\u2003- Prolonged labor\n\u2003- Hemorrhage\n\u2003- Fever\n\u2003- Convulsion\n\u2003- Edema of the face, hands or legs, or\n\u2003- Blurred vision\n2. Promote birth planning, including:\n(1) Planning for delivery at a health facility\n(2) If facility is not feasible, choosing a trained birth attendant; preparing the site of delivery in the house; obtaining birth kit or boiling the blade and the pieces of thread; planning for emergency transport; and saving money for emergency\n3. Distribute: clean delivery kit, obtained from Bangladesh Rural Advancement Committee (NGO) free-of-charge, at the second antenatal visit for use by birth attendant\n4. Promote newborn-care preparedness, including:\n(1) Choosing a household member to take care of the newborn right after birth\n(2) Drying and wrapping the baby from head to toe soon after delivery and before the delivery of placenta; using 2 pieces of cloth to wrap the newborn; holding the baby at all times during and immediately after the delivery; avoiding any contact of the newborn with the floor; not keeping the newborn in an unclean or cold place; applying gentle stimulation or refer for resuscitation of the newborn if he/she does not breathe immediately after birth; and practicing wrapping the baby using a doll during CHW visits\n(3) Feeding colostrum to the newborn; initiating breastfeeding immediately after birth; practicing exclusive breastfeeding up to six months; and feeding the newborn frequently in the proper position day and night\n(4) Delaying bathing of the newborn for 72 hours\n(5) Umbilical area care: keeping the cord clean and dry; and avoiding applying anything to the umbilical stump\n(6) Monitoring the baby for signs of infection; and seeking care immediately from CHW or health facility if the newborn has any of the following danger signs:\n\u2003- No cry or breathing at birth,\n\u2003- Convulsions\n\u2003- Unconsciousness\n\u2003- Difficulty breathing\n\u2003- Feeling hot or cold to the touch\n\u2003- Skin pustules or blisters\n\u2003- Umbilical pus or redness\n\u2003- Weak, abnormal or absent cry\n\u2003- Lethargic or less than normal movement\n\u2003- Yellow colour of the body, or\n\u2003- Feeding problem\nPOSTNATAL: Four home visits on postnatal days 0, 2, 5, and 8 to:\n1. Reinforce newborn care messages provided through prenatal visits\n2. Provide counseling for routine breastfeeding and for breastfeeding difficulties\n3. Surveillance of newborn illness: Identify sick neonates based on a clinical algorithm. For identified sick neonates, recommend referral-level evaluation at Kumudini hospital or, if referral fails, continue monitoring according to the clinical algorithm.\n\n\nDefinition of neonatal complications used to measure conditional care seeking: baseline and endline survey.\nBaseline survey* | Endline survey\u2020\n1. Fever | 1. Fever (temp more than 101F)\n2. Trouble breathing | 2. Difficulty in breathing or fast breathing\u2021\n3. Jaundice | 3. Jaundice\n4. Diarrhea | 4. Diarrhea\n5. Umbilical infection or discharge | 5. Pus in the umbilicus or redness of the umbilicus\u2021\n6. Convulsion | 6. Convulsion\n7. Stopped breast feeding | 7. Poor feeding or unable to suck\n8. Body became excessive cold | 8. Hypothermia (temp 95.5\u201397.5 F)\n9. Retention of urine | 9. Doesn't pass urine\n10. Unconsciousness | 10. Unconscious\n\n*Persistent vomiting was included in the baseline survey.\nFollowing additional signs were included in the endline survey: (1) red eye/passage of pus from eyes, (2) skin lesion with infection, (3) baby doesn't cry/breath, (4) chest in drawing, (5) doesn't pass stool, (6) cold/cough, and (7) others.\nListed as 2 separate signs in the survey questionnaire.\n\nIndividual signs included in the prenatal, labor/delivery, and postpartum danger sign knowledge scores.\n | Danger signs\nPrenatal | 1. Severe headache\n | 2. Blurred vision\n | 3. Fetal movement absent\n | 4. High blood pressure\n | 5. Edema of the face/swelling\n | 6. Edema of the hands/leg swelling\n | 7. Convulsions\n | 8. Excessive vaginal bleeding\n | 9. Severe lower abdominal pain\n | 10. Leaking fluid (meconium stained)\nLabor/delivery | 1. Excessive vaginal bleeding\n | 2. Foul-smelling discharge\n | 3. High fever\n | 4. Baby's hand or feet coming out first\n | 5. Baby is in abnormal position\n | 6. Prolong labor (>12 hours)\n | 7. Retained placenta\n | 8. Rupture uterus\n | 9. Cord prolapse\n | 10. Cord around neck\n | 11. Convulsion\nPostpartum | 1. Excessive vaginal bleeding\n | 2. Foul-smelling discharge\n | 3. High fever\n | 4. Inverted nipples\n | 5. Tetanus\n | 6. Retained placenta\n | 7. Severe abdominal pain\n | 8. Convulsions\n | 9. Engorged breasts/swelling of breasts\nNeonatal | 1. Poor feeding or unable to suck\n | 2. Diarrhea\n | 3. Redness around the cord\n | 4. Red eye/discharging eyes\n | 5. Difficult breathing\n | 6. Yellow coloration of the skin/jaundice\n | 7. Hypothermia/shivering\n | 8. Blisters on skin/Skin lesion\n | 9. Baby doesn't cry\n | 10. Fever\n | 11. Unconscious\n | 12. Fast breathing\n | 13. Chest indrawing\n | 14. Doesn't pass urine\n | 15. Doesn't pass stool\n | 16. Convulsions\n\nOne additional prenatal danger signs (high fever) and 1 additional newborn danger sign (cold/cough) were included in the endline survey. We excluded those in creating the knowledge scores in order to maintain comparability across surveys.\n\nChanges in program coverage by community health workers in the intervention arm, among women who had a pregnancy outcome in the 8-month period before each survey.\nService | Adequacy survey 1 |  | Adequacy survey 2 |  | Endline survey | \n | (Apr 2004\u2013Nov 2004) |  | (Dec 2004\u2013Jul 2005) |  | (Aug 2005\u2013Mar 2006) | \nAntenatal BNCP, among all pregnant women | (N\u200a=\u200a565) |  | (N\u200a=\u200a564) |  | (N\u200a=\u200a1096) | \nHome visit at least once | 91.3 | (89.0\u201393.7) | 87.0 | (84.2\u201389.8) | 93.0 | (91.5\u201394.6)\nHome visits twice | 86.9 | (84.1\u201389.7) | 83.8 | (80.8\u201386.9) | 91.0 | (89.3\u201392.7)\nDelivery, among all women who delivered at home | (N\u200a=\u200a447) |  | (N\u200a=\u200a385) |  | (N\u200a=\u200a800) | \nLabor notification to CHWs | 28.8 | (24.6\u201333.0) | 44.4 | (39.4\u201349.3) | 34.8 | (31.5\u201338.1)\nDelivery attendance by CHWs | 12.0 | (9.0\u201315.0) | 20.0 | (16.0\u201324.0) | 13.8 | (11.4\u201316.2)\nPostnatal, among all women who delivered live births at home | (N\u200a=\u200a433) |  | (N\u200a=\u200a379) |  | (N\u200a=\u200a790) | \nHome visits at least once during postnatal day 0\u201327 | 75.5 | (71.4\u201379.5) | 83.7 | (79.9\u201387.4) | 79.7 | (76.9\u201382.5)\nHome visits at least once during postnatal day 0\u20136 | 65.6 | (61.1\u201370.1) | 77.7 | (73.4\u201381.9) | 73.8 | (70.7\u201376.9)\nHome visits at least once during postnatal day 0\u20131 | 53.6 | (48.8\u201358.3) | 73.6 | (69.1\u201378.1) | 69.1 | (65.9\u201372.4)\nTime of first home visit (day)\u2020 | 2.4 | (1.9\u20133.0) | 1.6 | (1.0\u20132.2) | 1.5 | (1.1\u20131.8)\nTotal number of home visits\u2020 | 2.6 | (2.5\u20132.8) | 2.8 | (2.6\u20133.0) | 3.2 | (3.0\u20133.3)\n\nBNCP: Birth and newborn care preparedness; CHWs: Community health workers.\n\u2020Among those who had at least 1 postnatal home visit during postnatal days 0\u201327: n\u200a=\u200a325 (Adequacy survey 1); n\u200a=\u200a318 (Adequacy survey 2); and n\u200a=\u200a628 (Adequacy survey 1).\n\nMaternal demographic and household economic characteristics by study arm and time, among all women had live births between 2001\u20132002 (baseline) and between 2004\u20132005 (endline).\n | Comparison |  |  |  | Intervention |  |  | \n | Baseline |  | Endline |  | Baseline |  | Endline | \n | (n\u200a=\u200a5166) |  | (n\u200a=\u200a5143) |  | (n\u200a=\u200a4822) |  | (n\u200a=\u200a4498) | \n | % | (95% CI) | % | (95% CI) | % | (95% CI) | % | (95% CI)\nMother's age at birth (year) |  |  |  |  |  |  |  | \nmean | 25.2 | (25.0\u201325.3) | 25.0 | (24.8\u201325.1) | 25.4 | (25.2\u201325.5) | 25.3 | (25.1\u201325.4)\n<20 years | 14.4 | (13.4\u201315.4) | 15.6 | (14.6\u201316.6) | 12.7 | (11.7\u201313.6) | 13.1 | (12.2\u201314.2)\n\u226530 years | 20.2 | (19.1\u201321.3) | 17.9 | (16.9\u201319.0) | 20.3 | (19.2\u201321.4) | 18.7 | (17.6\u201319.9)\n\u226535 years | 6.2 | (5.6\u20136.9) | 6.5 | (5.8\u20137.2) | 6.3 | (5.7\u20137.0) | 6.5 | (5.8\u20137.3)\nMaternal education |  |  |  |  |  |  |  | \nEver attended school | 61.7 | (60.4\u201363.1) | 72.4 | (71.1\u201373.6) | 64.7 | (63.3\u201366.1) | 75.6 | (74.4\u201376.9)\nCompleted primary school | 48.9 | (47.5\u201350.3) | 58.1 | (56.7\u201359.4) | 52.0 | (50.6\u201353.5) | 62.5 | (61.1\u201363.9)\nCompleted high school | 14.6 | (13.6\u201315.6) | 20.0 | (18.9\u201321.1) | 16.5 | (15.4\u201317.6) | 21.9 | (20.7\u201323.1)\nHousehold wealth* |  |  |  |  |  |  |  | \nWealth index score | \u22120.19 | (\u22120.2\u20130.1) | 0.44 | (0.4\u20130.5) | \u22120.36 | (\u22120.4\u20130.3) | 0.44 | (0.4\u20130.5)\nPoor | 22.7 | (21.6\u201323.9) | 15.0 | (14.0\u201316.0) | 25.7 | (24.5\u201327.0) | 14.7 | (13.7\u201315.8)\nRich | 17.6 | (16.6\u201318.7) | 26.2 | (25.0\u201327.4) | 15.7 | (14.6\u201316.7) | 26.5 | (25.2\u201327.8)\n\n*Wealth index created based on pooled baseline and endline surveys, using principal component analysis of durable goods, electricity, toilet facility, sources of drinking water, and housing materials. Poor refers to the lowest quintile of the wealth index and rich is the highest quintile of the wealth index.\n\nAdjusted predicted mean of knowledge and practice indicators by study arm and time, among women who had a pregnancy outcome in the 1-year period before each survey.*\n\n | Comparison |  | Intervention |  | \n | Baseline | Endline | Baseline | Endline | \n | (May 2002\u2013 Jul 2003) | (Feb 2005\u2013 Apr 2006) | (May 2002\u2013 Jul 2003) | (Feb 2005\u2013Apr 2006) | \nPRACTICE (percent of target population practicing a behavior) |  |  |  |  | \nPrenatal care and birth preparedness among all women |  |  |  |  | \nNumber of women | 2644 | 1759 | 2371 | 1732 | \nHad \u22651 antenatal care visits from a qualified provider\u2020 | 47.8 | 49.1 | 47.4 | 68.8 | \u2225\nReceived \u22651 tetanus immunization | 76.4 | 73.8 | 77.1 | 77.5 | \nReceived \u22652 tetanus immunizations | 56.9 | 41.0 | 54.7 | 39.8 | \nReceived iron supplementation | 45.8 | 42.7 | 47.9 | 55.7 | \u2225\nFacility-based delivery |  |  |  |  | \nNumber of women | 2644 | 1759 | 2371 | 1732 | \nDelivered at medical facilities | 12.5 | 16.5 | 12.1 | 20.2 | \u2225\nImmediate newborn care among all home-born live births |  |  |  |  | \nNumber of home-born live births | 2006 | 1322 | 1805 | 1231 | \nSterile cord cut\u2021 | 59.2 | 66.9 | 63.3 | 95.1 | \u2225\nNot applying anything on the newly-cut cord | 95.1 | 86.0 | 94.8 | 94.3 | \u2225\nDrying/wiping the baby before delivery of placenta | 2.1 | 3.0 | 2.2 | 14.4 | \u2225\nWrapping the baby before delivery of placenta | 2.4 | 2.7 | 2.9 | 13.5 | \u2225\nDelaying bath to the 3rd day or later | 1.5 | 13.4 | 1.6 | 77.8 | \u2225\nBreastfeeding initiation within 1 hour after birth | 41.2 | 55.0 | 40.9 | 80.0 | \u2225\nBreastfeeding prior to any food/liquid | 28.9 | 50.5 | 29.3 | 87.3 | \u2225\nCare seeking among neonates with complications |  |  |  |  | \nNumber of neonates with 1 or more of the 10 complications\u00a7 | 812 | 400 | 733 | 355 | \nReceived any treatment | 93.7 | 95.9 | 92.9 | 97.3 | \nReceived treatment from a qualified provider\u2020 | 27.4 | 34.6 | 30.7 | 55.7 | \u2225\nKNOWLEDGE (mean danger sign knowledge scores) [range] |  |  |  |  | \nNumber of women | 2644 | 1759 | 2371 | 1732 | \nMaternal danger sign knowledge score: antenatal [0\u201310] | 1.0 | 2.2 | 1.1 | 2.9 | \u2225\nMaternal danger sign knowledge score: labor/delivery [0\u201311] | 1.1 | 1.9 | 1.2 | 2.4 | \u2225\nMaternal danger sign knowledge score: postpartum [0\u20139] | 1.0 | 2.0 | 1.0 | 2.5 | \u2225\nNeonatal danger sign knowledge score [0\u201315] | 2.3 | 2.4 | 2.3 | 2.8 | \u2225\n\n*Adjusted for mother's age at birth (<20 years, 20\u201329 years (reference group), or \u226530 years), maternal educational attainment (<primary school completion vs. \u2265 primary school completion), and the household wealth (the lowest quintile, middle 3 quintiles (reference), or the highest quintile of the wealth index), using linear probability regression models.\nIncludes check-up/treatment by a qualified provider (e.g., doctor, nurse, family welfare visitor, or medical assistant) or at a medical facility (e.g., satellite clinic, Family Welfare Centre, Upazila health complex, qualified doctor's chamber, clinic, or hospital).\nBlade from a clean delivery kit or other instrument that was boiled before use.\nSee Table 2 for the list of complications.\nSignificant differential change over time by study arm (i.e., significant interaction) (p-value <0.05).\n\nLevels and causes of neonatal mortality by time and study arm.\n |  | Baseline |  |  |  | Endline |  | \n |  | Comparison |  | Intervention |  | Comparison |  | Intervention\nLEVELS |  |  |  |  |  |  |  | \nPeriod (calendar years) |  | 2001\u20132002 |  | 2001\u20132002 |  | 2004\u20132005 |  | 2004\u20132005\nNumber of live births | 5292 |  | 4951 |  | 5241 |  | 4616 | \nNumber of deaths | 131 |  | 125 |  | 146 |  | 111 | \nNeonatal mortality rate (per 1000 live births) [95% CI] | 24.8 | [20.7\u201329.4] | 25.2 | [21.0\u201330.1] | 27.9 | [23.5\u201332.8] | 24.0 | [19.8\u201329.0]\nAdjusted hazard ratio* [95% CI] | 1 | (reference) | 1.02 | [0.80\u20131.30] | 1 | (reference) | 0.87 | [0.68\u20131.12]\nCAUSES, based on a hierarchical algorithm\u2020 |  |  |  |  |  |  |  | \nPeriod (calendar years) |  | 2002 |  | 2002 |  | 2004\u20132005 |  | 2004\u20132005\nNumber of live births | 2636 |  | 2463 |  | 5241 |  | 4616 | \nNumber of deaths | 72 |  | 51 |  | 146 |  | 111 | \nNumber of deaths with complete verbal autopsy data (% of total neonatal deaths) | 67 | (93.1) | 42 | (82.4) | 129 | (88.4) | 93 | (83.8)\nNumber of deaths by cause (% of neonatal deaths with complete verbal autopsy data [95% CI]) |  |  |  |  |  |  |  | \nCongenital malformations | 6 | (9.0 [3.4\u201318.5]) | 4 | (9.5 [2.7\u201322.6]) | 10 | (7.8 [3.8\u201313.8]) | 0 | (0.0 [0.0\u20133.9]) \u2021\nTetanus | 3 | (4.5 [0.9\u201312.5]) | 2 | (4.8 [0.6\u201316.2]) | 3 | (2.3 [0.5\u20136.6]) | 1 | (1.1 [0.0\u20135.8])\nPreterm | 16 | (23.9 [14.3\u201335.9]) | 15 | (35.7 [21.6\u201352.0]) | 33 | (25.6 [18.3\u201334.0]) | 25 | (26.9 [18.2\u201337.1])\nBirth asphyxia | 20 | (29.9 [19.3\u201342.3]) | 11 | (26.2 [13.9\u201342.0]) | 63 | (48.8 [39.9\u201357.8]) | 46 | (49.5 [38.9\u201360.0])\nBirth injury | 0 | (0.0 [0.0\u20135.4]) \u2021 | 0 | (0.0 [0.0\u20138.4]) \u2021 | 0 | (0.0 [0.0\u20132.8]) \u2021 | 3 | (3.2 [0.7\u20139.1])\nInfection | 11 | (16.4 [8.5\u201327.5]) | 8 | (19.0 [8.6\u201334.1]) | 15 | (11.6 [6.7\u201318.5]) | 11 | (11.8 [6.1\u201320.2])\nDiarrhea | 0 | (0.0 [0.0\u20135.4]) \u2021 | 0 | (0.0 [0.0\u20138.4]) \u2021 | 1 | (0.8 [0.0\u20134.2]) | 1 | (1.1 [0.0\u20135.8])\nCause not assigned | 11 | (16.4 [8.5\u201327.5]) | 2 | (4.8 [0.6\u201316.2]) | 4 | (3.1 [0.9\u20137.7]) | 6 | (6.5 [2.4\u201313.5])\nCause-specific neonatal mortality rate (per 1000 live births) [95% CI] |  |  |  |  |  |  |  | \nCongenital malformations | 2.3 | [0.8\u20135.0] | 1.6 | [0.4\u20134.2] | 1.9 | [0.9\u20133.5] | 0.0 | [0.0\u20130.8] \u2021\nTetanus | 1.1 | [0.2\u20133.3] | 0.8 | [0.1\u20132.9] | 0.6 | [0.1\u20131.7] | 0.2 | [0.0\u20131.2]\nPreterm | 6.1 | [3.5\u20139.9] | 6.1 | [3.4\u201310.0] | 6.3 | [4.3\u20138.8] | 5.4 | [3.5\u20138.0]\nBirth asphyxia | 7.6 | [4.6\u201311.7] | 4.5 | [2.2\u20138.0] | 12.0 | [9.2\u201315.4] | 10.0 | [7.3\u201313.3]\nBirth injury | 0.0 | [0.0\u20131.4] \u2021 | 0.0 | [0.0\u20131.5] \u2021 | 0.0 | [0.0\u20130.7] \u2021 | 0.6 | [0.1\u20131.9]\nInfection | 4.2 | [2.1\u20137.5] | 3.2 | [1.4\u20136.4] | 2.9 | [1.6\u20134.7] | 2.4 | [1.2\u20134.3]\nDiarrhea | 0.0 | [0.0\u20131.4] \u2021 | 0.0 | [0.0\u20131.5] \u2021 | 0.2 | [0.0\u20131.1] | 0.2 | [0.0\u20131.2]\nCause not assigned | 4.2 | [2.1\u20137.5] | 0.8 | [0.1\u20132.9] | 0.8 | [0.2\u20132.0] | 1.3 | [0.5\u20132.8]\n\n*Analysis of baseline and endline combined, stratified by study arm (n\u200a=\u200a10553, intervention arm, and n\u200a=\u200a9567, intervention arm). Mortality risk at baseline (2001\u20132002) was the reference group in each arm. Estimates were controlled for: sex, mother's age at birth [<20 years, 20\u201329 years [reference group]), or \u226530 years]), maternal educational attainment [<primary school completion vs. \u2265 primary school completion]), and the household wealth [the lowest quintile, middle 3 quintiles [reference]), or the highest quintile of the wealth index]), using a survival-time model with a Weibull survival distribution.\nCauses are presented in the order of the hierarchy. Case definitions for the seven causes are presented elsewhere in detail.\n\nOne-sided 97.5% confidence interval.",
        "label": "low",
        "id": "task4_RLD_test_462"
    },
    {
        "paper_doi": "10.3390/healthcare2040417",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallelEthics and informed consent: yes\nFollow-up period: 6 weeksSample size estimate: pilot studyITT analysis: yes, number randomised: 92, number analysed: 87Funding: non-industryPreregistration: yes\n\n\nParticipants: Location: Queensland, Australia\nIntervention group: n = 35,control group: n = 35Mean age: intervention group = 30.6 years (IQR 5.5),control group = 30.7 years (IQR 5.0)\nInclusion criteria: booked for elective caesarean section; pre-pregnancy BMI >= 30; able to provide consent.\nExclusion criteria: women whose condition changed to require urgent caesarean section; previous participation in the trial; existing infection.\n\n\nInterventions: Aim/s: to assess the feasibility of a definitive RCT to test the effectiveness and safety of prophylactic NPWT in obese women after caesarean section.\n Group 1 (NPWT) intervention: PICO dressing applied over the primarily closed incision by the surgeon in the operating room. Dressing was left on for 4 days or longer if drainage continued, unless soiled or dislodged.Group 2 (control) intervention: Comfeel dressing applied over the primarily closed incision by the surgeon in the operating room. Dressing was left on for 4 days or longer if drainage continued, unless soiled or dislodged.\nStudy date/s: July 2012 to April 2014\n\n\nOutcomes: Surgical site infectionType of SSIHospital readmissionDehiscence; blistersHaematomaValidity of measure/s: CDC definitions and criteria for superficial, deep, and organ/space SSI were used for the primary outcome and SF-12 for quality of life.Time points: 1, 2, 3, and 4 weeks postsurgery\n\n\nNotes: \n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "Obese women undergoing caesarean section (CS) are at increased risk of surgical site infection (SSI).\nNegative Pressure Wound Therapy (NPWT) is growing in use as a prophylactic approach to prevent wound complications such as SSI, yet there is little evidence of its benefits.\nThis pilot randomized controlled trial (RCT) assessed the effect of NPWT on SSI and other wound complications in obese women undergoing elective caesarean sections (CS) and also the feasibility of conducting a definitive trial.\nNinety-two obese women undergoing elective CS were randomized in theatre via a central web based system using a parallel 1:1 process to two groups i.e., 46 women received the intervention (NPWT PICO\u2122 dressing) and 46 women received standard care (Comfeel Plus\u00ae dressing).\nAll women received the intended dressing following wound closure.\nThe relative risk of SSI in the intervention group was 0.81 (95% CI 0.38\u20131.68); for the number of complications excluding SSI it was 0.98 (95% CI 0.34\u20132.79).\nA sample size of 784 (392 per group) would be required to find a statistically significant difference in SSI between the two groups with 90% power.\nThese results demonstrate that a larger definitive trial is feasible and that careful planning and site selection is critical to the success of the overall study.\n1. Introduction\nBetween 187 and 281 million surgical procedures are performed around the world each year, or one for every 25 people.\nIn Australia in 2008/9, 1.8 million elective surgeries were performed with one elective surgery for every 12.4 people.\nSurgical site infections (SSIs) are defined by the Centers for Disease Control and Prevention (CDC) as infections occurring up to 30 days after surgery that affect the incision, deep tissue at the operation site or involve the organs or body space.\nSSIs have many negative effects including pain, increasing the risk of morbidity and mortality, prolonging hospitalisation and increasing costs.\nOf concern is that SSIs occur in up to 30% of all surgical procedures, and are the third most commonly reported hospital acquired infection.\nObesity is an independent predictor of SSI, thus it has significant safety and cost implications.\nObesity, defined as a body mass index (BMI) \u226530, is a growing global public health problem in developed nations.\nIn 2007\u20132008, 28%\u201343% of 18\u201344 year old Australian women of childbearing age were obese.\nObese women are more likely to have a caesarean section (CS).\nOne meta-analysis of 16 studies identified the odds ratio for overweight or obese women (BMI \u2265 25) having a CS as 2.0 (95% CI 1.9\u20132.2) compared to non-overweight women, similar to results of an Australian analysis of 11,252 women giving birth.\nPost-operative infection is a potential complication of all surgeries including CS, but overweight and obese women are at particular risk.\nA meta-analysis of 6 studies showed the odds ratio for overweight or obese CS women having an infection was 3.3 (95% CI 2.7\u20134.1) compared to non-overweight women, consistent with individual studies.\nGiven that SSI extends hospital length of stay by up to 6 days in women undergoing obstetric and gynaecologic surgery, increasing hospital costs by US$14,000 for each SSI, it has significant implications for women and the health system.\nNegative Pressure Wound Therapy (NPWT), also known as vacuum assisted closure, has been used to aid healing since the late 1990s.\nIt is based on a closed sealed system that applies negative pressure to the wound surface.\nThe wound is covered or packed with an open-cell foam or gauze dressing and sealed with an occlusive drape.\nIntermittent or continuous suction is maintained by connecting suction tubes from the wound dressing to a vacuum pump and liquid waste collector.\nStandard negative pressure rates are 50\u2013125 mm Hg. Despite limited evidence of its effectiveness, Tipton and colleagues report \u201cvacuum therapy can be included as an option for management of abdominal wounds, but evidence from randomized controlled trials in obese women undergoing cesarean is not available\u201d.\nOthers note NPWT is increasingly being used in closed incisions to prevent SSI and dehiscence.\nAdditionally, one retrospective cohort study of 48 women receiving standard dressings compared to 21 women receiving NPWT found fewer wound complications in the NPWT group, but this difference was not statistically significant.\nLimitations of Mark et al.\u2019s study such as the small sample size, use of historical controls, lack of control over the dressings used in the control group and reliance on coded medical record data suggests the findings should be interpreted very cautiously.\nFinally, a recent Cochrane Review of NPWT notes limited evidence for its effectiveness and recommends high quality trials to be undertaken.\nThus, this limited evidence base became the impetus to undertake a pilot trial in preparation for a larger, definitive trial of NPWT in obese women undergoing elective CS.\n2. Aim\nThe aim of this pilot randomized controlled trial (RCT) was to assess the feasibility of conducting a larger trial in terms of measurement of potential outcomes, recruitment, intervention fidelity and retention.\nThe hypothesis tested was \u201cIn obese women undergoing elective CS, those who receive a NPWT dressing will have significantly better outcomes than those receiving the standard dressing\u201d.\nData from this pilot study will assist researchers to determine sample size requirements and potential primary and secondary outcomes to be used in a larger, definitive trial.\n3. Methods\nA parallel group pilot RCT was undertaken (Australian and New Zealand Trial Registation number ACTRN12612000171819).\nEthics approval was granted by the hospital and university office of human research ethics committees.\nAn interim analysis of the first 48 women enrolled in this pilot showed 87% of women approached agreed to be part of the trial and there was 94.2% retention.\nAll women received the dresssings they were randomized to, and inter-rater reliability for the outcome SSI was 0.87 (citation masked for blinded peer review).\n3.1. Participants and Setting\nThis study took place in one Australian hospital.\nAs this was a pilot study, the target sample size was set at 80\u2013100.\nInclusion criteria were: (i) women booked for elective CS surgery; (ii) recorded pre-pregnancy BMI of \u226530 and (iii) able to provide written informed consent.\nExclusion criteria were: (i) women whose condition changes to warrant an urgent or emergency CS; (ii) previous participation in this trial; (iii) existing infection after admission to hospital and prior to CS; and (iv) unable to speak or understand English with no interpreter present.\n3.2. Outcomes\nThe primary outcome for this study was surgical site infection (SSI), as defined by the Centers for Disease Control and Prevention.\nSecondary outcomes included: (1) type of SSI\u2013superficial incision, deep incision or organ/body space using the CDC criteria; (2) wound complications (i.e., dehiscence, haematoma, bleeding, seroma, blisters); (3) hospital length of stay (HLOS); and (4) hospital readmissions (within 28 days).\nAll outcomes except HLOS and readmission were assessed daily while the women were in hospital and weekly for 4 weeks after hospital discharge.\nNo changes in the proposed trial outcomes occurred during the study.\n3.3. Intervention and Control\nAt the completion of skin closure, those randomly allocated to the NPWT had, a PICO\u2122 (Smith and Nephew, Hull, UK) applied by the obstetrician under sterile conditions.\nWomen in the control arm had, Comfeel Plus\u00ae (Coloplast, City, Denmark) dressing applied per manufacturer\u2019s recommendations after skin closure.\nIn both groups, the dressing remained in place until day 4, unless it became soiled or dislodged, in which case a new dressing of the same type was applied.\nTo ensure consistency, obstetricians, nurses and midwives received trial-specific education (Negative Pressure Wound Therapy (NPWT) and Comfeel Plus standard dressing).\nThe research assistant (RA) was available to clinical staff via telephone and in person to provide ongoing training and support about correct use of the dressings as well as monitor dressing changes and complete documentation daily to assess protocol compliance and outcomes.\n3.4. Procedure\nPotential participants were screened between the 32nd and 38th weeks of gestation by either the attending doctor or midwife in the antenatal clinic.\nAn RA who was a Registered Nurse recruited participants during their 36th week outpatient visit, providing potential participants with an information summary of the research.\nIf women agreed to participate, they signed a consent form.\nOn the day of the elective CS, the RA confirmed ongoing consent from the women.\nRandomization was via a computer-generated 1:1 ratio, and had blocks of randomly varying sizes.\nRandomization occurred by the RA in the operating room.\nA centralized web-based randomization service was accessed which ensured allocation concealment.\nThe RA collected all outcome data daily while the women were in hospital.\nFollowing hospital discharge women were contacted weekly until the study end-point, at 28 days.\nField notes were recorded that provided narrative information regarding the conduct of the trial and the care women received.\nA separate person, experienced in assessing for SSI, assessed the outcome SSI and was blinded to group allocation.\nAssessment of the data for SSI occurred at two intervals during the course of the study, firstly data on 35 women was assessed prior to preliminary analsysis (9 months into the trial) and the remaining 52 women\u2019s data was examined on completion of the study.\nAll women had completed 28 days of data collection at time of outcome assessment.\n3.5. Data Analysis\nDescriptive and infererntial statistics were used to analyse the data.\nContinous variables were summarized using mean and standard deviation (SD) or median and inter quartile range (IQR) based on normality assumptions.\nNormal continous varibles were compared between the intervention and control groups, using independent t-test while those that were not normal were analysed using Mann Whitney U test.\nCategorical variables were described using frequency and percentages.\nTesting of hypotheses of categorical varibles were evaluated using Chi-square test or Fisher\u2019s exact test as appropriate.\nPrimary and secondary outcome variables were compared by computing the risk in each group and risk ratio (RR) and 95% confidence interval (CI).\nWe did not expect statistical signficance between the groups for the oucome measures but point estimates (RR) were expected to show the direction and approximate magnitude of effect, if the study were to have been sufficienty powered.\nWith the intention of conducting a larger trial, we used these data for a power calculation.\nMost data analyses were carried out using SPSS version 21, MedCalc was used for risk computations and confidence intervals, and PASS version 12 was used for sample size calcuations.\n4. Results\nRecruitment occurred from July 2012 to April 2014.\nAs identified in the flow diagram (Figure 1), a total of 111 women were recruited but 19 (17%) were subsequently excluded prior to randomisation.\nThere was incomplete outcome data on 5 (5%) women, therefore the final analysis included 87 women.\nFour of the five women dropped out before the final data collection point and the fifth was transferred inter-hospital and had no outcome data.\nAll women in the intervention and all women in the control group received the dressing to which they were randomized.\nOne (2.2%) woman in the intervention group had a subsequent dressing change that resulted in a standard (rather than NPWT) dressing being used for the replacement (contamination).\nIn successive dressing changes, none of the control women received the intervention (NPWT) dressing.\nWomen were analysed according to their randomized dressing irrespective of whether they received a different dressing to the group to which they were allocated during the study period.\nTable 1 displays the characteristics of the sample.\nWhile the two groups were similar, the length of surgery was longer in the control group and this group also had more smokers.\nThe number of women requiring a dressing change was significantly different, with the 5/43 (11.6%) of the control group and 16/44 (36.3%) of the intervention group having at least one dressing change (p = 0.006).\nTable 2 shows the comparison of primary and secondary outcomes.\nIn total, 27.9% of the control group and 22.7% of the intervention group had a SSI, but this difference did not reach statistical significance, due to smaller sample size.\nHowever the RR of 0.81 (95% CI 0.39; 1.68) shows the risk of SSI was almost 20% lower in the NPWT group (10/44) compared to the control group (12/43) which may be clinically important, although not statistically significant with this sample size.\nAs identified in Table 2, there were no statistically significant differences in the other outcomes, although there was a trend towards reduced bruising but increased blistering in NPWT.\nNo women in either group experienced a seroma or dehiscence.\nFigure 2 provides a power curve based on the SSI data, demonstrating the various sample sizes required for trials powered at 80%, 90% and 95%.\nA sample size of 392 per group would be required to find a statistically significant difference in SSI between the two groups with 90% power.\n5. Discussion\nThis pilot study was undertaken to assess the feasibility of a larger definitive trial comparing the use of NPWT to standard dressings including the likely sample size required to detect a significant difference in SSI.\nOur findings showed a trend towards fewer SSI in the NPWT group, although this difference was not statistically significant, likely due to the small sample.\nHowever, if this trend were to be supported in an adequately powered trial, it could have important implications for clinical practice.\nPreventing some SSIs has a number of benefits including improving women\u2019s recovery from surgery, decreasing the need for treatment and hospital length of stay as well as saving valuable health care dollars.\nArgubly, despite NPWT being more expensive than conventional dressings, the benefits in preventing SSI may outweigh these dressing costs.\nIt would be important for larger trials to incorporate some form of economic analysis, given the limited evidence in this area.\nThis recommendation is also supported by a Cochrane Review of NPWT, which only identified one abstract of a study that considered costs but the full paper was not available.\nThere is however some costing research in other patient populations, with one study of patients with diabetic foot wounds finding that the average cost to achieve healing was less in the NPWT group (although this was a small study).\nWithout good quality RCT evidence of effect or cost-benefit, it would be premature to recommend using NPWT for surgical wounds in obese women undergoing elective CS.\nDespite our sample being randomized, there were group differences in both the smoking status of the women and the length of surgery, with women in the control group reporting smoking more and their surgery was longer.\nBoth factors are recognised risk factors for SSI, thus it is always possible that these results explain the trend towards more SSI in the control group.\nIt would be expected that randomising women in a trial with a larger sample size would see at least the difference in smoking status between the two groups disappear.\nIn a larger sample, the length of surgery, a potential confounder, could be handled statistically, by entering it into the analysis as a covariate.\nThe independent effect of NPWT could then be assessed while controlling for length of surgery.\nHowever, the clinical significance of an average of 10 min longer surgery is unknown.\nThe NPWT dressing was acceptable to both the women in the trial and the healthcare teams providing care.\nFor example we were able to enrol 87% of the women approached to participate in the study and all women allocated to the NPWT received it (i.e., clinicians did not prevent the NPWT dressing from being used in women recruited to the study).\nHowever, a general lack of familiarity with the NPWT dressing meant that ongoing education of both the medical and nursing staff was required.\nAdditionally, at times when the research team was not available, some NPWT dressings were changed and in one instance, was replaced by the standard (control) dressing.\nIn fact, 36% of the NPWT group had at least one dressing change, as compared to 12% in the control group.\nThis unexpected finding requires further understanding as it has implications for future research, treatment costs and clinical practice.\nFor example, this may indicate the need for additional staff training to ensure unnecessary dressing changes do not occur or it may indicate some other factor that can be addressed such as poor application technique.\nAn important consideration for RCTs is the extent to which the control and intervention groups receive similar care.\nIn this hospital, women having elective CS followed a standardised care pathway during their admission, which should have standardised important aspects of care.\nHowever, the RA observed and documented subtle differences in certain aspects of surgical care such as antibiotic timing in theatre, surgeons\u2019 preference for wound closure including suture materials, and type of standard dressing.\nEach of these issues could influence the findings of a definitive trial.\nClinical practice guidelines and systematic reviews recommend pre-operative prophylactic antibiotics for clean contaminated wounds such as CS.\nIncluding these recommendations during education sessions related to a larger trial may help to standardize practice.\nIn terms of wound closure, the 2008 National Institute of Health and Clinical Excellence guidelines note there is no high quality evidence to recommend one practice over another, but a recent meta-analysis found closure with staples had a twofold higher risk of wound infection than closure with subcuticular sutures.\nThus, including the use of sutures rather than staples for wound closure in future trial protocols could reduce the potential impact of this potential confounder, although a small study of 63 women undergoing CS found surgeons preferred staples over sutures.\nFinally, in terms of what dressings were used in the control group, a 2011 Cochrane review found no evidence to suggest one dressing type was better than others for the prevention of SSI.\nIt could be that there are differences that have yet to be demonstrated.\nIn future trials, standardizing the dressing type in the control group may be prudent, but some variation in clinical practice does not mean that subsequent trials without standardisation cannot be completed in a rigorous manner.\nIt does however suggest that future trials should be a pragmatic (versus explanatory) trial.\nSackett suggests that pragmatic trials answer the question \u201cDoes this treatment improve patient-important outcomes when applied by typical clinicians to typical patients?\u201d.\nThere are a number of features of pragmatic trials that make them particularly well suited for testing interventions such as wound dressings in the clinical environment.\nFirst, pragmatic trials focus on effectiveness in usual circumstances or practice.\nSecond, the intervention is applied in a flexible way, as it would be in clinical practice.\nFinally, the findings of the research are generally directly relevant to patients, clinicians and decision makers.\nAs part of the feasibility component of this trial, we generated information to estimate a range of possible sample sizes for the primary outcome of SSI, required for a larger definitive trial.\nUsing this approach reflects best practice and has added to the methodological rigor of this pilot trial.\nHowever, as there may be some uncertainties around sample size estimates obtained through pilot trials, it is also recommended to discuss estimates with clinicians to obtain additional information around clinically meaningful effect sizes.\nOur results indicate that a definitive trial would require an overall sample size of 784 (i.e., 392 per group in a two arm trial) to have 90% power to find a difference between groups if the primary outcome was the absence or presence of a SSI.\nClearly, if SSI remains the primary outcome it will require a multi-site study.\nIn our pilot study we measured a number of other complications including bleeding, bruising, blister, seroma and dehiscence, but only noted whether they were present or absent and not the extent of each.\nClinically, a small amount of bleeding, bruising or blistering would likely have little effect on the women or their ongoing care, but if more extensive, would likely require corrective action.\nInterestingly, there were no cases of either seroma or dehiscence reported but it is always possible this could occur in a larger sample.\nThere was a trend towards more blistering in the NPWT group but none in the control group developing blisters.\nIn one trial of 60 patients undergoing total knee arthroplasty, the rate of blisters in the NPWT group was so high (63%; RR 18.3 95% CI 4.3\u201377.6), the trial was stopped.\nA recent review suggests skin blisters are common in orthopaedic surgery when adhesive dressings are used because of the swelling/oedema that occurs.\nClearly, blistering is an important safety consideration for both future trials and when the NPWT dressings are used in clinical practice.\nAn alternative option for selecting the primary outcome measure for the definitive trial is to develop a \u201ccomposite\u201d outcome such as \u201cany wound complication\u201d used in some previous research.\nA composite measure involves aggregating the scores of several variables into an overall score.\nThe use of composite measures versus single outcome measures has been debated for some time.\nUsing a composite measure of \u201cany wound complication\u201d as the primary outcome in a definitive trial would likely result in a smaller sample size being required to demonstrate statistical significance.\nNonetheless, there are also a number of limitations to such an approach.\nFor example, grouping more serious complications like SSI and wound dehiscence with minor blistering or bleeding could make interpretation of the research findings including their clinical relevance difficult.\nOther considerations for the larger definitive trial include standardizing training across sites especially proper application of the NPWT dressing, a clear monitoring plan to ensure the trial is proceeding as planned and additional data collection about site specific processes.\nGiven the challenges associated with the real-world clinical settings, and the large number of health care providers involved in the clinical management of this population, using a pragmatic approach to trial design is appropriate.\n6. Conclusions\nThis pilot study of 87 women showed that a larger definitive trial is feasible.\nAlmost 90% of women approached agreed to be in the trial and 95% completed it.\nA sample size of 784 women would be required to detect a 20% difference in SSI at 90% power.\nA pragmatic trial, and associated process evaluation may be an appropriate approach if a definitive trial is undertaken in the future.\nParticipant Flow Diagram.\nPower Curve.Effect size = 0.10, Baseline rate = 0.30, Alpha = 0.05, 2 sided.\n\nCharacteristics of the Sample (n = 87).\nCharacteristic | Intervention Group n = 44 | Control Group n = 43 | p-Value\n | Median (IQR) | Median (IQR) | \nAge | 30.6 (5.5) | 30.7 (5.0) | 0.925\nBody Mass Index | 35.7 (4.5) | 36.8 (5.8) | 0.538\nLength of surgery (minutes) | 45.0 (16.0) | 53.0 (16.0) | 0.002\n | Frequency (%) | Frequency (%) | \na Co-morbidities (yes/no) | 30 (68.1) | 30 (69.7) | \u22120.145 z score\nNumber of co-morbidities |  |  | \n0 | 14 (31.8) | 13 (30.2) | \n1 | 18 (40.9) | 21 (48.8) | \n2 | 10 (22.7) | 5 (11.6) | \n3 | 1 (2.3) | 3 (7.0) | \n4 | 1 (2.3) | 1 (2.3) | \nPrevious CS (yes/no) | 37 (84.0) | 40 (93.0) | \u22120.188 z score\nNumber of previous CS |  |  | \n0 | 7 (15.9) | 3 (7.0) | \n1 | 24 (54.5) | 28 (65.1) | \n2 | 7 (15.9) | 11 (25.6) | \n3 | 5 (11.4) | 0 (0.0) | \n4 | 1 (2.3) | 1 (2.3) | \nSmoker | 3 (6.8) | 10 (23.3) | 0.032\nDiabetic (any type) | 13 (29.5) | 12 (27.9) | 0.290\n\na Comorbidities included: Anaemia, Diabetes Mellitus, Gestational Diabetes, Hypercholesterol, Hypertension, Immuno-compromised, Nutritional deficiency, Thromboembolytic disease, Smoking, Other.\n\nRelative Risk of Outcomes (n = 87).\nOutcome | Interventionn = 44Frequency (%) | Controln = 43Frequency (%) | RR | 95% CI | p Value\nSurgical site infection | 10 (22.7) | 12 (27.9) | 0.81 | 0.39\u20131.68 | 0.579\nType of SSI |  |  |  |  | a 0.928\nSuperficial incision | 5 (11.4) | 7 (16.3) | 0.70 | 0.24\u20132.03 | 0.509\nDeep incision | 4 (9.1) | 4 (9.3) | 0.98 | 0.26\u20133.66 | 0.972\nOrgan/space | 1 (2.3) | 1 (2.3) | 0.98 | 0.06\u201315.13 | 0.987\nNumber of wound complications (excluding SSI) | 6 (13.6) | 6 (14.0) | 0.98 | 0.34\u20132.79 | 0.966\nNumber of complications (including SSI) | 14 (31.8) | 17 (39.5) | 0.80 | 0.46\u20131.42 | 0.454\nType of wound complication |  |  |  |  | a 0.147\nBleeding | 1 (2.3) | 1 (2.3) | 0.98 | 0.06\u201315.13 | 0.987\nBruising | 1 (2.3) | 4 (9.3) | 0.24 | 0.03\u20132.10 | 0.199\nb Other | 4 (9.1) | 1 (2.3) | 3.91 | 0.46\u201333.58 | 0.214\nHospital readmission | 1 (2.3) | 1 (2.3) | - | - | 0.987\n | Median (IQR) | Median (IQR) |  |  | \nHospital length of stay (days) | 3.0 (1.0) | 3.0 (1.0) | - | - | 0.724\n\na May be inaccurate due to number of cells with small expected values; b All other complications in the intervention group were blisters and the one other complication in the control group was erythema.",
        "label": "unclear",
        "id": "task4_RLD_test_694"
    },
    {
        "paper_doi": "10.1371/journal.pmed.1002357",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Open-label RCT in Haiti\n\n\nParticipants: 762 participantsInclusionAge >=18 years.Ability and willingness of participant to give written informed consent.CD4 cell count <= 500 cells/mm3.WHO stage 1 or 2 disease.ExclusionAny use of ART in the past.Pregnancy or breastfeeding at the screening visit.Psychologically unprepared to start ART, based on ART readiness survey.Plans to transfer care to another clinic during the study period.WHO stage 3 or 4 disease.\n\n\nInterventions: InterventionDay of presentation: HIV testing, CD4 count, physician evaluation, first adherence counselling visit with social worker; physician visit for ART initiation and ART initiation.Days 3, 10 and 17: first, second and third adherence counselling visits with social worker; physician visit to assess for opportunistic infections and provide adherence counselling.Day 24 and Week 7: physician visits for medical assessment and adherence counselling.Participants also had prophylactic treatment with trimethoprim-sulphamethoxazole and isoniazid. Field workers phoned participants who missed a visit and attempted a home visit for those not reachable by phone.ControlDays 7, 14: first and second adherence counselling visits with social worker; physician visit to assess for opportunistic infections and provide adherence counselling.Day 21: third adherence counselling visit with social worker; physician visit for ART initiation.Week 5: fourth adherence counselling visit with social worker; physician visit to assess for opportunistic infections and provide adherence counselling.Week 7: physician visit for medical assessment and adherence counselling.\n\n\nOutcomes: Retention in care at 12 months, viral suppression at 12 months, adherence to ART, uptake of ART, cost-effectiveness of standard and same-day ART initiation\n\n\nNotes: \n\n",
        "objective": "To assess the effects of interventions for rapid initiation of ART (defined as offering ART within seven days of HIV diagnosis) on treatment outcomes and mortality in people living with HIV. We also aimed to describe the characteristics of rapid ART interventions used in the included studies.",
        "full_paper": "Background\nAttrition during the period from HIV testing to antiretroviral therapy (ART) initiation is high worldwide.\nWe assessed whether same-day HIV testing and ART initiation improves retention and virologic suppression.\nMethods and findings\nWe conducted an unblinded, randomized trial of standard ART initiation versus same-day HIV testing and ART initiation among eligible adults \u226518 years old with World Health Organization Stage 1 or 2 disease and CD4 count \u2264500 cells/mm3.\nThe study was conducted among outpatients at the Haitian Group for the Study of Kaposi\u2019s Sarcoma and Opportunistic infections (GHESKIO) Clinic in Port-au-Prince, Haiti.\nParticipants were randomly assigned (1:1) to standard ART initiation or same-day HIV testing and ART initiation.\nThe standard group initiated ART 3 weeks after HIV testing, and the same-day group initiated ART on the day of testing.\nThe primary study endpoint was retention in care 12 months after HIV testing with HIV-1 RNA <50 copies/ml.\nWe assessed the impact of treatment arm with a modified intention-to-treat analysis, using multivariable logistic regression controlling for potential confounders.\nBetween August 2013 and October 2015, 762 participants were enrolled; 59 participants transferred to other clinics during the study period, and were excluded as per protocol, leaving 356 in the standard and 347 in the same-day ART groups.\nIn the standard ART group, 156 (44%) participants were retained in care with 12-month HIV-1 RNA <50 copies, and 184 (52%) had <1,000 copies/ml; 20 participants (6%) died.\nIn the same-day ART group, 184 (53%) participants were retained with HIV-1 RNA <50 copies/ml, and 212 (61%) had <1,000 copies/ml; 10 (3%) participants died.\nThe unadjusted risk ratio (RR) of being retained at 12 months with HIV-1 RNA <50 copies/ml was 1.21 (95% CI: 1.04, 1.38; p = 0.015) for the same-day ART group compared to the standard ART group, and the unadjusted RR for being retained with HIV-1 RNA <1,000 copies was 1.18 (95% CI: 1.04, 1.31; p = 0.012).\nThe main limitation of this study is that it was conducted at a single urban clinic, and the generalizability to other settings is uncertain.\nConclusions\nSame-day HIV testing and ART initiation is feasible and beneficial in this setting, as it improves retention in care with virologic suppression among patients with early clinical HIV disease.\nTrial registration\nThis study is registered with ClinicalTrials.gov number NCT01900080\nIn a randomized unblinded trial in Port-au-Prince, Haiti, Serena Koenig and colleagues investigate whether initiating ART on the day of HIV diagnosis improved retention in care and viral suppression.\nAuthor summary\nWhy was this study done?\nMultiple visits for counseling, laboratory testing, and other procedures to prepare patients for initiation of antiretroviral therapy (ART) are burdensome and contribute to the high rate of attrition during the period from HIV testing to ART initiation.\nThe World Health Organization (WHO) recently changed their guidelines to recommend ART for all persons living with HIV, facilitating ART initiation.\nThis study was conducted to determine if ART initiation on the day of HIV diagnosis could improve treatment initiation rates, retention in care, and HIV viral suppression for patients with asymptomatic or minimally symptomatic HIV disease.\nWhat did the researchers do and find?\nWe randomly assigned patients who presented for HIV testing at a clinic in Port-au-Prince, Haiti to standard ART initiation or same-day HIV testing and ART initiation (356 in the standard and 347 in the same-day groups).\nThe standard group had 3 weekly visits with a social worker and physician and then started ART 21 days after the date of HIV diagnosis; the same-day ART group initiated ART on the day of HIV diagnosis.\nAll participants in the same-day ART group and 92% of participants in the standard group initiated ART.\nAt 12 months after HIV testing, a higher proportion of participants in the same-day ART group were retained in care (80% versus 72%), and a higher proportion were retained in care with viral load <50 copies/ml (53% versus 44%) and viral load <1,000 copies/ml (61% versus 52%).\nWhat do these findings mean?\nThis study demonstrates that it is feasible to initiate ART on the day of HIV diagnosis for patients with early HIV clinical disease and that same-day treatment leads to increased ART uptake, retention in care, and viral suppression.\nThough same-day ART initiation improves outcomes, retention in care and viral suppression remain suboptimal, so further interventions to maximize long-term outcomes will be essential.\nThe study is limited by being conducted at 1 clinic in urban Haiti.\nFurther study will be necessary to determine if this strategy will be effective in other settings.\nIntroduction\nThe Joint United Nations Programme on HIV/AIDS (UNAIDS) 90-90-90 targets state that 90% of HIV-infected persons know their status, 90% initiate antiretroviral therapy (ART), and 90% achieve virologic suppression by the year 2020 to curb the AIDS epidemic.\nIn 2015, the World Health Organization (WHO) updated their guidelines to recommend ART for all persons living with HIV based on evidence that earlier treatment improves outcomes and decreases transmission.\nTo achieve these goals, patients must be promptly linked to HIV services, initiated on ART, and retained in lifelong care.\nAttrition rates are particularly high during the period from HIV testing to ART initiation, with one-quarter to one-third of patients lost in the process of starting ART.\nEven if many of these patients re-engage in care at a later date, they will return with more advanced disease.\nThough there are many factors that contribute to pretreatment attrition, the current standard of care in most settings, which requires multiple sequential visits for HIV testing and counseling, laboratory testing, and adherence counseling prior to ART initiation, creates barriers to treatment initiation.\nAs of June 2016, WHO guidelines note inadequate evidence to support a recommendation of same-day HIV testing and ART initiation.\nHowever, the availability of point-of-care tests, the fact that CD4 cell counts are no longer necessary prior to ART initiation, and the provision of same-day counseling can accelerate treatment initiation, potentially reducing attrition.\nWe conducted a randomized trial in Haiti to determine whether same-day HIV testing and ART initiation, as compared with standard ART initiation, improves retention in care with viral suppression.\nMethods\nStudy design and setting\nWe conducted an unblinded, randomized controlled trial of standard ART initiation versus same-day HIV testing and ART initiation among HIV-infected adults at the Haitian Group for the Study of Kaposi\u2019s Sarcoma and Opportunistic infections (GHESKIO) in Port-au-Prince, Haiti.\nHaiti is the poorest country in the Western Hemisphere, with adult HIV prevalence of 1.7%.\nGHESKIO is a Haitian nongovernmental organization and the largest provider of HIV care in the Caribbean, treating up to 700 patients per day for HIV and/or tuberculosis (TB).\nAll care is provided free of charge.\nThe study was approved by the institutional review boards at Partners Healthcare, GHESKIO, Weill Cornell Medical College, and Florida International University.\nSee supporting information files S1 Text for the study protocol and S2 Text for the CONSORT checklist.\nParticipants\nParticipants were recruited from the HIV voluntary counseling and testing center at GHESKIO from August 2013 to October 2015.\nThey received HIV testing and posttest counseling; those with a positive HIV test were referred for same-day physician evaluation, CD4 count (FACS Count, Becton-Dickinson, Franklin Lakes, New Jersey), WHO staging, and chest radiograph.\nPatients were eligible for study inclusion if they were infected with HIV-1, \u226518 years of age, and had WHO Stage 1 or 2 disease and CD4 count \u2264500 cells/mm3.\nInitially, enrollment was limited to patients with CD4 count \u2264350 cells/mm3, but in February 2014, the cutoff was increased to \u2264500 cells/mm3 in response to revised WHO and Haitian guidelines.\nPatients were excluded if they were already aware of their HIV diagnosis, had received ART previously, were pregnant or breastfeeding, lived outside of the greater Port-au-Prince metropolitan area, planned to transfer care during the study period, or failed to demonstrate preparedness on an ART readiness survey, which was administered by a social worker prior to study enrollment.\nThe survey includes a 5-point scale, with respondents ranking their preparedness from \u201cnot at all ready\u201d to \u201ccompletely ready\u201d in response to 7 questions.\nStudy inclusion required a response of \u201csomewhat ready\u201d or \u201ccompletely ready\u201d for all 7 questions (S3 Text).\nRandomization and masking\nAfter the patients had provided written informed consent, the study team performed a screening evaluation for study exclusion criteria, and eligible participants were enrolled and randomized on the day of HIV testing.\nParticipants were randomly assigned with the use of a computer-generated random-number list to either standard ART or same-day ART initiation in a 1:1 ratio, with allocation concealment.\nThe randomization sequence was generated by a computer in the GHESKIO data management unit by a data manager who had no other involvement in study procedures.\nParticipants were enrolled in the study and assigned to groups by a study physician.\nParticipants, site personnel, and study statisticians were not masked to group assignment.\nProcedures\nAfter randomization, the standard group participants received ART initiation procedures that mirror national guidelines.\nParticipants were referred to return on Day 7 for baseline laboratory tests (creatinine, alanine aminotransferase, aspartate aminotransferase, complete blood count, purified protein derivative [PPD]), physician evaluation, and counseling with a social worker.\nOn Day 10, they received interpretation of PPD results, and on Days 14 and 21, they were seen by a physician and social worker for additional counseling, test results, and ongoing evaluations for opportunistic infections.\nParticipants started ART on Day 21 and had an additional social worker and physician visit at Week 5 (Fig 1).\nThe ART regimen was the same as that for nonstudy patients at GHESKIO.\nFirst-line therapy included a single combination tablet including tenofovir disoproxil fumarate, lamivudine, and efavirenz.\nThe same-day ART group had identical laboratory tests as the standard ART group, a 30-minute counseling session with a social worker, and physician evaluation, and then initiated the same ART regimen as the standard ART group.\nThey returned on Day 3 for physician and social worker visits and receipt of baseline laboratory test results; those with creatinine clearance <50 mL/minute as calculated by the Cockcroft-Gault equation were switched from tenofovir to zidovudine or abacavir.\nThey returned on Days 10 and 17 for additional physician and social worker visits and on Day 24 for a physician visit.\nThe same number of scheduled physician visits and counseling sessions were provided to each group so that the only difference in care was in the schedule of visits during the first 5 weeks of the study and the timing of ART initiation.\nAll care was delivered by GHESKIO clinic staff, and the same providers (physicians, nurses, social workers, pharmacists, and field workers) cared for both groups.\nA counseling manual was followed with an outline for the social workers to follow at each scheduled counseling visit; these were identical between groups, except for the timing of ART initiation, and each session took about 30 minutes.\nAll counseling was provided for individual patients, rather than for groups.\nThe counseling sessions were audiotaped and systematically evaluated for quality control purposes.\nIf a participant in either group missed a study visit that included a scheduled social worker counseling session, the counseling was provided at the next visit.\nParticipants in both groups had monthly physician visits throughout the follow-up period and received the same package of services provided to all HIV-infected patients at GHESKIO, including prophylactic treatment with trimethoprim-sulfamethoxazole and isoniazid.\nField workers phoned patients who missed a visit and attempted a home visit for those not reachable by phone.\nParticipants received a transportation subsidy of 100 Haitian gourdes (US$1.70) per visit.\nOutcomes\nThe primary endpoint was retention in care with HIV-1 RNA <50 copies/ml at 12 months after HIV testing.\nRetention was defined as attending the 12-month visit (1 clinic visit between 12 and 15 months after HIV testing).\nLost to follow-up (LTFU) was defined as failure to attend the 12-month visit.\nDeaths were ascertained by review of medical records or report from family members.\nA National Institutes of Health Division of AIDS Expedited Adverse Event Form was filled out within 48 hours after the study team became aware of any death.\nTransfers were ascertained by confirmation that the participant was receiving care at a different site.\nSecondary outcomes include survival, ART initiation, retention in care with HIV-1 RNA <1,000 copies/ml at 12 months after HIV testing, adherence as measured by pharmacy refill records and self-report, and cost and cost-effectiveness of standard and same-day ART; the adherence and cost-effectiveness evaluations will be reported in separate manuscripts.\nStatistical analysis\nDemographic, clinical, and laboratory data from the electronic medical record and study forms were de-identified, entered into an Excel spreadsheet, and exported into Stata v14 software (StataCorp, 2011, College Station, Texas) for analysis.\nAfter study completion, all participants who were LTFU were recontacted to determine their vital status.\nThe study was powered to detect a 10% absolute difference in the rate of retention with virologic suppression between the 2 groups at 12 months after enrollment (65% in the standard and 75% in the same-day ART group).\nAt the \u03b1 = 0.05 significance level, we estimated that we would need to enroll 349 participants per group (698 in total) to achieve 80% power to detect this difference.\nBecause patients who transferred during the study period were excluded, we increased the total sample size to 762 participants.\nFor all analyses, a modified intention-to-treat approach was used, in which all patients were analyzed according to their assignment group, excluding patients who transferred to another facility during the follow-up period, according to protocol.\nBaseline characteristics were summarized using simple frequencies and proportions and medians with interquartile ranges (IQRs) stratified by treatment arm.\nAmong participants who died, baseline CD4 count was compared using the Wilcoxon rank-sum test.\nWe compared the proportion of participants who were retained in care with HIV-1 RNA <50 copies/ml (primary endpoint), retained with HIV-1 RNA <1,000 copies/ml, retained regardless of HIV-1 RNA, initiated ART, and died (secondary endpoints) at 12 months after enrollment using a chi-square test.\nWe conducted multivariable logistic regression including all covariates listed in Table 1 to control for any residual confounding.\nWe present unadjusted and adjusted risk ratios (RR) with 95% confidence intervals.\nBecause of the change in enrollment criteria mid-study, we conducted a sensitivity analysis that included only the participants who met the original enrollment criteria of CD4 count \u2264350 cells/mm3.\nIn response to a reviewer\u2019s request, we also plotted retention in care, regardless of viral load, for both groups and compared the distributions with the log-rank test.\nThe study is registered with ClinicalTrials.gov number NCT01900080.\nResults\nA total of 821 patients were screened, and 762 were enrolled in the study and underwent randomization (Fig 2).\nAfter randomization, 59 participants (28 in the standard ART and 31 in same-day ART group) transferred to another clinic and were excluded from all analyses, as per protocol.\nThe median age was 37 years old (IQR: 30\u201345 years), 347 (49%) were women, and the median CD4 count was 248 cells/mm3 (IQR: 148, 345).\nOf the 356 participants in the standard group, 256 (72%) were retained in care, 20 (6%) died, and 80 (23%) were LTFU (Table 2).\nAmong the 256 participants retained in the standard ART group, 156 (61% of retained and 44% overall) had HIV-1 RNA <50 copies/ml.\nOf the 347 participants in the same-day ART group, 277 (80%) were retained in care, 10 (3%) died, and 60 (17%) were LTFU.\nAmong the 277 participants retained in the same-day ART group, 184 (66% of retained and 53% overall) had HIV-1 RNA <50 copies/ml.\nThe unadjusted RR of being retained in care at 12 months and achieving HIV-1 RNA <50 copies/ml was 1.21 (95% CI: 1.04, 1.38; p = 0.015) for the same-day ART group compared to the standard group (Table 3); the adjusted RR for this comparison was 1.24 (95% CI: 1.06, 1.41; p = 0.008).\nIn the standard ART group, 184 (72% of retained and 52% overall) participants who were retained in care had HIV-1 RNA <1,000 copies/ml.\nIn the same-day ART group, 212 (77% of retained and 61% overall) participants who were retained in care had HIV-1 RNA <1,000 copies/ml.\nThe unadjusted RR of being retained in care at 12 months and achieving HIV-1 RNA <1,000 copies/ml was 1.18 (95% CI: 1.04, 1.31; p = 0.012) for the same-day ART group compared to the standard ART group (Table 3); the adjusted RR for this comparison was 1.20 (95% CI: 1.05, 1.33; p = 0.008).\nIn the sensitivity analysis that included only participants who met the original enrollment criteria (CD4 count \u2264350 cells/mm3), the adjusted RR of being retained in care at 12 months and achieving HIV-1 RNA <50 copies/ml was 1.19 (95% CI: 0.99, 1.38; p = 0.060), and the adjusted RR of being retained in care at 12 months and achieving HIV-1 RNA < 1,000 copies/ml was 1.18 (95% CI: 1.01, 1.34; p = 0.035).\nVital status at the end of the study was known for 328 (92%) participants in the standard ART group and 329 (95%) in the same-day ART group.\nThe unadjusted RR for mortality was 0.51 (95% CI: 0.24, 1.08; p = 0.073) for the same-day group compared to the standard group; the adjusted RR for this comparison was 0.43 (95% CI: 0.19, 0.94; p = 0.033).\nIn the sensitivity analysis that included only participants with CD4 count \u2264350 cells/mm3, the adjusted RR for mortality was 0.41 (95% CI: 0.18, 0.93; p = 0.033).\nAmong the participants who died, the median baseline CD4 count was 100 cells/mm3 (IQR: 45, 192) in the standard and 207 cells/mm3 (IQR: 112, 291) in the same-day ART group (p = 0.078).\nEight of 20 (40%) deaths in the standard ART group occurred in participants who were LTFU prior to ART, 8 (40%) deaths occurred in those LTFU after starting ART, and 4 (20%) occurred while in care; the causes of death for those in care were stroke, trauma, and cancer in 3, and the fourth had pain and died after seeing a traditional healer.\nThree of the 10 (30%) deaths in the same-day ART group occurred in participants who were LTFU after starting ART; among the 7 (70%) participants who died while in care, 1 of each died of stroke, pneumonia, malaria, renal failure, and sudden death, and 2 died of gastroenteritis.\nNo deaths for those in care were attributed to immune reconstitution syndrome or an opportunistic infection that was missed at ART initiation.\nIn Fig 3, the Kaplan-Meier curve plots the retention in care, regardless of viral load, for both groups.\nThe log-rank test comparing the curves between the standard and same-day ART group indicates a significant difference (p = 0.028).\nIn the same-day ART group, 344 of 347 (99%) participants started ART on the day of HIV testing, and the remaining 3 patients started ART within the subsequent week.\nDuring the Day 3 follow-up visit, 13 patients (4%) in the same-day ART group had adjustments in their ART regimens (replacement of tenofovir with zidovudine or abacavir) because they had creatinine clearance <50 mL/minute on baseline testing.\nIn the standard group, 281 (79%) participants initiated ART by Day 28, the end of the time window for the 3-week ART initiation visit.\nThirty-six (10%) standard group participants initiated ART from Day 29 to Day 90, and 12 (3%) initiated ART after Day 90 due to late or missed visits.\nTwenty-seven (8%) standard group participants never started ART during the study period because they were LTFU or died prior to initiating treatment.\nIsoniazid prophylaxis was initiated for 337 (95%) participants in the standard group and 340 (98%) in the same-day group.\nEight cases of TB were diagnosed during the first 3 months after ART initiation; 6 of these occurred in the standard group and 2 in the same-day ART group.\nDiscussion\nThe results of this randomized controlled trial show that among HIV-infected adults with early WHO Stage disease and CD4 count \u2264500 cells/mm3, same-day HIV testing and ART initiation, as compared to standard care, improves retention in care with virologic suppression and, in the multivariable analysis, decreases mortality.\nThese results are important given recent WHO 2016 guidelines stating the lack of evidence in support of same-day ART initiation.\nOur findings suggest that ART initiation as soon as possible after HIV testing may be beneficial for clinically stable patients.\nIn resource-poor settings with fragile delivery systems, such as Haiti, the provision of immediate support by care providers at the time of HIV diagnosis can have both structural and individual impact.\nIn addition to making treatment initiation logistically easier for patients, we believe that same-day counseling and ART initiation increase the sense of hope, optimism, and overall connectedness to the healthcare system for patients, which have been shown to be important for retention.\nOur findings are consistent with the results of the RapIT study, a randomized trial that included participants in South Africa with WHO Stage 3 or 4 disease or CD4 count \u2264350 cells/mm3.\nParticipants in the standard group in that study generally started ART at the sixth visit, and 72% of participants in the rapid group started ART on the day of study enrollment.\nRapid ART initiation resulted in a 17% improvement in retention and 13% improvement in viral suppression.\nA stepped-wedge cluster-randomized trial in Uganda found an increase in ART initiation within 2 weeks after eligibility by implementing a multicomponent intervention to streamline ART initiation that included training healthcare workers, providing point-of-care CD4 count testing platforms, eliminating mandatory multiple preinitiation sessions, and giving feedback to facilities on their ART initiation rates.\nA weighted proportion of 80% in the intervention group had started ART within 2 weeks after eligibility compared with 38% in the control group.\nA cohort study of same-day ART initiation in pregnant women in South Africa also found high rates of treatment initiation, with 91% initiating ART on the day of referral to the service.\nIn the intervention group of the Sustainable East Africa Research on Community Health (SEARCH) HIV test-and-treat study, a cluster-randomized controlled trial conducted in Kenya and Uganda, HIV-infected patients who were identified through community testing were referred to HIV care upon diagnosis and then offered immediate ART initiation; retention was high (89%) among patients newly linking to care.\nAt ART initiation, it is critical that patients are ready to start lifelong therapy, that TB screening is conducted, and that renal function is evaluated to avoid the use of tenofovir in patients with renal insufficiency.\nIn this study, ART readiness was remarkably high, with over 99% of patients screened for the study reporting they were ready to start lifelong ART.\nThis is a particularly significant and timely finding for the provision of recommended universal ART because the majority of people living with HIV have early clinical disease, and there has been prior concern that healthier patients may be less willing to accept lifelong therapy.\nMost patients with early clinical disease do not have TB symptoms (cough, fever, night sweats, or weight loss), so they do not require further work up to exclude TB, according to WHO guidelines.\nWith the exclusion of patients with a baseline chest x-ray that was suspicious for TB, we found that less than 1% of participants in the same-day ART group had TB that was missed at the time of ART initiation.\nWe found that 4% of participants in the same-day ART group had creatinine clearance <50 mL/minute; ART regimens were adjusted on Day 3 for these patients.\nBoth groups in our study received high-level care, with multiple counseling and physician visits in the first month, followed by monthly physician visits.\nAt the time the study was started, this was the standard of care in Haiti.\nHowever, this standard has shifted over the past few years towards decreased frequency of visits and nonphysician providers.\nWe believe that same-day ART can be provided with fewer follow-up visits if proper counseling is provided during the early period after ART initiation.\nHowever, clinic-level procedures play a major role in the effectiveness of accelerated ART initiation strategies, as illustrated in Malawi, where among nearly 22,000 pregnant women who started ART for mother-to-child prevention, LTFU rates ranged from 0% to 58% between facilities and were highest among women who initiated ART on the day of HIV testing at large clinics.\nThough lower than anticipated, retention in both groups in our study was higher than reports of standard ART initiation from other resource-poor settings.\nTwo studies from South Africa found that approximately one-third of patients remained in care from HIV testing through 12 months of ART, and systematic reviews of African studies have found high rates of pre-ART attrition.\nIn Haiti, data on pre-ART outcomes are limited, but 12-month retention after ART initiation is 73% nationwide.\nWe attribute the higher retention in our study in large part to faster ART initiation, even in the standard group, compared to many other HIV programs.\nWe surmise that retention would have been lower in the standard group if there had been longer delays in ART initiation.\nThe rates of retention with viral suppression in our study are lower than those reported from clinical trial cohorts, including at GHESKIO.\nIn the GHESKIO Clinical Trials Unit, with a median monthly average of 483 subjects participating in NIH-funded clinical trials, retention is 97%.\nWe attribute the lower retention and viral suppression rates in our study to 2 major reasons.\nFirst, nearly all patients meeting WHO stage and CD4 criteria were enrolled in the study on the day of HIV testing, including those with substantial barriers to retention in care and adherence.\nIn contrast, over one-third of patients are generally lost to care prior to ART initiation or enrollment in clinical trials.\nSecond, the care that was provided in this study was similar to that received by nonstudy patients at GHESKIO, with the aim of producing findings that could be reproduced in other resource-poor settings.\nIn order to achieve the UNAIDS 90-90-90 targets, it will be important to evaluate reasons for attrition and implement new strategies to improve retention in care.\nOne approach that has been successful in a cohort of nonresearch patients at GHESKIO has been expedited follow-up care, with fewer visits of shorter duration for clinically stable patients.\nStreamlined care has also been associated with high rates of retention in the SEARCH study, which is described above.\nOur study was conducted in a large urban clinic, which may limit the generalizability of our findings.\nIn addition, though our study included patients with early clinical disease, the CD4 counts in our population were lower than would be expected with the provision of universal ART.\nIt is possible that patients with higher CD4 counts may experience less benefit from same-day ART.\nIt is also noteworthy that we conducted a chest x-ray prior to enrollment; if same-day ART is provided without a chest x-ray, it is possible that TB cases will be missed.\nOur study was not blinded.\nAll participants in both groups received the same number of visits and the same retention plan, but we cannot exclude the possibility that awareness of study group impacted provider behavior.\nIn conclusion, in a population of asymptomatic or minimally symptomatic HIV-infected patients, same-day HIV testing and ART initiation decreased mortality and improved the rate of retention in care with virologic suppression compared with standard ART initiation.\nFurthermore, human and material resources provided to each group were similar, so same-day ART is not expected to increase treatment costs.\nThe new WHO recommendations to provide ART to all HIV-infected patients should facilitate same-day test and treat.\nStudy interventions for the standard ART and same-day ART groups.\nScreening, randomization, and follow-up.\nRetention in care by study group.\n\nBaseline characteristics of study participants by group.\nCharacteristic | Standard Group (n = 356) | Same-Day ART Group (n = 347)\nAge (years)\u2014Median (IQR) | 37 (30, 45) | 37 (29, 46)\nFemale sex\u2014no. (%) | 181 (51) | 166 (48)\nEducation\u2014no. (%)\n\u00a0\u00a0\u00a0\u00a0No school | 90 (25) | 93 (27)\n\u00a0\u00a0\u00a0\u00a0Primary school | 110 (31) | 111 (32)\n\u00a0\u00a0\u00a0\u00a0Secondary school or more | 156 (44) | 143 (41)\nIncome\u2014no. (%)\n\u00a0\u00a0\u00a0\u00a0No income | 92 (26) | 90 (26)\n\u00a0\u00a0\u00a0\u00a0>$0 to $1/day | 176 (49) | 159 (46)\n\u00a0\u00a0\u00a0\u00a0>$1 to $2/day | 67 (19) | 76 (22)\n\u00a0\u00a0\u00a0\u00a0>$2/day | 21 (6) | 22 (6)\nMarital status\u2014no. (%)\n\u00a0\u00a0\u00a0\u00a0Single | 71 (20) | 71 (20)\n\u00a0\u00a0\u00a0\u00a0Currently married/living with partner | 222 (62) | 211 (61)\n\u00a0\u00a0\u00a0\u00a0Formerly married | 63 (18) | 65 (19)\nWHO Stage\u2014no. (%)\n\u00a0\u00a0\u00a0\u00a0WHO Stage 1 | 117 (33) | 101 (29)\n\u00a0\u00a0\u00a0\u00a0WHO Stage 2 | 239 (67) | 246 (71)\nCD4 count (cells/mm3)\u2014Median (IQR) | 247 (150, 349) | 249 (143, 336)\nBody mass index\u2014Median (IQR)* | 21.6 (19.7, 23.9) | 20.9 (19.3, 23.5)\n\n* Body mass index differed significantly between the 2 groups (p = 0.025).\nART, antiretroviral therapy; IQR, interquartile range, WHO, World Health Organization.\n\nStudy outcomes by group.\nOutcome | Standard ART Group (n = 356) | Same-Day ART Group (n = 347) | Unadjusted Risk Difference (95% CI) | p-value\nPrimary Outcome\nRetained in care at 12 months with VL <50 copies/ml | 156 (43.8%) | 184 (53.0%) | 9.2% (1.8%, 16.6%) | 0.015\u2020\nSecondary Outcomes |  |  |  | \nRetained in care at 12 months with VL <1,000 copies/ml | 184 (51.7%) | 212 (61.1%) | 9.4% (2.1%, 16.7%) | 0.012\u2021\nRetained in care at 12 months, regardless of VL results | 256 (71.9%) | 277 (79.8%) | 7.9% (1.6%, 14.2%) | 0.014\u2020\u2020\nDied | 20 (5.6%) | 10 (2.9%) |  | \nLost to follow-up | 80 (22.5%) | 60 (17.3%) |  | \n\n\u2020 p-value comparing the proportion of all patients who were retained in care with viral load <50 copies/ml between the 2 arms.\n\u2021 p-value comparing the proportion of all patients who were retained in care with viral load <1,000 copies/ml between the 2 arms.\n\u2020\u2020 p-value comparing the proportion of all patients who were retained in care between the 2 arms.\nART, antiretroviral therapy; VL, viral load.\n\nUnadjusted and adjusted risk ratios of study outcomes.\n | Unadjusted | Adjusted for All Baseline Co-variates\n | RR | 95% CI | p-value | RR | 95% CI | p-value\n | Retained in care with viral load <50 copies/ml\nStandard ART Group | 1.0 |  |  | 1.0 |  | \nSame-Day ART Group | 1.21 | (1.04, 1.38) | 0.015 | 1.24 | (1.06, 1.41) | 0.008\n | Retained in care with viral load <1,000 copies/ml\nStandard ART Group | 1.0 |  |  | 1.0 |  | \nSame-Day ART Group | 1.18 | (1.04, 1.31) | 0.012 | 1.20 | (1.05, 1.33) | 0.008\n | Mortality during study period\nStandard ART Group | 1.0 |  |  | 1.0 |  | \nSame-Day ART Group | 0.51 | (0.24, 1.08) | 0.073 | 0.43 | (0.19, 0.94) | 0.033\n\nART, antiretroviral therapy; RR, risk ratio.",
        "label": "high",
        "id": "task4_RLD_test_92"
    },
    {
        "paper_doi": "10.5830/cvja-2016-035",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Participants \"were randomised to an intervention and control group using a web-based computer system that ensured assignment concealment\".\"Upon completion of the research period, the blinding was unveiled and data were presented for statistical analysis\".\n\n\nParticipants: 33 HIV-positive people with definite or probable tuberculous pericarditis at a secondary level hospital in the Northern Cape of South Africa.All participants received standard treatment according to the South African National Tuberculosis Management Guidelines, that is weight-adjusted antituberculosis drugs and oral corticosteroids for 4 weeks. Participants also had pericardial \"aspiration until dryness\", and antiretroviral therapy.\n\n\nInterventions: InterventionColchicine 1.0 mg per day for 6 weeks.ComparisonPlacebo for 6 weeks.Participants were followed up with serial echocardiography for 16 weeks.\n\n\nOutcomes: Primary outcomeConstrictive pericarditis.\n\n\nNotes: Study location: Kimberley, South Africa\n\n",
        "objective": "To assess the effects of treatments for tuberculous pericarditis.",
        "full_paper": "Summary\nIntroduction\nTuberculous (TB) pericarditis carries significant mortality and morbidity rates, not only during the primary infection, but also as part of the granulomatous scar-forming fibrocalcific constrictive pericarditis so commonly associated with this disease.\nNumerous therapies have previously been investigated as adjuvant strategies in the prevention of pericardial constriction.\nColchicine is well described in the treatment of various aetiologies of pericarditis.\nThe aim of this research was to investigate the merit for the use of colchicine in the management of tuberculous pericarditis, specifically to prevent constrictive pericarditis.\nMethods\nThis pilot study was designed as a prospective, double-blinded, randomised, control cohort study and was conducted at a secondary level hospital in the Northern Cape of South Africa between August 2013 and December 2015.\nPatients with a probable or definite diagnosis of TB pericarditis were included (n = 33).\nStudy participants with pericardial effusions amenable to pericardiocentesis underwent aspiration until dryness.\nAll patients were treated with standard TB treatment and corticosteroids in accordance with the South African Tuberculosis Treatment Guidelines.\nPatients were randomised to an intervention and control group using a webbased computer system that ensured assignment concealment.\nThe intervention group received colchicine 1.0 mg per day for six weeks and the control group received a placebo for the same period.\nPatients were followed up with serial echocardiography for 16 weeks.\nThe primary outcome assessed was the development of pericardial constriction.\nUpon completion of the research period, the blinding was unveiled and data were presented for statistical analysis.\nResults\nTB pericarditis was found exclusively in HIV-positive individuals.\nThe incidence of pericardial constriction in our cohort was 23.8%.\nNo demonstrable benefit with the use of colchicine was found in terms of prevention of pericardial constriction (p = 0.88, relative risk 1.07, 95% CI: 0.46\u20132.46).\nInterestingly, pericardiocentesis appeared to decrease the incidence of pericardial constriction.\nConclusion\nBased on this research, the use of colchicine in TB pericarditis cannot be advised.\nAdjuvant therapy in the prevention of pericardial constriction is still being investigated and routine pericardiocentesis may prove to be beneficial in this regard.\nIntroduction\nSouth Africa, a land of stark contrasts, contains a diverse natural beauty that can easily be compared with some of the world\u2019s most majestic outdoor scenes.\nOne of the new seven wonders of the natural world, Table Mountain, parades its splendour to the capital of South Africa, Cape Town.\nUnfortunately, South Africa is also considered by many to be one of the tuberculosis (TB) capitals of the world.\nThe incidence of TB in South Africa is estimated to have increased by over 400% in the past 15 years.\nThis is confounded by a staggering co-infection rate of approximately 73% with the human immunodeficiency virus (HIV).\nOne of the most dreaded complications of TB pericarditis is pericardial scar formation.\nDue to scarring, the pericardium becomes calcified and contracts over the cardiac chambers, thereby encasing the heart in a fibrocalcific skin that impedes diastolic filling.\nConstrictive pericarditis (CP) is the natural consequence of about 17 to 40% of cases of TB pericardial infection.\nThe definitive treatment of CP is surgical removal of the pericardium, a procedure with a significant peri-operative mortality rate of approximately 15%.\nSouth Africa is on the forefront of research on TB heart disease and has recently published the large, multi-centre IMPI trial.\nOne of the goals of the IMPI trial was to assess the impact of corticosteroids in the management of TB pericarditis.\nThe major findings of the study included (1) corticosteroids had no impact on mortality rates in patients with TB pericarditis, (2) corticosteroids decreased the incidence of pericardial constriction by 46%, and (3) HIV-positive patients who received corticosteroids had a significantly increased risk of developing HIV-associated malignancies.\nIn established TB, early and effective treatment with shortcourse anti-TB therapy is the mainstay of management.\nVarious strategies have been investigated as adjuncts to anti-TB drugs in the prevention of pericardial constriction.\nThe ongoing discussions and numerous investigations into a wide array of agents as possible \u2018magic bullets\u2019 in the prevention of pericardial constriction (post-TB infection) illustrates both the interest in the field, and also the lack of a satisfying solution to this problem.\nThe following strategies have previously been evaluated: Mycobacterium indicus pranii immunotherapy, corticosteroids, pericardiocentesis,6 open surgical drainage (pericardial window), thalidomide, instilling intrapericardial fibrinolytic therapies, and a wide array of non-steroidal anti-inflammatory medication.\nNot one of these therapies has, to date, been internationally recognised as an acceptable standard of therapy, and the choice of adjuvant treatment varies significantly among experts in the field.\nColchicine is an inhibitor of microtubule polymerisation.\nIt acts by binding to tubulin and is registered for the acute treatment of gout crystal arthropathies.\nThe plant source of colchicine, the autumn crocus (Colchicum autumnale), was described as treatment for arthritis in the Ebers Papyrus in 1500 BC.\nIn modern medicine, colchicine has however played a wider role in the treatment of pericarditis of various aetiologies, both acute and chronic.\nThis has been investigated in a prospective, randomised trial named COPE (Colchicine for Acute Pericarditis),13 and the major findings concluded that colchicine significantly reduced the recurrence rates and symptom persistence due to pericarditis.\nTo date however, the use of colchicine has, to the best of our knowledge, never been systematically assessed in the context of pericardial TB.\nThe purpose of this research was to assess the merit for the use of colchicine in the context of TB pericarditis.\nMethods\nThis research was conducted in the Northern Cape province of South Africa at a secondary-level hospital in Kimberley between August 2013 and April 2015.\nThe research was approved by the ethics committee of the University of the Free State and the study was registered with the National Health Research Committee.\nThe research was conducted in accordance with the Declaration of Helsinki.\nThis pilot study was designed as a prospective, doubleblind, randomised, control cohort.\nAll patients presenting to the Kimberley Hospital complex (KHC) with pericardial effusions were assessed for inclusion and exclusion criteria.\nIn the absence of contra-indications, patients underwent therapeutic pericardiocentesis if the procedure was deemed safe and possible.\nStandard therapy was initiated in accordance with the South African National Tuberculosis Management Guidelines:14 weight-adjusted anti-TB drugs (Rifafourf\u00ae) and oral corticosteroids.\n(prednisone: 1.5 mg/kg per day for four weeks; 1.0 mg/kg per day for two weeks; 0.5 mg/kg per day for one week; 0.25 mg/kg per day for one week).\nHIV co-infected patients not previously on treatment were initiated on fixeddose combination (FDC) antiretroviral treatment six weeks after initiation of TB treatment (FDC: Tenofovir Disoproxil Fumarate 300 mg, Emtricitabine 200 mg and Efavirenz 600 mg).\nPatients were randomly assigned to the intervention group with the use of a web-based randomisation system that ensured assignment concealment.\nThe intervention group received colchicine (dose 1.0 mg per day) for a total of six weeks, whereas the control group received a placebo for the same period (Fig. 1).\nPatients subsequently underwent serial echocardiographic examinations on an out-patient basis and adherence checks, including pill counts, were done at follow-up visits.\nThe primary outcome assessed was the development of pericardial constriction and this diagnosis was made echocardiographically at four months post initial presentation.\nUpon completion of the follow-up period of all patients, the blinding was unveiled and data were presented for statistical analysis.\nTwo groups of patients were included: (1) definite TB pericarditis: the presence of TB bacilli was observed on microscopic examination of pericardial fluid; cultures of pericardial fluid were positive for Rifampicin-sensitive Mycobacterium tuberculosis (MTB); pericardial fluid was positive for MTB on direct polymerase chain reaction (PCR) (Gene Xpert); and (2) probable TB pericarditis: proof of TB was found elsewhere (positive cultures for MTB on sputum or cerebrospinal fluid); pericardial fluid with adenine deaminase (ADA) level > 40 U/l; a total diagnostic index score > 6 on using the Tygerberg clinical prediction score (Table 1).\nThe exclusion criteria were: patients with renal or hepatic impairment (creatinine clearance rate > 85 ml/min or transaminases > 1.5 upper limit of normal); and pregnant patients or patients intending to become pregnant within four months.\nThe gold-standard diagnostic test for the diagnosis of CP is the demonstration of increased interventricular interdependence during cardiac catheterisation.\nDoppler echocardiography and other novel echocardiographic techniques have provided us with reliable non-invasive alternatives to the diagnosis of CP.\nIn this study, the diagnosis of CP was made by means of echocardiography by adhering to the principles in the article by Dal-Bianco et al. on the echocardiographic diagnosis of CP.\nInitial echocardiographic assessment ensured that no features of constriction were present at the time of enrolment in the study.\nFollow-up echocardiograms were performed four months after the initiation of therapy.\nThe echocardiograms were performed and co-reviewed by two experienced echocardiographers (who had both attended a dedicated workshop at a tertiary-level academic hospital aimed at the echocardiographic diagnosis of CP).\nA GE Vivid E6\u00ae ultrasound machine was used to perform a systematic examination according to the basic minimum standards as stipulated by the British Society of Echocardiography.\nNumerous other echocardiographic parameters were assessed, including the presence of a septal shudder, respiratophasic septal shift, left atrial enlargement and echocardiographic features of pericardial thickening (Figs 2\u20134).\nStatistical analysis\nStatistical analysis was performed by the Department of Biostatistics of the University of the Free State, Bloemfontein, South Africa.\nThe SAS Version 8.3 was used.\nGroups were compared regarding outcomes using frequency tables with appropriate hypothesis testing (chi-squared of Fisher\u2019s exact test) and 95% confidence intervals for differences in percentages.\nThe standard deviation value p < 0.05 was considered significant.\nResults\nThirty-three patients met the initial inclusion criteria.\nThree patients passed away while in hospital and an additional three passed away during the follow-up period.\nIn-patient deaths were due to neutropenic sepsis, cerebrovascular incident and nosocomial pneumonia, respectively.\nIn all out-patient deaths, the cause was undetermined.\nFive patients were lost to follow up and one patient was removed from the study due to presumed drug side effects.\nA total of 21 patients completed the follow-up period (Fig. 5).\nThe study population had a female preponderance (66% females) and the mean age of the studied patients was 31 years.\nDisseminated pericardial tuberculosis was found to be a disease exclusive to the immune-compromised in this cohort; all 21 patients were HIV positive.\nThe median CD4+ count was 162 and 346 cells/mm3 in the colchicine and placebo groups, respectively.\nOf the 21 eligible participants, 12 had been assigned to the treatment group and the remaining nine were in the placebo group.\nThe diagnosis of definite pericardial tuberculosis was made in 23.8% of the patients, while the remaining 76.2% were diagnosed on the basis of suggestive clinical and biochemical features (see inclusion criteria).\nOf the studied patients, 47.6% underwent pericardiocentesis, whereas the remaining 52.4% could not undergo safe pericardiocentesis.\nThe average volume of fluid drained via single pericardial aspiration was 622 ml.\nThe macroscopic appearance of the fluid varied from serosanguineous to haemorrhagic, reflecting the different pathological stages of development.\nMycobacterium tuberculosis was proven on pericardial aspirates in 50% of cases, either by positive culture (30%) or by direct PCR technique (Gene Xpert) (20%) (Table 2).\nPericardial constriction is the natural sequela of approximately 17 to 40% of TB pericardial infections.\nIn our cohort, the incidence of pericardial constriction (demonstrated by echocardiography) four months after the initial diagnosis was 23.8%.\nOf the five patients who developed pericardial constriction, two were in the control group and the remaining three were in the group treated with colchicine.\nOf those who did not develop pericardial constriction, nine were in the colchicine group and seven were in the placebo group.\nThe data from Table 3 yields a p-value of 0.88.\nThe relative risk for developing constriction in the colchicine group compared to the intervention group was 1.07 (95% CI: 0.46\u20132.46).\nThere was therefore no statistically demonstrable correlation between the use of colchicine and pericardial constriction in this study cohort.\nThe side effects among the patients using colchicine were usually minor; 56% of the initial 19 patients who were in the colchicine group reported self-limiting diarrhoea during their hospital stay.\nSerious side effects were observed in one patient who developed hepatitis during his course of treatment.\nThe patient was removed from the study and daily liver function testing showed a rapid recovery.\nAlthough the study was neither empowered nor designed to evaluate the effect of pericardiocentesis on the subsequent development of pericardial constriction, a very apparent and interesting finding was observed.\nWe found that, with the exception of one patient, all those who developed pericardial constriction were in the group that did not undergo pericardiocentesis.\nConversely, in the group that underwent pericardiocentesis, only one participant developed pericardial constriction.\nPericardiocentesis therefore seemed to be very effective in the prevention of pericardial constriction and in this cohort only one patient (10%) who underwent pericardiocentesis developed constriction.\nThese findings are observational and disregard the initial group allocations.\nDiscussion\nThe proverbial \u2018eureka moment\u2019 in the management of TB pericarditis seems to be elusive.\nNumerous interventions have been postulated and investigated in an attempt to prevent the devastating post-inflammatory changes in the pericardium following TB pericarditis.\nIn this pilot study, the merit of adding colchicine to the current management guidelines was investigated in a systematic manner.\nAs all the participants of this study were HIV positive, the findings can only be applied to this subgroup of patients with TB pericarditis.\nThere was a notable difference in the median CD4+ lymphocyte count between the treatment and placebo groups, but when assessed as an independent variable, no correlation could be demonstrated between degree of immunocompetency, as measured by CD4+ count, and the risk for development of constriction.\nThis pilot study could not demonstrate any benefit derived from the addition of colchicine to the routine management of HIV-positive patients with TB pericarditis.\nThe power of this pilot trial was insufficient to detect small differences in outcome; however, it appears that colchicine use has no correlation with the prevention or formation of post-TB CP.\nThis pilot trial could not assess the beneficial effects of colchicine in the HIV-negative patient with TB pericarditis.\nAfter considering the findings of this pilot research, the costs of the drug, the polypharmacy these patients are exposed to, drug\u2013drug interactions and side effects (albeit mild), this study would advise against the use of colchicine in the management of HIV-positive patients with TB pericarditis.\nThe implementation of a pericardiocentesis until dryness (with or without extended drainage) was up to this point never studied in a controlled or comparative manner.\nResearch conducted by Reuter et al. in 2007 found the first evidence to suggest the benefit of a pericardiocentesis until dryness with extended drainage.\nIn their research, 162 patients with TB pericarditis underwent pericardiocentesis, and over a followup period of six years, only two patients (1.23%) developed fibrous pericardial constriction.\nThe research concluded that echocardiographic-guided pericardiocentesis with extended drainage is a safe and effective management option, and when combined with short-course anti-tuberculous therapy, it almost completely prevents the development of CP.\nA few leading centres are employing a routine \u2018pericardiocentesis until dryness\u2019 approach based on this literature, whereas most do not.\nThe interesting observation made in our pilot study was that the findings made by Reuters et al. in 2007 were reproducible on a much smaller scale.\nPericardial constriction, although having a low incidence, was almost exclusively seen in the group that did not undergo pericardiocentesis (observational \u2013 disregard original group allocation).\nAs suggested by some expert opinion and as supported by the data published by Reuters et al. and observational findings of our pilot trial, the practice of routine pericardiocentesis until dryness in the absence of contraindications appears to be the preferred management option and this might well be the long-awaited \u2018eureka moment\u2019, in an attempt to halt the development of pericardial constriction.\nLinitations of the study\nThe diagnosis of pericardial constriction was made with echocardiography, whereas the gold standard for diagnosing CP is invasive haemodynamic studies.\nWork done by Oh et al. and Boonyaratevej et al. demonstrated that one of the most characteristic findings of CP, a respiratory variation in early transmittal inflow velocity, is neither perfect in its sensitivity nor specificity for the diagnosis.\nIn patients with markedly elevated left atrial pressures, the respiratory variation in the inflow velocities may be less than 25%.\nFurthermore, in patients with chronic obstructive pulmonary disease and severe right ventricular dysfunction, the variation may be elevated in the absence of CP.\nThis research emphasises the importance of using a variety of recognised echocardiographic diagnostic tools to confirm a non-invasive diagnosis of CP.\nThe duration of follow up was only four months.\nSome comparative research had follow-up periods of up to six years.\nMost patients who develop CP, do so in a period of three to four months.\nThere may however be patients who will only develop constriction after four months.\nResearch to address this aspect may be valuable.\nCorticosteroids were used as part of the standard therapy in all patients.\nHowever, subsequent to the initiation of the research, the IMPI trial brought to light their findings that corticosteroids should not be used in TB pericarditis in HIV-infected patients.\nThe South African National TB guidelines published in 2014 still advised the use of corticosteroids in all patients and the findings of the IMPI trial had not yet been incorporated into current South African National Tuberculosis Management Guidelines.\nConclusion\nBased on current research, the use of colchicine in addition to standard antituberculous therapy cannot be advised in the context of TB pericarditis in the HIV-positive population.\nThe jury is still out on which adjuvant strategies may prove to be beneficial in the prevention of CP, especially in the HIV-coinfected subgroup.\nBased on observations from this research and some other studies, routine pericardiocentesis until dryness with extended drainage may prove to be the long-awaited solution to the common dilemma of post-TB CP.\nFlow diagram illustrating study methodology.\nA. Tissue Doppler of the medial aspect of the mitral valve annulus demonstrating early diastolic tissue velocity of 0.14 m/s. B. Tissue Doppler of the lateral aspect of the mitral valve annulus showing early diastolic tissue velocity of 0.12 m/s. The lower tissue velocity on the lateral aspect is the opposite of the normal phenomenon (annulus reversus).\nPulse-wave Doppler at the level of the mitral valve leaflet tips demonstrating a respiratophasic variation in the early diastolic transmitral inflow velocities in excess of 25%.\nDilated and distended inferior vena cava. No respiratory variation was observed.\nScreening, randomisation, follow up and analysis of the study patients.\n\nThe Tygerberg clinical prediction score for the diagnosis of TB pericarditis. A total diagnostic score > 6 yields a sensitivity of 82% and a specificity of 76% for the diagnosis of TB pericarditis\nAdmission variable | Diagnostic index\nWeight loss | 1\nNight sweats | 1\nFever | 2\nSerum globulin > 40 g/l | 3\nLeukocyte count < 10 \u00d7 109 | 3\n\n\nPericardial fluid biochemistry\nBiochemical parameter | Average\nProtein (g/l) | 62.7\nADA (U/l) | 96.6\nLDH (U/l) | 4494\npH | 7.3\nGlucose (mmol/l) | 2.8\n\n\nTwo-by-two table demonstrating the primary study outcome\n | Colchicine | Placebo | Total\nConstriction | 3 | 2 | 5\nNo constriction | 9 | 7 | 16\nTotal | 12 | 9 | 21\n",
        "label": "low",
        "id": "task4_RLD_test_1"
    },
    {
        "paper_doi": "10.1371/journal.pmed.1001630",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: cRCT with 2 intervention armsUnit of allocation: clusters (villages)Number of units: 25 randomized villages in each arm. A subset of 20 villages per arm was used for entomological assessment.Outcome assessment/surveillance type: see below in 'Outcomes' sectionLength of follow-up: 3 postintervention cross-sectional household surveys were undertaken in 2012. Survey A (23 February to 31 March) was after the short rainy season and 2 months after the first spray round. Survey B (25 June to 31 July) was after the long rainy season, 6 months after the first spray round, and 2 months after the second spray round. Survey C (25 October to 4 December) was 6 months after the second spray round and 10 months after the first. Baseline surveys were conducted in 2011 during the same periods as surveys A and B.Adjustment for clustering: yes\n\n\nParticipants: Number of participants: for each of the surveys, a different number of participants were used in each cohortSurvey A: 2192 children in control arm, 2348 in intervention armSurvey B: 2045 children in control arm, 2207 in intervention armSurvey C: 2101 children in control arm, 2303 in intervention armPopulation characteristics: cohort of children aged 6 months to 14 years, villages had to be sprayed with IRS in the baseline year.Withdrawal and loss to follow-up: 82.2-84.4% of intervention participants tested in each survey. 78.3-80.8% of control participants tested\n\n\nInterventions: IRSActive ingredient and dosage: bendiocarb 400 mg/m2Formulation: 80% WPFrequency of spraying: 2 rounds of spraying (December 2011 to January 2012) and (April 2012 to May 2012), timed to precede the peak in malaria cases that normally occurs at the end of each rainy season.Coverage: survey A: 92.1% (95% CI 88.4% to 94.7%) (1215); survey B: 89.5% (95% CI 84.0% to 93.2%) (1138); survey C: 89.3% (95% CI 83.6% to 93.2%) (1209)Buffer size between clusters: each village was divided into a core surveillance area consisting of >= 200 houses and approximately 1 km radius, where the surveys were conducted, and an outer buffer zone of approximately 1 km width which also received treatment but in which there was no outcome monitoring.ITNActive ingredient and dosage: permethrin 2% w/w (Olyset Net)Coverage measured as % of households with >= 1 ITN per sleeping space: survey A: 57.2 (range 53.6-60.7) (1215); survey B: 57.4 (range 54.0-60.9) (1142); survey C: 56.8 (range 51.7-61.8) (1211)Coverage measured as % of households with >= 1 ITN: survey A: 89.0 (range 87.1-90.6) (1216); survey B: 88.2 (range 85.7-90.3) (1142); survey C: 83.8 (range 79.9-87.1) (1211)Compliance measured as % of study children that reported sleeping under an ITN the night previous to the survey: survey A: 53.0 (range 47.5-58.3) (2349); survey B: 44.1 (range 39.2-49.2) (2207); survey C: 36.1 (range 31.0-41.5) (2303)ControlITN only as aboveCoverage measured as % of households with >= 1 ITN per sleeping space: survey A: 52.2 (range 47.8-56.5) (1178); survey B: 51.6 (range 47.0-56.0) (1094); survey C: 52.8 (range 47.6-58.0) (1168)Coverage measured as % of households with >= 1 ITN: survey A: 85.8 (range 83.7-87.7) (1177); survey B: 82.5 (range 78.7-85.7) (1096); survey C: 78.2 (range 74.3-81.6) (1170)Compliance measured as % of study children that reported sleeping under an ITN the night previous to the survey: survey A: 46.6 (range 41.7-51.6) (2193); survey B: 40.7 (range 34.7-47.0) (2045); survey C: 36.0 (range 29.8-42.6) (2101)Cointerventions: none reported\n\n\nOutcomes: P falciparum parasite rate in children aged 6 months to 14 years, 80 households in each cluster. Up to 3 children per household selected. Aimed for a mean of 80 children per cluster. Tested with RDT (Carestart (Pan) Malaria, DiaSys)Anaemia in children aged < 5 yearsMean Hb in children aged < 5 years. Tested with HemoCue Hb 201+ (Aktiebolaget Leo Diagnostics)EIR: 20/25 clusters per arm were monitored for 1 night each month from April 2011 to December 2012. 8 randomly selected houses in each clusterSporozoite rate\n\n\nLocation profile: Study location: north-west Tanzania, Muleba Distract, Kagera Region, the study area included 68,108 households at an altitude of 1100-1600 m above sea level. Rainfall occurred in 2 seasons: the 'short rains' in October-December (mean monthly rainfall 160 mm) and the 'long rains' in March-May (mean monthly rainfall 300 mm).Malaria endemicity: perennial with peaks after the rainy seasonEIR: baseline characteristics measured by the study reported a mean per month in the control arm of 1.1 (range 0.4-2.8) and 1.3 (range 0.4-4.4) in the intervention armPopulation proximity/density: not reportedPlasmodium spp:P falciparum\n\n\nVector profile: Primary (and secondary) vector species:An gambiae s.s. and An arabiensisVector behaviour (nature, stability, adult habitat, peak biting times, exophilic/endophilic, exophagic/endophagic, anthropophilic/zoophilic): not reportedPhenotypic resistance profile: resistance to pyrethroids in An gambiae s.s.Genotypic resistance profile: not reportedMethod of mosquito collection: CDC light traps indoors\n\n\nNotes: For inclusion in the review meta-analyses, we calculated adjusted risk ratios for prevalence from the reported adjusted odds ratios following the methodology stated in Section 12.5.4.4 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a)\n\n",
        "objective": "To summarize the effect on malaria of additionally implementing IRS, using non\u2010pyrethroid\u2010like or pyrethroid\u2010like insecticides, in communities currently using ITNs.",
        "full_paper": "Philippa West and colleagues compare Plasmodium falciparum infection prevalence in children, anemia in young children, and entomological inoculation rate between study arms.\nPlease see later in the article for the Editors' Summary\nBackground\nInsecticide-treated nets (ITNs) and indoor residual spraying (IRS) of houses provide effective malaria transmission control.\nThere is conflicting evidence about whether it is more beneficial to provide both interventions in combination.\nA cluster randomised controlled trial was conducted to investigate whether the combination provides added protection compared to ITNs alone.\nMethods and Findings\nIn northwest Tanzania, 50 clusters (village areas) were randomly allocated to ITNs only or ITNs and IRS.\nDwellings in the ITN+IRS arm were sprayed with two rounds of bendiocarb in 2012.\nPlasmodium falciparum prevalence rate (PfPR) in children 0.5\u201314 y old (primary outcome) and anaemia in children <5 y old (secondary outcome) were compared between study arms using three cross-sectional household surveys in 2012.\nEntomological inoculation rate (secondary outcome) was compared between study arms.\nIRS coverage was approximately 90%.\nITN use ranged from 36% to 50%.\nIn intention-to-treat analysis, mean PfPR was 13% in the ITN+IRS arm and 26% in the ITN only arm, odds ratio\u200a=\u200a0.43 (95% CI 0.19\u20130.97, n\u200a=\u200a13,146).\nThe strongest effect was observed in the peak transmission season, 6 mo after the first IRS.\nSubgroup analysis showed that ITN users were additionally protected if their houses were sprayed.\nMean monthly entomological inoculation rate was non-significantly lower in the ITN+IRS arm than in the ITN only arm, rate ratio\u200a=\u200a0.17 (95% CI 0.03\u20131.08).\nConclusions\nThis is the first randomised trial to our knowledge that reports significant added protection from combining IRS and ITNs compared to ITNs alone.\nThe effect is likely to be attributable to IRS providing added protection to ITN users as well as compensating for inadequate ITN use.\nPolicy makers should consider deploying IRS in combination with ITNs to control transmission if local ITN strategies on their own are insufficiently effective.\nGiven the uncertain generalisability of these findings, it would be prudent for malaria control programmes to evaluate the cost-effectiveness of deploying the combination.\nTrial registration\nwww.ClinicalTrials.gov NCT01697852\nPlease see later in the article for the Editors' Summary\nEditors' Summary\nBackground\nEvery year, more than 200 million cases of malaria occur worldwide, and more than 600,000 people, mainly children living in sub-Saharan Africa, die from this parasitic infection.\nMalaria parasites, which are transmitted to people through the bites of infected night-flying mosquitoes, cause a characteristic fever that needs to be treated promptly with antimalarial drugs to prevent anaemia (a reduction in red blood cell numbers) and organ damage.\nPrompt treatment also helps to reduce malaria transmission, but the mainstays of global malaria control efforts are the provision of insecticide-treated nets (ITNs) for people to sleep under to avoid mosquito bites, and indoor residual spraying (IRS) of houses with insecticides, which prevents mosquitoes from resting in houses.\nBoth approaches have been scaled up in the past decade.\nAbout 54% of households in Africa now own at least one ITN, and 8% of at-risk populations are protected by IRS.\nAs a result of the widespread deployment of these preventative tools and the increased availability of effective antimalarial drugs, malaria-related deaths in Africa fell by 45% between 2000 and 2012.\nWhy Was This Study Done?\nSome countries have chosen to use ITNs and IRS in combination, reasoning that this will increase the proportion of individuals who are protected by at least one intervention and may provide additional protection to people using both interventions rather than one alone.\nHowever, providing both interventions is costly, so it is important to know whether this rationale is correct.\nIn this cluster randomised controlled trial (a study that compares outcomes of groups of people randomly assigned to receive different interventions) undertaken in the Muleba District of Tanzania during 2012, the researchers investigate whether ITNs plus IRS provide more protection against malaria than ITNs alone.\nMalaria transmission occurs throughout the year in Muleba District but peaks after the October\u2013December and March\u2013May rains.\nNinety-one percent of the district's households own at least one ITN, and 58% of households own enough ITNs to cover all their sleeping places.\nAnnual rounds of IRS have been conducted in the region since 2007.\nWhat Did the Researchers Do and Find?\nThe researchers allocated 50 communities to the ITN intervention or to the ITN+IRS intervention.\nDwellings allocated to ITN+IRS were sprayed with insecticide just before each of the malaria transmission peaks in 2012.\nThe researchers used household surveys to collect information about ITN coverage in the study population, the proportion of children aged 0.5\u201314 years infected with the malaria parasite Plasmodium falciparum (the prevalence of infection), and the proportion of children under five years old with anaemia.\nIRS coverage in the ITN+IRS arm was approximately 90%, and 50% of the children in both intervention arms used ITNs at the start of the trial, declining to 36% at the end of the study.\nIn an intention-to-treat analysis (which assumed that all study participants got the planned intervention), the average prevalence of infection was 13% in the ITN+IRS arm and 26% in the ITN arm.\nA per-protocol analysis (which considered data only from participants who received their allocated intervention) indicated that the combined intervention had a statistically significant protective effect on the prevalence of infection compared to ITNs alone (an effect that is unlikely to have arisen by chance).\nFinally, the proportion of young children with anaemia was lower in the ITN+IRS arm than in the ITN arm, but this effect was not statistically significant.\nWhat Do These Findings Mean?\nThese findings provide evidence that IRS, when used in combination with ITNs, can provide better protection against malaria infection than ITNs used alone.\nThis effect is likely to be the result of IRS providing added protection to ITN users as well as compensating for inadequate ITN use.\nThe findings also suggest that the combination of interventions may reduce the prevalence of anaemia better than ITNs alone, but this result needs to be confirmed.\nAdditional trials are also needed to investigate whether ITN+IRS compared to ITN reduces clinical cases of malaria, and whether similar effects are seen in other settings.\nMoreover, the cost-effectiveness of ITN+IRS and ITN alone needs to be compared.\nFor now, though, these findings suggest that national malaria control programs should consider implementing IRS in combination with ITNs if local ITN strategies alone are insufficiently effective and cannot be improved.\nAdditional Information\nPlease access these websites via the online version of this summary at http://dx.doi.org/10.1371/journal.pmed.1001630.\nInformation is available from the World Health Organization on malaria (in several languages), including information on insecticide-treated bed nets and indoor residual spraying; the World Malaria Report 2013 provides details of the current global malaria situation\nThe US Centers for Disease Control and Prevention provides information on malaria, on insecticide-treated bed nets, and on indoor residual spraying; it also provides a selection of personal stories about malaria\nInformation is available from the Roll Back Malaria Partnership on the global control of malaria and on the Global Malaria Action Plan (in English and French); its website includes fact sheets about malaria in Africa and about nets and insecticides\nMedlinePlus provides links to additional information on malaria (in English and Spanish)\nIntroduction\nIn the past decade, insecticide-treated net (ITN) distribution has been scaled up across Africa in line with the Abuja Declaration in 2000.\nThe percentage of households that owned at least one ITN in Africa increased from 3% in 2000 to 54% in 2013.\nThe World Health Organization (WHO) policy that ITNs should be provided to everyone in malaria risk areas (universal coverage) has been adopted by 34 of the 44 malaria endemic countries in Africa.\nIndoor residual spraying (IRS) of houses, the second major vector control tool used to prevent malaria, has similarly been scaled up.\nThe proportion of at-risk populations protected by IRS increased from less than 5% in 2005 to 8% in 2012.\nAs a result of the increase in the deployment of these preventive tools and the increased availability and use of artemisinin-based combination therapies, malaria-related mortality fell by 45% between 2000 and 2012 in Africa, but there remained an estimated 165 million cases and 562,000 deaths due to malaria in 2012.\nIn an attempt to reduce the malaria burden further, a number of countries have chosen to use ITNs and IRS in combination.\nFifty-seven countries, 31 of which are in Africa, use both IRS and ITNs, in at least some areas.\nApplying ITNs and IRS in the same area can increase the proportion of individuals who are protected by at least one intervention or, more optimally, may provide additional protection for those protected by both interventions compared to those receiving one method alone.\nSince the cost of implementing both IRS and universal coverage of ITNs is much greater than the cost of implementing only one of the interventions, it is important to know what extra protection is gained by adding a second intervention, to help national malaria control programmes and international funding agencies such as the President's Malaria Initiative (PMI) and the Global Fund to Fight AIDS, Tuberculosis and Malaria make decisions that are based on evidence of likely impacts and costs.\nThis is particularly significant now, since it is estimated that global funding for malaria is less than half of what is needed to attain universal coverage of malaria vector control, i.e., access to either ITNs or IRS.\nIt is unclear from current evidence whether combined use of ITNs and IRS provides an additional benefit compared to using either intervention alone, and whether this will be similar across transmission settings.\nA recent trial in Benin found no added benefit to using IRS in combination with ITNs compared to ITNs alone.\nHowever, this trial had a relatively small sample size, and its findings may be applicable to only a particular transmission setting in west Africa.\nTo help define future malaria control policy in Africa, the PMI decided to sponsor an independent two-arm cluster randomised controlled trial (CRT) to compare the protective effectiveness of IRS in combination with high coverage of ITNs with high coverage of ITNs alone for malaria transmission control.\nTanzania has a high malaria disease burden, with a national average of 9% of children under 5 y being infected with malaria parasites.\nMalaria control activities have been scaled up nationally since 2005.\nA universal coverage campaign (UCC) primarily funded by the Global Fund to Fight AIDS, Tuberculosis and Malaria distributed long-lasting insecticidal nets (LLINs) free of charge in 2011 to top up coverage from previous distributions.\nIRS, funded by the PMI, commenced in 2007 in two districts of Kagera Region, in northwest Tanzania, and has since been extended to cover 18 districts.\nBecause IRS is costly and logistically intensive, there is an urgent need to know whether it is necessary to continue with IRS after an ITN UCC has been successfully completed.\nThe trial was carried out in 109 rural villages in Muleba District (1\u00b045\u2032S 31\u00b040\u2032E), Kagera Region.\nThe study area includes 68,108 households at an altitude ranging from 1,100 to 1,600 m above sea level.\nRainfall occurs in two seasons: the \u201cshort rains\u201d in October\u2013December (average monthly rainfall 160 mm) and the \u201clong rains\u201d in March\u2013May (average monthly rainfall 300 mm), with malaria transmission occurring throughout the year and peaking after the rainy seasons.\nAnnual rounds of IRS with the pyrethroid lambda-cyhalothrin (ICON 10CS, Syngenta) were conducted between 2007 and 2011 in Muleba District, i.e., in the entire study area.\nThe predominant malaria vectors are Anopheles gambiae s.s. and An. arabiensis .\nTests of mosquito susceptibility using standard WHO bioassays showed resistance to pyrethroids in An. gambiae s.s. in 2011.\nAs a result, IRS policy was changed to use the carbamate insecticide bendiocarb (Ficam 80% wettable powder, Bayer) by the PMI in 2012.\nMethods\nEthics and Community Sensitisation\nThe trial was approved by the ethics review committees of the Kilimanjaro Christian Medical College, the Tanzanian National Institute for Medical Research, and the London School of Hygiene and Tropical Medicine.\nWritten informed consent was obtained from all respondents.\nPrior to the baseline surveys, village and hamlet leaders were invited to sensitisation sessions conducted by district health officers.\nThe trial was registered with ClinicalTrials.gov (registration number NCT01697852) in September 2012.\nThe trial was not registered earlier because the authors were not aware of journal requirements for prospective registration.\nAll authors have affirmed that any trials they are involved in on the same or a related drug or intervention are registered.\nAn accurate summary of the trial's results has been submitted to ClinicalTrials.gov.\nStudy Design\nA CRT was conducted, comparing the Plasmodium falciparum prevalence rate (PfPR) in children 0.5\u201314 y old between communities targeted to receive both high-coverage IRS and high coverage of ITNs (ITN+IRS arm) and communities targeted for high coverage of ITNs only (standard-care control arm).\nSecondary outcomes were moderate/severe anaemia (haemoglobin <8 g/dl) in children under 5 y old and entomological inoculation rate (EIR) due to An. gambiae s.l.\nPower calculations showed that 25 clusters per study arm were required, with 80 children per cluster, to give 80% power to detect a true absolute difference in PfPR of at least 3% between study arms (relative difference 31%) with 5% significance (two-sided), based on an expected prevalence in the ITN only arm of 9% (PfPR in first baseline survey).\nThe between-cluster coefficient of variation (k) was calculated as 0.25 from the first pre-randomisation baseline survey.\nEach cluster consisted of at least one village and was divided into a core surveillance area consisting of at least 200 houses and approximately 1 km radius, where the surveys were conducted, and an outer buffer zone, 1 km in width, which also received the allocated treatment but in which no outcome monitoring was done.\nVillages were eligible for inclusion in the study if they were within daily commuting distance for survey work and had been sprayed with IRS in the baseline year.\nAll clusters received LLINs from the UCC in 2011.\nTwenty-five clusters were randomly allocated to receive IRS, in addition to ITNs, using restricted randomisation to limit potential imbalance between study arms.\nBaseline surveys provided data on seven criteria for which the study arms were balanced by constraining the randomisation (Table 1).\n200,000 random allocations were generated.\nMean values for each arm were calculated from cluster summaries for each of the seven restriction variables; 25,119 randomisations fulfilled the restriction criteria and were therefore eligible.\nThese allocations were tested for independence between any two clusters.\nThe large number of acceptable allocations, of which one was randomly selected, ensured that the restriction did not affect the validity of inference.\nThere was no evidence of dependence between any pair of clusters.\nInterventions\nHouseholds in the study area with children aged under 5 y received LLINs from a national distribution campaign in 2009.\nIn 2011, the district health authority, supported by Mennonite Economic Development Associates, completed a UCC that distributed 144,000 LLINs (Olyset, Sumitomo Chemicals) to the population of Muleba District, including all study clusters.\nThe campaign aimed to top up net coverage, so that every sleeping place had one ITN.\nAfter the UCC, 91% of households owned at least one ITN, and 58% of households owned enough ITNs to cover all their sleeping places.\nSpraying was conducted by RTI International on behalf of PMI in the ITN+IRS study arm.\nThe interior walls of each dwelling were sprayed with the carbamate insecticide bendiocarb (Ficam 80% wettable powder, Bayer) at 400 mg/m2 between December 2011 and January 2012 (round 1), and between April and May 2012 (round 2).\nSpray rounds were timed to precede the peak in malaria cases that normally occurs at the end of each rainy season, taking into account the relatively short residual duration of bendiocarb.\nBendiocarb is a carbamate insecticide recommended by WHO for IRS.\nIt is one of the few insecticides evaluated and approved by the WHO Pesticide Evaluation Scheme that has the potential to control pyrethroid-resistant mosquitoes, is odour-free, and is safe to house occupants at the recommended application rate.\nBefore obtaining WHO approval, all IRS insecticides are subject to risk assessment by WHO toxicologists.\nBendiocarb is an acetylcholinesterase inhibitor, but no serious adverse effects due to bendiocarb IRS have been reported in the recent medical literature.\nSurveys\nThree post-intervention cross-sectional household surveys were undertaken in 2012 (see Figure 1).\nSurvey A (23 February\u201331 March) was after the short rainy season and 2 mo after the first spray round.\nSurvey B (25 June\u201331 July) was after the long rainy season, 6 mo after the first spray round, and 2 mo after the second spray round.\nSurvey C (25 October\u20134 December) was 6 mo after the second spray round and 10 mo after the first.\nBaseline surveys were conducted in 2011 during the same periods as surveys A and B.\nFor each survey, 80 households were randomly selected in the core area of each cluster.\nHouseholds were eligible for the study if they had children aged 0.5\u201314 y.\nAny child aged 0.5\u201314 y was eligible to be included in the study.\nUp to three children per household were randomly selected for testing.\nAllowing for ineligible households, absence on the day of the survey, and refusals at the household and individual level, it was estimated that this would provide on average 80 children for testing per cluster.\nThe household head or another responsible adult from the household was interviewed, after seeking written informed consent.\nData on IRS coverage, bed net ownership and usage, demographics of household members, and other household characteristics were gathered using an adapted version of the standard Malaria Indicator Survey.\nSelected children were tested on the following day for malaria parasites using a rapid diagnostic test (RDT) (CareStart [Pan] Malaria, DiaSys) and had haemoglobin levels measured using HemoCue Hb 201+ (Aktiebolaget Leo Diagnostics).\nIndividuals testing positive by RDT were treated with artemether/lumefantrine (Artefan 20/120, Ajanta Pharma) following national treatment guidelines.\nEntomological surveillance was carried out in the core surveillance areas of a subset of 40 of the 50 clusters from April 2011 to December 2012.\nFor one night of each month US Centers for Disease Control and Prevention light traps for mosquito collections were set up in eight randomly selected houses in each cluster (320 houses per month).\nAnopheles mosquitoes collected were identified to species using a simplified morphological key adapted from Gillies and Coetzee.\nA sub-sample of An. gambiae s.l. individuals were tested using real-time PCR TaqMan assay to distinguish between the two sibling species An. gambiae s.s. and An. arabiensis .\nMosquitoes were also tested for P. falciparum sporozoites (P. falciparum circumsporozoite protein) using ELISA.\nStatistical Analysis\nStatistical analysis was done in Stata 12 (Statacorp) and R version 2.13.1 (R Foundation for Statistical Computing).\nThe odds of PfPR and moderate/severe anaemia for individuals were compared between study arms in intention-to-treat (ITT) analysis using logistic regression.\nMean haemoglobin was compared between the study arms using linear regression.\nA robust variance estimator was used to calculate standard errors to adjust for within-cluster correlation of responses (Stata survey commands, first-order Taylor-series linearization method).\nPfPR was considered as P. falciparum alone or mixed infections as detected by the RDT.\nThe overall odds ratio (OR) for the three surveys combined was calculated accounting for survey.\nAn adjusted Wald test was performed to test whether there was evidence for effect modification between study arm and survey round.\nA sensitivity analysis was conducted excluding one cluster from the ITN only arm that mistakenly received IRS, to assess the impact of this protocol violation on the results of ITT analysis.\nBecause of the wide variation in cluster-level estimates of PfPR at baseline, an OR for ITN+IRS versus ITN alone was calculated adjusting for baseline PfPR.\nA secondary per-protocol analysis was performed, in which individuals from the ITN+IRS arm who used an ITN and lived in a house sprayed in the most recent round of IRS were compared to individuals who used an ITN in the ITN only arm.\nThe cluster that violated the protocol was excluded from the per-protocol analysis.\nThe monthly EIR was calculated as the daily EIR found during the one night collection multiplied by the number of days in the month.\nMean EIRs were compared between study arms using negative binomial regression and adjusting for within-cluster correlation.\nResults\nAt baseline, PfPR, anaemia, ITN ownership, ITN usage, and mean EIR per month (Table 2) were similar in the two study arms.\nPfPR in children aged 6 mo to 14 y old was 9.3% (95% CI 5.9%\u201314.5%) after the short rains (survey A, February\u2013March) and 22.8% (95% CI 17.3%\u201329.4%) after the long rains (survey B, June\u2013July).\nAnaemia in children 0.5\u20134 y was 6.2% (95% CI 4.5%\u20138.5%) after the long rains.\nOf the 2,000 houses selected in each study arm for each post-intervention survey, 20% to 24% had no children between 0.5 and 14 y old (were ineligible), 13% to 18% were vacant on the day of survey, fewer than 1% refused to participate, and 55% to 61% participated in the survey (Figure 2).\nOf the children selected for RDT, 81%\u201384% were tested.\nPost-intervention IRS coverage reported by householders was 92.1% after the first spray round and 89.5% after the second (Table 3).\nIn the intervention year, the percentage of houses with sufficient ITNs for each sleeping place remained stable over successive surveys and was similar between study arms (range 52%\u201357%; Table 3).\n82.2% and 87.0% of households owned at least one ITN in the ITN only arm and the ITN+IRS arm, respectively (all surveys combined), with weak evidence that the percentage of households that owned at least one ITN was lower in the ITN only arm, and that it decreased from survey A to survey C in both arms (Table 3).\nITN usage in children was similar between study arms but declined from 50% in survey A to 36% in survey C.\nThe primary outcome PfPR was lower in the ITN+IRS arm than in the ITN only arm in all three surveys in the intervention year (Table 4).\nFor all three surveys combined, the overall OR was 0.43 (95% CI 0.19\u20130.97), with weak evidence that the intervention effect differed between surveys (interaction p\u200a=\u200a0.08).\nThe strongest effect was observed in survey B (OR 0.33, 95% CI 0.15\u20130.75), which was conducted at the peak of malaria transmission after the long rains, 6 mo after the first IRS and 2 mo after the second IRS.\nThe evidence for an effect was weaker in survey A (OR 0.51, 95% CI 0.24\u20131.09), conducted shortly after the first IRS round, and in survey C (OR 0.48, 95% CI 0.18\u20131.24), conducted several months after the main transmission season and 6 mo after last spray round.\nThe range of cluster-specific estimates for PfPR was 0% to 92% in the ITN only arm and 0% to 68% in the ITN+IRS arm.\nThe sensitivity analysis showed that excluding the cluster from the ITN only arm that had received IRS did not affect the results of the ITT analysis (Table S1).\nThe overall OR for all three surveys combined was very similar after adjusting for baseline PfPR, OR\u200a=\u200a0.41, but the precision of the estimate was increased (95% CI 0.29\u20130.59, p<0.0001).\nPrevalence of moderate to severe anaemia in children under 5 y old, a secondary outcome, was lower in the ITN+IRS arm in all post-intervention surveys, but the difference was statistically significant only in survey B (Table 5).\nMean haemoglobin was higher in children under 5 y old in the ITN+IRS arm than in the ITN only arm in all three surveys.\nThe evidence for an effect was greatest in survey B (0.49 g/dl, 95% CI 0.10\u20130.89, p\u200a=\u200a0.016), with a non-significant result in survey A (0.28 g/dl, 95% CI \u22120.02 to 0.59, p\u200a=\u200a0.065) and survey C (0.36 g/dl, 95% CI \u22120.02 to 0.73, p\u200a=\u200a0.060).\nMean EIR per month, a secondary outcome, was 0.22 in the ITN+IRS arm and 1.26 in the ITN only arm (rate ratio\u200a=\u200a0.17, 95% CI 0.03\u20131.08, p\u200a=\u200a0.059; Table 6).\nThe between-cluster coefficient of variation (k) was 0.20, 0.28, and 0.26 in the three post-intervention surveys, respectively.\nFor each survey, k was similar in the two arms.\nFor all surveys, per-protocol analysis showed statistically significant evidence for a protective effect of the combined intervention on PfPR (survey A: OR 0.39, 95% CI 0.18\u20130.81; survey B: OR 0.21, 95% CI 0.09\u20130.49; and survey C: OR 0.27, 95% CI 0.10\u20130.73; Table 7).\nDiscussion\nThis is the first randomised trial to our knowledge that provides evidence that IRS, when used in combination with ITNs, can give significant added protection against malarial infection compared to ITN use alone.\nThere was also some evidence that anaemia prevalence was lower in communities with the combination.\nExposure to infectious mosquito bites was about one-sixth in communities with the combined intervention compared to those in the ITN only arm.\nTwo rounds of IRS with bendiocarb were conducted to overcome the short residual activity of the insecticide and to ensure that there was active ingredient on the walls of sprayed homes throughout the transmission season.\nIRS coverage in the ITN+IRS arm was high at approximately 90% in both spray rounds, which would have optimised its effectiveness.\nOn the other hand, whilst 85% of households owned at least one ITN, use of ITNs was modest, declining to 36% by the end of the study.\nThe low usage of ITNs means that the addition of IRS may have simply protected those who were not using an ITN, thus compensating for low ITN usage rather than offering additional protection to net users.\nThis interpretation is contradicted by the results of a per-protocol analysis, which excluded those not using ITNs, showing strong evidence that ITN users whose houses were sprayed were additionally protected by IRS.\nThe estimated reduction in PfPR associated with the combination of interventions was greater in the per-protocol analysis than in the ITT analysis in each survey.\nPer-protocol analysis excludes non-compliers (for IRS and ITN) and therefore may have been influenced by confounders.\nIt is likely that the observed overall effect of the intervention combination was a result of both IRS protecting those not using ITNs, and IRS additionally protecting ITN users.\nA potential negative impact of the combination of interventions is that having their house sprayed may encourage some residents to stop sleeping under an ITN.\nThis was not observed in this study; ITN usage was similar between the villages with and without IRS in each post-intervention survey.\nITN usage and ownership was slightly higher at baseline in the ITN+IRS arm compared to the ITN only arm, but the 95% confidence intervals for these estimates overlapped.\nThis non-significant difference could have led to a slight overestimation of the effect size.\nPfPR was slightly lower at baseline in the ITN+IRS arm compared to the ITN only arm, but the effect size did not change after adjusting for PfPR at baseline.\nThis suggests that baseline PfPR was not confounding the relationship between study arm and PfPR (the outcome).\nIn the baseline year, malaria prevalence was higher in June\u2013July after the long rainy season than in February\u2013March after the short rains.\nIn the intervention year, the prevalence similarly increased in June\u2013July (survey B) in the ITN only arm, but prevalence in the ITN+IRS arm remained low, suggesting IRS and ITNs in combination prevented the seasonal increase in infections.\nThe added protective effect of IRS peaked in the second survey, at the height of transmission after the long rains.\nThis was probably the optimal time for the insecticide to reduce the abundance of the mosquito population (N. Protopopoff, personal communication) and thus to observe the impact of IRS on the prevalence of malarial infections.\nThe limited residual activity of bendiocarb IRS has been shown to reduce its protective effectiveness 3\u20135 mo after spraying, which probably accounts for the loss of added benefit seen in the third survey, which was 6 mo after the last spray round at the beginning of the short rains.\nImplementing IRS with long-lasting insecticide formulations might be necessary to maintain the effectiveness of the combination throughout the year.\nAlternatively, the time between IRS rounds could be reduced, but this would considerably raise the cost of the combined intervention.\nThe secondary outcomes anaemia and EIR also pointed to added protection being provided by the combination of IRS and ITNs, but the evidence for these endpoints was weaker.\nThe combination intervention was associated with higher haemoglobin levels in children under 5 y, particularly at the peak of the transmission season.\nThe study had been powered to show a difference in the primary outcome (PfPR), and therefore may have been underpowered for these secondary outcomes.\nNevertheless, the results for all outcomes are consistent.\nOne of the limitations of this study is that clinical incidence of malaria could not be recorded in addition to infection prevalence because recording of confirmed malaria cases was unreliable because of stock-outs of RDTs at health facilities.\nImplementing both IRS and universal coverage of ITNs is obviously considerably more costly than ITNs alone.\nEstimating the cost-effectiveness of the combination compared to ITNs alone was beyond the scope of this particular research.\nAlthough IRS is known to be highly cost-effective, the marginal cost per case averted through using IRS in combination with ITNs should ideally be assessed in future studies.\nThis is particularly important in light of the funding gap that has been identified for meeting the demand for universal coverage of vector control for populations in malaria endemic regions.\nPrevious studies have investigated the combined use of multiple vector control methods versus one method alone, but the results have been inconsistent.\nThe only published trial data are from a 28-cluster, four-arm CRT carried out in Benin that compared (1) targeted coverage of LLINs (pregnant women and children only), (2) universal coverage of LLINs, (3) targeted coverage of LLINs combined with bendiocarb IRS, and (4) universal coverage of LLINs combined with bendiocarb-treated wall linings.\nThe study found no difference in malaria incidence, geometric mean parasite density, or mosquito abundance between any of the study arms.\nThe lack of any evidence of an added benefit of the combined interventions over the use of LLINs alone has to be viewed against the modest sample size, and hence potentially low power of this trial, and the lack of a comparator arm with universal coverage of ITNs.\nThere are a number of differences between the Benin trial and the current study that may have contributed to the discordant results.\nIn the Benin trial, the interval between IRS rounds was 8 mo, whereas it was only 4 mo in the current study, as IRS was timed according to the seasonal peaks in cases, and taking account of its short residual duration on walls.\nThe first two cross-sectional surveys for the current trial were timed to coincide with the seasonal peaks in cases and were only 2 mo after each IRS round, whereas in Benin the cases were recorded at 6-wk intervals for 18 mo, so that the measured effect of the additional IRS may include a period when the insecticide, which is known to have a short residual duration, was no longer effective.\nIn the Benin trial, LLINs were given only to target groups in the reference arm and in the study arm with IRS, whereas in the current trial ITNs were distributed to all age groups.\nLarge CRTs have recently been conducted in the Gambia and in Sudan comparing villages with IRS and LLINs to villages with only LLINs, but the results have not yet been published.\nEvidence of an added benefit from the combination intervention compared to IRS or ITNs alone has been shown in a number of observational studies.\nFor example, children 2\u201314 y old consistently received added personal protection from using nets in addition to IRS on the island of Bioko, Equatorial Guinea (OR 0.71, 95% CI 0.59\u20130.86), and in Zambezia, Mozambique (OR 0.63, 95% CI 0.50\u20130.79).\nIn Pakistan, nets provided added protection against P. vivax and P. falciparum in refugee camps where IRS was conducted.\nHowever, other studies observed no additional benefit from the combination compared to one intervention alone.\nOne interpretation of these divergent conclusions is that if the intervention present in both study arms is compromised or poorly implemented, the second method compensates for the deficiency of the first, providing apparent added protection that would otherwise not be seen.\nOn the other hand, if the reference arm intervention is well implemented and efficacious in both study arms, there may be little or no scope for additional protection by a second intervention.\nITN usage in the present trial was moderate, and hence the IRS protected many people who were not using a net in the ITN+IRS arm, whilst non-users in the ITN only arm remained unprotected.\nAny community or \u201cmass effect\u201d of ITNs on mosquito population size would have been limited because of the low community net usage.\nTherefore, the protective effect of ITNs in this study was possibly suboptimal.\nIn Bioko, ITNs provided personal protection in the presence of IRS that was rendered only partially effective by moderate coverage (77%\u201379%) and use of an insecticide that did not outlast the long malaria season.\nProtopopoff et al. reported that in Burundi there was no additional reduction in infection prevalence in children from adding LLINs to IRS because high coverage (90%) of IRS had already reduced the sporozoite rate to a level where nets had no further impact.\nIn Sao Tome, where the IRS programme was poorly implemented, with low coverage and long intervals between spray rounds, there was an additional benefit from using ITNs and IRS compared to IRS alone.\nHowever, on the neighbouring island of Principe, where IRS coverage was high (85%) and implemented on schedule, there was no added protection from ITNs in combination with IRS compared to IRS alone.\nInsecticide resistance may be another reason why differences have been seen for the effectiveness of the combination of IRS and ITNs, resulting in either an apparent \u201cadded\u201d effect of the second effective intervention, if the first was ineffective due to insecticide resistance, or no added effect if the second intervention was ineffective due to insecticide resistance.\nIn the study area of this trial, there was evidence for high levels of resistance to pyrethroids in An. gambiae s.s.\nThe epidemiological impact of pyrethroid resistance on the effectiveness of ITNs is currently not known.\nHowever, if the effectiveness of the ITNs was compromised because of insecticide resistance, this would have enhanced our estimate of the additional benefit of non-pyrethroid IRS.\nIf pyrethroid-treated nets were to be rendered partially ineffective in the presence of resistance, there would be a compelling case for combining ITNs with non-pyrethroid IRS.\nAn experimental hut trial in an area of Tanzania where the main vector is An. arabiensis found that if ITNs were used, the addition of IRS using insecticides with high irritancy such as dichlorodiphenyltrichloroethane (DDT) or lambda-cyhalothrin did not increase mosquito mortality or repel mosquitoes from the house.\nHowever, the addition of IRS using pirimiphos-methyl, an organophosphate that has high toxicity and low irritancy, did increase mosquito mortality.\nThese findings underscore that the interaction between the two interventions is complex and that the added protective effect will be dependent on the feeding and resting behaviours of particular malaria vectors, on the type of IRS insecticide used, on the susceptibility of local vectors to each of the insecticides in the combination, and on ITN usage.\nAs a result, added protection may not be observed in all situations.\nA systematic review of all the trial results estimating the effectiveness of the combination of ITNs and IRS should be undertaken once the results of the trials in Sudan and the Gambia are available.\nNevertheless, this trial provides encouraging evidence for an additional benefit from applying IRS in combination with ITNs compared to ITNs alone.\nTo our knowledge it is the first CRT to do so.\nThe added protection from the supplementary use of IRS may in the case of bendiocarb be limited to only a few months, raising the question of whether residual insecticides of short duration are cost-effective when used in combination with ITNs.\nThis study was conducted as an effectiveness study and not an efficacy study.\nThe LLINs were distributed by a national UCC and therefore represented a real-life malaria control programme, including the challenges faced in achieving high coverage and usage of ITNs.\nIn conclusion, national malaria control programmes should consider implementing IRS in combination with ITNs if local ITN strategies alone are insufficiently effective and cannot be improved.\nA key consideration would be the additional cost of providing the combined intervention.\nGiven the inconsistent trial evidence and the unproven generalisability of the findings of all studies that have investigated this question, it would be prudent for malaria control programmes implementing the two methods simultaneously to monitor the impact and cost-effectiveness of the combination to verify whether the additional resources have the desired effect.\nStudy timetable.Surveys 1 and 2 are baseline surveys. Surveys A, B, and C are post-intervention.\nTrial profile for study households and children in the ITN only and ITN+IRS study arms.Survey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray. *No children 0.5\u201314 y old. 1Dwelling vacant for survey duration. 2Includes not found (91.0%), not visited (2.4%), and missing data (6.6%). 3Households (HH) that were included and where children attended for testing.\n\nRestriction variables for randomisation and realisation of balance between the study arms.\nVariable | Maximum Difference in Means between Study Armsa | ITN Arma | ITN+IRS Arma | Actual Difference\nPfPRb in February\u2013March 2011c | 3% | 9.9% | 9.3% | 0.5%\nPfPR in June\u2013July 2011d | 3% | 22.4% | 19.6% | 2.7%\nHousing densitye | 20 HH/km2 | 165.1 HH/km2 | 152.6 HH/km2 | 12.5 HH/km2\nMean elevation | 50 m | 1,364.8 m | 1,330.7 m | 34.1 m\nITN usaged,f | 5% | 35.0% | 30.4% | 4.6%\nAdequate LLIN ownershipe,g | 5% | 61.3% | 56.3% | 5.0%\nClusters with entomological surveillance | Count of 2 | 20 clusters | 20 clusters | 0 clusters\n\nMeans for each study arm were calculated from cluster summaries.\n\nPfPR from RDTs.\nRecorded in baseline survey 1(February\u2013March 2011).\nRecorded in baseline survey 2 (June\u2013July 2011) after the UCC.\nHousing density in surveillance area of clusters.\nNet used the night before the survey in all age groups.\nPercentage of households with at least one LLIN per two people.\nHH, household.\n\nBaseline characteristics of individuals and households by study arm, Muleba District, 2011.\nCharacteristic | ITN Only ArmPercent [95% CI] (n) | ITN+IRS ArmPercent [95% CI] (n)\nPfPR in March 2011a,b,c | 10.3 [5.2\u201319.3] (2,487) | 8.4 [4.5\u201315.3] (2,655)\nPfPR in July 2011a,b,d | 24.6 [17.0\u201334.3] (2,121) | 21.0 [13.8\u201330.5] (2,185)\nModerate/severe anaemiaa,d,e | 6.4 [3.9\u201310.2] (785) | 6.1 [4.1\u20138.9] (841)\nMean haemoglobin (g/dl)a,d, | 10.6 [10.4\u201310.9] (785) | 10.6 [10.4\u201310.9] (841)\nITN use in all age groupsa,d,f | 53.3 [48.2\u201358.3] (6,755) | 58.2 [53.8\u201362.5] (6,913)\nHouseholds with adequate ITNsd,g,h | 54.5 [49.5\u201359.5] (1,243) | 62.3 [57.3\u201367.1] (1,250)\nHouseholds with \u22651 ITNd,g | 88.9 [86.0\u201391.3] (1,248) | 92.6 [90.8\u201394.0] (1,251)\nHouseholds received IRS in 2011c,g,i | 94.4 [91.3\u201396.5] (1,598) | 95.5 [93.5\u201396.9] (1,640)\nMean An. gambiae mosquitoes per house per nightg,j | 3.1 [1.0\u20139.6] (1,055) | 2.2 [0.5\u20139.1] (1,120)\nSporozoite ratea,k | 1.1 [0.8\u20131.4] (1,359) | 2.0 [1.4\u20132.8] (1,466)\nMean EIR/monthl | 1.1 [0.4\u20132.8] | 1.3 [0.4\u20134.4]\n\nCalculated from individual-level data.\n\nPfPR from RDTs.\nRecorded in baseline survey 1 (February\u2013March 2011).\nBaseline survey 2 (June\u2013July 2011) after the UCC.\nHaemoglobin <8 g/dl.\nReported sleeping under an ITN the night previous to the survey.\nCalculated from household-level data.\nAt least one ITN per sleeping place.\nApproximately 1 mo after spraying.\nArithmetic mean.\nProportion of mosquitoes positive for P. falciparum sporozoites.\nNumber of infective bites per month.\n\nIRS coverage, ITN ownership, and ITN usage in the intervention year, Muleba District, 2012.\nSurvey | Arm | Reported IRS CoverageaPercent [95% CI] (nb) | Adequate ITN OwnershipcPercent [95% CI] (nb) | \u22651 ITN OwneddPercent [95% CI] (nb) | ITN UseePercent [95% CI] (nf)\nSurvey A | ITN only | 3.3 [1.8\u20135.9] (1,177) | 52.2 [47.8\u201356.5] (1,178) | 85.8 [83.7\u201387.7] (1,177) | 46.6 [41.7\u201351.6] (2,193)\n | ITN+IRS | 92.1 [88.4\u201394.7] (1,215) | 57.2 [53.6\u201360.7] (1,215) | 89.0 [87.1\u201390.6] (1,216) | 53.0 [47.5\u201358.3] (2,349)\nSurvey B | ITN only | 5.2 [1.3\u201318.6] (1,094) | 51.6 [47.0\u201356.0] (1,094) | 82.5 [78.7\u201385.7] (1,096) | 40.7 [34.7\u201347.0] (2,045)\n | ITN+IRS | 89.5 [84.0\u201393.2] (1,138) | 57.4 [54.0\u201360.9] (1,142) | 88.2 [85.7\u201390.3] (1,142) | 44.1 [39.2\u201349.2] (2,207)\nSurvey C | ITN only | 13.0 [6.6\u201324.1] (1,165) | 52.8 [47.6\u201358.0] (1,168) | 78.2 [74.3\u201381.6] (1,170) | 36.0 [29.8\u201342.6] (2,101)\n | ITN+IRS | 89.3 [83.6\u201393.2] (1,209) | 56.8 [51.7\u201361.8] (1,211) | 83.8 [79.9\u201387.1] (1,211) | 36.1 [31.0\u201341.5] (2,303)\n\nSurvey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\nReported spray status of household in the spray round preceding the survey.\nHouseholds.\nPercentage of households with sufficient ITNs for at least one per sleeping place.\nPercentage of households with at least one ITN.\nPercentage of study children that reported sleeping under an ITN the night previous to the survey. ITN usage in all age groups was very similar to ITN use in the study children.\nIndividuals.\n\n\nPfPR in children 0.5\u201314 y old in the ITN only and ITN+IRS arms (intention to treat) in survey A, B, and C, Muleba District, Tanzania, 2012.\nSurvey | Arm | PfPRaPercent [95% CI] (n) | OR [95% CI], p-Value\nSurvey A | ITN only | 23.6 [15.4\u201334.2] (2,191) | 1.00\n | ITN+IRS | 13.6 [8.3\u201321.4] (2,342) | 0.51 [0.24\u20131.09], p\u200a=\u200a0.082\nSurvey B | ITN only | 30.5 [20.2\u201343.4] (2,033) | 1.00\n | ITN+IRS | 12.7 [7.4\u201321.0] (2,204) | 0.33 [0.15\u20130.75], p\u200a=\u200a0.009\nSurvey C | ITN only | 24.5 [14.2\u201338.9] (2,091) | 1.00\n | ITN+IRS | 13.4 [7.3\u201323.4] (2,285) | 0.48 [0.18\u20131.24], p\u200a=\u200a0.127\nAll three surveys combined | ITN only | 26.1 [16.7\u201338.4] (6,315) | 1.00\n | ITN+IRS | 13.3 [7.9\u201321.5] (6,831) | 0.43 [0.19\u20130.97], p\u200a=\u200a0.043b\n\nSurvey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\n\nPfPR from RDTs.\nAdjusted for survey.\n\nAnaemia and mean haemoglobin in children under 5+IRS arms (intention to treat), for survey A, B, and C, Muleba District, Tanzania, 2012.\nSurvey | Arm | Anaemia Prevalencea | Mean Haemoglobin (g/dl)\n |  | Percent [95% CI] (n) | OR [95% CI], p-Value | Mean [95% CI] (n) | Difference [95% CI], p-Value\nSurvey A | ITN only | 6.0 [4.1\u20138.7] (815) | 1.00 | 10.6 [10.4\u201310.8] (815) | \n | ITN+IRS | 3.9 [2.5\u20136.2] (864) | 0.64 [0.34\u20131.19], p\u200a=\u200a0.155 | 10.9 [10.7\u201311.1] (864) | 0.28 [\u22120.02 to 0.59], p\u200a=\u200a0.065\nSurvey B | ITN only | 4.7 [2.6\u20138.6] (737) | 1.00 | 10.9 [10.6\u201311.2] (737) | \n | ITN+IRS | 2.2 [1.3\u20133.6] (784) | 0.44 [0.20\u20131.01], p\u200a=\u200a0.053 | 11.4 [11.2\u201311.6] (784) | 0.49 [0.10 to 0.89], p\u200a=\u200a0.016\nSurvey C | ITN only | 3.2 [1.8\u20135.7] (739) | 1.00 | 10.8 [10.6\u201311.1] (739) | \n | ITN+IRS | 2.6 [1.6\u20134.4] (831) | 0.81 [0.37\u20131.77], p\u200a=\u200a0.590 | 11.2 [11.0\u201311.4] (831) | 0.36 [\u22120.02 to 0.73], p\u200a=\u200a0.060\nAll three surveys combined | ITN only | 4.7 [3.2\u20136.9] (2,291) | 1.00 | 10.8 [10.5\u201311.0] (2,291) | \n | ITN+IRS | 2.9 [2.0\u20134.3] (2,479) | 0.62 [0.34\u20131.10], p\u200a=\u200a0.102b | 11.2 [11.0\u201311.3] (2,479) | 0.37 [0.07 to 0.68], p\u200a=\u200a0.017b\n\nSurvey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\nPrevalence of moderate/severe anaemia (haemoglobin <8 g/dl).\nAdjusted for survey.\n\nMean number of An. gambiae mosquitoes per household, sporozoite rate, and EIR in the ITN only and ITN+IRS arms during the post-intervention period, Muleba District, Tanzania, 2011\u20132012.\nArm | Mean or Percent [95% CI] (n)a | Effect [95% CI], p-Value\nMeanbAn. gambiae per house per night\nITN only | 1.7 [0.5\u20136.4] (1,892) | \nITN+IRS | 0.4 [0.1\u20131.4] (1,893) | Rate ratio\u200a=\u200a0.23 [0.04\u20131.44], p\u200a=\u200a0.113\nSporozoite ratec\nITN only | 2.5 [2.1\u20133.1] (3,059) | \nITN+IRS | 1.8 [0.5\u20136.2] (717) | OR\u200a=\u200a0.72 [0.21\u20132.53], p\u200a=\u200a0.600\nMean EIR/monthd\nITN only | 1.3 [0.3\u20134.6] | \nITN+IRS | 0.2 [0.1\u20130.8] | Rate ratio\u200a=\u200a0.17 [0.03\u20131.08], p\u200a=\u200a0.059\n\nData are mean [95% CI] (number of houses) for mean An. gambiae per house per night and percent [95% CI] (number of An. gambiae) for sporozoite rate.\nArithmetic mean.\nProportion of mosquitoes positive for P. falciparum sporozoites.\nNumber of infective bites per month.\n\nPer-protocol analysis of PfPR in children 0.5\u201314 y old and anaemia in children under 5 y old in surveys A, B, and C.\nSurvey | Arm | PrevalencePercent [95% CI] (n) | OR [95% CI], p-Value\nPfPRa |  |  | \nSurvey A | ITNb | 26.7 [17.5\u201338.6] (954) | 1.00\n | ITN+IRSc | 12.3 [7.8\u201318.9] (1,142) | 0.39 [0.18\u20130.81], p\u200a=\u200a0.013\nSurvey B | ITNb | 35.5 [23.2\u201350.2] (782) | 1.00\n | ITN+IRSc | 10.2 [5.7\u201317.7] (892) | 0.21 [0.09\u20130.49], p\u200a=\u200a0.001\nSurvey C | ITNb | 29.4 [16.7\u201346.4] (707) | 1.00\n | ITN+IRSc | 10.1 [5.4\u201318.2] (770) | 0.27 [0.10\u20130.73], p\u200a=\u200a0.011\nAnaemiad |  |  | \nSurvey A | ITNb | 5.9 [3.5\u20139.7] (390) | 1.00\n | ITN+IRSc | 3.8 [1.8\u20137.5] (453) | 0.62 [0.25\u20131.55], p\u200a=\u200a0.301\nSurvey B | ITNb | 5.4 [2.2\u201312.5] (295) | 1.00\n | ITN+IRSc | 1.9 [0.8\u20134.1] (374) | 0.33 [0.10\u20131.12], p\u200a=\u200a0.076\nSurvey C | ITNb | 4.0 [2.2\u20137.0] (303) | 1.00\n | ITN+IRSc | 2.3 [1.0\u20135.0] (305) | 0.57 [0.21\u20131.55], p\u200a=\u200a0.264\n\nMuleba, Tanzania, 2012; analysis restricted to ITN users in both study arms. Survey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\n\nPfPR from RDTs.\nITN used by the individual the night preceding the survey in the ITN only arm.\nITN used by the individual the night preceding the survey, and household with IRS in the ITN+IRS arm. One cluster that was allocated to be in the ITN only arm but received IRS in the second spray round was excluded from this analysis.\nPrevalence of moderate/severe anaemia (haemoglobin <8 g/dl).",
        "label": "low",
        "id": "task4_RLD_test_275"
    },
    {
        "paper_doi": "10.1371/journal.pmed.1002433",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Non-blinded, cluster-RCT. 10 clinics in Mozambique were randomized to either intervention (CIS) or control (standard care). A pre-post intervention 2-sample design was nested within the intervention arm to assess the additional effectiveness of an enhanced version of the CIS, referred to as CIS+. Consequently, the standard care arm enrolled 1 cohort of patients, while the intervention arm enrolled 2 sequential cohorts of patients (CIS and CIS+). CIS+ participants were enrolled after CIS enrolment was completed at each facility randomized to the intervention arm.\n\n\nParticipants: 5327 participants5 clinics were selected from urban areas and 5 from rural areasInclusionAll adults testing HIV-positive in the VCT clinics within the participating health facilitiesExclusion< 18 years of age.Pregnant.Planned to move from their community of residence in the next 12 months.Had enrolled in HIV care or initiated ART in the past 6 months.Did not understand Portuguese or Xitsua.Were incapable of providing informed consent.\n\n\nInterventions: Intervention arm: 4 evidence-based interventions that simplified the clinic flow and encouraged linkage to and retention in care:Pima (Inverness Medical Innovations) CD4 assay machines in the VCT clinics to enable HIV testing counsellors to provide real-time, POC CD4 test results immediately following diagnosis.Participants with Pima CD4 cell count < 350 cells/mm3 were provided with rapid ART initiation. These individuals received an individual ART preparatory counselling session in the VCT clinic immediately following CD4 testing, on the day of diagnosis. Facility receptionists were instructed to expedite appointments for these participants when they presented to schedule their clinical consultations. Clinicians were encouraged to initiate ART at the first clinical visit.Participants received health messages and appointment reminders via SMS messaging.participants in the CIS+ cohort received the CIS interventions plus a series of non-cash FIs in the form of prepaid cellular air-time cards.Control: standard care - participants were managed as per prevailing Ministry of Health guidelines.Individuals diagnosed with HIV received post-test counselling in the VCT clinic and were referred verbally to HIV services.Participants presenting to the facility receptionist to schedule a clinical consultation for HIV care were referred to the laboratory for CD4 cell count, chemistry, and haematology testing, and provided with an appointment 2-4 weeks later to allow sufficient time for the laboratory results to be received.ART eligibility was determined at that first clinical consultation based on CD4 cell count < 350 cells/mm3 and/or WHO stage 3/4.Those found to be eligible for ART received at least 1 individual counselling session before initiating treatment.For ART-eligible participants, the time interval between enrolment in HIV care and ART initiation was estimated at 1-2 months at the time the study started.Participants initiating ART were requested to return every 2 weeks for the first month, at 2 months, at 6 months, and every 6 months thereafter.ART-ineligible participants were instructed to return at 6 months for repeat clinical evaluation and laboratory testing.\n\n\nOutcomes: Mortality and viral suppression at 12 months, time to ART initiation, linkage to care at 1 month, retention in care at 6 months, disease progression\n\n\nNotes: \n\n",
        "objective": "To assess the effects of interventions for rapid initiation of ART (defined as offering ART within seven days of HIV diagnosis) on treatment outcomes and mortality in people living with HIV. We also aimed to describe the characteristics of rapid ART interventions used in the included studies.",
        "full_paper": "Background\nConcerning gaps in the HIV care continuum compromise individual and population health.\nWe evaluated a combination intervention strategy (CIS) targeting prevalent barriers to timely linkage and sustained retention in HIV care in Mozambique.\nMethods and findings\nIn this cluster-randomized trial, 10 primary health facilities in the city of Maputo and Inhambane Province were randomly assigned to provide the CIS or the standard of care (SOC).\nThe CIS included point-of-care CD4 testing at the time of diagnosis, accelerated ART initiation, and short message service (SMS) health messages and appointment reminders.\nA pre\u2013post intervention 2-sample design was nested within the CIS arm to assess the effectiveness of CIS+, an enhanced version of the CIS that additionally included conditional non-cash financial incentives for linkage and retention.\nThe primary outcome was a combined outcome of linkage to care within 1 month and retention at 12 months after diagnosis.\nFrom April 22, 2013, to June 30, 2015, we enrolled 2,004 out of 5,327 adults \u226518 years of age diagnosed with HIV in the voluntary counseling and testing clinics of participating health facilities: 744 (37%) in the CIS group, 493 (25%) in the CIS+ group, and 767 (38%) in the SOC group.\nFifty-seven percent of the CIS group achieved the primary outcome versus 35% in the SOC group (relative risk [RR]CIS vs SOC = 1.58, 95% CI 1.05\u20132.39).\nEighty-nine percent of the CIS group linked to care on the day of diagnosis versus 16% of the SOC group (RRCIS vs SOC = 9.13, 95% CI 1.65\u201350.40).\nThere was no significant benefit of adding financial incentives to the CIS in terms of the combined outcome (55% of the CIS+ group achieved the primary outcome, RRCIS+ vs CIS = 0.96, 95% CI 0.81\u20131.16).\nKey limitations include the use of existing medical records to assess outcomes, the inability to isolate the effect of each component of the CIS, non-concurrent enrollment of the CIS+ group, and exclusion of many patients newly diagnosed with HIV.\nConclusions\nThe CIS showed promise for making much needed gains in the HIV care continuum in our study, particularly in the critical first step of timely linkage to care following diagnosis.\nTrial registration\nClinicalTrials.gov NCT01930084\nIn a cluster-randomized trial done in Mozambique, Batya Elul and colleagues study a combined intervention for linkage to and retention of people with HIV in care.\nAuthor summary\nWhy was this study done?\nIn sub-Saharan Africa, HIV testing, care, and treatment programs have been widely scaled up over the past decade, but suboptimal outcomes across the HIV care continuum\u2014particularly with regards to timely linkage to and sustained retention in care\u2014compromise their effectiveness.\nPatients experience multiple barriers to linkage to and retention in HIV care including health system barriers, structural barriers, and behavioral barriers, yet prior studies have largely evaluated individual interventions targeting a single barrier to care.\nOur study was designed specifically to examine the effectiveness of a combination intervention strategy (CIS) composed of several scalable evidence-based interventions targeting the multiple and prevalent health system, structural and behavioral barriers that patients face across the HIV continuum.\nWhat did the researchers do and find?\nWe randomly assigned 10 primary health facilities in the city of Maputo and Inhambane Province in Mozambique to provide the standard of care (SOC) or the CIS, which included point-of-care CD4 testing at the time of diagnosis, accelerated ART initiation, and short message service (SMS) health messages and appointment reminders.\nA pre\u2013post intervention 2-sample design was nested within the intervention arm to assess the effectiveness of CIS+, an enhanced version of the CIS that additionally included conditional non-cash financial incentives for linkage and retention.\nWe enrolled 2,004 adults diagnosed with HIV in the voluntary counseling and testing clinics of participating health facilities, and compared the proportion who achieved a combined outcome of linkage to HIV care within 1 month of diagnosis and retention in care at 12 months across the 3 study groups.\nWe found an increased likelihood of achieving the combined outcome in the CIS group compared to the SOC group, driven primarily by very large increases in same-day linkage, but no difference between the CIS+ and CIS groups.\nWhat do these findings mean?\nThe CIS may help improve outcomes across the HIV care continuum in high-burden settings, particularly in the critical first step of timely linkage to care following diagnosis.\nFurther research is needed to understand whether financial incentives can be optimized in this setting, given their effectiveness in enhancing other health outcomes.\nIntroduction\nAlthough the extraordinary scale-up of HIV testing, care, and treatment programs in sub-Saharan Africa over the past decade has resulted in more than 19 million persons accessing antiretroviral therapy (ART), the effectiveness of these programs has been significantly hindered by high levels of attrition across the HIV care continuum.\nObservational studies and systematic reviews have repeatedly reported disturbing gaps in care as patients move from HIV testing clinics to HIV care clinics (i.e., linkage to care) and that patient dropout among those enrolled in HIV care is far too common, both before and after ART initiation (i.e., retention in care).\nIndeed, available data suggest that less than 1/3 of individuals who are diagnosed with HIV are successfully linked to and remain engaged in HIV care 12 months later.\nBarriers to timely linkage to and sustained retention in HIV care have been well documented, and include health system barriers (e.g., multiple HIV clinic visits for counseling and clinical and laboratory assessments prior to ART initiation), structural barriers (e.g., transport costs and distances, work and childcare constraints), and behavioral barriers (e.g., forgetting appointments, lack of understanding of required care).\nPrior studies have overwhelmingly evaluated individual interventions targeting a single barrier at a single point in the HIV care continuum such as mobile phone short message service (SMS) messaging to augment linkage to care following diagnosis, or point-of-care CD4 testing to enhance retention among patients enrolled in HIV care.\nHowever, it is increasingly recognized that multi-component approaches composed of several practical, evidence-based interventions that simultaneously target the multiple and recurrent barriers that patients face as they navigate across the HIV care continuum are needed to maximize individual and population health.\nFurther, implementation science research that evaluates proposed multi-component approaches in real-world settings is needed to assess not only effectiveness, but also implementation outcomes including reach, adoption, and sustainability.\nTo this end, we designed a combination intervention strategy (CIS) composed of several scalable evidence-based interventions targeting prevalent health system, structural, and behavioral barriers across the HIV care continuum, and determined its effect on a combined outcome of linkage to and retention in HIV care among adults newly diagnosed with HIV in Mozambique, while also collecting information on its implementation and potential for broader scale-up.\nData regarding intervention feasibility and patient acceptability have been published, and thus we present here the effectiveness results.\nBecause the interventions included in the CIS are expected to be implemented at the facility level, as opposed to targeted at specific individuals, should they be scaled up, we evaluated effectiveness using a cluster design, which best mirrors this implementation approach.\nMethods\nA detailed description of the study protocol has been published.\nEthics statement\nEthical approval was provided by Mozambique\u2019s National Committee for Bioethics for Health and Columbia University\u2019s institutional review board (IRB) (protocol AAAL1354).\nInformed written consent was obtained from all participants.\nStudy design\nBetween April 22, 2013, and June 30, 2016, we conducted a 2-arm cluster-randomized study (effectiveness\u2013implementation hybrid design, Type 1) in health facilities in Maputo and Inhambane Province in Mozambique in order to assess the effectiveness of the CIS.\nAdditionally, a pre\u2013post intervention 2-sample design was nested within the intervention arm to assess the additional effectiveness of an enhanced version of the CIS, referred to as CIS+.\nConsequently, the standard of care (SOC) arm enrolled 1 cohort of patients, while the intervention arm enrolled 2 sequential cohorts of patients (CIS and CIS+).\nCIS+ participants were enrolled after CIS enrollment was completed at each facility randomized to the intervention arm.\nStudy setting\nThe city of Maputo, the nation\u2019s capital, has an area of 300 km2 and an estimated population of 1,225,868, with an HIV prevalence of 16.9% among those aged 15 to 59 years.\nThe Maputo City Health Network has a total of 37 health facilities, 32 of which offered comprehensive HIV care and treatment services at the time of study implementation.\nIn contrast, Inhambane is a rural province, with an estimated 1,475,318 people spread across 68,615 km2.\nHIV prevalence among adults aged 15 to 59 years is 14.1%.\nThe ratio of doctors to population (5.96/100,000) is one of the lowest in the country.\nOf the 135 health facilities in the province, 76 offered HIV care and treatment services when our study was initiated.\nSuboptimal health facility infrastructure, long distances to facilities, and weak referral systems in the province are all believed to compromise health service uptake.\nRandomization\nPrimary health facilities providing HIV testing, care, and treatment services and operated by the Ministry of Health with technical support from the Center for Collaboration in Health, a local PEPFAR implementing partner, were the unit of randomization.\nWe focused on primary health facilities, rather than larger provincial hospitals, to reflect the increasingly decentralized nature of HIV service delivery in Mozambique.\nTen facilities in Maputo (N = 4) and Inhambane Province (N = 6) were selected from the 66 primary health facilities receiving technical support from the Center for Collaboration in Health in those regions.\nParticipating facilities were purposely chosen because they had the highest volume of adults testing HIV positive and enrolling in HIV care in the year prior to study start and thus were expected to have sufficient participants for appropriate power.\nFacilities were matched into pairs by region (Maputo or Inhambane), level of urbanicity (urban versus rural), and average number of patients testing HIV positive in voluntary counseling and testing (VCT) in the year prior to study initiation (high versus low), resulting in 5 matched pairs.\nMatched pairs were randomized by one of the authors (MRL) using a computerized random number generator to either the CIS arm or the SOC arm using matched-pair randomization.\nSequences were concealed until interventions were assigned.\nThe study was non-blinded.\nStudy population\nParticipants were enrolled in the SOC group beginning on April 22, 2013, and in the CIS group beginning on April 25, 2013.\nThe last patient was enrolled in the SOC group on November 20, 2014, and the last patient in the CIS group was enrolled on February 11, 2015.\nEnrollment in the CIS+ group began after each clinic randomized to the intervention arm completed CIS enrollment, and ran from June 16, 2014, through June 30, 2015.\nAll participants were followed for 12 months, with the last patient completing follow-up on June 30, 2016.\nBroad inclusion criteria were used to reflect as accurately as possible the population of adults newly diagnosed with HIV in VCT clinics at the participating health facilities.\nWe focused on individuals newly diagnosed in VCT clinics, as opposed to those diagnosed in antenatal clinics and tuberculosis clinics, because the latter groups of patients typically follow a modified clinic flow.\nAll adults testing HIV positive in the VCT clinics within the participating health facilities were informed of the study by HIV testing counselors following diagnosis, and those who were interested were referred to study staff for further information, eligibility screening, and consent procedures.\nPatients were excluded if they were less than 18 years of age, were pregnant, planned to move from their community of residence in the next 12 months, had enrolled in HIV care or initiated ART in the past 6 months, did not understand Portuguese or Xitsua, or were incapable of providing informed consent.\nStudy participants agreed to be referred to HIV care and treatment services at the same facility where they were diagnosed (referred to as the \u201cdiagnosing facility\u201d); to complete a baseline, 1-month, and 12-month interview; to be traced at their homes if they could not be reached by phone for follow-up interviews; to provide contact information for a family member or friend who could provide information on their vital status if they could not be located for a follow-up interview; and, if they enrolled in HIV care and treatment services at the diagnosing facility, to have their clinical data abstracted from the facility\u2019s existing electronic medical records.\nStudy interventions\nStandard of care\nParticipants at health facilities randomized to receive the SOC were managed as per prevailing Ministry of Health guidelines.\nIndividuals diagnosed with HIV received post-test counseling in the VCT clinic and were referred verbally to HIV services, typically in the diagnosing facility.\nPatients presenting to the facility receptionist to schedule a clinical consultation for HIV care were referred to the laboratory for CD4 cell count, chemistry, and hematology testing, and provided with an appointment 2\u20134 weeks later to allow sufficient time for the laboratory results to be received.\nART eligibility was determined at that first clinical consultation based on CD4 cell count \u2264 350 cells/mm3 and/or WHO stage 3/4.\nThose found to be eligible for ART received at least 1 individual counseling session before initiating treatment.\nFor ART-eligible patients, the time interval between enrollment in HIV care and ART initiation was estimated at 1\u20132 months at the time the study started.\nParticipants initiating ART were requested to return every 2 weeks for the first month, at 2 months, at 6 months, and every 6 months thereafter.\nART-ineligible patients were instructed to return at 6 months for repeat clinical evaluation and laboratory testing.\nCombination intervention strategy\nAt facilities randomized to the intervention arm, we introduced 4 evidence-based interventions that simplified the clinic flow and encouraged linkage to and retention in care.\nThese interventions targeted several known health system, structural, and behavioral barriers across the HIV care continuum, and were adapted for the on-the-ground realities\u2014including practice norms, physical space, and available staffing\u2014at the health facilities.\nFirst, we introduced Pima (Inverness Medical Innovations) CD4 assay machines in the VCT clinics to enable HIV testing counselors to provide real-time, point-of-care CD4 test results immediately following diagnosis, and thus addressed a health system barrier by reducing the number of visits required for CD4 testing.\nWe also hypothesized that receipt of additional information on one\u2019s health at the time of diagnosis would advance patient understanding of the need for care, a documented behavioral barrier.\nAll patients regardless of CD4 count were provided with a paper-based referral to on-site HIV services that included their CD4 count, and were instructed to present for their first clinical consultation within 1 week.\nSecond, to address additional health system barriers, patients with Pima CD4 cell count \u2264 350 cells/mm3 were provided with accelerated ART initiation, with the ultimate goal of decreasing the HIV morbidity and mortality that contributes to significant attrition among ART-eligible patients.\nThese individuals received an individual ART preparatory counseling session in the VCT clinic immediately following CD4 testing, on the day of diagnosis.\nFacility receptionists were instructed to expedite appointments for these patients when they presented to schedule their clinical consultations.\nAlthough the patients were directed to the laboratory to have their blood drawn for baseline laboratory tests required by national ART guidelines, clinicians were encouraged to initiate ART at the first clinical visit rather than await the results of the laboratory tests unless the patient presented with comorbid conditions.\nPatients who initiated ART received a 2-week supply and followed the visit schedule dictated by national guidelines, similar to the SOC procedures.\nOnce baseline laboratory results were available, they were reviewed by clinic staff, and if abnormalities were noted, the participant was contacted to return to the clinic.\nThird, participants received health messages and appointment reminders via SMS messaging to address behavioral barriers associated with deferring care engagement and forgetting appointments.\nThe messages were sent from the central study office to the participant\u2019s phone or to a friend or relative\u2019s phone per participant preference, and did not refer to HIV or a specific health facility or reveal any personal information.\nThe health messages encouraged participants to care for their health, and were sent weekly for 1 month following diagnosis and then monthly (e.g., \u201cHi.\nYour health is the most important thing.\nPlease remember to come to the health center for health services.\u201d).\nAppointment reminders were sent only to participants who linked to care at the diagnosing facility, and were sent 3\u20137 days before each scheduled clinic visit (e.g., \u201cHi.\nYour health is the most important thing.\nWe expect to see you at your upcoming appointment scheduled for the day ___.\u201d).\nParticipants were not asked to confirm receipt or reply to the messages.\nFinally, patients in the CIS+ cohort received the CIS interventions plus a series of non-cash financial incentives (FIs) in the form of prepaid cellular air-time cards to offset structural barriers associated with the direct and indirect costs of coming to the health facility to receive HIV care.\nAir-time cards rather than cash were selected as the incentive based on discussion with the Ministry of Health.\nEach card was valued at approximately US$5 and was provided conditionally upon the following achievements: linkage to care within 1 month of diagnosis, retention in care 6 months after diagnosis, and retention in care 12 months after diagnosis, for a total of approximately US$15.\nParticipants who completed each achievement received the card when presenting for routine services.\nParticipants without cellular phones could opt to give them to a family member, sell them for cash, or trade them for other goods.\nBoth the point-of-care CD4 testing and accelerated ART initiation interventions were provided by health facility staff to all individuals diagnosed with HIV in the VCT clinic regardless of whether they were enrolled in the study, while the SMS messages and FIs were provided by study staff and only to study participants.\nData collection and outcomes\nSite assessments\nData on the configuration of HIV services at the 10 participating study sites were collected at the beginning and at the end of the study using a standardized site assessment form.\nThe purpose of the site assessments was to identify important similarities and differences between participating health facilities, as well as to better understand how services at the site could impact study implementation.\nBaseline interview\nParticipants completed closed-ended questionnaires administered by trained research assistants at the time of study enrollment.\nThe questionnaire took about 30 minutes to complete, and gathered information on sociodemographic characteristics, social and family support, mental health, alcohol use, HIV testing history, HIV knowledge and beliefs, and anticipated stigma and barriers to care.\nAnticipated stigma was assessed through 6 items adapted from the 12-item anticipated HIV stigma index developed by Earnshaw and Chaudoir.\nStigma scores were summed, then dichotomized into 2 groups: highest (>75th percentile) versus lower anticipated stigma.\nMental health was assessed via a 7-question evaluation based on the Kessler 10-item scale for psychological distress.\nMental health scores were summed, then dichotomized into 2 groups: highest (<75th percentile) versus lower level of distress.\nPerceived availability of social support was assessed with 4 questions adapted from a 9-item scale by Wortman and colleagues.\nSocial support scores were summed, then dichotomized into 2 groups: higher (>50th percentile) versus lower social support.\nQuestions assessing HIV-related knowledge and attitudes were based on those used by one of the authors in a previous study.\nHIV knowledge scores were summed, then dichotomized into 2 groups: higher (>50th percentile) versus lower knowledge.\nBaseline interview data were double-entered into a study database, and a computer program identified discrepant double-entered results for correction against the paper-based forms.\nPatient tracing and follow-up interviews\nOne and 12 months after enrollment, up until June 30, 2016, trained research assistants contacted participants by phone to ascertain their vital status and HIV care status, and to administer follow-up questionnaires.\nIf the participant could not be contacted by phone after 3 attempts, research assistants visited the participant\u2019s home up to 3 times.\nParticipants who were located completed closed-ended interviews that gathered updated information on key domains from the baseline questionnaire, as well as self-reported information on linkage to (1- and 12-month questionnaires) and retention in HIV care (12-month questionnaire only), reasons for linkage/non-linkage (1- and 12-month questionnaires) and retention/non-retention (12-month questionnaire only), ART status, hospitalizations, and anticipated stigma.\nIn cases where the participant could not be located, research assistants contacted a friend or family member as specified by the participant at study enrollment.\nResearch assistants did not refer to HIV or the health facility during contact tracing but rather attempted to determine whether the participant was alive or dead.\nFor those whose vital status could not be determined through contact tracing, research assistants searched existing electronic medical records at other primary health facilities supported by the Center for Collaboration in Health in the same district to assess whether patients had enrolled in HIV care at another facility, and reviewed death registers at the municipal and provincial levels to ascertain their vital status.\nSimilar data entry and reconciliation procedures to those used for the baseline interview data were used for the tracing and follow-up data.\nAbstraction of clinical data for patients linking to HIV care at the diagnosing facility\nAs part of routine clinical practice for HIV patients, clinicians documented patient information at every clinic visit on national HIV care forms, and trained data clerks entered those data into an Access-based electronic medical record.\nIn its role as a PEPFAR implementing partner supporting the study sites, the Center for Collaboration in Health assessed the completeness and accuracy of these electronic data every 4 months and initiated targeted interventions to enhance data quality if there was greater than 15% disagreement on key data elements between the electronic and paper-based systems.\nDuring the study period, research assistants reviewed the electronic medical records to identify study participants who had linked to care at their diagnosing facility.\nFor those located, we extracted the complete electronic medical record, capturing information on visit dates, vital status, transfer status, ART status, laboratory test results, and opportunistic infections.\nOutcomes\nThe primary outcome was a combined outcome of linkage to HIV care within 1 month of diagnosis plus retention in care 12 months after diagnosis measured at the individual level.\nWe used a combined outcome to reflect the fact that improvements are needed across the HIV care continuum in order to maximize individual and population health.\nLinkage to care was defined by at least 1 clinical consultation for HIV that included assessment of the patient\u2019s medical history and a physical exam.\nRetention in care was defined by a clinic visit in the 90 days prior to the end of the 12-month study follow-up period, with no documentation that the patient had transferred to another facility or had died.\nWe assessed the combined outcome from the perspective of the diagnosing health facility using data from the electronic medical records maintained by the HIV clinics.\nAll study participants were included in these analyses, including those who did not complete follow-up interviews.\nParticipants whose electronic medical records were not located were considered not to have achieved the combined outcome for this analysis.\nAs a secondary approach, we evaluated the combined outcome from the perspective of the Mozambican health program by supplementing data from the electronic medical records with patient reports of linkage to and retention in care at HIV clinics at different health facilities (obtained during follow-up interviews) and information obtained from electronic medical records at other health facilities.\nIn these analyses, participants whose self-reported linkage and retention status suggested they were linked to and/or retained at a health facility other than their diagnosing clinic were considered to have achieved the respective linkage/retention outcomes.\nParticipants who either did not complete follow-up interviews or did not self-report linkage to or retention at another clinic maintained their initial outcome designation.\nAll study participants were included in these analyses.\nSecondary outcomes included linkage to care at several predefined time points, ART eligibility assessment (defined as receipt of WHO staging and/or CD4 cell count), ART initiation, disease progression (defined as a new WHO stage 3/4 condition or hospitalization noted in the electronic medical records or self-reported during follow-up interviews), retention in care 6 and 12 months after diagnosis regardless of the timing of linkage, and death.\nStatistical analysis\nThe trial was designed and powered to measure outcomes at the individual level, with outcomes assessed within each cluster (5 clusters per arm).\nIn our initial power calculations, we anticipated that an average of 200 patients per clinic (in the CIS and SOC arms) would be eligible for enrollment based on historical data on the annual number of adults testing positive in the VCT clinics at the participating health facilities.\nWith 5 facilities per study arm, an average of 200 patients per facility, an intraclass correlation coefficient (ICC) of 0.05, and an alpha of 0.05 and assuming that 35% of participants in the SOC arm would achieve the primary outcome, we estimated that the study would have 80% power to detect as statistically significant 55% of participants in the CIS group achieving the primary outcome, and greater than 80% power to detect as statistically significant 75% of participants in the CIS+ group achieving the primary outcome.\nBecause enrollment proceeded slower than originally planned, at study midpoint we assessed the implications for power if each health facility enrolled an average of 150 participants rather than 200.\nOur calculations revealed minimal change in power with this reduction in the number of participants per health facility.\nCalculations were performed using PASS 8.0 software for 2 independent proportions in a cluster randomization study design and a 2-sided Farrington and Manning Likelihood Score Test.\nOur power estimations and statistical analyses did not take into account the pair matching prior to randomization but rather followed recommendations from Diehr et al. to break matches in statistical analyses of clustered studies when the number of pairs is between 3 and 9.\nAn intent-to-treat analysis determined the relative risk (RR) of achieving study outcomes between the CIS and SOC groups, and between the CIS+ and CIS groups.\nFor analyses of the primary outcome, we used random-intercept multilevel log-Poisson models to account for clustering within health facilities with an empirical variance adjustment for small numbers of sampling units described by Morel et al..\nWe also assessed whether the primary outcome differed after adjustment for patient-level factors by constructing propensity scores that estimated the probability of inclusion in the CIS, CIS+, and SOC groups by age, sex, region, education, income, employment status, marital status, religion, prior year history of being away from home for more than 1 month, travel time to clinic, tuberculosis status, past hospitalizations, diagnosis history, and whether another family member was known to be living with HIV.\nThe propensity score was included as a covariate in the multivariable log-Poisson models (adjusted analyses).\nIn post hoc analyses, we further estimated the likelihood of key subgroups achieving the primary outcome using interaction contrast ratios.\nThe subgroups assessed included subgroups based on baseline age, sex, region of health facility, employment status, marital status, whether the participant was away from home for more than 1 month in the year prior to study enrollment, travel time to clinic, whether a household member was known to be HIV positive, and dichotomous variables based on scales for self-reported anticipated stigma, HIV knowledge, mental health, and perceived social support as described earlier.\nFor analyses of secondary outcomes, log-Poisson models were used for dichotomous outcomes, and t tests and 2-way median tests as appropriate for continuous outcomes, adjusting for clustering but not for patient-level differences.\nResults\nHealth facility characteristics\nAs noted above, 10 primary health facilities participated in the study, 4 in Maputo and 6 in Inhambane.\nAt study start, the 5 health facilities randomized to the intervention arm reported that they had experienced disruptions of 3 or more days in VCT services in the prior 12 months, while only 1 facility randomized to the SOC arm reported experiencing a similar disruption.\nBy study end, no facilities\u2014whether in the intervention or SOC arm\u2014had experienced such disruptions.\nThroughout the study, only intervention sites conducted point-of-care CD4 testing using Pima machines in the VCT clinic.\nTwo SOC sites reported that they had Pima machines available in their laboratories but only used them to monitor CD4 counts after patients had enrolled in HIV care.\nNone of the SOC sites used SMS messaging for health messages or appointment reminders on a routine basis for all patients, but 2 sites sent SMS appointment reminders for patients participating in community ART groups.\nThough the 2013 national HIV treatment guidelines stipulate that 1 ART preparatory counseling session is required for ART-eligible patients, all the facilities participating in the study typically conducted 2 to 3 sessions prior to ART initiation, with a slight reduction in the number of sessions observed between study start and end.\nEnrollment and participant characteristics\nFig 1 shows the enrollment, exclusion, and flow of the patients by study group.\nDuring the study period, 5,327 adults \u226518 years of age were diagnosed with HIV in the VCT clinics at the 10 study facilities.\nA total of 265 of those individuals were not referred to the study staff for further information on the study because they informed the HIV testing counselor that they were not interested in the study, were already receiving HIV services, or were not willing to be referred to the diagnosing health facility.\nAmong the 5,062 who were referred to the study staff for further information, 3,058 did not meet study eligibility criteria.\nThe main reasons for exclusion were inability to provide informed consent due to distress following diagnosis (19%), inability to understand Portuguese or Xitsua (12%), and refusal to be referred to the diagnosing health facility for HIV services (10%).\nA total of 2,004 adults \u226518 years of age enrolled in the study at the 10 health facilities: 744 (37%) in the CIS group, 493 (25%) in the CIS+ group, and 767 (38%) in the SOC group.\nThe majority of participants were female (64%), and the median age of participants was 34 years of age, with no meaningful differences observed by study group (Table 1).\nMore than half of the participants (53%) were living with a partner at the time of diagnosis, and 65% of participants had a primary or lower level of education.\nMost participants (74%) were employed, and 43% had a monthly income of less than 1,500 meticais (approximately US$50).\nOne-quarter (27%) reported that another household member was living with HIV.\nWhile no serious adverse events were reported during the study period, there was 1 unanticipated event of a female participant reporting intimate partner violence.\nThe Mozambican National Committee for Bioethics for Health and the Columbia University IRB were informed of this event, and the participant asked to remain in the study but to conduct all study interviews at the facility (i.e., no follow-up phone calls).\nIntervention effect on linkage to and retention in HIV care at the diagnosing facility\nAs shown in Table 2, the CIS was associated with statistically significant improvements in the combined outcome of linkage to care within 1 month of diagnosis and retention in care 12 months following diagnosis when compared to the SOC.\nAnalyses using data from electronic medical records to examine linkage to and retention at the diagnosing health facility showed that 57% of participants in the CIS group achieved the primary outcome versus 35% of those in the SOC group (RRCIS vs SOC = 1.58, 95% CI 1.05\u20132.39).\nPost hoc calculation of the ICC for the primary outcome according to the methods of Snijders and Bosker for binary outcome data estimated an ICC of 0.066, similar to but slightly higher than the assumed ICC of 0.05 used in power and sample size estimation.\nThese results were robust to adjustment for patient-level differences (adjusted RR [aRR]CIS vs SOC = 1.55, 95% CI 1.07\u20132.25).\nAs shown in Fig 2, the greatest intervention effects were observed among young adults age 18\u201324 years (RRCIS vs SOC = 2.39, 95% CI 1.51\u20133.80, p-value for interaction between age and treatment arm = 0.07), those in Maputo (RRCIS vs SOC = 2.31, 95% CI 1.90\u20132.79, p-value for interaction between region and treatment arm < 0.0001), those who did not report that another household member was living with HIV (RRCIS vs SOC = 1.81: 95% CI 1.52\u20132.16, p-value for interaction between household member with HIV and treatment arm = 0.11), and those reporting high levels of anticipated stigma at enrollment (RRCIS vs SOC = 1.95, 95% CI 1.53\u20132.49, p-value for interaction between stigma and treatment arm = 0.10).\nEighty-nine percent of participants in the CIS group linked to the diagnosing facility on the same day as diagnosis compared to 16% (RRCIS vs SOC = 9.13, 95% CI 1.65\u201350.40) in the SOC group, 91% within 1 week compared to 46% (RRCIS vs SOC = 2.43, 95% CI 0.70\u20138.41), and 94% within 1 month compared to 63% (RRCIS vs SOC = 1.48, 95% CI 0.93\u20132.35).\nBy 12 months, nearly all CIS participants (96%) had linked to care compared to 77% (RRCIS vs SOC = 1.23, 95% CI 1.03\u20131.48) of SOC participants.\nAmong those linking to care, the median (interquartile range [IQR]) time from diagnosis to linkage was 0 days (0\u20130) in the CIS group and 3 days (1\u201326) in the SOC group (median test p < 0.001 for CIS versus SOC).\nThe effect of the intervention on retention in care, regardless of the timing of linkage, was more modest but statistically significant (6-month retention: 62% CIS versus 53% SOC, RRCIS vs SOC = 1.18, 95% CI 1.00\u20131.39; 12-month retention: 58% CIS versus 44% SOC, RRCIS vs SOC = 1.32, 95% CI 1.12\u20131.54).\nIn analyses restricted to the participants initiating ART, the median (IQR) time from diagnosis to ART initiation in the CIS and SOC groups was 32 (12\u2013135), and 63 (33\u2013230) days, respectively, while the median (IQR) time from enrollment in HIV care to ART initiation was 32 (11\u2013127), and 50 (15\u2013205) days, respectively.\nMedian time from ART eligibility to ART initiation for the CIS, CIS+, and SOC groups was 21 (9\u201340), and 25 (11\u201356) days, respectively.\nThere was no additional benefit of adding FIs to the CIS, with 55% (RRCIS+ vs CIS = 0.96, 95% CI 0.81\u20131.16; aRRCIS+ vs CIS = 0.94, 95% CI 0.76\u20131.18) of those in the CIS+ group achieving the primary outcome; 95% (RRCIS+ vs CIS = 1.00, 95% CI 0.83\u20131.13) linking to HIV care within 1 month of diagnosis, regardless of retention at 12 months; and 55% (RRCIS+ vs CIS = 0.95, 95% CI 0.79\u20131.13) being retained in care 12 months after diagnosis, regardless of the timing of linkage to care.\nIntervention effect on linkage to and retention in care at any health facility\nAnalyses supplementing data from electronic medical records from participating facilities with data from patient interviews and other health facilities in the study regions to examine linkage to and retention at any health facility showed similar effects of the intervention package.\nA total of 74% (RRCIS vs SOC = 1.47, 95% CI 1.08\u20132.01) of participants in the CIS group and 47% in the SOC group were found to have linked to HIV care at any health facility within 1 month of diagnosis and were retained in HIV care 12 months after diagnosis (Table 2).\nAdjustment for patient-level differences did not result in any change in this finding (aRRCIS vs SOC = 1.46, 95% CI 1.05\u20132.04).\nInclusion of FIs in the CIS also showed no additional benefit for linkage to and retention at any health facility, with 73% (RRCIS+ vs CIS = 0.98, 95% CI 0.85\u20131.15; aRRCIS+ vs CIS = 0.96, 95% CI 0.83\u20131.11) of those in the CIS+ group known to have linked to and been retained in HIV care at any health facility compared to the CIS group.\nIntervention effect on ART eligibility and initiation, disease progression, and death\nData from electronic medical records at study sites indicated that compared to patients in the SOC group, patients in the CIS group were more likely to ever have their ART eligibility assessed (100% versus 76.9%, RRCIS vs SOC = 1.29, 95% CI 1.08\u20131.54), be identified as ART eligible (75% versus 60%, RRCIS vs SOC = 1.24, 95% CI 1.07\u20131.43), and initiate ART (65% versus 54%, RRCIS vs SOC = 1.20, 95% CI 1.00\u20131.43) (Table 3).\nVery few participants were diagnosed with a new WHO stage 3/4 event at the diagnosing facility or self-reported a hospitalization in the 12 months after HIV diagnosis.\nThose in the CIS group had a non-significantly but modestly decreased risk compared to those in the SOC group (1% versus 3%, RRCIS vs SOC = 0.38, 95% CI 0.07\u20132.03), while similar results were observed between the CIS and CIS+ groups (1% versus 1%, RRCIS+ vs CIS = 0.65, 95% CI 0.12\u20133.64).\nNeither the CIS nor the CIS+ interventions had a significant effect on mortality within 12 months of diagnosis, with 6%, 5%, and 7% of participants in the CIS, CIS+, and SOC groups, respectively, known to have died during study follow-up (RRCIS vs SOC = 0.87, 95% CI 0.40\u20131.91; RRCIS+ vs CIS = 0.88, 95% CI 0.45\u20131.74).\nThe CIS also did not have a significant impact on mortality before (3%, RRCIS vs SOC = 0.78, 95% CI 0.46\u20131.32) or after ART initiation (3%, RRCIS vs SOC = 0.96, 95% CI 0.26\u20133.48); participants in the CIS+ group were less likely to die, though non-significantly so, before initiating ART compared to those in the CIS group (1% versus 3%, RRCIS+ vs CIS = 0.34, 95% CI 0.09\u20131.29).\nDiscussion\nWe conducted a cluster-randomized study in Mozambique to examine the effectiveness of a multi-component approach to increase linkage to and retention in HIV care\u20142 critical elements of the HIV care continuum\u2014among adults newly diagnosed with HIV.\nThe operational model of the CIS that we evaluated addresses known structural, biomedical, and behavioral barriers across the HIV care continuum and was composed of evidence-based, practical, and scalable interventions, including CD4 testing in VCT clinics with immediate turnaround of results, accelerated ART initiation for eligible individuals, and SMS health messages and appointment reminders.\nAn enhanced version of the CIS additionally included FIs.\nIn the spirit of implementation science, 2 of the interventions were implemented by existing health facility staff, rather than study staff, providing information on the real-world successes and challenges associated with the CIS that can be extrapolated to a range of settings with similar implementation contexts.\nOur study showed that participants receiving the CIS were 1.58 times more likely to link to HIV care at their diagnosing facility within 1 month of diagnosis and be retained in care at that same facility 12 months following diagnosis, representing not only a statistically significant but also a programmatically meaningful improvement.\nParticularly impressive gains were observed in timely linkage to care at the diagnosing facility: 89% of CIS participants linked to care on the day of diagnosis, representing a greater than 5-fold improvement compared to the SOC, and nearly universal linkage (96%) was achieved within 1 month of diagnosis.\nNotably, the intervention effect was greatest in subpopulations documented to have particularly poor outcomes across the HIV care continuum, including young adults and those with high stigma perceptions.\nThe intervention also had beneficial effects on other important milestones in the HIV care continuum in the 12 months following diagnosis, including the likelihood of patients having their ART eligibility assessed and initiating ART.\nWhile the intervention significantly increased retention in HIV care at both 6 and 12 months following diagnosis, retention in the CIS group remained concerningly low and far short of what is needed to end the HIV epidemic in Mozambique and other high-burden countries.\nWe found no additional gain in effectiveness from adding FIs to the CIS.\nPrior studies examining the effect of FIs in enhancing outcomes across the HIV care continuum among people living with HIV have shown inconsistent results.\nStudies from India, Uganda, and Democratic Republic of the Congo reported reductions in time to ART initiation and improvements in retention with the provision of incentives, while in the United States, randomized trials did not show any effect of FIs on linkage to care or viral load suppression.\nWhile 89% of participants in the current study reported that the type of FI provided and the amount of the FIs (i.e., mobile phone air-time vouchers worth approximately US$5 at 3 points in time) were adequate, it is possible that the FIs were not sufficiently optimized to affect behaviors.\nIndeed, as reported elsewhere, patient reactions to the FIs were surprisingly tepid, with only 21% reporting it to be the \u201cmost useful\u201d intervention for retention in care 12 months following diagnosis.\nAdditionally, fidelity to the FI component of the intervention package was imperfect, with, for example, 86% of participants eligible to receive the first incentive actually receiving it, which may have further limited the effect of this intervention.\nHowever, given the benefits of FIs in other health sectors, further research is needed to understand whether and how they may be optimized to enhance outcomes across the HIV care continuum.\nThis study has several important strengths.\nIt is among the first studies to evaluate the impact of a multi-component approach on 2 important HIV care and treatment indicators: timely linkage to care following an HIV diagnosis and sustained retention in care.\nImproving performance for these 2 elements of the HIV care continuum is critical for realizing the individual and population benefits of HIV programming in sub-Saharan Africa.\nFurther, while studies have examined the effectiveness of multi-component intervention packages that include FIs on HIV care outcomes, this study is the first to our knowledge to use a design that permits estimation of the additional benefit of including FIs as part of such a package.\nOur study also had limitations.\nFirst, in alignment with recent recommendations for implementation science studies, we used existing electronic medical records in the HIV clinics at the study sites to ascertain outcomes at the diagnosing facility, but such records may have limited data quality.\nHowever, data quality assessments were conducted regularly during the study period and ensured at least 85% concurrence between paper-based and electronic medical records on key data elements.\nSecond, aside from the FI, we cannot unpack the effect of individual intervention components.\nThird, the relevance of point-of-care CD4 count testing may change as countries adopt \u201ctreatment for all\u201d strategies, although our results suggest that providing people living with HIV with additional information on their health status immediately following diagnosis may be important in facilitating same-day linkage to care and likely same-day ART initiation.\nFourth, the CIS+ cohort was enrolled once the target sample size had been reached in the CIS cohort, thus introducing the potential for secular trends to have biased the comparison of the CIS and CIS+ packages.\nHowever, because we found no difference in the primary outcome between the CIS+ and CIS groups, secular trends would have had to have operated in the direction of reducing overall linkage and retention for this bias to result in the failure to observe an additional benefit of FIs for linkage and retention.\nWhile this is plausible, we do not have any evidence that a substantial reduction in overall linkage and retention occurred over the relatively limited time frame of the study.\nFinally, while the study was implemented in 2 contrasting settings within Mozambique, study facilities were located primarily in urban and semi-urban areas within the city of Maputo and Inhambane Province, which may limit generalizability.\nIndeed, settings with lower education and cell phone coverage than those included in our study may experience greater challenges implementing the SMS health messages and appointment reminders.\nSimilarly, while we set broad inclusion criteria, we did exclude people who did not understand Portuguese or Xitsua, were planning on leaving the community, or were not willing to receive services at the diagnosing facility, all factors that may have reduced generalizability.\nFinally, due to slower-than-expected enrollment, we enrolled fewer participants in the CIS+ group than intended, which decreased our power to detect statistically significant differences in study outcomes between the CIS+ and CIS groups.\nHowever, as the proportion achieving the combined outcome in the 2 groups was extremely similar (CIS 57% versus CIS+ 55%), it is unlikely that the inability to detect significant differences was primarily due to lack of power.\nConclusion\nMulti-component intervention strategies have been proposed to address troubling gaps in the HIV care continuum.\nTo our knowledge, this is amongst the first studies to rigorously evaluate such an approach.\nThe CIS we examined, comprising 3 evidence-based, practical, and scalable interventions, holds great promise as an approach to make much needed gains in the HIV care continuum in sub-Saharan Africa, particularly in the critical first step of timely linkage to care following diagnosis.\nFlow chart for study participation.CIS, combination intervention strategy; SOC, standard of care; VCT, voluntary counseling and testing.\nRelative risk of the CIS compared to the SOC on the primary outcome at the diagnosing health facility by patient characteristics.a Fifteen patients with missing information were excluded from this estimate. A description of the variables examined and categories used are provided in the Methods section.\n\nParticipant characteristics at study enrollment in the 3 study groups (N = 2,004).\nCharacteristic | TotalN = 2,004 | CISN = 744 | CIS+N = 493 | SOCN = 767 | p-Value\nRegion |  |  |  |  | \nMaputo | 1,077 (54%) | 396 (53%) | 275 (56%) | 406 (53%) | 0.58\nInhambane | 927 (46%) | 348 (47%) | 218 (44%) | 361 (47%) | \nSex |  |  |  |  | 0.50\nFemale | 1,292 (64%) | 490 (66%) | 319 (65%) | 483 (63%) | \nMale | 712 (36%) | 254 (34%) | 174 (35%) | 284 (37%) | \nAge (years) | 34.2 (9.6) | 34.9 (9.8) | 33.8 (9.9) | 33.8 (9.3) | 0.045\n18\u201324 | 265 (13%) | 90 (12%) | 70 (14%) | 105 (14%) | 0.12\n25\u201339 | 1,233 (62%) | 440 (59%) | 301 (61%) | 492 (64%) | \n40\u201349 | 348 (17%) | 148 (2%) | 87 (18%) | 113 (15%) | \n50+ | 158 (8%) | 66 (9%) | 35 (7%) | 57 (7%) | \nMarital status |  |  |  |  | <0.001\nMarried/partner and living together | 1,068 (53%) | 376 (51%) | 255 (52%) | 437 (57%) | \nMarried/partner, but not living together | 222 (11%) | 101 (14%) | 86 (17%) | 35 (5%) | \nSingle | 713 (36%) | 266 (36%) | 152 (31%) | 295 (38%) | \nMissing/refused | 1 (0%) | 1 (0%) | 0 (0%) | 0 (0%) | \nEducation |  |  |  |  | 0.003\nNone | 164 (8%) | 59 (8%) | 33 (7%) | 72 (9%) | \nPrimary | 1,149 (57%) | 442 (59%) | 256 (52%) | 451 (59%) | \nSecondary | 471 (24%) | 164 (22%) | 130 (26%) | 177 (23%) | \nAbove secondary | 219 (11%) | 78 (1%) | 74 (15%) | 67 (9%) | \nMissing/refused | 1 (0%) | 1 (0%) | 0 (0%) | 0 (9%) | \nEmployment |  |  |  |  | 0.46\nEmployed | 1,473 (74%) | 537 (72%) | 361 (73%) | 575 (75%) | \nUnemployed | 531 (26%) | 207 (28%) | 132 (27%) | 192 (25%) | \nMonthly income |  |  |  |  | <0.001\n\u22641,500 meticais | 871 (43%) | 342 (46%) | 165 (33%) | 364 (47%) | \n>1,500 meticais | 936 (47%) | 343 (46%) | 271 (55%) | 322 (42%) | \nMissing/refused | 197 (1%) | 59 (8%) | 57 (12%) | 81 (11%) | \nAnother household member has HIV |  |  |  |  | 0.28\nYes | 550 (27%) | 187 (25%) | 144 (29%) | 219 (29%) | \nNo | 913 (46%) | 361 (49%) | 219 (44%) | 333 (43%) | \nDon\u2019t know | 539 (27%) | 196 (26%) | 130 (26%) | 213 (28%) | \nMissing/refused | 2 (0%) | 0 (0%) | 0 (0%) | 2 (0%) | \n\nData given as N (percent).\nCIS, combination intervention strategy; SOC, standard of care.\n\nLinkage to and retention in HIV care: CIS versus SOC and CIS+ versus CIS.\nCategory | Outcome | CISN = 744 | CIS+N = 493 | SOCN = 767 | RR1 (95% CI), p-Value | aRR2 (95% CI), p-Value\nN | Percent | N | Percent | N | Percent | CIS versus SOC | CIS+ versus CIS | CIS versus SOC | CIS+ versus CIS\nPrimary outcome |  |  |  |  |  |  |  |  |  |  | \nAt diagnosing facility | Linked to care within 1 month of diagnosis and retained 12 months after diagnosis | 425 | 57% | 273 | 55% | 268 | 35% | 1.58 (1.05\u20132.39)p = 0.03 | 0.96 (0.81\u20131.16)p = 0.66 | 1.55 (1.07\u20132.25)p = 0.04 | 0.94 (0.76\u20131.18)p = 0.52\nAt any health facility | Linked to care within 1 month of diagnosis and retained 12 months after diagnosis | 547 | 74% | 360 | 73% | 363 | 47% | 1.47 (1.08\u20132.01)p = 0.02 | 0.98 (0.85\u20131.15)p = 0.91 | 1.46 (1.05\u20132.04)p = 0.03 | 0.96 (0.83\u20131.11)p = 0.52\nSecondary outcomes |  |  |  |  |  |  |  |  |  |  | \nLinkage at diagnosing facility | Same day as HIV test | 659 | 89% | 457 | 93% | 120 | 16% | 9.13 (1.65\u201350.40)p = 0.02 | 1.04 (0.92\u20131.20)p = 0.38 | N/A | \n | Within 1 week of HIV test | 678 | 91% | 461 | 94% | 349 | 46% | 2.43 (0.70\u20138.41)p = 0.14 | 1.03 (0.91\u20131.16)p = 0.59 | N/A | \n | Within 1 month of HIV test | 703 | 94% | 467 | 95% | 482 | 63% | 1.48 (0.93\u20132.35)p = 0.09 | 1.00 (0.89\u20131.13)p = 0.96 | N/A | \n | Within 12 months of HIV test | 716 | 96% | 467 | 95% | 592 | 77% | 1.23 (1.03\u20131.48)p = 0.03 | 0.98 (0.87\u20131.11)p = 0,74 | N/A | \nRetention at diagnosing facility | 6 months after diagnosis | 462 | 62% | 322 | 65% | 405 | 53% | 1.18 (1.00\u20131.39)p = 0.06 | 1.05 (0.88\u20131.26)p = 0.48 | N/A | \n | 12 months after diagnosis | 435 | 58% | 273 | 55% | 341 | 44% | 1.32 (1.12\u20131.54)p = 0.004 | 0.95 (0.79\u20131.13)p = 0.45 | N/A | \n\n1RR accounts for clustering within sites using random-intercept log-Poisson regression with empirical standard error estimates.\n2aRR adjusts for patient-level differences using propensity scores.\naRR, adjusted relative risk; CIS, combination intervention strategy; N/A, not applicable; RR, relative risk; SOC, standard of care.\n\nART determination and initiation, disease progression, and death: CIS versus SOC and CIS+ versus CIS.\n\u00a0Outcome | CIS(N = 744) | CIS+(N = 493) | SOC(N = 767) | RR1 (95% CI), p-value\nN | Percent | N | Percent | N | Percent | CIS versus SOC1 | CIS+ versus CIS1\nART eligibility assessed | 744 | 100% | 493 | 100% | 590 | 77% | 1.29 (1.08\u20131.54)p = 0.01 | 1.00 (0.89\u20131.12)p = 1.00\nIdentified as ART eligible | 557 | 75% | 372 | 75% | 464 | 60% | 1.24 (1.07\u20131.43)p = 0.01 | 1.01 (0.85\u20131.19)p = 0.91\nInitiated ART | 484 | 65% | 332 | 67% | 416 | 54% | 1.20 (1.00\u20131.43)p = 0.05 | 1.03 (0.88\u20131.22)p = 0.59\nNew WHO stage 3/4 or hospitalization | 7 | 1% | 3 | 1% | 23 | 3% | 0.38 (0.07\u20132.03)p = 0.22 | 0.65 (0.12\u20133.64)p = 0.53\nDeath within 12 months | 46 | 6% | 27 | 5% | 54 | 7% | 0.87 (0.40\u20131.91)p = 0.69 | 0.88 (0.45\u20131.74)p = 0.63\nDeath before ART initiation | 22 | 3% | 5 | 1% | 29 | 4% | 0.78 (0.46\u20131.32)p = 0.31 | 0.34 (0.09\u20131.29)p = 0.09\nDeath after ART initiation | 24 | 3% | 22 | 4% | 25 | 3% | 0.96 (0.26\u20133.48)p = 0.94 | 1.38 (0.62\u20133.07)p = 0.33\n\n1RR accounts for clustering within sites using random-intercept log-Poisson regression with empirical standard error estimates.\nART, antiretroviral therapy; CIS, combination intervention strategy; RR, relative risk; SOC, standard of care.",
        "label": "low",
        "id": "task4_RLD_test_80"
    },
    {
        "paper_doi": "10.5021/ad.2013.25.1.17",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Trial designMulticentre, randomised, double-blind, half-sided trialTrial registration numberNot reportedSettingMulticentre; it does not specifically mention location however all authors are from South Korea and the trial was approved by Institutional Review Boards at 5 South Korean hospitals.Date trial conductedNot reportedDuration of trial participation15 daysAdditional design detailsNoneInclusion criteriaPatients with moderate-severe symmetrical eczematous skin lesionsExclusion criteriaPatients < 4 years oldPatients currently undergoing treatment with systemic glucocorticoids, antibiotics or immunosuppressive agentsPatients treated with UV radiationPatients with other chronic non-eczematous skin diseases, also those with infectious dermatosesPatients with a chronic medical illness such as diabetesPatients who were pregnant or lactatingPatients with skin lesions involving the face or genital areaPatients with other severe dermatoses or scarsNotesPrior to the start of the trial, participants taking a systemic corticosteroid had a washout period of 4 weeks, and participants who applied a TCS had a washout period of 1 week.\n\n\nParticipants: Total number randomised175 participants (350 sides of the body)AgeFor the 159 participants who were analysed the mean age was 32.32 years (SD 19.86, range 5-79).SexOf 159 participants who contributed data there were 76 male (47.80%) and 83 female (52.20).Race/ethnicityNot reportedDuration of eczemaPaper states that 25 (15.72%) participants had \"past skin disease history\", whilst 134 (84.28%) did not.Severity of eczemaBaseline IGA of clinical response was 7.46 +- 3.11 in the mometasone furoate group, whilst 7.51 +- 3.18 in the methylprednisolone aceponate group. The index was calculated from assessment of 4 signs/symptoms: erythema, vesiculation, pruritus, and burning/pain where the physician rated each parameter on a 0-3 scale: 0 = no symptoms, 1 = mild, 2 = moderate and 3 = severe. The paper does not describe how this was calculated; possibly it was summed for each participant and a mean calculated.Filaggrin mutation statusNot reportedNumber of withdrawals15 participants were excluded due either to violation of protocols or adverse reactions, and 1 participant was excluded due to a screening criteria violation. It is unclear which group they were allocated to.NotesNone\n\n\nInterventions: Run-in detailsNAGroupsMometasone furoate cream; applied in a multi-lamellar emulsion cream to 1 side of the body once daily for 2 weeks. Concurrent treatment: noneMethylprednisolone aceponate cream; applied to 1 side of the body once daily for 2 weeks. Concurrent treatment: noneAdherenceNot reportedCo-interventionsThe mometasone furoate preparation also contained multi-lamellar emulsions (paper suggests this can aid restoration of the barrier function of the skin) and hexylene glycol (an antimicrobial excipient).NotesThe concentrations of the 2 TCS preparations were not stated in the paper. Therefore, the methylprednisolone preparation was assumed to be 0.1% and mometasone assumed to be 0.1%. This was  because these were the only concentrations identified in the reference sources that we used to identify potency.\n\n\nOutcomes: Adverse events (included those deemed treatment-related or not) at up to day 15*TEWL: measured by Tewameter(r) to evaluate epidermal permeability barrier function (Courage & Khazaka, Cologne, Germany). The TEWL improvement ratio was calculated as: TEWL improvement ratio (%) = [(TEWLday1-TEWLdayn)/ TEWLday1] x 100 (%) at day 1, 4, 8, and 15.IGA. The IGA index was calculated from assessment of 4 signs/symptoms: erythema, vesiculation, pruritus, and burning/pain where the physician rated each parameter on a 0-3 scale: 0 = no symptoms, 1 = mild, 2 = moderate and 3 = severe. The IGA improvement ratio was calculated as: IGA improvement ratio (%) = [(IGAday1-IGAdayn)/IGAday1] x 100 (%) at day 1, 4, 8, and 15.*VAS for pruritus: improvement of pruritus after treatment was scored subjectively, \"using 10 visual analogue scales that the patients scored\". We assume this means participants marked the severity of the itch on a 10 mm VAS. The VAS improvement ratio was calculated as: VAS improvement ratio (%) = [(VASday1-VASdayn)/VASday1] x 100 (%) at day 1, 4, 8, and 15.**denotes relevance to this review\n\n\nFunding source: None stated\n\n\nDeclarations of interest: None declared\n\n\nNotes: The mometasone furoate preparation also contained multi-lamellar emulsions (the paper suggests this can aid restoration of the barrier function of the skin) and hexylene glycol (an antimicrobial excipient)\n\n",
        "objective": "To establish the effectiveness and safety of different ways of using topical corticosteroids for treating eczema.",
        "full_paper": "Background\nTopical application of corticosteroids also has an influence on skin barrier impairment.\nPhysiological lipid mixtures, such as multi-lamellar emulsion (MLE) containing a natural lipid component leads to effective recovery of the barrier function.\nObjective\nThe purpose of this study was to conduct an evaluation of the therapeutic efficacy and skin barrier protection of topical mometasone furoate in MLE.\nMethods\nA multi-center randomized, double-blind, controlled study was performed to assess the efficacy and safety of mometasone furoate cream in MLE for Korean patients with eczema.\nThe study group included 175 patients with eczema, who applied either mometasone furoate in MLE cream or methylprednisolone aceponate cream for 2 weeks.\nTreatment efficacy was evaluated using the physician's global assessment of clinical response (PGA), trans-epidermal water loss (TEWL), and visual analogue scale (VAS) for pruritus.\nPatients were evaluated using these indices at days 4, 8, and 15.\nResults\nComparison of PGA score, TEWL, and VAS score at baseline with those at days 4, 8, and 15 of treatment showed a significant improvement in both groups.\nPatients who applied mometasone furoate in MLE (74.8%) showed better results (p<0.05) than those who applied methylprednisolone aceponate (47.8%).\nThe TEWL improvement ratio was higher in the mometasone furoate in MLE group than that in the methylprednisolone aceponate group, and VAS improvement was also better in the mometasone furoate in MLE group.\nConclusion\nMometasone furoate in MLE has a better therapeutic efficacy as well as less skin barrier impairment than methylprednisolone aceponate.\nINTRODUCTION\nTopical steroids are an important and common treatment modality for eczematous skin disorders.\nThe choice of topical steroid varies based on skin disease severity, potency, and delivery vehicle used.\nAppropriate use of a topical steroid makes it possible to treat an eczematous disorder effectively.\nHowever, they are many notable side effects such as skin atrophy, acneiform eruptions, hypertrichosis, hypopigmentation, and development of a cutaneous infection.\nSuch topical glucocorticoid side effects make many patients hesitant to use topical steroids.\nEfforts to minimize the side effects of topical steroids have been attempted.\nOne approach is to develop novel steroids such as non-halogenated double-ester-type glucocorticoids.\nAnother technique is to use topical steroid together with an anti-atrophogenic substance.\nIn particular, topical materials that not only minimize side effects of topical steroid but enhance the physiological lipid mixture have been investigated.\nAnother important aspect of eczema treatment is normalizing the defective skin barrier, because impaired skin barrier function is an important factor in the pathogenesis of eczema.\nSkin barrier function is affected by multiple factors, including downregulation of filaggrin and locogrin, reduced ceramide levels, increased proteolytic enzyme levels, and enhanced trans-epidermal water loss (TEWL).\nThe epidermal barrier is composed of a combination of corneocytes and intercellular lipids.\nThe stratum corneum (SC) provides a mechanical protection to the skin.\nIt functions as a barrier to water loss and permeation of soluble substances from the environment.\nRegulation of permeability, desquamation, antimicrobial peptide activity, toxin exclusion, and selective chemical absorption are all primary functions of the extracellular lipid matrix.\nIn contrast, mechanical reinforcement, hydration, cytokine-mediated initiation of inflammation, and protection from ultraviolet (UV) damage are all provided by corneocytes.\nAmong various factors, hydrophobic intercellular lipids are the most important factor for skin barrier function, and diminished ceramides create a leaky barrier.\nParticularly in the case of atopic dermatitis (AD), abnormal barrier function results from disruption of the multi-lamellar structure due to a significant reduction in the amount of intercellular lipid ceramides in the SC.\nA disrupted skin barrier can be replenished by a physiological lipid mixture.\nMulti-lamellar emulsions (MLEs) containing pseudoceramide has a multi-lamellar structure similar to intercellular lipids of the SC.\nLee et al. reported that patients with AD treated only with a MLE improved more than those applying a commercial moisturizing cream in an objective assessment and subjective satisfaction scores for symptoms and signs.\nAhn et al. reported on a co-application of MLE for topical steroid-protected skin barrier function, reinforcing the skin barrier permeability.\nTopical application of ceramides results in restoration of barrier function by reducing TEWL.\nEffective enhancement of skin barrier function with a physiological lipid mixture has been reported recently.\nChamlin et al. reported on alleviation of childhood AD using a ceramide-dominant, barrier-repair lipid, and attributed the improvement seen in their patients to a normalization of barrier function, which in turn dampened the cytokine cascade that initiates and sustains AD.\nMometasone furoate is a potent topical steroid with proven efficacy, similar to betametasone.\nMometasone furoate contains hexylene glycol, which has antimicrobial properties.\nThe effects of hexylene glycol on microorganisms may potentially be beneficial for treating eczematous disorders, with effects on microorganisms, possibly leading to better treatment and a longer relapse-free period.\nIt is also a safe and effective method for a long-term use to treat chronic, recurrent disease.\nThis study was performed to evaluate the clinical efficacy of this physiological lipid mixture as a vehicle of mometasone furoate in patients with eczema.\nWe designed a multi-center, randomized, double-blind controlled assessment to compare mometasone furoate in MLE with methylprednisolone aceponate.\nMATERIALS AND METHODS\nPatients\nPatients with eczema and moderate to severe manifestations were enrolled.\nAll patients showed eczematous skin lesions, presenting symmetrically.\nPatients <4 years old were excluded.\nThe following patients were also excluded from the study: patients currently undergoing treatment with systemic glucocorticoids, antibiotics or immunosuppressive agents; those treated with UV radiation; those with other chronic non-eczematous skin diseases; those with infectious dermatoses, chronic medical illness such as diabetes; those pregnant or lactating; those with skin lesions involving the face or genital area; and those with other severe dermatoses or scars.\nWe used a wash-out period for patients who had undergone treatment with topical and systemic corticosteroids.\nPatients taking a systemic corticosteroid had a wash-out period of 4 weeks.\nPatients who applied a topical corticosteroid had a wash-out period of 1 week.\nThis study was approved by the Institutional Review Boards at Incheon St. Mary's Hospital, Kangnam Sacred Heart Hospital, Seoul National University Hospital, Severance Hospital, and Kwandong University Myongji Hospital.\nStudy design\nAfter informed consent was obtained, the patients were randomly assigned to apply mometasone furorate in MLE on one side and apply methylprednisolone aceponate on the other side.\nThe topically applied formulations were mometasone furoate in MLE and methylprednisolone aceponate.\nTogether on each randomly assigned side, mometasone furoate in MLE was applied on one side for 2 weeks, and methylprednisolone aceponate was applied on the other side for 2 weeks.\nMometasone furoate in MLE or methylprednisolone aceponate was applied topically to skin lesions once daily by all subjects.\nThe study included a baseline visit before treatment, and on days 1, 4, 8, and 15 after treatment initiation.\nPatients were observed and assessed by one physician on days 1, 4, 8, and 15 throughout the trial.\nThe physician's global assessment of clinical response (PGA) score was adopted for an objective assessment of the clinical response to treatment.\nThe PGA index was calculated from scales of erythema, vesiculation, pruritus, and burning/pain.\nEach parameter was judged on a 0~3 scale: 0=no symptoms, 1=mild, 2=moderate, and 3=severe.\nClinical efficacy was assessed by the PGA improvement ratio.\nThe PGA improvement ratio was calculated as:\nPGA improvement ratio (%)=[(PGAday1-PGAdayn)/PGAday1]\u00d7100 (%)\nThe TEWL was measured by Tewameter\u00ae to evaluate epidermal permeability barrier function (Courage & Khazaka, Cologne, Germany) on days 1, 4, 8, and 15.\nClinical efficacy of the improved skin barrier function was assessed by the TEWL improvement ratio.\nThe TEWL improvement ratio was calculated as:\nTEWL improvement ratio (%)=[(TEWLday1-TEWLdayn)/TEWLday1]\u00d7100 (%)\nIn addition, improvement of pruritus after treatment was assessed subjectively, using 10 visual analog scales that the patients scored.\nThe visual analogue scale (VAS) improvement ratio was calculated as:\nVAS improvement ratio (%)=[(VASday1-VASdayn)/VASday1] \u00d7100 (%)\nAll adverse events were recorded, and whether they were treatment related or not was also noted.\nStatistical analysis\nMcnemar's t-test was used to compare the PGA improvement ratio on days 1, 4, 8, and 15 of treatment.\np<0.05 were regarded as statistically significant.\nA paired t-test and the Wilcoxon signed-rank test were used to verify significant differences in TEWL and VAS scores between groups A and B during the follow-up period.\nRESULTS\nSummary of patients\nA total of 175 patients were initially enrolled.\nFifteen patients were excluded due either to violation of protocols or adverse reactions, and one patient was excluded due to a screening criteria violation.\nIn total, 159 patients were analyzed (76 males and 83 females; age range, 5~79 years; mean age, 32.32\u00b119.86, mean\u00b1standard deviation years old).\nNo clinically significant differences were observed in the PGA score, TEWL, or VAS between the mometasone furoate in MLE group and the methylprednisolone aceponate group.\nBasal demographic characteristics of the study groups and basal results of the PGA, TEWL, and VAS scores are summarized in Table 1 and 2.\nClinical efficacy of mometasone furoate in MLE\nWe performed a data analysis on patients who completed 15 days of treatment to assess efficacy.\nComparison of the PGA score improvement ratio at baseline with days 4, 8, and 15 of treatment showed a significant increase for all follow-up periods.\nAt baseline, the mean PGA score was 7.46\u00b13.11 in the mometasone furoate in MLE group and 7.51\u00b13.18 in the methylprednisolone group.\nAfter 15 days, the PGA improvement ratio in the mometasone furoate in MLE group was 82.62\u00b121.62%, and that in the methylprednisolone acetonate group was 68.32\u00b124.05% (p\u22640.0001).\nThe PGA improvement ratio for days 4, 8, and 15 is summarized in Fig. 1.\nTEWL improvement in the mometasone furoate in MLE group\nAt baseline, the baseline TEWL score in the mometasone furoate in MLE group was 33.73\u00b122.47 g/h/m2 and 33.05\u00b121.96 g/h/m2 in the methylprednisolone acetonate group.\nNo significant difference was observed between the groups.\nAfter 15 days of treatment, the TEWL improvement ratio increased in both groups for all follow-up periods.\nAfter 15 days of treatment in the mometasone furoate in MLE group, the TEWL improvement ratio, which was 48.30\u00b168.04% at baseline, was statistically significant; and in the methylprednisolone acetonate group, the TEWL improvement ratio increased 32.74\u00b150.07% from baseline after 15 days.\nAlthough the TEWL improvement ratio increased in both groups, the TEWL improvement ratio in the mometasone furoate in MLE group was superior to that in the methylprednisolone acetonate group (p\u22640.0001).\nThe TEWL improvement ratio for days 4, 8, and 15 is summarized in Fig.\n2. The intergroup differences were statistically significant at every point.\nVAS improvement in the mometasone furoate in MLE group\nThe subjective VAS score was measured at every visit to evaluate improvement of pruritus.\nThe initial VAS score was 5.83\u00b12.31 in the mometasone furoate in MLE group and 5.99\u00b12.29 in the methylprednisolone aceponate group (p>0.05).\nThe VAS improvement ratio increased in both groups for all follow-up periods.\nAfter 15 days of treatment, the VAS improvement ratio score increased 83.28\u00b123.47% from baseline in the mometasone furoate in MLE group and 75.41\u00b127.24% in the methylprednisolone group.\nIn addition, pruritus showed a more significant improvement in the mometasone furoate in MLE group than that in the methylprednisolone group (p\u22640.0001).\nThe VAS improvement ratio at days 4 and 8 also showed an increase from baseline; however, no statistically significant intergroup difference was observed (p=0.2117, p=0.1131).\nThe VAS improvement ratio on days 4, 8, and 15 is summarized in Fig.\n3. The difference was statistically significant at day 15 (p\u22640.0001).\nAdverse effects\nAn itching sensation was observed in two patients (1.15%) who applied mometasone furoate in MLE and in 4 patients (2.30%) who applied methylprednisolone aceponate.\nUrticaria was observed in one patient (0.57%).\nHerpes virus infection was observed in one patient (0.57%) at a non-applied site.\nFive patients were excluded from the study due to pruritus and urticaria.\nDISCUSSION\nBoth mometasone furoate in MLE and methylprednisolone aceponate were effective treatments for eczema.\nStatistically significant differences in the PGA score were observed between the values at baseline and at day 15 of treatment.\nPatients who applied mometasone furoate in MLE showed better results in the PGA improvement ratio than those who applied methylprednisolone aceponate.\nBoth agents are potent group II corticosteroids, and they had shown similar efficacy in previous studies.\nHowever, mometasone furoate in MLE showed superior efficacy, likely due to the effects of the physiological lipid mixture on enhanced skin barrier function.\nThis was supported by the results showing TEWL improvement, as it mirrors the skin barrier function.\nThe TEWL improvement ratio was also higher in those who applied mometasone furoate in MLE.\nProlonged treatment with a topical steroid creates structural defects in the epidermis, which has been attributed to a disturbance in epidermal differentiation and thinning of the SC.\nA disrupted skin barrier due to abnormalities in intercellular lipid lamellae, which are thought to mediate transcutaneous water loss, results in an increase in TEWL score.\nThe VAS improvement ratio was higher with the physiological lipid mixture.\nApplying mometasone furoate in MLE resulted in significant improvement in clinical symptoms and signs of eczematous disorder.\nSubjectively, patients felt less of an itching sensation over the treatment period.\nTopical corticosteroids not only have antiproliferative effects but also suppress differentiation of the epidermal layer, resulting in defects in the epidermis.\nTherefore, long-term use of topical glucocorticoid causes weakening of the skin barrier.\nMany studies have been performed to minimize the side effects of glucocorticoids.\nOne attempt involves reinforcing the skin barrier function using a physiological lipid mixture.\nMan MQ et al. reported that topical application of a physiological lipid mixture results in accelerated recovery of barrier function, whereas an incomplete lipid mixture may inhibit the normal recovery response.\nAdditionally, topical application of an MLE containing a pseudoceramide results in significantly decreased TEWL and skin pH.\nAhn et al. reported that concurrent application of MLE with steroid significantly reduces skin atrophy induced by steroid.\nThe physiological lipid mixture using a vehicle allows for the penetration of topical steroid into the skin and enhances delivery.\nMometasone furorate in MLE showed a more beneficial effect than that of methylprednisolone aceponate, which has similar potency but not the physiological lipid mixture.\nAs such, the treatment period for a topical steroid would be shortened using the same potency.\nThe side effects of topical steroids are dependent on duration and frequency of application.\nIt is possible that the lower steroid potency causes similar effects to a higher potency steroid.\nWe suggest that a physiological lipid mixture as a topical steroid vehicle enhances the effects of a topical steroid and diminishes the side effects.\nUse of moisturizers is an important treatment method to improve skin hydration, even when overt disease is not observed.\nDaily treatment with topical moisturizers prevents exacerbation of skin lesions by reducing elevated TEWL.\nThe MLE-containing moisturizers are effective for hydration of eczema and have a safety profile.\nA physiological lipid mixture using pseudoceramides is presented in an orthorhombic lipid phase, which chiefly exists on the human SC lipid, while non-physiological lipid mixture moisturizers only show a liquid crystalline phase and a hexagonal phase.\nAn increase in TEWL results in skin dehydration, and xerosis cutis itself aggravates eczematous disease and makes pruritus more severe.\nPruritus, the most common and important symptom of an eczematous skin disorder, causes a vicious itching-scratching cycle.\nTherefore, rapid amelioration of pruritus shortens disease duration and enhances treatment compliance.\nIn recent studies, a physiological lipid mixture using pseudoceramides not only restored the skin barrier but also had an anti-inflammatory effect.\nKang et al. reported that topical application of MLE using pseudoceramides results in diminished mRNA expression of interleukin (IL)-4 and tumor necrosis factor-\u03b1 in murine atopic dermatitis-like skin lesions.\nIL-4 and IL-13 play a major role in allergic reactions.\nA suppressed immune response diminishes the cytokine cascade involved in the pathogenesis of pruritus.\nIn our study, only a few local side effects were observed such as pruritus and urticaria.\nThe intensity of the itching sensation was mild, and the duration was a few hours to a day.\nAs such, local side effects rarely required discontinuation of treatment.\nIn conclusion, mometasone furoate in MLE was a more effective treatment for moderate to severe skin eczema than that of methylprednisolone.\nAs mometasone furoate in MLE induced improvement in skin barrier function, and we recommend more studies using of physiological lipid mixtures as a vehicle for topical therapeutic agents.\nThe physician's global assessment of clinical response (PGA) improvement ratio at day 4, 8, 15. Comparison of the PGA improvement ratio between themometasone furoate in multi-lamella emulsion (MLE) group and the methylprednisolone group. A significant difference was observed in the PGA improvement ratio at all follow-up periods (p\u22640.0001).\nThe trans-epidermal water loss (TEWL) improvement ratio. Comparison of the TEWL improvement ratio between the mometasone furorate in multi-lamella emulsion (MLE) group and the methylprednisolone group. A significant difference was observed in the TEWL improvement ratio at all follow-up periods (p\u22640.0001).\nThe visual analog scale (VAS) improvement ratio. Comparison of the VAS improvement ratio between the mometasone furorate in multi-lamella emulsion (MLE) group and the methylprednisolone group. A significant difference was observed in the VAS improvement ratio at day 15 (p\u22640.0001).\n\nBase demographic characteristics of the study group\nValues are presented as mean\u00b1standard deviation or number (%).\n\nBaseline results of PGA, TEWL, and VAS score\nValues are presented as mean\u00b1standard devation. PGA: physician's global assessment of clinical response, TEWL: transepidermal water loss, VAS: visual analog scale, MLE: multilamellar emulsion.",
        "label": "unclear",
        "id": "task4_RLD_test_457"
    },
    {
        "paper_doi": "10.1007/bf01992163",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: randomized\nTime period/duration of trial: unclear, before 1993\nDuration of follow-up: 2 months\n\n\nParticipants: Setting: unclear\nLocation: unclear\nAge: 42 adult participants > 16 years (mean age: 28.2 years in ceftriaxone group, 26.8 years in ciprofloxacin group)\nGender: no details given\nHealth status of participants: not recorded\nInclusion criteria:blood culture positiveacute S typhi infectionExclusion criteria:inability to take oral medicationpossible or proven pregnancylack of fever at time of admission\n\n\nInterventions: Ceftriaxone: IV, 3 g once daily for 7 daysCiprofloxacin: oral, 500 mg twice daily for 7 days\n\n\nOutcomes: Clinical cure: clinical failure defined as fever > 38 degC after 7 days of therapy or who deteriorated clinically after 5 full days of therapy; cure if patient afebrile and asymptomatic on or before day 7 and did not require additional therapy during 2 months of follow-upRelapse: readmission for typhoid within 2 months of discharge with a positive blood or stool culture for S typhi with the same antibiogram as previousConvalescent faecal carriage: not defined, but stool culture positivity assessed at 28 days post-enrolment\n\n\nOrganism type and antimicrobial susceptibility: S typhi = 42 (MDR S typhi = 22), S paratyphi = 0\n\n\nNotes: The trial was terminated when the clinicians involved in the trial felt that it was no longer ethical to randomize patients to receive ceftriaxone, given the higher cost, need for IV access and perceived lower efficacy of this regime\n\n",
        "objective": "To evaluate the effectiveness of cephalosporins for treating enteric fever in children and adults compared to other antimicrobials.",
        "full_paper": "iprofloxacin versus Ceftriaxone in the Treatment of ultlresistant Typhoid Fever\niprofloxacin versus Ceftriaxone in the Treatment of ultlresistant Typhoid Fever\nA randomized trial comparing ceftriaxone (3 g given parenterally per day for 7 days) to ciprofloxacin (500 mg given orally twice a day for 7 days) in the treatment of blood culture positive typhoid fever was conducted.\nTwenty patients were openly randomized to receive ciprofloxacin and 22 to receive ceftriaxone.\nThe outcome was classified as clinical failure in 6 patients (27 %) in the ceftriaxone group, but in none in the ciprofloxacin group (p = 0.01).\nThe mean duration of fever was four days in the ciprofloxacin group and about five days in the ceftriaxone group (p = 0.04).\nIn the six patients in the ceftriaxone group who experienced failure, therapy was switched to ciprofloxacin and the patients became afebrile and asymptomatic within 48 hours.\nPatients with resistant strains of Salmonella typhi and patients with sensitive strains responded equally well to ciprofloxacin therapy.\nAnalysis of a subset of 12 of the multiresistant strains revealed that resistance was encoded for by a transferable 180 kilobase plasmid.\nCiprofloxacin represents a useful treatment option in areas where multiresistant strains are likely to be encountered.\nTyphoid fever is traditionally treated with a two-Week course of either chloramphenicol, cotrirnoxazole or ampicillin/amoxiciilin.\nDespite minor differences in toxicity, duration of fever, carriage and relapse rate, these agents are roughly equal in clinical efficacy (1).\nResistance to these agents has occurred sporadically over the past two de-Cades in a variety of locations (1,2), but beginning in 1989, Salmonella typhi strains resistant to all three standard antimicrobial agents have been reported with alarming frequency from locations as diverse as the Indian subcontinent, the Arabian (Persian) Gulf, the UK and China (1)(2)(3)(4).\nThese multiresistant strains are fully pathogenic, often Causing illness more severe than that due to sensitive strains (5).\nTreatment of typhoid caused by multiresistant Salmonella typhi strains is not standardized.\nBoth third-generation cephalosporins (particularly ceftriaxone) and the new fluoroquinolones have been used with some success (1,2), but no study has yet been published which makes a direct comparison of these two classes of antibiotics in the therapy of typhoid fever.\nIn response to the rapid dissemination of multiresistant Salmonella typhi in Bahrain (4), we initiated a trial comparing oral ciprofloxacin with parenteral ceftriaxone for the treatment of typhoid fever.\nWe also investigated the nature of antibiotic resistance in selected multiresistant Salmonella typhi recently introduced into the Arabian Gulf area.\nPatients and Methods\nPatients.\nResults\nPatients and Bacterial Strains: A total of 43 patients met the study entry criteria.\nOne patient was subsequently excluded when he proved to have active tuberculosis in addition to Salmonella typhi bacteremia and failed to become afebrile after more than two weeks of therapy with both ciprofloxacin and ceftriaxone.\nOf 42 evaluable patients, 20 were randomized to receive ciprofloxacin and 22 to receive ceftriaxone.\nThere were no significant differences when the treatment groups were compared by age, duration of fever prior to admission, prevalence of multiresistant strains, white blood cell count, serum sodium level, hematocrit, or proportion of patients with diarrhea, constipation, splenomegaly or occult blood in the stool (Table 1).\nAll Salmonella typhi isolates were sensitive to both ceftriaxone and ciprofloxacin when tested by the Kirby-Bauer technique.\nTriresistant strains were defined as those resistant to cotrimoxazole, ampicitlin and chloramphenicol, whereas sensitive strains were sensitive to all three antibiotics.\nNo isolate had an intermediate pattern of sensitivity.\nOutcome of Therapy.\nThere were no cases of clinical failure in the ciprofloxacin group, whereas there were six cases of failure in the ceftriaxone group (p = 0.01).\nAll ciprofloxacin patients were asymptomatic and afebrite by day 6 (mean: 4 days of fever).\nThe ceftriaxone group required a significantly longer time for resolution of fever (mean: about 5 days, p = 0.04).\nFive of the six cases of failure in the ceftriaxone group were thus classified because of persistent fever after a full seven days of therapy; the sixth patient had fever and persistent severe neuropsychiatric symptoms on day 6 of ceftriaxone therapy and was deemed a case of clinical failure by the investigators (Table 2).\nAll six cases of failure in the ceftriaxone group\nWere subsequently allocated to receive ciprofloxacin therapy.\nThese six patients were afebrile and asymptomatic within 48 hours, and all patients had an uneventful recovery while completing a one-week course of ciprofloxacin.\nBlood cultures were done on day 3 of initial therapy in all 42 study patients and on day 8 in the six patients receiving ciprofloxacin after failure of ceftriaxone; all cultures were negative.\nOne patient in the ceftriaxone group experienced relapse four weeks after therapy, both blood and Stool cultures being positive for a SalmoneUa typhi strain with the same antibiogram as the initial isolate.\nOne patient in the ciprofloxacin group was readmitted with fever eight weeks after discharge.\nThis patient's stool grew a sensitive Salmonella typhi; however, her prior isolate was triresistant and infection was attributed to reinfection rather than relapse.\nAll other patients had negative stool cultures four weeks after therapy and did not relapse within a two-month follow-up period.\nThe study was terminated when the clinicians in-Volved in the study felt that it was no longer ethical to randomize patients to receive ceftriaxone, given the higher cost, need for intravenous access and lower efficacy of this regimen.\nPhage Types and Drug Resistance.\nSeven of the 12 randomly selected strains belonged to Vi-phage type El, three to type M1, one to type A, and one to type 51.\nAll these strains were resistant to chloramphenicol, ampicillin and trimethoprim, but were sensitive to ceftriaxone, nalidixic acid and ciprofloxacin.\nThe complete spectrum of resistance was transferable at 28 \u00b0C but not at 37 \u00b0C.\nIn all cases, the spectrum of resistance was en-\nDiscussion\nIn this study, ciprofloxacin given orally produced more rapid and reliable resolution of fever than parenteral ceftriaxone.\nProlonged fever in ceftriaxone treated typhoid patients has been observed in other studies (10,11) and may reflect the relatively poor intracellular penetration of cephalosporins.\nCiprofloxacin, with its excellent intracellular penetration, has been almost uniformly successful in the treatment of typhoid caused by both sensitive and resistant Salmonella typhi isolates (1,12,13).\nThe short course (7 days) of ciprofloxacin used in our study was efficacious and not associated with a high rate of stool carriage or relapse, thus having significant advantages compared to the longer two-week course of traditional agents.\nBecause of these advantages, ciprofloxacin has recently been advocated in the UK as the drug of choice for treatment of typhoid in patients with a high pretreatment likelihood of infection with strains resistant to traditional agents (14).\nDespite its advantages, ciprofloxacin does have appreciable drawbacks.\nIts use in children and pregnancy is controversial due to concern about possible cartilage injury.\nWhile much less expensive than ceftriaxone, ciprofloxacin is still more expensive than oral drugs of choice used in the past.\nAlthough touted as a drug for treatment of typhoid carriers (15), its failure to reliably eliminate stool carriage in a recent outbreak of Salmonella java is also disquieting (16).\nMost disconcerting of all is a report from India of decreasing susceptibility of Salmonella typhi to ciprofloxacin and the need for higher doses (1.5 g/day) to achieve a cure (17).\nIn spite of these potential problems, on the basis of the findings of this study, oral ciprofloxacin (500 mg orally b.i.d.) can be recommended for the initial therapy of typhoid in areas where resistant strains are responsible for a sizeable proportion of cases of typhoid fever.\nIt is an effective oral drug which can cure typhoid in a one-week course of therapy.\nThe rapid spread of multiresistant typhoid fever over large geographic areas presents multiple challenges, especially in less developed countries where access to newer and more expensive antimicrobial agents may be limited.\nFurther research efforts must continue to focus on oral agents with good intracellular penetration which can be used for short courses of therapy with the chances of a high cure rate.\nTable 1 : Comparison of the two treatment groups at study entry.\nCipro- | Ceftri-\nfloxacin | axone\n(n = 20) | (n = 22)\n,...._.\nTable 2 : Outcome of therapy in the two treatment groups.\n | Cipro-Ceftriaxone P value\n | floxacin | (n = 22) | \n | (n=20) |  | \nClinical failure | 0/20 | 6/22 | 0.01\nRelapse | 0 | 1 | NS\nDays to resolution |  |  | \nof fever (mean) | 4.0 | 5.2 | 0.04\nNS: not significant. |  |  | \nAcknowledgements\nThe authors kindly thankDr. E.C, Oldfield Ili for his critical review.\nWe also thank Ms. C. Cesefia for her help in manuscript preparation and the entire staff of Salmaniya Medical Centre for their cooperation in this study.\nSpecial thanks to Dr. A. L. Bourgeois for his help with the bacterial isolates.\nThe Chief, Navy Bureau of Medicine and Surgery, Washington, DC, Clinical Investigation Program, sponsored this study No. 84-16-1968-366, Operation Desert Storm Scholarly Work, as required by HSETCINST 6000.41.\nThis work was also supported by the Naval Medical Research and Development Command, Naval Medical Command, National Capital Region, Bethesda, MD (work unit No. 3M162770AR122), the Ministry of Health of Bahrain, and Naval Medical Research Unit No. 3, Cairo, Egypt (work unit No. 3M162787A870AN121).\nThe opinions and assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the views of the United States Department of the Navy, the Department of Defense or the United States Government.",
        "label": "unclear",
        "id": "task4_RLD_test_204"
    },
    {
        "paper_doi": "10.7189/jogh.09.020402",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: DesigncRCTAllocation of clusters50 schools randomized to intervention, 50 to control\n\n\nParticipants: 9258 primary school-aged children\n\n\nInterventions: Broad multiple\n\n\nOutcomes: Any STH\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Background\nWater, sanitation, and hygiene (WASH) in schools is promoted by development agencies as a modality to improve school attendance by reducing illness.\nDespite biological plausibility, the few rigorous studies that have assessed the effect of WASH in schools (WinS) interventions on pupil health and school attendance have reported mixed impacts.\nWe evaluated the impact of the Laos Basic Education, Water, Sanitation and Hygiene Programme \u2013 a comprehensive WinS project implemented by UNICEF Lao People\u2019s Democratic Republic (Lao PDR) in 492 primary schools nationwide between 2013 and 2017 \u2013 on pupil education and health.\nMethods\nFrom 2014-2017, we conducted a cluster-randomized trial among 100 randomly selected primary schools lacking functional WASH facilities in Saravane Province, Lao PDR.\nSchools were randomly assigned to either the intervention (n\u2009=\u200950) or comparison (n\u2009=\u200950) arm.\nIntervention schools received a school water supply, sanitation facilities, handwashing facilities, drinking water filters, and behavior change education and promotion.\nComparison schools received the intervention after research activities ended.\nAt unannounced visits every six to eight weeks, enumerators recorded pupils\u2019 roll-call absence, enrollment, attrition, progression to the next grade, and reported illness (diarrhea, respiratory infection, conjunctivitis), and conducted structured observations to measure intervention fidelity and adherence.\nStool samples were collected annually prior to de-worming and analyzed for soil-transmitted helminth (STH) infection.\nIn addition to our primary intention-to-treat analysis, we conducted secondary analyses to quantify the role of intervention fidelity and adherence on project impacts.\nResults\nWe found no impact of the WinS intervention on any primary (pupil absence) or secondary (enrollment, dropout, grade progression, diarrhea, respiratory infection, conjunctivitis, STH infection) impacts.\nEven among schools with the highest levels of fidelity and adherence, impact of the intervention on absence and health was minimal.\nConclusions\nWhile WinS may create an important enabling environment, WinS interventions alone and as currently delivered may not be sufficient to independently impact pupil education and health.\nOur results are consistent with other recent evaluations of WinS projects showing limited or mixed effects of WinS.\nSchool-aged children in low-income settings are at substantial risk for water, sanitation, and hygiene (WASH)-related infections such as pathogens causing diarrheal diseases, soil-transmitted helminths (STH), and trachoma.\nCrowded, unsanitary conditions may facilitate the spread of pathogens, and increase pupils\u2019 risk for disease.\nImproved access to WASH facilities combined with sufficient behavior change may not only prevent the spread of pathogens within the school domain but also lead to beneficial WASH habits at home and throughout the life course.\nThe limited data available indicate that only 69% of schools worldwide have access to sanitation facilities, while only 66% have access to water.\nWASH in schools (WinS) targets and indicators have been included in the Sustainable Development Goals.\nDespite the biological plausibility of WinS interventions to reduce illness and subsequently school absence, evidence of impact has been mixed.\nSome WinS efficacy studies, such as those assessing intensive handwashing programs in China and Egypt, reported reductions in absence and absence due to illness.\nHowever, with only 6- and 3-month follow up periods, respectively, and with soap being continuously supplied by the intervention or school administration, respectively, the long-term sustainability of handwashing behaviors linked to these impacts is unknown.\nEffectiveness trials of WinS projects have not replicated this success.\nA matched-control evaluation of a comprehensive WinS program in Mali revealed reductions in pupil-reported diarrhea, symptoms of respiratory infection, and absence due to diarrhea, but higher odds of absence overall among pupils enrolled in beneficiary schools.\nHowever, there were imbalances between the beneficiary and comparison groups at baseline, and the study was further limited by inconsistent fidelity to the intervention by implementing partners and participating schools.\nA randomized controlled trial (RCT) of a WinS program in Kenya reported a 44% reduction in odds of Ascaris lumbricoides reinfection, but no overall impact on absence or diarrhea.\nProgram impact differed by intervention arm (as individual and combined WASH interventions were employed) and subsets of the sample population.\nAbsence among girls in the hygiene promotion and water treatment arm reduced by 58%.\nIn water-scarce schools that received a comprehensive WASH intervention, including water supply improvements, risk of diarrhea among pupils reduced by 61%, while diarrhea among pupils\u2019 siblings under 5 years old reduced by 56%.\nHowever, program impact may have been affected by incomplete and inconsistent intervention delivery (fidelity) and uptake and use by the target population (adherence).\nA WinS intervention in Lao People\u2019s Democratic Republic (Lao PDR), Cambodia, and Indonesia had no impact on STH infection or being underweight, but reported evidence of improvement in dental cavities.\nAgain, this evaluation was potentially limited by incomplete fidelity and adherence to the program, as well as a non-randomized design and contamination from concurrent programming in control schools.\nHere, we present results from the Water, Sanitation, and Hygiene for Health and Education in Laotian Primary Schools (WASH HELPS) study, a cluster-RCT designed to measure the impact of a comprehensive WinS project \u2013 water supply, sanitation, handwashing, and behavior change - in Lao PDR on pupil absence, diarrhea, respiratory infection, and STH infection.\nGiven past challenges in program fidelity and adherence to project outputs and behaviors, we also apply two analyses that have previously been used to evaluate the role of intervention fidelity and adherence on WinS project impacts.\nMETHODS\nStudy setting and intervention\nThe Laos Basic Education, Water, Sanitation and Hygiene Programme was implemented by UNICEF in 492 primary schools across thirteen provinces between 2013 and 2017.\nThe WASH HELPS Study, a research component of the intervention, was conducted between September 2014 and May 2017 in Saravane Province, which was selected because it was the only province in which intervention activities had not yet occurred, thus allowing a randomized intervention trial.\nThe study setting, baseline results, intervention components, intervention outputs and outcomes, and their fidelity and adherence have been described in detail elsewhere.\nKey outputs and outcomes of the project are listed in Table 1.\nBriefly, the comprehensive WinS project included provision of a school water supply, sanitation facilities, handwashing facilities (individual and group), drinking water filters, and behavior change education and promotion.\nThe project was implemented in two phases; lessons learned from Group 1 schools (n\u2009=\u200952; intervention started in 2014) were applied to improve the project for Group 2 schools (n\u2009=\u200948; intervention started in 2015), leading to different levels of achievement at output and outcome levels between groups, as well as different durations of follow-up.\nStudy design, sampling, and data collection\nWe conducted a cluster-randomized, controlled trial among 100 primary schools (50 intervention, 50 comparison).\nStudy design, sampling, and data collection methods have been previously published.\nWe used stratified random sampling to help ensure equal representation of control and intervention schools in each district, and that the number of schools selected in each district was proportional to the number of eligible schools in each district.\nWe selected up to 40 pupils from grades 3-5 in each school using systematic stratified sampling, with grade and sex as the stratification variables.\nPupils selected at baseline were followed throughout the entire study period; pupils who left the school due to abandonment or transfer were replaced at the beginning of the following academic year, maintaining equal grade and sex ratios when possible.\nPupils who progressed from fifth to the sixth grade were replaced with pupils from grade three the following academic year.\nA total of 3993 pupils were enrolled throughout the study period.\nData were collected over three or two school years (Group 1 and 2 schools, respectively) to measure uptake and sustainability of facilities and behavior change.\nTo account for variabilities across time and season, data were collected throughout the school year, which consists of 33 weeks across two semesters (September-January and February-May), with five to six hours of instruction per day.\nTrained enumerators visited study schools every six to eight weeks during the school year through March 2017, for a total of 11 (Group 1) or 7 (Group 2) visits per school.\nAll visits were unannounced and during school hours.\nAt each visit, enumerators conducted a roll call of all students enrolled in the school using sex- and grade-specific ledgers; interviewed the school directors; interviewed sampled pupils in grades 3\u20135; observed conditions and functionality of WinS hardware; and observed individual and group handwashing practices.\nEach year, stool samples were collected from up to 50 pupils per school prior to distribution of preventative chemotherapy as part of the National School Deworming Programme.\nStool samples were tested for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus) using the Kato Katz technique.\nMeasures\nOur primary impact of interest was pupil absence measured by school-wide roll-call at each visit.\nAt the beginning of each data collection visit, enumerators visited each classroom with a roster of all students enrolled in the school, stratified by grade and sex.\nAt each visit, enumerators confirmed with the head teacher whether there were any new students since the last visit or if any students had left the school.\nNew students were added to the roster.\nDropout was recorded for students who had dropped out since the last visit.\nAbsences that were followed by a designation of dropout or transfer were removed from roll-call analysis.\nSecondary educational impacts included enrollment, dropout, and progression.\nEnrollment was calculated at each visit by summing the count of pupils on the roll-call roster and subtracting those who had dropped out or transferred.\nIn addition to student-level dropout recorded in the roll-call register, an aggregate school-level count of dropout was reported by the school at the end of each school year.\nPupils who transferred to another school were not considered to have dropped out.\nProgression was school-reported at the end of each academic year as the count of students who passed the national exam and progressed to the next grade level.\nAll secondary educational impacts were stratified by grade and sex.\nSecondary health impacts included diarrhea, symptoms of respiratory infection, and conjunctivitis/non-vision related eye illness and were collected through pupil interviews.\nAll health impacts were binary and self-reported with a one week recall period.\nPupils were asked if they had had diarrhea using local terminology and were also asked how many times they had defecated each day; a pupil was considered to have had diarrhea if he or she had reported having diarrhea and had defecated three or more times in a 24-hour period.\nPupils were considered to have symptoms of respiratory infection if they reported cough, runny nose, stuffy nose, or sore throat.\nDuring the last visit we included negative-control questions about self-reported cuts/scrapes and toothache.\nThese questions served as a measure of respondent bias, as there is no biological plausibility of an association between a WinS program and cuts/scrapes or toothache.\nData on STH infection were collected yearly.\nAny sample testing positive for the hookworms, A. lumbricoides, or T. trichuria considered positive for STH infection.\nIntervention fidelity and adherence for this study has been described previously.\nTo measure fidelity- defined as how the intervention was delivered per the stated design- we created an index score in which one point was given for each of the 20 output criteria fulfilled (Table 1).\nFor each visit, the minimum intervention fidelity score was zero and the maximum score was 20.\nTo measure adherence \u2013 defined as achievement of behavioral outcomes promoted by the intervention \u2013 a similar index score was created.\nAlthough there were five behavioral outcomes of interest (Table 1), we excluded group compound cleaning from the index given that reported participation in group compound cleaning was nearly universal among both intervention and comparison schools at baseline (97.9%), therefore the adherence score ranged from 0-4.\nA behavior was considered to be achieved when >75% of pupils reported or were observed to complete the behavior except for group handwashing, which was binary (either the school performed group handwashing or did not).\nAnalysis\nData were analyzed using Stata Statistical Software: Release 13 (StataCorp LP, College Station, TX, USA) and SAS version 9.4 (SAS Institute, Cary, NC, USA).\nIntention to treat analysis\nOur primary analysis was an intention-to-treat (ITT) analysis, which was used on all primary and secondary impacts.\nFor binary impacts (roll-call absence, diarrhea, symptoms of respiratory infection, conjunctivitis/non-vision related eye illness, STH infection, toothache, cuts/scrapes), we estimated relative risk using a \u201cmodified Poisson\u201d approach.\nThis is a validated method to produce relative risk ratios for binary data using a multi-level mixed Poisson model with robust error variances, and was chosen for this analysis because Stata does not support the use of log-linear binomial regression when using mixed effects generalized linear models.\nOdds ratios were obtained when the modified Poisson model did not converge for a specific impact (eg, toothache).\nRandom intercepts at the school and pupil levels were included to account for clustering of pupils within schools and for repeated measures of pupils over time, respectively.\nFor count impacts (enrollment, abandonment, progression), we estimated relative risk using Poisson regression models.\nAs these data were aggregated at the school-level, we included a random intercept at the school level only.\nAll ITT models compared intervention schools to comparison schools as they were randomly allocated to intervention and comparison groups, without regard to project fidelity or adherence.\nIntervention and comparison schools were balanced on key indicators at baseline, therefore intervention schools were included in the analysis once UNICEF documented that full intervention implementation (eg, both hardware and behavior change components) was complete.\nSince full implementation generally occurred at the same time in each district, comparison schools were also included once implementation occurred in their respective districts.\nModels included several design variables, including the district and visit number, and controlled for the following fixed effects, determined a priori based on biological plausibility of affecting impacts: pupil sex, pupil grade, school enrollment size, season (rainy or dry).\nThe rice crop calendar (planting, growing, harvesting) was included as a fixed effect in the absence model because rice agriculture is the predominant economic activity in the province and the need to stay home and support the family was the leading cause of pupil-reported absence.\nFully adjusted models were used to produce adjusted risk ratios (aRR) for each of the associations of interest.\nThese fixed effects, as well as whether the school was concurrently receiving aid from the World Food Program (WFP) school feeding program, were also assed for effect modification.\nCovariates were determined to be effect modifiers if an interaction term between the covariate and intervention group was significant in the full model.\nIntervention fidelity and adherence are important considerations when evaluating the impact of WASH programs.\nAssessing these factors along the causal \u2018theory of change\u2019 allows us to understand not only if but why and how that intervention succeeded or not in that context (ie, was there theory failure?).\nFurther, assessment of the process can determine if the intervention followed the intervention protocol to activate that theory of change (ie, was there intervention failure?).\nIn contexts where fidelity and adherence to the intervention is imperfect, ITT results may underestimate the causal effect for the potential impact of changes to outputs or outcomes, resulting in null or mixed effects.\nGiven the suboptimal fidelity and adherence of the intervention based on our monitoring data, we conducted a secondary analysis to quantify the impact of the project as implemented by UNICEF and adhered to by schools and pupils on the primary impact (roll-call absence) and select secondary impacts (diarrhea, respiratory infection, and STH prevalence).\nWe explore two modeling frameworks that have been previously used to evaluate the role of fidelity and adherence to a school WASH intervention on project impacts: As-treated (AT) analysis and Structural Nested Models (SNMs).\nEach framework operates under different assumptions and differ in robustness and efficiency; as such, comparing estimates lends a more informed picture of project impact.\nWe conducted a sensitivity analysis to identify a meaningful threshold of fidelity and adherence.\nThe scale of 20 outputs (fidelity) were categorized with cut-points at each 10th percentile and the scale of four outcomes (adherence) were unadjusted.\nWe observed lower risk of absence among schools with 70%-80% intervention fidelity and higher, but there was no clear evidence of a threshold for any other association (Figure 1).\nWe thus selected a threshold of 75%, which is consistent with previous research on fidelity to WinS projects.\nOnly the SNM requires specifying a threshold of fidelity/adherence, however, we also applied the 75% threshold to the AT models for comparability between the two approaches.\nAs-treated analysis\nThe AT analysis groups subjects according to the treatment received and does not consider the treatment intended (as is the case with ITT analysis).\nAdvantages to the AT approach are that it is analytically straightforward and easily supports our clustered and longitudinal study design.\nDisadvantages are that characteristics of schools with good fidelity or students who adhere to behaviors may be fundamentally different from those with poor fidelity/adherence, which can lead to confounding.\nThis confounding may be remedied by controlling for the prognostic factors that led participants to choose to adhere, but only if those prognostic factors are known, which is often not the case.\nFor the AT analysis, we ran two separate models that were structurally identical to the ITT models.\nHowever, instead of using intervention status as the primary predictor, as in the ITT analysis, schools were grouped according to intervention fidelity (ie, fulfilling \u226575% of outputs or not) and adherence (ie, fulfilling \u226575% of outcomes or not), respectively.\nAT models included the same covariates as the ITT models, with a priori identified fixed effects and random intercepts at the school and pupil levels.\nOnly data collected after the implementor reported intervention delivery was complete were included.\nAT models were stratified by effect modifiers identified in the ITT analysis.\nStructural nested model analysis\nSecond, we assessed the role of fidelity and adherence using SNMs, an instrumental variable approach.\nSNMs resolve the potential confounding issue presented by AT models because they do not break the randomization of intervention status.\nInstead, SNMs create a counterfactual for each study participant in order to compare the risk of an impact among adherers against the risk of the impact had the same individual not adhered.\nUnlike the ITT and AT models, to control for relevant covariates, a weighted distribution of population data are produced in order to remove the association between population-level confounders and randomization.\nWhile SNMs are advantageous because they account for unknown or unmeasured confounders, drawbacks are that they are more computationally intensive and rely on strong assumptions.\nSNM assumptions are described in detail elsewhere; briefly, they are as follows: (1) Exclusion restriction \u2013 randomization has no direct effect on the outcome; (2) Consistency \u2013 observed outcomes are possible under the fidelity/adherence level actually observed; (3)\nThe potential outcomes used to estimate the SNM effects are independent of randomization; (4) No interaction \u2013 the model\u2019s causal effect is consistent across randomization groups.\nOur code was derived from Garn et al and adapted for a 2-arm trial.\nBecause the SNM methodology we used does not accommodate repeated measures, we averaged time-varying pupil-level data (eg, grade, absence, reported diarrhea, reported symptoms of respiratory infection) and school-level data (output index, behavior index) across the data collection period.\nAs such, binary variables such as absence, reported diarrhea, and reported symptoms of respiratory infection became a continuous variable between zero and one, in which zero indicated never being absent, reporting diarrhea, or reporting symptoms of respiratory infection, whereas one indicated always being absent, reporting diarrhea, or reporting symptoms of respiratory infection.\nSimilar to the ITT and AT models described above, observations were included only after full implementation had been achieved.\nModels were adjusted using the same covariate variables as we used in the ITT and AT models.\nAs with the AT models, achievement of \u226575% of outputs and \u226575% of outcomes were considered achieving fidelity and adherence, respectively.\nSNMs were stratified by effect modifiers identified in the ITT analysis.\nFor all analyses, results were considered statistically significant if the P-value was <0.05.\nEthics\nThe WASH HELPS Study was approved by Emory University\u2019s Institutional Review Board (IRB0076404) and the Lao Ministry of Health\u2019s National Institute of Public Health National Ethics Committee (No. 043 NIOPH/NECHR).\nBoth Institutional Review Boards approved consent in loco parentis (in the place of the parent) signed by the school director.\nPupils who were selected for the pupil interview and/or stool collection provided informed verbal assent prior to any data collection.\nAll consent/assent procedures occurred after randomization.\nThe intervention was delivered to comparison schools in April 2017, after research activities ended.\nThe study is registered in ClinicalTrials.gov (NCT02342860).\nRESULTS\nBaseline results and intervention fidelity and adherence\nA total of 100 schools (n\u2009=\u200950 intervention, n\u2009=\u200950 comparison) were randomized, received the intervention, and included in the analysis (Figure 2).\nThere were no significant differences in key pupil-level or school-level indicators between intervention and comparison groups at baseline, indicating that the cluster-randomization was successful in creating balanced groups.\nFollowing full intervention implementation, intervention fidelity was 30.9% across all schools and visits and intervention adherence was 29.4%.\nData on fidelity to specific project outputs and adherence to specific project behaviors across the evaluation period have been previously published.\nIntention-to-treat analysis\nWe found no impact of the intervention on the primary impacts (roll-call absence) or secondary impacts (enrollment, progression, pupil-reported diarrhea, pupil-reported symptoms of respiratory infection, pupil-reported conjunctivitis, STH infection; Table 2).\nThere was some evidence of effect modification.\nRisk of diarrhea was higher in the rainy season compared to the dry season; when stratified by season, there was no significant impact of the intervention on diarrhea in either season (Dry season adjusted risk ratio (aRR)\u2009=\u20090.69, 95% confidence interval (CI)\u2009=\u20090.44, 1.10; Rainy season aRR\u2009=\u20091.14, 95% CI\u2009=\u20090.65, 1.99).\nPupil sex, pupil grade, school enrollment size, receiving support from the WFP school feeding program, and the rice crop calendar (absence model only) did not modify the effect of any primary or secondary impacts.\nWe found no difference in reported prevalence of toothache or cuts/scrapes (the negative control questions) among pupils attending intervention vs comparison schools (toothache adjusted odds ratio (aOR\u2009=\u20090.64, 95% CI\u2009=\u20090.23, 1.84; cuts/scrapes aRR\u2009=\u20091.06, 95% CI\u2009=\u20090.66, 1.72), indicating that any respondent bias that may have been present occurred equally between groups.\nAs-treated analysis\nAT results are presented in Table 3.\nIntervention fidelity \u2013 meeting \u226575% of output indicators associated with water supply, toilets, handwashing facilities, promotion of group hygiene activities, group handwashing facilities, and filtered drinking water\u2013 was associated with roll call absence and prevalence of STH.\nCompared to students attending schools without intervention fidelity, students attending schools with intervention fidelity had a 23% lower risk of absence (aRR\u2009=\u20090.76, 95% CI\u2009=\u20090.64, 0.91) and a 20% higher risk of STH prevalence (aRR\u2009=\u20091.20, 95% CI\u2009=\u20091.01, 1.43).\nDiarrhea was significantly higher during the rainy season, but when stratified there was no significant difference by fidelity status (Dry season aRR\u2009=\u20090.84, 95% CI\u2009=\u20090.48, 1.49; Rainy season aRR\u2009=\u20091.65, 95% CI\u2009=\u20090.82, 3.33).\nIntervention adherence \u2013 meeting outcome indicators associated with toilet use, handwashing with soap after toilet use, daily group handwashing, and daily group toilet cleaning \u2013 was not significantly associated with any impacts.\nStructural nested model analysis\nResults from the SNMs are presented in Table 3.\nDiarrhea was the only impact associated with fidelity or adherence.\nWhen stratified by season, diarrhea was lower in the dry season among students attending schools with intervention fidelity (aRR\u2009=\u20090.45, 95% CI\u2009=\u20090.24, 0.85) and adherence (aRR\u2009=\u20090.42, 95% CI\u2009=\u20090.21, 0.87); there was no significant difference in diarrhea between groups during the rainy season.\nDISCUSSION\nIn the primary analysis, we found no evidence that the intervention had an effect on absence, school enrollment, dropout, grade progression, pupil-reported diarrhea, pupil-reported symptoms of respiratory infection, pupil-reported conjunctivitis, or prevalence of STH.\nThese results contribute to the growing body of research showing limited or mixed impacts of WinS effectiveness trials on pupil health and education.\nSince 2010, access to WASH has been a fundamental human right recognized by the United Nations General Assembly.\nAs such, regardless of its potential education and health impacts, WinS access is an important objective, evidenced by its inclusion in the Sustainable Development Goals.\nHowever, if improvements in education and health indicators are to be achieved, results from this and other rigorously evaluated WinS programs suggest that WinS interventions alone, and as currently delivered in many contexts, may be insufficient to achieve anticipated education and health impacts.\nThe theory of change for WinS programs posits that improved WASH access leads to reductions in pathogen exposure at the school level and the habitualization of hygiene behaviors that can be practiced both at school at and home, which in-turn leads to reduced illness and thus reduced school absence.\nNumerous factors influence school absence, such as household wealth, distance to school, and number of siblings.\nLao PDR is a least-developed country, with over 65% of the population working in agriculture.\nIn Saravane Province, where over half of the population lives in poverty, the school calendar largely coincides with rice planting and harvesting seasons, and children are often kept home from school to assist in the fields and with other household chores.\nIndeed, in the current study, the leading pupil-reported cause of school absence was the need to stay home to support the family in economic activities (9.4% of pupils in intervention group and 8.7% of pupils in comparison group across all visits), not illness (5.1% of pupils in intervention group and 5.8% of pupils in comparison group across all visits), which may explain the lack of an impact of the intervention on absence.\nThus, the role of school WASH in supporting an enabling environment may be critical, but ultimately not sufficient to reduce absence when other factors like household economic needs or food security is the main driver of truancy from school.\nComplementary approaches to WinS may be necessary to achieve improvements in absence and other educational impacts.\nFor example, WinS may be successful in combination with school feeding programs or conditional cash transfers, both of which have been associated with reduced absence and increased enrollment in other low- and middle-income contexts.\nAlthough our results did not reveal a significant interaction between the WFP school feeding program and absence or enrollment, our study was not designed or adequately powered to detect a difference.\nAlthough there are potential mechanisms by which improved WASH may impact illness independently of measurable impacts on absence, we found no overall impact of the WinS intervention on pupil illness.\nThese results contrast to previous WinS research that reported overall reductions in diarrhea, respiratory infection, and absence due to illness, but are consistent with results from a WinS intervention in Lao PDR, Cambodia, and Indonesia that found no impact of the intervention on STH or underweight.\nOne explanation for the lack of an effect of the WinS intervention on pupil illness is low household WASH access; in this study context, the health benefits linked to improvements in school WASH conditions and behaviors provided by this intervention were likely not sufficient to overcome other potential transmission pathways at home or elsewhere in the community.\nEnvironmental improvements in both the domestic and public domains may be required for successful control of infections targeted by environmental improvements, such as diarrhea.\nAs such, WinS alone may not achieve significant health gains without concurrent community and household WASH improvements.\nFidelity and adherence are fundamental antecedents to achieving intervention effects.\nIt is possible that the lack of an effect of the intervention could be due, in part, to sub-optimal or unsustained fidelity and adherence.\nHowever, our secondary analyses yielded limited evidence of an effect of the intervention, even at high levels of intervention fidelity and adherence.\nAdditionally, our sensitivity analysis showed no clear trend in impacts across the fidelity/adherence continuum.\nWith two exceptions \u2013 the association between fidelity and lower absence (AT analysis) and the association between fidelity and adherence and lower diarrhea during the dry season (SNM analysis) \u2013 we did not find that fidelity and adherence led to improved education or health.\nThese results support the above conclusion that factors other than WinS \u2013 such as low household WASH access or household economics \u2013 may supersede health and education benefits of a WinS intervention in low-income contexts.\nHowever, the AT evidence should be should be interpreted cautiously due to the limited potential for causal inference resulting from breaking the randomization assignment in the AT analysis.\nThe two fidelity and adherence analyses results were inconsistent and sometimes yielded estimates of effect in opposite directions (eg, associations between adherence and diarrhea, respiratory infection, and STH), which is likely due to unaccounted for confounding in the AT analysis.\nIV analyses are known to yield estimators with high variance, especially when compliance is low, which may also partially explain differences between the AT and SNM results.\nThe choice of which method to use depends on numerous factors, including study design, plausibility of meeting analysis assumptions, and available analytical resources; our conflicting estimates highlight the importance of testing the sensitivity of multiple fidelity analysis options.\nStrengths and limitations\nThe design, methods and approach of the WASH HELPS study were robust.\nRandomized controlled trials offer the greatest potential for causal inference.\nThe longitudinal design allowed us to collect data across three full school years of in Group 1 schools and two full school years of in Group 2 schools, allowing us to capture inter-seasonal and inter-year variations in the outputs, outcomes, and impacts.\nAll data were collected during unannounced school visits so that schools could not prepare for the visit and bias observations.\nOur primary measure of impact \u2013 roll-call absence - is an objective measure of school absence.\nThis impact evaluation was conducted by external researchers, to foster an unbiased assessment of the project impact.\nOur field team was composed of experienced Laotian enumerators to ensure the tools were designed and delivered with cultural and contextual appropriateness.\nThis robust study design lends strong internal validity, and results may be generalized to the larger, nationwide WinS project.\nThis was an effectiveness trial evaluating an intervention as conducted in a real-world setting.\nThe lessons from this project, taken with other recent WinS trials, reveal heterogeneity of findings that can inform programming across contexts.\nLastly, in addition to comparing two methods to analyze the effect of intervention fidelity on WinS impacts, our fidelity analysis also examines adherence to intervention behaviors, which has not been previously included in WinS fidelity analyses.\nThere are a number of limitations to this evaluation.\nFirst, the secondary health impact measures (diarrhea, symptoms of respiratory infection, conjunctivitis) were based on self-report by pupils, which may be subject to bias, and this evaluation was not blinded for either the beneficiaries or data collectors.\nMore objective and robust measures of pupil health, such as molecular methods to detect enteric infection in stool samples, would improve our confidence in the reported impacts, though these measures can be costly, time consuming, and require specialized equipment and laboratory staff.\nAs a way to measure potential reporting bias, we included a negative control question about symptoms of illness unrelated to WASH access (cuts/scrapes and toothache) at the last survey visit.\nDifferences in reported symptoms of these illnesses between intervention and comparison groups would indicate a potential reporting bias, but we found no evidence to suggest that any bias may have existed to a greater degree among either the intervention group or the comparison group.\nAdditionally, schools in the comparison group did not have functional WASH facilities, so it is unlikely that the null results could be explained by a change in behaviors among the comparison group.\nSecond, the intervention was delivered across two different school years, so Group 1 schools had one more year of surveillance than Group 2 schools.\nFollowing a single cohort of schools over the same time period would have provided a more accurate measure of WinS hardware and software performance, sustainability, and impact.\nThird, implementation was delayed in many Group 1 schools.\nThe intervention was fully implemented in Group 1 schools at visit 4, with the exception of Samoui district, in which the intervention was fully implemented at visit 9.\nOur analysis excludes visits prior to full intervention implementation, thus power may have been limited by dropping observations under incomplete intervention delivery.\nLast, we were unable to account for the quality of intervention design or dose of the intervention received, which are important components of fidelity and adherence.\nCONCLUSIONS\nOur findings and those of other rigorous WinS trials suggest that WinS programs \u2013 as currently designed and delivered \u2013 do not have a population-level benefit on education and health.\nIn this context, the WinS improvements alone were not sufficient to address the other powerful causes of absenteeism, enrollment, and dropout that are not related to \u2013 but possibly more influential than \u2013 school WASH.\nWe believe this likely holds in many similar settings.\nSimilarly, WinS improvements, though potentially critical for the enabling environment, may not be sufficient to overcome disease transmission in areas where community and household WASH coverage is poor.\nWinS, independent of its stated purpose of improving education and health, is an important objective for dignity, inclusivity, and development.\nHowever, if intended impacts are to be achieved, improving intervention fidelity and adherence and including other complementary approaches for WASH may be required.\nTo better understand how to improve intervention fidelity and adherence, evaluations of WinS interventions need to better understand and adapt to contextual drivers of key impacts and outcomes, further develop and test theories of change, and conduct rigorous process evaluations to understand where along the causal pathways interventions are falling short.\nAssociation between intervention fidelity and adherence continuum and intervention impacts.\nFlow diagram of school and pupil selection.\n\nIntervention outputs and behavioral outcomes and their measurement indicators\nOutput | Indicator and criteria\nWater supply | \u2022 Improved* water point on school compound\n\u00a0\u00a0\u00a0\u00a0- Water point functional in the previous year (director reported)\n\u2022 Water tank to supply toilet and handwashing stations\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water observed in tank\nToilets | \u2022 At least one improved* toilet compartment\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is sex separated (by designation)\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is unlocked\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is clean\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet has water available inside compartment for flushing\nHandwashing facilities | \u2022 At least one individual handwashing station available to pupils\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water available at individual handwashing station\n\u00a0\u00a0\u00a0\u00a0\u00a0- Soap available at individual handwashing station\nPromotion of daily group hygiene activities | \u2022 Daily group handwashing schedule posted in at least one classroom or near toilet\n\u2022 Daily group compound cleaning schedule posted in at least one classroom or near toilet\n\u2022 Daily group toilet cleaning schedule posted in at least one classroom or near toilet\nGroup handwashing | \u2022 Group handwashing facility available to pupils\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water available at group handwashing facility\n\u00a0\u00a0\u00a0\u00a0\u00a0- Soap available at group handwashing facility\nWater filters | \u2022 At least one drinking water filter available in a classroom for pupil use\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water in filter\nOutcome | Indicator\nToilet use | \u2022 Percentage of students using toilet for defecation during school hours (pupil-reported)\nHandwashing (individual) | \u2022 Percentage of students washing hands with soap and water upon exiting toilet (observation)\nDaily group handwashing | \u2022 School conducted daily group handwashing the day of visit (observation)\nDaily group toilet cleaning | \u2022 Percentage of students participating in daily group toilet cleaning within the previous five school days (pupil-reported)\nDaily group compound cleaning | \u2022 Percentage of students participating in daily group compound cleaning within the previous five school days (pupil-reported)\n\n*Defined according to Joint Monitoring Programme (JMP) standards.\n\nAssociation between WinS intervention and health and educational impacts, Saravane Province, Lao People\u2019s Democratic Republic, 2014-2017 (n\u2009=\u2009100 schools)\nImpact | Comparison* | Intervention* | Adjusted risk ratio | 95% confidence interval\nRoll-call absence\u2020 | 6024 (32.2%) | 7147 (29.9%) | 1.01 | 0.84, 1.20\nEnrollment\u2021 | 68.2 (49.7) | 71.6 (50.0) | 1.07 | 0.84, 1.35\nDropout\u2021 | 0.8 (2.6) | 0.4 (1.0) | 0.56 | 0.25, 1.24\nGrade progression\u2021 | 64.4 (48.6) | 67.3 (48.6) | 1.07 | 0.91, 1.25\nDiarrhea\u2020,\u00a7 | 1032 (21.1%) | 947 (14.7%) | 0.80 | 0.51, 1.26\nSymptoms of respiratory infection\u2020,\u2016 | 1414 (28.9%) | 2064 (32.1%) | 1.08 | 0.95, 1.23\nConjunctivitis\u2020,\u00a7 | 41 (0.8%) | 48 (0.8%) | 0.89 | 0.53, 1.52\nPrevalence of any STH\u2020,\u00b6 | 1833 (39.8%) | 1935 (41.6%) | 1.00 | 0.85, 1.17\n\n*Data are n (%) for impacts at the pupil level (roll-call absence, diarrhea, symptoms of respiratory infection, conjunctivitis, and prevalence of STH) and mean (SD) for impacts at the school-level (enrollment, dropout, grade progression) across study period.\n\u2020Risk ratios were calculated using a Poisson model with robust error variances and random intercepts at the school and pupil level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size, and season (rainy or dry). Absence model additionally controlled for and rice crop calendar (planting, growing, harvesting).\n\u2021Risk ratios were calculated using a Poisson model with random intercepts at the school level. All models adjusted for district and visit number.\n\u00a7Pupil-reported in previous week.\n\u2016Pupil-reported cough, runny nose, stuffy nose, or sore throat in previous week.\n\u00b6Samples testing positive for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus).\n\nAssociation between WinS intervention fidelity and adherence and absence, diarrhea, respiratory infection, and soil-transmitted helminth infection (STH), Saravane Province, Lao PDR, 2014-2017 (n\u2009=\u2009100 schools)\n | As-treated analysis | Structural nested model analysis\n | Adjusted risk ratio* | 95% confidence interval | Adjusted risk ratio\u2020 | 95% confidence interval\nRoll-call absence:\nFidelity\u2021 | 0.76 | 0.64, 0.91 | 0.97 | 0.33, 2.81\nAdherence\u2021 | 0.91 | 0.79, 1.05 | 0.96 | 0.19, 4.97\nDiarrhea:\u00a7\nFidelity, dry season\u2021 | 0.84 | 0.48, 1.49 | 0.45 | 0.24, 0.85\nAdherence, dry season\u2021 | 1.00 | 0.70, 1.44 | 0.42 | 0.21, 0.87\nFidelity, rainy season\u2021 | 1.65 | 0.82, 3.33 | 1.03 | 0.42, 2.51\nAdherence, rainy season\u2021 | 1.41 | 0.61, 3.26 | 0.50 | 0.19, 1.30\nSymptoms of respiratory infection\u2016:\nFidelity\u2021 | 1.00 | 0.89, 1.14 | 1.41 | 0.93, 2.13\nAdherence\u2021 | 0.97 | 0.84, 1.11 | 2.30 | 0.54, 8.87\nPrevalence of any STH:\u00b6\nFidelity\u2021 | 1.20 | 1.01, 1.43 | 1.10 | 0.57, 2.13\nAdherence\u2021 | 0.93 | 0.77, 1.12 | 1.18 | 0.37, 3.73\n\nSTH \u2013 soil-transmitted helminth\n*Risk ratios calculated using a Poisson model with robust error variances and random intercepts at the school and pupil level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size, season (rainy or dry). Absence models additionally controlled for rice crop calendar (planting, growing, harvesting).\n\u2020Risk ratios calculated using a Structural Nested Model with random intercepts at the school level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size.\n\u2021Fulfilling \u226575% of intervention outputs was considered fidelity. Fulfilling \u226575% of intervention outcomes was considered adherence.\n\u00a7Pupil-reported in previous week.\n\u2016Pupil-reported cough, runny nose, stuffy nose, or sore throat in previous week.\n\u00b6Samples testing positive for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus).",
        "label": "low",
        "id": "task4_RLD_test_779"
    },
    {
        "paper_doi": "10.1371/journal.pone.0006857",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Non-randomised controlled cluster trial\n\n\nParticipants: Country: TanzaniaSetting (coverage): 2 intervention districts, 1 control districtOutlets: Small drug shops (duka la dawa baridi)Age group: All age groups\n\n\nInterventions: Intervention: Subsidised ACT (AL)Comparison: No ACT subsidy (control)Supportive interventions: Behavior change communication (e.g. local radio advertisements, wall paintings, themed cultural shows) emphasising the importance of using ACTs and their availability in private shops\n\n\nOutcomes: ACT uptake, availability and price\n\n\nNotes: The project managers procured AL from the manufacturer, Novartis, and sold them to a pharmaceutical wholesaler in Dar es Salaam at an average of US$ $0.11 per dose, 88% below the price offered to public buyers.In one of the intervention districts (Kongwa), the suggested retail price intended to inform consumers of the maximum amount they should pay was set at 300, 600, 900, and 1200 Tanzanian shillings (0.25, 0.50. 0.75. and 1 USD respectively) for the four weight packs respectively; no suggested retail price was included on drugs distributed to Maswa in order to test its effect on price outcomes\n\n",
        "objective": "To assess the effect of programmes that include ACT price subsidies for private retailers on ACT use, availability, price and market share.",
        "full_paper": "Background\nWHO estimates that only 3% of fever patients use recommended artemisinin-based combination therapies (ACTs), partly reflecting their high prices in the retail sector from where many patients seek treatment.\nTo overcome this challenge, a global ACT subsidy has been proposed.\nWe tested this proposal through a pilot program in rural Tanzania.\nMethods/Principal Findings\nThree districts were assigned to serve either as a control or to receive the subsidy plus a package of supporting interventions.\nFrom October 2007, ACTs were sold at a 90% subsidy through the normal private supply chain to intervention district drug shops.\nData were collected at baseline and during intervention using interviews with drug shop customers, retail audits, mystery shoppers, and audits of public and NGO facilities.\nThe proportion of consumers in the intervention districts purchasing ACTs rose from 1% at baseline to 44.2% one year later (p<0.001), and was significantly higher among consumers purchasing for children under 5 than for adults (p\u200a=\u200a0.005).\nNo change in ACT usage was observed in the control district.\nConsumers paid a mean price of $0.58 for ACTs, which did not differ significantly from the price paid for sulphadoxine-pyrimethamine, the most common alternative.\nDrug shops in population centers were significantly more likely to stock ACTs than those in more remote areas (p<0.001).\nConclusions\nA subsidy introduced at the top of the private sector supply chain can significantly increase usage of ACTs and reduce their retail price to the level of common monotherapies.\nAdditional interventions may be needed to ensure access to ACTs in remote areas and for poorer individuals who appear to seek treatment at drug shops less frequently.\nTrial Registration\nControlled-Trials.com ISRCTN39125414.\nIntroduction\nArtemisinin-based combination therapies (ACTs) have become a mainstay of malaria treatment because of their high efficacy and their potential to delay the development of antimalarial resistance.\nYet despite availability of substantial donor funding, the proportion of fevers treated with ACTs remains minimal.\nThis partly reflects limited ACT availability outside the public sector.\nIn many countries, 40\u201360% of people seek treatment for fever or malaria from private vendors, such as pharmacies, drug shops and general stores.\nHowever, ACTs are typically sold at retail prices 20\u201340 times those of common alternatives such as amodiaquine and sulphadoxine-pyrimethamine (SP), restricting their uptake by consumers, particularly in rural areas.\nAs a result, most retail sector anti-malarial customers continue to use older therapies for which widespread resistance has been reported or artemisinin monotherapies, which are strongly discouraged by the World Health Organization because their use is likely to accelerate the development of artemisinin resistance.\nMany others purchase antipyretics only.\nAnticipating this challenge, in 2004 the Institute of Medicine recommended a global subsidy of ACTs as the best means to achieve high coverage and prolong the efficacy of these drugs.\nIt argued that reducing the ex-factory price of ACTs to that of common alternatives (roughly $0.05) would ensure their widespread distribution through private channels and crowd out other drugs.\nThis concept was further developed by the Roll Back Malaria Partnership and launched by the Board of the Global Fund in November 2008 as the Affordable Medicines Facility-malaria (AMFm).\nThe AMFm is scheduled to be launched in 11 countries within the coming year.\nHowever, the dearth of evidence on the likely impact of such a subsidy has hindered its design and raised numerous concerns about investing public money in an unproven mechanism.\nTo fill this evidence gap, we piloted the AMFm model in two rural Tanzanian districts starting in October 2007.\nThis report presents the results from the pilot and assesses their implications for implementation of the AMFm and other large-scale subsidies.\nMethods\nStudy design\nThe intervention was conducted in two rural districts of Tanzania: Maswa in Shinyanga region and Kongwa in Dodoma region.\nA third district, Shinyanga Rural in Shinyanga region, served as a control (see Figure 1).\nWe conducted a detailed analysis of all districts in the country according to a range of key indicators, including malaria endemicity, population per health facility, employment, prevalence of private drug shops, and bed net ownership.\nThe selected districts were among the few roughly comparable across all indicators, with high malaria transmission, large numbers of private drug shops and, importantly, no malaria-related trials (e.g., vaccines) underway.\nSocioeconomic status in the three districts was below national averages as evidenced by comparison of key assets such as housing materials, toilet facilities, and availability of electricity.\nThe selected districts were randomly assigned to one of the three arms in the study design: subsidy, subsidy plus suggested retail price, and no subsidy (control).\nAs two of the qualified districts were adjacent, randomization was limited so that one of the adjacent districts served as the control.\nThe project centered on the distribution of ACTs at highly subsidized prices.\nThe project managers procured quantities of artemether-lumefantrine (AL), the recommended first-line ACT in Tanzania, from the manufacturer, Novartis, and sold them to a pharmaceutical wholesaler in Dar es Salaam at an average of $0.11 per dose, 88% below the price offered to public buyers (private buyers are typically offered substantially higher prices).\nThe wholesaler, Nufaika Distributors Limited, was selected due to its extensive distribution network and interest in selling ACTs, among other factors, following due diligence of 10 similar businesses.\nThe wholesaler received no instructions other than to sell the ACTs to drug shops in the two intervention districts according to its standard practices.\nIt was made clear that the wholesaler would not be monitored or held accountable for its pricing, stocking, or other practices.\nSmall drug shops, known as duka la dawa baridi (DLDB), were the primary retail outlet for subsidized ACTs in the project.\nPast studies have found that these shops are the most important commercial source of anti-malarials in Tanzania and it is estimated that there are more than 8,000 in operation nationwide in rural and urban areas.\nUnder Tanzanian law, these shops are permitted to sell over-the-counter medicines, but not products requiring a prescription (an exception was granted for ACTs for this project).\nSmall volumes of anti-malarials are also sold through general stores alongside common staples, but distribution to these outlets was not encouraged as they are not legally permitted to sell pharmaceuticals.\nAL was distributed in four weight-specific packs, in redesigned packaging with simplified dosing instructions in the local language, Kiswahili.\nACTs distributed to Kongwa were marked with a suggested retail price (SRP) intended to inform consumers of the maximum amount they should pay.\nThe SRP was set at 300, 600, 900, and 1200 Tanzanian Shillings (0.25, 0.50.\n0.75. and 1 USD respectively) for the four weight packs respectively based on an analysis of costs in the supply chain; no SRP was included on drugs distributed to Maswa in order to test its effect on price outcomes.\nThe AMFm will support countries to conduct accompanying interventions such as training, behavior change communication (BCC), and regulatory strengthening to facilitate the effective distribution and use of subsidized ACTs.\nThis project accordingly included some of those activities in the intervention districts.\nPrior to the initiation of the subsidy, the Tanzania Food and Drug Authority (TFDA) conducted a one-day training of DLDB attendants focusing on malaria symptoms and ACT dispensing and dosing.\nPopulation Services International conducted a range of BCC activities, including local radio advertisements, wall paintings, and themed cultural shows, throughout the project.\nThe activities emphasized the importance of using ACTs and their availability in private shops, as well as basic messages on the dangers of malaria and the importance of prompt treatment-seeking.\nData collection\nThe study's primary outcomes were the proportion of antimalarial consumers visiting DLDB who purchased subsidized ACTs and the price they paid for the drugs.\nSecondary outcomes included the proportion of DLDB stocking the product, the socioeconomic status of the consumer, the age of the intended patient, and ACT distribution by public facilities during the same period.\nWe employed four data collection methodologies: exit interviews, retail audits, mystery shoppers, and public facility audits.\nAll four methodologies were conducted together five times during the project: once prior to the launch of the subsidy in August 2007 and four times throughout the intervention in November 2007 and March, August, and November 2008.\nData were collected at all DLDB and public facilities in the three districts.\nDLDB were initially identified through TFDA records, with unregistered shops captured through discussions with local informants and systematic physical reconnaissance throughout each district.\nGeographical positions of all shops were recorded using hand-held GPS units (Garmin Etrex).\nFor the exit interviews, data collectors positioned themselves near a DLDB and remained there for the full business day.\nAll customers emerging were asked to answer a short questionnaire on the products bought, and the brand of the product was visually verified.\nTo assess the customers' socio-economic status, interviewees were asked a series of questions about their household assets from the 2003\u201304 Tanzania HIV/AIDS Indicator Survey.\nDLDB retail audits were conducted twice during every data collection period, at a four week interval.\nCollectors recorded the volume of all anti-malarial stocks present.\nA short questionnaire was also administered to the owner or attendant to determine the amount of each product newly purchased and disposed of (e.g., due to expiry or damage) during the previous four weeks.\nStock levels were then compared between the two audits and purchases and disposals subtracted to determine the volume of sales during that period.\nDLDB pricing and dispensing practices were also observed by having collectors visit each shop once per survey posing as a consumer seeking malaria treatment.\nTwo such \u201cmystery shopper\u201d scenarios were employed \u2013 adults purchasing for themselves and purchasing for a nine-month old child at home with malaria symptoms.\nData collectors also visited all public and non-governmental organization (NGO) health facilities in each survey period to review ACT stocks and dispensing records.\nIn this paper we focus primarily on data from the baseline in August 2007 and follow-up in August 2008 because of potential seasonal variation in malaria transmission and treatment seeking.\nTo enable robust analysis of the impact of the SRP, pricing data are pooled across intervention surveys and compared to the baseline.\nA full set of results for all data collection periods is available at http://www.cshor.org/TanzaniaPilotData.xls.\nData analysis\nTo assess geographical variation in outcomes, the competition level of all DLDB was calculated using the fixed radius approach.\nThe competitive space of each DLDB was defined as 1 kilometer and each shop was assigned to a competition index category between 0 and 5 based on the number of other DLDB within that radius.\nExit interviewees were allocated to wealth quintiles using the asset weights and quintile break-points generated through principal components analysis of 53 variables from the 2003\u20132004 HIV/AIDS Indicator Survey.\nDue to operational challenges associated with administering asset ownership questions at shops (as opposed to at household level as is typical), several variables from the original survey were not captured, including additional forms of household lighting, water source, and land ownership status.\nAsset weights and national quintiles were therefore calculated using only those variables captured in the exit interviews.\nTo enable comparison of price across products, the exact number of pills purchased (syrups and injectables were excluded) by interviewees or mystery shoppers was recorded and the price paid for a full appropriate dose for the intended patient then extrapolated using the standard dosing schedule according to Tanzania national treatment guidelines or product registration with the TFDA.\nSimilarly, to compare sales across products, we converted the total number of pills sold to adult equivalent doses based on the recommended dosing.\nSurvey data were analyzed using SPSS v.14/16 (Chicago, Illinois) and SAS v.9.1 (Cary, NC), and GPS data using MapInfo v.7.8 (Troy, New York).\nProportions were compared using chi-square tests.\nStudent t-tests were used to compare means and the Satterthwaite t-test was used when variances were significantly unequal.\nA repeated measures multivariate regression model was used to compare differences in purchase price while controlling for potentially confounding factors and adjusting for clustering of multiple purchases in the same shops.\nEthical Considerations\nAs a pilot project of the Tanzania Ministry of Health and Social Welfare to prepare for a national program, the interventions were developed with the Tanzania Food and Drug Authority and National Malaria Control Program according to the policies and guidelines of the Ministry and approved by the Chief Medical Officer accordingly.\nNo additional interventions were added as part of the subsequent evaluation.\nOral informed consent was obtained from all consumers emerging from drug shops as well as from drug shop owners prior to the administration of retail audits.\nNo ethnic or individual identifying information was captured.\nThis study complied with the guidelines of the Declaration of Helsinki.\nResults\nTable 1 summarizes the observations recorded for each methodology.\nThe total number of DLDB audited increased from 200 in August 2007 to 216 in August 2008 due to the opening of new shops, with 30 (13%) and 39 (15%) additional shops closed or refusing to participate at these two time periods respectively.\nA similar number of all shops observed were in areas of high competition (36% and 38% in categories 4 and 5 in August 2007 and August 2008 respectively) as in areas of low competition (35% in categories 0 and 1 in both periods).\nHigh competition DLDB were located in population centers while those in low competition categories were 20\u201325 km from major roads.\nThe majority of consumers interviewed in all districts in August 2007 and August 2008 were from the two least poor SES quintiles (59% and 68% respectively).\nStocking\nThere was a pronounced increase in the proportion of shops stocking ACTs in the intervention districts, from 0/133 in August 2007 to 109/151 (72.2%) in August 2008 (p<0.001).\nShops stocking ACT in the control district changed negligibly from 1/67 (1%) to 0/65 over the same period.\nShops with two or more other shops in their competition radius were significantly more likely to stock ACTs in August 2008 (82/101, 81.2%) than those with 0 or 1 competitor (27/50, 54.0%; p<0.001).\nBy comparison, stocking of some other common anti-malarials was more consistent across competition categories: 75/101 (74.3%) of shops in category 2 and above and 34/50 (68.0%) in categories 0 and 1 stocked amodiaquine, a non-significant difference.\nPricing\nInterviewed consumers paid a mean price of $0.58 for all ACTs (SD\u200a=\u200a$0.28) during the study period (Figure 2).\nConsumers purchasing ACTs for children under 5 paid significantly less than those buying for adults (16+ years), at a mean expenditure of $0.35 (SD\u200a=\u200a$0.19) and $0.70 (SD\u200a=\u200a$0.28) respectively (p<0.001).\nOverall, the average price paid for ACTs for adults did not differ significantly from expenditures on SP ($0.67, SD\u200a=\u200a$0.34), but was significantly higher than for amodiaquine ($0.48, SD\u200a=\u200a$0.27; p<0.001).\nThe mean price paid for ACTs for children under 5 was significantly less than for both SP ($0.51, SD\u200a=\u200a$0.28; p\u200a=\u200a0.001) and AQ ($0.86, SD\u200a=\u200a$0.30; p<0.001).\nControlling for the age of the intended recipient, the district in which the drug was purchased, and clustering of multiple purchases in shops, the price paid for ACTs did not vary significantly by either the SES quintile of the consumer or the competition category of the shop.\nConsumers paid significantly more for the three largest of the four AL weight packs in Kongwa than in Maswa (all p<0.001), while no significant difference was observed for the lowest (5\u201315 kg) dose.\nIn Kongwa, the mean price paid for two AL doses, infant (5\u201315 kg) at $0.34 (SD\u200a=\u200a$0.24; p<0.001) and child (15\u201325 kg) at $0.55 (SD\u200a=\u200a$0.15; p\u200a=\u200a0.034), were significantly higher than the SRP.\nMean prices did not differ significantly from the SRP for the other two doses.\nUptake and Sales\nTable 2 presents exit interview data on uptake for AL, artemisinin monotherapy, and the two other most commonly purchased antimalarials, amodiaquine and SP.\nThe proportion of anti-malarial consumers in the intervention districts who purchased ACTs increased strikingly during the project, from 4/399 (1.0%) in August 2007 to 201/455 (44.2%) in August 2008 (p<0.001).\nThis proportion subsequently declined insignificantly to 227/572 (39.7%) in November 2008.\nOver the same period, purchases of SP and amodiaquine in the intervention districts declined significantly, from 232/399 (58.2%) to 163/455 (35.8%; p<0.001) and 146/399 (36.6%) to 75/455 (16.5%; p<0.001) respectively.\nUse of artemisinin monotherapies remained negligible throughout the study.\nPurchases by mystery shoppers followed a similar trend, with the proportion of shoppers offered ACTs in intervention districts rising from 6/133 (4.5%) to 83/146 (56.9%; p<0.001) and those offered SP declining from 83/133 (62.4%) to 38/146 (26.0%; p<0.001).\nWhen restricted to only shops stocking ACTs, the proportion of interviewed consumers and mystery shoppers buying ACTs was significantly higher than when all shops were included: 55.5% v. 44.2% (p\u200a=\u200a0.001) and from 76.4% v. 56.9% (p\u200a=\u200a0.001) respectively.\nNo change in ACT purchasing was observed in the control district.\nIn August 2008, 44/83 (53.0%) of consumers purchasing anti-malarials for a child under 5 bought ACTs compared to 104/291 (35.7%) of those purchasing for an adult (p\u200a=\u200a0.005).\nThere was no correlation between the SES of the consumer and the likelihood of buying ACTs, with ACTs comprising 44.4% of purchases by consumers in the poorest two quintiles (n\u200a=\u200a45) compared to 42.4% by those in the least poor quintiles (n\u200a=\u200a328).\nSimilarly, there was no significant difference in the proportion of consumers buying ACTs between high and low competition stores.\nRetail audits showed that 9,786 adult equivalent doses of ACTs (60.3% of all adult equivalent anti-malarial doses) were sold by DLDB in a 4-week period in July/August 2008, while ACT sales in the control district remained negligible (see Figure 3).\nOverall distribution of ACTs in the intervention districts, including dispensing from public and NGO facilities, increased 62% between November 2007 (the first collection following intervention) and August 2008 from 30,946 to 69,068 doses, with ACTs distributed through the private sector accounting for 38% of that growth.\nDiscussion\nThe introduction of subsidized ACTs resulted in a rapid and pronounced increase in the proportion of people accessing ACTs from private shops from close to zero to 44% after one year.\nDistribution of ACTs through the public sector rose at the same time, indicating that the intervention contributed to increases in overall volumes of the drug distributed.\nImportantly, the greatest increases in ACT usage at DLDB were by those seeking treatment for children under 5, the group at the greatest risk of malaria mortality.\nHowever, as other studies have found, purchases at DLDB for children under 5 were modestly underrepresented compared to the estimated fever incidence for this age group.\nThe rise in ACT usage appears to have crowded out the use of some sub-optimal therapies such as SP and AQ, although a substantial number of consumers continued to purchase these drugs at the end of the study.\nThere was no significant change in ACT uptake during the last 3 months of the study, but this does not suggest that a long-term plateau in ACT purchasing had been reached.\nThe intervention is targeting a fundamental shift in the market for anti-malarials, which will require years to fully realize.\nAs shown in other studies, poorer individuals appear to have sought treatment for malaria at DLDB substantially less frequently than wealthier ones, suggesting that additional interventions may be needed to increase ACT access among this population.\nThe subsidy had the intended effect of reducing the retail price of ACTs to levels similar to commonly used alternatives.\nOn average, consumers paid 93% less than ACT prices regularly observed in private outlets in rural and urban areas across Tanzania.\nContrary to competition theory and common concern about the AMFm, consumers did not pay significantly more for ACTs at more remote shops facing less competition.\nThese results also contradict another common critique of large-scale ACT subsidies: that businesses would apply excessive mark-ups thereby minimizing the benefit to consumers.\nThere were almost no instances of such \u201cprice gouging,\u201d with 86% of all purchases within $0.08 of the SRP for the dose and the highest price paid still 81% below typical ACT prices in Tanzanian shops.\nThe SRP appears to have had the opposite of the intended effect, artificially inflating prices in Kongwa above those determined by the market in Maswa.\nThe SRP levels were determined based on estimated costs and profit margins in the private anti-malarial supply chain derived from interviews with more than 50 businesses.\nThat those levels were substantially above the retail market prices suggests that an SRP should be used with caution, based on a more detailed understanding of pricing practices and only in cases where unreasonable profit margins are being charged.\nAlthough stocking of ACTs rose substantially during the intervention, it was skewed towards shops in towns and other population centers.\nWhen analysis was restricted to those shops stocking ACTs, purchases of the drug increased markedly, indicating that availability may have served as a major limitation on overall uptake.\nMany shopkeepers described high consumer demand for ACTs but expressed frustration at problems in obtaining the drugs.\nThis suggests that addressing supply chain issues should be a central focus of large-scale subsidy plans.\nIn particular, since wholesalers often lack an inherent financial incentive to distribute to remote outlets, public sector means of creating such incentives, such as providing a substantial rebate to wholesalers for achieving certain coverage levels in remote areas, should be explored.\nCaution should be used in directly applying these findings to other settings.\nSocioeconomic factors, malaria treatment-seeking behavior, and the structure of private supply chains all vary widely between and within countries.\nThis study operated through only one wholesaler and one type of retail outlet, while national-scale subsidies will employ multiple of both.\nAnd although concerted measures were put in place to limit the Hawthorne Effect, it is possible that businesses and consumers were influenced by the presence of the study team.\nNevertheless, these results are cause for cautious optimism that subsidies applied at the top of the private supply chain can lead to rapid and dramatic increases in ACT usage.\nConsumer demand for ACTs was high and average retail prices remained low due to businesses applying modest mark-ups on the product, similar to those for older drugs such as SP and amodiaquine.\nThe study also highlights key areas for further research on this topic.\nOverall changes in ACT coverage and treatment-seeking behavior, including among different SES groups, should be robustly assessed through studies collecting household level data.\nMoreover, the use of privately distributed ACTs for non-malarious fevers and opportunities for effectively introducing diagnosis into the private sector should be explored.\nHowever, the need for further research should not delay implementation.\nNo amount of piloting will fully recreate the conditions of a national or global subsidy, and some \u201clearning by doing\u201d will be inevitable.\nOur findings should provide sufficient confidence for large-scale implementation of the subsidy model as a central part of the global effort to increase ACT access from current dismal levels towards the Roll Back Malaria target of 80% of patients by the end of 2010.\nLocation of project districts by role in study.The red areas denote the two intervention districts, while the green area shows the control district.\nPrice paid for ACTs and most common alternative anti-malarials by interviewed consumers.Prices of subsidized ACTs and the most commonly purchased alternative anti-malarial \u2013 amodiaquine (AQ) or sulphadoxine-pyrimethamine (SP) \u2013 observed in the two intervention districts between November 2007 and November 2008 are displayed by intervention district and age of intended recipient. The thick blue line denotes the median and the red X the mean, with the surrounding box delineating the interquartile range (IQR). The lines extending from each box mark 1.5 times the IQR, with dots showing outliers that do not fall within this range.\nStocking and sales of subsidized ACTs at drug shops in intervention districts.All duka la dawa baridi (DLDB) in Maswa (top) and Kongwa (bottom) districts are mapped as either white (ACTs not in stock at time of survey) or black (ACTs in stock) circles. Data is shown at baseline (August '07) and two periods during implementation (November '07 and August '08). Districts are divided into 10 km2 squares, with the total volume of adult equivalent doses of ACTs sold in that area over the preceding four weeks shown by the color of shading as follows: Tan\u200a=\u200a1 to 50 doses sold; Orange\u200a=\u200a51 to 250 doses; Light red\u200a=\u200a251 to 500 doses; and Dark red\u200a=\u200a501 to 10,000 doses.\n\nRecorded observations by methodology, characteristic, and district, August 2007 and August 2008.\n | August 2007 | August 2008\n | Maswa | Kongwa | Control | Maswa | Kongwa | Control\nDLDB Audited | 73 | 60 | 67 | 83 | 68 | 65\nComp Index 0 | 12 | 7 | 15 | 13 | 20 | 14\nComp Index 1 | 12 | 15 | 9 | 10 | 7 | 12\nComp Index 2\u20133 | 21 | 18 | 20 | 28 | 18 | 13\nComp Index 4\u20135 | 3 | 5 | 6 | 0 | 10 | 18\nComp Index 5+ | 25 | 15 | 17 | 32 | 13 | 8\nExit Interviewees | 346 | 53 | 181 | 167 | 288 | 118\nBuying for Ages 16+ | 275 | 37 | 125 | 79 | 212 | 76\nBuying for Ages 5\u201316 | 28 | 7 | 10 | 39 | 42 | 10\nBuying for Ages<5 | 43 | 9 | 46 | 49 | 34 | 32\nSES Quintile 1 (poorest) | 3 | 4 | 21 | 8 | 12 | 4\nSES Quintile 2 | 27 | 7 | 40 | 16 | 9 | 8\nSES Quintile 3 | 62 | 8 | 66 | 32 | 50 | 9\nSES Quintile 4 | 137 | 26 | 45 | 56 | 112 | 52\nSES Quintile 5 (least poor) | 117 | 8 | 9 | 55 | 105 | 45\nMystery Shoppers | 73 | 60 | 67 | 81 | 65 | 65\nPublic/NGO facilities audited | 38 | 33 | 34 | 35 | 36 | 36\n\n\nPurchase of anti-malarials by exit interview customers by district, August 2007 and August 2008.\n | August 2007 | August 2008\n | Maswa | Kongwa | Total Intervention Districts | ControlDistrict | Maswa | Kongwa | Total Intervention Districts | Control District\nAdults (ages 16+)\nAntimalarial purchases of which: | 275 | 37 | 312 | 125 | 79 | 212 | 291 | 76\nACT | 3 (1%) | 1 (3%) | 4 (1%) | 0 (0%) | 38 (48%) | 66 (31%) | 104 (35%) | 0 (0%)\nSP | 187 (68%) | 26 (70%) | 213 (68%) | 78 (62%) | 30 (38%) | 118 (56%) | 148 (51%) | 63 (83%)\nAQ | 71 (26%) | 10 (27%) | 81 (26%) | 41 (33%) | 8 (10%) | 23 (11%) | 31 (11%) | 12 (16%)\nChildren (ages<5)\nAntimalarial purchases of which: | 37 | 6 | 43 | 35 | 49 | 34 | 83 | 32\nACT | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 20 (41%) | 24 (71%) | 44 (53%) | 2 (6%)\nSP | 1 (3%) | 2 (33%) | 3 (7%) | 3 (9%) | 1 (2%) | 2 (6%) | 3 (4%) | 3 (9%)\nAQ | 35 (95%) | 4 (67%) | 39 (91%) | 32 (91%) | 22 (45%) | 8 (24%) | 30 (36%) | 22 (69%)\n",
        "label": "high",
        "id": "task4_RLD_test_357"
    },
    {
        "paper_doi": "10.1186/s40608-014-0021-5",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Setting: university conference room, USADesign: randomised controlled trialRecruitment: on a university campus through announcements in lectures and through electronic bulletin boardsAllocation to groups: randomised using a random number generator\n\n\nParticipants: Overweight or obese women (n = 62)Mean age: 21.87 (SD 3.03), range 18-33Ethnicity: 45.16% Hispanic or Latino; 27.42% Black/African American; 4.84% Caribbean non-Hispanics; 8.06% Asian/Pacific Islander; 3.23% white Non-Hispanic; 9.68% mixed race; 1.61% don't know/not sureMean BMI: 28.42 kg/m2 (SD 3.10)Education: 82% of all participants had a high school degree/equivalency, some college or a 2-year college degree\n\n\nInterventions: Intervention 1: menu with energy (kcal) information (n = 20)Intervention 2: menu with energy (kcal) information and with exercise equivalents (n = 20)Control: menu with no energy information (n = 22)\n\n\nOutcomes: Energy (kcal) consumption during a meal; measured by weighing leftover food\n\n\nNotes: Food offered was fast food from Burger King. Participants attended twice for baseline and intervention meal.Subgroup analysis was conducted by the study author for restrained vs unrestrained eaters. A repeated measures analysis of variance was conducted. Data were combined from intervention 1 and 2 and compared with the control. This study was a thesis dissertatio\n\n",
        "objective": "To assess the impact of nutritional labelling for food and non\u2010alcoholic drinks on purchasing and consumption of healthier items. Our secondary objective was to explore possible effect moderators of nutritional labelling on purchasing and consumption.",
        "full_paper": "Background\nBetter techniques are needed to help consumers make lower calorie food choices.\nThis pilot study examined the effect of menu labeling with caloric information and exercise equivalents (EE) on food selection.\nParticipants, 62 females, ages 18-34, recruited for this study, ordered a fast food meal with menus that contained the names of the food (Lunch 1 (L1), control meal).\nOne week later (Lunch 2 (L2), experiment meal), participants ordered a meal from one of three menus with the same items as the previous week: no calorie information, calorie information only, or calorie information and EE.\nResults\nThere were no absolute differences between groups in calories ordered from L1 to L2.\nHowever, it is noteworthy that calorie only and calorie plus exercise equivalents ordered about 16% (206\u00a0kcal) and 14% (162\u00a0kcal) fewer calories from Lunch 1 to Lunch 2, respectively; whereas, the no information group ordered only 2% (25\u00a0kcal) fewer.\nConclusions\nMenu labeling alone may be insufficient to reduce calories; however, further research is needed in finding the most effective ways of presenting the menu labels for general public.\nBackground\nPoint-of-purchase menu labeling, particularly at fast food restaurants, has been of special interest in the fight against obesity.\nAs fast food consumption has been correlated with obesity and other negative health outcomes, these food outlets are being targeted for change.\nIn 2010, the US federal health care reform bill was signed into law and includes a requirement that restaurant chains with at least 20 outlets nationwide post calorie labels on menu boards.\nIt has been suggested that knowledge of the calories contained in foods is essential to choosing and consuming an energy-balanced diet.\nWhile consumer polls show a desire for calorie information at the point-of-purchase in restaurants, research on the actual effects has shown mixed results.\nOne possible reason for inconsistent effectiveness may be the lack of understanding of the value of a calorie or the lack of a reference amount for a calorie.\nExercise equivalents have been discussed by nutrition experts as a potential method to inform consumers about calorie values.\nExercise equivalents are defined as the amount of time doing particular physical activities that would be needed to burn off calories in foods.\nFor example, burning off a 300-calorie hamburger would require about 75\u00a0minutes of walking, after expending the calories needed for daily subsistence.\nExercise equivalents could potentially simplify food and/or restaurant nutrition labels, increase understanding of calories and of energy imbalance, and facilitate a decrease in overall energy intake.\nLiterature exploring the use of exercise equivalents is limited.\nA new study by Dowray et al., explored the potential effect of exercise equivalents on menu labels.\nThis study was a web-based survey, and asked participants to \u201cimagine they are in a fast food restaurant\u201d, and order a meal from an online menu.\nParticipants were randomly assigned to see one of four menus (calories only, calories and number of minutes to walk to burn off that amount of calories, calories and number of miles to walk off that amount of calories, or no information).\nThe results from this study are significant; calories were significantly different based on menu type (p\u2009=\u20090.02), with the calories and exercise equivalents in mileage group ordering significantly less calories than the other three groups (p\u2009=\u20090.0007).\nAdditionally, 82% of their participants reported a preference for exercise equivalents on menu labeling.\nThis study shows a positive impact in using exercise equivalents to aid in the understanding of a calorie, and potential to help consumers order lower calorie food items.\nThese findings are consistent with an earlier study, which showed that exercise equivalents helped reduce purchases of sugar-sweetened beverages among low-income black adolescents.\nA study by Fitch et al. assessed the influence of calorie information versus exercise equivalents on food selection amongst adolescents and adults who ate at fast food restaurants regularly.\nThis study indicated that calorie labels were preferred to exercise equivalents overall (71%), and some cited the latter as demotivating; however, the Fitch study has limitations.\nThey tested the impression of exercise equivalents rather than their actual effect, examined exercise equivalents as an alternative to, rather than addition to calories, and had a predominantly white sample population, many of whom were not overweight or obese.\nThey also found that exercise equivalents had a more favorable impression among non-whites than whites, and, along with another study, among younger persons.\nThe current study aimed to test the actual effect of exercise equivalents on fast food point-of-purchase behaviors.\nResearch has shown that non-white, overweight and obese individuals are more likely to consume fast food and thus be at increased risk of negative health outcomes.\nFor this reason, we recruited young, predominantly non-white overweight and obese women for our study and presented them with exercise equivalents alongside calories at the point-of-choice.\nWe compared the effect of the exercise equivalents with the provision of simple caloric information or no information at all.\nAdditionally, we sought to evaluate the impact of restrained eating on point of purchase and consumption behaviors.\nThe current study was a pilot.\nWhile the researchers acknowledge the small sample size as a limitation, the goal of the study was to test a new design: the potential use of exercise equivalents for public health outreach, with hopes that other researchers can utilize for future studies on this topic.\nThinking of new ways to promote healthy behaviors at the point-of-purchase is important for nutrition researchers, educators, and policymakers.\nWe believe the novelty of our experimental design, with emphasis on using exercise equivalents as a nutrition intervention for at risk individuals, furthers thought on point-of-purchase interventions.\nMethods\nStudy overview\nA three-group repeated-measures experimental study was conducted to determine whether providing information about calories and exercise equivalents at the point-of-choice for a fast food meal would decrease calories ordered or consumed among overweight and obese 18-34-year-old women at a public university in southern Florida in 2009, and to investigate any correlation with consumption with prior dieting history, qualified as dietary restraint in this study.\nAll participants were asked to participate in two sessions during a two-week period.\nThe Florida International University Institutional Review Board approved this study.\nAll persons gave their informed consent prior to inclusion in the study.\nStudy participants\nA total of 62 overweight or obese female participants were recruited on a south Florida college campus.\nTelephone and in-person screening determined whether participants met the inclusion criteria: female, age 18-34 years old, BMI at least 25 and less than 40, as calculated from researcher-measured height and weight, ate fast food at least \u201coccasionally\u201d, and able to read and speak English.\nParticipants were also screened at this time for dietary restraint for randomization into the three study groups.\nPersons were excluded for dieting in the last three months; requiring a special diet such as vegetarian, kosher, or accommodating a food allergy or health condition; being pregnant or giving birth in last year; having a chronic disease such as heart disease or diabetes; having current self-reported depression, self-reported alcohol or drug abuse, or eating disorder; being a health major; not typically eating lunch; and participating in a previous food-related study.\nExclusion criteria were set to ensure participants were healthy and able to partake in a food-related study and to help avoid any bias gained from previous food-related studies.\nIn order to help further blind participants to the menu manipulation aspect of the study, participants were told that the purpose of the study was to \u201cbetter understand fast food meal choices\u201d.\nExperimental design\nParticipants attended two meal sessions, Lunch 1 and one week later Lunch 2.\nThe food choices were from a fast food restaurant located on the university campus.\nThe restaurant is part of a national chain specializing in hamburgers and French fries.\nThe foods were in their original portion-controlled wrappers or packaging, which allowed the researcher to easily record choices made by participants.\nThe study took place in a controlled setting within the university, at a private conference room in the University\u2019s Graham Center nearby the student union where the students normally eat.\nIncentives for the participants included $5 for completion of the screening questions, the two free lunches, and a $20 gift card at each lunch.\nAt the start of each Lunch, participants were given a menu.\nThe paper menus were in a similar format to menu boards at fast food restaurants.\nThe food items were those available for lunch at Burger King on the dates of the experiment.\nThe participants were able to choose entr\u00e9es (e.g. Hamburger, Whopper, TenderGrill, BK Veggie Burger or TenderGrill), a garden salad, side dishes (i.e., fries, onion rings), condiments (ketchup, mayonnaise, fat free ranch dressing, or honey mustard dressing) and a drink (i.e., water, Coca-Cola, diet Coca-Cola, or apple juice).\nThe observer recorded the quantity of the food ordered and eaten by using a digital food scale, weighing the remaining portions and using a measuring cup for the liquids.\nThe researcher ensured that all participants had finished eating and had left the study site prior to weighing and measuring left-over foods and drinks.\nCalories consumed were derived by taking food waste and weighing on a digital scale and calculating total calories eaten by the following formula: Total Calories For Food Item Chosen\u2013Food Waste = Calories Consumed.\nLunches were served from 11:30\u00a0a.m. until 3:00\u00a0p.m., and participants made appointments at 30-minute increments.\nAll participants were told in advance that they would not be able to leave the study site with any leftover food, to limit the possibility that participants would order more food than they intended to consume.\nDuring Lunch 1, participants were given a menu, similar in format to menu boards at fast food restaurants.\nParticipants were able to order any foods and beverages from the menu, which listed only names of items, no calories or exercise equivalents.\nThe experimental manipulation took place one week later at Lunch 2.\nAll participants were randomly assigned to one of three groups.\nEach group received different information on their menus: no information on calorie or exercise equivalents, calories only, or calories and exercise equivalents.\nColumn headers for the exercise equivalents and calories described the numbers (\u201cminutes to burn off food in walking\u201d, \u201ccalories\u201d), as did labels after each values.\nThe exercise equivalence of calories was calculated based on an intensity level of 3.3 METs for walking at the moderate pace of 3.0 mph on a firm surface, and a body weight of 160 pounds.\nData collection\nAt the research table, participants completed standardized questions on the following demographic information: age, marital status, education, income, race, religion, and whether the participant was a smoker.\nBody Mass Index was assessed by the investigators at the research table using a standardized height and weight measurement procedure as outlined in Third National Health and Nutrition Examination Survey (NHANES III) Anthropometric Procedures Manual.\nDietary restraint was determined using the TFEQ.\nScores on the TFEQ restraint sub-scale range from 0 to 21, with restrained eaters defined as those who have a score of 13 or above.\nParticipants were blocked by restraint in order to test whether unrestrained and restrained eaters responded differently, since restraint has been shown to influence food choice and the reading of nutrition labels, and were then randomly assigned to one of three study groups.\nStatistical analysis\nANOVA and chi-square tests were conducted to compare demographic information by study group.\nBoth the foods ordered and the foods consumed were analyzed.\nWithin each study group, a paired t-test was conducted to test for the change in calories ordered or consumed from Lunch 1 to Lunch 2.\nThe subsequent change by study group was calculated as the mean (plus or minus the standard error) of the changes from Lunch 1 to Lunch 2 for each of the group\u2019s individual members.\nProportionate change for calories order and calories consumption from Lunch 1 to Lunch 2 were calculated as the mean (plus or minus the standard error) of the proportionate changes of each of the group\u2019s individual members.\nThe t-tests were used to examine whether the proportionate changes are significant.\nAnalysis of covariance (ANCOVA) was conducted using General Linear Model in SPSS 17.0 statistical software (SPSS Inc., 2008) to test for differences between study groups in difference from Lunch 1 to Lunch 2 in calories ordered or consumed.\nThe effect size (partial eta squared) and observed power (using alpha\u2009=\u20090.05) are also calculated using SPSS.\nTwo general linear models were created, both controlling for age, BMI, and dietary restraint, and with study group as the fixed factor.\nFor model 1, the response variable was the difference from Lunch 1 to Lunch 2 in total calories ordered, and an additional covariate was calories ordered in Lunch 1 (control meal).\nFor model 2, the response variable was the difference from Lunch 1 to Lunch 2 in total calories consumed, and an additional covariate was calories consumed in Lunch 1.\nBoth models assumed that the response variables were continuous, residuals were normally distributed, and the subjects were independent.\nTotal number of items ordered was also compared by study group.\nResults\nStudy participants had a mean age of 21.9\u2009\u00b1\u20093.03\u00a0years and BMI of 28.4\u2009\u00b1\u20093.10.\nSeventy-three percent (n\u2009=\u200945) were black or Hispanic and 63% (n\u2009=\u200939) were unrestrained eaters (Table\u00a01).\nDemographic information was comparable across study groups (all p values\u2009>\u20090.05) (Table\u00a01).\nAll study groups decreased the number of calories ordered from Lunch 1 to Lunch 2 (Table\u00a02).\nWhile the current study is under-powered to ascertain statistical non-significance or significance, calorie only and calorie plus exercise equivalents ordered about 16% (206\u00a0kcal) and 14% (162\u00a0kcal) fewer calories from Lunch 1 to Lunch 2, respectively; whereas, the no information group ordered only 2% (25\u00a0kcal) fewer (Table\u00a02).\nIn all study groups, both restrained and unrestrained eaters had an average decrease in number of calories ordered from Lunch 1 to Lunch 2, with the exception of restrained eaters in the group receiving no calorie or exercise equivalent information at Lunch 2.\nThe greatest proportionate decrease in calories ordered was seen in restrained eaters in the calories-only group (24.7% decrease; p\u2009=\u20090.05).\nUnrestrained eaters in the calories and exercise equivalents group ordered an average of 275 fewer calories at Lunch 2 compared with Lunch 1, with a proportionate decrease of 14.0%, although this was not statistically significant (p\u2009=\u20090.24).\nUnrestrained eaters in the calories and exercise equivalents group had greater absolute and proportionate decreases in calories ordered from Lunch 1 to Lunch 2 than unrestrained eaters in the other two study groups.\nAdditional analyses examining number of items ordered revealed no significant differences by study group (data not shown).\nDuring the exit questionnaire, 57 participants (92%) said they believed that a combination of calories and exercise equivalents would influence the foods they ordered at a fast food restaurant.\nDiscussion\nConsumption of fast foods is common in the US.\nTo reduce negative effects and mitigate public health disparities in food environments, interventions may be especially critical in populations of persons who eat at fast food restaurants.\nCalorie information presented at point-of-purchase is a relatively new concept nationwide; however, research has shown mixed results at the point of purchase.\nThere is a potential for exercise equivalents as a supplemental guide to novice calorie counters and those unaware of the negative health implications in consuming fast food.\nWhile the current study was underpowered, we believe the novelty of the design of the experiment, the emphasis on utilizing exercise equivalents for a potential nutrition intervention for at risk individuals, will heighten awareness for future researchers on the need for further investigating point of purchase interventions, specifically exercise equivalents.\nThere have been several real-world studies that have shown an impact on calories ordered using sales data.\nFor example, using a randomization design, Roberto et al. reported that calorie information on restaurant menu did reduce the total amount of calories people ordered and consumed.\nAnother quasi-experimental design study examined the sales data before and after provision of point-of-selection nutrition labels found that the nutrition labels reduced average energy content of entr\u00e9e purchased without reducing overall sales.\nAdditionally, using data from Starbucks, Bollinger et al. found that mandatory calorie posting in chain restaurants resulted in 6% decrease in calories per transaction.\nDumanovsky et al. conducted a cross-sectional survey and assessed consumer purchases in 2007, before caloric information was mandated by chain restaurants, and again in 2009, after the menu labeling legislation was passed.\nAlthough they did not find an overall change in calories consumed, they did observe a significant decrease in the calories consumed at specific chain restaurants including McDonald\u2019s, Au Bon Pain and KFC.\nWith the rollout of the new law mandating fast food restaurants list caloric value for all menu items pending, understanding the potential implications is important.\nCalorie information at the point-of-purchase for restaurants has been required by law for chain restaurants in New York City since 2008, in California, Oregon and Maine since 2009 and has also been adopted in many other cities and counties.\nA recent study by Krieger et al. is one of the first to investigate the effect of the nationwide menu labeling bill.\nThis cross-sectional study surveyed fast food patrons both before the menu label regulation was implemented, and again 18\u00a0months later, post-regulation.\nInterestingly, they found a significant decrease in calories ordered in coffee and taco establishments, but not in burger and sandwich shops; and a decrease in calories ordered by women, but not men.\nThe effectiveness of nutrition labels on point-of-choice food purchasing has provided mixed results.\nSimilar to the present study, prior studies that have looked at point-of-purchase at fast food restaurants and the other at nutrition labels have also failed to show statistical significance.\nThis study provided a real-world setting was created to measure actual point-of-purchase behavior.\nThe unique strength of this study is that the study design provides a potential alternative or addition to the soon-to-be implemented national menu labeling law as a public health intervention.\nThis study illustrates a novel design to test the effectiveness of adding exercise equivalents to provide a frame of reference for consumers.\nUsing exercise equivalents on food labels and food served away from home could provide consumers with a context for the term, \u201ccalorie\u201d, and, thus, contribute to the understanding of the nutrition labels for better food choice and selection.\nPresentation of caloric information of fast food translated into exercise equivalents did not have a statistically significant impact on the food choices of overweight and obese women who were restrained eaters or unrestrained eaters.\nHowever, unrestrained eaters presented with calorie information and exercise equivalents combined had a larger decrease in calories ordered compared to those with calorie information only and for those with no information.\nThe impact of calorie information with exercise equivalents on unrestrained eaters should be further examined as unrestrained eaters generally do not deliberately attempt to limit their food intake.\nThere are several limitations in this exploratory study.\nThe small study sample size is a major limitation of this study.\nAdditionally, the study was limited to female college students thus limiting its generalizability.\nAnother limitation is that individuals were getting food at no cost, which might have influenced the total number of food items, and hence amount of calories chosen.\nThe average calories for the foods chosen for Lunch 1 and Lunch 2 were 1215.16 and 1087.50, respectively.\nDumanovsky and colleagues (2009) established baseline data on mean calorie intake at Burger King of 926.2.\nParticipants in the current study chose approximately 225 more calories per meal on average than participants in Dumanovsky study, perhaps because the food was free.\nThis study did not collect pre- and post-intervention food diaries.\nIt is possible that participants who chose lower calorie foods during the intervention may have increased their intake later in the day to compensate.\nAdditionally, there is a potential limitation in using exercise equivalents, or calories alone, to promote lower calorie food choices as opposed to nutritionally dense options.\nLower calorie foods do not necessarily make a food nutritionally \u201cbetter\u201d than another.\nDespite this, although nutritionally and calorically dense foods such as tree nuts, avocado, and fatty fish are touted as an important part of a healthful diet, they are not commonly offered at fast food restaurants.\nAs such, exercise equivalents and calories listed can potentially provide meaningful reference points for fast-food patrons.\nConclusions\nThe current study presented an intervention designed to improve the effectiveness of calorie information on point-of-choice.\nThe concept of menu labeling, exercise equivalents and other point-of-purchase messages are a potentially useful way to reach consumers at the point of their food decision.\nThis research, combined with previous studies, suggest that in addition to calorie labeling information, there is a need for further research of point-of-purchase interventions to find the most effective ways of presenting the menu labels for general public.\nAlthough this study was not powered to see statistical differences, the concept that behaviors may differ based on calorie and exercise information should be further explored in a larger study.\n\n\nParticipant characteristics by study group, in a group of overweight or obese women\n\n | Study group for menu type at Lunch 2 (experiment meal)\nNo calorie or exercise equivalent information | Calories only | Calories and exercise equivalents | p1\nTotal | N | % | N | % | N | % | \n22 | 100% | 20 | 100% | 20 | 100% | \nAge (years; mean, SD) | 21.9\u2009\u00b1\u20093.5 |  | 21.6\u2009\u00b1\u20092.3 |  | 22.2\u2009\u00b1\u20093.2 |  | 0.82\nWeight (pounds; mean, SD) | 167.9\u2009\u00b1\u200926.5 |  | 171.2\u2009\u00b1\u200926.6 |  | 165.6\u2009\u00b1\u200925.8 |  | 0.79\nBMI (kg/cm2; mean, SD) | 27.9\u2009\u00b1\u20093.1 |  | 28.7\u2009\u00b1\u20093.0 |  | 28.7\u2009\u00b1\u20093.3 |  | 0.64\nRace/Ethnicity (N) |  |  |  |  |  |  | 0.90\nHispanic/Latino | 8 | 36% | 10 | 50% | 10 | 50% | \nBlack/African American | 7 | 32% | 5 | 25% | 5 | 25% | \nOther | 7 | 32% | 5 | 25% | 5 | 25% | \nDietary restraint2 |  |  |  |  |  |  | 0.66\nRestrained | 7 | 32% | 7 | 35% | 9 | 45% | \nUnrestrained | 15 | 68% | 13 | 65% | 11 | 55% | \n\n\n1Using ANOVA for age, weight and BMI, and Chi-square test for dietary restraint and race/ethnicity.\n\n2Classified using restraint subscale of TFEQ; score of <13 indicates restrained eater, >\u2009=\u200913 indicates unrestrained eater.\n\n\nCalories ordered and consumed (mean \u00b1 SE) by meal and study groups\n\n | Model 1: Calories ordered (mean \u00b1 SE) by meal1\nStudy group | n | Lunch 13 | Lunch 2 | Difference | Proportionate change (%)4 | p | Cohen\u2019s d\nNo calorie or exercise equivalent information | 22 | 1,201.4\u2009\u00b1\u2009100.0 | 1,176.1\u2009\u00b1\u200999.5 | -25.2\u2009\u00b1\u200995.2 | 9.3\u2009\u00b1\u200911.6 | 0.43 | 0.17\nCalories only | 20 | 1,282.8\u2009\u00b1\u200989.7 | 1,077.0\u2009\u00b1\u2009114.0 | -205.8\u2009\u00b1\u2009110.6 | -14.4\u2009\u00b1\u20097.3 | 0.06 | 0.45\nCalories and exercise equivalents | 20 | 1,162.8\u2009\u00b1\u2009141.1 | 1,000.5\u2009\u00b1\u200998.2 | -162.3\u2009\u00b1\u2009132.5 | 1.6\u2009\u00b1\u200913.3 | 0.90 | 0.02\n | Model 2: Calories consumed (mean \u00b1 SE) by meal2\nStudy Group | n | Lunch 13 | Lunch 2 | Difference | Proportionate change (%)5 | p | Cohen\u2019s d\nNo calorie or exercise equivalent information | 22 | 986.6\u2009\u00b1\u200984.1 | 995.4\u2009\u00b1\u200991.5 | 8.8\u2009\u00b1\u200983.9 | 10.7\u2009\u00b1\u200911.6 | 0.36 | 0.2\nCalories only | 20 | 1,059.6\u2009\u00b1\u200972.7 | 898.8\u2009\u00b1\u200987.6 | -160.7\u2009\u00b1\u2009106.3 | -9.3\u2009\u00b1\u200910.7 | 0.39 | 0.2\nCalories and exercise equivalents | 20 | 840.9\u2009\u00b1\u200988.6 | 841.3\u2009\u00b1\u200982.0 | 0.5\u2009\u00b1\u200976.9 | 11.9\u2009\u00b1\u200913.7 | 0.40 | 0.2\n\n\n1ANCOVA p-value\u2009=\u20090.43, controlling for age, BMI, race, dietary restraint, and calories ordered at Lunch 1; partial eta squared\u2009=\u20090.03, observed power\u2009=\u20090.19.\n\n2ANCOVA p-value\u2009=\u20090.31, controlling for age, BMI, race, dietary restraint, and calories consumed at Lunch 1; Partial eta squared\u2009=\u20090.04, observed power\u2009=\u20090.25.\n\n3All persons received menus with no calorie or exercise equivalent information at Lunch 1.\n\n4Overall mean \u00b1 SE of individual proportionate changes, each calculated as (calories ordered in Lunch 2-calories ordered in Lunch 1)/calories ordered in Lunch 1.\n\n5Overall mean \u00b1 SE of individual proportionate changes, each calculated as (calories consumed in Lunch 2-calories consumed in Lunch 1)/calories consumed in Lunch 1.",
        "label": "unclear",
        "id": "task4_RLD_test_127"
    },
    {
        "paper_doi": "10.1186/1471-2334-6-16",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Randomized (computer-generated) to PQ or bulaquine (1:2 ratio) after primary treatment with quinine + doxycycline\n\n\nParticipants: 93 participants in IndiaInclusion criteria> 16 years.Male.Uncomplicated P. falciparum only.> 55 P. falciparum gametocytes/mL on admission.Exclusion criteriaAntimalarial treatment in previous two weeks.Allergy to trial drug.G6PD deficient.\n\n\nInterventions: All patients: quinine days 1 to 7: 30 mg/kg/day (10 mg/kg/day three times per day) + 100 mg doxycyclineRandomization and treatment on day 4PQ.Bulaquine.\n\n\nOutcomes: Gametocyte prevalence, density and viability on days 1, 4, 15, 22, and 29Adverse events\n\n\nNotes: Gametocyte viability assessed by Shute's technique (ex flagellation\n\n",
        "objective": "To assess the effects of single dose or short\u2010course PQ (or an alternative 8AQ) alongside treatment for people with P. falciparum malaria.",
        "full_paper": "Background\nThe WHO recommends that adults with uncomplicated P. falciparum successfully treated with a blood schizonticide receive a single dose of primaquine (PQ) 45 mg as a gametocytocidal agent.\nAn earlier pilot study suggested that 75 mg of bulaquine (BQ), of which PQ is a major metabolite, may be a useful alternate to PQ.\nMethods\nIn a randomized, partial blind study, 90 hospitalized adults with Plasmodium falciparum malaria that was blood schizonticide-responsive and a gametocytemia of > 55/\u03bcl within 3 days of diagnosis were randomized to receive single doses of either PQ 45 mg or BQ 75 mg on day 4.\nWe assessed gametocytemia on days 8, 15, 22 and 29 and gametocyte viability as determined by exflagellation (2\u00b0 end point) on day 8.\nResults\nOn day 8, 20/31 (65%) primaquine recipients versus 19/59 (32%) bulaquine recipients showed persistence of gametocytes (P = 0.002).\nAt day 15 and beyond, all patients were gametocyte free.\nOn day 8, 16/31 PQ and 7/59 BQ volunteers showed gametocyte viability (p = 0.000065).\nConclusion\nBQ is a safe, useful alternate to PQ as a Plasmodium falciparum gametocytocidal agent and may clear gametocytemia faster than PQ.\nIntroduction\nMalaria remains the most important parasitic infection with 300\u2013500 million people affected yearly and 1.5\u20132.7 million deaths each year.\nWorld over, malaria control has focused on pharmacological intervention, vector control, curtailing irrational and indiscriminate use of antimalarials, and the development of vaccines.\nOf these strategies, pharmacological intervention remains the most effective way to combat malaria\n8 aminoquinolines like primaquine are unique antimalarials in that they exhibit activity against multiple life cycle stages of Plasmodia that infect humans.\nPrimaquine 45 mg as a single dose is recommended by the World Health Organization (WHO) and the National Antimalarial Programme (NAMP) of India for its gametocytocidal activity in P. falciparum.\nIn 1998, Gogtay et al found that the efficacy of PQ 45 mg as a falciparum gametocytocidal agent in patients sensitive to chloroquine in India to be approximately 77% at Day 29 follow up.\nDuring the past several years, attempts have been made to produce primaquine analogs with improved anti-malarial activity and lower toxicity.\nBulaquine, formerly called CDRI 80/53, is metabolized to PQ and differs from PQ only by the 2,4 dihydrofuran group present in the basic side chain anchored onto the quinoline nucleus in the 8 position.\nBulaquine is currently licensed only for use in India for the radical cure of vivax malaria dosed at 25 mg/day for 5 days, but not as a gametocytocidal agent IN Plasmodium falciparum.\nThe results of a pilot study by our group assessing the P. falciparum gametocytocidal effect of a single dose of bulaquine 75 mg in India suggested it may be more effective than PQ 45 mg.\nHere, in a larger population of adults with P falciparum malaria successfully treated with blood schizonticides, we compared the gametocytodical activity of BQ and PQ.\nVolunteers and methods\nProtocol\nThe protocol was approved by the institutional ethics committee and the Drugs Controller General of India.\nThe study was conducted between January 2002 and April 2004.\nEnrollment and procedures\nPatients who were at least 16 years old with uncomplicated Plasmodium falciparum infection, provided written informed consent, and had a gametocyte count > 55/\u03bcl within 72 hours of diagnosis, regardless of asexual parasite counts, were eligible for enrollment.\nThe minimal gametocyte count was chosen based on infectivity to mosquitoes.\nPatients who were pregnant or lactating, had received antimalarial treatment in the previous 2 weeks, had co-infection with Plasmodium vivax, claimed an allergy to primaquine or bulaquine, or were G6PD deficient were excluded.\nOn admission, patients were initially assessed by thick and thin blood films stained using the Jaswant Singh and Bhattacharji (JSB) field stain.\nSubsequently, Giemsa stained blood smears were used to determine the number of asexual and sexual parasites/\u03bcl, assuming a white blood cell count of 8000/\u03bcl.\nEnrolled patients were admitted to hospital on Day 1 and treated under observation with quinine orally 10 mg/kg/day thrice daily for a total of 7 days and doxycycline 100 mg once daily for 7 days.\nAt day 4, consecutive patients were randomly allocated in a 1:2 fashion to receive an observed single dose of either PQ 45 mg or bulaquine 75 mg, based on a computer generated randomization code.\nUnequal allocation was used because of earlier studies suggesting the superiority of bulaquine.\nThe test articles were administered on day 4 because the incidence of nausea and vomiting is higher in the first few days of schizonticidal therapy and this was given regardless of parasite clearance.\non day 8, all patients were assessed for gametocytemia, discharged, and asked to follow up on days 15, 22, and 29 for further safety and parasitologic checks.\nGiemsa stained malaria blood smears during hospitalization were done twice a day for the first 72 hours and once a day thereafter until discharge and on the follow-up days.\nThe slide readers were blinded to the treatment.\nOutcomes\nEfficacy was assessed by gametocytemia (primary end point) and gametocyte viability (secondary end point) on admission and all follow up days.\nThe latter was assessed by the modified Shute's technique.\nThis technique depicts exflagellating microgametes in blood films that have been kept moist at 21\u201325\u00b0C for 1 hour with complete RPMI medium and AB positive serum and then Giemsa stained.\nOne or more exflagellating microgametes was considered a positive test.\nThese end points were identical to the previous study.\nSafety was monitored by routine clinical hematological and biochemical laboratories and an electrocardiogram on days 1 and 8.\nAdverse event recording was focused only on nausea, vomiting, and epigastric distress and were recorded only if not reported on Day 1 or if a symptom worsened after Day 1.\nSample size and statistical analysis\nThe estimated sample size was calculated using Casagrande's method based on a previous study by Gogtay et al comparing the two drugs.\nAssuming a 30% difference in efficacy on Day 8, at 5% significance and 90% power, a sample size of 28 patients and 56 patients are required in the primaquine and bulaquine group, respectively, to demonstrate the superiority of bulaquine.\nP values \u2264 0.5 were considered significant.\nResults\nA total of 93 male patients were enrolled.\nWomen with malaria are not inclined to get admitted, especially because of hardships related to hospitalization and supervised drug administration.\nThe age of the patients ranged from 16\u201372 years: 31.47 \u00b1 11.62 years).\nThere were three drop outs, two in the Bulaquine arm and one in primaquine arm.\nThese patients did not return for any follow up visit after discharge from the hospital and were omitted from analysis.\nAt admission, gametocytaemia between the 2 groups was similar.\nAt day 8, 20/31 (65%) PQ recipients and 19/59 (32%) BQ recipients had gametocytes on blood smear (p = 0.002).\nAt days 15, 22, and 29, all patients in both treatment groups were free of gametocytes.\nAt day 8, 16/31 (52%) PQ recipients and 7/59 (12%) BQ recipients had viable gametocytes on exflagellation testing (Table)(p = 0.000065).\nAll patients with viable gametocytes had smear positive gametocytemia.\nThere were no important clinical hematology or biochemical laboratory values, and all electrocardiograms were within normal limits.\nDiscussion\nA single dose of 45 mg primaquine is given along with or after schizonticidal therapy in areas where malaria is endemic as a transmission blocking strategy and currently is the only option available for this indication.\nThe present study carried out in 91 cases of uncomplicated Plasmodium falciparum malaria assessed the efficacy of bulaquine, the parent compound of primaquine.\ngiven as a single dose of 75 mg for its gametocytocidal effect.\nOn day 8 of therapy, fewer patients with bulaquine had gametocytes as compared to those who received 45 mg primaquine.\nThis suggests superior efficacy of bulaquine or its ability to clear gametocytes more rapidly, as all patients were free of gametocytes by day 15 and beyond.\nThis clearance of gametocytes completely by Day 15 in both groups in the present study contrasts with our previous study with 45 mg primaquine where persistent gametocytemia was seen in 23% patients responsive to chloroquine up to Day 29.\nThis may in part be due to use of different schizonticidal agents in the two studies, with chloroquine being used in the former and quinine in the latter, with quinine leading to greater clearance of asexual parasites.\nBased on the results of the present study, bulaquine in a single dose of 75 mg may represent yet another treatment option for gametocytocidal effect in addition to primaquine and the current licensing of the drug in the country could be changed to include gametocytocidal effect apart from anti-relapse effect.\nWhether increasing the dose of primaquine from 45 mg to 60 mg will improve efficacy needs to be addressed in future studies.\nIn this study, the Shute technique, which measures the ability of the male gametocyte to exflagellate was used as a surrogate marker for assessing transmission blocking.\nAssessment of true gametocytocidal efficacy of any drug will depend upon demonstration of the ability to block transmission to mosquitoes.\nThis in turn can be assessed by only checking for the presence of the oocyst and ookinete in the mosquito midgut, not done in the study.\nIn the first study where we first reported the declining efficacy of primaquine as a gametocytocidal agent, chloroquine versus coartemether was studied, since chloroquine then was (and still remains) the first line therapy for the country.\nThe results of the study showed a high degree of chloroquine resistance (> 50%), and we shifted to using quinine as first line for uncomplicated falciparum malaria in the hospital, since there are only isolated reports of quinine resistance in the country.\nDrugs that can block the spread of malarial parasites by killing or preventing the transmission or maturation of gametocytes, represent important tools in malaria control.\nSupputtamongkol et al compared the efficacies of mefloquine-artesunate and mefloquine-primaquine on the subsequent development of gametocytemias.\nThe latter combination was not found to be effective in either clearance of existing gametocytemias or the prevention of new gametocytemias.\nPrimaquine has an extremely short half life, and as such may not be completely able to eliminate gametocytes and needs to be used with an effective schizonticidal agent.\nPersistence of gametocytemia is one of the predictors of treatment failure and use of an effective anti-malarial drug to eradicate asexual forms will still remain the most effective means to prevent gametocytemias.\n\nGametocytaemia of patients at different days of follow-up in the two groups\nDays of follow-up | Primaquine n = 31. Gametocyte density/\u03bcl(mean \u00b1 SD) | Bulaquine n = 59. Gametocyte density/\u03bcl(mean \u00b1 SD)\n1 | 80-6040(1342 \u00b1 182.80) | 80-4160(1064.2 \u00b1 182.80)\n4 | 120-8000(1494 \u00b1 385.04) | 80-3260(1003.7 \u00b1 118.35)\n8 | 80-2120(543 \u00b1 128.75) | 32-1820(161.01 \u00b1 32.76)\n15,22,29 | Nil | Nil\n\n\nResults of Efficacy.\nDays of follow -up | No of patients positive for gametocytes | Patients sample positive for exflagellation\n | Pts treated with PQN = 31 | Pts treated with BQN = 59 | Pts treated with PQN = 31 | Pts treated with BQN = 59\n\n1 | 31/31100% | 59/59100% | 31/31100% | 59/59100%\n4 | 31/31100% | 59/59100% | 31/31100% | 59/59100%\n8 | 20/31 *65% | 19/59 *32% | 16/31 *52% | 7/59 *12%\n15 | Nil | Nil | Nil | Nil\n22 | Nil | Nil | Nil | Nil\n29 | Nil | Nil | Nil | Nil\n\n* statistically significant P < 0.05\nAcknowledgements: This study was carried out as a project under the Center for Advanced Research in Clinical Pharmacology, and was funded by the Indian Council of Medical Research, New Delhi. We thank Nicholas Piramal India Ltd for the supply of bulaquine capsules.",
        "label": "unclear",
        "id": "task4_RLD_test_915"
    },
    {
        "paper_doi": "10.1371/journal.pone.0011880",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: An open-label (non-inferiority) RCTFollow-up: Particiopants were managed as outpatients unless local practice dictated otherwise (some centres used hospital stays of between 3 and 28 days). Outpatients were asked to return on days 1, 2, 3, 7, 14, 21, 28, 35, 42, 49, 56, and 63, and any time they felt unwell. Blood smears were performed at each visit.Adverse event monitoring: Blood and urine samples were taken for analysis on days 0, 28, 63 (if abnormal on day 28) and on the day of any recurrent parasitaemia. Twelve-lead ECGs were performed at days 0, 2, 7, 28 (if abnormal on day 7), 63 and on the day of any recurrent parasitaemia.\n\n\nParticipants: Number of participants: 1150Inclusion criteria: Age 3 months to 65 years (>=18 years in India), P. falciparum mono-infection (80 to 200,000 parasites/uL) or mixed infection, weight >= 5 kg, fever (>= 37.5 degC) or history of fever, informed consent.Exclusion criteria: Severe malaria, treatment with MQ in the 60 days before screening, treatment with DHA-P in the 3 months before screening, > 4% parasitised red blood cells, pregnancy or lactation.\n\n\nInterventions: 1. DHA-P, fixed dose combination, adult tablets 40 mg/320 mg, child tablets 20 mg/160 mg (Eurartesim(r): Sigma Tau)One dose daily for 3 days2.25 mg/kg DHA and 18 mg/kg piperaquine per doseDose rounded up to the nearest half tablet2. Artesunate plus mefloquine, loose dose combination, AS 50mg tablets, MQ 250 mg tablets (Mepha Ltd)AS 4mg/kg once daily for 3 daysMQ none on day 0, then 15 mg/kg once on day 1 and 10 mg/kg once on day 2All doses supervised.\n\n\nOutcomes: Cure rate at days 28, 42, and 63, PCR corrected and uncorrectedMean change in Hb day 0 to day 63Gametocyte carriagePerson-gametocyte-weeksAdverse eventsNot included in this review:Fever clearanceParasite clearance\n\n\nNotes: Country: Thailand (six sites), Laos (two centres), and India (three centres).Setting: Hospitals and research units.Transmission: Varied across trial regions. Trial regions in Thailand had unstable, low and seasonal malaria transmission; trial regions in Laos had seasonal transmission with a peak just after the heavy rainy months of July to August; trial regions in India included areas with perennial transmission, perennial transmission with a seasonal peak from June to September, and transmission active in post monsoon months.Resistance: All sites had notable CQ resistance (estimates of 28 day treatment failure at the Indian sites ranged from 32% to 67% between 2002 and 2007). The Thai sites also had multi-drug resistant P. falciparum.Dates: Jun 2005 to Feb 2007.Funding: Medicines for Malaria Venture, Sigma Tau, and Oxford University\n\n",
        "objective": "To evaluate the effectiveness and safety of DHA\u2010P compared to other ACTs for treating uncomplicated P. falciparum malaria in adults and children.",
        "full_paper": "Background\nThe artemisinin-based combination treatment (ACT) of dihydroartemisinin (DHA) and piperaquine (PQP) is a promising novel anti-malarial drug effective against multi-drug resistant falciparum malaria.\nThe aim of this study was to show non-inferiority of DHA/PQP vs. artesunate-mefloquine (AS+MQ) in Asia.\nMethods and Findings\nThis was an open-label, randomised, non-inferiority, 63-day follow-up study conducted in Thailand, Laos and India.\nPatients aged 3 months to 65 years with Plasmodium falciparum mono-infection or mixed infection were randomised with an allocation ratio of 2\u22361 to a fixed-dose DHA/PQP combination tablet (adults: 40 mg/160 mg; children: 20 mg/320 mg; n\u200a=\u200a769) or loose combination of AS+MQ (AS: 50 mg, MQ: 250 mg; n\u200a=\u200a381).\nThe cumulative doses of study treatment over the 3 days were of about 6.75 mg/kg of DHA and 54 mg/kg of PQP and about 12 mg/kg of AS and 25 mg/kg of MQ.\nDoses were rounded up to the nearest half tablet.\nThe primary endpoint was day-63 polymerase chain reaction (PCR) genotype-corrected cure rate.\nResults were 87.9% for DHA/PQP and 86.6% for AS+MQ in the intention-to-treat (ITT; 97.5% one-sided confidence interval, CI: >\u22122.87%), and 98.7% and 97.0%, respectively, in the per protocol population (97.5% CI: >\u22120.39%).\nNo country effect was observed.\nKaplan-Meier estimates of proportions of patients with new infections on day 63 (secondary endpoint) were significantly lower for DHA/PQP than AS+MQ: 22.7% versus 30.3% (p\u200a=\u200a0.0042; ITT).\nOverall gametocyte prevalence (days 7 to 63; secondary endpoint), measured as person-gametocyte-weeks, was significantly higher for DHA/PQP than AS+MQ (10.15% versus 4.88%; p\u200a=\u200a0.003; ITT).\nFifteen serious adverse events were reported, 12 (1.6%) in DHA/PQP and three (0.8%) in AS+MQ, among which six (0.8%) were considered related to DHA/PQP and three (0.8%) to AS+MQ.\nConclusions\nDHA/PQP was a highly efficacious drug for P. falciparum malaria in areas where multidrug parasites are prevalent.\nThe DHA/PQP combination can play an important role in the first-line treatment of uncomplicated falciparum malaria.\nTrial Registration\nControlled-Trials.com ISRCTN81306618\nIntroduction\nThe bisquinoline piperaquine (PQP) was first synthesised in the 1960s independently by teams in China and France.\nIn 1978, because of its greater potency and tolerability relative to chloroquine, PQP replaced chloroquine in the Chinese National Malaria Control Programme.\nSuch was the success of the programme, that the following decade saw the use of PQP decrease because of the decrease in patients suffering from malaria.\nIn 1990, Chinese scientists working to develop alternative anti-malarial treatments that might offer higher cure rates with better tolerability profiles included PQP phosphate in an artemisinin-based combination treatment (ACT), known as China-Vietnam 4 (CV4\u00ae), that also included dihydroartemisinin (DHA), trimethoprim and primaquine phosphate.\nDuring the development programme, doses of each component of the combination were modified resulting in a new formulation, CV8\u00ae, which was tested in Vietnam and administered in the Vietnamese Malaria Control Programme in 2000.\nThis programme was widely successful despite the presence of chloroquine resistance in Vietnam.\nHowever, concerns about the association of red cell haemolysis with primaquine in populations with glucose-6-phosphate dehydrogenase deficiency and the questionable anti-malarial potency of trimethoprim led to the removal of these two components of the drug combination in CV8\u00ae.\nThe remaining two components of the regimen, the artemisinin DHA and PQP (Artekin\u00ae), provide a combination that is relatively inexpensive and has been shown to be effective both in curing malaria and preventing re-infection.\nSince their introduction in the 1990s, ACTs have been found to be highly effective treatments for malaria.\nAcross Asia, Africa and South America, clinical and parasitological responses to the combination of DHA and PQP have generally exceeded the 95% value that the World Health Organization (WHO) recommend for anti-malarial treatments.\nHowever, in recent years there have been suggestions that the overall efficacy of ACTs in Thailand and Cambodia may be declining.\nEvidence of falling efficacy has been characterised by reductions in the proportions of patients clearing their parasitaemia by day 2 of treatment in the Thailand-Myanmar border region, and by reductions in polymerase chain reaction (PCR)-corrected cure rates at 28- and 42-day follow-up assessments in the Cambodia-Thailand border region.\nRelatively low cure rates were also reported from Papua New Guinea and, most recently, there has been evidence of reduced artesunate susceptibility in Western Cambodia.\nHistorically, anti-malarial drug resistance has spread westwards from Cambodia through South Asia to Africa.\nConsequently, the recent reports of potential resistance to artemisinins alone and ACTs are of great concern.\nIn order to assess the safety and efficacy of the treatment combination of DHA and PQP in Asia, we conducted a randomised trial in Thailand, Laos and India comparing DHA/PQP with another ACT, artesunate (AS) plus mefloquine (MQ).\nAll study sites were located in areas of notable chloroquine resistance.\nDihydroartemisinin plus PQP was administered as a single tablet (DHA/PQP) and AS plus MQ were administered as separate loose tablets (AS+MQ).\nThis study had the largest sample size to date of any study assessing DHA/PQP in South East Asia.\nMethods\nThe protocol and amendments for this trial and supporting CONSORT checklist are available as supporting information; see Protocol S1 to S8 and Checklist S1.\nStudy regions\nThis study was conducted in six centres in Thailand (Hospital for Tropical Diseases, Faculty of Tropical Medicine, Mahidol University, Bangkok; Suanphung Hospital, Ratchaburi; Proppra Hospital, Proppra District, Tak; Shoklo Malaria Research Unit, Mae Sod District, Tak; Mae Sod Hospital, Muang District, Tak; and Mae Ramat Hospital, Mae Ramat District, Tak), two centres in Laos (Phalanxay District Hospital, Savannakhet Province; Xepon District Hospital, Savannakhet Province) and three centres in India (Down Town Hospital, Guwahati, Assam; Goa Medical College and Hospital, Goa; and Wenlock District Government Hospital, Mangalore) over two malaria seasons, with patients screened from June 2005 to February 2007.\nThe study regions provided a range of transmission conditions and standard treatments.\nStudy regions in Thailand were located in malarious forest on the Thai-Myanmar border in regions of unstable, low and seasonal malaria transmission that used AS+MQ as a first-line treatment.\nNearly all infections with Plasmodium falciparum and P. vivax in the region are symptomatic and P. falciparum is multi-drug resistant, with high levels of resistance to chloroquine.\nMalaria transmission in the Phalanxay and Xepon districts in Laos was seasonal with a peak just after the heavy rainy months of July to August.\nPrevious studies in the Phalanxay district showed high levels of chloroquine resistance.\nThe first-line treatment was artemether/lumefantrine.\nIn India, study regions included areas of perennial transmission (Assam), perennial transmission with a seasonal peak from June to September (Goa), and transmission active in post-monsoon months (Mangalore).\nChloroquine resistance was present at all study sites, with site records showing 28-day treatment failure rates of 32% in Assam (in 2002), 54% in Goa (in 2007) and 67% in Mangalore City (in 2005).\nArtemisinin-based combination treatments are used as standard treatments in Goa, Mangalore and Assam (specifically AS plus sulphadoxine/pyrimethamine).\nPatients\nMales and females aged from 3 months to 65 years weighing at least 5 kg with fever or history of fever (\u226537.5\u00b0C) and microscopically confirmed mono-infection with P. falciparum (asexual forms parasitaemia \u226580 per \u00b5L \u2264200,000 per \u00b5L) or mixed infection were eligible for the study.\nLocal regulations were followed in India that required all recruited patients to be aged 18 years and over.\nKey exclusion criteria were: severe malaria, treatment with MQ in the 60 days prior to screening, treatment with DHA/PQP in the 3 months prior to screening and >4% parasitised red blood cells.\nPregnant or lactating women were not eligible for the study.\nStudy design\nThis was a randomised, Phase III, open-label study with two treatment arms: DHA/PQP and the active comparator AS+MQ.\nAt the time the study was designed, there was no single first line treatment in the countries involved in the study.\nHowever, AS+MQ was first line treatment in Thailand, was well characterized and was widely used in all the considered countries.\nThis made AS+MQ suitable as the control treatment against which the efficacy and safety of DHA/PQP would be tested.\nAs the study was not blinded, to limit bias, the following procedures were put in place: 1) Randomisation was conducted under blinded conditions: the blind to the investigator and patient in the randomisation process was maintained by the use of sealed envelopes.\n2) Evaluation of the PCR test results was blinded (centralised at the Prince Leopold Institute of Tropical Medicine, Antwerp, Belgium, with quality control at Shoklo Malaria Research Unit, Mae Sod, Thailand).\n3) The chairman and the statistician of the independent Data Monitoring Committee reviewed the most relevant safety data and participated in the final blinded data review meeting where all decisions about assessment of the primary outcome and patient allocation to the pre-defined populations were made under blinded conditions.\nThe randomisation list was generated by an external contract research organisation (MDS Pharma Services) using the plan procedure of SAS (Cary, NC, USA).\nPatients were allocated to receive either DHA/PQP or AS+MQ following a 2\u22361 randomisation schedule ratio.\nThe unbalanced ratio was chosen to increase the chance of detecting rare adverse reactions and to provide more precise estimates of cure rates in the DHA/PQP arm.\nDihydroartemisinin/PQP\n(Eurartesim\u2122, Sigma-Tau, Italy) was given once daily, on days 0, 1 and 2 of the study, at the standard dosage of 2.25 mg/kg and 18 mg/kg per dose of DHA and PQP, respectively, rounded up to the nearest half tablet.\nTo facilitate the correct dosing of DHA/PQP, two formulations were used (DHA 20 mg + PQP 160 mg and DHA 40 mg + PQP 320 mg).\nOver the 3 days, the cumulative doses were of about 6.75 mg/kg of DHA and 54 mg/kg of PQP.\nArtesunate and MQ (Mepha Ltd, Switzerland) were administered as separate tablets containing AS 50 mg and MQ 250 mg; AS was administered at 24-h intervals on days 0, 1 and 2 with a daily dose of 4 mg/kg and MQ was administered at 15 mg/kg on day 1 and 10 mg/kg on day 2 at 24-h intervals, but was not administered on day 0.\nAs part of their routine treatment, febrile patients who attended the study centres underwent a thick blood smear test for malaria before any anti-malarial treatment was administered.\nIf the smear was found to be positive for P. falciparum, patients were told about the study and offered the opportunity to participate.\nEligible patients who agreed to participate were given detailed explanations of the trial from study staff, including the information that they would be given one of two different anti-malaria treatments on a random basis.\nParticipating patients all provided written informed consent.\nPatients who declined to participate were provided with treatment for uncomplicated P. falciparum malaria, standard for the area in which they lived.\nParticipating patients were managed as outpatients and doses were given under medical supervision, although at some centres patients could be hospitalised according to local practice, which ranged from 3 days at Mae Ramat Hospital, Thailand and at centres in Laos and India to 28 days at Mahidol University, Thailand.\nIf not in hospital, patients were encouraged to return to the study centre for assessments on days 1, 2, 3, 7, 14, 21, 28, 35, 42, 49, 56 and 63, and they were also told that they could make unscheduled visits on any day on which they felt unwell.\nPatients were followed-up for 63 days.\nIn accordance with WHO (WHO 2003 and 2009) and standard practice in malaria clinical trials, the primary endpoint of the study was the PCR-corrected cure rate, based on the adequate clinical and parasitological response (ACPR), which was defined as the absence of parasitaemia irrespective of the patient's body temperature, with the patient not meeting any of the pre-defined criteria of early treatment failure or late clinical or parasitological failure (see below).\nIn more detail, the primary endpoint was defined using two methods.\nThe first method was based purely on the standard definitions of early/late clinical and parasitological failure as defined by the WHO.\nThis endpoint is referred to as the true treatment failure and is defined as the sum of early treatment failures and late recrudescences, which included late treatment failures that were assessed as recrudescences according to PCR analysis (100 minus the true treatment failure rate provides the WHO cure rate).\nThe second method, agreed with the Data Monitoring and the Clinical Development Committees, was based on a pre-defined procedure that expanded the WHO definitions with a set of rules allowing the evaluation of each individually randomised patient (for definitions see Table 1).\nThis approach was considered to be primary because it was deemed to be in line with the requirements of the most stringent regulatory authorities.\nAll cases not strictly matching the WHO definitions and/or the described procedure were reviewed individually at the final blinded data review meeting.\nDay 63 was chosen as the primary time-point because of the long half lives of piperaquine and mefloquine and because the risk of new infection is lower than in other parts of the world.\nSecondary endpoints included PCR-corrected adequate clinical and parasitological response on days 28 and 42, PCR-uncorrected adequate clinical and parasitological response (PCR-uncorrected), proportion of patients with early and late treatment failure, proportion of aparasitaemic patients, proportion of afebrile patients, number of new infections, gametocyte carriage and the safety profile of the two treatments including adverse events and 12-lead electrocardiograph parameters (QT interval corrected according to the methods of Bazett, QTc(B), and Fridericia, QTc(F)).\nProcedures\nThe presence of P. falciparum was verified at all study visits including screening, using thick and thin Giemsa-stained blood smears obtained from the patient to calculate parasite density, which was initially calculated by counting the number of asexual parasites per 500 leukocytes in the thick blood film, based on an assumed white cell count of 8,000 cells per \u00b5L.\nBlood smears were obtained from a finger prick applied directly to a microscope slide to create the blood film.\nParasite density per \u00b5L was calculated as: (number of parasites counted \u00d78,000)/(number of leukocytes counted).\nFor samples with higher levels of parasitaemia (>3 parasites/1000 red blood cells), parasite density was calculated from the thin film per 1000 red blood cells as: (number of P. falciparum trophozoites per 1000 red blood cells x haematocrit \u00d7125.6).\nGametocyte prevalence was also evaluated.\nBoth the thick and thin blood smear readings were done locally following the above described procedure, while the gametocyte assessments were carried out in accordance with standard practice at each individual site.\nAll technicians who read the slides had undergone appropriate training in malaria-related microscopy and had at least 5 years experience in reading blood smears.\nA process of quality control was used to monitor the values being provided by the local laboratories.\nOne in five of every samples was sent to a central, independent laboratory (Department of Parasitology and Medical Entomology, Muhimbili University of Health and Allied Sciences, Dar Es Salaam, Tanzania) for review.\nThree spots of blood were collected on 3MM filter paper (Whatman, UK) at the enrolment visit and at any visit after day 7 for PCR analysis.\nFilter papers were dried and individually stored in a plastic bag containing silica gel.\nAll filter papers were subsequently transferred to the Institute of Tropical Medicine (Antwerp, Belgium) where centralised genotyping was conducted; deoxyribonucleic acid was purified as described elsewhere.\nThree polymorphic genetic markers, MSP1, MSP2 and GluRP were used to distinguish recrudescence from new infections.\nRecrudescence was defined as at least one identical allele for each of the three markers in the pre-treatment and post-treatment samples.\nNew infections were diagnosed when all alleles for at least one of the markers differed between the two samples.\nAn independent expert read all gels under blinded conditions (National Museum of Natural History, Paris, France).\nFor quality control, 20% of the filter papers were re-analysed and read by an independent laboratory (Shoklo Malaria Research Unit, Mae Sot District, Tak, Thailand).\nTwelve-lead electrocardiograms were recorded at days 0, 2, 7, 28 (if abnormal on day 7), 63 and on the day of any recurrent parasitaemia that occurred using the CarTouch device with CarTouch version 1.4.1 software (Cardionics SA, Brussels, Belgium).\nThe electrocardiogram was viewed during recording and then transmitted via modem to MDS Pharma Services Central Telemedicine Department (Paris, France) for interpretation and reporting.\nThe data were analysed using both Bazett's (QTc(B)) and Fridericia's (QTc(F)) correction methods.\nElectrocardiogram results were tabulated as normal, borderline or prolonged according to gender and age normal ranges (adult males and children (1\u201312 years): normal: <430 msec; borderline: 430\u2013450 msec; prolonged: >450 msec.\nAdult females: normal: <450 msec; borderline: 450\u2013470 msec; prolonged: >470 msec).\nBlood samples were taken for standard laboratory assessments of haematology, biochemistry and urinalysis on days 0, 28, 63 (if abnormal on day 28) and on the day of any recurrent parasitaemia.\nEthical approval and informed consent\nStudy staff conducted all interviews with patients, and children's parents, in their native language and they explained their rights under International Conference on Harmonisation-Good Clinical Practice (ICH-GCP).\nWhen obtaining informed consent from patients the signature of a witness was obtained if patients were unable to write.\nConsideration was given to the ethical implications of the randomisation ratio assigning more patients to receive DHA/PQP than AS+MQ during the design of the study.\nAs the safety and tolerability profile of DHA and PQP has been well characterised at the doses proposed for the present study, it was considered acceptable to adopt this approach in the present study that would result in more patients receiving DHA/PQP.\nThe study protocol was approved by the following ethics committees: Institutional Ethics Committee, National Institute of Malaria Research (ICMR), Delhi, India; Goa Medical College and Hospitals Local Ethics Committee, Goa, India; Institutional Ethics Committee, Kasturba Medical College, Mangalore, India; Laos National Ethics Committee for Health Research (NECHR), National Institute of Public Health, Vientiane, Laos; Oxford Tropical Research Ethics Committee (OXTREC), University of Oxford, Headington, United Kingdom; The Ethical Review Committee for Research in Human Subjects, Ministry of Public Health, Nontaburi, Thailand; Tropical Medicine Ethics Committee (TMEC), Mahidol University, Rachathewi District, Bangkok, Thailand.\nThe Food and Drug Department, Government of Laos PDR, approved the use of DHA/PQP in that country.\nThe trial was conducted under the provisions of the Declaration of Helsinki (1964 and its subsequent amendments up to 2002) and in accordance with ICH-GCP.\nA Study Steering Committee, a Data Monitoring Committee and a Clinical Development Committee were created prior to the beginning of the trial, and worked independently to harmonise and monitor the study.\nThe trial was registered prior to the enrolment of the first patient in the International Standard Randomised Controlled Trials Register, number ISRCTN81306618, at http://www.controlled-trials.com/isrctn/trial/l/0/81306618.html.\nStatistical methods\nThe statistical analysis was conducted according to a pre-defined data analysis plan.\nThree populations were prospectively planned.\nThe intention-to-treat (ITT) population was defined as all randomised patients who took at least one dose of study treatment.\nThe per protocol population was defined as all randomised patients who were eligible according to the main (pre-defined) protocol inclusion and exclusion criteria, received at least 80% of the study medication, underwent the day 63 assessment, took no other anti-malarial drugs (excluding rescue therapies) and, in the presence of asexual parasite stages on thick or thin blood smears, had an evaluable PCR test.\nPatients who missed visits from day 0 to day 2 were excluded from the per protocol population.\nThe third population, referred to as modified ITT, was midway between the two described populations.\nAlthough this was pre-defined as the co-primary population (together with the per protocol), results in this paper are presented only for the two most extreme populations, i.e., ITT and per protocol populations, because these are standard and the findings were very similar in all populations analysed.\nThe analysis of the PCR-corrected and uncorrected ACPR (as defined in Table 1) was based on simple proportions and 97.5% (one-sided) confidence intervals (CIs) computed on the difference between these proportions of the test and reference treatments.\nIf the lower limit of this CI was greater than \u22125% DHA/PQP could be considered non-inferior to AS+MQ.\nThe primary time point was day 63 but this analysis was repeated also at days 42 and 28.\nVarious sensitivity analyses were carried-out on the primary endpoint for verifying the robustness of findings towards different assumptions.\nThese included an analysis where all patients with missing parasitaemia were treated as failures, an analysis where patients with new infections as detected by PCR were excluded and an analysis on two enlarged per protocol populations where all patients/all early failures \u201cnot receiving at least 80% of study treatment\u201d or \u201cfailing to attend a visit at days 0\u20132\u201d were not excluded.\nThe true treatment failures defined in accordance with the WHO handbook were analysed by means of the survival analysis (Kaplan-Meier estimates).\nAll sources of uncertainty (i.e. withdrawals, new infections, patients with PCR results either not available or indeterminate) were censored.\nSurvival analysis techniques were also applied to the analysis of time to parasite clearance and the estimation of the rate of new infections (in the latter analysis recrudescent infections were censored).\nFor descriptive purposes, the proportions of early, late and true treatment failures were also computed as simple rates for each treatment arm together with the relevant CIs.\nBy-country heterogeneity in cure rates was assessed by the Breslow-Day test or by logistic regression, when the former was not applicable.\nThe study was conducted over two malaria seasons with different centres working in the two study periods.\nTo evaluate how the 63-day PCR-corrected cure rates varied across the cohorts from the two seasons, a logistic regression model was fitted with cohort and treatment as the explanatory variables.\nThe treatment by cohort interaction was evaluated as a candidate to enter this model through a residual score test.\nAll tests for heterogeneity and interaction were evaluated at the 10% significance level.\nRates of person-fever-days and person-gametocyte-weeks were calculated as the number of weeks in which fever was present or blood slides were positive for gametocytes, respectively, divided by the number of follow-up weeks and expressed per 100 person-weeks.\nThe safety population, which coincides with the ITT population, was used for all safety assessments.\nAdverse events were coded using the MedDRA dictionary (MedDRA V8.1).\nProportions of patients experiencing at least one adverse event were compared between treatments using the Pearson Chi square test.\nTo determine the sample size of this study, the PCR-corrected cure rate at day 63 was estimated to be at least 92% for AS+MQ in the ITT population using a literature search.\nExpert opinion was used to define the non-inferiority margin, which was set at 5%.\nThe planned sample size of 700 in the DHA/PQP arm and 350 in the AS+MQ arm (1050 patients in total) provided 80% power to show non-inferiority with a non-inferiority margin of \u22125% (test minus reference) and a one-sided \u03b1 level of 2.5%.\nThe rate of patient attrition in the per protocol population was expected to be 20% compared with the ITT population, but the PCR-corrected cure rate was expected to be higher than in the ITT population (95% in the per protocol population), therefore the projected power of 80% was maintained also for the analysis on the per protocol population.\nWhen India was included in the study to increase the speed of recruitment, the total sample size was increased by 150 patients to ensure that 100 Indian patients were treated with DHA/PQP, in accordance with Indian requirements.\nConsequently, the power of the study exceeded 80%.\nResults\nPatient disposition and demographic characteristics\nOne thousand two hundred and thirty-nine patients were screened, with 769 patients randomised to DHA/PQP and 381 patients randomised to AS+MQ.\nFigure 1 illustrates the number of patients completing the study and included in the different study populations.\nSixty-one percent of patients were recruited in Thailand, 26% in Laos and 13% in India (Table 2).\nThere were no notable differences in demographic parameters between treatment arms overall or by country (Table 2).\nChildren \u22645 years of age comprised approximately 8% of the study population, with most of the study population over 18 years of age (approximately 75%).\nMedian [range] doses received by patients of the ITT population are presented by age class in Table 3.\nOne hundred and forty-four patients were excluded from the per protocol population (Table 2).\nThere were no notable differences between the treatment arms in the number of patients excluded from the per protocol population, the most frequent reasons for exclusion were lost to follow-up before or at day 63 and PCR missing or indeterminate before or at day 63 (DHA/PQP: 35 patients: AS+MQ: 19 patients).\nPrimary Endpoint\nThe analysis of the PCR-corrected cure rate at day 63 confirmed that DHA/PQP was non-inferior to AS+MQ.\nFor the ITT population, PCR-corrected cure rates were 87.9% for DHA/PQP and 86.6% for AS+MQ (97.5% CI: >\u22122.87%; Table 4).\nAs expected, better absolute but comparatively similar results were obtained for the per protocol population with PCR-corrected cure rates of 98.7% for DHA/PQP and 97.0% for AS+MQ (97.5% CI: >\u22120.39%).\nSensitivity analyses confirmed these results, with similar treatment differences between cure rates and relevant CIs compared with those described above.\nSecondary endpoints\nNon-inferiority was also proved for the uncorrected cure rates at day 63: the cure rates were 67.3% in the DHA/PQP arm and 59.6% in the AS+MQ arm (ITT population) with a CI >1.75%.\nThis treatment difference was also statistically significant (Chi-squared test: p\u200a=\u200a0.010; Table 4).\nSimilar results were obtained for the per protocol population (Table 4).\nSignificantly more patients who received AS+MQ experienced new infections compared with those receiving DHA/PQP.\nKaplan-Meier estimates for the proportion of patients with new infection at day 63 were 22.7% for DHA/PQP and 30.3% for AS+MQ (ITT population; p\u200a=\u200a0.0042).\nSimilar results were obtained for the other populations.\nNon-inferiority of DHA/PQP versus AS+MQ was also shown at days 28 and 42 in all study populations (Table 4).\nOn days 28 and 42, in the per protocol population, the PCR-corrected cure rates defined in Table 1 were significantly greater for DHA/PQP than AS+MQ (p\u200a=\u200a0.001 and p\u200a=\u200a0.031, respectively) while these differences were not statistically significant in the ITT population (Table 4).\nA logistic regression test to assess by-country heterogeneity in the PCR-corrected cure rates (evaluated according to Table 1) showed no significant differences between countries (ITT: p\u200a=\u200a0.688; per protocol: p\u200a=\u200a0.988;).\nSimilar results were obtained for a Breslow-Day test conducted to assess by-country heterogeneity in the uncorrected cure rate (ITT: p\u200a=\u200a0.896; per protocol: p\u200a=\u200a0.728).\nThe 95% (two-sided) CIs at each individual country level for the PCR-corrected cure rate at day 63 for the per protocol population are shown in Figure 2A.\nNo effect due to the cohorts from the two seasons (per protocol: p\u200a=\u200a0.760) was shown in this study (Figure 2B shows the relevant CIs) and no difference was observed by age class (per protocol: p\u200a=\u200a0.998).\nClassification by age was: \u22642 years; 2\u201312 (included) years; 12\u201318 (included) years and >18 years (relevant CIs are shown in Figure 2C).\nResults in the other populations for country, cohort (season) of enrolment and age groups were similar.\nWhen the PCR-corrected cure rates were assessed and analysed according to the WHO recommendations, i.e. as 100 minus the true treatment failure rate, the results were similar to those in the PCR-corrected cure rate as defined in Table 1, although cure rates were always slightly higher.\nThe Kaplan-Meier estimates of the cure rates were 97.6% (DHA/PQP) versus 96.5% (AS+MQ) in the ITT population and 98.2% (DHA/PQP) versus 96.8% (AS+MQ) in the per protocol population.\nThe simple rates of early, late and true treatment failures on days 63, 42 and 28 are shown in Table 5 together with the relevant CIs.\nKaplan-Meier estimates of median time to parasite clearance were 2 days for each treatment (ITT population), while the Kaplan-Meier estimate of the rate of aparasitaemic patients at day 3 (i.e. 24 h after completion of study treatment) was 97.6% for DHA/PQP and 97.6% for AS+MQ.\nSimilar results were obtained for the per protocol population.\nProportions of afebrile patients showed very similar profiles to aparasitaemic patients, with profiles being virtually superimposable.\nFever incidence measured in person-fever-days (per 100 person-days) was similar for DHA/PQP and AS+MQ: 1065/5315 (20.04%) versus 524/2636 (19.88%), p\u200a=\u200a0.929 (Table 6; ITT population).\nThe overall gametocyte prevalence throughout the study (from day 7 to day 63), measured as person-gametocyte-week rates, was significantly higher in the DHA/PQP arm in all populations (DHA/PQP: 76/749, 10.15%; AS+MQ: 18/369, 4.88%; p\u200a=\u200a0.003; Table 6, ITT population).\nHowever, the treatment difference for prevalence was not statistically significant from day 35 onwards (Table 6; ITT population).\nThere were no deaths during the study.\nIn the DHA/PQP group, there were 12 (1.56%) events judged by the investigator to be serious, which occurred in 12 patients.\nSix (0.78%) of these events were considered by the investigators to be related to DHA/PQP.\nThese were two cases of anaemia, one from day 7 to day 90, the other from day 7 to day 35; one viral infection (possibly Dengue fever) from day 15, fully recovered; and one Wolf Parkinson White (WPW) syndrome from day 2 to day 90.\nWolf Parkinson\nWhite syndrome is a congenital defect of accessory conduction pathways.\nElectrocardiograms for this patient were submitted to two cardiologists to provide expert opinions.\nBoth considered that the failure to diagnose WPW at baseline could be attributed to the increased heart rate caused by the patients' fever which hid the electrocardiographic characteristics of WPW.\nOther events considered serious and related were one convulsion on day 0, and one encephalitis on day 45 which resulted in a left-sided hemiplegia.\nThe patient was diagnosed with P. falciparum malaria on day 48, which was treated with i.v. artesunate and oral mefloquine.\nThe other six (0.8%) serious events were judged unrelated to DHA/PQP by the investigator: two cases of pyelonephritis, one case of aspiration pneumonia, and three cases of P. falciparum malaria.\nIn the AS+MQ group three (0.8%) events in three patients were judged to be serious and all were judged related to study treatment: one case of anaemia from day 8, one case of convulsion on day 1, and one case of encephalitis from day 16 to day 31.\nAdverse event profiles for DHA/PQP and AS+MQ were very similar in terms of type and frequency of events and were consistent with those expected in adult patients with acute malaria.\nMost patients in the study experienced adverse events: 69.4% (532/767) of patients in the DHA/PQP arm and 72.4% (276/381) of patients in the AS+MQ arm.\nThere was no statistically significant difference between the treatments in the incidence of adverse events (p\u200a=\u200a0.282, Chi-square test).\nThe most frequently reported adverse events were malaria symptoms, with headache the most commonly reported (Table 7).\nThe frequencies of individual adverse events were generally similar between treatments, although the frequencies of nausea, vomiting and dizziness appeared to be higher in the AS+MQ arm (Table 7).\nApproximately 3% of patients in each arm experienced at least one adverse event related to skin or subcutaneous tissue; the most frequent was pruritus (DHA/PQP: 17 patients [2.2%]; AS+MQ: 8 patients [2.1%]).\nOne patient in each arm experienced allergic dermatitis.\nAs would be expected in patients with malaria, anaemia and thrombocytopenia were common in each arm, approximate proportions were 30% for low red blood cells (normal range for >15 years of age: 3.5\u22125.5\u00d71012/L for women and 4\u22125.5\u00d71012/L for men), 50% for low haemoglobin (normal range for >15 years of age: 110\u2212150 g/L for women and 120\u2212160 g/L for men), and 67% for low platelets (normal range for all ages: 100\u2212300\u00d7109/L).\nAt the end of the study following treatment, these proportions had decreased to approximately, 20%, 40% and 17%, respectively.\nThere was no apparent difference between treatments.\nIncreases in mean haemoglobin levels were observed over the 63-day follow-up period.\nIn the ITT population, mean (standard deviation) haemoglobin values on day 0 were 118.7 (24.4) g/L for DHA/PQP and 119.8 (23.4) g/L for AS+MQ.\nAt day 63, mean (standard deviation) changes from day 0 were 12.8 (22.2) g/L for DHA/PQP and 14.21 (21.2) g/L for AS+MQ.\nSimilar results were observed for the per protocol population.\nOther than elevated liver parameters, as might be expected in this population, there were no relevant changes in biochemistry parameters.\nAt baseline, using the QTc(B) method, there was a statistically significant difference between treatments (Mantel-Haenszel Chi-square test, p\u200a=\u200a0.026), with a higher proportion of patients in the DHA/PQP group having borderline QTc(B) values (16.6% vs. 12.2%; p\u200a=\u200a0.066).\nNo statistically significant difference between treatments was observed at baseline for QTc(F) (2.9% for DHA/PQP vs. 1.6% in AS+MQ).\nOn day 2, there was a statistically significant difference between treatments (Mantel-Haenszel Chi-square test p<0.001), with a higher proportion of patients in the DHA/PQP group having borderline (21.4%; p\u200a=\u200a0.043) or prolonged (8.6%; p\u200a=\u200a0.007) QTc(B) intervals than in the AS+MQ group (16.3% and 4.2%, respectively).\nThis difference was also observed for the QTc(F) method: borderline, 13.0% for DHA/PQP vs. 5.3% for AS+MQ (p<0.001); prolonged, 4.7% for DHA/PQP vs. 5.3% for AS+MQ (p<0.001).\nBy day 7, there was no difference between treatments.\nThe proportion of patients with QTc(B) increase >60 msec from baseline to day 2 was 0.9% for DHA/PQP vs. 0.8% for AS+MQ (not significant), and 4.6% for DHA/PQP vs. 2.9% for AS+MQ when the same increase was assessed with QTc(F) (p<0.001).\nHowever, QTc and QT prolongation were reported as adverse events by 43 (5.6%) patients in the DHA/PQP group and 16 (4.20%) patients in the AS+MQ group; these were judged by the investigator to be related to study treatment for 28 (3.65%) patients in the DHA/PQP group and 13 (3.41%) patients in the AS+MQ group.\nMean QTc(F)values on day 0 were 387.70 msec for DHA/PQP and 385.54 msec for AS+MQ.\nMean increases from baseline to day 2 were 22.93 msec and 14.65 msec, respectively.\nThis difference between treatments was statistically significant (p <0.001).\nOn day 7, the mean increase from day 0 for DHA/PQP had fallen to 10.47 msec with the value for AS+MQ being 13.39 msec; this difference was not statistically significant (p\u200a=\u200a0.075).\nDiscussion\nIn this study conducted in Thailand, Laos and India, we have shown that both the DHA/PQP and AS+MQ treatment combinations are efficacious treatments of P. falciparum malaria.\nThe day-63 PCR-corrected cure rates (as defined in Table 1) were 87.9% for DHA/PQP and 86.6% for AS+MQ in the ITT population and 98.7% for DHA/PQP and 97.0% for AS+MQ in the per protocol population.\nIn terms of this primary outcome variable, DHA/PQP was non-inferior to AS+MQ.\nReview of the data by country found no differences between the primary outcome measures, further confirming that DHA/PQP was similarly active against the P. falciparum found in India, Laos and Thailand, relative to AS+MQ.\nThe PCR-corrected cure rates observed for DHA/PQP in this study were in line with the day-63 PCR-corrected cure rates noted in a previous study conducted in Thailand, day-42 rates observed in Myanmar (Burma) and Laos and day-28 cure rates observed in Cambodia.\nLikewise, PCR-corrected cure rates for AS+MQ were also similar to previous studies of AS+MQ conducted in Thailand, Laos and India, which generally showed rates ranging from approximately 95% to 100% in the per protocol population.\nPolymerase chain reaction-corrected cure rates for both DHA/PQP and AS+MQ exceeded 95% on days 28, 42 and 63.\nThe 95% threshold on day 28 is the level the WHO recommend for adoption of a new treatment.\nThis is the first study of the use of DHA/PQP in the Indian population, with 101 patients receiving the combination.\nAlthough this was only 13% of the patients in our study, we feel that it provides sufficient numbers to enable an initial assessment of the response to DHA/PQP in communities in Assam, Goa and Mangalore.\nIndia, Thailand and Laos have different backgrounds in terms of parasite transmission, resistance and seasonality of infection, yet a heterogeneity analysis showed no statistically significant difference in the relative responses to the two treatments between India and the other countries in the study.\nIndian communities that we sampled in this study had notable levels of chloroquine resistance, with historical site records showing treatment failure rates for chloroquine treatment ranging from 32% to 67% (28 days of follow-up).\nThe sites in Laos and Thailand were also in regions with high levels of chloroquine resistance.\nBased on the structural similarities of chloroquine and PQP, there was the potential for cross-resistance and a low response to DHA/PQP may have been expected.\nHowever, the DHA/PQP combination appeared to be unaffected by any potential cross-resistance and PCR-corrected cure rates were similar for India (96.5%), Thailand (96.6%) and Laos (98.0%).\nThe DHA/PQP combination exerted a significant post-treatment prophylactic effect in this study.\nThis is supported by, first, significant reductions in the incidence of new infections for DHA/PQP compared with AS+MQ and second, by higher uncorrected cure rates in the DHA/PQP arm than AS+MQ, as the uncorrected cure rate endpoint includes new infections as well as recurrence of infection.\nThis effect is thought to be modulated by levels of drug remaining in the blood because of the long half-life of PQP.\nThe post-treatment prophylaxis was significantly better for DHA/PQP reflecting the differential terminal half lives of PQP and MQ (4\u20135 weeks for PQP and 14 days for MQ).\nThe extended post-treatment prophylactic effect is of particular importance in countries with a high risk of new infection and can also reduce the risk of anaemia by allowing patients more time for haematological recovery between infections.\nThe DHA/PQP combination, like all ACTs, rapidly reduces parasite biomass in the patient through the brief yet potent activity of DHA (the artemisinin component).\nSubsequent removal of uncleared parasites is achieved by the less active but more slowly eliminated partner drug; in this case PQP with a half-life of 4\u20135 weeks.\nThe slightly shorter half-life of MQ may explain the difference observed in the post-treatment prophylactic effect of the two regimens.\nThe DHA/PQP regimen has previously been shown to exert a superior post-treatment prophylactic effect to another ACT, artemether-lumefantrine.\nOverall gametocyte prevalence was significantly higher in the DHA/PQP arm than AS+MQ.\nThe viability of gametocytes remaining post-treatment was not assessed in this study.\nOne potential point of interest for future studies would be to determine whether the viability of any remaining gametocytes is compromised by treatment with DHA/PQP.\nAlthough formal statistical comparison of the tolerability profile of the two combinations was not planned or conducted, some individual adverse events appeared to be reported more frequently in patients receiving AS+MQ than those receiving DHA/PQP.\nMefloquine has been linked with adverse events of gastrointestinal and central nervous system origin.\nIn this study, the frequency of reporting of nausea, vomiting and dizziness was 2\u20133 fold higher in the AS+MQ arm than the DHA/PQP arm.\nVomiting soon after dosing is an important determinant of treatment efficacy as attaining and maintaining effective systemic levels of anti-malarial drugs is essential to disease outcome.\nThis is of particular importance where there are powerful influences on adherence to dosing regimens.\nIncreases in QTc(F) interval were seen for both DHA/PQP and AS+MQ after the start of treatment, but a statistically significant increase from baseline was observed in QTc interval for DHA/PQP relative to AS+MQ on day 2.\nAn important element in assessing the clinical importance of QTc prolongation is the extent of change from baseline: an increase in QTc interval from baseline of >60 msec could be clinically significant and this is the recognised threshold for drug-induced arrhythmias in non-cardiovascular indications (ICH E14 guideline).\nThere was a statistically significant treatment difference for the proportion of patients with such an increase in QTc(F), while no treatment differences were observed for the same increase in QTc(B) and for the incidence of cardiac events.\nFinally, by day 7 there were no significant differences between treatments.\nBased on our results, it is difficult to attribute a particular clinical relevance to the QTc increase observed with DHA/PQP.\nIn fact, it is expected that malaria shortens the QT interval during the acute illness, leading to increases after the start of treatment.\nMost studies of antimalarial drugs report some transient prolongation of the QT interval in the days after the start of treatment.\nNo warnings were found in the literature about QTc prolongation with DHA/PQP.\nFor patients to access DHA/PQP via public sector healthcare systems, this fixed dose combination requires regulatory approval.\nOne requirement for registration is that the formulation must be manufactured according to GMP.\nThis study was conducted as one of a series of pharmacokinetic and ICH-GCP-compliant randomised, controlled trials at sites across Africa and South East Asia using DHA/PQP manufactured according to GMP.\nIn conclusion, the fixed dose DHA/PQP combination tablet in this study emerged as a highly efficacious treatment for P. falciparum malaria.\nThe effects were observed across the three Asian countries in which the study was conducted, with no country effect.\nThe combination of DHA/PQP provided greater protection against new infections than AS+MQ.\nResults from this study indicate that, although there may be evidence that suggests that the overall efficacy of ACTs may be falling, DHA/PQP can play an important role in, possibly, a new policy of multiple first-line treatments of uncomplicated falciparum malaria.\nFlow chart of patient disposition.\nNinety-five percent (two-sided) confidence intervals of PCR-corrected cure rates in the ITT population, by country, cohort of enrolment from the two seasons and age group.\n\nRules used to determine patient outcome for the ITT and per protocol populations for the day 63 uncorrected adequate clinical and parasitological response (ACPR) and polymerase chain reaction (PCR)-corrected ACPR.\nStep | Event to be assessed | ITT | Per Protocol\nDay-63 uncorrected adequate clinical and parasitological response\n1 | Informative withdrawal before or at day 63: any reason except lost to follow-up | Failure | Failure or excluded depending on reason\n2 | Non-informative withdrawal before or at day 63: lost to follow-up | Failure | Excluded\n31 | Presence of major protocol violation | No effect | Excluded\n4 | ETF2, LCF3, and LPF4 in accordance with the WHO criteria | Failure | Failure\n51 | Data collected on CRF (such as adverse events) raising the suspicion of recurrence of malaria | Failure | Failure\n61 | Presence of missing parasitaemia at two or more consecutive scheduled visits or presence of an isolated missing parasitaemia not preceded and followed by a negative parasitaemia | Failure | Failure\n71 | Administration of drugs with a known or suspected anti-malaria action as rescue treatment | Failure | Failure\n81 | Administration of drugs with a known or suspected anti-malaria action as non rescue treatment | Failure | Excluded\nPCR-corrected adequate clinical and parasitological response\n9 | PCR: non interpretable or missing or not done at or after day 4 | Failure | Excluded\n10 | PCR\u200a=\u200anew infection or uncorrected ACPR\u200a=\u200afailure for non-falciparum Plasmodia | Success | Success\n11 | PCR\u200a=\u200arecrudescence | Failure | Failure\n\nITT \u200a=\u200a intention to treat.\nCases in these categories were individually revised at the blind data review meeting. Protocol violations were pre-defined.\nETF \u200a=\u200a early treatment failure, defined as development of danger signs (recent convulsions, altered consciousness, lethargy, unable to drink or breast feed, recurrent vomiting, unable to stand/sit due to weakness) or severe malaria (unarousable coma, repeated convulsions, severe anaemia, respiratory distress, jaundice) on days 0, 1, 2 or 3, and the presence of parasitaemia; parasitaemia with a parasite count on day 2 greater than that on day 0 irrespective of body temperature; parasitaemia on day 3 with fever (temperature \u226537.5\u00b0C); or parasitaemia on day 3\u226525% of count on day 0.\nLCF \u200a=\u200a late clinical failure, defined as development of danger signs or severe malaria after day 3 in the presence of parasitaemia or presence of parasitaemia and temperature \u226537.5\u00b0C (or history of fever) on any day from day 4 to day 63, without previously meeting the criteria of early treatment failure.\nLPF \u200a=\u200a late parasitological failure, defined as reappearance of parasitaemia after initial clearance between day 7 and day 63 and temperature <37.5\u00b0C, without previously meeting the criteria of early treatment failure or late clinical failure.\n\nBaseline characteristics (ITT population).\nVariable | DHA/PQP | AS+MQ\n | N\u200a=\u200a767 | N\u200a=\u200a381\nSample size by country; n (%) |  | \nThailand | 466 (61) | 234 (61)\nLaos | 200 (26) | 98 (26)\nIndia | 101 (13) | 49 (13)\nGender; n : n |  | \nMale : Female | 582\u2236185 | 295\u223686\nAge; years |  | \nMean \u00b1 SD | 25.4\u00b113.3 | 25.8\u00b113.7\n\u22645 years, n (%) | 57 (7) | 32 (8)\n>5\u2013\u226412 years, n (%) | 68 (9) | 31 (8)\n>12\u2013\u226418 years, n (%) | 76 (10) | 31 (8)\n>18\u2013\u226464 years, n (%) | 566 (74) | 287 (75)\nWeight; kg |  | \nMean \u00b1 SD | 44.3\u00b115.1 | 44.6\u00b115.1\nRace; n (%) |  | \nAsian | 767 (100) | 381 (100)\nPresence of fever |  | \nn (%) | 509 (66) | 258 (68)\nTemperature in \u00b0C |  | \nMean \u00b1 SD | 37.9 (1.01) | 37.9 (1.02)\nParasite density |  | \nGeometric mean | 7923.8 | 9735.4\nHaemoglobin; g/L |  | \nNormal range (min - max)*: (105\u2013180) |  | \nMean \u00b1 SD | 118.2\u00b124.5 | 120.0\u00b123.2\nn Missing (%) | 4 (0.52) | 2 (0.52)\nn <100 g/L (%) | 162 (21.12) | 72 (18.90)\nn\u2265100 g/L (%) | 601 (78.36) | 307 (80.58)\nWhite cells; \u00d7109/L |  | \nNormal range (min - max)*: (3.6\u201311.0) |  | \nMean \u00b1 SD | 6.3\u00b12.6 | 6.3\u00b12.3\nPlatelets; \u00d7109/L |  | \nNormal range (min - max)*: (140\u2013500) |  | \nMean \u00b1 SD | 127.7\u00b170.6 | 124.8\u00b165.5\nALT; U/L |  | \nNormal range (min - max)*: (0\u201350) |  | \nMean \u00b1 SD | 31.1\u00b129.3 | 32.9\u00b141.5\nTotal bilirubin; mg/dl |  | \nNormal range (min - max)*: (0\u20131.5) |  | \nMean \u00b1 SD | 1.19\u00b10.85 | 1.20\u00b10.78\nCreatinine; \u00b5mol/L |  | \nNormal range (min - max)*: (0\u2013150.28) |  | \nMean \u00b1 SD | 75.5\u00b128.7 | 76.2\u00b130.7\nStudy populations; n (%) |  | \nSafety/ITT | 767 (99.7) | 381 (100.0)\nPer protocol | 668 (86.9) | 336 (88.2)\n\nDHA/PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine; SD \u200a=\u200a standard deviation; Hb \u200a=\u200a haemoglobin; ALT \u200a=\u200a alanine aminotransferase.\n*Since the normal laboratory reference ranges vary across centres, the minimum of the lower limits and the maximum of the upper limits are reported.\n\nDoses (mg/kg) received for each age class (ITT population).\nAge range (years) | Doses received over 3 days; median [range]\n | DHA | PQP | AS | MQ\n\u22645 | 6.7 [1.8\u20139.2] | 53.3 [14.5\u201373.8] | 12.5 [10.7\u201313.6] | 24.0 [21.7\u201327.8]\n>5\u2013\u226412 | 7.2 [5.0\u201310.0] | 57.3 [40.0\u201380.0] | 12.5 [10.7\u201313.2] | 25.0 [22.3\u201326.3]\n>12\u2013\u226418 | 7.9 [2.9\u20139.7] | 63.3 [22.9\u201377.8] | 12.0 [11.4\u201312.5] | 25.0 [24.2\u201331.3]\n>18\u2013\u226464 | 7.1 [2.1\u201310.0] | 56.5 [16.8\u201380.0] | 12.0 [3.9\u201312.5] | 25.0 [0.0\u201326.5]\n\nDHA \u200a=\u200a dihydroartemisinin; PQP \u200a=\u200a piperaquine; AS \u200a=\u200a artesunate; MQ \u200a=\u200a mefloquine.\n\nPolymerase chain reaction (PCR)-corrected and uncorrected cure rates (ITT and per protocol populations).\n | DHA/PQP | AS+MQ | Lower limit of the 97.5% | p-value2\n | % (n) | % (n) | (one-sided) CI1 | \nITT | N\u200a=\u200a767 | N\u200a=\u200a381 |  | \nDay 63 |  |  |  | \nPCR-corrected cure rate | 87.9 (674) | 86.6 (330) | \u22122.87 | 0.544\nUncorrected cure rate | 67.3 (516) | 59.6 (227) | 1.75 | 0.010\nDay 42 |  |  |  | \nPCR-corrected cure rate | 90.5 (694) | 88.2 (336) | \u22121.56 | 0.228\nUncorrected cure rate | 83.2 (638) | 77.4 (295) | 0.79 | 0.019\nDay 28 |  |  |  | \nPCR-corrected cure rate | 93.7 (719) | 91.9 (350) | \u22121.36 | 0.236\nUncorrected cure rate | 92.3 (708) | 88.2 (336) | 0.37 | 0.022\nPer protocol | N\u200a=\u200a668 | N\u200a=\u200a336 |  | \nDay 63 |  |  |  | \nPCR-corrected cure rate | 98.7 (659/668) | 97.0 (326/336) | \u22120.39 | 0.074\nUncorrected cure rate | 75.5 (504/668) | 66.4 (223/336) | 3.07 | 0.002\nDay 42 |  |  |  | \nPCR-corrected cure rate | 99.3 (663) | 97.6 (328) | \u22120.12 | 0.031\nUncorrected cure rate | 91.2 (609) | 85.4 (287) | 1.41 | 0.006\nDay 28 |  |  |  | \nPCR-corrected cure rate | 99.9 (667) | 97.9 (329) | 0.38 | 0.001\nUncorrected cure rate | 98.2 (656) | 93.8 (315) | 1.68 | <0.001\n\nDHA-PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine; CI \u200a=\u200a confidence interval.\nConfidence interval for the difference DHA/PQP minus AS+MQ.\nChi-squared test.\n\nEarly, late and true treatment failure rates1 as defined by WHO (ITT and per protocol populations).\n | DHA/PQP | AS+MQ | 95% CI\n | % (n) | % (n) | \nITT\nEarly treatment failures | 0.52 (4) | 0.26 (1) | \u22120.46, 0.98\nDay 63 |  |  | \nLate treatment failures | 13.17 (101) | 14.96 (57) | \u22126.10, 2.52\nTrue treatment failures | 2.09 (16) | 2.62 (10) | \u22122.44, 1.36\nDay 42 |  |  | \nLate treatment failures | 5.74 (44) | 9.71 (37) | \u22127.37, \u22120.58\nTrue treatment failures | 1.17 (9) | 2.10 (8) | \u22122.56, 0.70\nDay 28 |  |  | \nLate treatment failures | 1.17 (9) | 5.25 (20) | \u22126.44, \u22121.71\nTrue treatment failures | 0.65 (5) | 1.84 (7) | \u22122.65, 0.28\nPer protocol |  |  | \nEarly treatment failures | 0 | 0.30 (1) | \u22120.89, 0.28\nDay 63 |  |  | \nLate treatment failures | 12.87 (86) | 15.48 (52) | \u22127.23, 2.02\nTrue treatment failures | 1.65 (11) | 2.98 (10) | \u22123.39, 0.73\nDay 42 |  |  | \nLate treatment failures | 5.84 (39) | 9.82 (33) | \u22127.63, \u22120.34\nTrue treatment failures | 0.75 (5) | 2.38 (8) | \u22123.39, 0.12\nDay 28 |  |  | \nLate treatment failures | 0.90 (6) | 4.76 (16) | \u22126.25, \u22121.48\nTrue treatment failures | 0.15 (1) | 2.08 (7) | \u22123.49, \u22120.38\n\nThe rates reported in this table are simple rates.\n\nPrevalence of gametocytes and fever according to day of follow-up (ITT population).\n | DHA/PQP | AS+MQ | p-value1\n | N\u200a=\u200a767 | N\u200a=\u200a381 | \n | n/N (%) | n/N (%) | \nGametocyte prevalence2 |  |  | \nDay 7 | 59/749 (7.88) | 15/369 (4.07) | 0.016\nDay 14 | 30/742 (4.04) | 3/365 (0.82) | 0.003\nDay 21 | 16/733 (2.18) | 0/362 | 0.005\nDay 28 | 9/722 (1.25) | 0/353 | 0.035\nDay 35 | 1/715 (0.14) | 0/339 | 1.000\nDay 42 | 1/692 (0.14) | 0/328 | 1.000\nDay 49 | 2/666 (0.30) | 0/316 | 1.000\nDay 56 | 1/651 (0.15) | 1/312 (0.32) | 0.543\nDay 63 | 1/623 (0.16) | 2/301 (0.66) | 0.249\nOverall (from day 7 up to day 63) | 76/749 (10.15) | 18/369 (4.88) | 0.003\nPerson-gametocye-weeks3 (/100 person-weeks) | 130/6420 (2.02) | 23/3108 (0.74) | 0.014\nFever prevalence2 |  |  | \nDay 0 | 509/767 (66.36) | 258/381 (67.72) | 0.646\nDay 1 | 244/767 (31.81) | 129/381 (33.86) | 0.486\nDay 2 | 80/765 (10.46) | 44/379 (11.61) | 0.555\nDay 3 | 51/764 (6.68) | 21/379 (5.54) | 0.457\nDay 7 | 40/753 (5.31) | 16/373 (4.29) | 0.458\nOverall (from day 0 up to day 7) | 566/767 (73.79) | 298/381 (78.22) | 0.102\nPerson-fever-days4 (/100 person-days) | 1065/5315 (20.03) | 524/2636 (19.88) | 0.929\n\nDHA/PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine.\nPearson Chi-square or Fisher's exact test, as appropriate.\nCalculated, at a given time, as number of patients with gametocytes, or fever \u226537.5\u00b0C, at that time divided by the number of patients having reached that time.\nCalculated as number of weeks in which blood slides were positive for gametocytes during the whole study (up to a maximum duration of 70 days) divided by the number of all follow-up weeks and expressed per 100 person-weeks. Withdrawal patients were analysed up to the date of withdrawal recorded in the efficacy dataset even if they performed additional gametocyte assessments.\nCalculated as number of days in which temperature was greater or equal to 37.5 during the first study week divided by the number of all first follow-up weeks and expressed per 100 person-days. Withdrawal patients were analysed up to the date of withdrawal recorded in the efficacy dataset even if they performed additional assessments for temperature.\n\nMost frequently reported adverse events (>5% of either treatment arm; ITT population).\nEvent | DHA/PQP | AS+MQ | Chi-Square\n | N\u200a=\u200a767 | N\u200a=\u200a381 | p-value\n | n (%) | n (%) | \nHeadache | 138 (18.0) | 77 (20.2) | 0.3644\nMalaria1 | 111 (14.5) | 86 (22.6) | 0.0006\nP. falciparum infection | 103 (13.4) | 58 (15.2) | 0.4097\nPyrexia | 81 (10.6) | 43 (11.3) | 0.7092\nEosinophilia | 65 (8.5) | 38 (10.0) | 0.4026\nCough | 60 (7.8) | 37 (9.7) | 0.2786\nAnaemia | 55 (7.2) | 25 (6.6) | 0.7027\nMyalgia | 46 (6.0) | 22 (5.8) | 0.8801\nArthralgia | 42 (5.5) | 21 (5.5) | 0.9799\nProlonged QTc interval | 41 (5.4) | 16 (4.2) | 0.3999\nAbdominal pain | 40 (5.2) | 20 (5.3) | 0.9804\nAsthenia | 38 (5.0) | 29 (7.6) | 0.0705\nAnorexia | 38 (5.0) | 21 (5.5) | 0.6871\nNausea | 22 (2.9) | 26 (6.8) | 0.0016\nVomiting | 19 (2.5) | 24 (6.3) | 0.0013\nDizziness | 11 (1.4) | 24 (6.3) | <.0001\n\nDHA/PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine.\nReporting of malaria as an adverse event was not complete in this study. Some study centres chose not to report malaria as it was known that to enter the study all patients had to have Plasmodium falciparum infection.",
        "label": "low",
        "id": "task4_RLD_test_568"
    },
    {
        "paper_doi": "10.1186/s12889-017-4203-1",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: DesigncNON-RCTAllocation of clusters\n10 schools allocated to intervention, 10 to control (Cambodia)9 schools allocated to intervention, 9 to control (Indonesia)22 schools allocated to intervention, 22 to control (Lao PDR)\n\n\nParticipants: 478, 486, and 535 children ages 6 to 7\n\n\nInterventions: Broad multiple\n\n\nOutcomes: Any STH\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Background\nThe Fit for School (FIT) programme integrates school health and Water, Sanitation and Hygiene interventions, which are implemented by the Ministries of Education in four Southeast Asian countries.\nThis paper describes the findings of a Health Outcome Study, which aimed to assess the two-year effect of the FIT programme on the parasitological, weight, and oral health status of children attending schools implementing the programme in Cambodia, Indonesia and Lao PDR.\nMethods\nThe study was a non-randomized clustered controlled trial with a follow-up period of two years.\nThe intervention group consisted of children attending public elementary schools implementing the FIT programme, including daily group handwashing with soap and toothbrushing with fluoride toothpaste, biannual school-based deworming; as well as construction of group handwashing facilities.\nControl schools implemented the regular government health education curriculum and biannual deworming.\nPer school, a random selection of six to seven-year-old grade-one students was drawn.\nData on parasitological infections, anthropometric measurements, dental caries, odontogenic infections and sociodemographic characteristics were collected at baseline and at follow-up (24\u00a0months later).\nData were analysed using the \u03c72-test, Mann Whitney U-test and multilevel logistic and linear regression.\nResults\nA total of 1847 children (mean age\u00a0=\u00a06.7\u00a0years, range 6.0\u20138.0\u00a0years) participated in the baseline survey.\nOf these, 1499 children were available for follow-up examination \u2013 478, 486 and 535 children in Cambodia, Indonesia and Lao PDR, respectively.\nIn all three countries, children in intervention schools had a lower increment in the number of decayed, missing and filled permanent teeth between baseline and follow-up, in comparison to children in controls schools.\nThe preventive fraction was 24% at average.\nThe prevalence of soil-transmitted helminth infection (which was unexpectedly low at baseline), the prevalence of thinness and the prevalence of odontogenic infections did not significantly differ between baseline and follow-up, nor between intervention and control schools.\nConclusions\nThe study found that the FIT programme significantly contributed to the prevention of dental caries in children.\nThis study describes the challenges, learnings and, moreover, the importance of conducting real-life implementation research to evaluate health programmes to transform school settings into healthy learning environments for children.\nThe study is retrospectively registered with the German Clinical Trials Register, University of Freiburg (Trial registration number: DRKS00004485, date of registration: 26th of February, 2013).\nBackground\nImprovements of Water, Sanitation and Hygiene (WASH) are fundamental to promote child health in low and middle-income countries.\nWater scarcity, limited access to improved sanitation and lack of personal hygiene at home and in school significantly contribute to the immense burden of preventable childhood diseases, such as diarrhoea, acute respiratory infections, intestinal worms and dental caries.\nThese hygiene-related illnesses add to a vicious cycle of poverty and disease through their adverse impacts on children\u2019s school attendance, educational performance and productivity.\nImproving WASH in Schools (WinS) is a key intervention to increase children\u2019s prospects for a healthy development.\nIt contributes to a safe and healthy learning environment and is a prerequisite for teachers and students to develop and practice positive hygiene habits.\nWinS has gained increasing attention on political agendas, particularly in the development sector, as evidenced by the inclusion of WinS targets and respective indicators as part of the Sustainable Development Goals (SDGs).\nThe growth of the WinS sector is also visible in the area of research.\nThe benefit of school-based handwashing with soap is now well established; this intervention alone has been shown to prevent around one-third of diarrhoea episodes in children.\nThere is promising evidence from a few recent cluster-randomized trials that WinS programmes, such as school-based hygiene promotion, water treatment and improved sanitation, are effective in reducing pupil absenteeism by 21% to 58%, in some cases specifically for girls.\nHowever, there are not many studies evaluating the benefits of WinS programmes on children\u2019s health outcomes, which are generally challenging to measure.\nStrong evidence for such health effects, together with research on best methods of integrating WASH into school health programmes, would allow for stronger advocacy and foster the adoption of appropriate WASH policies within the education sector.\nThe Fit for School (FIT) approach is an integrated school health promotion and WinS concept, which has been developed as a response to the serious health problems of Southeast Asian children.\nThe FIT approach aims to improve child health through the institutionalization of a combination of simple, evidence-based preventive interventions, including improvement of WASH facilities, daily practice of group hygiene activities and school-based deworming.\nThe implementation of the FIT approach is conceptually based on the principles of \u2018simplicity\u2019, \u2018scalability\u2019, \u2018sustainability\u2019 and \u2018system-thinking\u2019, which address the concepts of intersectoral collaboration, sustainable financing mechanisms, active community involvement and the strengthening of school-based management.\nThe FIT approach originated in the Philippines in 2007/2008 where it was implemented by the Department of Education as the \u2018Essential Health Care Programme\u2019.\nIn 2011, commissioned by the German Federal Ministry for Economic Cooperation and Development, the Deutsche Gesellschaft f\u00fcr Internationale Zusammenarbeit (GIZ) and the South-East Asian Ministers of Education Organization Regional Centre for Educational Innovation and Technology (SEAMEO INNOTECH) partnered to expand the FIT approach to Cambodia, Indonesia and Lao PDR as the \u2018Regional Fit for School Programme\u2019.\nThe Regional FIT Programme supported the respective Ministries of Education (MoEs) to adapt the concept to local conditions and start implementation in model schools during a pilot phase from 2012 to 2014.\nThe piloting of the programme was accompanied by an extensive Fit for School Programme Assessment Study, which comprised three study components - a WASH survey, a behaviour study and a Health Outcome Study (HOS).\nThis paper describes the findings of the HOS, which aimed to assess the two-year-effect of the FIT interventions on the parasitological, weight and oral health status of children attending schools implementing the FIT programme in Cambodia, Indonesia and Lao PDR.\nThe hypothesised health outcomes of the study, supported by available evidence, are presented in Fig.\n1.\nThe findings of the WASH survey and the behaviour study will be reported in separate papers.\nMethods\nStudy design\nDaily handwashing with soap as a group activity,\nDaily toothbrushing with 0.3\u00a0ml of toothpaste (containing 1450\u00a0ppm free available fluoride) as a group activity,\nBiannual deworming with a single dose of albendazole or mebendazole (400\u00a0mg tablet) as part of the respective national government-coordinated deworming programme.\nThe study was designed as a non-randomized clustered controlled trial.\nIt describes a longitudinal cohort of children that were followed-up for a period of two years.\nThe intervention group consisted of public elementary schools implementing the FIT programme interventions, including:\nAccess to water and soap is a prerequisite for the practice of these daily hygiene activities.\nTherefore, the FIT programme supported the construction and maintenance of group washing facilities, which serve as a starting point for stepwise improvement of other aspects of WinS, such as availability of appropriate sanitation facilities.\nGroup washing facilities consisted of prefabricated facilities containing several water slots to accommodate many students for group hygiene activities.\nEducational staff in the intervention schools received practical guidance and training materials, but no further support to implement the programme activities.\nThe control group included public elementary schools that implemented nothing else than the regular government health education curriculum and biannual deworming as part of the national deworming programme.\nThe national deworming programme has been implemented since 2004, 1999, and 2005 in Cambodia, Indonesia and Lao PDR respectively.\nBaseline data were collected in 2012 - two weeks before the implementation of the FIT programme - and follow-up data were collected 24\u00a0months later in 2014.\nThe study\u2019s original methodology and protocol was developed in the Philippines in 2009.\nThe HOS in Cambodia, Indonesia and Lao PDR followed a similar methodology.\nStudy sample and procedure\nThe study included a total of 41 intervention schools implementing the FIT programme: 10 schools in Cambodia (Pnomh Penh, and the provinces Kampot, Takeo, Kampong Thom and Kampong Chhnang), 9 schools in Indonesia (Bandung City and Indramayu) and 22 schools in Lao PDR (Vientiane Capital and surroundings).\nSelection of the intervention schools was done by the respective MoEs on the basis of accessibility and support from the school administration.\nFor each intervention school, the nearest public elementary school with the same classification according to the size of the school (school population) was assigned as a control school.\nA power calculation indicated that samples of 600 children per country (300 children per group) were required.\nThe sample size was based upon detecting a 20% difference in mean caries increment between intervention and control schools after a 24-month period with a statistical power of 80% and a significance level of 5%.\nThis would also provide adequate power to detect a 15% difference in the proportion of thin children and children with soil-transmitted helminth (STH) infection.\nThe sample size was increased to 720 children per country to allow for a drop-out rate of 20% without compromising on the statistical power.\nPer school, a random selection of 36, 40 and 17 six- to seven-year-old children was drawn from the list of enrolled grade-one students for the baseline study in Cambodia, Indonesia and Lao PDR, respectively.\nThe same children were re-examined after 24\u00a0months.\nConsent for their participation was secured from the parents or guardians by school representatives.\nChildren with no parental or guardian consent were excluded from the study.\nData collection\nIn each country, data collection was performed by a team of local researchers from partner institutions, including the MoEs, the oral health and health offices of the Ministries of Health (MoHs), the Faculty of Dentistry of Universitas Padjadjaran in Indonesia and the University of Health Sciences in Lao PDR.\nPrior to data collection, examiners participated in a three-day training on standardised data collection methods and a calibration process.\nAll examiners were blind to which group the schools belonged.\nData collection took place on the school ground following a standard procedure: registration and stool specimen collection, anthropometric measurement, oral examination and a socio-demographic interview.\nParasitological examination\nChildren submitted a stool sample on the day of data collection.\nWithin the same day, labelled stool specimens were brought to the MoH Centre for Malaria, Parasitology and Entomology laboratory in Lao PDR and the West Java Provincial Health Laboratory and the Indramayu District Health Laboratory in Indonesia for examination.\nIn Cambodia stool specimens were directly examined on the school ground by staff from the MoH Centre for Malaria, Parasitology and Entomology laboratory.\nSamples were examined to determine the presence and intensity of STH infection (Ascaris species, hookworm and Trichuris species) using the Kato-Katz method.\nCut-off points defined by the WHO were used to classify light-, moderate-, and heavy-intensity infections.\nTen percent of stool samples were re-examined by a reference microscopist for quality control.\nAnthropometric measurement\nChildren\u2019s weight and height were measured in duplicate following standards described by Cogill\nThe average of two measurements was recorded.\nA SECA digital weighing scale (calibrated at the beginning of each day and after every 10th child) was used to measure weight to the nearest 0.1\u00a0kg.\nStanding height was measured to the nearest 0.1\u00a0cm using a microtoise.\nBody mass index (BMI) was computed as weigh/height2 (kg/m2) and converted to BMI for age z-scores using the 2007 WHO Growth reference for school-aged children and adolescents.\nThinness and overweight were defined as a BMI for age below and above 2SDs from the WHO growth reference median, respectively.\nOral examination\nIn each country, four calibrated dentists performed oral examinations in the schoolyard or inside a classroom to collect data on oral health status.\nOral health status referred to dental caries experience and the presence of odontogenic infections, which are the two most common oral diseases among children.\nChildren were placed in supine position on a classroom bench, table or series of chairs with their heads on a pillow placed on the lap of the dentist.\nMouth mirrors with illumination (Mirrorlite) and a CPI-ball-end probe were used to score dental caries according to WHO Basic Methods for Oral Health Surveys.\nDental caries experience was expressed as the dmft/DMFT-index by calculating the sum of decayed (d/D), missing (m/M) and filled (f/F) teeth (t/T).\nOdontogenic infections were scored according to the criteria of the pufa/PUFA-index, which records the presence of open pulp (p/P), ulceration (u/U), fistula (f/F) and abscesses (a/A).\nFor both indexes, uppercase letters indicate the permanent dentition, and lowercase letters indicate the primary dentition.\nKappa-scores for inter-examiner reliability ranged from 0.73 to 0.97 (mean k\u00a0=\u00a00.87) for dmft/DMFT and from 0.58 to 1.00 (mean k\u00a0=\u00a00.78) for pufa/PUFA.\nCovariates\nAll children completed an interview-questionnaire in native language to collect demographic information, including date of birth and gender.\nSocioeconomic status (SES) was assessed using six questions as proxy-indicators: \u2018Do you have a TV at home?\u2019 (Yes/No), \u2018Do you have a car at home?\u2019 (Yes/No), How many brothers do you have?\u2019, \u2018How many sisters do you have?\u2019, \u2018Did you eat breakfast today?\u2019 (Yes/No) and \u2018Did you eat lunch yesterday?\u2019 (Yes/No).\nOnly the number of siblings (family size) was used as an indicator of SES in this study, since the other variables showed little variance.\nChildren were also asked about the presence of mouth problems and abdominal pain at the time of examination.\nData from the WASH survey were used to obtain information on school characteristics, including the number of enrolees per school and the number of handwashing facilities with water and soap available.\nIn addition, data on sanitation were collected, including the number, functionality and cleanliness of toilets, as proxy indicators of school maintenance and cleanliness in general, and as a covariate since there is evidence that access and cleanliness of sanitation facilities influences children\u2019s parasitological health.\nData in the WASH survey were collected at baseline and follow-up through observations using an adapted version of the UNICEF WASH in Schools Monitoring Observational Tool.\nPer country, two researchers of the local research team were trained to conduct the WASH survey in the schools.\nToilets were scored as clean, partially clean or not clean, and as functional, partially functional and not functional.\nToilets were subsequently classified as both clean and functional, partially clean and/or functional, or not clean and/or functional.\nData from the WASH survey were solely used to describe the schools in the study sample, and for inclusion as potentially important covariates in the analysis of children\u2019s health outcomes.\nThe full findings of the WASH survey will be reported in a separate paper.\nStatistical analysis\nData were analyzed using STATA 13 (Stata Corp, College Station, Texas, USA).\nA P-value of <0.05 was regarded as significant.\nComplete case analysis was used to handle missing data.\nDifferences in health outcomes between intervention and control schools were analyzed for each country separately, and for the overall sample.\nThe \u03c72-test was used to assess differences in the prevalence of STH infection, the prevalence of thinness and the prevalence of dental caries and odontogenic infections (all expressed in percentages); the Mann Whitney U-test was used to assess differences in the mean DMFT increment and PUFA increment.\nThe preventive fraction for DMFT was calculated, which is the difference in mean DMFT increment between the intervention and control schools expressed as a percentage of the mean DMFT increment in the control group.\nSince existing dental caries lesions cannot disappear or decrease through intervention, and primary teeth are exfoliating at the age of children\u2019s examination, the analysis of caries progression and odontogenic infection was limited to the permanent dentition only.\nFurthermore, multilevel logistic and linear regression analyses with backward selection were performed to explore which factors (including the FIT programme) were associated with STH infection at follow-up (no infection vs infection), thinness at follow-up (normal weight or overweight vs thinness) and DMFT increment between baseline and follow-up (continuous).\nVariables considered in the models were the FIT programme (control schools vs. intervention schools), sociodemographic characteristics and health parameters of the children, and school characteristics.\nBecause children (first level) were nested in schools (second level), which were in turn nested in the three countries (third level), multilevel analyses were used to control for the possible effect of clustered differences within the sample.\nFor each model, the intraclass correlation coefficient (ICC) was calculated to indicate the percentage of total variance that was due to differences between schools or the countries.\nIn eight schools in Lao PDR (four intervention and four control schools), the simultaneous implementation of an oral health programme from the Japanese International Cooperation Agency, providing dental treatment (restorations) and fluoride rinsing, interfered with the implementation of the Fit for School programme without knowledge of the study investigators.\nDuring the follow-up data collection the fact was revealed and it was decided to exclude these eight schools from the analysis of oral health outcomes.\nEthical approval\nThe study received ethical approval from the National Ethics Committee for Health Research of the MoHs in Cambodia and Lao PDR, and from the Health Research Ethics Committee of the University of Padjadjaran, Indonesia.\nThe study is registered with the German Clinical Trials Register, University of Freiburg (Trial registration number: DRKS00004485, date of registration: 26th of February, 2013).\nParents of participating children provided written informed consent.\nResults\nCharacteristics of the study sample\nA total of 1847 children participated in the baseline survey.\nOf these, 1499 children were available for follow-up examination \u2013 478 children in Cambodia (241 in intervention schools and 237 in control schools), 486 children in Indonesia (248 in intervention schools and 238 in control schools) and 535 children in Lao PDR (279 in intervention schools and 256 in control schools).\nThe follow-up rate was 76.6%, 85.3% and 81.0% in Cambodia, Indonesia and Lao PDR, respectively, with an average follow-up rate of 81.2%.\nParasitological, anthropometric and oral health parameters of the dropout children were similar to those children who were followed-up.\nThe mean time interval between baseline and follow-up was 23.9\u00a0\u00b1\u00a00.3\u00a0months.\nThe child characteristics of the study sample are described in Table 1.\nThe mean age of the children at baseline was 6.7\u00a0\u00b1\u00a00.5\u00a0years (range 6.0\u20138.0\u00a0years) in intervention schools and 6.8\u00a0\u00b1\u00a00.5\u00a0years (range 6.0\u20138.0\u00a0years) in control schools (P\u00a0<\u00a00.05), and 48.4% and 53.9% were boys in intervention and control schools, respectively (P\u00a0<\u00a00.05).\nAround one-third of children came from large families with three or more siblings \u2013 a proxy indicator of lower SES.\nTable 2 describes the characteristics of the intervention and control schools.\nAll schools in Indonesia were located in urban areas; Cambodia and Lao PDR also included schools in rural provinces.\nIn all three countries, the mean number of handwashing slots with water and soap was significantly higher in intervention schools than in control schools, due to the construction of group washing facilities as part of the programme implementation.\nConsequently, substantially less children had to share one water slot in intervention schools compared to the control schools (Table 2).\nAccess to toilets was similar in intervention and control schools.\nHowever, toilet conditions in terms of functionality and cleanliness were slightly better in intervention schools, although this was only statistically significant for Lao PDR and the overall sample.\nParasitological status\nThe prevalence of STH infection in the overall sample was 8.1% at baseline, and this remained the same at follow-up.\nLess than 1% of the children had moderate to heavy STH infection.\nIn all three countries, the STH prevalence at baseline and at follow-up did not significantly differ between intervention schools and control schools (Table 3).\nHookworm accounted for more than 80% of the STH infection.\nTable 4 shows the factors that were significantly associated with STH infection at follow-up.\nChildren with STH infection at baseline were nine times more likely to be (re-)infected at follow-up.\nThe odds of STH infection were also higher for older children and children from larger families, while children who attended schools in urban areas had lower odds of STH infection.\nEvery 10 % increase in the percentage of clean and functional toilets at school was associated with 0.91 (0.83; 1.00) lower odds of STH infection at follow-up.\nIn other words, the odds of having STH infection were more than two times higher for children in schools with zero clean and functional toilets compared to children from schools where all toilets are fully clean and functional (1/(OR10)\u00a0=\u00a01/(0.9110)).\nThe school level random effects variance and ICC show that 15.6% of the variance in STH infection at follow-up occurred between schools.\nWeight status\nAt baseline, the prevalence of thinness in the overall sample was 8.7% and this increased to 11.6% at follow-up.\nIn Cambodia, both intervention and control schools showed a significant increase in the percentage of thin children between baseline and follow-up, but in Indonesia and Lao PDR the prevalence of thinness remained stable over the 2-year period.\nIn all three countries, the prevalence of thinness did not significantly differ between intervention and control schools at baseline, nor at follow-up (Table 3).\nThe prevalence of overweight increased from 3.4% to 5.6% in Cambodia (P\u00a0=\u00a00.002), from 13.9% to 22.7% in Indonesia (P\u00a0<\u00a00.001) and from 5.5% to 8.8% in Lao PDR (P\u00a0=\u00a00.001).\nTable 5 shows the factors that were significantly associated with thinness at follow-up.\nChildren who were thin at baseline were 57 times more likely to remain thin at follow-up.\nThe odds of being thin at follow-up were also higher for children who were stunted at baseline and for children with more decayed, missing and filled teeth (DMFT) at follow-up, while the odds were lower for children attending urban schools.\nThe school- and country-level variance was close to zero, which indicates that thinness at follow-up was independent of the children\u2019s school or country.\nOral health status\nAt baseline, 94.4% of the children in the overall sample had dental caries in the primary dentition (with a mean dmft of 9.2\u00a0\u00b1\u00a04.4) and 73.2% of them had odontogenic infections (with a mean pufa of 3.8\u00a0\u00b1\u00a02.6).\nThe baseline oral health status of the permanent dentition was comparable between children from intervention schools and control schools (Table 3).\nIn all three countries, children in intervention schools had a lower prevalence of dental caries in the permanent dentition at follow-up and a lower increment in DMFT between baseline and follow-up in comparison to children in controls schools, although these were only statistically significant in the overall sample.\nThe preventive fraction for DMFT was 23.9% in the overall sample, and 18.3%, 22.4%, 38.0% in Cambodia, Indonesia and Lao PDR, respectively (Table 3).\nThere were no significant differences in the prevalence of odontogenic infections and PUFA increment between intervention and control schools.\nTable 6 describes the factors that were significantly related with DMFT increment.\nThe DMFT increment was significantly lower in intervention schools compared to control schools in the full model.\nChildren who had more permanent teeth at baseline, younger children and children attending schools in urban areas had higher increment in DMFT between baseline and follow-up.\nThe random effects variance and ICC show that 14.3% of the variance in DMFT increment occurred between schools and 9.9% between countries.\nDiscussion\nWASH and school health programmes have strong potential to contribute to better health of children inlow and middle-income countries.\nThe essence of the FIT programme lies in the institutionalization of a package of simple interventions within the education sector to establish hygiene habits and address some of the most prevalent childhood diseases in Southeast Asia.\nThe FIT interventions \u2013 namely handwashing with soap, toothbrushing with fluoride toothpaste, biannual deworming and improved WASH infrastructure \u2013 are all underpinned by ample evidence for their respective effectiveness for improving child health in controlled settings.\nStill, it is critical to conduct programme evaluation and impact research of such proven health and WASH interventions under real-life conditions where there is typically less or no control of possible cofounding factors.\nA government-run school programme, implemented by education staff without further external support provides such a real-life setting.\nResults of such research help to understand factors that facilitate the translation of the evidence from controlled settings into realistic public health promotion strategies, and, ideally, whether they can be applied at scale.\nThis study evaluated the 2-year effect of the FIT programme on parasitological, weight and oral health status in children attending elementary schools in Cambodia, Indonesia and Lao PDR participating in the Fit for School Programme.\nThe study found that the FIT interventions significantly reduced the development of dental caries in children (by 24% at average).\nDue to specific circumstances that were not anticipated prior to the study, no significant decreases in the prevalence of STH infection and thinness were observed, which does not mean that the interventions did not work.\nThe following paragraphs discuss this aspect in more detail.\nDiscussion of parasitological results\nThe FIT programme was expected to improve children\u2019s parasitological health via three mechanisms:.\nbiannual administration of albendazole or mebendazole is an efficacious method to treat existing worm infection, handwashing with soap interrupts transmission of helminthiases from contaminated soil or infected faeces and improved access to WASH contributes to a reduction of helminthiases in the school environment.\nThis was supported by findings from the former HOS in the Philippines, where the FIT programme led to a significant reduction in the prevalence of moderate to heavy STH infection.\nHowever, this study did not show a similar effect.\nThe general prevalence of STH infection was surprisingly low at baseline (unlike in the Philippines), which did not leave much room for further improvement.\nThe low baseline prevalence was unexpected in view of previously reported estimates of the parasitic disease burden in Southeast Asian countries, and could be indicative of effective implementation of the already existing national deworming programmes.\nYet, it should be noted that the prevalence rates in this study are clearly not a representative reflection of national prevalence estimates, which revealed prevalences up to 86%, 90% and 66% in Cambodia, Indonesia and Lao PDR, respectively (with a mean prevalence of approximately 30%), and that other geographical areas within the countries may show different results.\nInterestingly, this study found that children with STH infection at baseline were nine times more likely to be re-infected at follow-up despite the mass drug administration scheme.\nThis corresponds with previous studies reporting that anthelmintic drugs only provide a temporal reduction in morbidity, but do not prevent rapid reinfection.\nThis emphasises the need for complementary environmental STH control interventions, including improvement of sanitation (e.g. access to clean latrines and latrine maintenance) and interventions that promote hygiene habit formation in schools and in the home environment (e.g. handwashing with soap prior to eating and after defecation).\nDiscussion of weight status results\nThis study did not find a programme effect on children\u2019s weight status, which is not entirely surprising in view of the parasitological findings discussed above.\nWeight status was chosen as a relevant health indicator under the assumption that successful treatment of worm infection would enhance weight gain (\u201ccatch-up\u201d).\nThis was based on the Philippine HOS and other studies, where children\u2019s BMI significantly increased after 1\u00a0year of regular deworming, possibly as a result of the substantial reduction in moderate to heavy worm infestation, although the actual evidence for this mechanism is not strongly supported by literature.\nCircumstances in the current study were unexpectedly different, with less than 1% of children suffering from heavy worm infection at baseline and follow-up, so that no programme effect on parasitic health status was found.\nThe major causes of thinness are poor nutrition and infectious diseases, including diarrhoea.\nHandwashing with soap and safe sanitation are obvious measures to prevent infectious diseases, although the evidence of their effect on weight status and child growth is weak due to a lack of high quality randomised trials.\nLiterature suggests that interventions to promote hand hygiene, water and sanitation are necessary, but they do not have sufficient impact to address the chronic and persistent burden of underweight and thinness if they are not combined with efforts to tackle the root causes, including nutritional interventions and changes to the broader living environment.\nSevere dental caries is another strong determinant of underweight, as also confirmed by the findings of this study, which suggests that prevention and treatment of dental decay should be considered among priority interventions addressing the burden of malnutrition.\nNotably, data of this study suggest that obesity is on the rise in Southeast Asian countries, in particular in Indonesia.\nThe emergence of obesity as a major public health problem in developing countries has been frequently reported.\nThis paradoxical coexistence of both, childhood obesity and childhood undernutrition (also termed as the \u201cdouble burden of disease\u201d), will have important implications for the planning and redirection of health promotion strategies in the school context.\nA healthy nutrition environment at school (with nutrition choices low in sugar and fat), combined with regular physical activity, could be effective strategies to address obesity in school health programmes.\nDiscussion of oral health results\nThe daily toothbrushing intervention of the FIT programme significantly contributed to the prevention of dental caries.\nThe prevented fraction was 24% in the overall sample, ranging from 18% in Cambodia, 22% in Indonesia to 38% in Lao PDR.\nThe findings are in concordance with findings from a Cochrane meta-analysis on the caries-preventive benefits of fluoride toothpaste in children.\nThe difference in prevented fraction between individual countries, might be attributed to implementation quality.\nConsidering that the toothbrushing intervention was only performed once daily on schooldays, more dental caries may be prevented if the frequency of toothbrushing is increased to at least twice daily according to the recommendations of evidence-based guidelines.\nThe study results of all countries highlight that the burden of oral diseases in the Southeast Asian region is extremely high.\nMore than 70% of children had severe dental infection, which can lead to eating and sleeping impacts, poor quality of life, school absenteeism and growth retardation.\nIn order to reduce this neglected burden of oral diseases, full integration of oral health into public health promotion strategies and school health programmes is essential.\nThe primary focus should be on preventing the disease via regular toothbrushing with fluoride toothpaste, since it is a proven, simple and realistic intervention that does not require involvement of health professionals.\nMost important is the development of regular habits for toothbrushing.\nNext level interventions may include additional fluoridation, such as the use of fluoride gel, and oral urgent treatment to address pain and suffer among the children.\nChallenges of real life implementation research and implications for future research\nLimitations of the study design\nReal-life implementation research brings along challenges in study design and the extent to which research conditions can be controlled, which should be considered in the interpretation of the study findings.\nFirst, randomisation of intervention and control schools was not possible.\nIntervention schools were purposively selected by the MoEs with the primary aim to test the FIT approach in a local context, and to create a country-specific model template for implementation and scale-up.\nThe purposive sampling method may have introduced selection bias towards more favourable school conditions in intervention schools.\nThe potential effect of this bias was minimized by assigning control schools with nearest location to the intervention schools, so that similar socioeconomic parameters could be assumed.\nHowever, this made the study more vulnerable to the effects of unplanned crossover or \u2018spontaneous programme scale-up\u2019 as, in fact, there were a few control schools that were implementing the FIT interventions on their own initiative.\nSecondly, an unplanned and previously un-reported overlapping oral treatment programme by Japanese International Cooperation Agency in Lao PDR interfered with the evaluation of the FIT programme, and therefore eight schools had to be excluded from the oral health analysis.\nChallenges in implementation\nThe success of the FIT programme fully depends on the quality of programme delivery, which is built upon commitment and capacity of teaching staff, school leadership and local participation.\nNo additional programme staff interacts with or supports the school staff on a daily basis.\nEven if interventions are proven highly effective, their full potential will not be reached if adherence to the protocol and implementation is poor.\nAlthough implementation quality was not assessed in the study, the data strongly suggest that there were big differences in implementation among schools, as illustrated by the example of oral health results.\nThe mean DMFT increment in the overall intervention sample was 0.48\u00a0\u00b1\u00a00.91, however, there was wide variation between the intervention schools.\nIn some intervention schools, the mean DMFT increment was close to zero (minimum: 0.07\u00a0\u00b1\u00a00.45), while in others the mean DMFT increased by more than 1 (maximum: 1.39\u00a0\u00b1\u00a01.20) (results not shown).\nThe DMFT increment also greatly differed between countries, with Lao PDR having the lowest caries increase on average and Cambodia the highest (Table 3).\nAll schools were supposed to implement the same intervention of once-daily toothbrushing with fluoride toothpaste, yet these results clearly show that DMFT increase greatly varied within schools and countries.\nThis points to differences in programme compliance, since the interventions\u2019 potential effect should in principle be similar irrespective of the setting, as long as the intervention is the same.\nTherefore, it is of utmost importance that school management structures and compliance with the protocol are considered in the implementation of the FIT programme or any other programme, particularly if health effects are to be achieved.\nGood implementation requires leadership within the education sector and relies on a collaborative effort between the local government, the school administration, teaching staff and the local community for which roles and responsibilities need to be clearly defined.\nChoice of health indicators\nSTH infection, weight status and dental caries were chosen as key indicators for the health effect evaluation of the FIT programme, based on positive experiences from the former HOS in the Philippines and available evidence.\nThese are relevant health outcomes from a programme and advocacy perspective.\nHowever, in retrospect, these indicators also have limitations to evaluate the programme\u2019s potential health effect, considering the previously described implementation challenges and the fact that these health outcomes are difficult to change within the study\u2019s short time interval.\nTherefore, it is important that future programme evaluation research also explores options of including relevant intermediary health outcomes in addition to health indicators, such as hygiene behaviour change, rather than solely measuring conventional health indicators.\nHealthy hygiene behaviours are key determinants of child health, and their benefits are most pronounced when they are practiced habitually on a regular basis.\nSuccessful and sustained incorporation of healthy hygiene habits in the lives of children would provide lifelong health benefits, which can even be transferred to their families, friends and communities.\nA valuable area for future research would therefore be to assess whether WinS and health programmes, including the FIT programme, are successful in the formation of children\u2019s healthy hygiene habits.\nConclusions\nThis study evaluated the two-year effect of the Fit for School programme on relevant health outcomes of children in Cambodia, Indonesia and Lao PDR.\nIt found that the toothbrushing intervention significantly contributed to the prevention of dental caries in children.\nA clear asset of the study was that it describes real-life implementation research to assess whether a combination of relevant and already proven health and WASH interventions is effective in improving child health when delivered in schools as an integrated hygiene promotion package.\nMoreover, the results and their interpretation clearly highlight that effect evaluation research of WinS and health programmes encounters many challenges.\nThese include restrictions in randomisation, the potential of crossover effects, challenges related to implementation quality, and unforeseen conditions that are beyond the researchers\u2019 control, such as the interference of other health programmes.\nThese challenges make it difficult to demonstrate the programme\u2019s full potential effect, and this likely explains why no direct effect on weight status and STH infection was observed.\nHerewith, the study provides important learnings for future evaluation research, which points the way forward for also incorporating intermediary measures of behavioural outcomes and indicators of implementation quality, in addition to health indicators, in order to evaluate and understand how WinS programmes possibly lead to health benefits through implementation processes and their potential effect on hygiene habit formation.\nThe study suggests that even the most effective and simplest of health interventions, such as toothbrushing with fluoride toothpaste, handwashing with soap or deworming plausibly depend on implementation quality to reach their full beneficial potential.\nThe traditional complexity of school health and WinS with multiple cross-sectoral roles and responsibilities calls for a governance and management simplification under the education sector\u2019s leadership.\nAs much as the education sector has been able to improve schooling rates and quality of education, it is overdue that ownership, governance and financing for school health and WinS are seen equally important for children\u2019s health and education attainment.\nThis includes regular monitoring and evaluation of WinS and School Health implementation quality through Education Management and Information Systems and other surveillance tools, as well as a supportive policy context.\nSchools as health-promoting settings can only be effective in achieving better hygiene behaviour, in providing preventive health or WASH services if they are able to manage, monitor and finance such services sustainably and consistently according to government guidelines.\nAchieving the ambitious targets of the SDGs in health, education and WASH will only be realistic with such a paradigm shift.\nHypothesised health outcomes of the Fit for School programme, based on available evidence. Grey boxes represent hypothesised health outcomes resulting from the FIT programme interventions. White boxes with dashed lines represent intermediate health outcomes that were not assessed in this study. White boxes with dotted lines represent intermediate behavioural outcomes that were not assessed in this study. Summary of related evidence: a Biannual deworming reduces the prevalence and severity of intestinal worm infection. b, c, d, e Handwashing with soap and improved access to WASH have been associated with lower prevalence of STH infections and other infectious diseases, such as diarrhoea. f, g Lower prevalence of worm infection and diarrhoea have been associated with weight gain. h Toothbrushing with fluoride toothpaste prevents dental caries and odontogenic infections. i Lower prevalence of dental caries and odontogenic infections are associated with lower prevalence of thinness and weight gain \n\nChild characteristics of the study sample in Cambodia, Indonesia, Lao PDR and the pooled regional countries\n | Cambodia | Indonesia | Lao PDR | Regional (pooled)\n | FIT(n\u00a0=\u00a0241) | Control(n\u00a0=\u00a0237) |  | FIT(n\u00a0=\u00a0248) | Control(n\u00a0=\u00a0238) |  | FIT(n\u00a0=\u00a0279) | Control(n\u00a0=\u00a0256) |  | FIT(n\u00a0=\u00a0768) | Control(n\u00a0=\u00a0731) | \nChild characteristics | mean\u00a0\u00b1\u00a0sd | mean\u00a0\u00b1\u00a0sd | P* | mean\u00a0\u00b1\u00a0sd | mean\u00a0\u00b1\u00a0sd | P* | mean\u00a0\u00b1\u00a0sd | mean\u00a0\u00b1\u00a0sd | P* | mean\u00a0\u00b1\u00a0sd | mean\u00a0\u00b1\u00a0sd | P*\nAge at baseline | 6.6\u00a0\u00b1\u00a00.4 | 6.7\u00a0\u00b1\u00a00.5 | 0.143 | 6.8\u00a0\u00b1\u00a00.4 | 6.8\u00a0\u00b1\u00a00.4 | 0.129 | 6.7\u00a0\u00b1\u00a00.5 | 6.8\u00a0\u00b1\u00a00.6 | 0.026 | 6.7\u00a0\u00b1\u00a00.5 | 6.8\u00a0\u00b1\u00a00.5 | 0.003\n | n (%) | n (%) |  | n (%) | n (%) |  | n (%) | n (%) |  | n (%) | n (%) | \nGender |  |  | 0.780 |  |  | 0.100 |  |  | 0.088 |  |  | 0.034\n\u2003Boys | 122 (50.6) | 123 (51.9) |  | 118 (47.6) | 131 (55.0) |  | 132 (47.3) | 140 (54.7) |  | 372 (48.4) | 394 (53.9) | \n\u2003Girls | 119 (49.4) | 114 (48.1) |  | 130 (52.4) | 107 (45.0) |  | 147 (52.7) | 116 (45.3) |  | 396 (51.6) | 337 (46.1) | \nFamily sizea, b |  |  | 0.115 |  |  | 0.350 |  |  | 0.751 |  |  | 0.166\n\u20031 or no siblings | 83 (34.4) | 61 (25.7) |  | 131 (53.0) | 122 (51.5) |  | 108 (38.7) | 91 (35.6) |  | 322 (42.0) | 274 (37.5) | \n\u20032 siblings | 64 (26.6) | 73 (30.8) |  | 77 (31.2) | 66 (27.9) |  | 95 (34.1) | 92 (35.9) |  | 236 (30.8) | 231 (31.6) | \n\u20033 or more siblings | 94 (39.0) | 103 (43.5) |  | 39 (15.8) | 49 (20.7) |  | 76 (27.2) | 73 (28.5) |  | 209 (27.2) | 225 (30.8) | \n\n* \u03c72-test\n\na Measured at follow-up\n\nb Missing values: Cambodia: 0, Indonesia: 2, Lao PDR: 1\n\nCharacteristics of the schools in the study sample in Cambodia, Indonesia, Lao PDR and the pooled regional countries\n | Cambodia | Indonesia | Lao PDR | Regional (pooled)\nSchool characteristicsa | FIT(n\u00a0=\u00a010) | Control(n\u00a0=\u00a010) | P* | FIT(n\u00a0=\u00a09) | Control(n\u00a0=\u00a09) | P* | FIT(n\u00a0=\u00a022) | Control(n\u00a0=\u00a022) | P* | FIT(n\u00a0=\u00a041) | Control(n\u00a0=\u00a041) | P*\nGeographical location (n)\n\u2003Rural school | 6 | 6 |  | - | - |  | 7 | 7 |  | 13 | 13 | \n\u2003Urban school | 4 | 4 |  | 9 | 9 |  | 15 | 15 |  | 28 | 28 | \nNo. of enrollees(mean n\u00a0\u00b1\u00a0sd) | 805\u00a0\u00b1\u00a0384 | 723\u00a0\u00b1\u00a0228 | 0.999 | 592\u00a0\u00b1\u00a0323 | 388\u00a0\u00b1\u00a0220 | 0.200 | 219\u00a0\u00b1\u00a0124 | 118\u00a0\u00b1\u00a064 | 0.003 | 505\u00a0\u00b1\u00a0346 | 391\u00a0\u00b1\u00a0295 | 0.046\nNo. of handwashing slots w/ water & soap(mean n\u00a0\u00b1\u00a0sd) | 200\u00a0\u00b1\u00a0128 | 9\u00a0\u00b1\u00a016 | 0.001 | 89\u00a0\u00b1\u00a057 | 1\u00a0\u00b1\u00a02 | <0.001 | 113\u00a0\u00b1\u00a069 | 17\u00a0\u00b1\u00a038 | <0.001 | 129\u00a0\u00b1\u00a093 | 11\u00a0\u00b1\u00a030 | <0.001\nStudent to handwashing slot ratio(mean ratio\u00a0\u00b1\u00a0sd) | 4:1\u00a0\u00b1\u00a01:1 | 55:1\u00a0\u00b1\u00a046:1 | <0.001 | 6:1\u00a0\u00b1\u00a01:1 | 74:1\u00a0\u00b1\u00a071:1 | <0.001 | 2:1\u00a0\u00b1\u00a02:1 | 66:1\u00a0\u00b1\u00a063:1 | 0.001 | 4:1\u00a0\u00b1\u00a02:1 | 65:1\u00a0\u00b1\u00a060:1 | <0.001\nStudent to toilet ratio(mean ratio\u00a0\u00b1\u00a0sd) | 93:1\u00a0\u00b1\u00a056:1 | 102:1\u00a0\u00b1\u00a053:1 | 0.496 | 99:1\u00a0\u00b1\u00a050:1 | 112:1\u00a0\u00b1\u00a061:1 | 0.691 | 63:1\u00a0\u00b1\u00a044:1 | 45:1\u00a0\u00b1\u00a024:1 | 0.218 | 79:1\u00a0\u00b1\u00a050:1 | 74:1\u00a0\u00b1\u00a052:1 | 0.593\nPercentage of clean & functional toilets(mean %\u00a0\u00b1\u00a0sd) | 7.6\u00a0\u00b1\u00a014.0 | 0.0\u00a0\u00b1\u00a00.0 | 0.068 | 62.2\u00a0\u00b1\u00a041.8 | 36.1\u00a0\u00b1\u00a039.7 | 0.219 | 37.6\u00a0\u00b1\u00a044.1 | 15.9\u00a0\u00b1\u00a035.9 | 0.026 | 35.6\u00a0\u00b1\u00a042.0 | 16.5\u00a0\u00b1\u00a033.8 | 0.007\n\n*Mann Whitney U-test\n\na Measured at follow-up\n\nParasitological status, weight status and oral health status of children in intervention schools and control schools in Cambodia, Indonesia, Lao PDR and the pooled regional countries at baseline and follow-up\n | Cambodia | Indonesia | Lao PDR | Regional (pooled)\nFIT | Control | P* | FIT | Control | P* | FIT | Control | P* | FIT | Control | P*\n | Parasitological status\nSTH-prevalence at baseline (n, %) | 22 (9.1) | 18 (7.7) | 0.573 | 3 (1.5) | 6 (3.0) | 0.298 | 25 (9.9) | 37 (15.6) | 0.059 | 50 (7.1) | 61 (9.1) | 0.195\nSTH-prevalence at follow-up (n, %) | 34 (15.9) | 14 (9.1) | 0.056 | 2 (1.0) | 6 (3.1) | 0.120 | 22 (8.4) | 25 (10.6) | 0.419 | 58 (8.4) | 45 (7.7) | 0.630\n | Weight status\nPrevalence of thinness at baseline (n, %) | 21 (8.9) | 23 (9.9) | 0.718 | 15 (6.1) | 25 (10.6) | 0.076 | 16 (5.9) | 19 (7.6) | 0.434 | 52 (7.5) | 67 (9.9) | 0.112\nPrevalence of thinness at follow-up (n, %) | 31 (13.2) | 36 (15.5) | 0.485 | 14 (5.7) | 23 (9.7) | 0.095 | 25 (9.1) | 21 (8.3) | 0.748 | 70 (10.7) | 80 (12.5) | 0.294\n | Oral health status (permanent dentition)a\nDental caries prevalence at baseline (n, %) | 32 (13.3) | 43 (18.1) | 0.149 | 26 (10.6) | 23 (9.7) | 0.741 | 36 (15.7) | 40 (18.9) | 0.382 | 94 (13.1) | 106 (15.4) | 0.223\nDental caries prevalence at follow-up (n, %) | 126 (52.7) | 137 (58.1) | 0.243 | 75 (30.5) | 83 (35.0) | 0.288 | 68 (29.8) | 80 (38.3) | 0.062 | 269 (37.7) | 300 (44.0) | 0.017\nPrevalence of PUFA at baseline (n, %) | 1 (0.4) | 3 (1.3) | 0.3701 | 0 (0.0) | 0 (0.0) | - | 1 (0.4) | 5 (2.4) | 0.1101 | 2 (0.3) | 8 (1.2) | 0.0601\nPrevalence of PUFA at follow-up (n, %) | 17 (7.1) | 23 (9.8) | 0.302 | 21 (8.6) | 18 (7.6) | 0.704 | 10 (4.4) | 16 (7.7) | 0.149 | 48 (6.7) | 57 (8.4) | 0.265\nDMFT at baseline (mean\u00a0\u00b1\u00a0sd) | 0.18\u00a0\u00b1\u00a00.51 | 0.30\u00a0\u00b1\u00a00.73 | 0.127 | 0.15\u00a0\u00b1\u00a00.50 | 0.13\u00a0\u00b1\u00a00.44 | 0.757 | 0.28\u00a0\u00b1\u00a00.78 | 0.37\u00a0\u00b1\u00a00.90 | 0.324 | 0.20\u00a0\u00b1\u00a00.61 | 0.26\u00a0\u00b1\u00a00.71 | 0.181\nDMFT at follow-up (mean\u00a0\u00b1\u00a0sd) | 1.00\u00a0\u00b1\u00a01.20 | 1.29\u00a0\u00b1\u00a01.49 | 0.066 | 0.50\u00a0\u00b1\u00a00.92 | 0.59\u00a0\u00b1\u00a00.96 | 0.257 | 0.54\u00a0\u00b1\u00a01.06 | 0.79\u00a0\u00b1\u00a01.28 | 0.026 | 0.68\u00a0\u00b1\u00a01.09 | 0.89\u00a0\u00b1\u00a01.29 | 0.003\nDMFT increment (mean\u00a0\u00b1\u00a0sd) | 0.82\u00a0\u00b1\u00a01.07 | 0.99\u00a0\u00b1\u00a01.30 | 0.262 | 0.35\u00a0\u00b1\u00a00.72 | 0.46\u00a0\u00b1\u00a00.79 | 0.168 | 0.26\u00a0\u00b1\u00a00.81 | 0.41\u00a0\u00b1\u00a01.11 | 0.373 | 0.48\u00a0\u00b1\u00a00.91 | 0.63\u00a0\u00b1\u00a01.12 | 0.049\nPUFA at baseline (mean\u00a0\u00b1\u00a0sd) | 0.00\u00a0\u00b1\u00a00.06 | 0.02\u00a0\u00b1\u00a00.16 | 0.309 | 0.00\u00a0\u00b1\u00a00.00 | 0.00\u00a0\u00b1\u00a00.00 | - | 0.01\u00a0\u00b1\u00a00.13 | 0.02\u00a0\u00b1\u00a00.15 | 0.084 | 0.00\u00a0\u00b1\u00a00.08 | 0.01\u00a0\u00b1\u00a00.13 | 0.049\nPUFA at follow-up (mean\u00a0\u00b1\u00a0sd) | 0.10\u00a0\u00b1\u00a00.37 | 0.13\u00a0\u00b1\u00a00.42 | 0.312 | 0.13\u00a0\u00b1\u00a00.44 | 0.09\u00a0\u00b1\u00a00.34 | 0.663 | 0.08\u00a0\u00b1\u00a00.47 | 0.11\u00a0\u00b1\u00a00.42 | 0.149 | 0.10\u00a0\u00b1\u00a00.43 | 0.11\u00a0\u00b1\u00a00.40 | 0.267\nPUFA increment (mean\u00a0\u00b1\u00a0sd) | 0.09\u00a0\u00b1\u00a00.35 | 0.11\u00a0\u00b1\u00a00.38 | 0.483 | 0.13\u00a0\u00b1\u00a00.44 | 0.09\u00a0\u00b1\u00a00.34 | 0.663 | 0.07\u00a0\u00b1\u00a00.39 | 0.09\u00a0\u00b1\u00a00.41 | 0.434 | 0.10\u00a0\u00b1\u00a00.40 | 0.10\u00a0\u00b1\u00a00.38 | 0.548\nPreventive fraction (DMFT) (%) | 18.3% |  |  | 22.4% |  |  | 38.0% |  |  | 23.9% |  | \n\n\u03c72-test for dichotomous variables, Mann-Whitney U-test for continuous variables. 1 Fisher\u2019s exact test for small numbers\n\na 117 children excluded from analysis in Lao PDR because of an overlapping intervention of the Japan International Cooperation Agency\nRespective missing values in Cambodia, Indonesia and Lao PDR: parasitological status (at baseline): 3, 76, 46, (at follow-up): 110, 78, 38; weight status (at baseline): 11, 3, 11, (at follow-up): 10, 2, 9; oral health status (at baseline): 1, 2, 6, (at follow-up): 3, 3, 10\n\nFactors that are significantly associated with STH infection at follow-up in children in Cambodia, Indonesia and Lao PDR (pooled)a\n\n | Model for parasitological status (n\u00a0=\u00a01162)\nOR (95% CI) | Pa\nNo STH infection at follow-up (reference) vs. STH infection at follow-up\nChild-level variables\n\u2003Age (years) | 1.90 (1.19; 3.04) | 0.007\n\u2003Family size\n\u2003\u20031 or no siblings (39.5%) | reference | \n\u2003\u20032 siblings (32.5%) | 1.53 (0.83; 2.84) | 0.175\n\u2003\u20033 or more siblings (28.0%) | 2.06 (1.12; 3.80) | 0.020\n\u2003STH infection at baseline\n\u2003\u2003No (92.4%) | reference | \n\u2003\u2003Yes (7.6%) | 9.09 (4.98; 16.45) | <0.001\nSchool-level variables\n\u2003Geographical location\n\u2003\u2003Rural (34.9%) | reference | \n\u2003\u2003Urban (65.1%) | 0.34 (0.18; 0.64) | 0.001\nPercentage of fully clean and functional toilets (per 10%) | 0.91 (0.83; 1.00) | 0.045\nRandom effects\n\u2003Country level variance (95% CI) | 0.00 (0.00; 0.00) | ICC (%): 0.0\n\u2003School level variance (95% CI) | 0.78 (0.48; 1.26) | ICC (%): 15.6\n\nVariables considered in the initial model: Child variables: FIT programme, age at follow-up, gender, number of siblings, STH infection at baseline, School variables: geographical location, number of enrolees at follow-up, number of water slots with water and soap, student to water slot ratio, percentage of clean and functional toilets\n\naMultilevel mixed-effects logistic regression\n\nFactors that are significantly associated with thinness at follow-up in children in Cambodia, Indonesia and Lao PDR (pooled)a\n\n | Model for weight status (n\u00a0=\u00a01454)\nOR (95% CI) | Pa\nNot thin at follow-up (reference) vs. Thin at follow-up\nChild-level variables\n\u2003Thin at baseline\n\u2003\u2003No (91.9%) | reference | \n\u2003\u2003Yes (8.1%) | 57.3 (34.5; 95.0) | <0.001\n\u2003Stunted at baseline\n\u2003\u2003No (69.5%) | reference | \n\u2003\u2003Yes (30.5%) | 1.98 (1.27; 3.09) | 0.003\nDMFT at follow-up | 1.28 (1.10; 1.50) | 0.001\nSchool-level variables\n\u2003Geographical location\n\u2003\u2003Rural (33.2%) | reference | \n\u2003\u2003Urban (66.8%) | 0.60 (0.38; 0.96) | 0.032\nRandom effects\n\u2003Country level variance (95% CI) | 0.00 (0.00; 0.00) | ICC (%): 0.0\n\u2003School level variance (95% CI) | 0.00 (0.00; 0.00) | ICC (%): 0.0\n\nVariables considered in the initial model: Child variables: FIT programme, age at follow-up, gender, number of siblings, thin at baseline, stunted at baseline, DMFT at follow-up, School variables: geographical location, number of enrolees at follow-up, number of water slots with water and soap, student to water slot ratio, percentage of clean and functional toilets\n\naMultilevel mixed-effects logistic regression\n\nFactors that are significantly associated with DMFT increment in children in Cambodia, Indonesia and Lao PDR (pooled)a\n\n | Model for oral health status (n\u00a0=\u00a01395)\n\u03b2 (95% CI) | P\nDMFT increment\nFIT Programme\n\u2003No (48.9%) | reference | \n\u2003Yes (51.1%) | \u22120.15 (\u22120.29; \u22120.01) | 0.036\nChild-level variables\n\u2003Age (years) | \u22120.12 (\u22120.23; \u22120.01) | 0.04\n\u2003Number of permanent teeth at baseline | 0.04 (0.02; 0.06) | <0.001\nSchool-level variables\n\u2003Geographical location\n\u2003\u2003Rural (35.4%) | reference | \n\u2003\u2003Urban (64.6%) | 0.39 (0.22; 0.57) | <0.001\nRandom effects\n\u2003Country level variance (95% CI) | 0.10 (0.02; 0.56) | ICC (%): 10.1\n\u2003School level variance (95% CI) | 0.04 (0.02; 0.08) | ICC (%): 14.0\n\nVariables considered in the initial model: Child variables: FIT programme, age at follow-up, gender, number of siblings, number of permanent teeth at baseline, School variables: geographical location, number of enrolees at follow-up, number of water slots with water and soap, student to water slot ratio, percentage of clean and functional toilets\n\naMultilevel mixed-effects linear regression\n\nb117 children excluded from analysis because of an overlapping intervention of the Japan International Cooperation Agency in Lao PDR",
        "label": "low",
        "id": "task4_RLD_test_790"
    },
    {
        "paper_doi": "10.3390/ijerph9113806",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 2043 individuals, of which 440 were children < 5, from 260 householdsInclusion criteria: households were eligible if there was at least one child < 5\n\n\nInterventions: Plastic Biosand filter (117 households, 1012 individuals)Primary drinking supply (143 households, 1031 individuals)\n\n\nOutcomes: Diarrhoeal incidenceMicrobiological water quality\n\n\nNotes: Location: six rural communities, Tamale, GhanaLength: three months follow-upPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "A randomized controlled trial of the plastic BioSand filter (BSF) was performed in rural communities in Tamale (Ghana) to assess reductions in diarrheal disease and improvements in household drinking water quality.\nFew studies of household water filters have been performed in this region, where high drinking water turbidity can be a challenge for other household water treatment technologies.\nDuring the study, the longitudinal prevalence ratio for diarrhea comparing households that received the plastic BSF to households that did not receive it was 0.40 (95% confidence interval: 0.05, 0.80), suggesting an overall diarrheal disease reduction of 60%.\nThe plastic BSF achieved a geometric mean reduction of 97% and 67% for E. coli and turbidity, respectively.\nThese results suggest the plastic BSF significantly improved drinking water quality and reduced diarrheal disease during the short trial in rural Tamale, Ghana.\nThe results are similar to other trials of household drinking water treatment technologies.\n1. Introduction\nMany communities, especially in rural sub-Saharan Africa, still face significant challenges to provide access to improved drinking water sources and are struggling to meet the Millennium Development Goals for water and sanitation.\nIn Ghana, it is estimated that 93% of the urban population and 76% of the rural population have access to improved water and about 17% of urban and less than 10% of rural population have access to improved sanitation.\nThe lack of access to improved water and sanitation contribute significantly to diarrheal disease in the population.\nSpecifically, the Ghana Demographic and Health Survey suggests that diarrheal disease is a significant cause of morbidity and mortality in children less than five years of age, with 1 in 5 having reported diarrheal disease in the two weeks preceding the survey.\nIn the Northern region of Ghana, the two-week prevalence of diarrheal disease was almost 33%.\nAn immediate solution to address the lack of access to safe water is household water treatment (HWT), which allows households to treat drinking water at the point of consumption to improve its quality.\nStudies of HWT have shown that it can reduce the risk of diarrheal disease by 35% or more for a wide range of technologies in different settings and populations.\nOne technology that shows promise for areas where water has higher turbidity is the Hydraid\u00ae BioSand Water filter (plastic BSF), originally invented by David Manz.\nThis filter, which functions similarly to a slow sand filter, has been modified for intermittent use.\nThe most studied BSFs have been those having concrete housings.\nThe BSF\u2019s advantages are many, including a simple design, durable materials, local fabrication of the concrete housing, and provision of abundant quantities of water.\nThere have been four peer-reviewed published trials examining the health impact of the concrete BSF.\nThese trials suggest that use of the concrete BSF can reduce diarrheal disease by 50% or greater.\nWhile the BSF most often implemented is a concrete version, it has a relatively slow rate of daily production, can weigh as much as 500 lbs.\nwhen fully installed and can be difficult to transport to remote locations.\nA version of the BSF having a plastic housing has recently been produced and may overcome the problems in production, distribution and transport of the concrete filter.\nThe plastic BSF is light in weight, stackable and can be produced rapidly in large quantities by injection molding.\nThe plastic version of the BSF tested in this study is licensed by Manz, manufactured by Cascade Engineering and has specific depth and design parameters.\nIts sand filter bed has a smaller surface area than that of the concrete filter and it has tapered rather than straight side walls (see Figure 1).\nFurthermore, there is limited evidence of how well it will work in the field to improve water quality and reduce diarrheal disease risks, especially for waters with high turbidity.\nTo address the lack of field evidence of the performance of the plastic BSF to improve drinking water quality and reduce diarrheal disease in turbid surface water used by a population with a high diarrheal disease burden, a cluster randomized controlled trial (RCT) was performed in rural communities located in the Northern Region, Tamale, Ghana.\nThis study is one of three RCTs simultaneously performed on the plastic BSF in different geographic regions (including Cambodia and Honduras).\nThis is the first trial of the plastic BSF in this region of the world and only the second BSF RCT in Sub-Saharan Africa.\nUnique to this region and due to water scarcity, the communities often rely on water stored in \u201cdugouts\u201d, which are shallow surface water impoundments,.\nThe purpose of the RCT was to document the ability of the plastic BSF to improve water quality for both fecal indicator bacteria and turbidity and to reduce diarrheal disease in a setting where the population relies heavily on contaminated, highly turbid surface water for drinking and where the lack of access to water and sanitation are likely to be contributing significantly to morbidity and mortality in children under five years of age.\n2. Materials and Methods\n2.1. Research Setting, Study Population, and Participant Recruitment\nThis study (a cluster RCT) of the plastic BSF was conducted in six rural communities in Tamale, Ghana.\nAll villages and the water quality testing laboratory were located within 25 km of the city of Tamale (see Figure 2).\nField data collection took place between May to December 2008.\nThe six study communities and their households were selected based on the following criteria: child under the age of five years old, stored drinking water in the home, use of surface water as their primary drinking water source, did not spend most of the day selling goods in Tamale, were within 60 minutes from Tamale during the rainy season, and households agreed to participate.\nStudy design and protocols were approved by the Institutional Review Board of the University of North Carolina (IRB #08-0063) and the Ethical Review Committee of the Ghana Health Service.\nThe initial study was powered to detect a 25% reduction in diarrheal disease between the two groups based on an initial prevalence of diarrheal disease of 5%, 80% power and, \u03b1 = 0.05.\nWe also took into account the clustering of diarrhea within individuals and households and assumed four months of follow up visits as well as four people per household.\nBased on these sample size parameter values, we estimated the need for approximately 100 households in each study arm.\nPrior to recruitment, village elders were approached and informed about the study.\nIf the village elders were interested in participating, individual households were then asked to participate.\nHouseholds were excluded from the study if they did not have a child less than five years of age and/or did not want to participate.\nHousehold recruitment began on 10 May 2008 and was finalized on 16 June 2008, with informed consent obtained during the initial household visit.\nThe purpose of the initial cross-sectional recruitment questionnaire was to collect data on diarrheal disease prevalence in the households, risk factors of diarrheal disease, main drinking water sources, education levels of household members, access to sanitation and presence and type of any drinking water treatment practices.\nAccess to sanitation was assessed via questionnaire and a visual inspection of the facility, if it was present.\nLack of access was characterized by the absence of any type of latrine, pour flush toilet or other appropriate sanitation technology.\nThe initial cross-sectional recruitment was completed in six communities and a total of 260 households were recruited.\nAfter the initial recruitment, a baseline period of observation was performed prior to intervention with the plastic BSF.\nThe purpose of the baseline data collection period was to characterize and compare diarrheal disease and water quality between what would become the randomly selected intervention (plastic BSF) and control (no plastic BSF) villages.\n2.2. Intervention\nData collection for the longitudinal portion of the prospective cohort began on 17 June 2008 and was finished on 23 December 2008.\nHouseholds were visited at one week intervals and asked questions about diarrheal disease and water management practices in the home.\nDuring this time period, household water quality was also monitored periodically for total coliforms, E. coli and turbidity.\nThe randomization of villages and installation of plastic BSFs took place during the last week of August and first week of September 2008.\nBased on discussions with village leaders, randomization at the household level was deemed unacceptable by the majority of the villages.\nTherefore, the randomization was performed at the village-level.\nNumbers were assigned to the six villages and three numbers were selected from a random number generator to be the intervention villages.\nDue to the unequal size of the villages, more households were selected into the control group.\nAt the time of randomization, the six villages ranged in size from 14 to 70 households with children <5 years of age.\nAs a result of randomization, three villages with 70, 58 and 14 households were selected into the control group and three villages with 58, 24 and 33 households were selected into the intervention group (and received the plastic BSF).\nThe intervention phase of the study included weekly household observations from September 2008 to December 2008, with a potential of 15 weeks of household observations completed during this time period.\nAll households in the intervention and control groups continued to provide detailed information on weekly levels of diarrheal disease.\nFor household water quality analysis, turbidity was measured every two weeks and bacterial concentrations were measured five times during the intervention period, usually every two weeks.\n2.3. Diarrheal Disease Surveillance\nA consistent system of diarrheal disease surveillance was developed where one person in the household was identified as the primary respondent during the recruitment interview.\nThis person was surveyed weekly about diarrheal disease for all members of the household.\nUsing the following questions \u201cHas anyone in your house had diarrhea in the past one week?\u201d and \u201cIf yes, how many times did that person go in one 24-hour period?\u201d, the primary respondents were asked to verbally report any occurrence of diarrhea in the household within the last 7 days.\nAdditional questions on stool appearance (including the appearance of blood), duration of symptoms of diarrhea and use of treatment were also asked.\nIf the case of diarrhea was ongoing at the time of the visit, the case was asked about during the next visit to determine if it had resolved as well as to determine the duration of the case.\nOverall, there were 12 possible visits prior to plastic BSF installation as the intervention and 15 possible visits after plastic BSF installation, for a total of 27 potential weeks of observation for diarrheal disease surveillance.\n2.4. Water Quality Sample Collection and Analysis\nFrom the 260 households that were enrolled, samples of drinking water were taken during household visits from both the control and plastic BSF household groups.\nDuring the plastic BSF intervention, control households provided a sample of water used for drinking and BSF households providing three water samples at each visit: pre-filtered or untreated water, water directly from the plastic BSF outlet tube, and plastic BSF-treated water that had been stored for drinking.\nWater samples were collected by field staff directly into 500 mL sterile plastic collection bottles.\nThese samples were stored on ice and transported to the World Vision Laboratory at Savelugu, where they were immediately processed (within six hours of collection).\nAll samples were tested for total coliforms and E. coli using the IDEXX Colilert\u00ae Quantitray 2000\u00ae system (IDEXX Laboratories, Westbrook, ME, USA).\nMost probable numbers (MPN) for total coliforms and E. coli were determined using the IDEXX provided MPN table.\nTurbidity was tested using the Hach 2100P Portable Turbidimeter (Hach Company, Loveland, CO, USA).\n2.5. Data Analysis\nData from the initial cross-sectional questionnaire were used to compare the plastic BSF and control groups.\nPearson chi-squared tests were used to assess the proportion in each group for the following variables: access to sanitation, main drinking water source, educational attainment, having multiple children <5 years of age in the household, and reported drinking water treatment practices.\nThe effect of the plastic BSF on diarrheal disease was determined by comparing the longitudinal prevalence of diarrheal disease for all participants in each group, intervention (received BSF) and control (no BSF) using longitudinal prevalence ratios (LPR) generated from Poisson regression.\nIn order to classify a case of diarrheal disease, we used the World Health Organization (WHO) definition of three or more loose or watery stools in at least a 24-hour period.\nTo adjust for clustering within households and the villages, multi-level Poisson regression was performed and all data reported are based on the LPR from the regression model adjusted for clustering.\nTo fit the model, days with diarrheal disease (as counts) and person-days of observation were totaled by individual participant.\nThe multi-level model was used to adjust for clustering because individuals belonged to households which belonged to villages.\nAll diarrheal disease data analysis was performed in Stata 10.0 (Stata, StataCorp, College Station, TX, USA).\nBacterial concentration and turbidity data were log10 transformed and analyzed in Microsoft Excel and Stata 10.0 for graphical presentation and means testing.\nThe bacterial and turbidity reductions achieved by the plastic BSF were calculated as log10 reductions: log10 reduction = log10 influent \u2013 log10 effluent (Equation 1).\nFiltered drinking water quality of plastic BSF households was compared both for water taken directly from the filter and for stored filtered water and compared to untreated water from BSF and control households.\nPaired and unpaired t-tests were used to compare geometric mean log10 E. coli concentrations and geometric mean turbidities between plastic BSF and control water samples.\n3. Results\n3.1. Study Enrollment and Completion\nDuring the initial cross-sectional recruitment interview, six villages and 260 households were recruited into the study.\nThree villages were later randomized into the BSF intervention group and a total of 117 plastic BSFs were installed in separate households.\nThree villages were selected to remain as the control villages.\nAll control households were asked to continue normal water management practices for the intervention period.\nAlthough this randomization resulted in a small number of clusters, this small number of villages was selected to facilitate weekly follow-up visits.\nA larger number of villages would have made it logistically challenging to complete all visits within one week (the desired diarrheal disease recall period).\nPrior to randomization, during the baseline period, nine households (3.4%) dropped out of the study.\nDuring the plastic BSF intervention period, a total of seven households (2.3%) dropped out of the study; two households from the control group and five households from the plastic BSF intervention group.\n3.2. Baseline Characteristics and Group Comparability\nA total of 1012 people in the 117 households randomized to the plastic BSF-intervention villages and a total of 1031 people in the 143 households randomized to the control villages were compared.\nShown in Table 1 and Table 2 are characteristics of the plastic BSF and control groups based on data collected during the initial cross-sectional questionnaire.\nWhen the two groups were compared based on the age and proportion of males and females in various age groups, the two groups differed (statistically) in the proportion of those that were <2, 2\u20134 and >4 years of age, although the proportions were similar.\nThe proportion of males to females and the number of respondents that reported currently attending school were not significantly different between BSF and control households.\nPlastic BSF and control group characteristics regarding water, sanitation, hygiene, and other household level characteristics are presented in Table 2.\nAn overwhelming majority of households reported using surface water (collected from earthen dams, called \u201cdugouts\u201d) for drinking water in both dry and rainy seasons (71\u201398%).\nThese dams or dugouts are typically shallow areas with slightly raised banks that capture rain or runoff water during the rainy season.\nThis is the most common source for drinking water in this region of the country.\nFewer control households (71%) compared to BSF households (94%) reported using surface water from earthen dams during the rainy season, a difference that was statistically significant.\nHouseholds that reported using a source other than earthen dams during the rainy season reported using rainwater.\nAlmost all households lacked access to any type of sanitation (97 and 98%, respectively for control and plastic BSF households).\nThe two groups were not found to be significantly different when compared for other water and sanitation or demographic variables listed in Table 2, such as the practice of cloth sieving for water treatment, the proportion of households with at least one person currently attending school or households with more than one child less than five years of age.\n3.3. Diarrheal Disease\nIn order to examine the impact of the plastic BSF on diarrheal disease of participants, we compared the longitudinal prevalence between the two groups, plastic BSF intervention and control (no BSF) for the age groups of <2 years, all <5 years and all ages, both prior to the intervention and after installation of the plastic BSF as the intervention (Table 3).\nBefore intervention, households that were randomly selected to receive plastic BSFs experienced slightly lower longitudinal prevalence of diarrheal disease than control households for all categories of age groups, a difference that was not statistically significant (adjusted LPR for all ages: 0.98, 95% CI: 0.23\u20133.94).\nDuring the plastic BSF intervention period, longitudinal prevalence of diarrheal disease of all age groups was significantly lower in the plastic BSF intervention group than in the control group.\nFor example, for all ages, the BSF intervention group had 0.40 times the longitudinal prevalence of reported diarrhea as the control group (95% CI: 0.05, 0.80).\nThis observation suggests significant protection from diarrheal disease by the plastic BSF during the four month intervention period.\nWhen stratified by age group, the difference in longitudinal prevalence of diarrheal disease between participants in control and BSF households was even greater in children less than five years of age than in all age groups, with an adjusted LPR of 0.26, (95% CI: 0.07\u20130.89), corresponding to an estimated 74% reduction in diarrheal disease for plastic BSF participants compared to control participants.\nThe level of diarrheal disease reduction for BSF participants compared to control participants was only slightly lower for children less than two years of age compared to all participants combined, with an (adjusted LPR of 0.37 (95% CI: 0.15, 0.90)), corresponding to an estimated 63% diarrheal disease reduction.\n3.4. Water Quality Analysis\nHousehold drinking water quality was compared over the entire study period for plastic BSF and control households.\nThe geometric mean concentrations of E. coli and mean turbidities of household drinking water for the baseline and intervention periods are compared in Table 4.\nBefore the intervention, plastic BSF households and control households had water with very high geometric mean concentrations of E. coli and turbidity: 724 and 832 MPN E. coli /100 mL, respectively (control and plastic BSF).\nLikewise, households in both groups had drinking water with high geometric mean turbidities; 95 and 85 NTU for control and plastic BSF, respectively.\nNeither E. coli nor turbidity levels of household water were statistically significantly different between the two groups prior to the plastic BSF intervention (two sample t-test for turbidity (p = 0.23) and for E. coli, (p = 0.22)).\nAfter the plastic BSF intervention, households with the plastic BSF demonstrated appreciable improvements in drinking water quality compared to the control household without the BSF (Table 4).\nDuring the intervention period, E. coli concentrations were similar in the water that was collected and stored in the household before treatment (490 and 437 MPN/100mL for control and BSF groups, respectively; p value = 0.27).\nFor the plastic BSF group, E. coli levels in water from the plastic BSF were significantly lower than those in the stored untreated water, (16 MPN/100mL and 437 MPN/100mL, in BSF direct filtrate and untreated stored water, respectively, p value < 0.0001).\nPlastic BSF treated and stored water was also significantly improved based on E. coli levels compared to untreated stored water in BSF households (76 MPN/100 mL and 437 MPN/100 mL, respectively; p < 0.0001).\nHowever, plastic BSF treated and stored water had significantly higher E. coli concentrations than water taken directly from the plastic BSF outlet, with 76 MPN/100 mL and 16 MPN/100 mL, respectively; p < 0.0001.\nAfter the plastic BSF was installed, both the control and plastic BSF intervention groups had collected stored water with lower turbidity as compared to the baseline phase (near 100 NTU prior to intervention and <50 NTU for both groups during the intervention phase).\nHowever, plastic BSF households had significantly lower turbidity for water treated by the plastic BSF (15 NTU compared to 27 NTU for control households).\nUnlike E. coli levels, turbidity levels of water did not change significantly during storage.\nControl and plastic BSF households were asked (at each household visit) whether or not they had performed treatment to their water.\nPrior to plastic BSF intervention, households reported sieving the water through a cloth prior to drinking for 97% of all observations.\nBoiling and chlorine were cited as treatment for 0.2% and 0.15% of the observations, respectively.\nDuring the intervention period, the control group reported high levels of cloth-sieving (93% of all observations) but no additional treatment.\nThe plastic BSF group did not report any additional treatment beyond filtration with the plastic BSF during the intervention period.\nIn an attempt to measure adherence to the plastic BSF intervention, households reported the frequency of weekly plastic BSF use and use of the plastic BSF filtered water for drinking.\nDuring the weekly observations, none of the plastic BSF intervention households reported not using the BSF to filter water in the previous 7 days.\nIn addition, in very few observations (3% all household observations in plastic BSF households) users report not drinking the plastic BSF filtered water.\n3.5. Plastic BSF Performance\nThe plastic BSF achieved a geometric mean 97% reduction of E. coli when untreated water in BSF households was compared to water direct from the BSF outlet.\nHowever, the E. coli reduction was much lower (85%) when comparing untreated water in BSF households to BSF-treated and stored water.\nA geometric mean reduction of 67% for turbidity was found comparing untreated water to water directly from the BSF; a similar reduction was found comparing untreated to BSF treated and stored water (66% geometric mean reduction).\nWhen compared on a categorical basis of order of magnitude concentration ranges as a basis for categorizing risk levels posed by the water, plastic BSF treated water had significantly fewer samples in high risk E. coli concentration categories as compared to untreated source water (p < 0.0001, Pearson\u2019s chi-squared test).\nAs shown in Figure 3, 44% and 15% of water samples taken directly from the plastic BSF outlet and the plastic BSF treated and stored water, respectively, had less than 10 MPN E. coli/100mL (considered low risk water) as compared to only 1.5% of water in BSF households prior to treatment.\nFurthermore, 77 and 56% of samples taken from the plastic BSF outlet directly or plastic BSF treated and stored, respectively, had fewer than 100 MPN E. coli/100mL (considered moderate risk) as compared to only 12% of all samples in this category prior to plastic BSF treatment in plastic BSF intervention households.\nOverall, there was significant improvement in categorical concentrations of E. coli for plastic BSF treated drinking water and plastic BSF treated and stored water compared to untreated water.\n4. Discussion\nTo our knowledge, this is the first study to assess the ability of plastic BSF to reduce the longitudinal prevalence of diarrheal disease in Ghana.\nThere are relatively few studies examining household water filtration in communities in this region of the world and in water sources as turbid as the ones found in this study.\nWe documented a significant reduction (60%) in diarrheal disease for households who were asked to use a plastic BSF compared to control households that were given no BSF during the study period.\nOne similar study, performed in Kenya with a concrete housing intermittently operated slow sand filter, found a 54% reduction in diarrhea prevalence.\nThey also found that the reduction of diarrheal disease was substantially higher (77%) when comparing households only using unimproved surface water sources for their drinking water supply.\nSimilarly high reductions (80%) in diarrheal disease were also reported in a randomized controlled trial of a ceramic water filter in South Africa and Zimbabwe.\nThe results from our study suggest reductions in diarrheal disease that are consistent with similarly designed trials on household water filters in other countries in this region.\nThe results from this study are also similar to the diarrheal disease reduction results we documented (as part of the three country trial of the plastic BSF) from an RCT in Cambodia, where diarrheal reductions in households using the plastic BSF were similar (59%) to those reported here during a five-month intervention trial.\nIn Honduras, the diarrheal disease reduction results of a plastic BSF RCT were similar in magnitude to those reported here although this was not statistically significant.\nOverall, the results for diarrheal disease reduction in BSF households compared to control households observed in this study suggest greater or comparable reductions compared to those obtained from trials of concrete BSFs in other regions of the world such as Cambodia and the Dominican Republic, which demonstrated 47%, and 54% reductions in diarrhea illness risks, respectively, in children <5 years of age.\nThese results also compare favorably to peer-reviewed studies of other HWT technologies such as chlorine disinfection, which has been found to achieve reductions in diarrheal disease of about an average 30% in many regions around the world.\nResearchers, however, have recently questioned the validity of results of unblinded, randomized controlled trials lacking a placebo such as this one, suggesting that there may be a significant source of bias where households with the intervention may under-report diarrheal disease.\nBecause this is the first study of the plastic BSF in Northern Ghana, we did not include a placebo filter.\nFurthermore, due to the high turbidity of the surface water used as drinking water by households, it would have been difficult to blind participants to the treatment.\nBoisson et al. recently performed a placebo controlled trial of a household filter in the Democratic Republic of Congo and experienced considerable challenges in designing and implementing a neutral filter as a placebo.\nWe are unaware of any study that successfully employed a placebo filter at point-of-use in any developing country setting.\nAdditional research on the plastic BSF and other HWT technologies should attempt to measure health impacts in more objective ways that can help to eliminate bias.\nThese may include incorporating diagnostic procedures to detect intestinal infections or including anthropometric measurements of children for longer-term health outcomes such as was performed in a recent trial of solar disinfection in Kenya.\nHowever, even child anthropometry can be subject to measurement error and may not completely eliminate the question of bias.\n4.1. Effect of the Plastic BSF on Household Drinking Water Quality\nDuring the four months of intervention with the plastic BSF, we documented significant improvements in household water quality for both E. coli and turbidity of plastic BSF treated water.\nBecause the lack of access to improved drinking water was high in these communities (71\u201398% used surface water which was collected from earthen dams) and the water accessed was highly microbiologically contaminated and often very turbid, this study provides important evidence regarding the potential application of the plastic BSF for locations where safe water access is limited.\nDuring the study, the plastic BSF demonstrated an average 97% reduction of E. coli in BSF-treated water, which is similar to reductions seen both in the laboratory and in the field for concrete BSFs and other filtration technologies.\nFor the untreated water of this study, when E. coli concentrations were \u22651000 MPN/100mL, the plastic BSF averaged 99% reduction (data not shown).\nAs demonstrated in laboratory studies of a similarly designed plastic BSF, protozoan parasites and bacterial removals can be as high as \u226599%.\nVirus removals may not be as robust (<90%) and may be more dependent on the conditions of the biological activity and the biofilm in the BSF.\nUnder the conditions examined during our study, we expect bacterial and protozoan pathogens to be effectively removed by the plastic BSF at levels similar to the removals of the bacterial indicators (97% or more).\nFurthermore, there is potential for virus removals greater than what has been documented in the laboratory due to the potential for virus attachment to particles in the turbid raw water and the likelihood of robust biofilms as the result of this highly turbid raw water.\nRotavirus have been documented as an important pathogen in the region and further study on the potential for this BSF to remove viruses under the conditions of this study is warranted.\nAlthough there were significant reductions of E. coli in household drinking water samples as a result of treatment with the plastic BSF, there was also evidence of recontamination during storage of plastic BSF-treated water (Figure 3, Table 4).\nThis type of recontamination has been documented in previous studies of the concrete BSF.\nAs shown in Figure 1, although the water may leave the outlet tube of the plastic BSF relatively uncontaminated, during storage of treated water there are ample opportunities for bacterial contamination to be re-introduced via hands, dippers and even the storage container itself.\nThe opportunity for bacterial recontamination has been documented for other treatment options that do not provide a residual disinfectant, such as boiling.\nFor these types of technologies, additional training regarding safe and hygienic storage of treated water should be included to reduce bacterial recontamination.\nWhile the turbidity reduction in household drinking water by the plastic BSF in the intervention group was significant compared to untreated water and to control household water turbidity, the average turbidity of treated drinking water of 14 NTU was still higher than the WHO suggested limit of 5 NTU.\nFew studies have examined treatment of water in regions where the main source of drinking water is highly turbid surface water.\nCrump et al. demonstrated a significant reduction in diarrheal disease, bacterial contamination and turbidity for households using a combined flocculant-disinfectant in Kenya, with turbidities reduced from >100 NTU to <5 NTU in 50\u201370% of water samples treated.\nMwabi et al. reported that a range of locally produced point-of-use water filters, including BSFs, consistently reduced the turbidities of surface waters with an average turbidity of about 40 NTU to <5 NTU in South Africa.\nAlso important to consider was that turbidity decreased for both control and plastic BSF households during the intervention period.\nThis may be due to water quality improvements at the point of collection, possibly associated with changes in rainfall.\nLimitations of this study include randomization at the village level which resulted in a small number of clusters, having participants unblinded to the exposure and lack of a placebo.\nAdditional limitations include use of self-reported diarrhea, with a 7-day recall of diarrhea, as the health outcome measure and the relatively short duration of the intervention.\nHowever, such limitations are common in other studies of HWT technologies such as the ceramic water filter and chlorination.\nThe short duration and frequent observations of this study are important limitations of its design because evidence suggests that many point-of-use household water interventions show decreased performance effectiveness over time.\nA recent study investigated the impact of household survey on health behaviors and found that the act of survey itself may impact respondents\u2019 behaviors.\nTherefore, studies of longer duration and those that are designed to eliminate household survey impact on respondents\u2019 behaviors should also be considered for the BSF in future studies.\nDesign and conduct of studies that evaluate implementation programs extending over longer periods of time would also be more informative of the longer term impacts of the plastic BSF on correct and consistent daily use, water quality and health.\nDespite the aforementioned limitations in this study, the results can still be compared to other rigorous and widely cited studies of HWT technologies and their impact on the diarrheal disease risks of the users.\nIn particular, these results confirm those of past RCTs of the concrete BSF by documenting that the plastic BSF has the ability to reduce the longitudinal prevalence of diarrheal disease and significantly improve drinking water quality, even in a cohort of households that is primarily using unimproved, turbid and microbiologically contaminated surface water from earthen dams for their main source of drinking water.\n5. Conclusions\nTo our knowledge, this is only one of two known health impact studies on the performance of the BSF in Sub-Saharan Africa and is the first on the plastic BSF on this continent.\nPositive results were found for the ability of the plastic BSF filter to reduce diarrheal disease and improve water quality in communities of the Northern Region of Ghana using highly turbid and microbiologically contaminated surface water collected from earthen dams as their drinking water source.\nHowever, more research is necessary to document plastic BSF performance, continued and effective use and overall sustainability after installation and in the absence of intensive sampling or other monitoring.\nTypical set-up of the plastic BSF in a household in rural Tamale, Ghana (2008).\nMap a (1:500,000) indicating the location for the Randomized Controlled Trial of the Plastic Biosand Filter in Tamale, Ghana (2008).\nComparison of categorical E. coli concentrations in different water samples from plastic BSF households during intervention period of the RCT of the plastic BSF in rural Tamale, Ghana (2008).\n\nAge (as of May 2008), sex and education status of participants in control (no BSF) and intervention (BSF) households during a randomized controlled trial of the plastic BSF in rural Tamale, Ghana 2008.\nIndividual Variables | Control | Intervention | p values\n | (N = 1031) | (N = 1012) | Pearson \u03c72 test\nAge | N (%) | N (%) | \nParticipants \u2265 5 years old | 809 (78.5) | 794 (78.5) | \nParticipants 2\u20134 | 146 (14.3) | 116 (11.5) | \nParticipants <2 | 76 (7.4) | 102 (10.2) | 0.027 a\nSex |  |  | \nMale (\u22655) | 436 (42) | 391 (39) | \nMale (<5) | 113 (11) | 116 (11) | 0.37 a\nFemale (\u22655) | 373 (36) | 401 (40) | \nFemale (<5) | 108 (11) | 104 (10) | 0.12 a\nEver or currently attending school | 261 (25.3) | 260 (25.7) | 0.85\n\na Pearson chi-squared test performed for proportions in all categories of age and gender comparing people in BSF and control groups.\n\nSelected characteristics regarding demographics, water and sanitation for control (no BSF) and intervention (BSF) households in a randomized controlled trial of the plastic BSF in rural Tamale, Ghana, June\u2013December 2008.\nHousehold Level Variables | Control | Intervention | p values\n(N = 143) | (N = 117) | (\u03c72 test)\nUse surface water during dry season | 95.0% | 98.3% | 0.16\nUse surface water during rainy season | 70.6% | 94.0% | <0.001\nReported sieving drinking water through cloth | 96.5% | 96.6% | 0.97\nAt least one person attending school in household | 70.4% | 74.8% | 0.39\nLack access to sanitation a | 97.1% | 98.3% | 0.67\nMulti-child household b | 56.7% | 55.6% | 0.86\n\na Lack of access was characterized by the absence of any type of latrine or pour flush toilet.\nb Household has at least two children less than five years of age participating in the study.\n\nAdjusted longitudinal prevalence ratios for diarrheal disease, stratified by age, during the pre-intervention and intervention phases of a randomized controlled trial of the plastic BSF in rural Tamale, Ghana (2008).\nData collection Period | Age stratum | Unadjusted LP a\u2014Control Villages | Unadjusted LP a\u2014Plastic BSF Villages | Adjusted LPR (95% CI)\nBaseline (May\u2013August 2008) b  | All | 0.024 | 0.020 | 0.98 (0.23, 3.95) c\n<2 years of age | 0.081 | 0.10 | 1.56 (0.25, 9.83) d\n<5 years of age | 0.078 | 0.074 | 1.38 (0.19, 10.17) d\nPlastic BSF Intervention (September\u2013December 2008) | All | 0.012 | 0.0063 | 0.40 (0.05, 0.80) c\n<2 years of age | 0.028 | 0.015 | 0.37 (0.15, 0.90) d\n<5 years of age | 0.034 | 0.018 | 0.26 (0.07, 0.89) d\n\na LP\u2014unadjusted longitudinal prevalence which was calculated as the total number of days with diarrheal disease over the total number of days observed; b Period of observation in villages prior to randomization and plastic BSF installation; c Longitudinal prevalence ratio and 95% confidence interval with plastic BSF as exposure adjusted for adjusted for categorical age of participant and clustering of diarrheal disease within household and villages; d Longitudinal prevalence ratio and 95% confidence interval with plastic BSF as exposure adjusted for clustering of diarrheal disease within household and villages.\n\nGeometric mean E. coli concentrations and turbidities of household drinking waters for the control and plastic BSF groups before and after plastic BSF intervention in a randomized control trial in rural Tamale, Ghana (2008).\nBaseline (May\u2013August 2008) | Plastic BSF Intervention (September\u2013December 2008)\nWater Quality Parameter | Control HH Water | Plastic BSF HH Water | Control HH Water | Plastic BSF Untreated Water b | Plastic BSF Direct Filtrate | Plastic BSF Stored Filtrate\nE. coli (95%CI) a | 724 (631\u2013851) | 832 (724\u2013977) | 490 (426\u2013549)  | 437 (380\u2013501)  | 16 (13\u201320) | 76 (62\u201391)\n(N = 516) | (N = 424) | (N = 587) | (N = 385) | (N = 382) | (N = 381)\nNTU | 95 (83\u2013109)  | 85 (74\u201398)  | 25 (23\u201327) | 47 (42\u201351)  | 15 (13\u201317) | 15 (14\u201318)\n(N = 523) | (N = 430) | (N = 787) | (N = 527) | (N = 524) | (N = 527)\n% E. coli Reductions c | -- | -- | -- | -- | 97 | 85\n% NTU Reductions c | -- | -- | -- | -- | 67 | 66\n\na Geometric mean and 95% confidence interval for E. coli (MPN/100mL) and turbidity of household (HH) drinking water; b Untreated water refers to water that was taken from households prior to any treatment. This is the assumed contamination level before the BSFs were used to treat the water.; c\nE. coli and NTU reduction calculated as log10 reduction = log10 influent \u2013 log10 effluent and then the log10 reductions were transformed into %.",
        "label": "high",
        "id": "task4_RLD_test_690"
    },
    {
        "paper_doi": "10.1186/1475-2891-10-120",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: RCT.Standard care: all participants received antituberculous therapy (6 months DOT: isoniazid 50 mg, rifampicin 200 mg, ethambutol 10 to 15 mg/kg, and pyrazinamide 20 to 30 mg/kg daily for 2 months; isoniazid 50 mg and rifampicin 200 mg daily for 4 months) and were visited at home by the study nurse every 2 weeks.Study dates and duration: May 2005 to September 2007. Follow-up 8 weeks.\n\n\nParticipants: Number: 255 enrolled and randomized; outcomes presented for 237.Inclusion criteria: aged 6 weeks to 5 years; loss of > 10% maximum weight or failure to gain weight for 2 months; cough with wheeze for >= 4 weeks; history of household contact with a probable or confirmed tuberculosis case in the past 6 months; pyrexia of unknown origin; painless swelling in a group of cervical lymph nodes; children who were diagnosed with tuberculosis in the past 5 years and have received antituberculous therapy for a period < 4 weeks. Positive TST (>= 10 mm induration in HIV-negative and >= 5 mm in HIV-positive; after 48 to 72 hours) or with chest X-ray indicative of tuberculosis (based on unequivocal lymphadenopathy or military tuberculosis) eligible for enrolment.Exclusion criteria: children who had been treated with antituberculous therapy exceeding 4 weeks in the past 1 year.Baseline characteristicsNutritional status: median (IQR) MUAC cm: 13.0 (11.9 to 14.25) multivitamin group versus 13.0 (11.2 to 14.0) placebo group; weight (kg): 8.48 multivitamin group versus 7.95 placebo group.HIV status: 39.1% HIV-positive in the multivitamin group, 29.1% HIV-positive in the placebo group.MDR/XDR-TB: not mentioned.Micronutrient status: not documented.\n\n\nInterventions: Group 1: multivitamin supplements daily for the first 2 months. Composition: vitamin B1 0.5 mg, vitamin B2 0.6 mg, niacin 4 mg, vitamin B6 0.6 mg, folate 130 mg, vitamin B12 1 mg, vitamin C 60 mg, and vitamin E 8 mg.Age < 6 months: 1 capsule daily.Age 7 to 36 months: 2 capsules daily.Age > 36 months: 3 capsules daily.Group 2: placebo daily for the first 2 months.\n\n\nOutcomes: Weight gain at 2 months.Deaths during treatment.Not included in this review: height, MUAC, and triceps skin-fold thickness changes at 2 months, clearance on chest x-ray at 2 months, immunological outcomes.\n\n\nNotes: Location: Dar Es Salaam, Tanzania.Setting: hospital paediatric clinic.Registration number: NCT00145184.Source of funding: National Institutes of Health Grant, the Harvard School of Public Health\n\n",
        "objective": "To assess the effects of oral nutritional supplements in people being treated with antituberculous drug therapy for active tuberculosis.",
        "full_paper": "Background\nChildren with tuberculosis often have underlying nutritional deficiencies.\nMultivitamin supplementation has been proposed as a means to enhance the health of these children; however, the efficacy of such an intervention has not been examined adequately.\nMethods\n255 children, aged six weeks to five years, with tuberculosis were randomized to receive either a daily multivitamin supplement or a placebo in the first eight weeks of anti-tuberculous therapy in Tanzania.\nThis was only 64% of the proposed sample size as the trial had to be terminated prematurely due to funding constraints.\nThey were followed up for the duration of supplementation through clinic and home visits to assess anthropometric indices and laboratory parameters, including hemoglobin and albumin.\nResults\nThere was no significant effect of multivitamin supplementation on the primary endpoint of the trial: weight gain after eight weeks.\nHowever, significant differences in weight gain were observed among children aged six weeks to six months in subgroup analyses (n = 22; 1.08 kg, compared to 0.46 kg in the placebo group; 95% CI = 0.12, 1.10; p = 0.01).\nSupplementation resulted in significant improvement in hemoglobin levels at the end of follow-up in children of all age groups; the median increase in children receiving multivitamins was 1.0 g/dL, compared to 0.4 g/dL in children receiving placebo (p < 0.01).\nHIV-infected children between six months and three years of age had a significantly higher gain in height if they received multivitamins (n = 48; 2 cm, compared to 1 cm in the placebo group; 95% CI = 0.20, 1.70; p = 0.01; p for interaction by age group = 0.01).\nConclusions\nMultivitamin supplementation for a short duration of eight weeks improved the hematological profile of children with tuberculosis, though it didn't have any effect on weight gain, the primary outcome of the trial.\nLarger studies with a longer period of supplementation are needed to confirm these findings and assess the effect of multivitamins on clinical outcomes including treatment success and growth failure.\nClinicaltrials.gov Identifier\nNCT00145184\nIntroduction\nMycobacterium tuberculosis is one of the most successful pathogens known to man-both in terms of its longevity as well as in its ability to infect and cause disease in humans.\nMolecular genetics and genome sequencing techniques estimate that early forms of M. tb were present in East Africa at least 3 million years ago; it remains the single most common curable infectious disease cause of mortality worldwide, despite the availability of effective anti-tuberculous chemotherapy.\nAn estimated 250,000 children develop tuberculosis (TB) and 100,000 die of it every year worldwide.\nAge and immune status of the child are two major determinants of progression to active TB after primary infection-the risk is highest in very young (< 2 years of age) and immunocompromised children.\nMalnutrition and HIV infection increase this risk further; for example, it is estimated that only one out of ten immunocompetent persons infected with TB develops active TB in his/her lifetime; whereas, one out of ten HIV-infected persons infected with TB will develop active TB every year.\nData from several studies indicate that TB is associated with weight loss and protein and calorie malnutrition and such poor nutritional status in TB patients is a strong predictor of adverse events including treatment failure and mortality.\nStudies among children without TB have shown a beneficial effect of multiple micronutrient supplementation on growth indices; for example, a meta-analysis showed that multiple micronutrient interventions improved linear growth (effect size: 0.09; 95% CI: 0.008, 0.17).\nIn addition, micronutrient supplementation can also lead to boosting of the immune system, which may help improve the response to TB treatment.\nThere is limited data on the prevalence of micronutrient deficiencies among children with tuberculosis in resource-limited settings; however, in a trial of multivitamin supplementation in Tanzania, 22% and 15% of children born to HIV-infected women, who did not receive multivitamin supplementation, were deficient in vitamins E (< 11.6 \u03bcmol/L) and B12 levels (< 150 pmol/L), respectively.\nHowever, there are no studies of multivitamin supplementation among children with TB.\nIn our previous work, we have also shown the benefits of maternal multivitamin (vitamins B-complex, C, and E) supplementation on child morbidity and mortality.\nTherefore, we hypothesized that multivitamin supplementation would improve weight gain, a predictor for future growth and adverse clinical outcomes, in children with TB.\nTo test this hypothesis, we conducted a randomized placebo-controlled trial among children with TB, both with and without HIV infection, in Dar es Salaam, Tanzania.\nMaterials and methods\nStudy Design and Population\nThis study was a randomized double-blind placebo controlled trial among 255 children between the ages of six weeks and five years with probable tuberculosis.\nA total of 467 children aged six weeks to five years attending the pediatric clinic between May 2005 and September 2007 at the Muhimbili National Hospital in Dar es Salaam, Tanzania, were screened for signs and symptoms of TB (Figure 1).\nThe inclusion criteria comprised of presenting with cough or wheezing for at least four weeks, fever of unknown origin, painless swelling in a group of cervical lymph nodes, loss of more than 10% of maximum weight, failure to gain weight for two months, or a history of household contact with a case of probable/confirmed TB in the past six months, and these children were diagnosed as having suspected TB.\nChildren who had received anti-TB treatment for more than 4 weeks in the past year were not eligible.\nA chest X-ray (Postero-Anterior view) was also obtained and a tuberculin skin test (TST) was conducted in all children.\nThe TST used the standard WHO-purified protein derivative (PPD) and it was read after 48-72 hours.\nChildren with a positive TST (induration greater than or equal to 10 mm in HIV-uninfected and 5 mm in HIV-infected) or with a chest X-ray indicative of TB (based on unequivocal lymphadenopathy or miliary TB) were categorized as probable TB (n > 275) and became eligible for enrolment in the study.\nThe chest X-rays were read both by the study radiologist and the study pediatrician.\nAny discrepancies were resolved by a joint review of the radiological findings.\nTrained research assistants obtained informed consent from the parents or guardians of the children in two stages-first, for eligibility testing and the second for participation in the trial.\nChildren with probable TB, written consent from parent/guardian, and no intent to leave the Dar es Salaam area over the next eight weeks were then randomly assigned to receive daily multivitamin supplements or placebo for the next two months (n = 255).\nAn off-site statistician generated the randomization sequence; a list from 1 to 400 was prepared according to this randomization sequence in blocks of 20.\nAt enrolment, each eligible child was assigned to the next numbered bottle of regimen at the site by the study staff.\nTo minimize the risk of unblinding, the regimen bottles, with no visual difference between active regimen capsules and placebo capsules, were received from the manufacturer (Nutriset, France) without any identification; the study staff then labeled the bottles with the patient's initials and identification number.\nBoth the clinicians and the patients were blinded to the study regimen, and the randomization list was kept confidential by the statistician, with the exception of the pharmaceutical company preparing the blinded treatment.\nEach multivitamin capsule contained vitamins B1 0.5 mg, B2 0.6 mg, Niacin 4 mg, B6 0.6 mg, Folate 130 \u03bcg, B12 1 \u03bcg, C 60 mg, and E 8 mg.\nThis composition was selected based on the demonstrated benefits of maternal multivitamin (vitamins B-complex, C, and E) supplementation on child morbidity and mortality, particularly in subgroups of children born to women who were immunologically or nutritionally compromised.\nZinc and iron, on the other hand, were excluded because of concerns related to their supplementation, particularly among HIV-infected populations.\nThe roles of zinc and iron need to be examined separately rather than as part of a multi-nutrient regimen.\nSimilarly, there is limited data on the utility of nutrients such as selenium and vitamin K in child health, and therefore, they were not included in the study regimen.\nChildren younger than six months in both the treatment and placebo groups received one capsule daily, whereas children between 7-36 months of age received 2 capsules daily, and children older than 36 months received three capsules daily.\nThe doses provided were in multiples (two to five times) of the recommended dietary allowance as have been used in our earlier studies.\nAmpoules of sterile water to dissolve the contents of the capsule were also given to the study participants.\nBoth placebo and the active regimen had a sweet taste when dissolved in sterile water.\nIn addition, all children received standard anti-TB treatment, according to the guidelines of the National Tuberculosis and Leprosy Control Programme of Tanzania.\nThe guidelines at the time of the trial recommended a six-month course of anti-tuberculous drugs (Isoniazid 50 mg, Rifampicin 200 mg, Ethambutol 10-15 mg/kg, and Pyrazinamide 20-30 mg/kg daily for 2 months, followed by Isoniazid 50 mg and Rifampicin 200 mg daily for 4 months) using Directly Observed Therapy (DOT).\nThe institutional review boards at Harvard School of Public Health, Boston, MA and Muhimbili University of Health and Allied Sciences, Dar es Salaam, Tanzania approved the study protocol.\nBaseline Assessment\nAt randomization, a trained research assistant obtained information on socio-economic and demographic characteristics.\nWeight, height/length, head circumference, mid-upper arm circumference (MUAC), and triceps skin-fold thickness, were measured by a study nurse according to standard methods.\nA physician also obtained a complete medical history and conducted a physical examination of the participating child at this visit.\nFollow-up procedures\nAll study subjects were followed up every two weeks; the study nurse visited the child at home during weeks 2 and 6 and the child was seen at the study clinic during weeks 4 and 8 of the study.\nDuring each visit, the study nurse enquired about the health of the child during the preceding two weeks and checked for compliance with anti-TB therapy and the study regimen through direct questioning of the parent/guardian and counting of the capsules left.\nDuring the clinic visits, the study nurse determined the anthropometric measurements, and a physician examined the child.\nLaboratory Methods\nVenous blood was obtained from the participating children at entry into the study and at the last visit at 8 weeks after starting anti-TB therapy.\nComplete blood counts, including hemoglobin concentration, and albumin levels (using Roche Cobas Integra 400 plus analyzer) were determined for all participants.\nTotal white blood cell count was measured using a Beckman Coulter AcT Diff II hematology analyzer and differential white blood cell count was also determined automatically.\nAbsolute counts of T-cell subsets were measured using the FACSCount system (Becton-Dickinson, San Jose, CA).\nHIV-1 (referred to as HIV henceforth) status was assessed through a double ELISA for children older than 18 months; any discrepancies were resolved using a Western blot test.\nFor children younger than 18 months of age, HIV status was determined through Amplicor HIV-1 DNA PCR test version 1.5 (Roche Diagnostics, Branchburg, NJ).\nStatistical Analysis\nThe study was designed to enroll 400 children with TB and was powered (power = 80%) to detect a 27% difference in the primary endpoint of weight gain between the supplemented and the placebo groups.\nDue to funding constraints, we could not extend the enrollment period to increase sample size beyond 255.\nThe attained sample size had adequate power to detect a 32.5% difference in weight gain between the supplemented and the placebo groups.\nIntention to treat analyses for all endpoints was used.\nDescriptive statistics were expressed as the median with the first and third quartiles (interquartile range, IQR) and non-parametric Wilcoxon test was used for comparative statistics.\nHodges-Lehmann 95% confidence limits were used to express the effect size of the difference between the two groups.\nWe also examined the effect of the supplements within strata of age (younger than six months, seven-36 months, and older than 36 months) and HIV status.\nTests of interaction were based on comparing the non-parametric estimate of the treatment effect between strata, using their standard errors.\nSAS version 9.2 (SAS Institute, Cary, NC) was used for all the analyses.\nResults\nThe baseline characteristics of the children enrolled in the trial are presented by treatment arm in Table 1.\nThere were no significant differences between the two treatment arms.\nThe median age was 18 months in both the placebo and the supplement groups.\nForty seven percent and 44% of the children enrolled were females in the placebo and the supplemented groups, respectively.\nTwenty nine percent of the children in the placebo group and 39% in the supplemented group were HIV-infected.\nOnly 3 children out of the 146 who had available gastric aspirate or sputum for culture had mycobacteria isolated in the sample.\nThere were no significant differences in weight gain or changes in height, MUAC, head circumference, and triceps skin-fold thickness between the two groups during follow-up (Table 2).\nChildren in both the placebo and the supplement groups gained a median of 0.84 kg in the two months since starting anti-TB therapy (n = 237; 95% CI = -0.15, 0.18; p = 0.82).\nHowever, there was significant effect modification by age group (p < 0.01); children younger than six months of age who received multivitamins gained a median of 1.08 kg, compared to 0.46 kg gained by children receiving placebo in that age group (95% CI = 0.12, 1.10; p = 0.01).\nThere was no difference in the gain in length/height during follow-up between the two groups (p = 0.53).\nHowever, HIV-infected children between the ages of 6 months and 3 years gained a median of 2 cm when receiving multivitamins, compared to 1 cm when receiving placebo (n = 48; 95% CI = 0.20, 1.70; p = 0.01; p for interaction by age group = 0.01).\nThe median increase in MUAC was 1.0 cm in the placebo group and 0.8 cm in the supplement group (p = 0.61).\nHead circumference increased by a median of 0.5 cm in both the placebo and the supplement groups (p = 0.86).\nThe change in triceps skin-fold thickness over the period of follow-up was an 1.1 cm increase in the placebo group and 0.7 cm increase in the supplement group (p = 0.37).\nThere was no difference in the clearance of chest x-ray or mortality in the two treatment arms.\n199 children had a chest X-ray available at the end of follow-up and 100 of them showed complete resolution.\n13 children died during the course of the trial.\nThere was no significant difference in the hemoglobin levels in children in the placebo and supplement groups at baseline (Table 3); however, a significant difference was noted by the end of follow-up (n = 223) with the children in supplemented group having a greater increase in the hemoglobin levels by a median of 1.0 g/dL, compared to children in the placebo group (median of 0.4 g/dL; p < 0.01).\nIn subgroup analyses by sex, this effect on hemoglobin was not statistically significant in female children (p = 0.11).\nNo significant differences were observed in changes in albumin, and CD4, CD8, and CD3 T-cell counts between the placebo and the supplemented groups over follow-up.\nHowever, significant effect modification by age group was observed in CD8 T-cell counts (p = 0.01).\nChildren 3 years and older showed a median increase in CD8 T-cell counts of 135 when receiving multivitamins, compared to a median decrease of 158 observed in children in the same age group receiving placebo (95% CI = 37.00, 1029.00; p = 0.03).\nOn restricting the analyses to only the 87 HIV-infected children, a similar relationship was observed for hemoglobin as in the overall cohort.\nChildren in the supplemented group gained a median of 1.2 g/dL of hemoglobin compared to no change in the median of the placebo group (p < 0.01).\nDiscussion\nWe found that daily multivitamin supplementation in children with TB in a resource-limited setting resulted in an improvement in hemoglobin levels after two months of follow-up in all age groups and irrespective of HIV status.\nHowever, there was no effect of the supplement on albumin levels and growth indices, including weight, length/height, mid-upper arm circumference, head circumference, and triceps skin-fold thickness in the overall cohort.\nIn subgroup analyses, significant weight gain among the youngest children (six weeks to six months) was observed.\nThe results were similar in children also co-infected with HIV; however, these children had a significantly higher increase in height if between the ages of 6 months and three years.\nThere have been no earlier studies of multivitamin supplementation among children with TB.\nHowever, similar beneficial effects of supplements on hemoglobin levels have been obtained with micronutrient supplementation in children in other parts of the world.\nA recent review by Allen et al reported that multiple micronutrient supplementation leads to a significant increase in hemoglobin in children (effect size 0.39; 95% CI 0.25, 0.53).\nA pooled analysis used data from intervention trials in Indonesia, Peru, South Africa, and Vietnam among 1134 infants who had been randomized to either a placebo, weekly multiple micronutrient supplement (including vitamins A, D, E, K, C, B-1, B-2, B-6, and B-12, niacin, folate, iron, zinc, copper, and iodine), daily multiple micronutrient supplement, or daily iron supplements.\nThe daily micronutrient supplement was found to be the most effective in controlling anemia and iron deficiency.\nThe results are biologically plausible since the vitamins included in the supplement may lead to better hematologic status through several mechanisms.\nFor example, vitamin C improves intestinal absorption of iron and may also enhance mobilization of iron stores and riboflavin is necessary for the synthesis of the globin component of hemoglobin.\nThe effects on growth indices have comparatively been less consistent; the review by Allen et al reported small yet statistically significant improvements in length/height and weight in children supplemented with multiple micronutrients.\nOn the other hand, the pooled analysis cited above found that infants receiving a daily micronutrient supplement had significantly greater weight gain, whereas there were no differences in height gain.\nIn another meta-analyses of effects of micronutrient interventions on growth of children under five years of age, Ramakrishnan et al. found that multiple micronutrient interventions improve linear growth only and had no effect on weight gain.\nAdditionally, a few studies of multiple micronutrient supplementation in adults with tuberculosis have also been equivocal in their results on weight gain.\nFor example, a study in Mwanza, Tanzania, found that multiple micronutrient supplementation (vitamins A, B-complex, C, D, and E, selenium, copper, and zinc) for the first two months of TB treatment led to reduced weight gain among the HIV-infected TB patients; the HIV-uninfected TB patients demonstrated a non-significant increase in weight at the end of follow-up.\nSeveral studies have found that serum albumin is lower among patients with TB; however, there are no studies assessing the effect of multivitamin supplementation on albumin levels among children with TB that we can directly compare our results to.\nThe increase in albumin in all children observed in this trial is probably a response to adequate treatment for TB.\nThe main limitations of our trial were the small sample size and a short period of supplementation and follow-up; this could have led us to miss a beneficial effect of multivitamins on growth indices in the overall cohort, the primary outcome.\nThe trial also was not designed to measure effects of multivitamins in subgroups such as those defined by age or HIV status; therefore, the findings of weight gain among the youngest children or height differences among HIV infected children between the ages of 6 months and 3 years cannot be treated as conclusive.\nFurther, it is possible that we may have included children with other diseases, as TB diagnosis was not optimal.\nIt is also possible that the nutrients such as zinc and vitamin D that were not included in our supplement are more essential for growth.\nThe results of this trial should be generalizable to children with TB, with or without HIV co-infection, in most resource-limited settings.\nThe multivitamin supplement that we used has the potential to have several beneficial effects including on immune function in growing children, particularly those with TB as they may have several underlying micronutrient deficiencies, including those of vitamins B-complex, C, and E.\nThese nutrients are extensively involved in the immune system and its ability to fight infectious diseases such as TB.\nFor example, the B-vitamins are involved in increasing lymphocyte production, cell-mediated cytotoxicity, delayed-type hypersensitivity responses, and antibody production.\nVitamin C helps improve T- and B-lymphocyte proliferative responses and reduces the concentration of proinflammatory cytokines.\nVitamin E is responsible for improving delayed type hypersensitivity skin response, increasing IL-2 production, neutrophil phagocytosis, lymphocyte proliferation, and antibody response to T-cell dependent vaccines, and reducing production of inflammatory cytokines such as TNF- \u03b1 and IL-6.\nHowever, we did not observe an association between immune markers such as CD4, CD8, and CD3 T-cell count with multivitamin supplementation, except for children older than 3 years of age.\nWe are not aware of any known age-specific effects of vitamins on CD8 cells in this age group that may explain this finding.\nIn conclusion, multivitamin supplementation had no effect on weight gain, the primary outcome of the trial.\nHowever, supplementation, even for a short period of eight weeks, improved the hematological profile of all children with TB and led to significant weight gain amongst the youngest patients (n = 22; age six weeks to six months).\nIt is possible that older children need even greater doses of such nutrients to demonstrate an effect and for longer periods of time.\nThe impact of multivitamin supplementation on other parameters such as treatment outcomes needs to be assessed in larger trials with a longer period of supplementation.\nIf proven to be efficacious, multivitamin supplementation could represent an inexpensive adjunct to anti-tuberculous therapy, particularly in resource-limited settings.\nTrial Profile.\n\nBaseline Characteristics of children with TB randomized in the trial (n = 255)\nVariable  | Placebo (n = 127) | Multivitamins (n = 128)\nAge, in years, Median (IQR)a | 1.50 (0.83, 2.42) | 1.54 (0.79, 2.92)\nShillingsb spent on food per household per day, Median (IQR)a | 2500.00(2000.00, 3000.00) | 2000.00(2000.00, 3000.00)\nAge categories, n (%) |  | \n\u2003\u2264 6 months | 9 (7.09%) | 13 (10.16%)\n\u20036 months to \u2264 3 years | 93 (73.23%) | 85 (66.40%)\n\u2003> 3 years | 25 (19.68%) | 30 (23.44%)\nFemales, n (%) | 60 (47.24%) | 56 (43.75%)\nHIV-infected, n (%) | 37 (29.13%) | 50 (39.06%)\n\u2003\u2264 6 months | 4 (44.44%) | 4 (30.77%)\n\u20036 months to \u2264 3 years | 23 (24.73%) | 29 (34.12%)\n\u2003> 3 years | 10 (40.00%) | 17 (56.67%)\nMarital Status of Mother, n (%) |  | \n\u2003Single | 15 (11.81%) | 15 (11.72%)\n\u2003Married | 81 (63.78%) | 70 (54.69%)\n\u2003Divorced | 2 (1.58%) | 3 (2.34%)\n\u2003Cohabiting | 20 (15.75%) | 32 (25.00%)\n\u2003Separated | 6 (4.72%) | 7 (5.47%)\n\u2003Widowed | 3 (2.36%) | 1 (0.78%)\nMother Literate, n (%) | 114 (89.76%) | 117 (91.41%)\nMother's Occupation, n (%) |  | \n\u2003Housewife | 75 (59.05%) | 71 (55.47%)\n\u2003Small business/Farm/Informal Income | 41 (32.28%) | 42 (32.81%)\n\u2003Business woman | 3 (2.36%) | 6 (4.69%)\n\u2003Public House/Restaurant | 1 (0.79%) | 1 (0.78%)\n\u2003Professional (Nurse/Teacher) | 2 (1.58%) | 2 (1.56%)\n\u2003Skilled Office Work | 0 (0.00%) | 2 (1.56%)\n\u2003Unskilled Employment | 3 (2.36%) | 4 (3.13%)\n\u2003Other | 2 (1.58%) | 0 (0.00%)\nMeat or fish consumption, n (%) |  | \n\u2003Never/Less than once per month | 2 (1.61%) | 4 (3.15%)\n\u2003Sometimes/1-3 times per month | 13 (10.48%) | 11 (8.66%)\n\u2003About once per week | 33 (26.61%) | 31 (24.41%)\n\u20032-4 times per week | 69 (55.65%) | 77 (60.63%)\n\u2003Everyday/5-7 times per week | 7 (5.65%) | 4 (3.15%)\nAdmitted to the Hospital, n (%) | 21 (16.54%) | 21 (16.41%)\n\na IQR: Inter-Quartile Range\nb 1000 Tanzanian Shillings \u2248 1 US Dollar at the time of the trial\n\nEffect of multivitamin supplementation on anthropometric measurements\nOutcome  | Time Point (n)a  | Placebo | Multivitamins | Hodges-Lehmann | p-valuec \n |  | Median (IQR)b  | Median (IQR)b  | 95% CI | \n |  |  |  | (Multivitamin-Placebo) | \nWeight, kg |  |  |  |  | \n | Baseline (255) | 7.95 (6.20, 10.75) | 8.48 (6.52, 10.81) | (-1.15, 0.36) | 0.31\n | Intermediate (245) | 8.52 (6.84, 11.43) | 9.38 (7.08, 11.81) | (-0.35, 1.28) | 0.30\n | Final (237) | 9.01 (7.20, 12.02) | 9.80 (7.74, 11.94) | (-0.38, 1.31) | 0.28\n | Change (237) | 0.84 (0.45, 1.30) | 0.84 (0.51, 1.30) | (-0.15, 0.18) | 0.82\nHeight, cm |  |  |  |  | \n | Baseline (255) | 74.00 (67.50, 87.00) | 77.50 (68.00, 86.55) | (-4.60, 1.00) | 0.28\n | Intermediate (245) | 75.00 (68.00, 87.00) | 77.45 (69.40, 88.00) | (-1.00, 4.80) | 0.23\n | Final (236) | 76.00 (69.00, 88.20) | 79.00 (70.00, 88.30) | (-1.00, 5.00) | 0.24\n | Change (236) | 1.10 (0.50, 2.00) | 1.50 (0.50, 2.50) | (-0.20, 0.50) | 0.53\nMUAC, cm |  |  |  |  | \n | Baseline (255) | 13.00 (11.20, 14.00) | 13.00 (11.90, 14.25) | (-0.70, 0.25) | 0.35\n | Intermediate (245) | 13.50 (12.00, 14.50) | 14.00 (12.50, 15.00) | (-0.10, 0.80) | 0.18\n | Final (237) | 13.80 (12.55, 15.00) | 14.00 (12.70, 15.00) | (-0.30, 0.50) | 0.56\n | Change (237) | 1.00 (0.20, 1.50) | 0.80 (0.20, 1.50) | (-0.35, 0.20) | 0.61\nHead Circumference, cm |  |  |  | \n | Baseline (254) | 45.50 (43.00, 48.00) | 46.00 (43.50, 48.00) | (-1.00, 0.50) | 0.58\n | Intermediate (245) | 46.00 (43.80, 48.00) | 46.45 (44.00, 48.00) | (-0.50, 1.00) | 0.59\n | Final (236) | 46.00 (44.10, 48.50) | 46.50 (44.40, 48.50) | (-0.50, 1.00) | 0.58\n | Change (235) | 0.50 (0.00, 1.00) | 0.50 (0.00, 1.00) | (-0.20, 0.25) | 0.86\nTriceps skin-fold thickness, cm |  |  |  | \n | Baseline (255) | 6.20 (5.20, 7.90) | 6.15 (5.15, 8.00) | (-0.60, 0.40) | 0.55\n | Intermediate (245) | 7.10 (5.60, 8.20) | 7.20 (6.10, 8.95) | (-0.20, 0.90) | 0.25\n | Final (237) | 7.50 (6.20, 8.60) | 7.20 (6.00, 8.90) | (-0.70, 0.40) | 0.61\n | Change (237) | 1.05 (-0.20, 2.10) | 0.70 (-0.30, 1.90) | (-0.70, 0.20) | 0.37\n\na Intermediate is at 4 weeks follow-up; Final at 8 weeks follow-up; Change between the final and the baseline measurement\nb IQR: Interquartile Range\nc p-value based on Wilcoxon 2-group comparison\n\nEffect of multivitamin supplementation on laboratory parameters\nOutcome  | Time Point (n)a  | Placebo | Multivitamins | Hodges-Lehmann | p-valuec \n |  | Median (IQR)b  | Median (IQR)b  | 95% CI | \n |  |  |  | (Multivitamin-Placebo) | \nHemoglobin, g/dL |  |  |  | \n | Baseline (251) | 8.70 (7.70, 9.80) | 8.40 (7.40, 9.50) | (-0.20, 0.60) | 0.28\n | Final (225) | 9.40 (8.45, 10.25) | 9.60 (8.70, 10.60) | (0.00, 0.70) | 0.05\n | Change (223) | 0.40 (-0.30, 1.40) | 1.00 (0.30, 2.30) | (0.30, 1.00) | 0.0001\nAlbumin, g/L |  |  |  | \n | Baseline (253) | 33.95 (29.90, 38.90) | 34.00 (29.80, 39.10) | (-1.70, 1.70) | 0.92\n | Final (232) | 40.20 (36.30, 42.90) | 38.90 (33.80, 42.50) | (-2.70, 0.10) | 0.08\n | Change (231) | 4.55 (0.10, 9.60) | 3.40 (0.60, 7.00) | (-2.60, 0.60) | 0.21\nCD4, cells/\u03bcL |  |  |  | \n | Baseline (242) | 1592.50 (1099.00, 2254.00) | 1351.50 (803.00, 2208.50) | (-410.00, 45.00) | 0.11\n | Final (212) | 1507.50 (1038.50, 2011.50) | 1410.50 (782.50, 1904.50) | (-373.00, 71.00) | 0.19\n | Change (207) | -63.00 (-506.00, 240.00) | -59.50 (-448.00, 261.00) | (-204.00, 140.00) | 0.73\nCD8, cells/\u03bcL |  |  |  | \n | Baseline (242) | 1233.00 (902.00, 2230.00) | 1492.00 (988.00, 2624.00) | (-16.00, 450.00) | 0.06\n | Final (212) | 1328.00 (921.50, 2208.50) | 1506.50 (1039.00, 2742.00) | (-98.00, 389.00) | 0.24\n | Change (207) | -130.00 (-489.00, 394.00) | -77.50 (-735.00, 432.00) | (-322.00, 193.00) | 0.68\nCD3, cells/\u03bcL |  |  |  | \n | Baseline (242) | 3317.50 (2423.00, 4616.00) | 3593.50 (2430.00, 5211.50) | (-311.00, 590.00) | 0.53\n | Final (212) | 3176.00 (2269.00, 4676.50) | 3540.50 (2871.00, 4614.50) | (-154.00, 681.00) | 0.21\n | Change (207) | -223.00 (-1149.00, 724.00) | -324.00 (-1179.00, 889.00) | (-519.00, 466.00) | 0.89\n\na Final at 8 weeks follow-up; Change between the final and the baseline measurement\nb IQR: Interquartile Range\nc p-value based on Wilcoxon 2-group comparison",
        "label": "low",
        "id": "task4_RLD_test_421"
    },
    {
        "paper_doi": "10.1186/1475-2875-12-254",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: An open label RCTFollow-up: Followed up for 42 days and asked to return on days 1, 2, 3, 7, 14, 21, 28, 35, and 42 after enrolment or at any day if ill. Clinical assessment and blood smear at each visit. Hb measured on days 0, 7, 14, 28, and 42.Adverse event monitoring: Not reported. \"Adverse events investigated and addressed\".\n\n\nParticipants: Number of participants: 274Inclusion criteria: Children aged 6 to 59 months with axillary temperature >= 37.5 degC or history of fever in preceding 48 hrs, weight >= 5.0 kg, parasitaemia, residing within 10 km of Siaya District Hospital, written informed consent.Exclusion criteria: Lethargy, convulsions, persistent vomiting, inability to drink, signs of severe malaria, severe anaemia (Hb < 5 g/dL), known hypersensitivity to trial drugs, presence of chronic medical conditions, treatment with any anti-malarial in preceding two weeks, previous enrolment in any malaria trial, severe malnutrition (weight-for-age <= 3 standard deviations below mean for gender according to WHO standards).\n\n\nInterventions: 1. DHA-P, fixed dose combination, 20 mg/160 mg tablets (DuoCotexin: Beijing Holley-Cotec)5 to 6 kg: one half tablet daily7 to 9 kg: one tablet daily10 to 14 kg: two tablets on day 0 then one tablet on days 1 and 215 to 19 kg: two tablets daily2. Artemether-lumefantrine, fixed dose combination, 20 mg/120 mg (Coartem: Novartis)5 to 14 kg 1 tablet twice daily for 3 days15 to 24 kg 2 tablets twice daily for 3 days25 to 34 kg 3 tablets twice daily for 3 daysAll doses, except AL evening doses, administered under direct supervision.\n\n\nOutcomes: ACPR at days 28 and 42, PCR-unadjusted and PCR-adjustedMean change in Hb from baseline to day 28Adverse eventsNot included in this review:Fever clearanceParasite clearance\n\n\nNotes: Country: KenyaSetting: district hospital in western KenyaTransmission: Holoendemic with high transmission and two seasonal peaks, April to July and November to DecemberResistance: Not reportedDates: Oct 2010 to Aug 2011Funding: KEMRI/CDC Research and Public Health Collaboration, Beijing Holley-Cotec provided DHA-P free of charge\n\n",
        "objective": "To evaluate the effectiveness and safety of DHA\u2010P compared to other ACTs for treating uncomplicated P. falciparum malaria in adults and children.",
        "full_paper": "Background\nArtemether-lumefantrine (AL) was adopted as first-line treatment for uncomplicated malaria in Kenya in 2006.\nMonitoring drug efficacy at regular intervals is essential to prevent unnecessary morbidity and mortality.\nThe efficacy of AL and dihydroartemisinin-piperaquine (DP) were evaluated for the treatment of uncomplicated malaria in children aged six to 59 months in western Kenya.\nMethods\nFrom October 2010 to August 2011, children with fever or history of fever with uncomplicated Plasmodium falciparum mono-infection were enrolled in an in vivo efficacy trial in accordance with World Health Organization (WHO) guidelines.\nThe children were randomized to treatment with a three-day course of AL or DP and efficacy outcomes were measured at 28 and 42 days after treatment initiation.\nResults\nA total of 137 children were enrolled in each treatment arm.\nThere were no early treatment failures and all children except one had cleared parasites by day 3.\nPolymerase chain reaction (PCR)-uncorrected adequate clinical and parasitological response rate (ACPR) was 61% in the AL arm and 83% in the DP arm at day 28 (p\u2009=\u20090.001).\nPCR-corrected ACPR at day 28 was 97% in the AL group and 99% in the DP group, and it was 96% in both arms at day 42.\nConclusions\nAL and DP remain efficacious for the treatment of uncomplicated malaria among children in western Kenya.\nThe longer half-life of piperaquine relative to lumefantrine may provide a prophylactic effect, accounting for the lower rate of re-infection in the first 28 days after treatment in the DP arm.\nBackground\nResistance to anti-malarials by Plasmodium falciparum has been an ongoing global public health concern since chloroquine resistance emerged in the 1960s.\nDrug pressure is considered a major factor driving parasite resistance, as a result of the increased availability of medications in the public and private sectors, increased prevalence of sub-therapeutic drug concentrations, and counterfeit medications containing inadequate amounts of active ingredients, among other factors.\nIn Kenya, chloroquine was first-line treatment for uncomplicated P. falciparum malaria until 1998, despite the presence of high levels of chloroquine resistance since the early 1990s.\nIn 1998, sulphadoxine-pyrimethamine (SP) was adopted as first-line treatment, but resistance rapidly emerged.\nBoth sulphadoxine and chloroquine have long half-lives and consequent prolonged parasite exposure to subtherapeutic drug levels, which can contribute to resistance.\nArtemisinins, the newest class of anti-malarials, have a very short half-life (<8 hours) and rapidly reduce parasite burden; their use in combination with other anti-malarials decreases the chance of resistance emergence.\nHowever, artemisinin resistance has developed along the Thai-Cambodian border, likely due to its use at sub-therapeutic doses and as monotherapy.\nArtemisinin-based combination therapy (ACT) has superior efficacy and the potential to prevent drug resistance by incorporating a partner drug to enhance parasite clearance.\nACT has been adopted as first-line treatment in most malaria-endemic countries.\nArtemether-lumefantrine (AL), a highly effective ACT, is commonly used in many African countries as first-line treatment.\nIn 2004, Kenya adopted AL as first-line malaria treatment, but it was not widely implemented until 2006.\nDihydroartemisinin-piperaquine (DP), another ACT regimen, has been studied in East Africa as an alternative to AL.\nIts advantages over AL include once-daily dosing and a longer half-life of the partner drug, which may prevent re-infection in areas of intense malaria transmission.\nStudies have shown equivalent safety and efficacy profiles for DP and AL.\nIn 2010, DP was adopted as second-line treatment for uncomplicated P. falciparum malaria in Kenya.\nRegular monitoring of anti-malarial efficacy is essential to better inform national malaria policies.\nThis study was conducted in western Kenya to determine if AL remains efficacious for the treatment of uncomplicated malaria after five years of its implementation and to evaluate the efficacy of DP in this population.\nMethods\nStudy site and enrolment\nThis study was conducted between October 2010 and August 2011 at Siaya District Hospital (SDH) in Nyanza Province, western Kenya.\nThis region is holoendemic for P. falciparum with high malaria transmission and two seasonal peaks, April to July and November to December.\nThe entomological inoculation rate (EIR) in this area, historically around 300 infectious bites per person per year, has recently been estimated to be ten infectious bites per person per year (Gimnig J, pers comm).\nStudy subjects were recruited from the outpatient paediatric department of SDH, which serves approximately 100 patients per day.\nSubjects\nChildren aged six to 59 months with P. falciparum mono-infection were enrolled.\nAdditional inclusion criteria were axillary temperature \u226537.5\u00b0C or history of fever in the previous 48 hours, weight \u22655.0 kg, parasitaemia 1-200,000 asexual forms per \u03bcL (initially 2,000-200,000 but protocol amended in January 2011 to include any parasitaemia <200,000 as these drug regimens are used to treat patients with any level of parasitaemia in Kenya), residence within 10 km of SDH, and written informed consent by caregiver.\nSubjects were excluded if any of the following were present: lethargy, convulsions, inability to drink, persistent vomiting, symptoms of severe malaria, severe malnutrition (weight-for-age \u22643 standard deviations below the mean for gender according to World Health Organization (WHO) standards), severe anaemia (haemoglobin (Hb) <5 g/dl), known hypersensitivity to study drugs, presence of febrile illness other than malaria (e.g. measles, pneumonia), presence of chronic medical conditions, treatment with any anti-malarial in the previous two weeks, or previous enrolment in any malaria study.\nEthical considerations\nThis study received ethical clearance from the US Centers for Disease Control and Prevention (CDC, Atlanta, USA) and the Kenya Medical Research Institute (KEMRI, Nairobi, Kenya).\nWritten informed consent was obtained from caregivers of enrolled subjects and a long-lasting insecticide-treated bed net (ITN) was provided to enrolled subjects.\nClinical and laboratory procedures\nThis was a 42-day, open-label in vivo trial.\nInitial screening was offered to patients with fever or history of fever.\nCaregivers were then asked about interest in study participation.\nAfter consenting patients were screened for inclusion criteria, a rapid diagnostic test (RDT) (SD Bioline malaria Pf/pan, Standard Diagnostics Inc, Yongin, South Korea) for malaria and Hb testing (Hemocue\u00ae Hb 201+, Hemocue AB, Angelholm, Sweden) were performed.\nIf the RDT was positive and Hb was \u22655.0 g/dL, two thick and thin blood films were collected to assess parasitaemia and confirm malaria species.\nBlood films were read independently by two microscopists by counting the number of asexual parasites against 500 white blood cells (WBCs).\nSlides were considered to be negative only after examining fields containing 1,000 WBCs.\nThe geometric mean of the two readings was considered in the analyses.\nSlides with parasite densities discordant by more than 50% or with positive and negative results were re-examined by a third microscopist; the mean of the third read and the closest of the first two slides was considered final.\nAll microscopists were blinded to the treatment arm and were certified as expert readers through a quality assurance programme at the South African National Institute for Communicable Diseases.\nCaregivers of enrolled children were interviewed and children were examined by a study clinician.\nChildren were block randomized in fixed blocks of ten to treatment with AL (Coartem\u00ae; Novartis, Basel, Switzerland) or DP (DuoCotexin\u00ae; Holley-Cotec Pharmaceuticals, Beijing, China).\nSamples of the AL and DP used in this study were sent to CDC laboratories for quality testing using high-performance liquid chromatography (HPLC) (Agilent Technologies, Waldbronn, Germany).\nBoth treatments were co-formulated, fixed-dose ACT regimens and were administered under direct observation by study staff at the study clinic (except AL evening doses) for three consecutive days.\nAL tablets, consisting of 20 mg of artemether and 120 mg lumefantrine, were administered twice daily according to patient weight: 5-14 kg: one tablet per dose; weight 15-24 kg: two tablets per dose; weight 25-34 kg: three tablets per dose.\nMorning doses were given with milk and directly observed in the study clinic.\nCaregivers were given evening doses to administer at home with food or milk.\nDP tablets, consisting of 20 mg dihydroartemisinin and 160 mg of piperaquine phosphate, were administered once daily by study staff according to patient weight: 5-6 kg: one-half tablet daily; 7-9 kg: one tablet daily; 10-14 kg: two tablets on day 0, then one tablet on days 1 and 2; 5-19 kg: two tablets daily.\nA full dose was re-administered if the patient vomited within 30 min or a half dose if vomiting occurred between 31 and 60 min.\nPatients with vomiting within 30 min of the second dose were referred for parenteral treatment and withdrawn from the study.\nFollow up\nChildren were followed for 42 days and asked to return on days 1, 2, 3, 7, 14, 21, 28, 35, and 42 following enrolment, as well as any day if ill.\nThe study clinic was open daily during regular hours; study personnel provided after-hours care at SDH.\nA clinical assessment was performed and blood smears were collected at each study visit.\nHb levels were measured on days 0, 7, 14, 28, and 42.\nA filter paper blood spot was collected on days 0, 3, and 7 and in case of suspected treatment failure for molecular analysis.\nAdverse events were investigated and addressed.\nOutcomes\nEfficacy was assessed by clinical and parasitological outcomes using WHO definitions.\nChildren were classified as early treatment failure (ETF) if any of the following criteria were met: development of severe malaria by day 3, day 2 parasitaemia\u2009>\u2009day 0 parasitaemia, presence of parasites on day 3 with axillary temperature \u226537.5\u00b0C, or day 3 parasitaemia >25% of day 0 parasitaemia.\nChildren not meeting ETF criteria with P. falciparum parasitaemia occurring between day 7 and 28 or 42 without fever were classified as late parasitological failure (LPF).\nThose with fever occurring between day 4 and day 28 or day 42 with parasitaemia were classified as late clinical failure (LCF).\nIf no failure was recorded by day 28 or day 42, the outcome was classified as adequate clinical and parasitological response (ACPR).\nAll treatment failures with uncomplicated malaria were treated with AL and treatment failures with severe malaria were treated with parenteral quinine.\nFollow-up ended once a study subject met one of the four classification criteria: ETF, LPF, LCP or ACPR.\nMolecular analysis\nTo differentiate between recrudescence and re-infection, a genotypic analysis based on merozoite surface protein-2 (msp2), glutamate-rich protein (glurp), and merozoite surface protein-1 (msp1) was performed by PCR using filter-paper blood spots.\nRecrudescence was defined as at least one identical allele for each of the three markers (msp2, glurp, and msp1) in the pre- and post-treatment samples.\nStatistical analysis\nPrimary efficacy outcomes included day 28 and day 42 ACPR, both PCR-corrected and PCR-uncorrected for each ACT regimen.\nSecondary outcomes included haematologic response, rates of fever clearance and parasite clearance by day 3, rates of ETF, LPF and LCF.\nAssuming a PCR-corrected ACPR of 95% and 20% loss to follow up at 42 days, a sample size of 137 children per study arm was chosen, 274 children in total; this allowed for a precision rate of +/- 4.5% at a 5% significance level.\nThis study was not powered to detect a difference in efficacy between treatment arms.\nStudy forms were scanned into a Microsoft Access 2000 (Microsoft, Redmond, USA) database.\nStatistical analysis was performed using SAS\u00ae 9.2 (SAS Institute, Cary, USA).\nPer protocol (PP) analysis of outcomes excluded those children withdrawn from the study for any reason.\nIntention-to-treat (ITT) analysis was performed using survival analysis.\nKaplan-Meier curves were estimated for both 28 and 42 days of follow up; the log-rank test was used for comparing the curves.\nFor ITT analysis, all withdrawals, losses to follow up and treatment failures were censored on the last day of follow up.\nComparisons were made using \u03c72 test for categorical variables and Student\u2019s t-test or Wilcoxon rank-sum test (for non-parametric data) for continuous variables.\nA two-sided p-value <0.05 was considered statistically significant.\nResults\nBaseline characteristics\nA total of 669 children with fever or history of fever were screened (Figure\u00a01).\nAmong those, 420 (63%) were RDT-positive and 324 (77%) of RDT-positive children had microscopy-confirmed P. falciparum mono-infection.\nOf these remaining 324 eligible children, 50 (15%) did not meet other inclusion criteria or did not give consent.\nTwo hundred and seventy-four children were enrolled in the study, 137 in each arm.\nAmong those enrolled, 224 were included in the day 42 analysis, 111 in the AL arm and 113 in the DP arm.\nBaseline characteristics of the children enrolled in the two arms were similar (Table\u00a01).\nThe percentage of children withdrawn from analysis was similar between the two arms, 18.2% in AL arm and 17.5% in the DP arm (p\u2009=\u20090.88).\nReasons for withdrawal are shown in Figure\u00a01.\nClinical and parasitological outcomes\nNo ETF was observed in either treatment arm (Table\u00a02).\nFollow up was completed for 116 children in each arm up to day 28.\nPCR-uncorrected ACPR was 61% in the AL arm and 83% in the DP arm (p\u2009=\u20090.001).\nPCR-uncorrected results showed day 28 LCF and LPF were 11% and 28% in the AL arm, respectively; in the DP arm, day 28 LCF and LPF were 3% and 14%, respectively.\nPCR analysis revealed the majority of LCF and LPF cases in both arms were due to re-infection, 34 (92%) of 37 in the AL arm and 18 (95%) of 19 in the DP arm.\nHowever, the re-infection rate by day 28 in the DP arm was significantly lower than that of the AL arm (p\u2009=\u20090.03).\nAt day 28, PCR-corrected ACPR was 97% for AL and 99% for DP (p\u2009=\u20090.5).\nThe nine children for whom PCR results were missing due to lost filter-paper samples were excluded from the PCR-corrected analysis (eight from the AL arm and one from the DP arm).\nFor day 42, analysis was completed for 111 children in the AL arm and 113 children in the DP arm.\nThe PCR-uncorrected ACPR was 44% for the AL arm and 54% for the DP arm (p\u2009=\u20090.14).\nPCR-uncorrected results showed that the day 42 LCF and LPF were 18% and 37% in the AL arm, respectively; while LCF was 10% and LPF was 36% in the DP arm.\nThe re-infection rate was similar for the two arms at day 42 (p\u2009=\u20090.7).\nAfter PCR correction, ACPR for both arms was 96%.\nSimilarly, the children for whom PCR results were missing were excluded from analysis (ten in AL arm and four in DP arm).\nTo ensure the results of the PCR-corrected analysis were not biased by missing samples, sensitivity analyses were performed assuming all missing PCR results were due to re-infection and all were due to recrudescence, resulting in a range of PCR-corrected ACPR.\nFor day 28, the maximum ranges of PCR-corrected ACPR for the treatment arms are 91-97% (95% CI: 85-100%) in the AL arm and 98-99% (95% CI: 96-100%) in the DP arm.\nFor day 42, ACPR would be 87-96% (95% CI: 81-100%) in the AL arm and 92-96% (95% CI: 87-100%) in the DP arm.\nAt day 28, the survival analysis using the ITT definition demonstrated a PCR-uncorrected cure rate of 67% for AL and 85% for DP (p\u2009=\u20090.0004) (Figure\u00a02).\nAt day 42, the PCR-uncorrected cure rates were 55% for AL and 62% for DP (p\u2009=\u20090.22).\nThe PCR-corrected cure rates at day 28 were 98% for AL and 99% for DP (p\u2009=\u20090.28).\nAt day 42, PCR-corrected cure rates were 97% for both drugs.\nLaboratory outcomes\nTreatment with either AL or DP resulted in rapid parasite clearance.\nAlthough >75% of children in both arms remained parasitaemic on day 1, only four (3%) and five (4%) children remained parasitaemic on day 2 in the AL and DP arms, respectively.\nOne child in the study remained parasitaemic on day 3 (AL arm).\nOver 90% of children were afebrile by day 1 in both treatment arms.\nMean Hb of children who were not re-infected increased from a baseline of 9.8 g/dL to 11.6 g/dL at day 42, whereas the mean Hb of those re-infected increased from a baseline of 9.9 g/dL to 11.1 g/dL on the last study day (p\u2009=\u20090.9).\nThe change in Hb from baseline to the study endpoint was similar among the two study arms.\nDrug samples tested for quality assurance contained adequate concentrations of active ingredients.\nAdverse events\nThere were three children in the DP arm who vomited the drug twice following enrolment and were referred for alternative treatment.\nThe rates of vomiting for the first dose of medication were similar for AL and DP (3.7% and 4.9%, respectively) and not associated with age.\nThree enrolled children developed severe malaria more than 28 days after treatment; two in the AL arm and one in the DP arm.\nThese children were hospitalized for parenteral treatment.\nThese outcomes are attributable to re-infection (confirmed by PCR analysis), not poor efficacy of treatment regimens.\nAll children recovered completely; no deaths occurred during the study.\nDiscussion\nBoth AL and DP are efficacious in treating uncomplicated P. falciparum malaria in children in western Kenya.\nRecurrent parasitaemia among children under five years is frequent in this area despite high coverage with ITNs (70% household ownership and 42% usage in children under five years) and is mostly secondary to re-infection with P. falciparum.\nAs seen in other African countries, recurrent parasitaemia occurs significantly more frequently in those children treated with AL in the first 28 days.\nThis is likely due to the longer half-life of the piperaquine component of DP, which provides long-lasting prophylactic effect.\nThis study was not powered to compare the efficacy of the two drug regimens; however, there was a significant difference in re-infection at day 28 between the two groups.\nData from similar studies conducted in western Kenya from 2005 to 2009 assessing the efficacy of AL and DP had similar results.\nData collected in 2005 on children with uncomplicated malaria and treated with AL showed PCR-uncorrected ACPR at day 28 of 71% and at day 42 of 41%, compared to 61% and 44% in this study.\nHowever, the high numbers of recurrent parasitaemia were likely due to re-infection and not recrudescence, as it was the case in this and other studies.\nA smaller study conducted in 2007, showed that only one (1.5%) out of 67 patients had recurrent parasitaemia at day 28 after treatment with AL and no episodes of recurrent parasitaemia were detected at day 28 in the DP group.\nWhen comparing these data to the higher rates of recurrent parasitaemia in our study, we do not believe it is due to declining drug efficacy, but rather to the high frequency of re-infection with P. falciparum in children in the study area.\nLastly, data collected in 2009 in the Mbita region of western Kenya had similar PCR-corrected ACPR in the AL group (98.6% at day 28 and 97.2% at day 42) as what was observed in this evaluation, which further demonstrates the continued efficacy of AL in the region.\nThere is growing concern about the emergence of artemisinin resistance, signalled by delayed parasite clearance, as observed in Southeast Asia.\nOne recent study from the Kenyan coast reports decreased parasite clearance rates in those treated with ACT since 2006.\nThis study showed that although over 75% of children remained parasitaemic on day 1, over 95% had cleared parasitaemia by day 2, and only one child had parasites on day 3.\nThese clearance rates are similar to those seen in an efficacy study with AL conducted in this region in 2006.\nAs WHO defines suspected artemisinin resistance as \u226510% of cases with parasitaemia at day 3 after ACT treatment initiation, this study provides no evidence to suggest artemisinin resistance in western Kenya.\nAlthough the annual EIR in western Kenya has decreased dramatically, from 300 infectious bites per person per year in 1990 to ten in 2010, malaria prevalence during the peak transmission season in children under five years of age remains high at 42% (KEMRI/CDC, unpublished data).\nA previous study demonstrated that in several areas in Africa with annual EIRs in the range of one to ten infectious bites per person per year, the mean parasite prevalence remained similarly high.\nIn fact, parasite prevalence was not meaningfully reduced until the annual EIR was well below one infectious bite per person per year.\nThe high parasite prevalence despite low EIR is multifactorial.\nSome plausible reasons include recurrent infections due to Plasmodium vivax and Plasmodium ovale, lack of access to care resulting in delayed treatment, asymptomatic carriers, and frequent anti-malarial stock-outs.\nThis epidemiological context likely accounts for the high re-infection rates observed in this study, despite a large reduction in EIR over the past two decades.\nBoth drug regimens were well tolerated with a low frequency of vomiting.\nAlthough the frequency of vomiting the first dose was similar between the two treatment arms, only those in the DP arm (n\u2009=\u20093) vomited the drug twice and required alternative treatment.\nIncreased vomiting in children aged six to 24 months with uncomplicated malaria treated with DP, compared to AL, has been noted previously.\nVomiting did not seem to be related to younger age in either group.\nHowever, this study was not powered to detect such a difference.\nThere are a few limitations to this study.\nFirst, the evening doses of AL were not directly observed.\nAlthough caregivers were asked to confirm administration of evening doses and children who missed doses were withdrawn, the efficacy of AL may be underestimated in this study if some missed doses were not reported.\nSimilar studies have used the same approach.\nHowever, PCR-corrected efficacy of AL was still high in this study.\nIn addition, PCR data are missing for some children with treatment failure.\nNevertheless, as recrudescence was responsible for <10% of failures in both arms, the ACPRs calculated considering only available PCR samples are likely accurate.\nAlso, parasitaemia range for study inclusion was altered to include all children with P. falciparum mono-infection.\nThis increased the risk of including false positives; however, only certified microscopists read slides and only nine subjects had parasitaemia <2,000 parasites/\u03bcL. Lastly, contrary to the original protocol, two months after initiation of recruitment, study staff excluded some children with any history of vomiting.\nThe number of children excluded for this reason is unknown, as only persistent vomiting was an original exclusion criterion.\nTherefore, this study may underestimate the incidence of vomiting associated with the treatment regimens, as well as inadvertently excluding some children with high parasitaemia or high fever.\nDP may benefit children in this high-transmission setting because the long half-life of piperaquine is associated with a decreased re-infection rate in the first 28 days after treatment compared to AL.\nThis longer prophylactic effect may allow more time for Hb recovery, thus decreasing the severity of re-infections.\nThe once-a-day dosing is another advantage of this regimen and may improve adherence.\nHowever, the higher cost of a treatment dose of DP compared to AL, US$4 and US$1, respectively, may be a barrier to its use as first line.\nIn addition, DP may be more prone to the development of resistance because of the long half-life of piperaquine.\nConclusions\nThe results of this study demonstrate that AL and DP remain efficacious treatment regimens for uncomplicated P. falciparum malaria in western Kenya.\nWith day 3 parasite clearance rates of nearly 100%, there is no evidence of delayed parasite clearance to indicate emerging artemisinin resistance.\nFollowing WHO recommendations, regular monitoring to evaluate anti-malarial efficacy at least every two years should be maintained to confirm the continued efficacy of first-line anti-malarial therapy.\nTrial profile, western Kenya, 2011. Legend: RDT: Rapid diagnostic test for malaria; f/u: follow up; neg: negative.\nSurvival curve of enrolled children by PCR-uncorrected data, western Kenya 2011. Legend: This graph is using intention to treat (ITT) analysis.\n\nBaseline characteristics of children upon enrolment for artemether-lumefantrine (AL) and dihydroartemisinin-piperaquine (DP), western Kenya 2011\nCharacteristic | AL (n\u2009=\u2009137)(95% CI) | DP (n\u2009=\u2009137)(95% CI) | p-value\nMean age (months) | 36.1 (33.8\u201338.5) | 33.5 (31.1\u201335.9) | 0.11\nMale (%) | 55 | 57 | 0.9\nMean weight (kg) | 13.3 (12.8\u201313.8) | 13.0 (12.5\u201313.4) | 0.3\nMean axillary temperature (\u00b0C) | 37.6 (37.4\u201337.8) | 37.7 (37.5\u201338.0) | 0.35\nMean haemoglobin (g/dL) | 9.7 (9.4\u201310.0) | 9.9 (9.7\u201310.2) | 0.28\nGeometric mean day 0 parasite density (parasites/ \u03bcL) (range) | 45,168 (34,506\u201347,190) (10\u2013148,027) | 49,248 (35,188\u201352,544) (54\u2013166,584) | 0.49\n\n\nClinical and parasitological response rates for artemether-lumefantrine (AL) and dihydroartemisinin-piperaquine (DP) using per protocol analysis, western Kenya 2011\nOutcome | AL | DP | p-value\n\u00a0 | % (95% CI) | % (95% CI) | \u00a0\nEarly treatment failure | 0% (0%\u20133%) (0/137) | 0% (0%\u20133%) (0/137) | 1\nDay 3 parasite clearance | 99% (96%\u201399%) (130/131) | 100% (97%\u2013100%) (126/126) | 0.34\nDay 28 PCR-uncorrected ACPR* | 61% (52%\u201370%) (71/116) | 83% (75%\u201389%) (96/116) | 0.001\nDay 28 PCR-corrected ACPR** | 97% (92%\u201399%) (105/108) | 99% (95%\u2013100%) (114/115) | 0.48\nDay 42 PCR-uncorrected ACPR | 44% (35%\u201354%) (49/111) | 54% (45%\u201363%) (61/113) | 0.14\nDay 42 PCR-corrected ACPR | 96% (90%\u201399%) (97/101) | 96% (91%\u201399%) (105/109) | 0.26\n\n*Polymerase chain reaction-uncorrected adequate clinical and parasitological response.\n**Polymerase chain reaction-corrected adequate clinical and parasitological response.",
        "label": "unclear",
        "id": "task4_RLD_test_526"
    },
    {
        "paper_doi": "10.1007/s10151-020-02372-w",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallel design, 2 armsEthics and informed consent: ethics approved, informed consent \nFollow-up period: 7 days and 14 days postoperativelySample size estimate: primary endpoint incidence in our population = 33%; \"to demonstrate that NPWT decreased WHC by 80%, a total sample size of 70 subjects was needed for an alpha of 0.05 and 80% power. Thus, with expectations of omissions, we sought a total sample size of 38 patients in each arm\".ITT analysis: available case analysisFunding: funded only by the Jagiellonian University Medical College own funds, without any financial or material support from the NPWT equipment producersPreregistration: clinicaltrials.gov (NCT04088162) \n\n\nParticipants: Location: Poland\nIntervention group: 38 participants allocated to the group, and 35 analysed; control group: 37 participants allocated to the group, and 36 analysedMean age: intervention group mean age 61.6 +- 11.3 years; control group mean age 62.4 +- 11.3 years\nInclusion criteria: patients aged >= 18 years with a history of surgery for colorectal cancer, including formation of the protective ileostomy, who were scheduled to undergo ileostomy closure as an elective procedure.\nExclusion criteria: emergency/urgent operation, active infection, operations other than ileostomy closure, or parastomal hernioplasty. Patients who required a second operation or transfer to the intensive care unit or other hospital wards because of non-infectious complications within the first week after surgery were excluded from analysis (retrospectively).\n\n\nInterventions: Aim/s: to assess the usefulness of protective negative-pressure wound therapy (NPWT) in the reduction of wound healing complications (WHC) and surgical site infections (SSI) after diverting ileostomy closure in patients who underwent surgery for colorectal cancer.Group A (NPWT) intervention: postoperative NPWT, NANOVA negative-pressure dressing placed over the entire length of the incision, which was taken out at 72 h.Group B (control) intervention: customary care (without postoperative NPWT), sterile wound dressing placed over the incision with the first dressing change at 48 h postoperatively, and then daily until the removal of sutures on postoperative day 7.\nStudy date/s: January 2016 to December 2018\n\n\nOutcomes: Wound healing complications (any wound condition requiring postoperative intervention; may include review eligible outcomes)Surgical site infectionsHaematomaSeromaValidity of measure/s: wound healing complications were defined as any condition of the wound that required postoperative intervention other than a change of dressing or removal of stiches. Incisional surgical site infection diagnosis was made according to the criteria of the Center for Disease Control (CDC) and European Centre for Disease Prevention and Control (ECDC) for diagnosis of surgical site infection.Time points: 7 days and 14 days postoperatively\n\n\nNotes: \n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "Background\nThe aim of this study was to assess the usefulness of protective negative-pressure wound therapy (NPWT) in the reduction of wound healing complications (WHC) and surgical site infections (SSI) after diverting ileostomy closure in patients who underwent surgery for colorectal cancer.\nMethods\nIn this prospective randomized clinical trial in a tertiary academic surgical center, patients who had colorectal cancer surgery with protective loop ileostomy and were scheduled to undergo ileostomy closure with primary wound closure from January 2016 to December 2018 were randomized to be treated with or without NPWT.\nThe primary endpoint was the incidence of WHC.\nSecondary endpoints were incidence of SSI, length of postoperative hospital stay (LOS), and length of complete wound healing (CWH) time.\nResults\nWe enrolled 35 patients NPWT (24 males [68.6%]; mean age 61.6\u2009\u00b1\u200911.3\u00a0years), with NPWT and 36 patients (20 males [55.6%]; mean age 62.4\u2009\u00b1\u200911.3\u00a0years) with only primary wound closure (control group).\nWHC was observed in 11 patients (30.6%) in the control group and 3 (8.57%) in the NPWT group (p\u2009=\u20090.020).\nPatients in the NPWT group had a significantly lower incidence of SSI (2 [5.71%] vs. 8 [22.2%] in the control group; p\u2009=\u20090.046) as well as significantly shorter median CWH (7 [7\u20137] days vs. 7 [7\u201315.5] days, p\u2009=\u20090.030).\nThere was no difference in median LOS between groups (3 [2.5\u20135] days in the control group vs. 4 [2\u20134] days in the NPWT group; p\u2009=\u20090.072).\nConclusions\nProphylactic postoperative NPWT after diverting ileostomy closure in colorectal cancer patients reduces the incidence of WRC and SSI.\nClinical trial registration\nclinicaltrials.gov (NCT04088162).\nIntroduction\nSurgeries for colorectal cancer, especially procedures involving the lower rectum, are associated with a very high percentage of complications.\nDespite the current recomsmendation for primary anastomosis after a resection procedure, such treatment is associated with a significant risk of anastomotic leak (up to 20%\u2009=\u2009).\nOne way to reduce the risk of leak is to use a diverting ileostomy.\nAlthough this is currently considered a standard treatment, especially in the group of patients undergoing neoadjuvant radiotherapy or chemoradiotherapy, the technique still has its drawbacks, such as the necessity to perform an additional surgical operation for ileostomy closure, which has a high risk of wound healing complications, particularly surgical site infections (SSI).\nTo reduce the frequency of these complications, novel guidelines published in 2017 recommend using the purse-string suture technique.\nHowever, this has been associated with as much as a fivefold increase in healing time, as well as less desirable cosmetic effects, in patients without SSI.\nNegative-pressure wound therapy (NPWT) is currently a widely used method of treatment in various types of infectious complications, potentially providing the opportunity to prevent infectious complications after surgery by combining the benefits of both postoperative wound closure techniques: reduced healing time, compared to primary closure, and reduced risk of infectious complications, compared to purse-string techniques.\nThe aim of this randomized controlled trial was to assess the usefulness of postoperative NPWT in the reduction of postoperative wound-healing complications (WHC) and SSI after diverting ileostomy closure in patients who underwent colorectal resection for cancer.\nMaterials and methods\nPatients\nA randomized controlled trial was conducted between January 2016 and December 2018 in a tertiary referral center, University Hospital (Krakow, Poland).\nThe trial was registered at clinicaltrials.gov (NCT04088162) after approval of the protocol by ethics committee of Jagiellonian University Medical College (#1072.6219.263.2019).\nThe trial was designed as a single-center, randomized controlled, superiority trial with two parallel intervention arms.\nParticipants\nPatients aged\u2009\u2265\u200918\u00a0years with a history of surgery for colorectal cancer, including formation of the protective ileostomy, who were scheduled to undergo ileostomy closure as an elective procedure, were randomly divided into two groups: Group 1 to undergo postoperative NPWT and Group 2, a control group to undergo customary care (without postoperative NPWT).\nPatients were enrolled after providing informed consent on admission.\nExclusion criteria were emergency/urgent operation, active infection, operations other than ileostomy closure, or parastomal hernioplasty.\nPatients who required a second operation or transfer to the intensive care unit or other hospital wards because of noninfectious complications within the first week after surgery were also excluded from the analysis.\nRandomization\nThe 1:1 randomization with concealment was achieved using a random number generator (even/odd).\nUntil the end of the operation, patients did not know to which group they were assigned.\nThe randomization process and assignment of the patients to the groups were performed by a trial researcher who was not directly involved in the operation or postoperative care of the patient.\nOperating surgeons were also blinded to the randomization.\nThe NPWT dressing was set up at the end of the operation in sterile conditions in the operating room by one and the same person (the designated surgeon, a member of the research team) not directly involved in the operation or postoperative patient care.\nSample size calculation\nOur previous observations found the primary endpoint incidence in our population to be 33%, which was consistent with the data available in the literature.\nAdditionally, a previous pilot study demonstrated that NPWT had decreased the incidence of WHC by 70\u201385%.\nTo demonstrate that NPWT decreased WHC by 80%, a total sample size of 70 subjects was needed for an alpha of 0.05 and 80% power.\nThus, with expectations of omissions, we sought a total sample size of 38 patients in each arm.\nProcedures\nPatients\u2019 demographics, possible SSI risk factors, including age, sex, body mass index (BMI), active smoking, preoperative immunosuppressive treatment, incidence of comorbidities, amount of intraoperative bleeding, and surgery duration were prospectively collected.\nSurgical technique\nAll patients enrolled in the study had previously undergone a resection procedure for colorectal cancer with the simultaneous formation of a diverting loop ileostomy 20\u201330\u00a0cm proximally from the ileocecal valve at the ileal loop, which was delivered through a circular incision on the right lower abdominal wall without mesenteric torsion.\nIleostomy closure, as a second operation, was performed approximately 6\u00a0months after initial surgery, after adjuvant chemotherapy (if necessary).\nPatients with American Joint Committee on Cancer\u2014AJCC\u2014stage 0 or 1 had the ileostomy removal procedure performed much earlier, as early as 14\u00a0days, after the histopathological examination results were obtained.\nAt ileostomy closure, a circumferential incision around the ileostomy was performed.\nAdhesions were gently detached from the abdominal wall with scissors.\nAfter small bowel mobilization, the short-segment small bowel resection (approximately 15\u201325\u00a0cm) was performed.\nAnastomosis was performed via end-to-end single polydioxanone (PDS) Plus 4\u20130 running suture.\nWound closure was done along three layers, including the peritoneum layer, rectus abdominis fascia, and subcutaneous layer.\n2\u20130 absorbable PDS Plus running suture was used for peritoneum and fascia layer, and 3\u20130 Vicryl Plus single sutures were used to close the subcutaneous layer.\nIn the five cases of parastomal hernia (three in the NPWT group and two in the control group) it was necessary to perform hernia repair using a polypropylene mesh, as in the sub-lay method.\nIn those patients, the silicone drain was placed in the subfascial layer and removed at 2 or 3 postoperative days (after exudation reduction to less than 30\u00a0ml/day).\nIn the control group, the skin was closed by six to eight single non-absorbable Monosyn 3\u20130 loose sutures every 7\u20139\u00a0mm, and a sterile wound dressing was placed.\nIn the NPWT group, the skin was closed with 4\u20136 single non-absorbable Monosyn 3\u20130 loose sutures every 1\u00a0cm.\nA NANOVA negative-pressure dressing was placed over the entire length of the incision.\nThanks to the use of NPWT, which stabilizes wound edges, we were able to place less skin sutures without the risk of wound dehiscence.\nNPWT also provides the opportunity to evacuate exudate from the subcutaneous tissue in sterile conditions and prevent the formation of a seroma or hematoma.\nIn the control group, the first dressing change was made 48\u00a0h after the operation, and thereafter dressings were changed daily until the removal of sutures on postoperative day 7.\nIn the NPWT group, the NANOVA dressing was taken out at 72\u00a0h.\nThree Steri-Strips were placed between the sutures, and a standard sterile dressing was placed.\nThe dressing was then changed every 24\u00a0h until the removal of sutures on postoperative day 7.\nPerioperative care\nAll patients received second-generation cephalosporin (20\u00a0mg/kg) 30\u00a0min before the incision.\nOn postoperative day 1, patients were fully mobilized and received a standard oral liquid diet with a volume restriction of 1 l.\nPatients with good diet tolerance were fed with a standard hospital diet from postoperative day 2.\nPerioperative care of patients in this study was compliant with the enhanced recovery after surgery (ERAS) protocol in colorectal surgery.\nHealing was evaluated during dressing changes in the ward and then during routine check-ups in the outpatient clinic on postoperative days 7 and 14.\nPatients who noticed any abnormalities related to wound healing contacted the outpatient clinic by telephone and were admitted for an additional visit.\nAfter 30\u00a0days, the patients were contacted by telephone to obtain information about possible abnormalities related to healing and were asked to send a photo of the healed wound by e-mail.\nForWHC, the frequency of monitoring visits was based on clinical status.\nThe last visit to the outpatient clinic was made approximately 2\u00a0weeks after CWH.\nEnd point criteria\nThe primary endpoint was the reduction of WHC after protective ileostomy closure.\nWHC were defined as any condition of the wound that required postoperative intervention other than a change of dressing or removal of stiches.\nSecondary endpoints were the incidence of SSI, postoperative length of hospital stay (LOS) and the duration of CWH.\nCWH was defined as complete closure of the wound without any secretion from the wound, as assessed at the outpatient clinic or reported by the patient.\nIncisional SSI diagnosis were made according to the criteria of the Center for Disease Control (CDC) and European Centre for Disease Prevention and Control (ECDC) for diagnosis of SSI.\nStatistical analysis\nContinuous data are presented as the medians and inter-quartile ranges, unless otherwise indicated.\nContinuous variables were compared using the Mann\u2013Whitney test and Student\u2019s t-test.\nCategorical variables were compared using the chi-square test, including Yates\u2019 correction or Fisher\u2019s exact test when necessary.\nThe level of significance was set at p\u2009<\u20090.05.\nLogistic regression models were used to detect possible risk factors for WHC and SSI incidence.\nIn the case of LOS and CWH, simple linear regressions were used to determine potentially relevant factors, and then multiple regression models were created.\nAnalyses were performed with Statistica 13.5.\nResults\nA total of 75 patients were randomized to the study.\nFour patients (5.3%) were lost to follow-up (two were lost as a result of reoperation, one was transferred to another ward, and one was excluded because of a technical problem with NPWT device\u2014difficulties with maintaining airtightness), and none of those patients developed WHC or SSI within 30\u00a0days.\nPatient flow through the study is presented in Fig.\n1.\nTable 1 shows patients\u2019 baseline characteristics before ileostomy closure.\nIn the two study groups, 35 patients were treated with postoperative NPWT (24 males [68.6%]; mean age 61.6\u2009\u00b1\u200911.3\u00a0years), and 36 patients (20 males[55.6%]; mean age 62.4\u2009\u00b1\u200911.3\u00a0years) were treated with suturing of the wound and traditional dressings (control group).\nNo significant differences between the two groups were observed in patient characteristics or preoperative treatments.\nSurgical outcomes are shown in Table 2, and WHC are shown in Table 3.\nWHC were observed in 3 (8.6%) patients in the NPWT group and 11 (30.6%) in the control group (p\u2009=\u20090.020).\nSSI was observed in two (5.7%) patients with NPWT and in eight (22.2%) patients in the control group (p\u2009=\u20090.046).\nThe median LOS was 3 (2\u20134) days in the NPWT group and did not significantly differ from that in the control group, 4 (2.5\u20135) days.\nThe median duration of wound healing in patients with NPWT and in control groups was 7 (7\u20137) and 7 (7\u201315.5) days, respectively (p\u2009=\u20090.030).\nTo identify potential risk factors for WHC, univariate logistic regression models were constructed, as presented in Table 4.\nThe univariate analyses revealed that only postoperative NPWT significantly decreased the odds ratio for WHC incidence.\nIn the case of risk factors for SSI, none of the factors analyzed in the univariate regression model were statistically significant.\nSimple linear regression models were built for the length of hospital stay and CWH time.\nIn the LOS regression model, the adjusted R2 for multiple regression was 60.68% with a p value\u2009<\u20090.001.\nIn this model, the factors identified as significantly prolonging LOS were: SSI by 4.58\u00a0days (2.62\u20136.54), other complications by 8.65\u00a0days (6.20\u201311.10), and BMI\u2009<\u200918.5\u00a0kg/m2 by 7.36\u00a0days (1.22\u201313.50).\nOf all the factors analyzed in the simple regression models, only postoperative use of NPWT was significant.\nNPWT use was associated with shortening CWH by 3.00\u2009\u00b1\u20091.24\u00a0days in simple regression model with an adjusted R2 of 6.42% (p value\u2009<\u20090.001).\nDiscussion\nTo the best of our knowledge, this is the first study confirming the usefulness of postoperative NPWT in reducing the number of WHC associated with elective ileostomy closure in patients after surgery for colorectal cancer.\nOur study showed that the use of postoperative NPWT after ileostomy reversal procedures significantly reduces the risk of complications.\nThis may have significant clinical implications, especially taking into considerations studies that suggest the benefits of early closure of the ileostomy, even before adjuvant chemotherapy.\nHowever, concern over complications after ileostomy closure often postpones this procedure until the end of adjuvant chemotherapy.\nPostoperative NPWT may be an adequate solution of that clinical problem.\nAlthough the risk of anastomotic leak after colorectal cancer surgery is reduced by diverting ileostomy, opponents argue it creates the need for another surgery with a relatively high risk of complications.\nAlthough those tend to be relatively mild and local complications, they may cause a delay in adjuvant chemotherapy, which diminishes the results obtained by oncological treatment.\nOne method to minimize the risk of complications after the ileostomy closure is protective NPWT placement.\nThere is a very large divergence in the reported incidences of infectious complications after ileostomy reversal.\nStudies that deal with other issues related to surgery report a very low incidence of these complications; however, research focusing on the impact of various surgical techniques on the incidence of SSI reports up to 40% risk of SSI in control groups.\nThis relatively high incidence of SSI has led the American College of Surgeons (ACS) to recommend closing wounds after ileostomy reversal using the purse-string technique.\nOur study was designed between 2015 and 2016, prior to the 2017 publication of the ACS recommendations to use the purse-string technique for closure of this type of wound.\nUntil then, simple suture technique was the standard, which is why in the control group simple suturing of the wound was applied.\nBecause the proposed technique significantly extends healing time and often, especially in obese people, gives an unsatisfactory cosmetic effect, many surgeons still close their ileostomy wounds with conventional primary closure.\nThe use of postoperative NPWT provides the opportunity to combine the benefits of both techniques.\nWe suppose, based on our experience, that the healing time and final cosmetic effect do not differ from those of the primary closure technique.\nAdditionally, NPWT offers a significant reduction in the occurrence of WMC and SSI compared with the effects of using the purse-string technique.\nSeveral clinical trials have investigated the usefulness of postoperative NPWT in reducing postoperative infections, but the majority of them have been conducted in fields other than gastrointestinal surgery.\nOnly a few studies have been conducted in the field of general, oncological, or digestive surgery, and only one article has described the use of NPWT as a postoperative dressing in patients after diverting ileostomy closure.\nThe results of this study differ from our observations.\nIn the cited study, the wound after ileostomy was closed using the purse-string suture technique; for this reason, even in the group without complications, the wound healing time was longer than 30\u00a0days.\nIn our study, even in the group of patients with an infectious complication, CWH was shorter than 30\u00a0days.\nIn our research, the NPWT application was limited to 72\u00a0h after the surgery and was used only to evacuate the exudate or hematoma from the wound.\nThe dressing was placed in sterile conditions of the operating room over the cleaned wound while the antibiotic prophylaxis used for the operation was still in effect.\nIn the cited study, a NPWT was installed 24\u00a0h after surgery and maintained for more than 2\u00a0weeks, which in our opinion may increase the potential for colonization of the wound and possible development of infectious complications.\nIn the previously cited meta-analysis of all other clinical trials of NPWT in the postoperative period, its use was shorter, ranging from 3 to 7\u00a0days, and the dressing was placed over the wound immediately after the surgery.\nAnother difference between our study that of Uchino et al. concerns the intervention population.\nOur research included patients who underwent surgery for colorectal cancer, mainly rectal cancer, which particularly exposes them to the risk of infectious complications.\nThe majority of them were elderly with numerous comorbidities and higher BMI.\nAdditionally, most of them underwent radio- and chemotherapy shortly before the surgery.\nIn the Uchino study, the population consisted of patients who had surgery as a result of ulcerative colitis, resulting in a significantly lower number of infectious complications in the control group than in our study.\nGiven these differences, we believe that these two studies are not directly comparable, as they concern different issues.\nTo increase the chance of achieving statistical significance while limiting the necessary study population, the cumulative index of WHC, not the occurrence of SSI, was established as the primary outcome.\nWith the incidence of SSI in our group of patients and expected reduction of SSI cases by approximately 50%, the minimal sample size would have required recruitment of 80 patients in each arm.\nHowever, the observed effect of intervention exceeded our expectations.\nA statistically significant effect of the use of NPWT on reducing the incidence of SSI was confirmed.\nOur study showed that the use of postoperative NPWT dressing is safe.\nIn the NPWT group, there was no increase in the percentage of postoperative complications, as well as no case of a complication that could be linked directly to NPWT use (postoperative bleeding or entero-cutaneous fistula formation).\nIn the group of patients with NPWT, both the time of hospitalization and the time of healing of the surgical wound were shortened.\nLimitations\nBecause of the actual incidence of WHC in this study differing from the values assumed during the sample size calculation, the post hoc analysis revealed that the study of the primary outcome, despite obtaining statistically significant differences, achieved 65% power and not the assumed 80%.\nThe obtained results are therefore underpowered.\nAlso, our study used only one type of NPWT device.\nComparison of the effectiveness and safety of other types of NPWT equipment in this application will require further research.\nWe did not specifically analyze the time period between chemotherapy and operation, but we observed no difference in the incidence of postoperative chemotherapy between groups, and the time from colorectal resection and ileostomy closure until chemotherapy did not differ between groups.\nWe did not perform an analysis of the cosmetic effect of our treatment.\nFurther research in this area is required.\nLastly, this study was a single-center study, which, on one hand, is a strength in the consistent treatment of all patients, but on the other hand, it will require confirmation in multi-center studies on larger groups of patients.\nConclusions\nProphylactic postoperative NPWT after diverting ileostomy closure in colorectal cancer patients reduces the incidence of WHC, SSI, and complete wound healing time.\nCONSORT flow diagram. NPWT negative-pressure wound therapy\n\nGroups characteristics\nParameter | Group 1 NPWT | Group 2 control | p value\nNumber of patients, n | 35 | 36 | n/a\nFemales, n (%) | 11 (31.4%) | 16 (44.4%) | 0.259\nMales, n (%) | 24 (68.6%) | 20 (55.6%)\nMean age, years\u2009\u00b1\u2009SD | 61.6\u2009\u00b1\u200911.3 | 62.4\u2009\u00b1\u200911.3 | 0.974\nBody mass index, kg/m2\u2009\u00b1\u2009SD | 26.2\u2009\u00b1\u20094.5 | 26.2\u2009\u00b1\u20094.3 | 0.794\nASA 1, n (%) | 1 (2.9%) | 1 (2.8%) | 0.943\nASA 2, n (%) | 22 (62.9%) | 24 (66.7%)\nASA 3, n (%) | 12 (34.2%) | 11 (30.5%)\nAny comorbidity, n (%) | 25 (71.4%) | 26 (72.2%) | 0.942\nCardiovascular disease, n (%) | 11(31.4%) | 7 (19.4%) | 0.252\nHypertension, n (%) | 18(51.4%) | 13 (36.1%) | 0.199\nDiabetes, n (%) | 5 (14.3%) | 5 (13.9%) | 0.962\nPulmonary disease, n (%) | 3 (8.6%) | 4 (11.1%) | 0.724\nRenal disease, n (%) | 2 (5.7%) | 2 (5.6%) | 0.977\nOther comorbidity, n (%) | 15 (42.8%) | 14 (38.9%) | 0.738\nSmoking, n (%) | 5 (14.3%) | 6 (16.7%) | 0.785\nImmunosuppressive treatment, n (%) | 3 (8.6%) | 1 (2.8%) | 0.297\nRadiotherapy n (%) | 27 (77.1%) | 29 (80.6%) | 0.729\nChemotherapy (pre- or postoperative), n (%) | 27 (77.1%) | 28 (77.8%) | 0.950\nAJCC Stage 0, n (%) | 7 (20%) | 11 (30.6%) | 0.319\nAJCC Stage I, n (%) | 7 (20.0%) | 9 (25.0%)\nAJCC Stage II, n (%) | 7 (20.0%) | 7 (19.4%)\nAJCC Stage III, n (%) | 13 (37.1%) | 6 (16.7%)\nAJCC Stage IV, n (%) | 1 (2.9%) | 3 (8.3%)\nMedian time between operations, days (IQR) | 148 (17\u2013687) | 139 (25\u2013517) | 0.296\nPrevious operation\n\u00a0Hemicolectomy, n (%) | 4 (11.4%) | 4 (11.1%) | 0.829\n\u00a0Colectomy, n (%) | 2 (5.7%) | 2 (5.5%)\n\u00a0Anterior resection of rectum n (%) | 8 (22.9%) | 13 (36.1%)\n\u00a0Intersphincter resection n (%) | 3 (8.6%) | 4 (11.1%)\n\u00a0TaTME, n (%) | 18 (51.4%) | 15 (41.7%)\n\nNPWT negative pressure wound therapy\nSD standard deviation\nASA American Society of Anaesthesiologists class\nAJCC American Joint Committee on Cancer tumor stage\nIQR inter\u2212quartile range\nTaTME transanal total mesorectum excision\n\nPeri- and postoperative outcomes in analysed groups\nParameter | Group 1 NPWT | Group 2 control | p value\nNumber of patients, n | 35 | 36 | n/a\nMedian operative time, minutes (IQR) | 57 (50\u201373) | 55 (52\u201370) | 0.856\nMean perioperative blood loss, ml\u2009\u00b1\u2009SD | 16\u2009\u00b1\u20099 | 18\u2009\u00b1\u200910 | 0.654\nPatients without any complications, n (%) | 30 (85.7%) | 23 (63.9%) | 0.035\nPatients with complications, n (%) | 5 (14.3%) | 13 (36.1%)\nClavien-Dindo grade 1, n (%) | 4 (11.4%) | 6 (16.7%) | n/a\nClavien-Dindo grade 2, n (%) | 1 (2.9%) | 5 (13.9%)\nClavien-Dindo grade 3, n (%) | \u2013 | 2 (5.6%)\nClavien-Dindo grade 4, n (%) | \u2013 | \u2013\nClavien-Dindo grade 5, n (%) | \u2013 | \u2013\nMedian length of postoperative hospital stay, days (IQR) | 3 (2\u20134) | 4 (2.5\u20135) | 0.072\nMedian duration of complete wound healing time days (IQR) | 7 (7\u20137) | 7 (7\u201315.5) | 0.030\n\nSignificant p-values were marked with bolded font\nNPWT negative pressure wound therapy\nSD standard deviation\nIQR inter\u2212quartile range\n\nPostoperative wound management complications (WMC)\nComplication | Group 1 NPWT | Group 2 Control | p value\nNumber of WMC, n (%) | 3 (8.57%) | 11(30.6%) | 0.020\nSurgical site infections (SSI), n (%) | 2 (5.71%) | 8 (22.2%) | 0.046\nSuperficial SSI, n (%) | 2 (5.71%) | 4 (11.1%) | n/a\nDeep SSI, n (%) | 0 | 3 (38.8%) | n/a\nOrgan SSI 1, n (%) | 0 | 1 (2.8%) | n/a\nHematoma, n (%) | 0 | 3 (8.3%) | n/a\nSeroma, n (%) | 1 (2.9%) | 1(2.8%) | 0.984\n\nSignificant p-values were marked with bolded font\n\nUnivariate logistic regression analyses of WMC incidence\nParameter | OR (95% CI) | p value\nPostoperative NPWT (yes vs. no) | 0.19 (0.04\u20130.97) | 0.041\nSex (female vs. male) | 0.74 (0.15\u20133.72) | 0.704\nAge (\u2265\u200975 vs.\u2009<\u200975\u00a0years) | 1.72 (0.16\u201318.91) | 0.650\nBMI (\u2265\u200930 vs.\u2009<\u200930\u00a0kg/m2) | 2.07 (0.33\u201313.18) | 0.429\nSmoking (yes vs. no) | 0.34 (0.03\u20134.68) | 0.719\nRadiotherapy (yes vs. no) | 1.02 (0.04\u201327.13) | 0.992\nChemotherapy (yes vs. no) | 0.92 (0.04\u201323.07) | 0.926\nParaostomal hernioplasty (yes vs. no) | 0.44 (0.03\u20135.98) | 0.532\nTime between previous operation (\u2265\u2009365 vs\u2009<\u2009365\u00a0days) | 3.08 (0.23\u201340.65) | 0.381\nCardiovascular disease (yes vs. no) | 0.67 (0.10\u20134.69) | 0.685\nDiabetes (yes vs. no) | 1.44 (0.19\u201311.13) | 0.719\nPulmonary disease (yes vs. no) | 0.79 (0.06\u201310.61) | 0.858\nRenal disease (yes vs. no) | 3.39 (0.13\u201390.93) | 0.457\nTaTME (yes vs. no) | 1.25 (0.25\u20136.34) | 0.780\nAJCC IV (yes vs. no) | 2.19 (0.09\u201351.96) | 0.619\nComplications not related to wound management | 0.83 (0.25\u20132.77) | 0.762\n\nSignificant p-value was marked with bolded font\nWMC wound management complication\nNPWT negative pressure wound therapy\nBMI body mass index\nTaTME transanal total mesorectum excision\nAJCC American Joint Committee on Cancer tumor stage",
        "label": "unclear",
        "id": "task4_RLD_test_729"
    },
    {
        "paper_doi": "10.1136/bmj.n256",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Design: RCTUnit: 20 clusters (slums)\n\n\nParticipants: Location/setting: study was carried out in Mumbai, India. Health workers who had passed 10th grade education were trained to conduct CBE within 4 weeks to perform on participants in the study.Sample size: 151,538 women aged 35 to 64Sex: only females were included.Mean age: mean age for enrolment for all trial participants in the screening and control arm was 44.84 and 44.92 years respectively.Inclusion criteria: women aged 35 to 64 years who did not have a history of breast cancer.Exclusion criteria: women outside of ages 35 to 64 years, and all women with a history of breast cancer.\n\n\nInterventions: Intervention (screening) group: (n = 75,360) of 10 clusters must undergo 9 rounds of biennial monitoring for breast cancer occurrence and mortality, 4 rounds of screening by CBE and cancer awareness education and 5 rounds of active surveillance.Control group: (n = 76,178) of 10 clusters must undergo 9 rounds of biennial monitoring for breast cancer occurrence and mortality, 1 round cancer awareness education, 8 rounds of active surveillance.Trial duration was 20 years.\n\n\nOutcomes: Primary outcomes: the down-staging of breast cancer at diagnosis and reduction in mortality from breast cancer.Secondary outcomes: CBE coverage, sensitivity, and specificity.Timing of outcome assessments: total trial duration was 20 years, but database was locked in March 2019 for analysis.\n\n\nNotes: Study start date: May 1998Study end date: March 2019Funding source: US National Institutes of Health (grant number: RO1CA074801)Conflicts of interest: author conflicts not declare\n\n",
        "objective": "To assess whether training in CBE affects the ability of health workers in LMICs to detect early breast cancer.",
        "full_paper": "Abstract\nObjective\nTo test the efficacy of screening by clinical breast examination in downstaging breast cancer at diagnosis and in reducing mortality from the disease, when compared with no screening.\nDesign\nProspective, cluster randomised controlled trial.\nSetting\n20 geographically distinct clusters located in Mumbai, India, randomly allocated to 10 screening and 10 control clusters; total trial duration was 20 years (recruitment began in May 1998; database locked in March 2019 for analysis).\nParticipants\n151\u2009538 women aged 35-64 with no history of breast cancer.\nInterventions\nWomen in the screening arm (n=75\u2009360) received four screening rounds of clinical breast examination (conducted by trained female primary health workers) and cancer awareness every two years, followed by five rounds of active surveillance every two years.\nWomen in the control arm (n=76\u2009178) received one round of cancer awareness followed by eight rounds of active surveillance every two years.\nMain outcome measures\nDownstaging of breast cancer at diagnosis and reduction in mortality from breast cancer.\nResults\nBreast cancer was detected at an earlier age in the screening group than in the control group (age 55.18 (standard deviation 9.10) v 56.50 (9.10); P=0.01), with a significant reduction in the proportion of women with stage III or IV disease (37% (n=220) v 47% (n=271), P=0.001).\nA non-significant 15% reduction in breast cancer mortality was observed in the screening arm versus control arm in the overall study population (age 35-64; 20.82 deaths per 100\u2009000 person years (95% confidence interval 18.25 to 23.97) v 24.62 (21.71 to 28.04); rate ratio 0.85 (95% confidence interval 0.71 to 1.01); P=0.07).\nHowever, a post hoc subset analysis showed nearly 30% relative reduction in breast cancer mortality in women aged 50 and older (24.62 (20.62 to 29.76) v 34.68 (27.54 to 44.37); 0.71 (0.54 to 0.94); P=0.02), but no significant reduction in women younger than 50 (19.53 (17.24 to 22.29) v 21.03 (18.97 to 23.44); 0.93 (0.79 to 1.09); P=0.37).\nA 5% reduction in all cause mortality was seen in the screening arm versus the control arm, but it was not statistically significant (rate ratio 0.95 (95% confidence interval 0.81 to 1.10); P=0.49).\nConclusions\nThese results indicate that clinical breast examination conducted every two years by primary health workers significantly downstaged breast cancer at diagnosis and led to a non-significant 15% reduction in breast cancer mortality overall (but a significant reduction of nearly 30%in mortality in women aged \u226550).\nNo significant reduction in mortality was seen in women younger than 50 years.\nClinical breast examination should be considered for breast cancer screening in low and middle income countries.\nTrial registration\nClinical Trials Registry of India CTRI/2010/091/001205; ClinicalTrials.gov NCT00632047.\nIntroduction\nThe incidence of breast cancer is rising in all countries of the world, but particularly so in low and middle income countries.\nFor example, in Mumbai, India, the incidence of breast cancer has risen by nearly 40% between 1992 and 2016, and breast cancer is the leading cause of death from cancer in women in most states of India.\nBreast cancers in low and middle income countries are frequently detected in advanced stages, and consequently, more than half the global deaths from breast cancer occur in these countries.\nWhile mammography is the established screening tool in developed countries, the screening modality that is appropriate for India and other low and middle income countries remains undetermined.\nBreast self-examination might not be useful as a general strategy,  largely because it is not feasible to ensure women perform it well.\nHowever, a case-control study based on data from the Canadian National Breast Screening Study showed that in a controlled setting, where the quality of breast self-examination was carefully evaluated, women who conducted the procedure benefitted well.\nMammography, which is widely practiced in Western countries, might not be an appropriate approach in low and middle income countries because of its cost and complexity.\nFurthermore, most women in low and middle income countries are younger than 50, and mammography is less effective in this age group.\nClinical breast examination (CBE) is an alternative screening method, and was one of the components of screening in two important randomised trials.\nThe Health Insurance Plan Study was conducted in greater New York, USA, in the 1960s during which 62\u2009000 women aged 40-64 were randomised to receive yearly CBE plus mammography or no screening.\nDuring the 1960s, mammography was in its early stages of development, and a disproportionately large number of breast cancers were detected by CBE.\nAn estimated two thirds of the reduction in breast cancer mortality in the Health Insurance Plan study could be attributed to CBE.\nTo determine the relative contributions of mammography and CBE in the reduction of breast cancer mortality, the Canadian National Breast Screening Study was initiated in the early 1980s.\nIn one of two parts of the study, women aged 50-59 were randomly allocated to receive either yearly CBE plus mammography or yearly CBE alone.\nThe trial had the potential to determine whether mammography provided any added benefit in terms of mortality reduction in addition to that provided by CBE.\nAfter 13 years of follow-up and five rounds of screening, deaths from breast cancer in the two arms were almost identical.\nThese results remained unchanged after 25 years of follow-up.\nThe findings of the Health Insurance Plan Study and Canadian National Breast Screening Study provided a strong argument for a randomised trial to compare CBE with no screening,  and formed the basis for the Mumbai study.\nThis study aimed to determine whether CBE plus provision of cancer awareness would downstage breast cancer at diagnosis and reduce mortality from the disease, compared with no screening.\nMethods\nThe Mumbai study had two components: screening for cervix cancer by visual inspection and screening for breast cancer by CBE.\nThe results of the cervical cancer component have been reported, as well as details of methodology to include design, mechanisms of community outreach, recruitment and informed consent, training of primary health workers and medical social workers, sample size estimation, adherence to screening (after three rounds), and mechanism of referral and treatment.\nThe above methodological aspects are summarised in this paper.\nDefinition of a cluster\nA cluster comprised of many closely situated dwellings in congested slum areas, defined by geographical boundaries such as railway lines, water pipelines, highways, roads, public parks, and canals.\nEach cluster had 9000 to 10\u2009000 dwellings with a population of 50\u2009000-65\u2009000, of which about 7500 women were aged 35-64.\nTransfer between control and intervention clusters was unlikely because the clusters were geographically separate, and because virtually none of the participants underwent breast screening outside the trial.\nThe standard of care in our study population was no screening.\nRandomisation method\nRandomisation was by cluster, where groups rather than individuals were chosen as units of randomisation.\nTwenty independent clusters were numbered 1-20 and randomly allocated to screening or control groups by a draw of lots.\nWith this procedure, 10 clusters were assigned as screening clusters and 10 as control clusters.\nTrial participants and intervention\nThe current study, a cluster randomised controlled trial, recruited 151\u2009538 women aged 35-64 from 20 clusters in Mumbai.\nWomen in the screening arm (n=75\u2009360) received four rounds of CBE conducted by trained female primary health workers and cancer awareness information every two years, followed by five rounds of active surveillance by way of home visits every two years.\nWomen in the control arm (n=76\u2009178) received one round of cancer awareness followed by eight rounds of active surveillance every two years.\nParticipants in both arms were eligible for free diagnostic evaluation and treatment at the Tata Memorial Hospital; women in both groups were provided with identical identity cards to obtain free treatment at the hospital.\nRecruitment started in May 1998 and was completed in April 2002.\nFour rounds of CBE were concluded in December 2007 and follow-up continued until May 2018.\nThe database was locked in March 2019 for analysis.\nSample size considerations\nWe based sample size calculations primarily on expected incidence and mortality data from breast and cervical cancer over the long duration of the study.\nIntracluster correlation was estimated using age, education status, and religion of women in the study.\nThe computation was done using MLWin Software.\nFor estimation of sample size, we considered two primary outcomes\u2014breast and cervical cancer mortality.\nSample size derived was 150\u2009000 women, which was calculated to detect 25% reduction in mortality from breast cancer with 80% power and 5% type I error, after adjusting for intracluster correlation and design effect (0.00013758 and 2.0408, respectively).\nWith these considerations, 230 deaths from breast cancer in the control group were required for mortality analysis to be recommended.\nThe smaller design effect observed in the study indicated that the sample size was adequate to estimate reduction in mortality with anticipated power.\nThree way data linkage\nTo capture information on death from any cause, the trial had a three way data linkage system.\nPrimary collection of data was done by trained medical social workers by home visits.\nData were matched with those of Mumbai Municipal Death Records and with the Mumbai Cancer Registry.\nMore information about the linkage systems and process has been provided in the supplementary material.\nBreast cancer deaths\nBreast cancer as the cause of death among women who were diagnosed with breast cancer was blindly ascertained by two independent experts.\nIf there was a discrepancy between the two experts, the records were blindly reviewed by a third independent reviewer.\nCause of death was assigned to breast cancer when at least two of the three reviewers concurred.\nCause of death could not be ascertained in 40 women.\nStatistical analysis\nWe calculated incidence rates in both arms by taking into account the number of person years determined from the date of entry into the trial to the date of diagnosis.\nThe number of person years for calculating mortality rates was determined from the date of entry in the trial to the date of death.\nData were censored during analysis for women who had migrated or were lost to follow-up, or who had died from other causes.\nAll deaths in both arms were included for all cause mortality estimates.\nWe used a Poisson regression model to estimate incidence and mortality rate ratios and their 95% confidence intervals.\nAdjustments were made for design effect.\nAll statistical tests were two sided, and P<0.05 was considered to be statistically significant.\nThe data were analysed on the basis of intention to screen (all women, irrespective of compliance), and when the predefined number of events (230 deaths) were documented in the control arm.\nAll analyses were carried out in Stata software version 12 (Stata, College Station, TX).\nThe study underwent several protocol amendments during its long course, particularly in the initial years.\nThe amendments were suggested by consultants or the data safety monitoring committee from time to time and were duly approved by the institutional review board.\nThese amendments were also approved by the funding agency (US National Cancer Institute).\nAll interpretations in the manuscript are aligned with the finally amended protocol.\nPatient and public involvement\nPatients and public were not involved in setting the research question, outcome measures, design, interpretation, or writing of the results.\nHowever, involvement of local community leaders was sought during recruitment of study participants and study implementation.\nResults\nThe CONSORT flow diagram depicting the overall trial schema is presented in figure 1.\nDemographic and breast cancer risk factors were well balanced between the two arms indicating that randomisation was without bias (supplementary table 1).\nCompliance, quality assurance, and breast cancer detection\nThe mean adherence to screening after four rounds was 67.07%, and mean adherence to hospital referral for confirmation of diagnosis was 76.21% (supplementary table 2); overall, 94.82% (n=71\u2009456) of the participants were screened at least once.\nThe average screen positivity rate was 1.28% in the four screening rounds (supplementary table 2).\nAfter four rounds of screening, 199 women with breast cancer were identified (supplementary table 3).\nBreast cancers included 114 screen detected cancers, 77 interval cancers, and eight cancers among women who did not adhere to screening in the preceding round (supplementary table 3).\nAs a quality assurance measure, a random sample of 5% of women (n=10\u2009021) was also examined by a qualified medical officer.\nThe \u0138 value for concordance was found to be 0.76, (95% confidence interval 0.72 to 0.81), indicating that the quality of CBE conducted by primary health workers met quality assurance requirements.\nAverage adherence to rounds 5-9 of active surveillance after CBE screening was 77.57%, which was similar to the average adherence to rounds 5-9 received by the control arm (77.57% v 76.22%, P=0.99; supplementary tables 4 and 5).\nOf 641 cancers detected in the screening arm overall, 199 (31%) were detected during screening rounds 1-4 and 442 (69%) were detected during the active surveillance rounds 5-9 after CBE screening (supplementary tables 2 and 4).\nAdherence to treatment and to evidence based guidelines was similar in both arms (supplementary table 6); mean adherence of these women to treatment was 98.88%.\nAdherence in the control arm to the first and the only round of cancer awareness was 90.88% (n=69\u2009231).\nAverage adherence to the subsequent eight rounds of active surveillance was 78.14% (supplementary table 5).\nAfter nine rounds of active surveillance, 655 breast cancer cases were recorded in the control arm (supplementary table 5).\nProgressively more breast cancers were detected in each round as the women aged.\nMean adherence of these women to treatment was 97.63%.\nAge at enrolment and age at diagnosis of breast cancer\nMean age at diagnosis of breast cancer in women in the screening arm was 55.18 (standard deviation 9.10 (95% confidence interval 54.47 to 55.88)).\nMean age at diagnosis in the control arm was 56.50 (9.10 (55.80 to 57.20)).\nThis difference indicated that screening had brought forward breast cancer diagnosis by 16 months (P=0.01; table 1).\nAt the time of recruitment, over 70% women in both the screening and control arms were younger than 50, whereas at the time of breast cancer diagnosis, this proportion was reversed with nearly 75% of women aged 50 and older in both arms (table 1).\nThese data implied that breast cancer was diagnosed predominantly in older women, or in younger women after they reached age 50.\nThis finding formed the basis for us to analyse the subsequent data relating to breast cancer downstaging and mortality by using age 50 as the cutoff threshold, although this threshold was not prespecified in the protocol and should be considered a post hoc analysis.\nDownstaging of breast cancer\nBiennial CBE led to significant downstaging of breast cancer in all women (P=0.001; table 2), as well as in women younger than 50 (P=0.005) and in those aged 50 and older (P=0.05).\nStaging information was unavailable in 41 women in the screening arm and 73 women in the control arm.\nHowever, we saw no difference when comparing the survival of these women with missing information (supplementary figure 1).\nBreast cancer incidence and absence of overdiagnosis\nAt the end of screening, we found 198 women with breast cancer in the screening arm and 151 in the control arm, which translated into a crude incidence rate of 60.57 and 45.30 per 100\u2009000 women years, respectively (rate ratio 1.34 (95% confidence interval 1.05 to 1.71); P=0.02; table 3).\nWe saw an excess of 47 diagnoses of breast cancer in the screening arm compared with the control arm (table 3).\nAfter a median follow-up of 18 years, the screening and control arms had 640 and 655 cases of breast cancer, respectively, which translated into a crude incidence rate of 62.76 and 64.43 per 100\u2009000 women years, respectively (0.97 (0.87 to 1.09), P=0.66; table 3).\nSupplementary table 7 shows that although, as expected, a higher incidence of breast cancer was seen in the screening group than in the control group up to study year 10 (that is, until the end of screening round 4), this difference reduced gradually from study year 12 onwards (starting surveillance round 1) and disappeared completely by study year 18 (surveillance round 5).\nBreast cancer mortality\nWe recorded 213 breast cancer deaths in the screening arm and 251 deaths in the control arm (rate ratio 0.85 (95% confidence interval 0.71 to 1.01), P=0.07; table 3).\nThus, overall, a 15% non-significant reduction in mortality was seen when women of all ages were considered.\nAmong women younger than 50, 149 breast cancer deaths were recorded in the screening arm and 158 deaths in the control arm (0.93 (0.79 to 1.09), P=0.37).\nAmong women aged 50 and older, 64 breast cancer deaths were recorded in the screening arm and 93 deaths in the control arm (0.71 (0.54 to 0.94), P=0.02; table 3).\nThis subset analysis based on the age 50 threshold was not stipulated in the protocol and was a post hoc analysis.\nThe cumulative breast cancer mortality in the screening and control arms over 20 years is shown in figure 2.\nAn excess of breast cancer deaths in the screened population was seen in both age subgroups (age <50 and \u226550) in the early years after randomisation (fig 2), which lasted for about 14 years in women younger than 50 and about six years in those aged 50 and older.\nWhen breast cancer mortality data were analysed on the basis of attendance to the number of CBE screening rounds, we found that even women younger than 50 who attended all four rounds of screening benefitted significantly in terms of mortality reduction (rate ratio 0.66 (95% confidence interval 0.53 to 0.83), P<0.001).\nBut this benefit did not exist if these women attended only three rounds (0.88 (0.60 to 1.27), P=0.48).\nWomen aged 50 and older, however, benefitted from attending both three as well as four rounds of screening (attendance to all four rounds (0.64 (0.45 to 0.93), P=0.02); attendance to three rounds (0.66 (0.44 to 1.00), P=0.05); supplementary table 8).\nAll cause mortality\nWhen we considered all cause mortality during the 20 year period, we saw a non-significant reduction of 5% in the screening arm.\nAll cause mortality rates were 1100.59 and 1162.25 per 100\u2009000 women years in the screened and controls arms, respectively (rate ratio 0.95 (95% confidence interval 0.81 to 1.10); P=0.49).\nThe subdivision of all cause mortality by age (<50 and \u226550) is also represented (table 3).\nBreast cancer deaths comprise less than 3% of deaths from all causes in women in India; and hence a reduction in all cause mortality was not expected.\nThe cumulative all cause mortality in the screening and control arms over 20 years is shown in supplementary figure 2.\nDiscussion\nStatement of principal findings\nWe report here results of our randomised trial that compared CBE screening with no screening.\nWe showed that biennial CBE performed by trained female primary health workers significantly advanced breast cancer diagnosis by 16 months, and also downstaged the disease with fewer stage III or IV cancers in screened women.\nOverall, CBE led to a non-significant 15% reduction in breast cancer mortality; however, a significant reduction of nearly 30%was observed in women aged 50 and older.\nIn women younger than 50, despite successful downstaging, no mortality reduction was observed.\nLack of mortality reduction in younger women is consistent with data reported in some mammography trials, and could point to undetermined biological factors.\nParticipant attendance to the number of screening rounds also appeared to be important in breast cancer mortality reduction for women younger than 50.\nWe found a 34% mortality reduction in this age group if the women attended all four rounds of screening (P<0.001).\nThis benefit, however, disappeared if they attended only three rounds (mortality reduction 13%, P=0.48).\nFor women aged 50 and older, however, we observed mortality reduction after attendance to three and four rounds of screening (34%, P=0.05 and 36%, P=0.02, respectively; supplementary table 8).\nStrengths and weaknesses in relation to other studies\nTwo other randomised trials have compared CBE screening with no screening.\nA cluster randomised controlled trial was initiated in Kerala, India, in 2006 where three rounds of CBE every three years was planned to evaluate whether CBE can reduce incidence of advanced breast cancers and mortality from the disease.\nEarly results have shown a higher proportion of early stage breast cancers in the intervention arm than in the control arm.\nAnother trial comparing CBE screening with no screening in the Philippines could not be satisfactorily concluded because of unacceptably low levels of adherence, possibly because of external investigators not fully anticipating cultural and psychosocial barriers.\nIn our study, an excess mortality from breast cancer was seen in the screening arm during the first few years of screening for the total study population as well as when stratified by age groups.\nSuch an excess mortality was also seen in the cervical cancer component of this trial.\nA meta-analysis of seven breast cancer screening trials suggested an excess breast cancer mortality up to the fifth year of screening in women younger than 50 and in the first year in older women.\nThis excess was, however, not apparent in a combined analysis of Swedish trials.\nThe possible finding of early excess cancer mortality needs exploring.\nThe theory of biological predeterminism (pre-existing micrometastases before diagnosis and surgery) fails to explain this excess mortality but could point towards an impact of events at the time of diagnosis and surgery on mortality.\nStrengths and weaknesses of this study\nOne crucial element of our study that led to its successful completion was that it was entirely indigenous.\nThe trial was conceived, designed and implemented by a team based in Mumbai and had full understanding of the psychosocial, geopolitical, and geographical ground realities that influence the conduct of complex, public health randomised trials in low and middle income countries.\nOur study was conducted in slum areas largely inhabited by socioeconomically disadvantaged women who often moved residence requiring our medical social workers to trace their new abodes, sometimes in far flung parts of the city.\nOwing to our medical social workers making innumerable home visits to a population that was often mobile, we were able to achieve a satisfactory compliance at all levels of screening.\nThe quality of CBE performed by our primary health workers was also of high standard, which was confirmed by comparing the screening findings with a specialist breast clinician.\nWe were also able to capture death records of a high proportion of cases because of the three way data linkage system.\nFinally, our study included near perfect randomisation for a cluster randomised controlled trial; all demographic and breast risk factors were equally distributed in the screening and control arms.\nProvision of timely treatment could have helped to improve quality of life in screened women by preventing advanced stage disease, including local recurrence.\nOur study also had some limitations.\nCancer staging data were unavailable from 41 women in the screening arm and 73 women in the control arm.\nThis limitation probably did not affect the study results because the survival curves of patients with missing staging information were similar in the screening and control arms (supplementary figure 1).\nHowever, a sensitivity analysis of patients with missing staging information, in which all 41 women from the screening arm were assigned cancer stages III or IV and all 73 women from the control arm were assigned to cancer stages I or II, led to loss of statistical significance in the downstaging effect of screening.\nAnother study limitation was that cause of death information was not available through death certificates and the available documents for some women.\nTo overcome this limitation, three independent experts reviewed the records of all women with breast cancer who had died.\nBreast cancer was assigned as a cause of death only when at least two reviewers concurred (213 (83%) of 258 in the screening arm and 251 (90%) of 278 in the control arm).\nOur blinded review process for assigning cause of death was based on similar mechanisms used in other screening trials.\nHowever, the possibility of some residual uncertainty cannot be excluded; some degree of variability is inevitable in screening trials when death certificates are often modestly accurate and medical records often incomplete.\nWe did not observe a significant reduction in all cause mortality.\nBut because breast cancer deaths comprise less than 3% of all deaths in women in India, we did not expect a reduction in all cause mortality in our study.\nMeaning of the study\u2014possible explanations and implications for clinicians and policymakers\nOur study validates CBE as an alternative modality of breast screening.\nIt demonstrates that CBE screening is effective in reducing breast cancer mortality in Indian women aged 50 and older without any overdiagnosis.\nIn our trial, we were able to use a vertical programme with dedicated staff that was centrally controlled.\nFurthermore, women in India and in many other low and middle income countries are relatively lean and have smaller breasts than women in Western countries.\nThe health workers who screened women with CBE in this trial had passed 10th grade education and could be trained to perform CBE within a minimal training period (about four weeks).\nWe believe that CBE screening by primary health workers is replicable in the general population, and CBE has already been implemented in other parts of India as pilot schemes.\nOur study suggests that implementation of population screening by CBE in low and middle income countries is feasible, provided that adequate training of screening providers, careful monitoring, and quality of performance are assured.\nWhether the use of CBE in low and middle income countries at the community level can lead to a reduction in breast cancer mortality is still unknown.\nIts success can only be ascertained several years after CBE has been implemented as public health programme.\nWhat is already known on this topic\nBreast cancer screening by mammography reduces mortality in women aged 50 and older, but its effectiveness in women younger than 50 is questionable\nBreast self-examination has not been proven to be an effective method for early detection of breast cancer\nWhether screening by clinical breast examination can reduce mortality from breast cancer is not known\nWhat this study adds\nIn a 20 year study, clinical breast examination conducted by trained female health workers in Mumbai led to a downstaging of breast cancer at diagnosis and reduced mortality from the disease by nearly 30% in women aged 50 and older, but with no mortality reduction seen in women younger than 50\nA 5% reduction in all cause mortality was seen in the screening arm compared with the control arm, but was not statistically significant\nClinical breast examination should be considered for breast cancer screening in low and middle income countries\nExtra material supplied by authors\nTrial flow diagram\nCumulative breast cancer mortality during 20 years of study\n\nAge at enrolment of all women and age at diagnosis of breast cancer\nArm | Age at enrolment (for all trial participants) |  | Age at diagnosis (for participants with breast cancer only)\nTotal No | No of women aged <50 (%) | No of women aged \u226550(%) | P value | Mean age (SD (95% CI)) | Difference (95% CI) | Total No | No of women aged <50 (%) | No of women aged \u226550(%) | P value | Mean age (SD (95% CI)) | Difference (95% CI)\nScreening | 75\u2009177* | 54\u2009212 (72.11) | 20965 (27.89) | 0.06 | 44.84 (7.90 (44.78 to 44.90)) | 0.078 (\u22120.002 to 0.158) |  | 640\u2020 | 161 (25.16) | 479 (74.84) | 0.01 | 55.18 (9.10 (54.47 to 55.88)) | 1.321(0.330 to 2.312)\nControl | 76\u2009097* | 54\u2009188 (71.21) | 21909 (28.79) | 44.92 (8.00 (44.86 to 44.97)) |  | 655 | 147 (22.44) | 508 (77.56) | 56.50 (9.10 (55.80 to 57.20))\n\nSD=standard deviation.\nInformation on age was not available for 183 women in the screening arm and 81women in the control arm among the total women enrolled.\nOf the 641 women with breast cancer in the screening arm, one had bilateral breast cancer, who was considered only once.\n\nStaging of breast cancer at diagnosis \nAge group | Randomised group | Stages I or II (No (%)) | Stages III or IV (No (%)) | Total No | Pearson x2 | Difference (%) in stages III+IV between screening and control arms (95% CI)\nAll women* | Screening arm | 379 (63) | 220 (37) | 599 | 11.757 (P=0.001) | 9.83 (4.208 to 15.368)\nControl arm | 311 (53) | 271 (47) | 582\n<50\u2020 | Screening arm | 271 (63) | 161 (37) | 432 | 8.034 (P=0.005) | 9.77 (3.008 to 16.423)\nControl arm | 206 (53) | 183 (47) | 389\n\u226550\u2021 | Screening arm | 108 (65) | 59 (35) | 167 | 3.906 (P=0.05) | 10.27 (0.094 to 20.092)\nControl arm | 105 (54) | 88 (46) | 193\n\nStaging information unavailable from 41 women in the screening arm and 73 women in the control arm.\nStaging information unavailable from six women in the screening arm and 12 women in the control arm.\nStaging information unavailable from 35 women in the screening arm and 61 women in the control arm.\n\nBreast cancer incidence, breast cancer mortality, and all cause mortality after 20 years since commencement of study\n | Screening arm |  | Control arm | Rate ratio(95% CI)\u2020 | P value\nTotal No of women | No of diagnoses or deaths | No of person years | Crude rate per 100\u2009000 person years (95% CI) | Total No of women | No of diagnoses or deaths | No of person years | Crude rate per 100\u2009000 person year (95% CI)\nBreast cancer incidence\nCompletion of active screening | 75\u2009360 | 198 | 326\u2009891.2 | 60.57 (49.87 to 74.62) |  | 76\u2009178 | 151 | 333\u2009346.7 | 45.30 (38.51 to 53.64) | 1.34 (1.05 to 1.71) | 0.02\nCompletion of 20 years of study | 75\u2009360 | 640 | 1\u2009019\u2009761 | 62.76 (57.02 to 69.35) |  | 76\u2009178 | 655 | 1\u2009016\u2009616 | 64.43 (60.43 to 68.90) | 0.97 (0.87 to 1.09) | 0.66\nBreast cancer mortality\nAll ages* | 75\u2009360 | 213 | 1\u2009023\u2009097 | 20.82 (18.25 to 23.97) |  | 76\u2009178 | 251 | 1\u2009019\u2009500 | 24.62 (21.71 to 28.04) | 0.85 (0.71 to 1.01) | 0.07\nAge <50 | 54\u2009212 | 149 | 763\u2009141.8 | 19.53 (17.24 to 22.29) |  | 54\u2009188 | 158 | 751\u2009367.0 | 21.03 (18.97 to 23.44) | 0.93 (0.79 to 1.09) | 0.37\nAge \u226550 | 20\u2009965 | 64 | 259\u2009955.2 | 24.62 (20.62 to 29.76) |  | 21\u2009909 | 93 | 268\u2009133.1 | 34.68 (27.54 to 44.37) | 0.71 (0.54 to 0.94) | 0.02\nAll cause mortality\nAll ages* | 75\u2009360 | 11\u2009261 | 1\u2009023\u2009180 | 1100.59 (989.98 to 1224.58) |  | 76\u2009178 | 11\u2009853 | 101\u20099831 | 1162.25 (1037.16 to 1303.45) | 0.95 (0.81 to 1.10) | 0.49\nAge <50 | 54\u2009212 | 4450 | 763\u2009177.7 | 583.09 (539.66 to 629.69) |  | 54\u2009188 | 4708 | 751\u2009508.2 | 626.47 (572.73 to 684.32) | 0.931 (0.829 to 1.045) | 0.23\nAge \u226550 | 20\u2009965 | 6811 | 260\u2009001.8 | 2619.6 (2456.3 to 2796.9) |  | 21\u2009909 | 7145 | 268\u2009323.2 | 2662.8 (2498.2 to 2835.8) | 0.984 (0.902 to 1.073) | 0.71\n\nInformation on age not available for 183 women in the screening arm and 81 women in the control arm among study participants of all ages.\nRate ratio calculated by Poisson regression model after adjusting for cluster design. ",
        "label": "unclear",
        "id": "task4_RLD_test_904"
    },
    {
        "paper_doi": "10.1371/journal.pone.0077887",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 36 clusters, 569 households, 845 children < 5Inclusion criteria: households were eligible if they had at least one child < 5\n\n\nInterventions: Chlorine disinfection (WaterGuard) (427 children < 5)Primary drinking supply (422 children < 5)\n\n\nOutcomes: Diarrhoeal episodes for children < 5Intervention compliance\n\n\nNotes: Location: rural communities, Kersa district, EthiopiaLength: 16 weeksPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "Background\nHousehold water treatment has been advocated as a means of decreasing the burden of diarrheal diseases among young children in areas where piped and treated water is not available.\nHowever, its effect size, the target population that benefit from the intervention, and its acceptability especially in rural population is yet to be determined.\nThe objective of the study was to assess the effectiveness of household water chlorination in reducing incidence of diarrhea among children under-five years of age.\nMethod\nA cluster randomized community trial was conducted in 36 rural neighborhoods of Eastern Ethiopia.\nHouseholds with at least one child under-five years of age were included in the study.\nThe study compared diarrhea incidence among children who received sodium hypochlorite (liquid bleach) for household water treatment and children who did not receive the water treatment.\nGeneralized Estimation Equation model was used to compute adjusted incidence rate ratio and the corresponding 95% confidence interval.\nResult\nIn this study, the incidence of diarrhea was 4.5 episodes/100 person week observations in the intervention arm compared to 10.4 episodes/100 person week observations in the control arm.\nA statistically significant reduction in incidence of diarrhea was observed in the intervention group compared to the control (Adjusted IRR\u200a=\u200a0.42, 95% CI 0.36\u20130.48).\nConclusion\nExpanding access to household water chlorination can help to substantially reduce child morbidity and achieve millennium development goal until reliable access to safe water is achieved.\nTrial Registration\nClinicalTrials.gov NCT01376440\nIntroduction\nDiarrheal disease kills 1.5 million people mostly children under the age of five years in developing countries each year.\nMany of infectious agents causing diarrhea are potentially water borne transmitted through contaminated water.\nEven though Millennium Development Goal (MDG) of drinking water is achieved, 780 million people lack access to improved water sources and 2.5 billion lack improved sanitation worldwide, rural population are disproportionately undeserved.\nEven water from improved source is not always safe.\nFurthermore, water collected from initially acceptable microbial quality, it often becomes contaminated with pathogens during transport and storage.\nTo overcome the difficulties in providing safe water, point-of-use water treatment has been advocated as a means to improve access to potable water and decrease the global burden of diarrheal diseases.\nHowever, the effect of household water disinfection with chlorine on diarrhea episode reduction is variable, ranging from no protective effect to 85 percent reduction.\nThe studies on effectiveness of water quality interventions in reducing diarrhea have been flawed due to responder observer biases.\nUptake and use is low among rural population who are more at risk of water borne disease.\nIt is difficult to identify the population that benefit most from the potential effect of the intervention.\nThus, the main aim of this intervention study was to determine the effectiveness of household water chlorination (point-of-use water treatment) in reducing diarrhea incidence among children under-five years of age in rural community of Eastern Ethiopia.\nMethods\nEthics Statement\nThe study was reviewed and approved by the Haramaya University, College of Health Science Ethical Review Committee.\nConsent was obtained from district administration, district health department, community leaders.\nWritten consent was also obtained from the primary caregivers of children.\nField workers provided Oral Rehydration Solution (ORS) obtained from Kersa district health department for the children with diarrhea and advised their caregiver to take them to the nearby health facility for further treatment.\nControl communities received the intervention after the completion of the study.\nThe protocol for this trial (Protocol S1) and supporting CONSORT checklist (Checklist S1) are available as supporting information.\nStudy Setting\nThe study was conducted in rural Kersa Demographic and Health Research Centre (KD-HRC) Field Site, Kersa district, Eastern Ethiopia.\nIt is located in kersa district which is about 482 kms far from the capital, Adis Ababa.\nAccording to the 2007 baseline survey, the study area has a population of 47,036 (8960 households) distributed in 10 kebeles (smallest administrative unit with a population of 5,000) of which 7870 are children under-five years of age.\nAll the households had no running water.\nFamilies collect water from springs, streams/rivers, or wells and store in 20 liter jerry-can.\nStudy Design and Procedure\nWe conducted randomized controlled, parallel group, field trial to assess the effectiveness of household chlorination in reducing diarrhea episode.\nThe study site was selected by Haramaya University in 2007 to serve as demographic surveillance and health research center.\nIt is one of the six Demographic Surveillance Sites (DSS) in the country.\nEthiopian statistics authority has demarcated population enumeration villages (clusters) for 2007 Ethiopian population census across the country.\nClusters are distinct neighborhoods with defined geographical boundaries.\nRural KDS-HRC has 64 clusters and all were eligible for the study.\nEthiopian central statistics authority statistician randomly selected 36 clusters from district population enumeration areas using computer generated random sampling.\nField workers conducted census identified households in the selected clusters that had at least one child under-five years of age.\nTwenty four children (median of 16 households) were selected in each cluster by using simple random sampling from the list of households that contain the number of children in the presence of community leaders and some residents for follow up.\nThe randomization of clusters was done in a meeting with community leaders and representative from the health department.\nEach cluster code was written on a separate paper in front of the community leaders and put in a box.\nThey agreed the first 18 draws to be assigned in the intervention arm and the remaining in the control.\nOne of the community leaders draw 18 clusters consecutively from the box and assigned in the intervention arm.\nThe remaining 18 clusters assigned in the control group.\nField workers approached residents of selected households and completed baseline survey.\nSodium hypochlorite (intervention) was distributed for all the households in the intervention arm.\nWe used cluster level randomization to avoid ethical concerns and minimize the potential transfer of the intervention between the two groups (figure 1).\nSample Size\nThe sample size was calculated using methods published by Hayes and Bennett, assuming 11% incidence of diarrhea among children in the control group based on previous study, 80% power, 10% drop out, 95% confidence interval and design effect of three from clustering.\nAccordingly, we aimed to enroll 18 clusters per arm with 24 children under the age of five years per cluster followed for 16 weeks (which would provide 6921 person weeks of observation in each group) to get sufficient power to detect 40% reduction in the incidence of diarrhea in the intervention group among children under-five years of age.\nIntervention\nThe intervention for this study was 1.25% sodium hypochlorite branded locally known as \u201cWaterGaurd\u201d and obtained from population service international (PSI) manufactured specifically for home water disinfection.\nLocal women were selected to distribute sodium hypochlorite for the intervention households with no charge to treat their stored water at home for 16-weeks from June to October 2011.\nThey also explained how to treat water with chlorine and demonstrated how to use the disinfectant.\nControl groups continued their usual practice with respect to drinking water.\nBoth intervention and control groups collect and store water with 20 liter jerry cans.\nIn many African countries 20 liter jerry cans are used to transport and store water at home and are good options for safe storage.\nData Collection\nA baseline survey was conducted on the demographic and socioeconomic condition, sources of water, access to and quality of water, water handling practice, sanitation, hygiene and pre-intervention diarrhea rates.\nThe questionnaire was translated from English to the local language, Oromifa, back translated in to English and administered to the mother/caregiver by Local language.\nThe primary outcome was the occurrence of diarrhea among children under-five years of age.\nDiarrhea is defined as three or more loose or watery stools in 24 hours or more frequently than normal for an individual.\nWe defined a new episode of diarrhea if it occurred after a period of three diarrhea free days.\nWe calculated the incidence of diarrhea as the number of new episodes divided by the total number of person\u2013weeks observation.\nThe survey instrument was pre-tested in the nearby villages and amended based on the comments from the pretest.\nField workers obtained data on the occurrence of diarrhea, water treatment practices and residual chlorine on weekly bases during the study.\nThe secondary outcome was compliance of the intervention.\nIt was assessed on two unannounced and regular weekly visits using free residual chlorine measured with residual chlorine test kit (Wagetech 225 comparator color disc).\nIt is a color wheel test kit that uses DPD (N,N diethyl-p-phenylene diamine) tablet.\nWater samples were collected for bacteriological analysis from half of the randomly selected household water storage containers from both the intervention and the control villages at the baseline and the end of the study.\nSterile 150 ml bottles were used for sample collection.\nThe samples were transported to Haramaya University laboratory using ice packs and reached to the laboratory within 6 hours of collection.\nMultiple tube fermentation technique was used for determination of Escherichia coli which are regarded as the most reliable indicators of fecal contamination.\nThis technique is one of the standard methods in microbiological drinking water quality analysis.\nStatistical Analysis\nData were double entered on to EPI data Version 3.1 and statistical analysis was performed using STATA Version 11.\nMean was calculated to determine average compliance with the intervention.\nIntention- to- treat analysis was used to compare the incidence of diarrhea among children under-five years of age between intervention and control arms.\nWe presented incidence data because is a better predictor of disease transmission, disease surveillance and control.\nGeneralized estimation equation (GEE) with log link Poisson distribution family was used to consider the repeated and clustered nature of the data.\nCrude and adjusted incidence rate ratio along with corresponding 95% confidence intervals was calculated to control the potential confounders.\nWe used mixed effect logistic regression to obtain intera-cluster correlation coefficient (\u03c1).\nResults\nParticipants and Baseline Characteristics\nA total of 427 children in the intervention and 422 children in the control arm enrolled in 36 clusters.\nThe number of children per cluster was 24.\nThe median number of participating households with children under-five years of age per cluster was 16.\nThree households refused to give consent.\nFour households were lost to follow up.\nData obtained from 425 children (284households) in the intervention group and 420 children (285 households) in the control group.\nThe Follow up started in June 2011 and ended in October 2011.\nData was collected on the occurrence of diarrhea for 13499 person week observation representing 98.03% and 97% of the total person week observation in intervention and controls groups respectively (figure 1).\nThere was no any harm observed in this study.\nThe mean ages of the respondents and that of the children under the age of five years were 29.06 years and 26 months respectively.\nOnly 8% mothers/caregivers had formal education.\nThere was no significant difference in baseline demographic and socio-economic characteristics between the intervention and the control households.\nSome of the households were getting water from well (49.2%), some from spring (40.6%) and the rest from stream (10.19%).\nAlmost all households store their drinking water using jerry-cans.\nMost of the households did not treat their drinking water.\nThere was no significant difference on source of water and water treatment practice between the two groups at the baseline.\nAbout 37.6% of the households had latrine.\nBefore the intervention, the two week prevalence of diarrhea was 24.3% in the intervention households and 25.2% in the control households.\nOverall, at the baseline, the intervention and the control households had similar sanitation, hygiene and water handling practices.\nThe two groups were similar in many potential confounding variables, as well (Table 1).\nDiarrhea Incidence\nFrom the control arm, 700 episodes of diarrhea (10.4 episodes per 100 person weeks observation) was reported but, from the intervention arm, it was 307 episodes (4.5 episodes per 100 person weeks observation).\nThe effect of the intervention was different among the children with different age groups.\nIt reduced 63% and 53% of the burden among children 3 to 4 years and 1 to 2 years of age respectively, while the reduction was lesser in less than one year children (44%) (Table 2).\nThe relationship between study weeks and diarrhea in the intervention and control groups is shown in figure 2.\nOn the multivariable analysis adjusted for age, sex of the child, availability of latrine, waste disposal, availability of soap at home and hand washing facility, children in the intervention arm had lower risk of diarrhea (Adjusted RR\u200a=\u200a0.42; 95% CI 0.37\u20130.49).\nThere was a 58% overall reduction in the incidence of diarrhea among the intervention group compared to the controls (Table 3).\nUse of the Intervention\nField workers measured the residual chlorine on a weekly basis throughout the study.\nDuring a weekly scheduled visit, an average 79.89% of the samples from the intervention households had free residual chlorine \u22650.2 mg/l (Figure 3).\nIn unannounced visit 75.94% and 77.13% of the households had free residual chlorine in the 5th and 12th week respectively.\nMembers of the households were not informed about the result of the free residual chlorine test.\nDuring the follow up 1.25% of the control households reported treating their drinking water.\nMicrobiological Quality\nAt the beginning and the end of the study drinking water samples were taken from the intervention and control household water storage containers.\nAt the baseline, 78% of the households from the intervention (median E.coli MPN 70 per 100 ml) and 81% of those from the control households (median E.coli MPN 90 per 100 ml) were contaminated with E.coli.\nAt the end of the study, however, 16.5% of the intervention households (median E.coli MPN 0 per 100 ml) and 65% control households (median E.coli MPN 60 per 100 ml) had E.coli in the sampled water.\nAt the baseline, there was no significance difference between the two groups.\nHowever, drinking water samples from the intervention households were more likely to meet the WHO guidelines for bacteriological quality than samples from the control households at the end of the study (P<0.001) (Table 4).\nDiscussion\nA community based cluster randomized trial was conducted to assess the effectiveness of household water treatment with chlorine in reducing diarrhea among children under-five years of age.\nHousehold water treatment with sodium hypochlorite has reduced the incidence of diarrhea among children under-five years of age assigned to the intervention compared to the control (RR\u200a=\u200a0.42 95% CI 0.37\u20130.49).\nThere was significant improvement in the quality of stored water in intervention households.\nOur result on reduction of diarrhea episode was consistent with similar studies conducted in Kenya and Haiti.\nBut, the finding was higher than other studies from Bolivia and Brazil.\nThe high magnitude of the protective effect in this study could be attributed to the high compliance of the intervention as observed during both scheduled and unannounced visits.\nOther study indicated that beneficial effect of such interventions may be greater in population where fecal contamination of drinking water is more likely.\nWe observed high fecal contamination of household stored water during the baseline thus that could be the reason for observing the desired effects of the intervention.\nThe household water treatment in this study was less effective among the younger children (less than one year) and the result is consistent with a study conducted in rural Guatemala.\nThis might be due to the high susceptibility and more chance of exposure to contaminated supplemental liquids among younger children.\nHigh level of intervention compliance was achieved in the current study as observed in Zambia and Gana.\nSub-group analysis within the intervention arm showed association between increased compliance and lower incidence of diarrhea (P\u200a=\u200a0.014).\nThus, the observed reduction in diarrheal episodes among the treatment group can be attributed to the intervention.\nHousehold water chlorination is the most cost effective among all water quality interventions.\nFor instance, in the study area, 125 ml 1.25% sodium hypochlorite was bought $0.28 per bottle from retail shops or drug vendors, which can serve an average family for a month.\nThis is about $ 0.00035 per liter that could be affordable for many of the households if they have access to the service.\nUsing local community for the distribution of the disinfectant is a promising strategy to reach the rural community where the risk of diarrhea and other water borne diseases is high.\nThe primary outcome for this study was self reported diarrhea, which could be sensitive to recall bias.\nTo minimize this recall bias, the data collection was on a weekly basis.\nThis study was not blinded due to taste and odor of chlorine as well as ethical issue.\nCourtesy bias and Howthrone effect may overstate effectiveness of the HWT intervention in unblinded trials.\nHowever, in this study, independent field workers were used for introducing the intervention and for collecting of data.\nMausezahl et al recently indicated that in the absence of blinding using independent data collectors and intervention implementers is critical to reduce bias in HWT studies.\nThe investigators were not involved in the implementation of the intervention and data collection.\nThe objective of the study was not stated for data collectors.\nWe believe that sources of bias are addressed and their effects, if any, on the reported estimates are minimal.\nIn conclusion, home treatment of water with sodium hypochlorite significantly reduced the incidence of diarrhea among under-five children in the rural population where fecal contamination was high.\nWe suggest that increasing access to such intervention would decrease child morbidity and mortality caused by diarrhea and help to achieve millennium development goal of 4 and 7.\nCommunity randomized trial flow of participants on household water chlorination, eastern Ethiopia, 2011.\nWeekly prevalence of diarrhea versus weeks of observation, Kersa district, Eastern Ethiopia, 2011.\nPercentage of residual chlorine during observation period, Kersa district, Eastern Ethiopia, 2011.\n\nBaseline characteristics of community and household of the randomized control trial, Kersa district, Eastern Ethiopia, 2011.\nVariable | Control | Intervention | P value\nNumber of clusters (neighborhoods) | 18 | 18 | \nNumber of households | 285 | 284 | \nNumber of under-five children | 420 | 425 | \nMean family size per household | 5.79 | 6.26 | 0.15\nMean age of the children | 26.16 | 26.56 | 0.65\nPrimary caregiver of children |  |  | \nMean age | 29.6 | 29.3 | 0.45\nNo formal education | 264 (92.6) | 260(91.5) | 0.63\nOccupation (housewives) | 276(96.8) | 282(99.3) | 0.31\nMain occupation of the head of the household as farmer | 278 (97.5) | 273(96.1) | 0.40\nEconomic indicators |  |  | \nOwn land | 275(96.5) | 279(98.2) | 0.19\nOwn watch | 119(41.8) | 123(43.3) | 0.70\nOwn mobile | 27(9.5) | 37(13) | 0.18\nOwn television | 10(3.5) | 13(4.6) | 0.51\nOwn radio | 86(30.2) | 90(31.7) | 0.69\nPrimary water source |  |  | \nWell | 142(49.8) | 138(48.6) | 0.76\nSpring | 118(41.4) | 113(39.8) | 0.69\nStream/river | 25(8.8) | 33(11.6) | 0.26\nDomestic water treatment and storage |  |  | \nTreat water before drinking (any method) | 3(1) | 5(1.7) | 0.47\nStorage water at home | 285(100) | 284(100) | NA*\nUse Jerry can to store water | 285(100) | 283(99.6) | 0.31\nSanitation and hygiene |  |  | \nPlace to wash hand | 72(25.3) | 68(23.9) | 0.71\nSoap available | 36(12.6) | 27(9.5) | 0.23\nWaste disposal (proper) | 79 (27.7) | 71(25) | 0.48\nLatrine present | 106(37.1) | 108(38) | 0.83\nTwo week prevalence of diarrhea | 106(25.2) | 103(24.3) | 0.73\n\nNA\u200a=\u200anot applicable.\n\nEffect of the intervention with different age group of under-five children, Kersa district, Eastern Ethiopia, 2011.\nAge group | Control groups (N\u200a=\u200a420) | Intervention groups (N\u200a=\u200a425) | % reduction in | P value\n | Number ofDD episode | PWO | DD incidence | Number ofDD episode | PWO | DD incidence | DD incidence | \n<1 year | 104 | 1055 | 9.8 | 53 | 957 | 5.5 | 44 | 0.001\n1\u20132 years | 274 | 2173 | 12.6 | 131 | 2218 | 5.9 | 53 | <0.001\n3\u20134 years | 322 | 3486 | 9.2 | 123 | 3610 | 3.4 | 63 | <0.001\nAll <5 years | 700 | 6714 | 10.4 | 307 | 6785 | 4.5 | 57 | <0.001\n\nDD\u200a=\u200adiarrhea diseases, PWO\u200a=\u200aPerson week of observation. The incidence of diarrhea was calculated as the number of new episodes divided by the total number of person\u2013weeks observation.\n\nMultivariable analysis of intervention effect on the incidence of diarrhea among under-five children, Kersa district, Eastern Ethiopia, 2011.\nFactors | Crude IRR(95% CI) | Adjusted IRR(95% CI) | P value\nIntervention | 0.43(0.37\u20130.50) | 0.42(0.36\u20130.48) | <0.001\nControl | 1 | 1 | \nAge of the child | 0.88(0.83\u20130.94) | 0.89(0.84\u20130.94) | <0.001\nSex of the child |  |  | \nFemale | 0.98(0.85\u20131.14) | 1.02(0.90\u20131.16) | 0.709\nMale | 1 | 1 | \nLatrine available |  |  | \nYes | 0.88(0.75\u20131.02) | 0.99(0.85\u20131.15) | 0.944\nNo | 1 | 1 | \nProper waste disposal |  |  | \nYes | 0.82(0.69\u20130.97) | 0.81(0.68\u20130.97) | 0.027\nNo | 1 | 1 | \nSoap available at home |  |  | \nYes | 0.83(0.64\u20131.07) | 0.88(0.68\u20131.14) | 0.363\nNo | 1 | 1 | \nHand washing facility available |  |  | \nYes | 0.93(0.78\u20131.10) | 0.90(0.77\u20131.05) | 0.218\nNo | 1 | 1 | \n\n\nHousehold stored water quality among intervention and control households at the baseline and end point of intervention.\n | Intervention households | Control households | P value\nNumber (%) of households with E.coli |  |  | \nBaseline | 109(78%) | 114(81%) | 0.53\nEnd point | 23(16.5%) | 92(65%) | <0.001\nMedian E.coli per 100 ml of drinking water |  |  | \nBaseline | 70(0\u20131600) | 90(0\u2013900) | 0.735\nEnd point | 0(0\u2013280) | 60(0\u2013500) | <0.001\n\nMedian E.coli between the intervention and control households was compared using Wilcoxon rank sum test. The number of households with E.coli contamination between the two arms was compared using t-test.",
        "label": "low",
        "id": "task4_RLD_test_671"
    },
    {
        "paper_doi": "10.3390/tropicalmed4040141",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Design\ncNON-RCT*\nAllocation of clusters1 village randomized* to intervention, 1 to control\n\n\nParticipants: 527 individuals ages 3 to 70\n\n\nInterventions: Broad multiple\n\n\nOutcomes: Any STH\n\n\nNotes: *The study may have used a random mechanism to allocate the intervention, but there was only 1 intervention area compared to 1 control area, so randomization in this case not likely to have reduced confounding or imbalances\n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Many latrine campaigns in developing countries fail to be sustained because the introduced latrine is not appropriate to local socio-economic, cultural and environmental conditions, and there is an inadequate community health education component.\nWe tested a low-cost, locally designed and constructed all-weather latrine (the \u201cBALatrine\u201d), together with community education promoting appropriate hygiene-related behaviour, to determine whether this integrated intervention effectively controlled soil-transmitted helminth (STH) infections.\nWe undertook a pilot intervention study in two villages in Central Java, Indonesia.\nThe villages were randomly allocated to either control or intervention with the intervention village receiving the BALatrine program and the control village receiving no program.\nSTH-infection status was measured using the faecal flotation diagnostic method, before and eight months after the intervention.\nOver 8 months, the cumulative incidence of STH infection was significantly lower in the intervention village than in the control village: 13.4% vs. 27.5% (67/244 vs. 38/283, p < 0.001).\nThe intervention was particularly effective among children: cumulative incidence 3.8% (2/53) for the intervention vs. 24.1% (13/54) for the control village (p < 0.001).\nThe integrated BALatrine intervention was associated with a reduced incidence of STH infection.\nFollowing on from this pilot study, a large cluster-randomised controlled trial was commenced (ACTRN12613000523707).\n1. Introduction\nThe global prevalence of infection with soil-transmitted helminths (STH) remains high, with 1.5 billion people infected worldwide, many of them children.\nOver two thirds of STH infections are in Asia, mostly in Southeast Asia.\nThe prevalence of STH infection in Indonesia is high at 45\u201365%, with areas having poor sanitation reaching 80% prevalence.\nIn Central Java, research into STH infection among elementary school children by Laksono and later the Health Department, found an infection prevalence of 84\u201396%.\nMore recently, a cross-sectional survey in Semarang, Central Java, found a prevalence of 34% among a cohort of 6466 people aged between two and 93 years.\nAnthelmintic drugs aimed at reducing morbidity are effective, but only temporarily, with a cure often followed by subsequent reinfection.\nIn rural areas, open defecation coupled with poorly constructed or inadequate latrines allows STH eggs to spread infection.\nTherefore, for long-term prevention, improved sanitation and community education are essential.\nA recent systematic review and meta-analysis concluded that \u201cintegrated control approaches emphasizing health education and environmental sanitation are needed to interrupt transmission of STH\u201d.\nIn particular, interventions that improve the hygienic disposal of faeces to reduce soil and/or water contamination have been identified as a key strategy to control transmission and prevent related diseases.\nIn Indonesia, open defecation is common, with 55% of the poorest and 18% of the richest households practicing open defecation.\nIn 2010, less than 40% of the people in rural areas had improved latrines, defined as facilities that hygienically separate human excreta from human contact.\nThe country did not reach its Millennium Development Goal of 75% sanitation coverage by 2015.\nIn rural areas, which include 118 million people, or 46.3% of the country\u2019s population, it has been estimated that 47% of the population have improved latrines, 12% shared latrines, 12% other unimproved latrines and 29% no latrines (i.e., practice open defecation).\nCompounding the problems caused by the lack of improved latrines is inappropriate hygiene-related behaviour, particularly related to hand washing, with 2007 National baseline data indicating that less than a quarter (23.2%) of the population had appropriate hand-washing behaviour.\nThe aim of this study was to develop and test an integrated approach to the prevention of STH infection and reduction of both transmission and reinfection.\nOur intervention included anthelmintic drugs, the construction and adoption of improved latrines, and effective education regarding hygienic and sanitary behaviour.\n2. Methods\n2.1. Study Design\nThis study was conducted in two villages, Palemon and Cepoko, in the Gunungpati sub-district of the city of Semarang, Central Java, Indonesia (see Figure 1), from July 2011 to June 2012.\nA random selection was made from these villages regarding which one should receive the integrated intervention and which one should be the control, by researchers who had no prior knowledge or contact with the villagers or village officials.\nThe two villages though similar in size and local topography, were not in close proximity to each other.\nThe study area is wooded and hilly and most of the houses, often made of local brick, are constructed by the householders themselves.\nMore than half of the households in the study villages did not have their own latrines.\nA randomly selected cohort (control: n = 244; intervention: n = 283) was followed over the eight-month duration of the study.\nA questionnaire was administered at baseline and follow-up to all village residents, with the eligibility criterion of being over two years of age.\nParticipants also provided two stool samples for parasitological examination.\nFollowing the baseline survey, all residents (regardless of infection status) were treated with anthelmintic medication and the incidence of STH infection was assessed at follow-up.\nFor ethical reasons, participants who were found to be STH positive at follow-up were re-treated.\n2.2. Ethics \nEthical approval was given by the Semarang City authorities (ref. 070/613/IV/2011), and from the Human Research Ethics Committee at Griffith University (ref. PBH/17/11/HREC).\n2.3. Procedure\nOur study procedure reflected the integrated model previously described (see also Figure 2), comprising chemotherapy, a locally constructed latrine (the \u201cBALatrine\u201d) and community health education.\nFollowing WHO Guidelines, a single oral dose of Albendazole (400mg) was administered immediately after the baseline survey.\nThe BALatrine is designed for resource poor rural communities and emergency situations, to be made by local people using local materials.\nTesting for cultural acceptance was conducted in the field through pilot studies in Pekalongan in Central Java and BALatrines were also used in an emergency refugee camp during the 2010 eruption of the Mt Merapi volcano, where they were proven appropriate for the level of technology available in the village context.\nThe BALatrine is a relatively simple squat latrine (Figure 3) that can be constructed by village residents using inexpensive materials.\nWhen water for flushing is available, a U-bend (\u2018goose-neck\u2019) water closet can be attached.\nWhen water is not available, such as during the dry season, the latrine can be used in a dry-pit configuration (with removal of the U-bend attachment), with a lid to isolate it from insects and to prevent odours from escaping.\nThus, it can function despite seasonal changes in water supply.\nBesides being inexpensive (cost at the time of the study was $50 USD for the latrine and local building materials; equating to approximately $80\u201390 at the time of publication), for people with limited financial and educational resources it is easy to copy.\nIts construction and use reflect critical resource, environmental and technical issues and due to local village input it also overcomes some major disincentives embodied in conventional latrines by being culturally familiar, simple and easy to use.\nThe community health education programme is an essential adjunct to latrine construction.\nIn the intervention village, all residents were given health education regarding hygiene, sanitation, and prevention of STH infections.\nThis health education component was delivered via community meetings in each village.\nAll village residents were invited and meetings were held in the village meeting hall.\nThe health education/health promotion component of the intervention was implemented through a two-hour village-wide mobilization meeting, which formed the project launch and was designed to mobilize households by consciousness-raising and provision of information about parasite infection and burden of STH.\nSubsequently, a series of small group workshops took place with the villagers in order to describe the BALatrine construction in detail and how to plan, construct, use, and maintain their latrines, as well as to discuss STH disease pathways.\nThe content of the health education programme comprised information about the dangers of STH infections and, using illustrated leaflets, how the transmission of STH infections can be prevented by the construction of latrines and with appropriate hygiene-related behaviours.\n2.4. Measurements and Analyses\nThe primary outcome measures of the integrated intervention were STH infection status at baseline and at follow-up.\nSTH infection status was measured through laboratory analysis of stool samples collected from each participant at baseline and at follow-up eight months after the BALatrines were constructed.\nThe samples were analysed microscopically for the presence of helminth eggs, according to the Willis-Mollay Flotation technique.\nAfter preparing the samples, a cover slip was placed over each sample tube and left for 10 minutes.\nAfter 10 minutes, a drop of eosin solution (2%) was added to a glass slide onto which the cover slip was then placed and observed using a light microscope at 10 \u00d7 magnification.\nA positive sample was where at least one STH egg was identified.\nA face-to-face questionnaire (the Helminth Education and Latrine Project (HELP) questionnaire) was also administered at baseline and follow-up and this provided information about villagers\u2019 demographic attributes.\nWe also assessed local village environmental contamination with faeces.\nThese findings have been published elsewhere.\nData were analysed using SPSS Version 22 (IBM, New York, United States), Microsoft Excel and the tools at Open Source Epidemiologic Statistics for Public Health.\nDifferences between participants in control and intervention villages were analysed using the unpaired t-test and the Pearson\u2019s chi-squared (\u03c72) test.\nLogistic regression analysis was performed and odds ratios calculated.\nTo compare the intervention village with the control village, we report both crude odds ratios and adjusted for age and sex.\n3. Results \n3.1. Characteristics of the Participants\nThere were 527 participants in the study at baseline, 244 in the control village and 283 in the intervention village.\nTheir ages ranged from three to 70 years, with a mean \u00b1 SD of 29.4 \u00b1 16.1 years in the control village and 32.2 \u00b1 17.9 years in the intervention village (Table 1; for the age difference between villages, p = 0.06).\nIn both villages, similar proportions of the participants had completed elementary education (control, 211/244, 86.5%; intervention, 253/283, 89.4%).\nIn the control village, 98.8% of the residents had a monthly income below 1 M Indonesian Rupiah (IDR) or about US$70, whereas in the intervention village the comparable percentage was 97.2% (p = 0.017).\nIn the control village, 38.5% of the residents lived in a home in which all floor spaces were dry, whereas in the intervention village the comparable percentage was 54.1% (p < 0.001).\nIn the intervention village greater than 90% of households adopted the BALatrine, as measured at the follow up.\n3.2. STH-Infection Status\nAt baseline, the prevalence of STH infection was almost the same in the two villages (Table 1).\nAt follow-up, the cumulative incidence of infection was much lower in the intervention village than in the control village (Table 2, 13.4% vs. 27.5%, p < 0.001).\nAfter adjustments for age and gender, the benefit of the intervention was still clear (adjusted odds ratio = 0.38, 95% CI 0.25\u20130.60, Table 2).\nThe intervention was particularly effective in children (adjusted odds ratio = 0.12, 95% CI 0.03\u20130.56, Table 2).\n4. Discussion\nImproving access to adequate sanitation is a critical step toward the sustainable interruption of STH transmission.\nYet this is an ongoing challenge in low resource settings where public sewerage system infrastructure rarely exists, on-site sanitation systems are either improperly designed or poorly functioning, and open defecation is seen as culturally acceptable, especially in rural areas.\nOften people must rely on shared sanitation facilities, which have been shown to increase the risk of adverse health outcomes compared to individual household latrines.\nCompounding the issue is limited access to materials and a lack of technical expertise to build or improve latrines.\nThe BALatrine was specifically designed to overcome many of these challenges whilst also taking into consideration cultural appropriateness and acceptability.\nUsing simple technology and locally sourced, inexpensive materials, the BALatrine is cheap, easy to build and maintain and adaptable for wet and dry conditions.\nBuilding the latrines through community mobilisation also helps keep the costs down and enables the householders themselves to take ownership over the latrines and their maintenance and consequently, benefit from improved health, helping alleviate the cycle of poverty that is often associated with STH infections.\nIn the present study, we evaluated the effectiveness of an integrated BALatrine intervention at reducing human worm burden through a pilot study in two villages in Central Java.\nPeople in the intervention village were 2.6 times less likely to be infected following the BALatrine-based intervention than those in the control village indicating that the BALatrine is associated with a reduced worm burden.\nHowever, it is important to note that we did not resurvey participants after baseline deworming to determine the efficacy of the Albendazole treatment nor did we differentiate between STH species, which are known to respond differently to treatment.\nWe have based our interpretation of the results on the assumption that treatment was effective at temporarily reducing infections to zero.\nConsequently, it is possible that the effect seen could be a result of differential cure rates between villages based on their STH profile at baseline.\nHowever, a prevalence study we conducted the following year (manuscript under review) across 16 villages in two neighbouring subdistricts of Semarang, including Gunung Pati, revealed Ascaris lumbricoides as the predominant species (mean prevalence of 26% vs. 7.9% and 1.8% for hookworm and Trichuris trichuria, respectively).\nWe therefore believe that it is not unrealistic to assume that our study villages also had similar burdens of each of the STH species at baseline and that treatment would be similarly effective across the two sites.\nOur study also found that nearly all households adopted the BALatrine suggesting a strong willingness among villagers to improve sanitation and a desire to improve the health of their families.\nAccess to improved latrines does not guarantee their use, however, particularly over time when old habits or cultural preferences can be difficult to overcome.\nIntegrating health education and promotion programmes to improve peoples\u2019 understanding and knowledge of the link between open defecation and ill-health and WASH-related behaviours, is therefore extremely important.\nPeople also need to understand the importance of properly cleaning and maintaining their latrines, particularly as this is associated with higher latrine use.\nIn the present study, a community health education programme was administered prior to the construction and installation of the latrines and aimed to raise awareness, improve hygiene behaviours and motivate the villagers to build and continue to use their new latrines, which may have led to the high uptake observed in this study.\nHowever, we did not assess the impact of this program on participants\u2019 knowledge and behaviour change or assess latrine use over time, which are key limitations of this study.\nIt is well established that chemotherapy alone will not break the transmission cycle.\nRecent studies have also shown that programmes solely focusing on WASH have limited effect on STH incidence and may provide no additional impact compared to mass drug administration programmes alone.\nIn contrast, health education and promotion programmes can be highly effective at reducing the incidence of STH infections if designed appropriately such as the highly successful \u201cMagic Glasses\u201d study, which resulted in a 50% reduction in STH infections.\nHowever, sustained reinforcement of health messages is required in order to increase their effectiveness over the longer term.\nUltimately, eliminating STH will be best achieved through integrated control programmes.\nThe current study adds to the growing body of research into the impact of integrated control programs on soil-transmitted helminthiases and demonstrates that sanitation interventions can be effective at reducing worm burden when designed appropriately for the local context and combined with health education and promotion.\nIn conclusion, our findings provide \u201cproof of principle\u201d that the BALatrine-based intervention is effective in preventing STH infection.\nWe will now undertake a full-scale randomized controlled trial and contribute much needed evidence based on WASH and STH.\nMap of Gunung Pati subdistrict in Semarang, Central Java (source: Wikipedia Indonesia, 2017).\nFlowchart of the study.\nSchematic presentation of the BALatrine.\n\nBaseline characteristics of participants.\nVillage Status | Control | Intervention\nVillage  | Cepoko | Palemon\nSample Size | 244 | 283\nMean Age (years) | 29.4 | 32.2\nPrevalence of STH infection: % (95% CI) | 21.7% (16.5\u201326.9) | 25.8% (20.7\u201330.9)\nSex Ratio (F/M) | 141/103 | 151/132\nPrevalence of STH infection by Sex (F/M)  | 22.0%/21.4% | 20.5%/31.8%\n\n\nInfection rates in the control and intervention villages.\nVariable | Control | Intervention | Odds Ratio | Odds Ratio\nCrude | p Value | Adjusted * | p Value\nAll participants | n = 244 | n = 283 |  |  |  | \nPrevalence of infection at baseline: % (95% CI) | 21.7 (16.5\u201326.9) | 25.8 (20.7\u201330.9) | - | - | - | -\nCumulative incidence of infection at follow-up: % (95% CI) | 27.5 (21.9\u201333.1) | 13.4 (9.5\u201317.4) | 0.41 (0.26\u20130.64) | <0.001 | 0.38 (0.25\u20130.60) | <0.001\nChildren | n = 54 | n = 53 |  |  |  | \nPrevalence of infection at baseline: % (95% CI) | 18.8 (8.2\u201328.9) | 18.9 (8.3\u201329.4) | - | - | - | -\nCumulative incidence of infection at follow-up: % (95% CI) | 24.1 (12.7\u201335.5) | 3.8 (0.0\u20138.9) | 0.12 (0.03\u20130.58) | 0.01 | 0.12 (0.03\u20130.56) | 0.01\n\n* The model for all participants was adjusted for age and sex. The model for children (<14 years) was adjusted for gender only.",
        "label": "high",
        "id": "task4_RLD_test_814"
    },
    {
        "paper_doi": "10.1186/1471-2334-6-16",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Randomized (computer-generated) to PQ or bulaquine (1:2 ratio) after primary treatment with quinine + doxycycline\n\n\nParticipants: 93 participants in IndiaInclusion criteria> 16 years.Male.Uncomplicated P. falciparum only.> 55 P. falciparum gametocytes/mL on admission.Exclusion criteriaAntimalarial treatment in previous two weeks.Allergy to trial drug.G6PD deficient.\n\n\nInterventions: All patients: quinine days 1 to 7: 30 mg/kg/day (10 mg/kg/day three times per day) + 100 mg doxycyclineRandomization and treatment on day 4PQ.Bulaquine.\n\n\nOutcomes: Gametocyte prevalence, density and viability on days 1, 4, 15, 22, and 29Adverse events\n\n\nNotes: Gametocyte viability assessed by Shute's technique (ex flagellation\n\n",
        "objective": "To assess the effects of single dose or short\u2010course PQ (or an alternative 8AQ) alongside treatment for people with P. falciparum malaria.",
        "full_paper": "Background\nThe WHO recommends that adults with uncomplicated P. falciparum successfully treated with a blood schizonticide receive a single dose of primaquine (PQ) 45 mg as a gametocytocidal agent.\nAn earlier pilot study suggested that 75 mg of bulaquine (BQ), of which PQ is a major metabolite, may be a useful alternate to PQ.\nMethods\nIn a randomized, partial blind study, 90 hospitalized adults with Plasmodium falciparum malaria that was blood schizonticide-responsive and a gametocytemia of > 55/\u03bcl within 3 days of diagnosis were randomized to receive single doses of either PQ 45 mg or BQ 75 mg on day 4.\nWe assessed gametocytemia on days 8, 15, 22 and 29 and gametocyte viability as determined by exflagellation (2\u00b0 end point) on day 8.\nResults\nOn day 8, 20/31 (65%) primaquine recipients versus 19/59 (32%) bulaquine recipients showed persistence of gametocytes (P = 0.002).\nAt day 15 and beyond, all patients were gametocyte free.\nOn day 8, 16/31 PQ and 7/59 BQ volunteers showed gametocyte viability (p = 0.000065).\nConclusion\nBQ is a safe, useful alternate to PQ as a Plasmodium falciparum gametocytocidal agent and may clear gametocytemia faster than PQ.\nIntroduction\nMalaria remains the most important parasitic infection with 300\u2013500 million people affected yearly and 1.5\u20132.7 million deaths each year.\nWorld over, malaria control has focused on pharmacological intervention, vector control, curtailing irrational and indiscriminate use of antimalarials, and the development of vaccines.\nOf these strategies, pharmacological intervention remains the most effective way to combat malaria\n8 aminoquinolines like primaquine are unique antimalarials in that they exhibit activity against multiple life cycle stages of Plasmodia that infect humans.\nPrimaquine 45 mg as a single dose is recommended by the World Health Organization (WHO) and the National Antimalarial Programme (NAMP) of India for its gametocytocidal activity in P. falciparum.\nIn 1998, Gogtay et al found that the efficacy of PQ 45 mg as a falciparum gametocytocidal agent in patients sensitive to chloroquine in India to be approximately 77% at Day 29 follow up.\nDuring the past several years, attempts have been made to produce primaquine analogs with improved anti-malarial activity and lower toxicity.\nBulaquine, formerly called CDRI 80/53, is metabolized to PQ and differs from PQ only by the 2,4 dihydrofuran group present in the basic side chain anchored onto the quinoline nucleus in the 8 position.\nBulaquine is currently licensed only for use in India for the radical cure of vivax malaria dosed at 25 mg/day for 5 days, but not as a gametocytocidal agent IN Plasmodium falciparum.\nThe results of a pilot study by our group assessing the P. falciparum gametocytocidal effect of a single dose of bulaquine 75 mg in India suggested it may be more effective than PQ 45 mg.\nHere, in a larger population of adults with P falciparum malaria successfully treated with blood schizonticides, we compared the gametocytodical activity of BQ and PQ.\nVolunteers and methods\nProtocol\nThe protocol was approved by the institutional ethics committee and the Drugs Controller General of India.\nThe study was conducted between January 2002 and April 2004.\nEnrollment and procedures\nPatients who were at least 16 years old with uncomplicated Plasmodium falciparum infection, provided written informed consent, and had a gametocyte count > 55/\u03bcl within 72 hours of diagnosis, regardless of asexual parasite counts, were eligible for enrollment.\nThe minimal gametocyte count was chosen based on infectivity to mosquitoes.\nPatients who were pregnant or lactating, had received antimalarial treatment in the previous 2 weeks, had co-infection with Plasmodium vivax, claimed an allergy to primaquine or bulaquine, or were G6PD deficient were excluded.\nOn admission, patients were initially assessed by thick and thin blood films stained using the Jaswant Singh and Bhattacharji (JSB) field stain.\nSubsequently, Giemsa stained blood smears were used to determine the number of asexual and sexual parasites/\u03bcl, assuming a white blood cell count of 8000/\u03bcl.\nEnrolled patients were admitted to hospital on Day 1 and treated under observation with quinine orally 10 mg/kg/day thrice daily for a total of 7 days and doxycycline 100 mg once daily for 7 days.\nAt day 4, consecutive patients were randomly allocated in a 1:2 fashion to receive an observed single dose of either PQ 45 mg or bulaquine 75 mg, based on a computer generated randomization code.\nUnequal allocation was used because of earlier studies suggesting the superiority of bulaquine.\nThe test articles were administered on day 4 because the incidence of nausea and vomiting is higher in the first few days of schizonticidal therapy and this was given regardless of parasite clearance.\non day 8, all patients were assessed for gametocytemia, discharged, and asked to follow up on days 15, 22, and 29 for further safety and parasitologic checks.\nGiemsa stained malaria blood smears during hospitalization were done twice a day for the first 72 hours and once a day thereafter until discharge and on the follow-up days.\nThe slide readers were blinded to the treatment.\nOutcomes\nEfficacy was assessed by gametocytemia (primary end point) and gametocyte viability (secondary end point) on admission and all follow up days.\nThe latter was assessed by the modified Shute's technique.\nThis technique depicts exflagellating microgametes in blood films that have been kept moist at 21\u201325\u00b0C for 1 hour with complete RPMI medium and AB positive serum and then Giemsa stained.\nOne or more exflagellating microgametes was considered a positive test.\nThese end points were identical to the previous study.\nSafety was monitored by routine clinical hematological and biochemical laboratories and an electrocardiogram on days 1 and 8.\nAdverse event recording was focused only on nausea, vomiting, and epigastric distress and were recorded only if not reported on Day 1 or if a symptom worsened after Day 1.\nSample size and statistical analysis\nThe estimated sample size was calculated using Casagrande's method based on a previous study by Gogtay et al comparing the two drugs.\nAssuming a 30% difference in efficacy on Day 8, at 5% significance and 90% power, a sample size of 28 patients and 56 patients are required in the primaquine and bulaquine group, respectively, to demonstrate the superiority of bulaquine.\nP values \u2264 0.5 were considered significant.\nResults\nA total of 93 male patients were enrolled.\nWomen with malaria are not inclined to get admitted, especially because of hardships related to hospitalization and supervised drug administration.\nThe age of the patients ranged from 16\u201372 years: 31.47 \u00b1 11.62 years).\nThere were three drop outs, two in the Bulaquine arm and one in primaquine arm.\nThese patients did not return for any follow up visit after discharge from the hospital and were omitted from analysis.\nAt admission, gametocytaemia between the 2 groups was similar.\nAt day 8, 20/31 (65%) PQ recipients and 19/59 (32%) BQ recipients had gametocytes on blood smear (p = 0.002).\nAt days 15, 22, and 29, all patients in both treatment groups were free of gametocytes.\nAt day 8, 16/31 (52%) PQ recipients and 7/59 (12%) BQ recipients had viable gametocytes on exflagellation testing (Table)(p = 0.000065).\nAll patients with viable gametocytes had smear positive gametocytemia.\nThere were no important clinical hematology or biochemical laboratory values, and all electrocardiograms were within normal limits.\nDiscussion\nA single dose of 45 mg primaquine is given along with or after schizonticidal therapy in areas where malaria is endemic as a transmission blocking strategy and currently is the only option available for this indication.\nThe present study carried out in 91 cases of uncomplicated Plasmodium falciparum malaria assessed the efficacy of bulaquine, the parent compound of primaquine.\ngiven as a single dose of 75 mg for its gametocytocidal effect.\nOn day 8 of therapy, fewer patients with bulaquine had gametocytes as compared to those who received 45 mg primaquine.\nThis suggests superior efficacy of bulaquine or its ability to clear gametocytes more rapidly, as all patients were free of gametocytes by day 15 and beyond.\nThis clearance of gametocytes completely by Day 15 in both groups in the present study contrasts with our previous study with 45 mg primaquine where persistent gametocytemia was seen in 23% patients responsive to chloroquine up to Day 29.\nThis may in part be due to use of different schizonticidal agents in the two studies, with chloroquine being used in the former and quinine in the latter, with quinine leading to greater clearance of asexual parasites.\nBased on the results of the present study, bulaquine in a single dose of 75 mg may represent yet another treatment option for gametocytocidal effect in addition to primaquine and the current licensing of the drug in the country could be changed to include gametocytocidal effect apart from anti-relapse effect.\nWhether increasing the dose of primaquine from 45 mg to 60 mg will improve efficacy needs to be addressed in future studies.\nIn this study, the Shute technique, which measures the ability of the male gametocyte to exflagellate was used as a surrogate marker for assessing transmission blocking.\nAssessment of true gametocytocidal efficacy of any drug will depend upon demonstration of the ability to block transmission to mosquitoes.\nThis in turn can be assessed by only checking for the presence of the oocyst and ookinete in the mosquito midgut, not done in the study.\nIn the first study where we first reported the declining efficacy of primaquine as a gametocytocidal agent, chloroquine versus coartemether was studied, since chloroquine then was (and still remains) the first line therapy for the country.\nThe results of the study showed a high degree of chloroquine resistance (> 50%), and we shifted to using quinine as first line for uncomplicated falciparum malaria in the hospital, since there are only isolated reports of quinine resistance in the country.\nDrugs that can block the spread of malarial parasites by killing or preventing the transmission or maturation of gametocytes, represent important tools in malaria control.\nSupputtamongkol et al compared the efficacies of mefloquine-artesunate and mefloquine-primaquine on the subsequent development of gametocytemias.\nThe latter combination was not found to be effective in either clearance of existing gametocytemias or the prevention of new gametocytemias.\nPrimaquine has an extremely short half life, and as such may not be completely able to eliminate gametocytes and needs to be used with an effective schizonticidal agent.\nPersistence of gametocytemia is one of the predictors of treatment failure and use of an effective anti-malarial drug to eradicate asexual forms will still remain the most effective means to prevent gametocytemias.\n\nGametocytaemia of patients at different days of follow-up in the two groups\nDays of follow-up | Primaquine n = 31. Gametocyte density/\u03bcl(mean \u00b1 SD) | Bulaquine n = 59. Gametocyte density/\u03bcl(mean \u00b1 SD)\n1 | 80-6040(1342 \u00b1 182.80) | 80-4160(1064.2 \u00b1 182.80)\n4 | 120-8000(1494 \u00b1 385.04) | 80-3260(1003.7 \u00b1 118.35)\n8 | 80-2120(543 \u00b1 128.75) | 32-1820(161.01 \u00b1 32.76)\n15,22,29 | Nil | Nil\n\n\nResults of Efficacy.\nDays of follow -up | No of patients positive for gametocytes | Patients sample positive for exflagellation\n | Pts treated with PQN = 31 | Pts treated with BQN = 59 | Pts treated with PQN = 31 | Pts treated with BQN = 59\n\n1 | 31/31100% | 59/59100% | 31/31100% | 59/59100%\n4 | 31/31100% | 59/59100% | 31/31100% | 59/59100%\n8 | 20/31 *65% | 19/59 *32% | 16/31 *52% | 7/59 *12%\n15 | Nil | Nil | Nil | Nil\n22 | Nil | Nil | Nil | Nil\n29 | Nil | Nil | Nil | Nil\n\n* statistically significant P < 0.05\nAcknowledgements: This study was carried out as a project under the Center for Advanced Research in Clinical Pharmacology, and was funded by the Indian Council of Medical Research, New Delhi. We thank Nicholas Piramal India Ltd for the supply of bulaquine capsules.",
        "label": "low",
        "id": "task4_RLD_test_910"
    },
    {
        "paper_doi": "10.1371/journal.pone.0190751",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Year: October 2011-March 2013Location: Groningen, the NetherlandsTrial method: RCT stratified into 3 risk profiles (robust, frail, or complex)\n\n\nParticipants: Total number randomised: 1456 (602 eligible for this review)Mean age: intervention group 80.6 years, control group 80.8 yearsMale/female proportion: control group 394 women; intervention group 405 womenFrailty status: assessed using Groningen Frailty IndicatorInclusion criteriaAge >=75 yearsRegistered with participating GPLiving at home or in a home for the elderlyExclusion criteriaLong-term admission to a NH (not just for rehabilitation)Alternative type of integrated careParticipation in another research study\n\n\nInterventions: Name of intervention: EmbraceWhy (aim): \"to evaluate the effects of the population-based, person-centred and integrated care service Embrace on patient-reported outcomes at 12 months\"What (materials): in-person meetingsWhat (procedures): Embrace is a \"person-centred and integrated care service for community-living older adults\" delivered through regular community meetings \"in which self-management abilities were encouraged and during which local healthcare and welfare organisations provided information on health maintenance, physical and social activities, and dietary recommendations. In addition, frail people and those with complex care needs received individual support from a case manager.\"Who: multidisciplinary team comprising nursing home physician and 2 care managers (district nurse and social worker) trained in the principles and methods of EmbraceHow: through regular Embrace community meetingsWhere: community settingWhen and how much: 12 monthsTailoring: not specifiedModifications: not specifiedHow well (planned): not specifiedHow well (actual): not specifiedComparison: care as usual\n\n\nOutcomes: Mortality: number of individuals who died (monitored)Change in place of residence to a nursing or residential home: not reportedQOL: EQ-5D, EQ-5D-VAS Serious adverse effects: not reportedFunction-physical: level of frailty, Groningen Frailty Indicator self-report V (15 items; primary outcome for complex care needs and frail clusters only), modified Katz scale (Katz-15) 15 items covering ADLs and IADLs; INTERMED for the Elderly Self Assessment (biopsychosocial and healthcare domains; primary outcome for the complex care needs group only); SM (self-management) scales x 2Function-cognitive: not reportedFunction-emotional: not reportedFunction-social: not reportedHealthcare use: not reportedSocial care use: not reported Healthcare costs: total costs, including health and care use, informal careSocial care costs: not reportedPatient satisfaction with care: not reported\n\n\nNotes: Time points measured: baseline and 12 monthsTime points reported: baseline and 12 monthsFunding: Netherlands Organisation for Health Research and DevelopmentEthical approval: Medical Ethical Committee of the University Medical Center of Groningen waived ethical approval ()Conflicts of interest: none declare\n\n",
        "objective": "To assess the effects of case management for integrated care of older people living with frailty compared with usual care.",
        "full_paper": "Objective\nTo evaluate the effects of the population-based, person-centred and integrated care service \u2018Embrace\u2019 at twelve months on three domains comprising health, wellbeing and self-management among community-living older people.\nMethods\nEmbrace supports older adults to age in place.\nA multidisciplinary team provides care and support, with intensity depending on the older adults\u2019 risk profile.\nA randomised controlled trial was conducted in fifteen general practices in the Netherlands.\nOlder adults (\u226575 years) were included and stratified into three risk profiles: Robust, Frail and Complex care needs, and randomised to Embrace or care as usual (CAU).\nOutcomes were recorded in three domains.\nThe EuroQol-5D-3L and visual analogue scale, INTERMED for the Elderly Self-Assessment, Groningen Frailty Indicator and Katz-15 were used for the domain \u2018Health.\u2019\nThe Groningen Well-being Indicator and two quality of life questions measured \u2018Wellbeing.\u2019\nThe Self-Management Ability Scale and Partners in Health scale for older adults (PIH-OA) were used for \u2018Self-management.\u2019\nPrimary and secondary outcome measurements differed per risk profile.\nData were analysed with multilevel mixed-model techniques using intention-to-treat and complete case analyses, for the whole sample and per risk profile.\nResults\n1456 eligible older adults participated (49%) and were randomized to Embrace (n(T0) = 747, n(T1) = 570, mean age 80.6 years (SD 4.5), 54.2% female) and CAU (n(T0) = 709, n(T1) = 561, mean age 80.8 years (SD 4.7), 55.6% female).\nEmbrace participants showed a greater\u2013but clinically irrelevant\u2013improvement in self-management (PIH-OA Knowledge subscale effect size [ES] = 0.14), and a greater\u2013but clinically relevant\u2013deterioration in health (ADL ES = 0.10; physical ADL ES = 0.13) compared to CAU.\nNo differences in change in wellbeing were observed.\nThis picture was also found in the risk profiles.\nComplete case analyses showed comparable results.\nConclusions\nThis study found no clear benefits to receiving person-centred and integrated care for twelve months for the domains of health, wellbeing and self-management in community-living older adults.\nIntroduction\nOlder adults prefer to remain living at home for as long as possible\u2013\u2018to age in place\u2019\u2013and to participate in society.\nHowever, this preference is compromised by age-related health problems, leading to an increasing level of dependency and service-use, a growing sense of loss of control and insecurity, and the threat of ultimate relocation to an institution.\nA major challenge is how to support older people to age in place in the face of increasing decline associated with ageing.\nThe current healthcare systems are insufficiently able to address these challenges for many ageing individuals and need to be reorganised in such a way that they promote ageing in place.\nA model of increasing importance and popularity in healthcare reform is the Chronic Care Model (CCM).\nThe CCM addresses the needs of chronically ill patients by offering comprehensive, person-centred, proactive and preventive care and support.\nIt encourages patients to be informed and activated, meaning that they should formulate personal goals and a plan to improve their health, and should have the motivation, information, skills, and confidence necessary to deal with the consequences of their diseases.\nTwo randomised controlled trials on the CCM targeted older adults, but both have limitations regarding the study populations included as only frail older adults with a limited health status were included and not those who were still healthy.\nIn order to provide care and support to the total community-living population of older adults, the CCM can be combined with a Population Health Management (PHM) model.\nPHM models assess an entire population in a community and not just those in need of urgent care.\nPHM-based care and support can be targeted to individual needs by classifying population subgroups into risk profiles.\nEmbrace is an integrated care service based on the complete CCM and a PHM model (Kaiser Permanente [KP] Triangle) targeting all community-living older adults.\nEmbrace\u2019s goal is to support older adults to age in place by providing person-centred, integrated, proactive and preventive care and support.\nEmbrace classifies older adults into three risk profiles based on the complexity of their care needs and their level of frailty.\nCare and support are tailored to the risk profile and the needs of the older adults.\nA qualitative study of Embrace has already shown that older adults felt safe, secure and more in control due to Embrace.\nIn this study, we intend to evaluate the effects of the population-based, person-centred and integrated care service Embrace on patient-reported outcomes at 12 months on three domains comprising health, wellbeing and self-management among community-dwelling older people.\nMethods\nStudy design and setting\nBetween October 2011 and March 2013, we conducted a randomised controlled trial (RCT) with stratification into three risk profiles based on the level of frailty and complexity of care needs and balanced allocation within general practitioner (GP) practices to the intervention (Embrace) or care as usual (CAU) groups.\nThe RCT was performed in three semi-rural municipalities in the province of Groningen (in the northern Netherlands).\nParticipants were followed for twelve months between January 2012 and March 2013.\nThe Medical Ethical Committee of the University Medical Center of Groningen has assessed the study proposal and concluded that approval was not required (Reference METc2011.108).\nThe study was performed in accordance with the tenets of the Declaration of Helsinki.\nAll participants gave informed consent.\nThe study protocol has been published previously.\nThe trial was registered at the Netherlands National Trial Register NTR3039 (http://www.trialregister.nl; see S1 and S2 Files).\nThe CONSORT statement was followed to report the findings and the checklist is available as supporting information (S3 File).\nStudy population and procedure\nFirst, we invited all GPs working in the three municipalities to participate in the study.\nRecruitment stopped after fifteen GPs\u2013proportionally distributed according to the size of the municipalities\u2013agreed to participate as they had enough eligible participants to obtain the sample size needed.\nNext, all older adults aged 75 and over who were registered with one of the participating GPs and were living at home or in a home for the elderly, were invited to participate.\nExclusion criteria at baseline were long-term admission to a nursing home (not just for rehabilitation), receiving an alternative type of integrated care and participating in another research study.\nEligible participants received a letter from their GP with general information about Embrace and the study.\nAfter having provided informed consent, participants completed self-report questionnaires at baseline (T0: October-December 2011) and twelve months after starting (T1: January-March 2013), with support by a family member, friend or volunteer if needed.\nWe sent reminders to non-respondents, followed by telephone calls to all persistent non-respondents.\nRespondents who submitted questionnaires with missing values were called by help desk assistants or visited by volunteers to complete the missing items.\nStratified randomisation and blinding\nWe first stratified participants into one of three risk profiles, using results of the baseline assessment of complexity of care needs (measured using the INTERMED for the Elderly Self-Assessment [INTERMED-E-SA]) and the level of frailty (measured using the Groningen Frailty Indicator [GFI]).\nThese risk profiles are \u2018Complex care needs\u2019 for participants with complex care needs and at risk for assignment to a hospital or nursing home (INTERMED-E-SA \u226516), \u2018Frail\u2019 for participants at risk of complex care needs (INTERMED-E-SA <16 and a GFI \u22655) and \u2018Robust\u2019 for participants at risk for the consequences of ageing (INTERMED-E-SA <16 and GFI <5).\nAfter stratification, we performed an anonymised and computerised balanced randomisation process within each GP practice.\nTherefore, participants were equally distributed within each GP practice to Embrace and CAU, taking into account predetermined patient characteristics deemed capable of affecting intervention outcomes, for example age, gender, number of chronic conditions and living situation.\nElderly Care Team members did not know if someone was randomised to CAU or had declined participation, but knew who was randomised to Embrace.\nParticipating older adults were informed in writing whether they were assigned to Embrace or CAU.\nData collectors (volunteers available when necessary for helping filling in questionnaires, and help desk assistants) were blinded for randomisation and stratification, as were the data analysts (SS and RU) until the point of data analysis.\nFor practical reasons, the data manager was not blinded.\nIntervention: Embrace\nEmbrace (in Dutch: SamenOud [ageing together]) is a person-centred and integrated care service for community-living older adults.\nA multidisciplinary Elderly Care Team\u2013consisting of the older adults\u2019 GP, a nursing home physician and two case managers (district nurse and social worker)\u2013provides care and support to older adults.\nThis care is in addition to care as usual.\nBefore starting the intervention, team members followed an intensive training program (three days for the GPs and nursing home physicians, eight days for the case managers) that focused on working according to the Embrace principles and methods.\nAlso, team members were coached during the intervention to support the cultural change in professionals\u2019 deep-rooted working patterns.\nThe intensity, focus, and individual or group approach of the care and support depended on the participant\u2019s risk profile.\nWe invited all participants to follow a self-management support and prevention program focusing on staying healthy and independent for as long as possible.\nThe program included regular Embrace community meetings, in which self-management abilities were encouraged and during which local healthcare and welfare organisations provided information on health maintenance, physical and social activities, and dietary recommendations.\nIn addition, frail people and those with complex care needs received individual support from a case manager.\nThey jointly developed an individual care and support plan targeting all health-related problems, which had to be agreed upon by the Elderly Care Team before implementation.\nThe case managers monitored changes in the medical, psychosocial, or living situation, and navigated the plan\u2019s delivery.\nThe Elderly Care Team discussed and evaluated the participants\u2019 health status and social situation in monthly meetings.\nIf necessary, they took proactive steps in dialogue with participants to prevent deterioration.\nPeople with a \u2018Robust\u2019 profile were encouraged to contact the team in the event of changes in their health or living situation.\nDetails of the implementation of Embrace have been published in the study protocol.\nCare as usual\nThe control group received care as usual as provided by their GPs and local health and community organisations.\nMunicipalities are in charge of social care, disease prevention and health promotion.\nOnce a health problem is found, patients enter the health care system\u2013in most cases with a visit to their GP.\nIn the Netherlands, GPs are family physicians who usually have a long-term relationship with their patients.\nThey act as gatekeepers for specialised services in the Dutch healthcare system: patients need a referral to enter specialised medical care.\nThe mean number of GP visits increases with age from six visits per year at age 45\u201364 to fifteen visits per year for people aged 75 years and older, and a regular GP visit takes about ten minutes.\nPatient-reported primary and secondary outcomes\nWe used eight different questionnaires to assess patient-reported outcomes in three domains: \u2018Health,\u2019 \u2018Wellbeing\u2019 and \u2018Self-management,\u2019 as these outcomes are important to ageing in place and to participation in society.\nPrimary and secondary patient-reported outcomes differed per risk profile, as we expected problems to vary per profile (see Table 1).\nHealth\nThe \u2018Health\u2019 domain included the outcomes \u2018Health status,\u2019 \u2018Complexity of care needs,\u2019 \u2018Level of frailty\u2019 and \u2018Limitations in Activities of Daily Living (ADL).\u2019\nWe measured Health status using the EuroQol-5D three-level version (EQ-5D-3L), which is a short self-report questionnaire measuring health in five dimensions in combination with a visual analogue scale (EQ-VAS).\nWe measured Complexity of care needs using the INTERMED-E-SA, which includes twenty questions in the biological, psychological, social and healthcare domains.\nWe measured Level of frailty in the physical, social, cognitive and psychological domains with the GFI self-report version (fifteen items).\nWe measured Limitations in ADL using the Katz-15, which measures independence in six physical ADLs (PADL), seven instrumental ADLs (IADL) and two additional ADL items.\nWe calculated ADL performance as the total number of disabilities.\nSubscale scores were calculated for PADL and IADL.\nWellbeing\nThe \u2018Wellbeing\u2019 domain included \u2018Wellbeing\u2019 and \u2018Quality of Life\u2019 (QoL).\nWellbeing was measured using the Groningen Well-being Indicator (GWI), covering eight sources of wellbeing in daily experiences: enjoying eating and drinking, sleeping and resting well, having good relationships and contacts, being active, managing oneself, being oneself, feeling healthy in body and mind, and living pleasantly.\nParticipants had to indicate whether each source of wellbeing was important to them and, if so, whether they were satisfied with that source.\nThe Well-being Satisfaction Score is the number of important sources divided by the number of satisfactory sources [unpublished manuscript].\nWe assessed QoL using two items derived from the self-perceived health questions of the RAND-36.\nThe first item measured self-rated QoL, while the second item compared the current self-rated QoL with QoL a year earlier.\nBoth questions are rated on a 5-point scale ranging from 1 to 5.\nSelf-management\nThe \u2018Self-management\u2019 domain included \u2018Self-management ability\u2019 and \u2018Self-management knowledge and behaviour\u2019.\nWe assessed Self-management ability using the Self-Management Ability Scale (SMAS-30) version 2, which contains thirty items and six subscales.\nThe total SMAS score was calculated as the average of the subscale scores.\nWe measured Self-management knowledge and behaviour with the culturally adapted and validated version of the Partners in Health scale (PIH): the PIH scale for older adults (PIH-OA).\nThe PIH-OA includes three subscales measuring eight items on an 8-point scale.\nOriginally, we defined the PIH as a secondary outcome measurement for quality of care.\nHowever, the new, adapted version\u2013PIH-OA\u2013measures self-management and is therefore included in the present study.\nAdaptations to the trial protocol\nWhen effects are found on an outcome measurement, follow-up analyses will be performed using the subscales of that particular measurement instrument\u2013if applicable.\nSample size\nWe used the primary outcome Health status (EQ-VAS) to calculate the sample size needed.\nWe considered a change in outcome of six points (SD 14 points) on the EQ-VAS of participants in the smallest sample, i.e. the risk profile \u2018Frail,\u2019 clinically relevant.\nWith a power of 80% (\u03b1 = 0.05, two-sided), a total number of 1062 older adults had to be included in the analysis.\nTaking into account an estimated non-response rate of 30% and a loss-to-follow-up rate of 30%, 2178 patients had to be invited to participate.\nStatistical analyses\nDifferences between respondents and non-respondents were tested using Chi-square tests for categorical variables and t-tests for continuous variables.\nDifferences in reasons for dropout in the intervention and control groups were tested using Chi-square tests.\nWe assessed differences in change between the intervention and control groups using multilevel mixed-effects analyses with regression coefficients (B) with 95% confidence intervals (CI) at \u03b1 = 0.05 (two-sided), with adjustment for age and sex.\nIndividual measurements (difference score per outcome, calculated as the difference between the T0 score and T1 score) were included as the first level and GP practices as the second level.\nWe estimated the clinical relevance of the effects using Cohen\u2019s effect sizes (ES) for statistically significant differences (p<0.05), with an ES of \u22650.20 reflecting a clinically relevant difference.\nWe performed intention-to-treat (ITT) analyses for the whole sample and per profile.\nMissing data were imputed at item level by multiple imputation techniques, with the fully conditional specification approach\u2013which uses the Bayesian framework.\nVariables group, risk profile, GP, sex, age, marital status, living situation, educational level, income and receiving help with completing the questionnaire were used as covariates of the missing predictor models, generating twenty imputed data sets.\nMissing scale scores due to loss to follow-up were imputed using the mean change in deterioration of completed cases, as we assumed that older adults deteriorate over time.\nThis process was performed per risk profile for each scale.\nITT outcomes were compared with those of complete case analyses including participants having both T0 and T1 measurements.\nWe performed all analyses using SPSS Statistics version 23.0 and used Mplus version 7.1 to impute the data.\nResults\nParticipants\nFig 1 presents the flow of participants in the study.\nWe included 1456 of the 2988 eligible older adults in the study and analyses (48.7%).\nThe main reasons for non-participation included poor health or having a partner with poor health, good health, questionnaire length and lack of interest.\nNon-respondents differed from respondents (all p-values <0.01) regarding gender (more women declined to participate), age (oldest older adults consented less often) and degree of urbanisation (more older adults living in rural areas declined to participate) (S1 Table).\nTable 2 shows the baseline characteristics of participants.\nThere were no statistically significant differences in the baseline characteristics between Embrace and CAU.\nAfter twelve months, 570 (76.3%) Embrace recipients and 561 (79.1%) CAU recipients completed the follow-up questionnaire.\nDropouts (Embrace n = 177,23.7%; CAU n = 148, 20.9%) were significantly (all p-values <0.01) older, more frail, with more complex care needs and with poorer health.\nThere were no significant differences in attrition rates between Embrace and CAU for the whole sample and per profile.\nDifferences in effects between Embrace and CAU\nWhole sample\nWe found no clear beneficial effects of Embrace in the whole sample as compared to CAU.\nRegarding the Health domain, Embrace participants showed a significantly greater deterioration in ADL (p = 0.047, ES = 0.10) and PADL performance (p = 0.011, ES = 0.13) compared to CAU\u2013although these effect sizes indicated not clinically relevant changes.\nWe found no differences in the changes observed between Embrace and CAU regarding Wellbeing outcomes (p>0.05, ES\u22640.05.\nRegarding Self-management, Embrace participants showed a significantly greater improvement in the \u2018Knowledge domain of self-management knowledge and behaviour\u2019 compared to CAU, but this difference did not reach clinical relevance (p = 0.009, ES = 0.14) Table 3 and S2 Table).\nComplex care needs\nWe found no significant differences in the changes observed in the domains of Health and Wellbeing after twelve months between Embrace and CAU.\nHowever, there was a significant and clinically relevant difference in change in the Self-management outcomes \u2018Self-management abilities,\u2019 \u2018Self-efficacy beliefs\u2019 and \u2018Investment behaviour\u2019, as Embrace participants performed worse after twelve months, whereas those in CAU showed a small improvement (Table 3 and S3 Table).\nFrail\nWe found no significant differences in the change observed between Embrace and CAU regarding Health and Wellbeing, but Embrace participants did show a significantly greater improvement in the \u2018Self-management knowledge and behaviour\u2019 Self-management outcome, as well as in its \u2018Knowledge\u2019 domain, compared to a deterioration for those in CAU (Table 3 and S4 Table).\nRobust\nWe found no significant differences in the Health domain, except for significantly worse ADL performance compared to CAU\u2013although this difference was not clinically relevant.\nFurthermore, Embrace participants showed a significantly larger deterioration in the Wellbeing outcome \u2018QoL comparison item\u2019 compared to CAU, but this difference was not clinically relevant either.\nWe found no differences in the changes observed between groups regarding Self-management (Table 3 and S5 Table).\nMissing data and sensitivity analyses\nMissing scale scores ranged from 0.0% to 12.7%, with 37 of the 42 scales and subscales having less than 5.0% missing values (Table 3).\nSensitivity analyses with complete cases showed the same pattern of results, except for 1) a significant deterioration in PADL performance of the complex Embrace participants, and 2) a no longer significant\u2013but still clinically relevant\u2013improvement on the total PIH-OA score for the frail Embrace participants (S6\u2013S10 Tables).\nDiscussion\nThis RCT examined the effects of \u2018Embrace,\u2019 a person-centred and integrated care service for older adults, for the total sample and by respective risk profiles.\nWe found no clear clinically relevant changes after receiving twelve months of care and support by Embrace on health, wellbeing and self-management in the total sample of community-living older adults and neither in the risk profiles.\nOverall, Embrace participants showed a greater\u2013but clinically irrelevant\u2013improvement in self-management knowledge and a greater\u2013but clinically irrelevant\u2013deterioration in ADL compared to CAU.\nThis heterogeneous picture was also found in the risk profiles.\nInterpretation of findings\nThe care and support offered by Embrace had fewer beneficial effects\u2013and sometimes even unbeneficial effects\u2013on the domains of Health, Wellbeing and Self-management than we anticipated, which confirms the heterogeneous outcomes previously reported in RCTs on integrated care programs for community-living older adults.\nOur finding of no clear benefits for Embrace on the outcomes measured could be due to the duration of the intervention, the nature of the intervention, the selection of outcomes or methodological limitations.\nFirstly, the intervention may not have worked or may not yet have worked.\nWe may have been dealing with an investment effect, as this multifaceted and complex intervention requires a cultural change in professionals\u2019 deep-rooted working patterns, which could take more time than only twelve months despite an intensive training and coaching program before and during the intervention.\nAssessment among participating professionals of whether the care and support provided was in accordance with the Chronic Care Model underlined this assumption.\nWe found a clinically relevant increase in the perceived level of implementation of integrated care from a \u2018basic level\u2019 at the start to a \u2018reasonably good level\u2019 after twelve months\u2013indicating clinically relevant improvements with room for further improvement.\nEvaluation of effects in the longer term is therefore needed, as well as follow-up coaching for further support of the cultural change in professionals\u2019 working behaviour and evaluation of protocol adherence.\nSecondly, the contrast between our intervention and CAU may have been too small to detect differences over the first twelve-month period.\nThe Dutch healthcare system is already of a quite high standard, as all inhabitants have health insurance and healthcare is easily accessible, leaving little room for improvement.\nThis was confirmed by our finding that only the frail Embrace participants showed a significant increase in self-management knowledge and behaviour.\nThese participants had received little or no care before the start of the intervention, in contrast with the complex participants, the majority of whom already received home care.\nThirdly, we had to deal with the heterogeneity and instability of the older population, which increased measurement error and thus reduced the likelihood of observing effects.\nFourthly, the measurement instruments for health and wellbeing may not have been specific enough for this type of intervention and may not have been sensitive enough to detect changes in clinical practice.\nThis could explain why we did find effects on two specifically developed measurement instruments: the PIH-OA, which is a version of the PIH for the evaluation of self-management knowledge and behaviour in older adults, and the PAIEC, which is used in another Embrace study for evaluation of perceived quality of integrated care and support.\nStrengths and limitations\nThe strengths of this study are its design\u2013a RCT targeting all community-living older adults\u2013and its stratification of participants into risk profiles, thereby enabling professionals to provide person-centred care and support.\nMoreover, we were able to perform predefined subgroup analyses to examine the effect of integrated care in subgroups at a higher risk of deterioration.\nWe must also acknowledge some potential limitations.\nWe randomised within GP practices, which increased the risk of contamination.\nAlthough we instructed GPs to provide care as usual to patients who were not assigned to the intervention, we may have underestimated the effect on CAU participants.\nHowever, regular GP visits are brief and only take about ten minutes, with little time to discuss the topic of concern\u2013let alone other health-related topics.\nMoreover, CAU participants did not receive any additional support that was part of the intervention.\nFurthermore, a potential limitation is the non-response rate at baseline of about 50%.\nThe differences between respondents and non-respondents concerning gender, age and degree of urbanisation may limit the generalisability of our findings to some extent.\nImplications for practice, policy and research\nThe present study showed that receiving twelve months of integrated care has no clear beneficial effect on patient-reported outcomes.\nBased on these results, the implementation of integrated care services for older adults cannot be recommended.\nHowever, a parallel study on Embrace showed that perceived quality of care improved.\nMoreover, in a qualitative study of Embrace, older adults indicated that they felt safe and secure due to Embrace care and support.\nThese results could contribute to decision-making and show the need for mixed method evaluations.\nMixed method evaluation could also offer an explanation for the absence of clear effects in the present study.\nFurthermore, future research should focus on the long-term effects of Embrace and should use outcomes\u2013for example on dependency, age-related fears and coping\u2013and specifically developed measurement instruments appropriate for this older population and type of intervention.\nA future cost-effectiveness study could help policy makers and professionals decide whether to implement Embrace.\nThe effects of Embrace should also be evaluated in other geographical areas and in other cultures with different healthcare systems.\nFinally, stratification into risk profiles was the starting point for delivering care and support at a suitable care intensity level.\nFuture studies could also target different risk profiles.\nConclusion\nThe present study showed that receiving twelve months of person-centred and integrated care and support from Embrace has no clear beneficial effect on patient-reported health status and neither on wellbeing and self-management outcomes.\nFuture research should provide insight into the long-term effects of Embrace.\nMoreover, specifically developed measurement instruments suitable for the target population and intervention should be used in future studies.\nAs this is the first CCM-based RCT to include a population-based sample of community-living older adults, it contributes to the design of future research on population-based integrated care.\nCONSORT flow diagram of the Embrace study.\n\nPrimary and secondary measurement instruments per risk profile.\n | Complex care needs | Frail | Robust\n | Primary | Secondary | Primary | Secondary | Primary | Secondary\nHealth |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0EQ-5D-3L | X |  | X |  | X | \n\u00a0\u00a0\u00a0\u00a0INTERMED-E-SA | X |  |  | X |  | X\n\u00a0\u00a0\u00a0\u00a0GFI | X |  | X |  |  | X\n\u00a0\u00a0\u00a0\u00a0Katz-15 |  | X |  | X |  | X\nWellbeing |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0GWI |  | X |  | X |  | X\n\u00a0\u00a0\u00a0\u00a0QoL |  | X |  | X |  | X\nSelf-management |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0SMAS-30 |  | X | X |  | X | \n\u00a0\u00a0\u00a0\u00a0PIH-OA |  | X | X |  | X | \n\nEQ-5D-3L = EuroQol-5D-3L including the EuroQol visual analogue scale; GFI: Groningen Frailty Indicator; GWI = Groningen Well-being Indicator; INTERMED-E-SA = INTERMED for the Elderly Self-Assessment; PIH-OA = Partners in Health scale for older adults; QoL = Quality of life; SMAS-30 = Self-Management Ability Scale version 2.\n\nBaseline characteristics of participants (n = 1456).\nValues are numbers (percentages) unless stated otherwise.\n\u00a0 | Whole sample | Complex care needs | Frail | Robust\n | (n = 1456) | (n = 365) | (n = 237) | (n = 854)\n | Embrace | CAU | Embrace | CAU | Embrace | CAU | Embrace | CAU\n\u00a0 | (n = 747) | (n = 709) | (n = 187) | (n = 178) | (n = 122) | (n = 115) | (n = 438) | (n = 416)\nAge in years, mean (SD) | 80.6 | (4.5) | 80.8 | (4.7) | 81.8 | (4.6) | 81.5 | (4.9) | 81.6 | (5.1) | 82.8 | (5.5) | 79.9 | (4.0) | 79.9 | (4.1)\nFemale | 405 | (54.2) | 394 | (55.6) | 121 | (64.7) | 115 | (64.6) | 82 | (67.2) | 80 | (69.6) | 202 | (46.1) | 199 | (47.8)\nWidowed/divorced/single | 320 | (42.8) | 290 | (41.0) | 87 | (46.5) | 79 | (44.4) | 77 | (63.1) | 72 | (63.2) | 156 | (35.6) | 139 | (33.5)\nIn sheltered accommodation/home for the elderly | 93 | (12.5) | 99 | (14.0) | 37 | (19.9) | 40 | (22.6) | 20 | (16.4) | 26 | (22.8) | 36 | (8.3) | 33 | (8.0)\nLow educational level1 | 370 | (49.9) | 374 | (53.4) | 106 | (57.0) | 116 | (66.3) | 66 | (54.1) | 69 | (60.0) | 198 | (45.7) | 189 | (46.0)\nLow income2 | 261 | (44.1) | 231 | (42.4) | 80 | (54.1) | 77 | (54.2) | 53 | (55.8) | 51 | (54.8) | 128 | (36.7) | 103 | (33.2)\nNumber of chronic conditions, median (IQR)\u00a0 | 2 | (1\u20133) | 2 | (1\u20133) | 3 | (2\u20135) | 3 | (2\u20135) | 3 | (1\u20134) | 3 | (2\u20134) | 1 | (1\u20132) | 1 | (1\u20132)\nReceiving home care | 89 | (12.1) | 69 | (9.8) | 47 | (26.4) | 42 | (23.9) | 24 | (20.0) | 14 | (12.4) | 18 | (4.1) | 13 | (3.2)\nReceiving help with filling in the questionnaire | 243 | (32.8) | 245 | (35.0) | 99 | (53.8) | 106 | (60.2) | 48 | (39.3) | 43 | (37.7) | 96 | (22.1) | 96 | (23.4)\n\nCAU = Care as usual; IQR = Interquartile range; SD = Standard deviation.\n1 Low: (Less than) primary school or low vocational training.\n2 Low: <\u20ac1350 per month.\nValues are based on complete data. There were no significant differences between CAU and Embrace\u2013neither for the whole sample nor per risk profile. This was tested using independent t-tests for continuous variables, Chi-square tests for categorical variables, and Mann-Whitney U tests for non-normally distributed continuous variables and ordinal variables.\n\nPatient-reported outcomes at 12-month follow-up in the Embrace study: Overview of the results of the intention-to-treat multilevel analyses for the whole sample and per risk profile.\n |  |  | Whole sample | Complex care needs | Frail | Robust\n |  |  | (n = 1456) | (n = 365) | (n = 237) | (n = 854)\n |  |  | Embrace | CAU |  |  | Embrace | CAU |  |  | Embrace | CAU |  |  | Embrace | CAU |  | \n | Scale scores(range) | Higher score* | Mean change | Mean change | p-value\u2020 | ES | Mean change | Mean change | p-value\u2020 | ES | Mean change | Mean change | p-value\u2020 | ES | Mean change | Mean change | p-value\u2020 | ES\nHealth | \u00a0 | \u00a0 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  | \nEQ-5D-3L | -0.33\u20131.00 | + | 0.00 | 0.00 | 0.670 | 0.02 | -0.02 | -0.01 | 0.521 | 0.07 | -0.02 | 0.0 | 0.223 | 0.16 | 0.01 | 0.01 | 0.630 | 0.03\nEQ-VAS | 0\u2013100 | + | -0.5 | -0.6 | 0.878 | 0.01 | -0.1 | 1.6 | 0.323 | 0.10 | -1.7 | -3.0 | 0.387 | 0.11 | -0.4 | -0.9 | 0.511 | 0.05\nINTERMED-E-SA | 0\u201360 | - | -0.1 | -0.2 | 0.597 | 0.03 | -1.9 | -2.6 | 0.149 | 0.15 | 1.4 | 1.3 | 0.608 | 0.06 | 0.3 | 0.5 | 0.540 | 0.04\nGFI | 0\u201315 | - | 0.1 | 0.1 | 0.998 | 0.00 | 0.1 | 0.0 | 0.552 | 0.06 | -0.6 | -0.7 | 0.586 | 0.07 | 0.4 | 0.5 | 0.411 | 0.06\nKatz-151 | 0\u201315 | - | 0.35 | 0.19 | 0.047 | 0.10 | 0.58 | 0.33 | 0.204 | 0.13 | 0.28 | 0.39 | 0.660 | 0.06 | 0.26 | 0.08 | 0.035 | 0.14\n\u00a0\u00a0\u00a0\u00a0PADL | 0\u20136 | - | 0.14 | 0.06 | 0.011 | 0.13 | 0.32 | 0.14 | 0.058 | 0.20 | 0.14 | 0.10 | 0.561 | 0.08 | 0.07 | 0.01 | 0.089 | 0.12\n\u00a0\u00a0\u00a0\u00a0IADL2 | 0\u20137 | - | 0.19 | 0.13 | 0.185 | 0.07 | 0.27 | 0.16 | 0.363 | 0.10 | 0.11 | 0.25 | 0.355 | 0.12 | 0.18 | 0.08 | 0.063 | 0.13\nWellbeing | \u00a0 | \u00a0 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  | \nGWI SF Score3 | 0\u20131 | + | -0.02 | -0.02 | 0.892 | 0.01 | -0.02 | -0.03 | 0.512 | 0.07 | -0.04 | 0.0 | 0.478 | 0.09 | -0.02 | -0.02 | 0.900 | 0.01\nQoL general | 0\u20135 | - | 0.08 | 0.10 | 0.636 | 0.02 | 0.17 | 0.14 | 0.587 | 0.06 | 0.12 | 0.09 | 0.818 | 0.03 | 0.03 | 0.08 | 0.289 | 0.07\nQoL vs 1 year ago | 0\u20135 | - | 0.08 | 0.04 | 0.320 | 0.05 | -0.04 | 0.01 | 0.471 | 0.08 | 0.11 | 0.17 | 0.425 | 0.10 | 0.13 | 0.02 | 0.018 | 0.16\nSelf-management | \u00a0 | \u00a0 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  | \nSMAS-30\u00a0 | 0\u2013100 | + | -1.1 | -0.8 | 0.411 | 0.04 | -2.0 | 0.2 | 0.015 | 0.26 | -0.4 | -0.7 | 0.705 | 0.05 | -0.9 | -1.2 | 0.664 | 0.03\n\u00a0\u00a0\u00a0\u00a0INIT | 0\u2013100 | + | -2.3 | -2.5 | 0.709 | 0.02 | -2.8 | -2.1 | 0.530 | 0.07 | -1.7 | -2.3 | 0.658 | 0.06 | -2.2 | -2.8 | 0.485 | 0.05\n\u00a0\u00a0\u00a0\u00a0SE | 0\u2013100 | + | -0.8 | -0.9 | 0.455 | 0.04 | -2.1 | 1.7 | 0.020 | 0.24 | 0.0 | -1.3 | 0.619 | 0.07 | -0.4 | -2.0 | 0.585 | 0.04\n\u00a0\u00a0\u00a0\u00a0INVEST | 0\u2013100 | + | -1.1 | 0.0 | 0.802 | 0.01 | -1.3 | 0.8 | 0.005 | 0.30 | -0.3 | 0.5 | 0.412 | 0.11 | -1.2 | -0.4 | 0.068 | 0.13\n\u00a0\u00a0\u00a0\u00a0POSITIVE | 0\u2013100 | + | -0.2 | 0.2 | 0.542 | 0.03 | -0.2 | 1.2 | 0.217 | 0.13 | -0.3 | 0.5 | 0.680 | 0.05 | -0.1 | -0.3 | 0.835 | 0.01\n\u00a0\u00a0\u00a0\u00a0MULT | 0\u2013100 | + | -0.8 | -0.4 | 0.124 | 0.08 | -1.9 | 1.1 | 0.126 | 0.16 | -1.1 | -1.7 | 0.609 | 0.07 | -0.2 | -0.6 | 0.383 | 0.06\n\u00a0\u00a0\u00a0\u00a0VAR | 0\u2013100 | + | -1.3 | -0.8 | 0.461 | 0.04 | -3.2 | -1.3 | 0.177 | 0.14 | 1.2 | 0.3 | 0.450 | 0.10 | -1.2 | -0.8 | 0.649 | 0.03\nPIH-OA4 | 8\u201364 | + | 0.8 | 0.4 | 0.285 | 0.06 | 1.1 | 1.1 | 0.976 | 0.00 | 1.7 | -0.8 | 0.020 | 0.31 | 0.4 | 0.4 | 0.936 | 0.01\n\u00a0\u00a0\u00a0\u00a0Knowledge | 2\u201316 | + | 0.8 | 0.3 | 0.009 | 0.14 | 0.8 | 0.3 | 0.113 | 0.17 | 1.0 | -0.2 | 0.015 | 0.32 | 0.7 | 0.4 | 0.245 | 0.08\n\u00a0\u00a0\u00a0\u00a0Management | 2\u201316 | + | 0.1 | 0.0 | 0.691 | 0.02 | 0.2 | 0.2 | 0.969 | 0.00 | 0.2 | -0.2 | 0.398 | 0.11 | -0.1 | -0.1 | 0.965 | 0.00\n\u00a0\u00a0\u00a0\u00a0Coping | 4\u201332 | + | 0.0 | 0.1 | 0.659 | 0.02 | 0.1 | 0.6 | 0.336 | 0.10 | 0.6 | -0.4 | 0.119 | 0.21 | -0.2 | 0.0 | 0.355 | 0.06\n\nCAU = Care as usual; EQ-5D-3L = EuroQol-5D-3L; EQ-VAS = EuroQoL-5D visual analogue scale; ES = Effect size d, thresholds <0.2 trivial, \u2265 0.2\u20130.5 small, \u22650.5\u20130.8 medium, \u2265 0.8 large; GFI = Groningen Frailty Indicator; GWI SF Score = Groningen Well-being Indicator Satisfaction Score; IADL = Instrumental Activities of Daily Living; INIT = Taking initiatives subscale; INTERMED-E-SA = INTERMED for the Elderly Self-Assessment; INVEST = Investment behaviour subscale; MULT = Multi-functionality of resources subscale; PADL = Physical Activities of Daily Living; PIH-OA = Partners in Health scale for older adults; POSITIVE = Positive frame of mind subscale; QoL = Quality of life; SE = Self-efficacy beliefs subscale; SMAS-30 = Self-Management Ability Scale version 2; VAR = Variety in resources subscale.\n* + Higher score means improvement;\u2014higher score means deterioration.\n\u2020 Values are corrected for age and sex; bold values indicate p<0.05.\n1 Percentage of missing items at baseline before imputation 7.4%\n2 Percentage of missing items at baseline before imputation 5.4% and 6.1% at follow-up.\n3 Percentage of missing items at baseline before imputation 12.7%\n4 Percentage of missing items at baseline before imputation 5.7%\nBold text and orange filling: Significant (p<0.05) or clinically relevant (ES \u22650.20) deterioration\nBold text and green filling: Significant (p<0.05) or clinically relevant (ES \u22650.20) improvement",
        "label": "low",
        "id": "task4_RLD_test_761"
    },
    {
        "paper_doi": "10.1186/1475-2875-11-153",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Trial design: A phase III, randomized, open label, multi-centre, comparative study.Follow-up: Limited physical exam on Days 1-2, 3, 7, 14, 12, 28, 35, and 42 and if clinically indicated. ECG 2 to 4 hours after drug administration and on follow-up Days 7, 28, and 42. Thick and thin smears eight hourly in Days 0 to 2, then Days 3, 7, 14, 21, 28, 35, and 42. Haematology and blood chemistry at Days 3, 7, and 28. Blood spot for PCR at Day 42 or failure. Urinalysis on Day 3. HCG for women on Days 0 and 28.Adverse event monitoring: \"adverse events collected each time\" and \"Safety was assessed through direct questioning, physical examinations, ECG abnormalities (prolongation QT-interval), and significant change from baseline clinical laboratory parameters [17]. Adverse events were followed up until the event had resolved.\"\n\n\nParticipants: Number: 401 randomized (153 P. falciparum only, 90 mixed, 158 P. vivax only).Inclusion criteria: adult, absence of severe malnutrition, axillary temperature > 37.5degC or a history of fever within the preceding 24 hours, asexual P. falciparum density 1000 to 200,000/mL, P. vivax and other malaria density >= 250/mL, able to take oral treatment, informed consent, uncomplicated P. falciparum or P. vivax mono-infection, or mixed infection.Exclusion criteria: severe vomiting, history or evidence of 'clinically systematic significant disorders', other febrile conditions, hypersensitivity or adverse reactions to antimalarials, history of use of any other antimalarial agent within four weeks of the start of the trial and confirmed by urine test, and pregnancy or lactating.\n\n\nInterventions: 1. Artemesinin-naphthoquine, fixed-dose combination, 250 mg/100 mg tablets (Arco, Kunming Pharmaceutical Corporation, China):4 tablets as a single dose2. DHA-P, fixed-dose combination, 40 mg/320 mg tablets (Duo-Cotecxin: Holey-Cotec Pharmaceutical Co. Ltd, China):Daily doses of dihydroartemisinin 2 to 4 mg/kg and piperaquine 16 to 32 mg/kgweight <= 60 kg 3 tablets each day for 3 daysweight > 60 kg 4 tablets each day for 3 daysAll doses supervised.\n\n\nOutcomes: ACPR at Day 42, PCR-adjusted and PCR-unadjustedGametocyte carriageAdverse eventsFever clearanceParasite clearance\n\n\nNotes: Country: IndonesiaSetting: Three Armed Forces hospitals in Jayapura (Marthen Indeys/Army, Soedibjo Sardadi/Navy, and Bhayangkara/Police Hospitals) and one public hospital in Maumere (St. Gabriel Hospital)Transmission: Not reportedResistance: Widely reported resistance of P. falciparum to chloroquine, sulphadoxine-pyrimethamine, and quinineDates: 2007 to 2008Funding: Kunming Pharmaceutical Corporatio\n\n",
        "objective": "To evaluate the efficacy and safety of the artemisinin\u2010naphthoquine combination for treating adults and children with uncomplicated P. falciparum malaria.",
        "full_paper": "Background\nA practical and simple regimen for all malaria species is needed towards malaria elimination in Indonesia.\nIt is worth to compare the efficacy and safety of a single dose of artemisinin-naphthoquine (AN) with a three-day regimen of dihydroartemisinin-piperaquine (DHP), the existing programme drug, in adults with uncomplicated symptomatic malaria.\nMethods\nThis is a phase III, randomized, open label using sealed envelopes, multi-centre, comparative study between a single dose of AN and a three-day dose of DHP in Jayapura and Maumere.\nThe modified WHO inclusion and exclusion criteria for efficacy study were used in this trial.\nA total of 401 eligible adult malaria subjects were hospitalized for three days and randomly treated with AN four tablets single dose on day 0 or DHP three to four tablets single daily dose for three days, and followed for 42 days for physical examination, thick and thin smears microscopy, and other necessary tests.\nThe efficacy of drug was assessed by polymerase chain reaction (PCR) uncorrected and corrected.\nResults\nThere were 153 Plasmodium falciparum, 158 Plasmodium vivax and 90 P. falciparum/P. vivax malaria.\nMean of fever clearance times were similar, 13.0\u2009\u00b1\u200910.3 hours in AN and 11.3\u2009\u00b1\u20097.3 hours in DHP groups.\nThe mean of parasite clearance times were longer in AN compared with DHP (28.0\u2009\u00b1\u200911.7 hours vs 25.5\u2009\u00b1\u200912.2 hours, p\u2009=\u20090.04).\nThere were only 12 PCR-corrected P. falciparum late treatment failures: seven in AN and five in DHP groups.\nThe PCR uncorrected and corrected on day \u221242 of adequate clinical and parasitological responses for treatment of any malaria were 93.7% (95% Cl: 90.3\u201397.2) and 96.3% (95% Cl: 93.6\u201399.0) in AN, 96.3% (95% Cl: 93.5\u201399.0) and 97.3% (95% Cl: 95.0\u201399.6) in DHP groups.\nFew and mild adverse events were reported.\nAll the abnormal haematology and blood chemistry values had no clinical abnormality.\nConclusion\nAN and DHP are confirmed very effective, safe and tolerate for treatment of any malaria.\nBoth drugs are promising for multiple first-line therapy policies in Indonesia.\nBackground\nMalaria remains a major public health problem.\nThe World Health Organization (WHO) estimated the number of reported cases from Indonesia were 2.5 million in 2006.\nMost cases were recorded from Papua and East Nusa Tenggara provinces, where about 300,000 and 70,000 clinical malaria cases were reported annually from its provinces.\nTo accelerate malaria control, one of the four key recommended interventions is to give appropriate anti-malarial drugs with artemisinin-based combination therapy (ACT) for patients with confirmed malaria.\nArtemisinin was discovered by Chinese scientists in the 1970s.\nArtemisinin derivatives are the most rapidly acting and efficacious anti-malarial drugs.\nThese drugs show rapid absorption and activity against many stages of Plasmodium, including trophozoites and early sexual forms (gametocytes).\nTheir short elimination half-life (<3.7 hours) protects them from resistance and their potent and broad specificity of action reduces gametocyte carriage and infectivity, and reduces the transmission of Plasmodium falciparum.\nTolerability of these drugs is also very good.\nThis makes artemisinin derivatives the ideal partner drugs.\nWHO recommends the use of artemisinins only in combination with another anti-malarial drug to prevent the occurrence of drug resistance and to address the issue of its relatively short half-life.\nACT is the best therapeutic option for treating drug-resistant malaria and retarding the development of resistance presently.\nSince 2004, the Indonesian Malaria Control Programme has chosen a non-fixed dose artesunate-amodiaquine (AA) as the programme drug for treatment of uncomplicated P. falciparum, which had widely reported resistant to chloroquine, sulphadoxine-pyrimethamine or quinine.\nAA is also effective for Plasmodium vivax, safe for all age groups and relatively cheap.\nThough combination of AA is palatable and given by single daily dose for three days, the compliance is poor because of the number of pills to be swallowed.\nThis ACT resulted a problem in its wide-scale implementation.\nMoreover, there was reported in vitro amodiaquine cross-resistance with chloroquine.\nConsequently, the efficacy of AA was varied (78\u201396%).\nTo overcome these problems, other ACT is needed for treatment of any malaria in Indonesia.\nFixed-dose combination of artemether-lumefantrine (AL) has been registered recently in Indonesia.\nThis ACT was safe and very effective with a cure rate of 95% for P. falciparum malaria, but its efficacy was modest (43%) for P. vivax.\nAL is not a practical regimen because it should be administered twice daily for three days and given with fatty food.\nMoreover, the cost is expensive, >10 US$ per treatment course.\nA better ACT that would be simple to use, effective and affordable for all types of malaria is needed to eliminate malaria in Indonesia.\nDihydroartemisinin-piperaquine (DHP) is a fixed-dose ACT and given single daily dose for three days.\nClinical trials for treatment of uncomplicated malaria proved more superior compared with the existing fixed-dose AL, and ACT programme AA in Papua.\nThis ACT is very effective with cure rates of \u226595% for all malaria and safe.\nDHP has been used widely as the first-line ACT for more than two years in Papua.\nIn addition, the cost of DHP is similar to AA per treatment course.\nDHP trials in other areas are needed to collect unexpected adverse events.\nArtemisinin-naphthoquine (AN) is a new fixed-dose ACT.\nNaphthoquine phosphate is an anti-malarial drug synthesized by Chinese Academy of Military Medical Science in late 1980s.\nThough naphthoquine has a similar structure to chloroquine, cross-resistance has not yet been reported.\nExisting pharmacological data indicate that naphthoquine is effective against erythrocytic phase of Plasmodium, even in chloroquine resistance cases.\nNaphthoquine has a longer half-life (276 hours), compared to chloroquine and mefloquine.\nIt has been shown to be effective against P. falciparum at doses of 12 mg/kg used alone or at 400 mg in combination with artemisinin for adult patients.\nNaphthoquine appears to be an ideal partner drug for artemisinin.\nAN is administered with a single dose of therapy and has few side effects.\nOf limited data, AN was reported safe and effective for both P. falciparum and P. vivax .\nIn a study in Chinese adult subjects, one dose of a fixed-dose artemisinin (1,000 mg)-naphthoquine (400 mg) for treatment of uncomplicated malaria showed a cure rate of 98% at day 28 in P. falciparum, and 90% at day 56 in P. vivax .\nMore studies should be done to confirm its efficacy, safety, and tolerability in an enlarge scale and another population.\nIndonesia has been committed to eliminate malaria by year 2030.\nA clinical trial of AN was conducted to improve compliance and find a practical and simple ACT for all malaria species.\nBased on the previous Indonesian ACT trials, DHP is the best ACT for treatment of uncomplicated malaria in multi-drug resistant areas.\nTherefore, a clinical trial of AN was compared to DHP for its efficacy and safety in Jayapura and Maumere.\nMethods\nTime and study location\nThe trial was carried out in 2007\u20132008 at four hospitals, three Armed Forces hospitals in Jayapura (Marthen Indeys/Army, Soedibjo Sardadi/Navy and Bhayangkara/Police Hospitals), and one public hospital in Maumere (St Gabriel Hospital).\nStudy design\nThe study was a phase III, randomized, open label, multi-centre, comparative of the efficacy, safety and tolerability of a single dose AN vs DHP in adults with uncomplicated symptomatic P. falciparum, P. vivax or mixed P. falciparum/P. vivax malaria.\nThis trial was approved in writing by the Ethics Committee of National Institute of Health Research and Development, Ministry of Health (No. LB.03.02/2/449/2007), and the Bureau of Food and Drug Control (No.PO.01.01.3.1.1682), Republic of Indonesia.\nSample size\nAssuming the failure rates of AN (P1) and DHP (P2) were 0.1% and 5.0% based on un-published African AN studies prior a trial and the WHO recommendation for choosing ACT programme.\nThe study estimated risk ratio of failure rate was 2% with \u03b1 (type I error) of 0.05 and power (1-\u03b2) of 80%.\nA minimum sample size was 166 subjects per treatment group, and adjusted 20% for follow-up losses and withdrawals.\nA total 401 subjects were recruited in this trial.\nThe formula for calculating the sample size as the following\nN = The sample size of each of the treatment groups; P1 = The failure rate in the Arco\u2122 group (0.1%); P2 = The failure rate in the Duo-Cotecxin\u2122 group (5.0%); R = The risk ratio of treatment failure (P1/P2\u2009=\u20090.02).\nProcedures\nThe population of this study was adult males and females aged 15\u201369 years, body weight 35\u201375 kg and presenting with acute, symptomatic, uncomplicated P. falciparum and/or P. vivax malaria.\nThey were recruited according to the modified WHO inclusion criteria (absence of severe malnutrition, axillary temperature of \u226537.5\u00b0C or history of fever in the last 24 hours, asexual P. falciparum density 1,000\u2013200,000/\u03bcl, P. vivax and other malaria density \u2265250/\u03bcl, and ability to swallow oral medication), and exclusion criteria [severe vomiting, history or evidence of clinically systematic significant disorders, other febrile conditions, hypersensitivity or adverse reactions to anti-malarials, history of use of any other anti-malarial agent within four weeks prior to start of the study and confirmed by urine test (Dill Glazko and Lignin tests), and pregnancy or lactating] for therapeutic efficacy study.\nSubject informed consent was requested prior the study.\nSubjects who withdrew early were not replaced.\nEligible subjects were blindly, randomly assigned equally to one of the two treatment groups using sealed envelopes.\nAN (Arco\u2122, Kunming Pharmaceutical Corporation with Chinese quality standards, one tablet contained 250 mg of artemisinin and 100 mg of naphthoquine) was administered four tablets single dose only.\nDHP (Duo-Cotecxin\u2122, Holey-Cotec Pharmaceutical Co.LTd, China, one tablet contained 40 mg of dihydroartemisinin and 320 mg of piperaquine) was administered three (body weight of \u226460 kg) to four tablets (body weight of >60 kg) single daily dose for three days based on dosage of dihydroartemisinin 2\u20134 mg/kg bw or piperaquine 16\u201332 mg/kg bw.\nSubjects were observed for one hour to ensure that the medications were not vomited.\nAll subjects were hospitalized for three days or until fever and parasite had cleared for at least 24 hours, returned to study site for follow-up at all scheduled visits to day 42, and they had additional primaquine for radical treatment on day 42.\nSubjects with treatment failure were withdrawn from the study and given a rescue malaria treatment.\nThey had no study investigations performed thereafter.\nClinical and laboratory assessments\nAll eligible subjects had medical history and detail demography completed at enrolment.\nA full physical examination, electrocardiography (ECG) and laboratory tests (malaria microscopy, haematology, blood chemistry, PCR genotyping and urinalysis) were performed at baseline (day 0, prior to dose).\nLimited physical examination was performed during hospitalization (days 1\u20132) and on follow-up days (3, 7 14, 21, 28, 35, and 42) and if clinically indicated as well as adverse events collected each time.\nA 12-lead resting ECG was obtained approximately 2 to 4 hours after study drug administration on day 0\u20132, and follow up days 7, 28 and 42.\nThick and thin smears were examined at screening, days 0\u20132 hospitalization: eight hourly, days 3, 7, 14, 21, 28, 35, and 42; haematology and blood chemistry at days 0, 3, 7, and 28; and blood spot for PCR at days 0 and 42 or failure.\nUrinalysis was assessed on days 0 and 3.\nHCG was tested for women of potential pregnancy at screening and day 28.\nMicroscopy results were blind cross-checked by certified microscopists, and treatment failures were corrected by PCR.\nPCR was performed for speciation of plasmodium and genotype of P. falciparum.\nThere were 3 loci genotype of P. falciparum tested (MSP1, MSP2 and GLURP).\nThe primary (P) and nested (N) primers are as following : MSP1 (P1: 5\u2032CAC ATG AAA GTT ATC AAG AAC TTG TC3\u2032, P2: 3\u2032GTA CGT CTA ATT CAT TTG CAC G5\u2032; N1: 5\u2032GCA GTA TTG ACA GGT TAT GG3\u2032, N2: 3\u2032GAT TGA AAG GTA TTT GAC5\u2032); MSP2 (P1: 5\u2032GAA GTT AAT TAA AAC ATT GTC3\u2032, P2: 3\u2032GAG GGA TGT TGC TGC TCC ACA G5\u2032, N1: 5\u2032CTA GAA CCA TGC ATA TGT CC3\u2032, N2: 3\u2032GAG TAT AAG GAG AAG TAT G5\u2032) and GLURP (P1: 5\u2032ACA TGC AAG TGT TGA TCC3\u2032, P2: 3\u2032GAT GGT TTG GGA GTA ACG5\u2032, N1: 5\u2032TGA ATT CGA AGA TGT TCA CAC TGA AC3\u2032, N2: 3\u2032TGT AGG TAC CAC GGG TTC TTG TGG5\u2032).\nTo date, there are no recommended markers to distinguish recrudescence, relapse and new infection of P. vivax malaria from P. vivax treatment failures.\nEfficacy assessment\nThe efficacy of AN and DHP was assessed in intent-to-treat (ITT), modified ITT, and evaluable or per-protocol (PP) population.\nITT population included all randomized subjects who had received any amount of study medication.\nModified ITT population only included correctly randomized subjects in analysis, and excluded wrongly randomized and those lost to follow-up.\nPP analysis consisted only of the efficacy evaluable (EE) subjects defined according to the 2003 WHO criteria and constituted as PP population that did not include subjects who failed to comply with per-protocol.\nThe efficacy was a proportion of subjects with PCR-corrected adequate clinical and parasitological response (ACPR) at day 42.\nACPR was defined as the absence of asexual parasitaemia on day 42 irrespective of the temperature and not meeting any of criteria of early treatment failure (ETF) or late clinical or parasitological failure (LCF or LPF), or as subjects with clearance of asexual parasitaemia within 42 days of initiation of study treatment.\nSubjects classified as failures by clinical and parasitological criteria were considered ACPR if the PCR analysis showed a new infection (all the alleles in parasites from the failure-treatment sample were different from those in the admission sample, for one or more loci tested) rather than a recrudescence .\nRecrudescence was defined as reappearance of asexual parasites of the same isolate as initial infection with or without clinical signs, after initial clearance of parasites from the peripheral blood with positive blood smear and PCR confirmation of the same isolate (presence of at least one matching alleles).\nThe early and late failures were classified according to the 2003 WHO guidelines.\nThe total treatment failure was defined as the sum of early and late treatment failures.\nSafety assessment\nThe safety population was defined as all randomized subjects who had received any amount of study medication.\nSafety was assessed through direct questioning, physical examinations, ECG abnormalities (prolongation QT- interval), and significant change from baseline clinical laboratory parameters.\nAdverse events were followed up until the event had resolved.\nData analysis\nData were double entered and validated including microscopic validation and PCR corrected treatment failure data using EpiData 3.02, and analysed using SPSS for Windows version 15.\nThe Mann\u2013Whitney U-test was used for non-parametric comparisons, and Student\u2019s t-test for parametric comparisons.\nProportions were examined using \u03c72 with Yates\u2019 correction or by Fisher\u2019s exact-test.\nThe efficacy was assessed by survival analysis in which the cumulative risk of failure was calculated by the Kaplan Meier product limit formula.\nResults\nBaseline characteristics of study subjects\nIn Jayapura, over 3,000 clinical malaria cases had been screened, only 301 could be enrolled for this trial, and 151 cases were treated with AN and 150 treated with DHP.\nIn Maumere, of a total 154 screened clinical malaria, only 100 cases could participate, 50 were treated with AN and the other 50 treated with DHP.\nThe Armed Forces Hospitals contributed to 75% sample size, so were mostly male subjects.\nThere were seven subjects had weight >75 kgs (four in AN and three in DHP), 56% with axillary temperature of \u226537.5\u00b0C.\nThe characteristics of study subjects in both treatment groups were not different (Table 1).\nThe clinical symptoms of subjects in AN and DHP groups were also not different.\nFever, nausea, headache and rigors were the common symptoms documented in this study.\nOther classical symptoms are shown on Figure 1.\nThe initial laboratory findings of haematology, blood chemistry, and parasitology showed no differences between AN and DHP groups (Table 2).\nThough some of the study subjects had abnormal values, only a few had significant clinical abnormalities.\nThe distribution of malaria subjects with anaemia (Hb\u2009<\u200911 g/dL), thrombocytopaenia (platelet\u2009<\u2009150,000/ul) or leucocytosis (>10,000/ul) were almost similar between the treatment groups (52.2% vs 47.8%, 71.6% vs 73%, 6% vs 5%).\nOverall, only pallor was documented as a significant clinical abnormality with abnormal values of haematology.\nSome study subjects had higher or lower values of blood chemistry, such as alanine aminotransferase (ALT) (23.4%), aspartate aminotransferase (AST) (25.9%), bilirubin (31.4%), albumin (44.6%), urea (5.7%), sodium (23.1%), potassium (25.9%), creatinine (28.2%), and chloride (32.6%).\nOnly jaundice was documented as a significant clinical abnormality with abnormal values of bilirubin.\nThere were 153 Plasmodium falciparum (Pf), 158 Plasmodium vivax (Pv) and 90 P.falciparum/P.vivax malaria (Figure 2).\nGametocyte carriages were detected 34.4% (52 of 151) in P. falciparum, 94.3% (148 of 157) in P. vivax and 82.2% (74 of 90) in mixed P.falciparum/P. vivax, respectively.\nThe range of gametocyte densities was 1\u20132,697/ul (Table 2).\nAnalysed population\nThe ITT (401), modified ITT (384) and PP (378) population in each treatment group were microscopically cross-checked.\nThere were three of 401 study subjects not eligible after cross-checking.\nAll protocol violation cases were in AN group (two P. falciparum cases having asexual parasitaemia <1,000/ul, and one P. vivax case had taken an anti-malarial drug/chloroquine prior to the study).\nWhile all three withdrew consent, the cases were in DHP group (one parasitaemic P. falciparum case on day 0, one withdrawn P. vivax case by family of subject on day 0, and the one P. vivax case felt discomfort and dizzy by day 4).\nDuring the follow up, 17 cases were documented lost to follow up, seven (two P. falciparum cases on days 14 and 35; three P. vivax cases on days 3, 21, and 22; and two mixed P. falciparum/P. vivax cases on days 7 and 42) in AN group, and 10 [five P. falciparum cases on days 3, 7, 15, 20, and 21; four P. vivax cases on days 3 (two), 21, and 22; and one mixed P. falciparum/P. vivax case on day 3)] in DHP group ( 2 and Table 3).\nTherapeutic efficacy\nBoth study drugs had rapid fever clearance.\nOver 90% of study subjects became afebrile by the first 16 hours after the first dose of treatment, and all cleared in 56 hours (Figure 3).\nThe mean of fever clearance times (FCTs) were 13.0\u2009\u00b1\u200910.3 hours in AN and 11.3\u2009\u00b1\u20097.3 hours in DHP groups, and did not significantly differ.\nAlmost all hospitalized study subjects (>80%) were asymptomatic when discharged.\nDuring study follow-up, only mild reported symptoms (headache, dizzy, cough, abdominal pain, myalgia, sleeping disturbance, and fatigue) resolved with or without a simple symptomatic treatment.\nMost subjects (>90%) had cleared asexual parasitaemia by day 1\u201316 hours (Figure 4).\nAN had longer mean of parasite clearance time (PCT: 28.0\u2009\u00b1\u200911.7 hours) compared with DHP groups (25.5\u2009\u00b1\u200912.2 hours) (p\u2009=\u20090.04).\nOverall, mean of PCT of these ACTs was 26.7 (8\u201372) hours.\nThere was 68.8% gametocyte carriages prior ACT treatment.\nHowever, the proportion of gametocyte carriages reduced by day of follow-up.\nIt became 28.7% and 25.8% in the first 24 hours post-treatment, and 18.6% and 17% by day 3 in AN and DHP groups, respectively.\nGametocyte carriage was still detected in 1.1% subjects on completion of the study on day 42.\nOf 401 randomized study subjects, there were 19 TFs, 12 in AN and seven in DHP groups.\nNo ETF was reported.\nThere were six documented as LCFs and 13 as LPFs (Table 3).\nAll the TF cases were from Jayapura, four P. falciparum, three P. vivax and 12 P. falciparum/P. vivax malaria.\nOnly three of 19 TFs occurred by day \u226428, two in the AN and one in the DHP group.\nPCR speciation of the 19 paired samples of TF cases showed three LPFs diagnosed as P. vivax (one in AN group by day 32 and two in DHP by day 35) detected as P. falciparum, and classified as protocol violation cases.\nThere were another two LPFs by days 35 and 42 diagnosed as P. falciparum, one LPF by day 35 as mixed P. falciparum/P. vivax, and another one LCF by day 42 as mixed P. falciparum/P. vivax detected as P. falciparum new infections in AN group,\nOf the four TFs detected as new infections, two TFs had been treated with quinine plus doxycycline and classified as protocol violations at day 35.\nThe PCR-corrected treatment outcomes are shown in Table 4.\nThe overall uncorrected and corrected PCR therapeutic efficacies of both ACT for any malaria were between 89% to 95% in ITT and modified ITT population, and 94% to 97% in PP population.\nDHP had slightly higher uncorrected and corrected PCR at day 42 ACPR (96.3% and 97.3%) compared to AN (93.7% and 96.3%) for treatment of any malaria (Tables 3 and 4).\nAll TFs\u2019 PCR-corrected differences were detected in >28 day.\nTherefore, the day 28 uncorrected and corrected ACPR were similar, 95.5% (192 of 201) and 93.0% (186 of 200) in AN and DHP ITT populations; 98.0% (192 of 196) and 97.9% (186 of 190) in AN and DHP modified ITT population; 99.0% (192 of 194) and 99.5% (186 of 187) in AN and DHP PP population.\nFigures 5, 6 and 7 show the survival curves of PCR-corrected cumulative risk of failures of AN and DHP for treatment of any malaria in ITT, modified ITT and PP population.\nThe hazard ratio of risk failures were no different between treatment groups in the three populations.\nSafety\nThere were no serious adverse events reported in malaria subjects treated with AN and DHP during the study.\nOnly few (<10%) and mild symptoms as adverse events were documented.\nThe common reported adverse events were headache, dizzy, and cough.\nThere were also no clinically significant effects on myocardial electrophysiology identified through a series of ECG examinations.\nIn both study groups, haemoglobin, haematocrit, red blood cell (RBC) and platelet counts were gradually increased to normal limit by day 28.\nMeans of haematocrit and RBC were slightly decreased on day 3, and 7, and became normal by day 28.\nIn contrast, means of platelet on days 3 and 7 were increased significantly and then slightly decreased to normal value by day 28.\nThe means of haematocrit, Hb, RBC, WBC and platelet at point of investigation were not statistical significant different between treatment groups.\nOnly few leucocytosis (5.5%) were found in this trial.\nInterestingly, there were 21.4% eosinophilia (eosinophil >3%) on day 0, and increased to 38.5% by day 28.\nAll the abnormal haematological values had no significant clinical abnormality.\nThere were mild liver impairment with or without mild renal impairment on day 0 however bilirubin, albumin, ALT, AST, creatinine, urea and electrolytes were gradually improved to normal limit by day 28 in malaria subjects treated with AN and DHP.\nThe means of these blood chemistry parameters were not different between treatment groups.\nDiscussion\nACT is recommended to be given for 3 days when given with slowly eliminated partner drug.\nIn a three-day regimen, the artemisinin component is present in body during only two asexual parasite life-cycles, except for Plasmodium malariae.\nIn each asexual cycle, artemisinin and its derivatives reduce parasite numbers by a factor of approximately 10,000.\nThis anti-malarial drug is a potent and rapidly acting blood schizontocide, and gametocytocide.\nIt is still rational to study one-day ACT with a new good partner drug to improve the compliance and have a simple and practical regimen.\nMoreover, the risk of the development of de novo resistance is increased by the greater time dividing asexual parasites are exposed to drugs.\nThis makes the long-term risk of resistance developing a concern for single dose ACT.\nTherefore, this study will be useful to confirm the findings of other previous studies with different types of one-day ACT (artesunate-amodiaquine, artesunate-sulphadoxine/pyrimethamine, or artesunate-mefloquine) and in different geographical settings.\nNaphthoquine is a tetra-aminoquinoline, synthetic blood schizontocide anti-malarial drug.\nThough this anti-malarial drug has a similar structure to chloroquine, cross-resistance with chloroquine has not yet been reported.\nOf the limited clinical studies in China, naphthoquine in combination with artemisinin given in a single dose, was effective and safe with cure rates of 97.5% for treatment of P. falciparum malaria by day 28, and 90.0% for the treatment of P. vivax malaria by day 56.\nThis preliminary findings is valuable data and a good start for clarification whether the three-day regimen ACT is mandatory.\nOf the existing forms of ACT which had been studied in Indonesia, a three-day dihydroartemisinin-piperaquine (DHP) is the best alternative for the Indonesian ACT programme.\nDHP is safe and effective for both P. falciparum and P. vivax malaria.\nThe cure rates of DHP for the treatment of P. falciparum and P. vivax were reported 95.2% and 92.7%.\nWhile the cure rates of AA, the first ACT programme were only 84% and 53.5%.\nIn addition, DHP had post treatment prophylactic effect in high transmission area.\nDHP is a good comparator for a trial of a new ACT.\nCurrently, multiple first-line therapies (MFT) policies against malaria have been introduced.\nIt was developed in the context of an evolutionary-epidemiological modelling framework for malaria resistance.\nThe benefits of using MFT against malaria yields a better clinical outcome, delay the emergence and slow the fixation of resistant strains, and allow a larger fraction of the population to be treated without trading off future treatment of cases that may be untreatable because of high resistance levels.\nDespite DHP, other forms of ACT should be identified to be chosen for MFT policies, including AN.\nIn this trial, one-day single dose of fixed dose regimen of AN and three-day regimen of DHP did not result ETF for treatment of any malaria.\nAll 19 TFs were as LTFs and reported from Jayapura only.\nOf the 19 TFs, 63.2% were related with clinical symptoms and classified as LCFs.\nThough 84.2% TFs were occurred by day >28, only 21.1% TFs (4 of 19) confirmed as new infections which were detected by days 35 and 42.\nThis findings support Papua as a highly multidrug resistance area.\nA significant bigger number of samples from Jayapura (75% of a total sample) probably gave a chance detected more TF cases.\nMoreover, a relatively small number of new infections in a moderate-high transmission Papua by days 35 and 42 could be because of long half-life of both study drugs AN and DHP which known have post treatment prophylactic effect.\nACT with a long half-life partner drug is a good choice for malaria in high transmission area, however there will be also an increased risk of selecting drug-resistant isolates.\nTherefore, monitoring drug efficacy should be routinely maintained to detect the spread of drug resistance.\nIn this trial, both drugs showed very effective for treatment of any malaria.\nAN and DHP had day-42 PCR corrected ACPR of \u226590% in ITT and modified population, and >95% in PP population.\nMoreover, the day-28 and day-42 ACPRs of AN and DHP were not statistically different (p\u2009>\u20090.05).\nThese findings are consistent with previous efficacy studies of AN for treatment of uncomplicated P.falciparum malaria in China, Myanmar and Papua New Guinea, and DHP for treatment of uncomplicated P.falciparum and P.vivax malaria in Indonesia and for treatment of uncomplicated P.falciparum malaria in other Asian countries [24-28].\nSimilarly to other forms of ACT, AN and DHP cleared fever and asexual parasites rapidly with means FCT 13.0 and 11.3 hours, and PCT 28.0 and 25.5 hours.\nThese findings were similar to the previous studies.\nBoth forms of ACT also resulted 66.5% haemoglobin recovery by day 28.\nThe anaemic study subjects with Hb\u2009<\u200911 g% were significantly decreased from 23.6% and 21.6% prior treatment and became 5.9% and 9.9% by day 28 in AN and DHP groups.\nThough many factors influence the recovery of haemoglobin, a long half-life of partner drugs naphthoquine and piperaquine have an important role to prevent re-infections or relapses, which might cause anaemia.\nBoth AN and DHP were well-tolerated, with no significant ECG changes identified.\nMost of the adverse events were mild and related with symptoms attributable to malaria.\nAll symptoms were recovered with or without simple treatment.\nThere were no significant differences in tolerability between the two study drugs.\nIn this clinical trial, eosinophilia was found in one third of study subjects treated with AN or DHP.\nSeveral studies also reported that pseudoeosinophilia was associated with malaria infection detected by Sysmex XE-2100 haematology analyzer due to the presence of haemozoin-containing neutrophils.\nEosinophilia may also represent a normal late response to malaria infection.\nHowever, all eosinophilia subjects had no significant clinically abnormality.\nThere were also slightly elevations of aminotransferase (ALT and AST) in study subjects treated with AN and DHP which gradually decreased to normal limit.\nThese findings were reported similar to Chinese studies.\nThis study has shown the efficacy and safety a single dose of AN and a three-day dose of DHP for treatment of any malaria in adult subjects.\nFurther analysis of the efficacy and safety specifically in P. falciparum and P. vivax, and others malaria should be performed and published to show the detail of study findings.\nIn addition, there was a study reported a high cure rate of twice daily one day of AN (100%) versus AL (98.4%) in African children with P. falciparum malaria.\nThe dosage used in that study was based on the children weight groups.\nDosage for children is crucial and will be safe determined by per kg body weight.\nCost of drug is also important factor for choosing programme malaria drug.\nA single dose ACT, such as AN actually should be cheaper compare with three-day regimen ACT.\nMoreover, there are study limitations to extend follow up day to D42 whereas D63 is recommended for AN with a long half lives, and the small sample size and consequent loss of power.\nConclusions\nBoth fixed-dose forms of ACT are confirmed very effective, safe and tolerate for treatment of any malaria in adults, and meet with the recent WHO recommendation for replacing ineffective drugs.\nTheir longer post-treatment prophylactic effect is useful in areas where transmission is intense.\nArtemisinin-naphthoquine and dihydroartemisinin-piperaquine are promising forms of ACT for MFT policy.\nPharmacokinetic and therapeutic efficacy study in children, and cost-effectiveness study should be carried out for the safety and effectiveness of large-scale use.\nProportions of malaria symptom and sign on enrolment in artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups were no significant difference (p\u2009>\u20090.05).\nClinical trial profile of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups.\nProportions of afebrile malaria subject treated with artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups were similar by point of observation (p\u2009>\u20090.05).\nProportions of aparasitaemic (asexual) malaria subject treated with artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups were significantly difference by Day 0\u201316 hr (21.8% vs 55.6%, p\u2009<\u20090.0001), Day 1\u20130 hr (57.5% vs 71.3%, p\u2009=\u20090.01), and Day 1\u20138 hr (79.8% vs 88.5%, p\u2009=\u20090.03).\nCumulative risk of failure of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) in ITT population infected with any malaria. The TF by day 28 and 42 were 4.5% (9 of 201) and 10% (20 of 201) in AN group, and 7% (14 of 200) and 10% (20 of 200) in DHP group. The hazard ratio of failure between AN and DHP groups was 0.98 (95% CI: 0.53\u20131.82) with p\u2009=\u20090.95.\nCumulative risk of failure of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) in modified ITT population infected with any malaria. The TF by day 28 and 42 were 2% (4 of 196) and 6.7% (13 of 194) in AN group, and 2.1% (4 of 190) and 5.3% (10 of 190) in DHP group. The hazard ratio of failure between AN and DHP groups was 1.28 (95% CI: 0.56 \u2013 2.92) with p\u2009=\u20090.56.\nCumulative risk of failure of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) in PP population infected with any malaria. The TF by day 28 and 42 were 1% (2 of 194) and 3.7% (7 of 188) in AN group, and 0.5% (1 of 187) and 2.7% (5 of 185) in DHP group. The Hazard ratio of failure between AN and DHP groups was 1.38 (95% CI: 0.44 \u2013 4.35) with p\u2009=\u20090.58.\n\nCharacteristics of study subjects by the study drug on enrolment\nCharacteristic | Artemisinin- naphthoquine(AN) | Dihydroartemisinin- piperaquine(DHP) | P | Overall \nNumber of subjects | 201 | 200 | \u00a0 | 401\nAge:mean \u00b1 SD (range) years | 27.6\u2009\u00b1\u200910.8 (15\u201369) | 26.2\u2009\u00b1\u20099.2 (15\u201367) | 0.18 | 26.9\u2009\u00b1\u200910.0 (15\u201369)\nBody weight: mean \u00b1 SD (range) kgs | 58.7\u2009\u00b1\u20098.5 (36\u201381) | 58.5\u2009\u00b1\u20098.5 (37\u201385) | 0.86 | 58.6\u2009\u00b1\u20098.5 (36\u201385)\nAxillary temperature: mean \u00b1 SD (range) \u00b0C | 37.9\u2009\u00b1\u20091.0 (36\u201340.3) | 37.9\u2009\u00b1\u20091.1 (35\u201340.2) | 0.84 | 37.9\u2009\u00b1\u20091.1 (35\u201340.3)\nBlood Pressure Systole: mean \u00b1 SD (range) mmHg | 113\u2009\u00b1\u200913 (90\u2013150) | 114\u2009\u00b1\u200913 (80\u2013170) | 0.44 | 113\u2009\u00b1\u200913 (80\u2013170)\nDiastole: mean \u00b1 SD (range) mmHg | 72\u2009\u00b1\u20099 (50\u201392) | 73\u2009\u00b1\u20099 (40\u2013120) | 0.39 | 72\u2009\u00b1\u20099 (40\u2013120)\nHeart rate: mean \u00b1 SD (range) per min | 86\u2009\u00b1\u200913 (52\u2013126) | 87\u2009\u00b1\u200913 (56\u2013126) | 0.59 | 86\u2009\u00b1\u200913 (52\u2013126)\nRespiration rate: mean \u00b1 SD (range) per min | 20\u2009\u00b1\u20093 (16\u201336) | 20\u2009\u00b1\u20093 (16\u201332) | 0.51 | 20\u2009\u00b1\u20093.0 (16\u201336)\nSex: male:female (%) | 181:20 (90:10) | 170:30 (85:15) | 0.17 | 351:50 (87.5:12.5)\nHistory of fever in the last 24 hours (%) | 195 (97.0) | 199 (99.5) | 0.12 | 394 (98.3)\nStudy subject with T.axillary \u226537.5\u00b0C (%) | 114 (56.7) | 109 (54.5) | 0.73 | 223 (55.6)\nHistory of experience malaria in the last year (%) | 160 (79.6) | 159 (79.5) | 1.00 | 319 (79.6)\nFrequency of malaria in the last year: mean \u00b1 SD (range) times | 2.6\u2009\u00b1\u20092.3 (1\u201320) | 2.4\u2009\u00b1\u20091.6 (1\u201312) | 0.53 | 2.7\u2009\u00b1\u20093 (1\u201320)\nHistory of taken antimalarial drugs in the last 4 weeks (%) | 55 (27.4) | 57 (28.5) | 0.89 | 112 (27.9)\n\n\nHaematology, blood chemistry and parasitology findings of study subjects on enrolment by the study drug\nParameter | Artemisinin- naphthoquine (AN) | Dihydroartemisinin- piperaquine (DHP) | P | Overall\nNumber of subjects | 201 | 200 | \u00a0 | 401\nHaematocrit: mean \u00b1 SD (range)% | 36.1\u2009\u00b1\u20097.1 (15.5\u201368.9) | 36.6\u2009\u00b1\u20096.6 (19.1\u201363.4) | 0.46 | 36.4\u2009\u00b1\u20096.8 (15.5\u201368.9)\nHaemoglobin: mean \u00b1 SD (range) g/dL | 12.5\u2009\u00b1\u20092.3 (6.2\u201323.1) | 12.6\u2009\u00b1\u20092.1 (6.6\u201321.1) | 0.67 | 12.6\u2009\u00b1\u20092.2 (6.2\u201323.1)\nRed Blood Cell: mean \u00b1 SD (range) per uL | 4.5\u2009\u00b1\u20090.8 (1.5\u20137.7) | 4.4\u2009\u00b1\u20090.8 (2.2\u20137.3) | 0.70 | 4.4\u2009\u00b1\u20090.8 (1.5\u20137.7)\nPlatelet: mean \u00b1 SD (range) 103/mm3 | 116.4\u2009\u00b1\u200968.7 (1.8\u2013381.0) | 116.5\u2009\u00b1\u200965.9 (1 .0\u2013601.0) | 0.98 | 116\u2009\u00b1\u200967.2 (1.0\u2013601.0)\nWhite Blood Cell: mean \u00b1 SD (range) per uL | 6.4\u2009\u00b1\u20092.2 (2.0\u201314.0) | 7.1\u2009\u00b1\u20096.0 (2.2\u201365.0) | 0.11 | 6.8\u2009\u00b1\u20094.5 (2.0\u201365.0)\nBilirubin: mean \u00b1 SD (range) mg/dL | 1.0\u2009\u00b1\u20090.6 (0.1\u20135.2) | 1.0\u2009\u00b1\u20090.5 (0.1\u20133.8) | 0.67 | 1.0\u2009\u00b1\u20090.5 (0.1\u20135.2)\nAlbumin: mean \u00b1 SD (range) g/dL | 4.1\u2009\u00b1\u20091.0 (2.0\u20137.4) | 4.2\u2009\u00b1\u20091.0 (2.2\u20136.2) | 0.25 | 4.2\u2009\u00b1\u20091.0 (2.1\u20137.4)\nALT (SGPT): mean \u00b1 SD (range) IU/L | 28.8\u2009\u00b1\u200913.0 (5\u201373) | 28.3\u2009\u00b1\u200914.2 (6.9\u201398.0) | 0.71 | 28.5\u2009\u00b1\u200913.9 (5.0\u201398.0)\nAST (SGOT): mean \u00b1 SD (range) IU/L | 29.4\u2009\u00b1\u200914.0 (2.0\u201387.0) | 28.5\u2009\u00b1\u200911.8 (2.6\u201391.0) | 0.49 | 28.9\u2009\u00b1\u200912.9 (2.0\u201391.0)\nCreatinine: mean \u00b1 SD (range) mg/dL | 1.0\u2009\u00b1\u20090.4 (0.2\u20132.5) | 0.9\u2009\u00b1\u20090.3 (0.2\u20132.0) | 0.15 | 0.9\u2009\u00b1\u20090.4 (0.2\u20132.5)\nUrea: mean \u00b1 SD (range) mg/dL | 27.5\u2009\u00b1\u200911.7 (2.7\u201380.0) | 27.6\u2009\u00b1\u200911.7 (8.6\u2013110.0) | 0.94 | 27.6\u2009\u00b1\u200911.7 (2.7\u2013110.0)\nDensity of asexual parasites: geometric mean (range) per uL | 6310 (304\u2013113550) | 6972 (372\u2013140084) | 0.21 | 6634 (304\u2013140084)\nGametocyte carriages (%) | 134 (67.3) | 140 (70.3) | 0.54 | 274 (68.8)\nDensity of gametocytes: geometric mean (range) per uL | 40 (1\u20132697) | 35 (2\u20132208) | 0.55 | 38 (1\u20132697)\n\n\nPCR uncorrected efficacy of AN vs DHP in all malaria, Indonesia\nOutcome | Artemisinin- naphthoquine (AN) | Dihydroartemisinin-piperaquine (DHP) | P | Overall\nACPR/Treatment Success (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 179 [89.1 (84.7\u201393.4)] | 180 [90.0 (85.8\u201394.2)] | 0.76 | 359 [89.5 (86.1\u201392.3)]\nModified ITT | 179 [92.3 (88.5\u201396.0)] | 180 [94.7 (91.6\u201397.9)] | 0.33 | 359 [93.5 (90.5\u201395.7)]\nPer Protocol | 179 [93.7 (90.3\u201397.2)] | 180 [96.3 (93.5\u201399.0)] | 0.26 | 359 [95.0 (92.3\u201396.9)]\nLate Clinical Failure-LCF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 5 [2.5 (0.3\u20134.6)] | 1 [0.5 (0.5\u20131.5)] | 0.10 | 6 [1.5 (0.6\u20133.2)]\nModified ITT | 5 [2.6 (0.3\u20134.8)] | 1 [0.5 (0.5\u20131.6)] | 0.10 | 6 [1.6 (0.6\u20133.4)]\nPer Protocol | 5 [2.6 (0.4\u20134.9)] | 1 [0.5 (0.5\u20131.6)] | 0.10 | 6 [1.6 (0.6\u20133.4)]\nLate Parasitological Failure-LPF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 7 [3.5 (0.9\u20136.0)] | 6 [3.0 (0.6\u20135.4)] | 0.78 | 13 [3.2 (1.7\u20135.5)]\nModified ITT | 7 [3.6 (1.0\u20136.2)] | 6 [3.2 (0.7\u20135.6)] | 0.81 | 13 [3.4 (1.8\u20135.7)]\nPer Protocol | 7 [3.7 (1.0\u20136.3)] | 6 [3.2 (0.7\u20135.7)] | 0.81 | 13 [3.4 (1.8\u20135.8)]\n\u201cOther Failures\u201d (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 10 [5.0 (2.0\u20138.0)] | 13 [6.5 (3.1\u20139.9)] | 0.51 | 23 [5.7 (3.7\u20138.5)]\nModified ITT | 3 [1.5 (0.2\u20133.3)] | 3 [1.6 (0.2\u20133.4)] | 0.98 | 6 [1.6 (0.6\u20133.4)]\nPer Protocol | 0 | 0 | \u00a0 | 0\n\nAll protocol violations, withdrawn consents, and lost to follow up were classified as \u201cother failures\u201d in ITT analysis.\nAll protocol violations and withdrawn consents were classified as \u201cother failures\u201d in modified ITT analysis.\n\nPCR corrected efficacy of AN vs Duocotecxin in all malaria, Indonesia\nOutcome | Artemisinin- naphthoquine (AN) | Dihydroartemisinin-piperaquine (DHP) | P | Overall\nACPR/Treatment Success (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 181 [90.0 (85.9\u201394.2)] | 180 [90.0 (85.8\u201394.2)] | 0.99 | 361 [90.0 (86.7\u201392.8)]\nModified ITT | 181 [93.3 (89.8\u201396.8)] | 180 [94.7 (91.6\u201397.9)] | 0.55 | 361 [94.0 (91.1\u201396.2)]\nPer Protocol | 181 [96.3 (93.6\u201399.0)] | 180 [97.3 (95.0\u201399.6)] | 0.58 | 361 [96.8 (94.4\u201398.3)]\nLate Clinical Failure-LCF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 4 [2.0 (0.1\u20133.9)] | 1 [0.5 (0.5\u20131.5)] | 0.18 | 5 [1.2 (0.4\u20132.9)]\nModified ITT | 4 [2.1 (0.1\u20134.1)] | 1 [0.5 (0.5\u20131.6)] | 0.18 | 5 [1.3 (0.4\u20133.0)]\n3. Per Protocol | 4 [2.1 (0.1\u20134.2)] | 1 [0.5 (0.5\u20131.6)] | 0.18 | 5 [1.3 (0.4\u20133.1)]\nLate Parasitological Failure-LPF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 3 [1.5 (0.2\u20133.2)] | 4 [2.0 (0.1\u20133.9)] | 0.70 | 7 [1.7 (0.7\u20133.6)]\nModified ITT | 3 [1.6 (0.2\u20133.3)] | 4 [2.1 (0.1\u20134.1)] | 0.68 | 7 [1.8 (0.7\u20133.7)]\nPer Protocol | 3 [1.6 (0.2\u20133.4)] | 4 [2.2 (0.1\u20134.3)] | 0.69 | 7 [1.9 (0.8\u20133.8)]\n\u201cOther Failures\u201d (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 13 [6.5 (3.1\u20139.9)] | 15 [7.5 (3.8\u201311.1)] | 0.68 | 28 [7.0 (4.7\u20139.9)]\nModified ITT | 6 [3.1 (0.7\u20135.5)] | 5[2.6 (0.4\u20134.9)] | \u00a0 | 11 [2.9 (1.4\u20135.1)]\nPer Protocol | 0 | 0 | 0.79 | 0\n\nAll protocol violations, withdrawn consents, and lost to follow up were classified as \u201cother failures\u201d in ITT analysis.\nAll protocol violations and withdrawn consents were classified as \u201cother failures\u201d in modified ITT analysis. PCR corrected was carried out only for P.falciparum recurrences.",
        "label": "low",
        "id": "task4_RLD_test_1048"
    },
    {
        "paper_doi": "10.7189/jogh.09.020402",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: DesigncRCTAllocation of clusters50 schools randomized to intervention, 50 to control\n\n\nParticipants: 9258 primary school-aged children\n\n\nInterventions: Broad multiple\n\n\nOutcomes: Any STH\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Background\nWater, sanitation, and hygiene (WASH) in schools is promoted by development agencies as a modality to improve school attendance by reducing illness.\nDespite biological plausibility, the few rigorous studies that have assessed the effect of WASH in schools (WinS) interventions on pupil health and school attendance have reported mixed impacts.\nWe evaluated the impact of the Laos Basic Education, Water, Sanitation and Hygiene Programme \u2013 a comprehensive WinS project implemented by UNICEF Lao People\u2019s Democratic Republic (Lao PDR) in 492 primary schools nationwide between 2013 and 2017 \u2013 on pupil education and health.\nMethods\nFrom 2014-2017, we conducted a cluster-randomized trial among 100 randomly selected primary schools lacking functional WASH facilities in Saravane Province, Lao PDR.\nSchools were randomly assigned to either the intervention (n\u2009=\u200950) or comparison (n\u2009=\u200950) arm.\nIntervention schools received a school water supply, sanitation facilities, handwashing facilities, drinking water filters, and behavior change education and promotion.\nComparison schools received the intervention after research activities ended.\nAt unannounced visits every six to eight weeks, enumerators recorded pupils\u2019 roll-call absence, enrollment, attrition, progression to the next grade, and reported illness (diarrhea, respiratory infection, conjunctivitis), and conducted structured observations to measure intervention fidelity and adherence.\nStool samples were collected annually prior to de-worming and analyzed for soil-transmitted helminth (STH) infection.\nIn addition to our primary intention-to-treat analysis, we conducted secondary analyses to quantify the role of intervention fidelity and adherence on project impacts.\nResults\nWe found no impact of the WinS intervention on any primary (pupil absence) or secondary (enrollment, dropout, grade progression, diarrhea, respiratory infection, conjunctivitis, STH infection) impacts.\nEven among schools with the highest levels of fidelity and adherence, impact of the intervention on absence and health was minimal.\nConclusions\nWhile WinS may create an important enabling environment, WinS interventions alone and as currently delivered may not be sufficient to independently impact pupil education and health.\nOur results are consistent with other recent evaluations of WinS projects showing limited or mixed effects of WinS.\nSchool-aged children in low-income settings are at substantial risk for water, sanitation, and hygiene (WASH)-related infections such as pathogens causing diarrheal diseases, soil-transmitted helminths (STH), and trachoma.\nCrowded, unsanitary conditions may facilitate the spread of pathogens, and increase pupils\u2019 risk for disease.\nImproved access to WASH facilities combined with sufficient behavior change may not only prevent the spread of pathogens within the school domain but also lead to beneficial WASH habits at home and throughout the life course.\nThe limited data available indicate that only 69% of schools worldwide have access to sanitation facilities, while only 66% have access to water.\nWASH in schools (WinS) targets and indicators have been included in the Sustainable Development Goals.\nDespite the biological plausibility of WinS interventions to reduce illness and subsequently school absence, evidence of impact has been mixed.\nSome WinS efficacy studies, such as those assessing intensive handwashing programs in China and Egypt, reported reductions in absence and absence due to illness.\nHowever, with only 6- and 3-month follow up periods, respectively, and with soap being continuously supplied by the intervention or school administration, respectively, the long-term sustainability of handwashing behaviors linked to these impacts is unknown.\nEffectiveness trials of WinS projects have not replicated this success.\nA matched-control evaluation of a comprehensive WinS program in Mali revealed reductions in pupil-reported diarrhea, symptoms of respiratory infection, and absence due to diarrhea, but higher odds of absence overall among pupils enrolled in beneficiary schools.\nHowever, there were imbalances between the beneficiary and comparison groups at baseline, and the study was further limited by inconsistent fidelity to the intervention by implementing partners and participating schools.\nA randomized controlled trial (RCT) of a WinS program in Kenya reported a 44% reduction in odds of Ascaris lumbricoides reinfection, but no overall impact on absence or diarrhea.\nProgram impact differed by intervention arm (as individual and combined WASH interventions were employed) and subsets of the sample population.\nAbsence among girls in the hygiene promotion and water treatment arm reduced by 58%.\nIn water-scarce schools that received a comprehensive WASH intervention, including water supply improvements, risk of diarrhea among pupils reduced by 61%, while diarrhea among pupils\u2019 siblings under 5 years old reduced by 56%.\nHowever, program impact may have been affected by incomplete and inconsistent intervention delivery (fidelity) and uptake and use by the target population (adherence).\nA WinS intervention in Lao People\u2019s Democratic Republic (Lao PDR), Cambodia, and Indonesia had no impact on STH infection or being underweight, but reported evidence of improvement in dental cavities.\nAgain, this evaluation was potentially limited by incomplete fidelity and adherence to the program, as well as a non-randomized design and contamination from concurrent programming in control schools.\nHere, we present results from the Water, Sanitation, and Hygiene for Health and Education in Laotian Primary Schools (WASH HELPS) study, a cluster-RCT designed to measure the impact of a comprehensive WinS project \u2013 water supply, sanitation, handwashing, and behavior change - in Lao PDR on pupil absence, diarrhea, respiratory infection, and STH infection.\nGiven past challenges in program fidelity and adherence to project outputs and behaviors, we also apply two analyses that have previously been used to evaluate the role of intervention fidelity and adherence on WinS project impacts.\nMETHODS\nStudy setting and intervention\nThe Laos Basic Education, Water, Sanitation and Hygiene Programme was implemented by UNICEF in 492 primary schools across thirteen provinces between 2013 and 2017.\nThe WASH HELPS Study, a research component of the intervention, was conducted between September 2014 and May 2017 in Saravane Province, which was selected because it was the only province in which intervention activities had not yet occurred, thus allowing a randomized intervention trial.\nThe study setting, baseline results, intervention components, intervention outputs and outcomes, and their fidelity and adherence have been described in detail elsewhere.\nKey outputs and outcomes of the project are listed in Table 1.\nBriefly, the comprehensive WinS project included provision of a school water supply, sanitation facilities, handwashing facilities (individual and group), drinking water filters, and behavior change education and promotion.\nThe project was implemented in two phases; lessons learned from Group 1 schools (n\u2009=\u200952; intervention started in 2014) were applied to improve the project for Group 2 schools (n\u2009=\u200948; intervention started in 2015), leading to different levels of achievement at output and outcome levels between groups, as well as different durations of follow-up.\nStudy design, sampling, and data collection\nWe conducted a cluster-randomized, controlled trial among 100 primary schools (50 intervention, 50 comparison).\nStudy design, sampling, and data collection methods have been previously published.\nWe used stratified random sampling to help ensure equal representation of control and intervention schools in each district, and that the number of schools selected in each district was proportional to the number of eligible schools in each district.\nWe selected up to 40 pupils from grades 3-5 in each school using systematic stratified sampling, with grade and sex as the stratification variables.\nPupils selected at baseline were followed throughout the entire study period; pupils who left the school due to abandonment or transfer were replaced at the beginning of the following academic year, maintaining equal grade and sex ratios when possible.\nPupils who progressed from fifth to the sixth grade were replaced with pupils from grade three the following academic year.\nA total of 3993 pupils were enrolled throughout the study period.\nData were collected over three or two school years (Group 1 and 2 schools, respectively) to measure uptake and sustainability of facilities and behavior change.\nTo account for variabilities across time and season, data were collected throughout the school year, which consists of 33 weeks across two semesters (September-January and February-May), with five to six hours of instruction per day.\nTrained enumerators visited study schools every six to eight weeks during the school year through March 2017, for a total of 11 (Group 1) or 7 (Group 2) visits per school.\nAll visits were unannounced and during school hours.\nAt each visit, enumerators conducted a roll call of all students enrolled in the school using sex- and grade-specific ledgers; interviewed the school directors; interviewed sampled pupils in grades 3\u20135; observed conditions and functionality of WinS hardware; and observed individual and group handwashing practices.\nEach year, stool samples were collected from up to 50 pupils per school prior to distribution of preventative chemotherapy as part of the National School Deworming Programme.\nStool samples were tested for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus) using the Kato Katz technique.\nMeasures\nOur primary impact of interest was pupil absence measured by school-wide roll-call at each visit.\nAt the beginning of each data collection visit, enumerators visited each classroom with a roster of all students enrolled in the school, stratified by grade and sex.\nAt each visit, enumerators confirmed with the head teacher whether there were any new students since the last visit or if any students had left the school.\nNew students were added to the roster.\nDropout was recorded for students who had dropped out since the last visit.\nAbsences that were followed by a designation of dropout or transfer were removed from roll-call analysis.\nSecondary educational impacts included enrollment, dropout, and progression.\nEnrollment was calculated at each visit by summing the count of pupils on the roll-call roster and subtracting those who had dropped out or transferred.\nIn addition to student-level dropout recorded in the roll-call register, an aggregate school-level count of dropout was reported by the school at the end of each school year.\nPupils who transferred to another school were not considered to have dropped out.\nProgression was school-reported at the end of each academic year as the count of students who passed the national exam and progressed to the next grade level.\nAll secondary educational impacts were stratified by grade and sex.\nSecondary health impacts included diarrhea, symptoms of respiratory infection, and conjunctivitis/non-vision related eye illness and were collected through pupil interviews.\nAll health impacts were binary and self-reported with a one week recall period.\nPupils were asked if they had had diarrhea using local terminology and were also asked how many times they had defecated each day; a pupil was considered to have had diarrhea if he or she had reported having diarrhea and had defecated three or more times in a 24-hour period.\nPupils were considered to have symptoms of respiratory infection if they reported cough, runny nose, stuffy nose, or sore throat.\nDuring the last visit we included negative-control questions about self-reported cuts/scrapes and toothache.\nThese questions served as a measure of respondent bias, as there is no biological plausibility of an association between a WinS program and cuts/scrapes or toothache.\nData on STH infection were collected yearly.\nAny sample testing positive for the hookworms, A. lumbricoides, or T. trichuria considered positive for STH infection.\nIntervention fidelity and adherence for this study has been described previously.\nTo measure fidelity- defined as how the intervention was delivered per the stated design- we created an index score in which one point was given for each of the 20 output criteria fulfilled (Table 1).\nFor each visit, the minimum intervention fidelity score was zero and the maximum score was 20.\nTo measure adherence \u2013 defined as achievement of behavioral outcomes promoted by the intervention \u2013 a similar index score was created.\nAlthough there were five behavioral outcomes of interest (Table 1), we excluded group compound cleaning from the index given that reported participation in group compound cleaning was nearly universal among both intervention and comparison schools at baseline (97.9%), therefore the adherence score ranged from 0-4.\nA behavior was considered to be achieved when >75% of pupils reported or were observed to complete the behavior except for group handwashing, which was binary (either the school performed group handwashing or did not).\nAnalysis\nData were analyzed using Stata Statistical Software: Release 13 (StataCorp LP, College Station, TX, USA) and SAS version 9.4 (SAS Institute, Cary, NC, USA).\nIntention to treat analysis\nOur primary analysis was an intention-to-treat (ITT) analysis, which was used on all primary and secondary impacts.\nFor binary impacts (roll-call absence, diarrhea, symptoms of respiratory infection, conjunctivitis/non-vision related eye illness, STH infection, toothache, cuts/scrapes), we estimated relative risk using a \u201cmodified Poisson\u201d approach.\nThis is a validated method to produce relative risk ratios for binary data using a multi-level mixed Poisson model with robust error variances, and was chosen for this analysis because Stata does not support the use of log-linear binomial regression when using mixed effects generalized linear models.\nOdds ratios were obtained when the modified Poisson model did not converge for a specific impact (eg, toothache).\nRandom intercepts at the school and pupil levels were included to account for clustering of pupils within schools and for repeated measures of pupils over time, respectively.\nFor count impacts (enrollment, abandonment, progression), we estimated relative risk using Poisson regression models.\nAs these data were aggregated at the school-level, we included a random intercept at the school level only.\nAll ITT models compared intervention schools to comparison schools as they were randomly allocated to intervention and comparison groups, without regard to project fidelity or adherence.\nIntervention and comparison schools were balanced on key indicators at baseline, therefore intervention schools were included in the analysis once UNICEF documented that full intervention implementation (eg, both hardware and behavior change components) was complete.\nSince full implementation generally occurred at the same time in each district, comparison schools were also included once implementation occurred in their respective districts.\nModels included several design variables, including the district and visit number, and controlled for the following fixed effects, determined a priori based on biological plausibility of affecting impacts: pupil sex, pupil grade, school enrollment size, season (rainy or dry).\nThe rice crop calendar (planting, growing, harvesting) was included as a fixed effect in the absence model because rice agriculture is the predominant economic activity in the province and the need to stay home and support the family was the leading cause of pupil-reported absence.\nFully adjusted models were used to produce adjusted risk ratios (aRR) for each of the associations of interest.\nThese fixed effects, as well as whether the school was concurrently receiving aid from the World Food Program (WFP) school feeding program, were also assed for effect modification.\nCovariates were determined to be effect modifiers if an interaction term between the covariate and intervention group was significant in the full model.\nIntervention fidelity and adherence are important considerations when evaluating the impact of WASH programs.\nAssessing these factors along the causal \u2018theory of change\u2019 allows us to understand not only if but why and how that intervention succeeded or not in that context (ie, was there theory failure?).\nFurther, assessment of the process can determine if the intervention followed the intervention protocol to activate that theory of change (ie, was there intervention failure?).\nIn contexts where fidelity and adherence to the intervention is imperfect, ITT results may underestimate the causal effect for the potential impact of changes to outputs or outcomes, resulting in null or mixed effects.\nGiven the suboptimal fidelity and adherence of the intervention based on our monitoring data, we conducted a secondary analysis to quantify the impact of the project as implemented by UNICEF and adhered to by schools and pupils on the primary impact (roll-call absence) and select secondary impacts (diarrhea, respiratory infection, and STH prevalence).\nWe explore two modeling frameworks that have been previously used to evaluate the role of fidelity and adherence to a school WASH intervention on project impacts: As-treated (AT) analysis and Structural Nested Models (SNMs).\nEach framework operates under different assumptions and differ in robustness and efficiency; as such, comparing estimates lends a more informed picture of project impact.\nWe conducted a sensitivity analysis to identify a meaningful threshold of fidelity and adherence.\nThe scale of 20 outputs (fidelity) were categorized with cut-points at each 10th percentile and the scale of four outcomes (adherence) were unadjusted.\nWe observed lower risk of absence among schools with 70%-80% intervention fidelity and higher, but there was no clear evidence of a threshold for any other association (Figure 1).\nWe thus selected a threshold of 75%, which is consistent with previous research on fidelity to WinS projects.\nOnly the SNM requires specifying a threshold of fidelity/adherence, however, we also applied the 75% threshold to the AT models for comparability between the two approaches.\nAs-treated analysis\nThe AT analysis groups subjects according to the treatment received and does not consider the treatment intended (as is the case with ITT analysis).\nAdvantages to the AT approach are that it is analytically straightforward and easily supports our clustered and longitudinal study design.\nDisadvantages are that characteristics of schools with good fidelity or students who adhere to behaviors may be fundamentally different from those with poor fidelity/adherence, which can lead to confounding.\nThis confounding may be remedied by controlling for the prognostic factors that led participants to choose to adhere, but only if those prognostic factors are known, which is often not the case.\nFor the AT analysis, we ran two separate models that were structurally identical to the ITT models.\nHowever, instead of using intervention status as the primary predictor, as in the ITT analysis, schools were grouped according to intervention fidelity (ie, fulfilling \u226575% of outputs or not) and adherence (ie, fulfilling \u226575% of outcomes or not), respectively.\nAT models included the same covariates as the ITT models, with a priori identified fixed effects and random intercepts at the school and pupil levels.\nOnly data collected after the implementor reported intervention delivery was complete were included.\nAT models were stratified by effect modifiers identified in the ITT analysis.\nStructural nested model analysis\nSecond, we assessed the role of fidelity and adherence using SNMs, an instrumental variable approach.\nSNMs resolve the potential confounding issue presented by AT models because they do not break the randomization of intervention status.\nInstead, SNMs create a counterfactual for each study participant in order to compare the risk of an impact among adherers against the risk of the impact had the same individual not adhered.\nUnlike the ITT and AT models, to control for relevant covariates, a weighted distribution of population data are produced in order to remove the association between population-level confounders and randomization.\nWhile SNMs are advantageous because they account for unknown or unmeasured confounders, drawbacks are that they are more computationally intensive and rely on strong assumptions.\nSNM assumptions are described in detail elsewhere; briefly, they are as follows: (1) Exclusion restriction \u2013 randomization has no direct effect on the outcome; (2) Consistency \u2013 observed outcomes are possible under the fidelity/adherence level actually observed; (3)\nThe potential outcomes used to estimate the SNM effects are independent of randomization; (4) No interaction \u2013 the model\u2019s causal effect is consistent across randomization groups.\nOur code was derived from Garn et al and adapted for a 2-arm trial.\nBecause the SNM methodology we used does not accommodate repeated measures, we averaged time-varying pupil-level data (eg, grade, absence, reported diarrhea, reported symptoms of respiratory infection) and school-level data (output index, behavior index) across the data collection period.\nAs such, binary variables such as absence, reported diarrhea, and reported symptoms of respiratory infection became a continuous variable between zero and one, in which zero indicated never being absent, reporting diarrhea, or reporting symptoms of respiratory infection, whereas one indicated always being absent, reporting diarrhea, or reporting symptoms of respiratory infection.\nSimilar to the ITT and AT models described above, observations were included only after full implementation had been achieved.\nModels were adjusted using the same covariate variables as we used in the ITT and AT models.\nAs with the AT models, achievement of \u226575% of outputs and \u226575% of outcomes were considered achieving fidelity and adherence, respectively.\nSNMs were stratified by effect modifiers identified in the ITT analysis.\nFor all analyses, results were considered statistically significant if the P-value was <0.05.\nEthics\nThe WASH HELPS Study was approved by Emory University\u2019s Institutional Review Board (IRB0076404) and the Lao Ministry of Health\u2019s National Institute of Public Health National Ethics Committee (No. 043 NIOPH/NECHR).\nBoth Institutional Review Boards approved consent in loco parentis (in the place of the parent) signed by the school director.\nPupils who were selected for the pupil interview and/or stool collection provided informed verbal assent prior to any data collection.\nAll consent/assent procedures occurred after randomization.\nThe intervention was delivered to comparison schools in April 2017, after research activities ended.\nThe study is registered in ClinicalTrials.gov (NCT02342860).\nRESULTS\nBaseline results and intervention fidelity and adherence\nA total of 100 schools (n\u2009=\u200950 intervention, n\u2009=\u200950 comparison) were randomized, received the intervention, and included in the analysis (Figure 2).\nThere were no significant differences in key pupil-level or school-level indicators between intervention and comparison groups at baseline, indicating that the cluster-randomization was successful in creating balanced groups.\nFollowing full intervention implementation, intervention fidelity was 30.9% across all schools and visits and intervention adherence was 29.4%.\nData on fidelity to specific project outputs and adherence to specific project behaviors across the evaluation period have been previously published.\nIntention-to-treat analysis\nWe found no impact of the intervention on the primary impacts (roll-call absence) or secondary impacts (enrollment, progression, pupil-reported diarrhea, pupil-reported symptoms of respiratory infection, pupil-reported conjunctivitis, STH infection; Table 2).\nThere was some evidence of effect modification.\nRisk of diarrhea was higher in the rainy season compared to the dry season; when stratified by season, there was no significant impact of the intervention on diarrhea in either season (Dry season adjusted risk ratio (aRR)\u2009=\u20090.69, 95% confidence interval (CI)\u2009=\u20090.44, 1.10; Rainy season aRR\u2009=\u20091.14, 95% CI\u2009=\u20090.65, 1.99).\nPupil sex, pupil grade, school enrollment size, receiving support from the WFP school feeding program, and the rice crop calendar (absence model only) did not modify the effect of any primary or secondary impacts.\nWe found no difference in reported prevalence of toothache or cuts/scrapes (the negative control questions) among pupils attending intervention vs comparison schools (toothache adjusted odds ratio (aOR\u2009=\u20090.64, 95% CI\u2009=\u20090.23, 1.84; cuts/scrapes aRR\u2009=\u20091.06, 95% CI\u2009=\u20090.66, 1.72), indicating that any respondent bias that may have been present occurred equally between groups.\nAs-treated analysis\nAT results are presented in Table 3.\nIntervention fidelity \u2013 meeting \u226575% of output indicators associated with water supply, toilets, handwashing facilities, promotion of group hygiene activities, group handwashing facilities, and filtered drinking water\u2013 was associated with roll call absence and prevalence of STH.\nCompared to students attending schools without intervention fidelity, students attending schools with intervention fidelity had a 23% lower risk of absence (aRR\u2009=\u20090.76, 95% CI\u2009=\u20090.64, 0.91) and a 20% higher risk of STH prevalence (aRR\u2009=\u20091.20, 95% CI\u2009=\u20091.01, 1.43).\nDiarrhea was significantly higher during the rainy season, but when stratified there was no significant difference by fidelity status (Dry season aRR\u2009=\u20090.84, 95% CI\u2009=\u20090.48, 1.49; Rainy season aRR\u2009=\u20091.65, 95% CI\u2009=\u20090.82, 3.33).\nIntervention adherence \u2013 meeting outcome indicators associated with toilet use, handwashing with soap after toilet use, daily group handwashing, and daily group toilet cleaning \u2013 was not significantly associated with any impacts.\nStructural nested model analysis\nResults from the SNMs are presented in Table 3.\nDiarrhea was the only impact associated with fidelity or adherence.\nWhen stratified by season, diarrhea was lower in the dry season among students attending schools with intervention fidelity (aRR\u2009=\u20090.45, 95% CI\u2009=\u20090.24, 0.85) and adherence (aRR\u2009=\u20090.42, 95% CI\u2009=\u20090.21, 0.87); there was no significant difference in diarrhea between groups during the rainy season.\nDISCUSSION\nIn the primary analysis, we found no evidence that the intervention had an effect on absence, school enrollment, dropout, grade progression, pupil-reported diarrhea, pupil-reported symptoms of respiratory infection, pupil-reported conjunctivitis, or prevalence of STH.\nThese results contribute to the growing body of research showing limited or mixed impacts of WinS effectiveness trials on pupil health and education.\nSince 2010, access to WASH has been a fundamental human right recognized by the United Nations General Assembly.\nAs such, regardless of its potential education and health impacts, WinS access is an important objective, evidenced by its inclusion in the Sustainable Development Goals.\nHowever, if improvements in education and health indicators are to be achieved, results from this and other rigorously evaluated WinS programs suggest that WinS interventions alone, and as currently delivered in many contexts, may be insufficient to achieve anticipated education and health impacts.\nThe theory of change for WinS programs posits that improved WASH access leads to reductions in pathogen exposure at the school level and the habitualization of hygiene behaviors that can be practiced both at school at and home, which in-turn leads to reduced illness and thus reduced school absence.\nNumerous factors influence school absence, such as household wealth, distance to school, and number of siblings.\nLao PDR is a least-developed country, with over 65% of the population working in agriculture.\nIn Saravane Province, where over half of the population lives in poverty, the school calendar largely coincides with rice planting and harvesting seasons, and children are often kept home from school to assist in the fields and with other household chores.\nIndeed, in the current study, the leading pupil-reported cause of school absence was the need to stay home to support the family in economic activities (9.4% of pupils in intervention group and 8.7% of pupils in comparison group across all visits), not illness (5.1% of pupils in intervention group and 5.8% of pupils in comparison group across all visits), which may explain the lack of an impact of the intervention on absence.\nThus, the role of school WASH in supporting an enabling environment may be critical, but ultimately not sufficient to reduce absence when other factors like household economic needs or food security is the main driver of truancy from school.\nComplementary approaches to WinS may be necessary to achieve improvements in absence and other educational impacts.\nFor example, WinS may be successful in combination with school feeding programs or conditional cash transfers, both of which have been associated with reduced absence and increased enrollment in other low- and middle-income contexts.\nAlthough our results did not reveal a significant interaction between the WFP school feeding program and absence or enrollment, our study was not designed or adequately powered to detect a difference.\nAlthough there are potential mechanisms by which improved WASH may impact illness independently of measurable impacts on absence, we found no overall impact of the WinS intervention on pupil illness.\nThese results contrast to previous WinS research that reported overall reductions in diarrhea, respiratory infection, and absence due to illness, but are consistent with results from a WinS intervention in Lao PDR, Cambodia, and Indonesia that found no impact of the intervention on STH or underweight.\nOne explanation for the lack of an effect of the WinS intervention on pupil illness is low household WASH access; in this study context, the health benefits linked to improvements in school WASH conditions and behaviors provided by this intervention were likely not sufficient to overcome other potential transmission pathways at home or elsewhere in the community.\nEnvironmental improvements in both the domestic and public domains may be required for successful control of infections targeted by environmental improvements, such as diarrhea.\nAs such, WinS alone may not achieve significant health gains without concurrent community and household WASH improvements.\nFidelity and adherence are fundamental antecedents to achieving intervention effects.\nIt is possible that the lack of an effect of the intervention could be due, in part, to sub-optimal or unsustained fidelity and adherence.\nHowever, our secondary analyses yielded limited evidence of an effect of the intervention, even at high levels of intervention fidelity and adherence.\nAdditionally, our sensitivity analysis showed no clear trend in impacts across the fidelity/adherence continuum.\nWith two exceptions \u2013 the association between fidelity and lower absence (AT analysis) and the association between fidelity and adherence and lower diarrhea during the dry season (SNM analysis) \u2013 we did not find that fidelity and adherence led to improved education or health.\nThese results support the above conclusion that factors other than WinS \u2013 such as low household WASH access or household economics \u2013 may supersede health and education benefits of a WinS intervention in low-income contexts.\nHowever, the AT evidence should be should be interpreted cautiously due to the limited potential for causal inference resulting from breaking the randomization assignment in the AT analysis.\nThe two fidelity and adherence analyses results were inconsistent and sometimes yielded estimates of effect in opposite directions (eg, associations between adherence and diarrhea, respiratory infection, and STH), which is likely due to unaccounted for confounding in the AT analysis.\nIV analyses are known to yield estimators with high variance, especially when compliance is low, which may also partially explain differences between the AT and SNM results.\nThe choice of which method to use depends on numerous factors, including study design, plausibility of meeting analysis assumptions, and available analytical resources; our conflicting estimates highlight the importance of testing the sensitivity of multiple fidelity analysis options.\nStrengths and limitations\nThe design, methods and approach of the WASH HELPS study were robust.\nRandomized controlled trials offer the greatest potential for causal inference.\nThe longitudinal design allowed us to collect data across three full school years of in Group 1 schools and two full school years of in Group 2 schools, allowing us to capture inter-seasonal and inter-year variations in the outputs, outcomes, and impacts.\nAll data were collected during unannounced school visits so that schools could not prepare for the visit and bias observations.\nOur primary measure of impact \u2013 roll-call absence - is an objective measure of school absence.\nThis impact evaluation was conducted by external researchers, to foster an unbiased assessment of the project impact.\nOur field team was composed of experienced Laotian enumerators to ensure the tools were designed and delivered with cultural and contextual appropriateness.\nThis robust study design lends strong internal validity, and results may be generalized to the larger, nationwide WinS project.\nThis was an effectiveness trial evaluating an intervention as conducted in a real-world setting.\nThe lessons from this project, taken with other recent WinS trials, reveal heterogeneity of findings that can inform programming across contexts.\nLastly, in addition to comparing two methods to analyze the effect of intervention fidelity on WinS impacts, our fidelity analysis also examines adherence to intervention behaviors, which has not been previously included in WinS fidelity analyses.\nThere are a number of limitations to this evaluation.\nFirst, the secondary health impact measures (diarrhea, symptoms of respiratory infection, conjunctivitis) were based on self-report by pupils, which may be subject to bias, and this evaluation was not blinded for either the beneficiaries or data collectors.\nMore objective and robust measures of pupil health, such as molecular methods to detect enteric infection in stool samples, would improve our confidence in the reported impacts, though these measures can be costly, time consuming, and require specialized equipment and laboratory staff.\nAs a way to measure potential reporting bias, we included a negative control question about symptoms of illness unrelated to WASH access (cuts/scrapes and toothache) at the last survey visit.\nDifferences in reported symptoms of these illnesses between intervention and comparison groups would indicate a potential reporting bias, but we found no evidence to suggest that any bias may have existed to a greater degree among either the intervention group or the comparison group.\nAdditionally, schools in the comparison group did not have functional WASH facilities, so it is unlikely that the null results could be explained by a change in behaviors among the comparison group.\nSecond, the intervention was delivered across two different school years, so Group 1 schools had one more year of surveillance than Group 2 schools.\nFollowing a single cohort of schools over the same time period would have provided a more accurate measure of WinS hardware and software performance, sustainability, and impact.\nThird, implementation was delayed in many Group 1 schools.\nThe intervention was fully implemented in Group 1 schools at visit 4, with the exception of Samoui district, in which the intervention was fully implemented at visit 9.\nOur analysis excludes visits prior to full intervention implementation, thus power may have been limited by dropping observations under incomplete intervention delivery.\nLast, we were unable to account for the quality of intervention design or dose of the intervention received, which are important components of fidelity and adherence.\nCONCLUSIONS\nOur findings and those of other rigorous WinS trials suggest that WinS programs \u2013 as currently designed and delivered \u2013 do not have a population-level benefit on education and health.\nIn this context, the WinS improvements alone were not sufficient to address the other powerful causes of absenteeism, enrollment, and dropout that are not related to \u2013 but possibly more influential than \u2013 school WASH.\nWe believe this likely holds in many similar settings.\nSimilarly, WinS improvements, though potentially critical for the enabling environment, may not be sufficient to overcome disease transmission in areas where community and household WASH coverage is poor.\nWinS, independent of its stated purpose of improving education and health, is an important objective for dignity, inclusivity, and development.\nHowever, if intended impacts are to be achieved, improving intervention fidelity and adherence and including other complementary approaches for WASH may be required.\nTo better understand how to improve intervention fidelity and adherence, evaluations of WinS interventions need to better understand and adapt to contextual drivers of key impacts and outcomes, further develop and test theories of change, and conduct rigorous process evaluations to understand where along the causal pathways interventions are falling short.\nAssociation between intervention fidelity and adherence continuum and intervention impacts.\nFlow diagram of school and pupil selection.\n\nIntervention outputs and behavioral outcomes and their measurement indicators\nOutput | Indicator and criteria\nWater supply | \u2022 Improved* water point on school compound\n\u00a0\u00a0\u00a0\u00a0- Water point functional in the previous year (director reported)\n\u2022 Water tank to supply toilet and handwashing stations\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water observed in tank\nToilets | \u2022 At least one improved* toilet compartment\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is sex separated (by designation)\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is unlocked\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet is clean\n\u00a0\u00a0\u00a0\u00a0\u00a0- Toilet has water available inside compartment for flushing\nHandwashing facilities | \u2022 At least one individual handwashing station available to pupils\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water available at individual handwashing station\n\u00a0\u00a0\u00a0\u00a0\u00a0- Soap available at individual handwashing station\nPromotion of daily group hygiene activities | \u2022 Daily group handwashing schedule posted in at least one classroom or near toilet\n\u2022 Daily group compound cleaning schedule posted in at least one classroom or near toilet\n\u2022 Daily group toilet cleaning schedule posted in at least one classroom or near toilet\nGroup handwashing | \u2022 Group handwashing facility available to pupils\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water available at group handwashing facility\n\u00a0\u00a0\u00a0\u00a0\u00a0- Soap available at group handwashing facility\nWater filters | \u2022 At least one drinking water filter available in a classroom for pupil use\n\u00a0\u00a0\u00a0\u00a0\u00a0- Water in filter\nOutcome | Indicator\nToilet use | \u2022 Percentage of students using toilet for defecation during school hours (pupil-reported)\nHandwashing (individual) | \u2022 Percentage of students washing hands with soap and water upon exiting toilet (observation)\nDaily group handwashing | \u2022 School conducted daily group handwashing the day of visit (observation)\nDaily group toilet cleaning | \u2022 Percentage of students participating in daily group toilet cleaning within the previous five school days (pupil-reported)\nDaily group compound cleaning | \u2022 Percentage of students participating in daily group compound cleaning within the previous five school days (pupil-reported)\n\n*Defined according to Joint Monitoring Programme (JMP) standards.\n\nAssociation between WinS intervention and health and educational impacts, Saravane Province, Lao People\u2019s Democratic Republic, 2014-2017 (n\u2009=\u2009100 schools)\nImpact | Comparison* | Intervention* | Adjusted risk ratio | 95% confidence interval\nRoll-call absence\u2020 | 6024 (32.2%) | 7147 (29.9%) | 1.01 | 0.84, 1.20\nEnrollment\u2021 | 68.2 (49.7) | 71.6 (50.0) | 1.07 | 0.84, 1.35\nDropout\u2021 | 0.8 (2.6) | 0.4 (1.0) | 0.56 | 0.25, 1.24\nGrade progression\u2021 | 64.4 (48.6) | 67.3 (48.6) | 1.07 | 0.91, 1.25\nDiarrhea\u2020,\u00a7 | 1032 (21.1%) | 947 (14.7%) | 0.80 | 0.51, 1.26\nSymptoms of respiratory infection\u2020,\u2016 | 1414 (28.9%) | 2064 (32.1%) | 1.08 | 0.95, 1.23\nConjunctivitis\u2020,\u00a7 | 41 (0.8%) | 48 (0.8%) | 0.89 | 0.53, 1.52\nPrevalence of any STH\u2020,\u00b6 | 1833 (39.8%) | 1935 (41.6%) | 1.00 | 0.85, 1.17\n\n*Data are n (%) for impacts at the pupil level (roll-call absence, diarrhea, symptoms of respiratory infection, conjunctivitis, and prevalence of STH) and mean (SD) for impacts at the school-level (enrollment, dropout, grade progression) across study period.\n\u2020Risk ratios were calculated using a Poisson model with robust error variances and random intercepts at the school and pupil level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size, and season (rainy or dry). Absence model additionally controlled for and rice crop calendar (planting, growing, harvesting).\n\u2021Risk ratios were calculated using a Poisson model with random intercepts at the school level. All models adjusted for district and visit number.\n\u00a7Pupil-reported in previous week.\n\u2016Pupil-reported cough, runny nose, stuffy nose, or sore throat in previous week.\n\u00b6Samples testing positive for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus).\n\nAssociation between WinS intervention fidelity and adherence and absence, diarrhea, respiratory infection, and soil-transmitted helminth infection (STH), Saravane Province, Lao PDR, 2014-2017 (n\u2009=\u2009100 schools)\n | As-treated analysis | Structural nested model analysis\n | Adjusted risk ratio* | 95% confidence interval | Adjusted risk ratio\u2020 | 95% confidence interval\nRoll-call absence:\nFidelity\u2021 | 0.76 | 0.64, 0.91 | 0.97 | 0.33, 2.81\nAdherence\u2021 | 0.91 | 0.79, 1.05 | 0.96 | 0.19, 4.97\nDiarrhea:\u00a7\nFidelity, dry season\u2021 | 0.84 | 0.48, 1.49 | 0.45 | 0.24, 0.85\nAdherence, dry season\u2021 | 1.00 | 0.70, 1.44 | 0.42 | 0.21, 0.87\nFidelity, rainy season\u2021 | 1.65 | 0.82, 3.33 | 1.03 | 0.42, 2.51\nAdherence, rainy season\u2021 | 1.41 | 0.61, 3.26 | 0.50 | 0.19, 1.30\nSymptoms of respiratory infection\u2016:\nFidelity\u2021 | 1.00 | 0.89, 1.14 | 1.41 | 0.93, 2.13\nAdherence\u2021 | 0.97 | 0.84, 1.11 | 2.30 | 0.54, 8.87\nPrevalence of any STH:\u00b6\nFidelity\u2021 | 1.20 | 1.01, 1.43 | 1.10 | 0.57, 2.13\nAdherence\u2021 | 0.93 | 0.77, 1.12 | 1.18 | 0.37, 3.73\n\nSTH \u2013 soil-transmitted helminth\n*Risk ratios calculated using a Poisson model with robust error variances and random intercepts at the school and pupil level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size, season (rainy or dry). Absence models additionally controlled for rice crop calendar (planting, growing, harvesting).\n\u2020Risk ratios calculated using a Structural Nested Model with random intercepts at the school level. All models adjusted for district, visit number, pupil sex, pupil grade, school enrollment size.\n\u2021Fulfilling \u226575% of intervention outputs was considered fidelity. Fulfilling \u226575% of intervention outcomes was considered adherence.\n\u00a7Pupil-reported in previous week.\n\u2016Pupil-reported cough, runny nose, stuffy nose, or sore throat in previous week.\n\u00b6Samples testing positive for Ascaris lumbricoides, Trichuris trichuria, and hookworm (Ancyclostoma duodenale and Necatur americanus).",
        "label": "low",
        "id": "task4_RLD_test_782"
    },
    {
        "paper_doi": "10.1371/journal.pone.0006857",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Non-randomised controlled cluster trial\n\n\nParticipants: Country: TanzaniaSetting (coverage): 2 intervention districts, 1 control districtOutlets: Small drug shops (duka la dawa baridi)Age group: All age groups\n\n\nInterventions: Intervention: Subsidised ACT (AL)Comparison: No ACT subsidy (control)Supportive interventions: Behavior change communication (e.g. local radio advertisements, wall paintings, themed cultural shows) emphasising the importance of using ACTs and their availability in private shops\n\n\nOutcomes: ACT uptake, availability and price\n\n\nNotes: The project managers procured AL from the manufacturer, Novartis, and sold them to a pharmaceutical wholesaler in Dar es Salaam at an average of US$ $0.11 per dose, 88% below the price offered to public buyers.In one of the intervention districts (Kongwa), the suggested retail price intended to inform consumers of the maximum amount they should pay was set at 300, 600, 900, and 1200 Tanzanian shillings (0.25, 0.50. 0.75. and 1 USD respectively) for the four weight packs respectively; no suggested retail price was included on drugs distributed to Maswa in order to test its effect on price outcomes\n\n",
        "objective": "To assess the effect of programmes that include ACT price subsidies for private retailers on ACT use, availability, price and market share.",
        "full_paper": "Background\nWHO estimates that only 3% of fever patients use recommended artemisinin-based combination therapies (ACTs), partly reflecting their high prices in the retail sector from where many patients seek treatment.\nTo overcome this challenge, a global ACT subsidy has been proposed.\nWe tested this proposal through a pilot program in rural Tanzania.\nMethods/Principal Findings\nThree districts were assigned to serve either as a control or to receive the subsidy plus a package of supporting interventions.\nFrom October 2007, ACTs were sold at a 90% subsidy through the normal private supply chain to intervention district drug shops.\nData were collected at baseline and during intervention using interviews with drug shop customers, retail audits, mystery shoppers, and audits of public and NGO facilities.\nThe proportion of consumers in the intervention districts purchasing ACTs rose from 1% at baseline to 44.2% one year later (p<0.001), and was significantly higher among consumers purchasing for children under 5 than for adults (p\u200a=\u200a0.005).\nNo change in ACT usage was observed in the control district.\nConsumers paid a mean price of $0.58 for ACTs, which did not differ significantly from the price paid for sulphadoxine-pyrimethamine, the most common alternative.\nDrug shops in population centers were significantly more likely to stock ACTs than those in more remote areas (p<0.001).\nConclusions\nA subsidy introduced at the top of the private sector supply chain can significantly increase usage of ACTs and reduce their retail price to the level of common monotherapies.\nAdditional interventions may be needed to ensure access to ACTs in remote areas and for poorer individuals who appear to seek treatment at drug shops less frequently.\nTrial Registration\nControlled-Trials.com ISRCTN39125414.\nIntroduction\nArtemisinin-based combination therapies (ACTs) have become a mainstay of malaria treatment because of their high efficacy and their potential to delay the development of antimalarial resistance.\nYet despite availability of substantial donor funding, the proportion of fevers treated with ACTs remains minimal.\nThis partly reflects limited ACT availability outside the public sector.\nIn many countries, 40\u201360% of people seek treatment for fever or malaria from private vendors, such as pharmacies, drug shops and general stores.\nHowever, ACTs are typically sold at retail prices 20\u201340 times those of common alternatives such as amodiaquine and sulphadoxine-pyrimethamine (SP), restricting their uptake by consumers, particularly in rural areas.\nAs a result, most retail sector anti-malarial customers continue to use older therapies for which widespread resistance has been reported or artemisinin monotherapies, which are strongly discouraged by the World Health Organization because their use is likely to accelerate the development of artemisinin resistance.\nMany others purchase antipyretics only.\nAnticipating this challenge, in 2004 the Institute of Medicine recommended a global subsidy of ACTs as the best means to achieve high coverage and prolong the efficacy of these drugs.\nIt argued that reducing the ex-factory price of ACTs to that of common alternatives (roughly $0.05) would ensure their widespread distribution through private channels and crowd out other drugs.\nThis concept was further developed by the Roll Back Malaria Partnership and launched by the Board of the Global Fund in November 2008 as the Affordable Medicines Facility-malaria (AMFm).\nThe AMFm is scheduled to be launched in 11 countries within the coming year.\nHowever, the dearth of evidence on the likely impact of such a subsidy has hindered its design and raised numerous concerns about investing public money in an unproven mechanism.\nTo fill this evidence gap, we piloted the AMFm model in two rural Tanzanian districts starting in October 2007.\nThis report presents the results from the pilot and assesses their implications for implementation of the AMFm and other large-scale subsidies.\nMethods\nStudy design\nThe intervention was conducted in two rural districts of Tanzania: Maswa in Shinyanga region and Kongwa in Dodoma region.\nA third district, Shinyanga Rural in Shinyanga region, served as a control (see Figure 1).\nWe conducted a detailed analysis of all districts in the country according to a range of key indicators, including malaria endemicity, population per health facility, employment, prevalence of private drug shops, and bed net ownership.\nThe selected districts were among the few roughly comparable across all indicators, with high malaria transmission, large numbers of private drug shops and, importantly, no malaria-related trials (e.g., vaccines) underway.\nSocioeconomic status in the three districts was below national averages as evidenced by comparison of key assets such as housing materials, toilet facilities, and availability of electricity.\nThe selected districts were randomly assigned to one of the three arms in the study design: subsidy, subsidy plus suggested retail price, and no subsidy (control).\nAs two of the qualified districts were adjacent, randomization was limited so that one of the adjacent districts served as the control.\nThe project centered on the distribution of ACTs at highly subsidized prices.\nThe project managers procured quantities of artemether-lumefantrine (AL), the recommended first-line ACT in Tanzania, from the manufacturer, Novartis, and sold them to a pharmaceutical wholesaler in Dar es Salaam at an average of $0.11 per dose, 88% below the price offered to public buyers (private buyers are typically offered substantially higher prices).\nThe wholesaler, Nufaika Distributors Limited, was selected due to its extensive distribution network and interest in selling ACTs, among other factors, following due diligence of 10 similar businesses.\nThe wholesaler received no instructions other than to sell the ACTs to drug shops in the two intervention districts according to its standard practices.\nIt was made clear that the wholesaler would not be monitored or held accountable for its pricing, stocking, or other practices.\nSmall drug shops, known as duka la dawa baridi (DLDB), were the primary retail outlet for subsidized ACTs in the project.\nPast studies have found that these shops are the most important commercial source of anti-malarials in Tanzania and it is estimated that there are more than 8,000 in operation nationwide in rural and urban areas.\nUnder Tanzanian law, these shops are permitted to sell over-the-counter medicines, but not products requiring a prescription (an exception was granted for ACTs for this project).\nSmall volumes of anti-malarials are also sold through general stores alongside common staples, but distribution to these outlets was not encouraged as they are not legally permitted to sell pharmaceuticals.\nAL was distributed in four weight-specific packs, in redesigned packaging with simplified dosing instructions in the local language, Kiswahili.\nACTs distributed to Kongwa were marked with a suggested retail price (SRP) intended to inform consumers of the maximum amount they should pay.\nThe SRP was set at 300, 600, 900, and 1200 Tanzanian Shillings (0.25, 0.50.\n0.75. and 1 USD respectively) for the four weight packs respectively based on an analysis of costs in the supply chain; no SRP was included on drugs distributed to Maswa in order to test its effect on price outcomes.\nThe AMFm will support countries to conduct accompanying interventions such as training, behavior change communication (BCC), and regulatory strengthening to facilitate the effective distribution and use of subsidized ACTs.\nThis project accordingly included some of those activities in the intervention districts.\nPrior to the initiation of the subsidy, the Tanzania Food and Drug Authority (TFDA) conducted a one-day training of DLDB attendants focusing on malaria symptoms and ACT dispensing and dosing.\nPopulation Services International conducted a range of BCC activities, including local radio advertisements, wall paintings, and themed cultural shows, throughout the project.\nThe activities emphasized the importance of using ACTs and their availability in private shops, as well as basic messages on the dangers of malaria and the importance of prompt treatment-seeking.\nData collection\nThe study's primary outcomes were the proportion of antimalarial consumers visiting DLDB who purchased subsidized ACTs and the price they paid for the drugs.\nSecondary outcomes included the proportion of DLDB stocking the product, the socioeconomic status of the consumer, the age of the intended patient, and ACT distribution by public facilities during the same period.\nWe employed four data collection methodologies: exit interviews, retail audits, mystery shoppers, and public facility audits.\nAll four methodologies were conducted together five times during the project: once prior to the launch of the subsidy in August 2007 and four times throughout the intervention in November 2007 and March, August, and November 2008.\nData were collected at all DLDB and public facilities in the three districts.\nDLDB were initially identified through TFDA records, with unregistered shops captured through discussions with local informants and systematic physical reconnaissance throughout each district.\nGeographical positions of all shops were recorded using hand-held GPS units (Garmin Etrex).\nFor the exit interviews, data collectors positioned themselves near a DLDB and remained there for the full business day.\nAll customers emerging were asked to answer a short questionnaire on the products bought, and the brand of the product was visually verified.\nTo assess the customers' socio-economic status, interviewees were asked a series of questions about their household assets from the 2003\u201304 Tanzania HIV/AIDS Indicator Survey.\nDLDB retail audits were conducted twice during every data collection period, at a four week interval.\nCollectors recorded the volume of all anti-malarial stocks present.\nA short questionnaire was also administered to the owner or attendant to determine the amount of each product newly purchased and disposed of (e.g., due to expiry or damage) during the previous four weeks.\nStock levels were then compared between the two audits and purchases and disposals subtracted to determine the volume of sales during that period.\nDLDB pricing and dispensing practices were also observed by having collectors visit each shop once per survey posing as a consumer seeking malaria treatment.\nTwo such \u201cmystery shopper\u201d scenarios were employed \u2013 adults purchasing for themselves and purchasing for a nine-month old child at home with malaria symptoms.\nData collectors also visited all public and non-governmental organization (NGO) health facilities in each survey period to review ACT stocks and dispensing records.\nIn this paper we focus primarily on data from the baseline in August 2007 and follow-up in August 2008 because of potential seasonal variation in malaria transmission and treatment seeking.\nTo enable robust analysis of the impact of the SRP, pricing data are pooled across intervention surveys and compared to the baseline.\nA full set of results for all data collection periods is available at http://www.cshor.org/TanzaniaPilotData.xls.\nData analysis\nTo assess geographical variation in outcomes, the competition level of all DLDB was calculated using the fixed radius approach.\nThe competitive space of each DLDB was defined as 1 kilometer and each shop was assigned to a competition index category between 0 and 5 based on the number of other DLDB within that radius.\nExit interviewees were allocated to wealth quintiles using the asset weights and quintile break-points generated through principal components analysis of 53 variables from the 2003\u20132004 HIV/AIDS Indicator Survey.\nDue to operational challenges associated with administering asset ownership questions at shops (as opposed to at household level as is typical), several variables from the original survey were not captured, including additional forms of household lighting, water source, and land ownership status.\nAsset weights and national quintiles were therefore calculated using only those variables captured in the exit interviews.\nTo enable comparison of price across products, the exact number of pills purchased (syrups and injectables were excluded) by interviewees or mystery shoppers was recorded and the price paid for a full appropriate dose for the intended patient then extrapolated using the standard dosing schedule according to Tanzania national treatment guidelines or product registration with the TFDA.\nSimilarly, to compare sales across products, we converted the total number of pills sold to adult equivalent doses based on the recommended dosing.\nSurvey data were analyzed using SPSS v.14/16 (Chicago, Illinois) and SAS v.9.1 (Cary, NC), and GPS data using MapInfo v.7.8 (Troy, New York).\nProportions were compared using chi-square tests.\nStudent t-tests were used to compare means and the Satterthwaite t-test was used when variances were significantly unequal.\nA repeated measures multivariate regression model was used to compare differences in purchase price while controlling for potentially confounding factors and adjusting for clustering of multiple purchases in the same shops.\nEthical Considerations\nAs a pilot project of the Tanzania Ministry of Health and Social Welfare to prepare for a national program, the interventions were developed with the Tanzania Food and Drug Authority and National Malaria Control Program according to the policies and guidelines of the Ministry and approved by the Chief Medical Officer accordingly.\nNo additional interventions were added as part of the subsequent evaluation.\nOral informed consent was obtained from all consumers emerging from drug shops as well as from drug shop owners prior to the administration of retail audits.\nNo ethnic or individual identifying information was captured.\nThis study complied with the guidelines of the Declaration of Helsinki.\nResults\nTable 1 summarizes the observations recorded for each methodology.\nThe total number of DLDB audited increased from 200 in August 2007 to 216 in August 2008 due to the opening of new shops, with 30 (13%) and 39 (15%) additional shops closed or refusing to participate at these two time periods respectively.\nA similar number of all shops observed were in areas of high competition (36% and 38% in categories 4 and 5 in August 2007 and August 2008 respectively) as in areas of low competition (35% in categories 0 and 1 in both periods).\nHigh competition DLDB were located in population centers while those in low competition categories were 20\u201325 km from major roads.\nThe majority of consumers interviewed in all districts in August 2007 and August 2008 were from the two least poor SES quintiles (59% and 68% respectively).\nStocking\nThere was a pronounced increase in the proportion of shops stocking ACTs in the intervention districts, from 0/133 in August 2007 to 109/151 (72.2%) in August 2008 (p<0.001).\nShops stocking ACT in the control district changed negligibly from 1/67 (1%) to 0/65 over the same period.\nShops with two or more other shops in their competition radius were significantly more likely to stock ACTs in August 2008 (82/101, 81.2%) than those with 0 or 1 competitor (27/50, 54.0%; p<0.001).\nBy comparison, stocking of some other common anti-malarials was more consistent across competition categories: 75/101 (74.3%) of shops in category 2 and above and 34/50 (68.0%) in categories 0 and 1 stocked amodiaquine, a non-significant difference.\nPricing\nInterviewed consumers paid a mean price of $0.58 for all ACTs (SD\u200a=\u200a$0.28) during the study period (Figure 2).\nConsumers purchasing ACTs for children under 5 paid significantly less than those buying for adults (16+ years), at a mean expenditure of $0.35 (SD\u200a=\u200a$0.19) and $0.70 (SD\u200a=\u200a$0.28) respectively (p<0.001).\nOverall, the average price paid for ACTs for adults did not differ significantly from expenditures on SP ($0.67, SD\u200a=\u200a$0.34), but was significantly higher than for amodiaquine ($0.48, SD\u200a=\u200a$0.27; p<0.001).\nThe mean price paid for ACTs for children under 5 was significantly less than for both SP ($0.51, SD\u200a=\u200a$0.28; p\u200a=\u200a0.001) and AQ ($0.86, SD\u200a=\u200a$0.30; p<0.001).\nControlling for the age of the intended recipient, the district in which the drug was purchased, and clustering of multiple purchases in shops, the price paid for ACTs did not vary significantly by either the SES quintile of the consumer or the competition category of the shop.\nConsumers paid significantly more for the three largest of the four AL weight packs in Kongwa than in Maswa (all p<0.001), while no significant difference was observed for the lowest (5\u201315 kg) dose.\nIn Kongwa, the mean price paid for two AL doses, infant (5\u201315 kg) at $0.34 (SD\u200a=\u200a$0.24; p<0.001) and child (15\u201325 kg) at $0.55 (SD\u200a=\u200a$0.15; p\u200a=\u200a0.034), were significantly higher than the SRP.\nMean prices did not differ significantly from the SRP for the other two doses.\nUptake and Sales\nTable 2 presents exit interview data on uptake for AL, artemisinin monotherapy, and the two other most commonly purchased antimalarials, amodiaquine and SP.\nThe proportion of anti-malarial consumers in the intervention districts who purchased ACTs increased strikingly during the project, from 4/399 (1.0%) in August 2007 to 201/455 (44.2%) in August 2008 (p<0.001).\nThis proportion subsequently declined insignificantly to 227/572 (39.7%) in November 2008.\nOver the same period, purchases of SP and amodiaquine in the intervention districts declined significantly, from 232/399 (58.2%) to 163/455 (35.8%; p<0.001) and 146/399 (36.6%) to 75/455 (16.5%; p<0.001) respectively.\nUse of artemisinin monotherapies remained negligible throughout the study.\nPurchases by mystery shoppers followed a similar trend, with the proportion of shoppers offered ACTs in intervention districts rising from 6/133 (4.5%) to 83/146 (56.9%; p<0.001) and those offered SP declining from 83/133 (62.4%) to 38/146 (26.0%; p<0.001).\nWhen restricted to only shops stocking ACTs, the proportion of interviewed consumers and mystery shoppers buying ACTs was significantly higher than when all shops were included: 55.5% v. 44.2% (p\u200a=\u200a0.001) and from 76.4% v. 56.9% (p\u200a=\u200a0.001) respectively.\nNo change in ACT purchasing was observed in the control district.\nIn August 2008, 44/83 (53.0%) of consumers purchasing anti-malarials for a child under 5 bought ACTs compared to 104/291 (35.7%) of those purchasing for an adult (p\u200a=\u200a0.005).\nThere was no correlation between the SES of the consumer and the likelihood of buying ACTs, with ACTs comprising 44.4% of purchases by consumers in the poorest two quintiles (n\u200a=\u200a45) compared to 42.4% by those in the least poor quintiles (n\u200a=\u200a328).\nSimilarly, there was no significant difference in the proportion of consumers buying ACTs between high and low competition stores.\nRetail audits showed that 9,786 adult equivalent doses of ACTs (60.3% of all adult equivalent anti-malarial doses) were sold by DLDB in a 4-week period in July/August 2008, while ACT sales in the control district remained negligible (see Figure 3).\nOverall distribution of ACTs in the intervention districts, including dispensing from public and NGO facilities, increased 62% between November 2007 (the first collection following intervention) and August 2008 from 30,946 to 69,068 doses, with ACTs distributed through the private sector accounting for 38% of that growth.\nDiscussion\nThe introduction of subsidized ACTs resulted in a rapid and pronounced increase in the proportion of people accessing ACTs from private shops from close to zero to 44% after one year.\nDistribution of ACTs through the public sector rose at the same time, indicating that the intervention contributed to increases in overall volumes of the drug distributed.\nImportantly, the greatest increases in ACT usage at DLDB were by those seeking treatment for children under 5, the group at the greatest risk of malaria mortality.\nHowever, as other studies have found, purchases at DLDB for children under 5 were modestly underrepresented compared to the estimated fever incidence for this age group.\nThe rise in ACT usage appears to have crowded out the use of some sub-optimal therapies such as SP and AQ, although a substantial number of consumers continued to purchase these drugs at the end of the study.\nThere was no significant change in ACT uptake during the last 3 months of the study, but this does not suggest that a long-term plateau in ACT purchasing had been reached.\nThe intervention is targeting a fundamental shift in the market for anti-malarials, which will require years to fully realize.\nAs shown in other studies, poorer individuals appear to have sought treatment for malaria at DLDB substantially less frequently than wealthier ones, suggesting that additional interventions may be needed to increase ACT access among this population.\nThe subsidy had the intended effect of reducing the retail price of ACTs to levels similar to commonly used alternatives.\nOn average, consumers paid 93% less than ACT prices regularly observed in private outlets in rural and urban areas across Tanzania.\nContrary to competition theory and common concern about the AMFm, consumers did not pay significantly more for ACTs at more remote shops facing less competition.\nThese results also contradict another common critique of large-scale ACT subsidies: that businesses would apply excessive mark-ups thereby minimizing the benefit to consumers.\nThere were almost no instances of such \u201cprice gouging,\u201d with 86% of all purchases within $0.08 of the SRP for the dose and the highest price paid still 81% below typical ACT prices in Tanzanian shops.\nThe SRP appears to have had the opposite of the intended effect, artificially inflating prices in Kongwa above those determined by the market in Maswa.\nThe SRP levels were determined based on estimated costs and profit margins in the private anti-malarial supply chain derived from interviews with more than 50 businesses.\nThat those levels were substantially above the retail market prices suggests that an SRP should be used with caution, based on a more detailed understanding of pricing practices and only in cases where unreasonable profit margins are being charged.\nAlthough stocking of ACTs rose substantially during the intervention, it was skewed towards shops in towns and other population centers.\nWhen analysis was restricted to those shops stocking ACTs, purchases of the drug increased markedly, indicating that availability may have served as a major limitation on overall uptake.\nMany shopkeepers described high consumer demand for ACTs but expressed frustration at problems in obtaining the drugs.\nThis suggests that addressing supply chain issues should be a central focus of large-scale subsidy plans.\nIn particular, since wholesalers often lack an inherent financial incentive to distribute to remote outlets, public sector means of creating such incentives, such as providing a substantial rebate to wholesalers for achieving certain coverage levels in remote areas, should be explored.\nCaution should be used in directly applying these findings to other settings.\nSocioeconomic factors, malaria treatment-seeking behavior, and the structure of private supply chains all vary widely between and within countries.\nThis study operated through only one wholesaler and one type of retail outlet, while national-scale subsidies will employ multiple of both.\nAnd although concerted measures were put in place to limit the Hawthorne Effect, it is possible that businesses and consumers were influenced by the presence of the study team.\nNevertheless, these results are cause for cautious optimism that subsidies applied at the top of the private supply chain can lead to rapid and dramatic increases in ACT usage.\nConsumer demand for ACTs was high and average retail prices remained low due to businesses applying modest mark-ups on the product, similar to those for older drugs such as SP and amodiaquine.\nThe study also highlights key areas for further research on this topic.\nOverall changes in ACT coverage and treatment-seeking behavior, including among different SES groups, should be robustly assessed through studies collecting household level data.\nMoreover, the use of privately distributed ACTs for non-malarious fevers and opportunities for effectively introducing diagnosis into the private sector should be explored.\nHowever, the need for further research should not delay implementation.\nNo amount of piloting will fully recreate the conditions of a national or global subsidy, and some \u201clearning by doing\u201d will be inevitable.\nOur findings should provide sufficient confidence for large-scale implementation of the subsidy model as a central part of the global effort to increase ACT access from current dismal levels towards the Roll Back Malaria target of 80% of patients by the end of 2010.\nLocation of project districts by role in study.The red areas denote the two intervention districts, while the green area shows the control district.\nPrice paid for ACTs and most common alternative anti-malarials by interviewed consumers.Prices of subsidized ACTs and the most commonly purchased alternative anti-malarial \u2013 amodiaquine (AQ) or sulphadoxine-pyrimethamine (SP) \u2013 observed in the two intervention districts between November 2007 and November 2008 are displayed by intervention district and age of intended recipient. The thick blue line denotes the median and the red X the mean, with the surrounding box delineating the interquartile range (IQR). The lines extending from each box mark 1.5 times the IQR, with dots showing outliers that do not fall within this range.\nStocking and sales of subsidized ACTs at drug shops in intervention districts.All duka la dawa baridi (DLDB) in Maswa (top) and Kongwa (bottom) districts are mapped as either white (ACTs not in stock at time of survey) or black (ACTs in stock) circles. Data is shown at baseline (August '07) and two periods during implementation (November '07 and August '08). Districts are divided into 10 km2 squares, with the total volume of adult equivalent doses of ACTs sold in that area over the preceding four weeks shown by the color of shading as follows: Tan\u200a=\u200a1 to 50 doses sold; Orange\u200a=\u200a51 to 250 doses; Light red\u200a=\u200a251 to 500 doses; and Dark red\u200a=\u200a501 to 10,000 doses.\n\nRecorded observations by methodology, characteristic, and district, August 2007 and August 2008.\n | August 2007 | August 2008\n | Maswa | Kongwa | Control | Maswa | Kongwa | Control\nDLDB Audited | 73 | 60 | 67 | 83 | 68 | 65\nComp Index 0 | 12 | 7 | 15 | 13 | 20 | 14\nComp Index 1 | 12 | 15 | 9 | 10 | 7 | 12\nComp Index 2\u20133 | 21 | 18 | 20 | 28 | 18 | 13\nComp Index 4\u20135 | 3 | 5 | 6 | 0 | 10 | 18\nComp Index 5+ | 25 | 15 | 17 | 32 | 13 | 8\nExit Interviewees | 346 | 53 | 181 | 167 | 288 | 118\nBuying for Ages 16+ | 275 | 37 | 125 | 79 | 212 | 76\nBuying for Ages 5\u201316 | 28 | 7 | 10 | 39 | 42 | 10\nBuying for Ages<5 | 43 | 9 | 46 | 49 | 34 | 32\nSES Quintile 1 (poorest) | 3 | 4 | 21 | 8 | 12 | 4\nSES Quintile 2 | 27 | 7 | 40 | 16 | 9 | 8\nSES Quintile 3 | 62 | 8 | 66 | 32 | 50 | 9\nSES Quintile 4 | 137 | 26 | 45 | 56 | 112 | 52\nSES Quintile 5 (least poor) | 117 | 8 | 9 | 55 | 105 | 45\nMystery Shoppers | 73 | 60 | 67 | 81 | 65 | 65\nPublic/NGO facilities audited | 38 | 33 | 34 | 35 | 36 | 36\n\n\nPurchase of anti-malarials by exit interview customers by district, August 2007 and August 2008.\n | August 2007 | August 2008\n | Maswa | Kongwa | Total Intervention Districts | ControlDistrict | Maswa | Kongwa | Total Intervention Districts | Control District\nAdults (ages 16+)\nAntimalarial purchases of which: | 275 | 37 | 312 | 125 | 79 | 212 | 291 | 76\nACT | 3 (1%) | 1 (3%) | 4 (1%) | 0 (0%) | 38 (48%) | 66 (31%) | 104 (35%) | 0 (0%)\nSP | 187 (68%) | 26 (70%) | 213 (68%) | 78 (62%) | 30 (38%) | 118 (56%) | 148 (51%) | 63 (83%)\nAQ | 71 (26%) | 10 (27%) | 81 (26%) | 41 (33%) | 8 (10%) | 23 (11%) | 31 (11%) | 12 (16%)\nChildren (ages<5)\nAntimalarial purchases of which: | 37 | 6 | 43 | 35 | 49 | 34 | 83 | 32\nACT | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 20 (41%) | 24 (71%) | 44 (53%) | 2 (6%)\nSP | 1 (3%) | 2 (33%) | 3 (7%) | 3 (9%) | 1 (2%) | 2 (6%) | 3 (4%) | 3 (9%)\nAQ | 35 (95%) | 4 (67%) | 39 (91%) | 32 (91%) | 22 (45%) | 8 (24%) | 30 (36%) | 22 (69%)\n",
        "label": "low",
        "id": "task4_RLD_test_360"
    },
    {
        "paper_doi": "10.1371/journal.pntd.0007323",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: DesigncRCTAllocation of clusters\n90 clusters randomized to water; 90 to sanitation; 90 to hygiene; 90 to WASH; 180 to control\n\n\nParticipants: 3685 and 1706 children ages 2 to 12\n\n\nInterventions: Single WASH aspect and broad multiple\n\n\nOutcomes: Any STH; Ascaris lumbricoides; Trichuris trichiura; hookworm\n\n\nNotes: \n\n",
        "objective": "To assess the effectiveness of WASH interventions to prevent STH infection.",
        "full_paper": "Background\nSoil transmitted helminths (STH) infect >1.5 billion people.\nMass drug administration (MDA) effectively reduces infection; however, there is evidence for rapid reinfection and risk of potential drug resistance.\nWe conducted a randomized controlled trial in Bangladesh (WASH Benefits, NCT01590095) to assess whether water, sanitation, hygiene and nutrition interventions, alone and combined, reduce STH in a setting with ongoing MDA.\nMethodology/Principal findings\nIn 2012\u20132013, we randomized 720 clusters of 5551 pregnant women into water treatment, sanitation, handwashing, combined water+sanitation+handwashing (WSH), nutrition, nutrition+WSH (N+WSH) or control arms.\nIn 2015\u20132016, we enrolled 7795 children, aged 2\u201312 years, of 4102 available women for STH follow-up and collected stool from 7187.\nWe enumerated STH infections with Kato-Katz.\nWe estimated intention-to-treat intervention effects on infection prevalence and intensity.\nParticipants and field staff were not blinded; laboratory technicians and data analysts were blinded.\nPrevalence among controls was 36.8% for A. lumbricoides, 9.2% for hookworm and 7.5% for T. trichiura.\nMost infections were low-intensity.\nCompared to controls, the water intervention reduced hookworm by 31% (prevalence ratio [PR] = 0.69 (0.50,0.95), prevalence difference [PD] = -2.83 (-5.16,-0.50)) but did not affect other STH.\nSanitation improvements reduced T. trichiura by 29% (PR = 0.71 (0.52,0.98), PD = -2.17 (-4.03,-0.38)), had a similar borderline effect on hookworm and no effect on A. lumbricoides.\nHandwashing and nutrition interventions did not reduce any STH.\nWSH and N+WSH reduced hookworm prevalence by 29\u201333% (WSH: PR = 0.71 (0.52,0.99), PD = -2.63 (-4.95,-0.31); N+WSH: PR = 0.67 (0.50,0.91), PD = -3.00 (-5.14,-0.85)) and marginally reduced A. lumbricoides.\nEffects on infection intensity were similar.\nConclusions/Significance\nIn a low-intensity infection setting with MDA, we found modest but sustained hookworm reduction from water treatment and combined WSH interventions.\nImpacts were more pronounced on STH species with short vs. long-term environmental survival.\nOur findings suggest possible waterborne transmission for hookworm.\nWater treatment and sanitation improvements can augment MDA to interrupt STH transmission.\nTrial registration\nNCT01590095.\nAuthor summary\nSoil-transmitted helminths (STH) are associated with a large disease burden worldwide.\nMass administration of deworming drugs for preventive chemotherapy is the cornerstone of global strategy for STH control, but treated individuals often rapidly become reinfected, and there is also concern about emerging drug resistance.\nInterventions to treat drinking water, wash hands at critical times and isolate human feces from the environment through improved sanitation could reduce STH transmission by reducing the spread of ova from the feces of infected individuals into the environment and subsequently to new hosts.\nNutrition improvements could reduce host susceptibility to infection.\nExisting evidence on the effect of these interventions on STH is scarce.\nIn a setting with ongoing mass drug administration, we assessed the effect of individual and combined water, sanitation, handwashing and nutrition interventions on STH infection in children.\nApproximately 2.5 years after the initiation of interventions, we found reductions in STH infection from water treatment and sanitation interventions; there was no reduction from the handwashing and nutrition interventions.\nWhile the reductions were modest in magnitude compared to cure rates achieved by deworming drugs, they indicated sustained reduction in environmental transmission.\nThe reductions were more pronounced for STH species that do not have long-term environmental reservoirs.\nThese findings suggest that water treatment and sanitation interventions can augment mass drug administration programs in striving toward elimination of STH.\nIntroduction\nOver 1.5 billion people globally are infected with soil transmitted helminths (STH), specifically 819 million with Ascaris lumbricoides (roundworm), 465 million with Trichuris trichiura (whipworm), and 439 million with Necator americanus and Ancylostoma duodenale (hookworms).\nDeworming with mass drug administration (MDA) for preventive chemotherapy is the cornerstone of global WHO policy for STH control and effectively reduces infection.\nHowever, risk of drug resistance threatens the effectiveness of MDA programs given the wide-scale use, inadequate monitoring and limited number of effective anthelminthics, and frequent anthelminthic resistance in livestock.\nAdditionally, without environmental interventions to interrupt transmission, rapid reinfection is common; a systematic review demonstrated that prevalence reverts to 94% of pre-treatment levels for A. lumbricoides, 82% for T. trichiura and 57% for hookworm within 12 months post-treatment.\nWater, sanitation and hygiene improvements could potentially complement MDA programs in reducing STH transmission.\nTwo systematic reviews found reduced STH infection associated with improved water, sanitation and hygiene conditions in observational studies; however, there are few randomized assessments of the effect of water, sanitation and hygiene interventions on STH.\nSchool-based hygiene education trials have had mixed effects on STH.\nA health education program in Chinese schools improved handwashing practices and reduced STH infection while health hygiene education in schools in Peru reduced infections with A. lumbricoides but not T. trichiura and hookworm.\nHandwashing with soap and fingernail clipping reduced parasite infections in children in a trial in Ethiopia.\nTwo trials in India found no STH reduction from sanitation improvements, potentially because they did not attain sufficiently high latrine usage.\nIt is also possible that persistent long-term environmental reservoirs of STH ova sustain infections given the prolonged survival of some STH species, such as A. lumbricoides, in soil.\nWhile sanitation improvements should reduce immediate fecal input into the environment, their protective effect against STH infections may not be apparent until pre-existing ova in the environment are naturally inactivated.\nCombined water, sanitation and hygiene improvements targeting multiple transmission routes might achieve a larger impact by complementing the primary barrier of sanitation with the secondary barriers of water treatment and handwashing.\nSchool-based provision of combined water, sanitation and hygiene hardware reduced reinfection with A. lumbricoides but not other STH in a Kenyan trial.\nA recent community-based trial of integrated water, sanitation and hygiene interventions in addition to deworming in Timor-Leste found no impact on STH infections compared to deworming alone.\nThe effect of nutrition on STH infections is also poorly understood.\nImpaired immune function from nutritional deficiencies could increase host susceptibility to STH infection or exacerbate infection severity while nutritional supplements could also increase infection severity as excess nutrients are available for pathogens.\nA systematic review found mixed impacts of nutritional supplements on STH infection, concluding that the evidence is scarce and low-quality.\nNutrition interventions alongside water, sanitation and hygiene improvements could achieve synergistic benefits against STH infections.\nWe conducted a cluster-randomized trial (WASH Benefits, NCT01590095) in Bangladesh to assess the impact of individual and combined water, sanitation, handwashing (WSH) and nutrition interventions on child diarrhea and growth (primary and secondary outcomes).\nThe trial found that all interventions except for the individual water intervention reduced reported diarrhea, and all interventions with a nutrition component improved linear growth.\nHere, we report trial findings on STH infections (pre-specified tertiary outcomes) and test the hypotheses that (1) individual and combined WSH and nutrition interventions reduce STH infections, (2) combined WSH interventions reduce STH infections more than individual WSH interventions, and (3) combined nutrition and WSH interventions reduce STH infections more than nutrition or WSH interventions alone.\nThis work provides a novel investigation of the effect of improved WSH and nutrition on STH infections in a population with ongoing MDA to inform policy dialogue on whether these can complement MDA programs.\nMethods\nStudy setting\nThe trial was conducted in the Gazipur, Mymensingh, Tangail and Kishoreganj districts of central rural Bangladesh, which were selected because they had low groundwater arsenic and iron (to not interfere with the trial\u2019s chlorine-based water intervention) and no other water, sanitation, hygiene or nutrition programs.\nThe majority of households in the study area relied on untreated tubewell water for drinking, and while most households had access to on-site sanitation, it was common for latrines to drain into the environment, and animal feces were also commonly observed in the compounds.\nSince 2008, the Bangladesh Ministry of Health has implemented a school-based MDA program that provides deworming to school-aged children while pre-school-aged children receive deworming through the Expanded Program on Immunization (EPI).\nThe school-based MDA offers a single dose of mebendazole biannually while the EPI uses albendazole.\nIn a single dose, both drugs are effective for A. lumbricoides but have lower cure rates for T. trichiura; for hookworm, albendazole has a high cure rate while mebendazole has a modest cure rate.\nA 2010 evaluation of the national MDA campaign in two districts (not included in the WASH Benefits trial) found that 63\u201373% of school-attending children, 11\u201314% of non-school-attending school-aged children and 60% of pre-school-aged children received deworming.\nWASH Benefits activities were implemented independently from the MDA and EPI programs.\nRandomization and masking\nThe WASH Benefits trial enrolled pregnant women in their first or second trimester intending to stay in their village for 24 months post-enrollment, with the objective of following the birth cohort (\u201cindex children\u201d) born to them.\nField staff screened the study area for pregnant women and collected the global positioning system (GPS) coordinates of the compounds they lived in.\nEight neighboring eligible women were grouped into clusters using their compounds\u2019 GPS coordinates.\nCluster dimensions were chosen such that one field worker could visit all participants in a cluster in one day.\nA minimum 1-km buffer was enforced between clusters to minimize spillovers of infections and/or intervention behaviors between study arms.\nEvery eight adjacent clusters enrolled formed a geographic block.\nAn off-site investigator (BFA) used a random number generator to block-randomize clusters into study arms, providing geographically pair-matched randomization.\nParticipants and field staff were not blinded as interventions entailed distinct hardware; blinded technicians enumerated STH outcomes and blinded analysts (AE, JBC) independently replicated data management and analysis.\nDetails of the study design have been previously described.\nThe study protocol, pre-specified analysis plan, and a CONSORT checklist of trial procedures have been provided (S1\u2013S3 Text).\nInterventions\nStudy arms included (1) water treatment: chlorination with sodium dichloroisocyanurate (NaDCC) tablets coupled with safe storage in a narrow-mouth lidded vessel with spigot, (2) sanitation improvements: upgrades to concrete-lined double-pit latrines and provision of child potties and sani-scoops for feces disposal, (3) handwashing promotion: handwashing stations with a water reservoir and a bottle of soapy water mixture at the food preparation and latrine areas, (4) combined water treatment, sanitation and handwashing (WSH), (5) nutrition improvements including exclusive breastfeeding promotion (birth to 6 months), lipid-based nutrient supplements (6\u201324 months), and age-appropriate maternal, infant, and young child nutrition recommendations (pregnancy to 24 months), (6) nutrition plus combined WSH (N+WSH), and (7) a double-sized control arm with no intervention (Fig 1).\nFurther details of the interventions have been provided elsewhere.\nThe WSH interventions aimed to reduce children\u2019s early-life exposure to fecal pathogens.\nBangladeshi households are clustered in compounds shared by extended families; in our study, the compound containing the household where the index child lived (\u201cindex household\u201d) had an average of 2.5 households (range: 1\u201311).\nThe interventions targeted the compound environment as we expected this to be the primary exposure domain for young children.\nInterventions were delivered at index child, index household and compound levels (Fig 2).\nThe nutrition intervention targeted index children only.\nThe water and handwashing interventions were delivered to the index household.\nThe sanitation intervention provided upgraded latrines, potties and scoops to all households in the compound; as the shared compound courtyard serves as play space for children, we aimed to improve sanitary conditions in this environment with compound-level latrine coverage.\nBecause of the eligibility criterion of having a pregnant woman, enrolled compounds represented approximately 10% of compounds in a given geographical area; as such, we did not provide exclusive community-level latrine coverage.\nThe delivery of interventions was initiated around the time of index children\u2019s birth.\nLocal women hired and trained as community health promoters visited intervention arm participants on average six times per month to deliver intervention products for free, replenish the supply of consumables (chlorine tablets, soapy water solution, nutrient supplements), resolve hardware problems and encourage adherence to the targeted WSH and nutrition behaviors; health promoters did not visit control arm participants (S4 Text).\nThe health promotion visits and supply of consumable intervention products spanned the full study duration, including the period of the STH assessment.\nAll interventions had high user adherence throughout the study as measured by objective indicators (S4 Text).\nFurther details of intervention adherence have been previously reported.\nOutcome assessment\nWe assessed STH infections in children living in WASH Benefits compounds approximately 2.5 years after intervention initiation.\nHouseholds with no live birth or index child death were excluded from intervention promotion and subsequently follow-up.\nThe following children were eligible to enroll in the STH assessment: (1) all index children (aged 30 months on average at follow-up), (2) up to two other children per enrolled compound aged 3\u201312 years, living either in the index household or another household in the compound, enrolled in the preferential order of sibling of index child, child living in index household, or child living in another household in the compound, depending on availability.\nWe did not measure pre-intervention STH prevalence as index children were not born at the time and also because detection of any infections would ethically have required treatment and thus prevented us from testing our hypotheses under the typical conditions of our study population.\nRandomization was expected to ensure baseline balance in STH outcomes across study arms.\nAdditionally, protozoan infections with Giardia duodenalis, Cryptosporidium spp., and Entamoeba histolytica were measured at baseline among children aged 18\u201327 months (the anticipated age range for index children at the time of the STH assessment) to assess baseline balance on parasite infections between arms.\nTo measure STH outcomes, field staff distributed sterile containers to primary caregivers of enrolled children, instructed them to collect stool from the following morning\u2019s defecation event, and retrieved the containers on the morning of defecation.\nIf any enrolled child was absent or failed to provide a specimen, field staff returned to the household twice before classifying them as lost to follow-up.\nAfter the completion of stool collection in a given compound, all compound members were offered a single dose of albendazole.\nSpecimens without preservatives were transported on ice to the field laboratory of the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b) and analyzed on the same day.\nLaboratory staff were trained at the icddr,b parasitology laboratory using the Vestergaard Frandsen protocol to perform double-slide Kato-Katz and enumerate ova of A. lumbricoides, hookworm and T. trichiura.\nTwo slides were prepared from each stool sample and enumerated within 30 minutes of slide preparation.\n10% of slides were counted by two technicians (within the 30 minute-window since slide preparation), and 5% were counted by a senior parasitologist (by sending the slides to the icddr,b parasitology laboratory in Dhaka 0\u20134 days following the original count at the field laboratory) for quality assurance.\nTwo independent technicians double-entered slide counts into a database.\nEthics\nPrimary caregivers of children provided written informed consent.\nChildren aged 7\u201312 years provided written assent.\nThe protocol was approved by human subjects committees at University of California, Berkeley (2011-09-3652), Stanford University (25863), and the icddr,b (PR-11063).\nA data safety monitoring committee at icddr,b oversaw procedures.\nRegistration\nWASH Benefits was registered at ClinicalTrials.gov (NCT01590095) in April 2012 before trial enrolment began in May 2012; this registration lists the trial\u2019s primary and secondary outcomes (diarrhea, child growth).\nThe trial design was published in June 2013 before the STH follow-up began in May 2015 and lists STH under tertiary outcomes in Appendix 3.\nThe pre-specified analysis plan for the STH outcomes was registered at Open Science Framework (OSF, https://osf.io/v2c8p/) in August 2016 before data analysis began.\nStatistical analysis\nOutcomes\nOur pre-specified outcome measures were the infection prevalence, infection intensity and moderate/heavy infection prevalence for each STH species and for any of the three target species (\u201cany STH\u201d).\nFor each species, we classified stool samples with any ova as positive.\nWe quantified infection intensity in eggs per gram (epg) by multiplying the sum of egg counts from the two duplicated slides by 12 (the equivalent of the standard practice of multiplying slide counts by 24 in single-slide Kato-Katz).\nWe defined moderate/heavy intensity infections based on WHO categories (\u22655,000 epg for A. lumbricoides, \u22652,000 epg for hookworm, and \u22651,000 epg for T. trichiura).\nWe assessed the interrater agreement between two independent technicians and between a given technician and the senior parasitologist by calculating the kappa statistic for slides classified as positive.\nSample size\nWASH Benefits was designed to detect effects on child length and diarrhea with a planned sample size of 5040 pregnant women.\nWe assumed that two children per pregnant woman would be eligible for the STH assessment and 70% of children would provide stool.\nWe estimated STH prevalence and intra-class correlation coefficients (ICC) from the literature.\nWith a two-sided \u03b1 of 0.05, we had 80% power to detect the following relative reductions in prevalence between any intervention arm vs. control: 41% for A. lumbricoides, 50% for hookworm, 39% for T. trichiura, and 18% for any STH.\nStatistical parameters and estimation strategy\nWe compared STH outcomes in (1) individual and combined water, sanitation, handwashing and nutrition arms vs. controls (primary hypothesis), (2) combined vs. single WSH intervention arms, and (3) N+WSH vs. WSH and nutrition arms.\nWe estimated prevalence ratios (PR), prevalence differences (PD) and fecal egg count reductions (FECR, defined as the epg ratio minus one) between arms.\nWe estimated FECRs using geometric and arithmetic means; while geometric means prevent extreme data points from skewing means, arithmetic means are more sensitive to high infection intensities thought to correlate with higher morbidity burden and transmission.\nRandomization led to extremely good covariate balance, and our primary analysis therefore relied on unadjusted estimates.\nWe estimated the unadjusted parameters using targeted maximum likelihood estimation (TMLE) with influence-curve based standard errors treating clusters as independent units of analysis.\nSecondary analyses adjusted for pre-specified covariates using data-adaptive machine learning (see analysis plan).\nAnalyses were intention-to-treat as user uptake of interventions was high.\nAs per our pre-specified analysis plan, we did not adjust effect estimates for multiple comparisons as Bonferroni corrections and other multiplicity adjustments can be overcorrections, especially if the outcomes are correlated.\nAll analyses were conducted using R (version 3.3.2).\nSubgroup analyses\nOur primary analysis included all enrolled children of all ages.\nAs different interventions were implemented at index child, index household and compound levels, we also conducted subgroup analyses for the following three categories of children: (1) index children, (2) other children living in index household, and (3) children living in other households in compound.\nThe subgroup analysis for index children was pre-specified and the analysis for other children living in the index household vs. the rest of the compound was added post-hoc.\nWe also conducted a pre-specified subgroup analysis by child age (pre-school-age vs. school-age) as well as additional pre-specified subgroup analyses by deworming status, household size, wealth, housing materials, and baseline sanitation conditions (see analysis plan for details of subgroup analyses).\nMissing outcomes\nIndividuals that were lost at follow-up or failed to submit a specimen were classified as missing.\nTo assess if the likelihood of missing data was differential by study arm and/or covariates, we compared the percentage of missing observations between arms and the enrollment characteristics of those with available vs. missing specimens.\nWe also assessed the balance of baseline covariates between arms for households captured at follow-up.\nWe conducted a complete-case analysis and an inverse probability of censoring-weighted (IPCW) analysis re-weighting the measured population to reflect the original enrolled population (see analysis plan).\nResults\nEnrolment\nFieldworkers identified 13279 pregnant women in the study area (Fig 1).\nBetween May 2012-July 2013, we enrolled and randomized 5551 women in 720 clusters; the rest were excluded to create between-cluster buffers (n = 7429), were ineligible (n = 219) or refused (n = 80).\nAt the STH follow-up in May 2015\u20132016, 1449 women (26%) were lost because of no live birth (n = 361), index child death (n = 235), relocation (n = 375), absence (n = 182) or withdrawal (n = 296) (Fig 1).\nThe control arm had higher attrition (33%) than intervention arms combined (24%) as they had more withdrawals (12% vs. 3%).\nAmong 4102 (74%) available women, we enrolled 7795 children in the STH assessment (an average of 1.8 children per compound and 5.4 compounds per cluster), and we successfully collected stool from 7187 (92%) (Fig 1).\nStool recovery was somewhat lower in controls (87%) than in intervention arms (94%).\nHousehold-level enrolment covariates measured at baseline were balanced between arms for index households captured at follow-up (Table 1) and for those with vs. without specimens (S1 Table).\nThe prevalence of protozoan parasites measured among children aged 18\u201327 months at baseline was balanced between arms.\nChild-level characteristics\nAverage age at follow-up was 30 months (range: 22\u201338) for index children and 7 years (range: 3\u201312) for non-index children.\nCaregivers reported that 60% of index children and 67% of non-index children had been dewormed in the six months prior to our data collection.\nFor index children, the predominant source (86%) of deworming drugs was individual purchase at a pharmacy.\nFor non-index children, the predominant sources of deworming drugs were schools (48%) and pharmacies (44%), and approximately half of children were reported to be dewormed through the MDA program.\nThe average time since deworming was 9 weeks (range: 0\u201324) for both index and non-index children.\nThe percentage of dewormed children and the average time since deworming was balanced across trial arms (Table 2).\nCaregivers reported that 26% of index children and 8% of non-index children ingested soil in the last week, and 15% of index children and 25% of non-index children were observed to be wearing shoes.\nInfection prevalence and intensity\nSTH prevalence among all children in the control arm was 36.8% (n = 563) for A. lumbricoides, 9.2% (n = 142) for hookworm, 7.5% (n = 115) for T. trichiura, and 43.4% (n = 664) for any STH.\nThe geometric mean egg count among controls was 5.2 epg (143.4 epg among positive samples) for A. lumbricoides, 0.6 epg (119.8 epg among positive samples) for hookworm and 0.4 epg (115.4 epg among positive samples) for T. trichiura.\nMost infections were low-intensity.\nModerate/heavy infection prevalence among controls was 4.2% (n = 65) for A. lumbricoides, 0.1% (n = 2) for hookworm and 0.4% (n = 6) for T. trichiura.\nThe ICC for any STH infection was 0.18 for children within the same compound (0.18 for A. lumbricoides, 0.03 for hookworm and 0.21 for T. trichiura) and 0.08 for children within the same cluster of compounds (0.08 for A. lumbricoides, 0.06 for hookworm and 0.13 for T. trichiura).\nThe quality control measures indicated high interrater reliability (S5 Text).\nInterventions vs. control\nAmong all enrolled children, the single water intervention reduced hookworm prevalence by 31% (PR = 0.69 (0.50, 0.95), PD = -2.83 (-5.16, -0.50)) from a control prevalence of 9.2% but had no effect on other STH (Fig 3, Table 3, S2 Table).\nThe sanitation intervention reduced T. trichiura prevalence by 29% (PR = 0.71 (0.52, 0.98), PD = -2.17 (-4.10, -0.24)) from a control prevalence of 7.5% and achieved a similar borderline reduction on hookworm but had no effect on A. lumbricoides.\nSingle handwashing or nutrition interventions did not reduce the prevalence of any STH compared to controls; there was a borderline increase in A. lumbricoides prevalence in these arms (Fig 3, Table 3, S2 Table).\nCombined WSH reduced hookworm prevalence by 29% (PR = 0.71 (0.52, 0.99), PD = -2.63 (-4.95, -0.31)) and N+WSH by 33% (PR = 0.67 (0.50, 0.91), PD = -3.00 (-5.14, -0.85)) compared to controls.\nWSH and N+WSH also marginally reduced A. lumbricoides by 7\u201310% compared to controls (WSH: PR = 0.93, (0.83, 1.05), PD = -2.47 (-6.54, 1.60); N+WSH: PR = 0.90 (0.81, 1.01), PD = -3.79 (-7.74, 0.17)) but had no effect on T. trichiura (Fig 3, Table 3, S2 Table).\nCombined vs. single interventions\nCompared with single water, sanitation and handwashing interventions, combined WSH reduced A. lumbricoides more than handwashing alone (PR = 0.85, (0.75, 0.96), PD = -6.21 (-10.96, -1.46); this reflects the increased A. lumbricoides prevalence in the handwashing arm.\nWe found no other benefit from combined WSH vs. its individual components (Fig 3, S3 Table).\nCombined N+WSH reduced A. lumbricoides and hookworm prevalence compared to nutrition alone (A. lumbricoides: PR = 0.82 (0.72, 0.93), PD = -7.31 (-12.06, -2.57)); hookworm: PR = 0.65 (0.47, 0.91), PD = -3.29 (-5.96, -0.61)) but did not achieve any reduction compared to WSH (Fig 3, S4 Table).\nOther effects\nEffects on any STH closely reflected A. lumbricoides results due to the high prevalence of A. lumbricoides compared to the other two species (Fig 3, S2\u2013S4 Tables).\nInterventions did not affect the prevalence of moderate/heavy infections (S5\u2013S7 Tables) but we had low power for these rare outcomes.\nEffects on infection intensity were similar to effects on prevalence, except for a modest reduction in T. trichiura intensity from handwashing (Fig 4, S8\u2013S10 Tables).\nArithmetic means yielded similar results with wider confidence intervals (S8\u2013S10 Tables).\nUnadjusted, adjusted and IPCW estimates were similar (S2\u2013S10 Tables).\nSubgroup analyses\nSubgroup analyses on index children, other children in the index household and children in other households in the compound yielded findings consistent with those using pooled data from all children.\nNon-index children had higher infection prevalence (Table 3), consistent with previous evidence on these age groups.\nThe water intervention, which was implemented in the index household and showed a reduction in hookworm when using data from all enrolled children, substantially reduced hookworm among the older children living in the index household (PR = 0.59 (0.39, 0.90)) but not among index children themselves who had lower infection prevalence and also might have consumed less water due to their younger age (PR = 0.95 (0.52, 1.71)), nor among children in other households in the compound whose own households did not receive the water intervention (PR = 0.75 (0.37, 1.51)).\nThe handwashing intervention, which was implemented in the index household and did not achieve a reduction when using data from all children, also did not achieve a reduction among children living in the index household (Table 3).\nSimilarly, the nutrition intervention, which was provided to index children only and did not achieve a reduction when using data from all enrolled children, also did not achieve a reduction among index children (Table 3).\nAs expected, the reduction on hookworm and T. trichiura from the compound-level sanitation intervention was similar among all children in the compound (Table 3); however, most confidence intervals crossed the null, reflecting the small sample sizes of the subgroups.\nPoint estimates suggested that the WSH and N+WSH interventions, which reduced hookworm among all children, achieved similar reductions in all three subgroups of children but, once again, most confidence intervals crossed the null due to small sample size (Table 3).\nOur full set of subgroup findings are reported elsewhere (https://osf.io/v2c8p/); we note that these analyses should be considered exploratory as they had limited statistical power.\nDiscussion\nEffect of water treatment\nIn all intervention arms that included water treatment (the individual water treatment, WSH and N+WSH arms), we found a significant reduction in hookworm but not in A. lumbricoides or T. trichiura.\nThese findings suggest possible waterborne hookworm transmission, and chlorine water treatment with safe storage is effective in reducing this transmission.\nWhile infections of A. lumbricoides or T. trichiura are transmitted by ingesting embryonated ova, hookworm larvae infect hosts by penetrating skin; however A. duodenale can also be transmitted by ingesting larvae.\nA study using molecular diagnostics on infant stool samples in Dhaka detected A. duodenale several times more frequently than N. americanus, suggesting that A. duodenale could be the dominant hookworm species in the region.\nSTH ova have been detected in drinking water in low-income countries, suggesting a potential reservoir.\nA systematic review reported that boiling and filtering water was associated with reduced STH infection.\nWhile chlorination is generally considered ineffective against STH ova, fragile hookworm ova and larvae could be more chlorine-susceptible than the hardier ova of A. lumbricoides or T. trichiura.\nThe safe storage container with a narrow mouth and lid would also reduce STH contamination of stored drinking water by eliminating contact with hands, which are known reservoirs of ova and larvae.\nAn observational study found increased STH infection associated with unhygienic water storage.\nStorage could also allow eggs to settle out of the water column before consumption.\nHowever, while safe storage should similarly affect all three STH species, A. lumbricoides or T. trichiura were not reduced by our water intervention, potentially suggesting that the hookworm reduction is due to chlorine rather than safe storage; there are scarce data on the effectiveness of chlorine on hookworm.\nAlternatively, waterborne transmission could be more important for hookworm in this setting than for A. lumbricoides or T. trichiura.\nEffect of sanitation\nThe WASH Benefits sanitation intervention with concrete-lined double-pit latrines, potties and scoops for feces management reduced T. trichiura and achieved a borderline reduction in hookworm but had no effect on A. lumbricoides.\nWhile the other two arms with a sanitation component (WSH, N+WSH) reduced hookworm to a similar degree as the single sanitation intervention, T. trichiura was not affected in these arms; the reduction in the sanitation arm for this species could thus be a chance finding and should be interpreted cautiously.\nTwo previous sanitation trials in India found no impact on STH; however, both studies entailed community-level programs with relatively low adherence.\nWASH Benefits implemented a compound-level intervention with high adherence.\nAt follow-up, >95% of respondents in the sanitation, WSH and N+WSH arms had a latrine with a functional water seal compared to <25% of controls.\nIn structured observations, >90% of adults in sanitation arms used a hygienic latrine vs. 40% of controls.\nLow adherence is therefore unlikely to explain any lack of intervention impact.\nHowever, it is possible that structured observations overestimated actual latrine use due to respondent reactivity.\nAlso, children continued open defecation despite sanitation access; only 37\u201354% of young children in sanitation arms were observed to defecate in a latrine or potty vs. 32% of controls.\nFinally, WASH Benefits intervened on roughly 10% of compounds within a geographical area and did not implement exclusive community-level latrine coverage.\nBangladesh has a high population density, and STH ova from surrounding non-study compounds could have entered intervention compounds on shoes/soles of compound residents or via surface runoff.\nBangladeshi families also use soil from outside the compound to coat walls and courtyards.\nCommunity-level sanitation coverage may be more instrumental in improving child health than individual household sanitation in rural settings; it is possible that high community-level sanitation coverage is needed to more broadly impact STH infections.\nThe lack of sanitation impact on A. lumbricoides could also be due to its prolonged survival in soil, providing a persistent reservoir to sustain infection.\nHookworm larvae survive in soil for weeks and T. trichiura ova for months.\nIn contrast, A. lumbricoides ova can survive in soil for several years in warm and saturated conditions.\nA pilot assessment among study households found ova of A. lumbricoides in 67% and T. trichiura in 36% of courtyard soil samples; hookworm ova were not detected in soil (the protocol was not designed to detect hookworm larvae).\nAmong positive samples, 70% of A. lumbricoides and 88% of T. trichiura ova developed larvae when incubated (i.e., were viable).\nSoil from study households also had human and animal fecal markers and high concentrations of fecal indicator bacteria, suggesting heavy fecal contamination.\nAny reductions in fecal input into the environment through sanitation interventions would plausibly be reflected in a more immediate reduction in infections with hookworm and T. trichiura whose larvae/ova are shorter-lived in the environment than those of A. lumbricoides, which is consistent with our findings.\nThe effect of sanitation on A. lumbricoides infections may not be apparent until existing ova in the environment from pre-intervention contamination are naturally inactivated.\nEffect of handwashing\nHandwashing did not reduce STH infection except for a modest reduction in T. trichiura intensity.\nPrevious hygiene programs have reduced STH infections.\nTwo of the previous studies were conducted in schools, which may have fewer sources of fecal contamination than the domestic environment and therefore lower risk of re-contamination of hands following handwashing.\nIt is also possible that these studies achieved better handwashing than WASH Benefits, potentially because our promotion primarily targeted caregivers rather than children, some of whom were too young to wash their own hands.\nA study in Bangladesh found A. lumbricoides ova in 51%, T. trichiura ova in 23% and hookworm larvae in 26% of fingernails, and nail clipping reduced parasite infections among Ethiopian children.\nWhile our intervention promotion mentioned washing with soap under fingernails, the practices adopted by participants may not have been sufficient to remove ova/larvae from under nails.\nIndeed, hand observations showed that children in the handwashing arm had cleaner finger pads and palms than children in the control arm but there was little difference in the visual cleanliness of fingernails.\nThis could also explain why the T. trichiura intensity but not prevalence was reduced in the handwashing arm; if the intervention reduced but did not eliminate T. trichiura ova on hands, this could lead to a reduced worm burden without affecting prevalence.\nEffect of combined WSH interventions\nWe found no added benefit from combining WSH interventions compared to individual interventions.\nWhile we did not power the study to statistically detect differences between combined vs. individual intervention arms, the effect estimates suggest that combined WSH and N+WSH achieved a similar magnitude of reduction in hookworm prevalence as the individual water and sanitation interventions.\nOne possible explanation is that combined interventions might have lower user adherence as they require more complex behavior change.\nHowever, adherence indicators were similar between individual and combined intervention arms in our study.\nIt is also possible that the primary barrier of sanitation (reducing spread of ova into water sources) and the secondary barrier of water treatment (reducing ingestion of ova/larvae) were operating on the same waterborne transmission pathway and there was thus no benefit from combining them.\nNonetheless, combined WSH was the only intervention that achieved a small (albeit borderline non-significant) reduction in A. lumbricoides compared to controls.\nEffect of nutrition\nLipid-based nutrient supplements, breastfeeding and complementary feeding promotion did not reduce STH prevalence/intensity alone or in combination with WSH, even when the analysis was restricted to index children directly receiving this intervention.\nThere was a borderline increase in A. lumbricoides prevalence in the nutrition arm.\nA recent study found increased hookworm in school children receiving micronutrient-fortified rice.\nOther studies found STH reductions from nutritional supplements.\nWASH Benefits showed improved child growth and micronutrient status and reduced anemia in the nutrition and N+WSH arms.\nOne reason for the lack of STH reduction despite improved nutrition could be the dual direction of possible biological associations between nutrition and STH infection.\nBreastfeeding and improved nutrition could decrease infection risk by improving immune response and cell repair; conversely, it could increase risk by making nutrients available to helminths.\nChronic heavy STH infections can lead to malnutrition and growth faltering.\nHookworm reductions in the water and WSH arms were not reflected by improved growth in these arms.\nHowever, children in the WSH arm had a borderline reduction in anemia and iron deficiency, which would be consistent with the reduction in hookworm prevalence in this arm.\nNutritional outcomes are likely interrelated with myriad causal effects and the impact of STH on growth should be further assessed.\nFindings in the context of MDA\nSTH control policies, which currently emphasize MDA programs, could be strengthened by complementing these with water, sanitation and hygiene interventions.\nApproximately two thirds of children in our study had received deworming within the 6 months prior to our STH assessment, and among these, the average time since deworming was approximately two months.\nIt is possible that, had more time elapsed since the last reported deworming episode at the time of our assessment, allowing time for more prevalent reinfection, our interventions might have demonstrated a larger difference between intervention and control groups.\nIndeed, the majority of infections we detected were low-intensity, suggesting that deworming successfully reduced the prevalence of heavy infections that drive the morbidity burden.\nHowever, despite recent deworming, 43% of children in the control arm were infected with STH (mostly A. lumbricoides), demonstrating ongoing transmission and suggesting that MDA alone is unlikely to break STH transmission in this setting.\nAgainst this backdrop, we found a 30% relative reduction in hookworm prevalence from water treatment and combined WSH interventions, as well as a borderline reduction of similar size from sanitation improvements.\nWhile the reductions were comparable in magnitude to the reductions in child diarrhea and protozoan infections achieved by WASH Benefits and other water treatment and hygiene trials in low-income countries, they are small compared to the typical cure rates from deworming.\nHowever, while re-infection rates following deworming can be as high as 94% within 12 months of drug administration, the effects we report were observed 2.5 years after intervention initiation, suggesting sustained reductions in environmental transmission in a population receiving biannual MDA.\nIt is also possible that the effect of water, sanitation and hygiene on STH depends on background transmission intensity.\nIn our study, WSH interventions had more pronounced effect on hookworm, which was relatively rare (9% control prevalence), than on A. lumbricoides, which was more common (37% control prevalence).\nThis is consistent with a school-based trial in Kenya that found reduction in A. lumbricoides (9\u201314% prevalence in the study population) but not the more prevalent hookworm (28\u201329% prevalence) from a combined water, sanitation and hygiene intervention.\nThese findings suggest that in settings where deworming has been successfully implemented to reduce infection intensity and morbidity, water, sanitation and hygiene interventions can complement MDA programs in striving toward elimination by interrupting environmental transmission.\nLimitations\nWe measured STH infection using Kato-Katz, which has poor sensitivity when infection intensity is low.\nA systematic review demonstrated a sensitivity of 55% for A. lumbricoides, 53% for hookworm, 80% for T. trichiura for double-slide Kato-Katz for low-intensity infections.\nAs 95% of infections in our study were low-intensity, this could yield substantial false negatives in our outcome measurements.\nRecently developed sensitive nucleic acid-based diagnostics can detect infections that are missed by Kato-Katz.\nWe preserved an additional stool aliquot for validation analysis by quantitative polymerase chain reaction (qPCR).\nPreliminary analyses in a validation study using a subset of our specimens suggest that double-slide Kato-Katz had low to moderate sensitivity for all three STH while it had moderate specificity for A. lumbricoides and high specificity for T. trichiura and hookworm (Benjamin-Chung et al. 2019, in prep).\nAssuming non-differential misclassification by arm, imperfect sensitivity and specificity would bias our estimated intervention effects toward the null (S6 Text).\nIf the interventions reduced infection intensity, imperfect sensitivity could also lead to differential misclassification by arm, where a larger proportion of cases in the intervention arms would go undetected by Kato-Katz and intervention effects would therefore be biased away from the null.\nAlso, WASH Benefits was designed around its primary outcomes (length-for-age Z-score and diarrhea) so there was only sufficient statistical power to detect relatively large effects on hookworm and T. trichiura given their low prevalence.\nPost-hoc calculations suggested a minimum detectable effect of 19% relative reduction for A. lumbricoides, 41% for hookworm, and 52% for T. trichiura.\nFuture studies in low-prevalence settings should enroll sample sizes large enough to detect small effects and use sensitive diagnostics.\nWe conducted multiple comparisons, increasing the risk of chance findings; the T. trichiura reduction in the sanitation but not WSH and N+WSH arms could indicate random error.\nHowever, most observed reductions followed consistent patterns that are unlikely to be explained by chance.\nHookworm prevalence and intensity showed internally consistent reductions of similar size in all arms with a water or sanitation component, while the only intervention that achieved a borderline reduction in A. lumbricoides was combined WSH\u2014the most biologically plausible intervention to reduce environmental transmission.\nAnother limitation is that we assessed STH outcomes 2.5 years after intervention initiation, which is a relatively short period of time to assess impact on A. lumbricoides given its long survival in soil.\nThis timeframe risks underestimating the long-term population benefit of reducing environmental soil contamination through improved sanitation.\nLonger-term follow-up of this population might provide a more accurate assessment of the long-term contribution of improved sanitation towards A. lumbricoides elimination.\nEnvironmental conditions such as temperature, humidity and soil type affect the fate and transport of STH ova and intervention effects are therefore likely to be setting-dependent.\nWe controlled for month in our analysis to adjust for seasonality.\nAlso, our geographically pair-matched randomization synchronized the timing of outcome measurement between arms, eliminating confounding from season as well as from unmeasured spatiotemporal factors.\nHowever, our findings may not be generalizable to other settings with different climatic and geological conditions, or different levels of fecal contamination in the ambient environment.\nSimilarly, our findings are relevant to other populations with MDA programs and relatively low intensity of STH infection.\nFuture studies should investigate the effect of water, sanitation and hygiene improvements on STH infection and how these can augment MDA programs in high-intensity infection settings.\nConclusions\nIn a setting with ongoing MDA and low-intensity infections, we found modest but sustained reductions in hookworm prevalence and intensity from water treatment and combined WSH interventions.\nThere was no STH reduction from handwashing and nutrition improvements.\nIntervention effects were more pronounced on hookworm than on A. lumbricoides and T. trichiura; this could be because of the short survival of hookworm in soil, precluding persistent environmental reservoirs of ova from pre-intervention contamination.\nOur findings suggest that drinking water can potentially be an overlooked transmission route for hookworm and that water treatment and sanitation interventions can augment MDA programs in striving towards breaking STH transmission.\nFlowchart of study participation.The STH assessment enrolled (1) the birth cohort born to the enrolled women (referred to as \u201cindex child\u201d), (2) children living in the same household as the index child (referred to as \u201cindex HH child\u201d) and (3) children living in the same multi-household compound as the index child (referred to as \u201cother HH child\u201d).\nInterventions implemented at index child, index household, and compound levels.Index child refers to the birth cohort born to enrolled pregnant women. Index household refers to the household where the index child lived. Each enrolled compound contained a single index household and an average of 2.5 households total. Index children were the primary intervention recipients of the nutrition intervention delivered in the nutrition and N+WSH arms. Index households were the primary recipients of the water treatment and handwashing interventions; in the water, WSH and N+WSH arms they received water treatment products, and in the handwashing, WSH and N+WSH arms they received handwashing stations. The sanitation intervention was delivered to the entire compound; all households in compounds enrolled in the sanitation, WSH and N+WSH arms received latrine upgrades, child potties and scoops.\nPrevalence and prevalence ratio (PR) for A. lumbricoides, hookworm, and T. trichiura infection in children aged 2\u201312 years measured with double-slide Kato-Katz 2.5 years after intervention initiation. Vertical bars indicate 95% confidence intervals.\nGeometric mean eggs per gram for A. lumbricoides, hookworm and T. trichiura in children aged 2\u201312 years measured with double-slide Kato-Katz 2.5 years after intervention initiation.\n\nHousehold-level enrolment characteristics by intervention group measured at baseline (for index households that participated in STH assessment at follow-up). a\n\u00a0 | Control | Water | Sanitation | Handwashing | WSH | Nutrition | N+WSH\nNo. of women: | (N = 929) | (N = 550) | (N = 547) | (N = 539) | (N = 523) | (N = 491) | (N = 523)\nMaternal |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Age, mean (range) | 24 (15\u201343) | 24 (15\u201343) | 24 (15\u201341) | 24 (15\u201360) | 25 (15\u201344) | 24 (15\u201345) | 24 (14\u201343)\n\u00a0\u00a0\u00a0\u00a0Years of education, mean (range) | 6 (0\u201315) | 6 (0\u201314) | 6 (0\u201317) | 6 (0\u201316) | 6 (0\u201314) | 6 (0\u201316) | 6 (0\u201314)\nPaternal |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Years of education, mean (range) | 5 (0\u201316) | 5 (0\u201316) | 5 (0\u201317) | 5 (0\u201316) | 5 (0\u201316) | 5 (0\u201316) | 5 (0\u201316)\n\u00a0\u00a0\u00a0\u00a0Works in agriculture, % (n) | 31.4 (292) | 31.5 (173) | 30.7 (168) | 37.5 (202) | 31.0 (162) | 33.6 (165) | 31.2 (163)\nHousehold |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Number of persons, mean (range) | 5 (2\u201317) | 5 (2\u201323) | 5 (2\u201317) | 5 (2\u201322) | 5 (1\u201314) | 5 (2\u201318) | 5 (2\u201314)\n\u00a0\u00a0\u00a0\u00a0Has electricity, % (n) | 57.9 (538) | 62.7 (345) | 60.5 (331) | 59.7 (322) | 63.1 (330) | 61.3 (301) | 60.6 (317)\n\u00a0\u00a0\u00a0\u00a0Has a cement floor, % (n) | 10.0 (93) | 12.0 (66) | 12.1 (66) | 8.0 (43) | 10.7 (56) | 8.6 (42) | 12.1 (63)\n\u00a0\u00a0\u00a0\u00a0Acres of agricultural land owned, mean (range) | 0.1 (0.0\u20132.5) | 0.1 (0.0\u20132.4) | 0.1 (0.0\u20133.2) | 0.1 (0.0\u20132.6) | 0.2 (0.0\u20133.1) | 0.2 (0.0\u20132.8) | 0.2 (0.0\u20138.9)\nDrinking water |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Shallow tubewell primary water source, % (n) | 76.5 (711) | 73.5 (404) | 75.1 (411) | 70.3 (379) | 79.0 (413) | 75.2 (369) | 74.0 (387)\n\u00a0\u00a0\u00a0\u00a0Stored water observed at home, % (n) | 46.6 (433) | 51.1 (281) | 47.4 (259) | 48.8 (263) | 41.5 (217) | 41.8 (205) | 48.0 (251)\n\u00a0\u00a0\u00a0\u00a0Reported treating water yesterday, % (n) | 0.3 (3) | 0.2 (1) | 0.0 (0) | 0.2 (1) | 0.0 (0) | 0.0 (0) | 0.4 (2)\nSanitation |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Daily defecating in the open, % (n) |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Adult men | 7.2 (67) | 5.3 (29) | 6.6 (36) | 9.8 (53) | 6.5 (34) | 7.3 (36) | 7.5 (39)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Adult women | 4.7 (44) | 2.6 (14) | 4.2 (23) | 5.2 (28) | 4.0 (21) | 5.30 (26) | 3.8 (20)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Children: 8-<15 years (N = 1743) | 10.2 (38) | 9.5 (21) | 8.9 (22) | 15.5 (37) | 8.3 (19) | 8.3 (17) | 9.5 (22)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Children: 3-<8 years (N = 2179) | 40.0 (197) | 36.0 (111) | 37.2 (109) | 37.7 (110) | 35.5 (99) | 35.1 (85) | 36.3 (99)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Children: 0-<3 years (N = 848) | 80.5 (157) | 85.6 (89) | 81.1 (86) | 85.5 (100) | 78.0 (92) | 82.5 (85) | 88.6 (93)\n\u00a0\u00a0\u00a0\u00a0Latrine, % (n) |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Owned b | 53.4 (496) | 52.9 (291) | 53.4 (292) | 54.6 (294) | 53.0 (277) | 54.2 (266) | 54.1 (283)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Concete slab | 90.4 (840) | 92.7 (510) | 88.3 (483) | 89.6 (483) | 90.1 (471) | 90.0 (440) | 90.3 (472)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Functional water seal | 25.3 (235) | 26.4 (145) | 26.0 (142) | 25.4 (137) | 21.0 (110) | 26.5 (130) | 22.6 (118)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Visible stool on slab or floor | 48.6 (451) | 44.9 (247) | 44.8 (245) | 43.8 (236) | 52.6 (275) | 46.2 (227) | 49.3 (258)\n\u00a0\u00a0\u00a0\u00a0Owned a potty, % (n) | 3.4 (32) | 3.8 (21) | 3.8 (21) | 5.0 (27) | 3.6 (19) | 5.3 (26) | 4.8 (25)\n\u00a0\u00a0\u00a0\u00a0Human feces observed in, % (n) |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0House | 9.0 (84) | 9.6 (53) | 7.7 (42) | 10.6 (57) | 7.1 (37) | 7.1 (35) | 6.9 (36)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Child's play area | 1.4 (13) | 1.1 (6) | 0.9 (5) | 1.1 (6) | 0.8 (4) | 0.6 (3) | 1.2 (6)\nHandwashing |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Has within 6 steps of latrine, % (n) |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Water | 12.8 (119) | 12.0 (66) | 11.9 (65) | 8.5 (46) | 8.2 (43) | 8.6 (42) | 11.7 (61)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Soap | 5.5 (51) | 6.9 (38) | 7.5 (41) | 4.8 (26) | 4.6 (24) | 4.5 (22) | 5.9 (31)\n\u00a0\u00a0\u00a0\u00a0Has within 6 steps of kitchen, % (n) |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Water | 8.5 (79) | 6.6 (36) | 7.3 (40) | 5.8 (31) | 8.2 (43) | 9.1 (45) | 8.8 (46)\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Soap | 2.4 (22) | 2.2 (12) | 2.0 (11) | 2.0 (11) | 2.1 (11) | 3.9 (19) | 3.3 (17)\n\na Household-level characteristics are only available for index households as we did not collect information on other households in the compound.\nb Households in these communities who do not own a latrine typically share a latrine with extended family members who live in the same compound.\n\nChild-level characteristics by intervention group measured at follow-up.\n | Control | Water | Sanitation | Handwashing | WSH | Nutrition | N+WSH\nIndex children | N = 823 | N = 522 | N = 525 | N = 513 | N = 496 | N = 463 | N = 489\nMale, % (n) | 49.1 (404) | 49.8 (260) | 49.7 (261) | 48.0 (246) | 51.8 (257) | 51.2 (237) | 46.6 (228)\nAge in months, mean (range) | 30 (25\u201337) | 30 (25\u201338) | 30 (25\u201335) | 30 (24\u201337) | 30 (22\u201335) | 30 (26\u201338) | 30 (25\u201335)\nDeworming within last 6 months, % (n) | 64.0 (527) | 61.9 (323) | 58.5 (307) | 59.5 (305) | 57.5 (285) | 58.5 (271) | 59.5 (291)\nDeworming as part of MDA, % (n) a | 7.4 (39) | 9.3 (30) | 7.2 (22) | 9.8 (30) | 5.3 (15) | 6.6 (18) | 4.8 (14)\nDeworming source, % (n) a |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Pharmacy | 87.7 (462) | 83.0 (268) | 89.3 (274) | 79.3 (242) | 87.0 (248) | 87.8 (238) | 88.0 (256)\n\u00a0\u00a0\u00a0\u00a0Clinic | 10.8 (57) | 14.6 (47) | 8.8 (27) | 19.0 (58) | 12.6 (36) | 11.1 (30) | 10.7 (31)\n\u00a0\u00a0\u00a0\u00a0School | 1.5 (8) | 1.9 (6) | 1.0 (3) | 1.6 (5) | 0.4 (1) | 0.7 (2) | 1.0 (3)\nWeeks since deworming, mean (range) a | 9 (1\u201324) | 9 (1\u201324) | 9 (1\u201322) | 9 (1\u201321) | 9 (1\u201323) | 10 (0\u201324) | 9 (1\u201324)\nChild ingested soil within last week, % (n) | 36.7 (302) | 19.2 (100) | 29.5 (155) | 19.5 (100) | 25.6 (127) | 21.8 (101) | 19.0 (93)\nChild wearing shoes, % (n) | 16.5 (136) | 15.7 (82) | 14.3 (75) | 16.2 (83) | 14.5 (72) | 13.2 (61) | 16.8 (82)\nNon-index children | N = 707 | N = 449 | N = 447 | N = 464 | N = 445 | N = 400 | N = 444\nMale, % (n) | 45.3 (320) | 47.9 (215) | 53.5 (239) | 47.4 (220) | 50.3 (224) | 48.0 (192) | 50.0 (222)\nAge in years, mean (range) | 7 (4\u201312) | 7 (4\u201312) | 7 (4\u201312) | 7 (3\u201312) | 7 (3\u201312) | 7 (4\u201312) | 7 (4\u201312)\nSchool-age, % (n) | 89.0 (629) | 87.5 (393) | 89.7 (401) | 88.6 (411) | 87.9 (391) | 85.3 (341) | 88.1 (391)\nDeworming within last 6 months, % (n) | 70.2 (496) | 67.7 (304) | 74.1 (331) | 66.4 (308) | 63.8 (284) | 63.3 (253) | 62.8 (279)\nDeworming as part of MDA, % (n) a | 48.2 (239) | 54.0 (164) | 50.2 (166) | 56.2 (173) | 51.1 (145) | 48.2 (122) | 45.5 (127)\nDeworming source, % (n) a |  |  |  |  |  |  | \n\u00a0\u00a0\u00a0\u00a0Pharmacy | 47.6 (236) | 40.5 (123) | 42.9 (142) | 39.3 (121) | 41.6 (118) | 45.5 (115) | 47.7 (133)\n\u00a0\u00a0\u00a0\u00a0Clinic | 7.3 (36) | 9.2 (28) | 5.7 (19) | 9.7 (30) | 9.5 (27) | 11.1 (28) | 6.8 (19)\n\u00a0\u00a0\u00a0\u00a0School | 45.0 (223) | 50.0 (152) | 50.1 (167) | 51.0 (157) | 48.9 (139) | 42.7 (108) | 45.5 (127)\nWeeks since deworming, mean (range) a | 9 (0\u201324) | 9 (1\u201324) | 9 (1\u201324) | 9 (1\u201324) | 9 (1\u201322) | 9 (0\u201320) | 9 (1\u201324)\nChild ingested soil within last week, % (n) | 16.0 (113) | 2.7 (12) | 11.0 (49) | 2.4 (11) | 9.7 (43) | 5.3 (21) | 6.3 (28)\nChild wearing shoes, % (n) | 28.2 (199) | 26.3 (118) | 26.0 (116) | 23.7 (110) | 20.7 (92) | 26.0 (104) | 24.6 (109)\n\na Among children reported to be dewormed within last 6 months.\n\nSTH prevalence ratio for intervention vs. control arms for index children, other children in index household and children in other households in compound, unadjusted analysis.\n | All observations | Index children a | Other children in index household b | Children in non-index households c\n | N | Prev. | Prevalence ratio | N | Prev. | Prevalence ratio | N | Prev. | Prevalence ratio | N | Prev. | Prevalence ratio\nA. lumbricoides |  |  |  |  |  |  |  | \nControl | 1530 | 36.8% |  | 823 | 31.3% |  | 496 | 44.4% |  | 211 | 40.3% | \nWater | 971 | 35.8% | 0.97 (0.86, 1.11) | 522 | 28.0% | 0.89 (0.74, 1.07) | 325 | 43.4% | 0.98 (0.82, 1.18) | 124 | 49.2% | 1.22 (0.91, 1.64)\nSanitation | 972 | 35.9% | 0.98 (0.88, 1.09) | 525 | 30.3% | 0.97 (0.83, 1.13) | 327 | 42.8% | 0.98 (0.84, 1.13) | 120 | 41.7% | 0.96 (0.73, 1.27)\nHandwashing | 977 | 40.5% | 1.10 (0.98, 1.24) | 513 | 36.5% | 1.16 (0.97, 1.40) | 322 | 45.7% | 1.05 (0.89, 1.23) | 142 | 43.7% | 1.05 (0.80, 1.39)\nWSH | 941 | 34.3% | 0.93 (0.83, 1.05) | 496 | 28.8% | 0.92 (0.78, 1.09) | 311 | 40.5% | 0.93 (0.78, 1.12) | 134 | 40.3% | 1.00 (0.74, 1.34)\nNutrition | 863 | 40.3% | 1.10 (0.98, 1.23) | 463 | 35.0% | 1.12 (0.96, 1.29) | 257 | 49.0% | 1.14 (0.95, 1.37) | 143 | 42.0% | 1.14 (0.84, 1.54)\nNutrition + WSH | 933 | 33.0% | 0.90 (0.80, 1.01) | 489 | 26.4% | 0.84 (0.72, 0.99) | 290 | 41.0% | 0.95 (0.80, 1.13) | 154 | 39.0% | 1.02 (0.75, 1.40)\nHookworm |  |  |  |  |  | \nControl | 1530 | 9.2% |  | 823 | 3.6% |  | 496 | 16.1% |  | 211 | 14.7% | \nWater | 971 | 6.4% | 0.69 (0.50, 0.95) | 522 | 3.4% | 0.95 (0.52, 1.71) | 325 | 9.5% | 0.59 (0.39, 0.90) | 124 | 10.5% | 0.75 (0.37, 1.51)\nSanitation | 972 | 7.0% | 0.76 (0.54, 1.06) | 525 | 2.7% | 0.73 (0.35, 1.51) | 327 | 12.5% | 0.78 (0.52, 1.17) | 120 | 10.8% | 0.78 (0.38, 1.58)\nHandwashing | 977 | 8.3% | 0.90 (0.66, 1.22) | 513 | 3.5% | 0.96 (0.49, 1.90) | 322 | 13.4% | 0.84 (0.61, 1.16) | 142 | 14.1% | 0.95 (0.54, 1.69)\nWSH | 941 | 6.6% | 0.71 (0.52, 0.99) | 496 | 3.0% | 0.83 (0.49, 1.40) | 311 | 11.3% | 0.70 (0.46, 1.08) | 134 | 9.0% | 0.63 (0.36, 1.09)\nNutrition | 863 | 9.5% | 1.03 (0.74, 1.43) | 463 | 5.0% | 1.36 (0.72, 2.59) | 257 | 14.0% | 0.90 (0.62, 1.30) | 143 | 16.1% | 1.37 (0.78, 2.42)\nNutrition + WSH | 933 | 6.2% | 0.67 (0.50, 0.91) | 489 | 3.1% | 0.84 (0.51, 1.38) | 290 | 9.7% | 0.62 (0.41, 0.94) | 154 | 9.7% | 0.58 (0.28, 1.20)\nT. trichiura |  |  |  |  |  | \nControl | 1530 | 7.5% |  | 823 | 5.2% |  | 496 | 9.5% |  | 211 | 11.8% | \nWater | 971 | 7.1% | 0.95 (0.68, 1.32) | 522 | 4.6% | 0.88 (0.54, 1.44) | 325 | 9.5% | 1.02 (0.63, 1.66) | 124 | 11.3% | 0.86 (0.42, 1.78)\nSanitation | 972 | 5.3% | 0.71 (0.52, 0.98) | 525 | 3.2% | 0.62 (0.38, 1.00) | 327 | 7.3% | 0.81 (0.49, 1.35) | 120 | 9.2% | 0.72 (0.41, 1.27)\nHandwashing | 977 | 6.0% | 0.80 (0.59, 1.10) | 513 | 3.7% | 0.71 (0.41, 1.21) | 322 | 9.3% | 0.98 (0.61, 1.59) | 142 | 7.0% | 0.72 (0.34, 1.51)\nWSH | 941 | 6.4% | 0.85 (0.59, 1.22) | 496 | 5.0% | 0.96 (0.63, 1.47) | 311 | 8.0% | 0.87 (0.54, 1.41) | 134 | 7.5% | 0.69 (0.32, 1.49)\nNutrition | 863 | 7.2% | 0.96 (0.69, 1.33) | 463 | 4.5% | 0.87 (0.55, 1.37) | 257 | 8.6% | 0.92 (0.57, 1.50) | 143 | 13.3% | 1.31 (0.61, 2.81)\nNutrition + WSH | 933 | 9.1% | 1.21 (0.89, 1.65) | 489 | 6.3% | 1.21 (0.79, 1.86) | 290 | 12.4% | 1.33 (0.88, 2.02) | 154 | 11.7% | 1.04 (0.52, 2.08)\nAny STH |  |  |  |  |  |  |  |  |  |  |  | \nControl | 1530 | 43.4% |  | 823 | 35.2% |  | 496 | 54.2% |  | 211 | 49.8% | \nWater | 971 | 42.5% | 0.98 (0.87, 1.10) | 522 | 33.0% | 0.94 (0.79, 1.11) | 325 | 52.6% | 0.98 (0.85, 1.14) | 124 | 56.5% | 1.14 (0.89, 1.46)\nSanitation | 972 | 40.6% | 0.94 (0.84, 1.04) | 525 | 32.6% | 0.92 (0.79, 1.07) | 327 | 50.2% | 0.95 (0.82, 1.09) | 120 | 50.0% | 0.97 (0.75, 1.24)\nHandwashing | 977 | 46.3% | 1.07 (0.96, 1.19) | 513 | 38.8% | 1.10 (0.93, 1.31) | 322 | 54.7% | 1.02 (0.89, 1.17) | 142 | 54.2% | 1.07 (0.85, 1.35)\nWSH | 941 | 39.3% | 0.91 (0.81, 1.01) | 496 | 32.3% | 0.92 (0.79, 1.06) | 311 | 47.6% | 0.90 (0.76, 1.06) | 134 | 46.3% | 0.93 (0.70, 1.23)\nNutrition | 863 | 45.1% | 1.04 (0.93, 1.16) | 463 | 38.2% | 1.08 (0.95, 1.24) | 257 | 53.7% | 1.02 (0.86, 1.21) | 143 | 51.7% | 1.14 (0.88, 1.48)\nNutrition + WSH | 933 | 38.8% | 0.89 (0.81, 0.99) | 489 | 30.1% | 0.85 (0.74, 0.98) | 290 | 49.3% | 0.93 (0.80, 1.08) | 154 | 46.8% | 0.95 (0.74, 1.23)\n\na Children born to enrolled pregnant women following enrolment. The nutrition intervention was delivered to index children only.\nb Other children living in index child\u2019s household (index household). These include siblings of index children and children from other mothers in the same household. The water and handwashing interventions were delivered to index households only.\nc Children living in other households in index child\u2019s compound. The sanitation intervention was delivered to all households in the compound.",
        "label": "low",
        "id": "task4_RLD_test_808"
    },
    {
        "paper_doi": "10.1371/journal.pone.0012613",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 190 children under 5, 1144 people, 240 householdsInclusion criteria: unimproved water sources that tested over 1000 thermotolerant coliforms (TTC)/100 ml, reported low use of household water treatment, were easily accessible all year round and were motivated to take part in the project\n\n\nInterventions: LifeStraw(r) Family filter\n\n\nOutcomes: Incidence of diarrhoea among young children in the preceding seven days (recorded monthly); cough and fever also recordedFilter and water quality monitoringCompliance\n\n\nNotes: Location: rural eastern province of Kasai, Democratic Republic of CongoLength: 12 monthsPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "Background\nHousehold water treatment can improve the microbiological quality of drinking water and may prevent diarrheal diseases.\nHowever, current methods of treating water at home have certain shortcomings, and there is evidence of bias in the reported health impact of the intervention in open trial designs.\nMethods and Findings\nWe undertook a randomised, double-blinded, placebo-controlled trial among 240 households (1,144 persons) in rural Democratic Republic of Congo to assess the field performance, use and effectiveness of a novel filtration device in preventing diarrhea.\nHouseholds were followed up monthly for 12 months.\nFilters and placebos were monitored for longevity and for microbiological performance by comparing thermotolerant coliform (TTC) levels in influent and effluent water samples.\nMean longitudinal prevalence of diarrhea was estimated among participants of all ages.\nCompliance was assessed through self-reported use and presence of water in the top vessel of the device at the time of visit.\nOver the 12-month follow-up period, data were collected for 11,236 person-weeks of observation (81.8% total possible).\nAfter adjusting for clustering within the household, the longitudinal prevalence ratio of diarrhoea was 0.85 (95% confidence interval: 0.61\u20131.20).\nThe filters achieved a 2.98 log reduction in TTC levels while, for reasons that are unclear, the placebos achieved a 1.05 log reduction (p<0.0001).\nAfter 8 months, 68% of intervention households met the study's definition of current users, though most (73% of adults and 95% of children) also reported drinking untreated water the previous day.\nThe filter maintained a constant flow rate over time, though 12.4% of filters were damaged during the course of the study.\nConclusions\nWhile the filter was effective in improving water quality, our results provide little evidence that it was protective against diarrhea.\nThe moderate reduction observed nevertheless supports the need for larger studies that measure impact against a neutral placebo.\nTrial Registration\nCurrent Controlled Trials ISRCTN03844341\nIntroduction\nDiarrhoea is responsible for 1.8 million deaths annually, mostly among children under five in developing countries.\nMuch of this disease burden is attributable to unsafe water, poor hygiene and sanitation.\nAn estimated 884 million people worldwide lack access to improved water sources; hundreds of millions more rely on improved sources that are not consistently safe for drinking.\nEven water that is safe at the point of distribution often becomes contaminated during collection, transport and storage within the home due to poor hygiene conditions and practices.\nWhile safe, reliable, piped-in water is an essential goal, treating water at the household or other point of consumption provides a means by which vulnerable populations can improve the quality of their own drinking water.\nThe practice is widespread, with hundreds of millions reporting that they usually treat their water at home before drinking it.\nThere is also evidence that household water treatment is protective against diarrhoea though research suggests that placebo effect and reporting bias play a role in the estimate of effect reported in open trials.\nPlacebo-controlled trials of chlorine-based interventions have been conducted, but apart from a recent study in Ghana, none have assessed the neutrality of the placebo or the effectiveness of the blinding, and other issues have been raised about their methodological quality.\nFilters are more difficult to blind among populations relying on unimproved water.\nIf the water is turbid, a placebo that contains no filter medium is readily identified by comparing its effluent with the effective filter.\nHowever, a placebo that removes turbidity to ensure blinding will probably also remove pathogens that tend to adhere to the suspended solids; it may also create adsorption sites or promote biofilm adhesion that will also render the \u201cplacebo\u201d at least somewhat effective in removing pathogens.\nTo date, the only placebo-controlled trials of household-based filters have been conducted in the United States with municipally treated water that is low in turbidity but also met WHO water quality standards.\nThus, these results cannot be generalised to settings with turbid and contaminated water.\nSeveral water treatment methods have been promoted in low-income settings, including disinfection, disinfection/flocculation, ceramic filtration, solar disinfection and boiling.\nEach has limitations in terms of microbiological effectiveness, cost, acceptability, environmental impact, and sustainability among target populations.\nMoreover, except for boiling, none of these interventions have achieved scale except in limited settings.\nThis has led to calls for alternative technologies that are effective against the full array of microbial pathogens, that can be deployed and used at a large scale with minimum programmatic support, and that will be embraced by the target population.\nThe Lifestraw Family\u00ae is a newly developed household-based gravity filter that employs hollow-fibre membranes to remove waterborne pathogens by ultrafiltration.\nIndependent laboratory testing has shown the device to meet the US Environmental Protection Agency (USEPA) standards for bacteria, viruses and protozoan cysts.\nThe device is designed to treat a minimum of 18,000 L of water and assumed to last for about three years.\nThe manufacturer, Vestergaard-Frandsen SA of Lausanne, Switzerland, plans to sell the filter in large volumes for about US$20.\nMethods\nStudy design; sample size calculation\nThe study was designed as a randomised, double-blinded, placebo-controlled trial.\nOur primary outcome was longitudinal prevalence of diarrhea defined as the number of weeks with diarrhea divided by the total number of weeks under observation.\nThe study was powered to detect a 30% reduction in the mean longitudinal diarrhoea between the two groups.\nThis was a conservative estimate in comparison with the pooled risk reduction of 63% calculated from six previous studies of household filters.\nThe calculation assumed 80% power, \u03b1\u200a=\u200a0.05, a baseline longitudinal prevalence of diarrhoea of 5%, and a coefficient of variation of 2.\nIn order to account for potential lost to follow-up (10%) as well as clustering of diarrhoea within household and intermittent surveillance (7-day period prevalence measured repeatedly once a month over the 12-month follow-up period) (10%), we estimated that we needed at least 600 individuals in each arm.\nAssuming a mean of 5 persons per household, the number of households to be recruited was approximately 120 per arm, or 240 households in total.\nThe protocol for this trial (Protocol S1) and CONSORT checklist (Checklist S1) are available as supporting information.\nSetting and participant eligibility\nThe study was conducted from April 2008 to July 2009 in the rural health zone of Bibanga, 80 km from the city of Mbuji-Mayi in the eastern province of Kasai, the Democratic Republic of Congo (DRC).\nDespite abundant water resources, more than three quarters of the population in rural areas in the DRC rely on unimproved water sources for drinking, mainly surface water and unprotected springs.\nWith the assistance of the Presbyterian Church of Kinshasa, which has been supporting community health programmes in this area for many years, and staff at the health zone level, we identified possible study sites.\nSelected communities relied on unimproved water sources that tested over 1000 thermotolerant coliforms (TTC)/100 ml, reported low use of household water treatment, were easily accessible all year round from the reference hospital of Bibanga where the field team was established, and were motivated to take part in the project.\nIn order to meet the sample size requirements, the study was conducted in two neighbouring villages.\nIntervention\nEach intervention household received a Lifestraw Family filter and each control household received a placebo.\nThe Lifestraw Family is a gravity-fed microbiological water purifier.\nWater is poured into a 2.5 L plastic vessel, passes through a 27-\u00b5m pre-filter, and flows down a 1 m long plastic pipe before passing through the filtration cartridge comprised of hollow-fibres with a 20-nm pore size.\nThe top vessel contains a slow eluding chlorine tablet designed to prevent biofilm formation and increase the life of the cartridge.\nTreated water is accessed from the side of the cartridge via a tap.\nThe device is cleaned daily by rinsing the pre-filter and backwashing the cartridge using a squeeze-pump and outlet valve mounted on the bottom of the cartridge.\nThe device is designed to treat at least 18,000 L of water with a flow rate of approximately 150 ml per minute or 9 L per hour.\nIn the laboratory, the filter was found to meet the USEPA standards for microbiological water purifiers by reducing bacteria by 6.9 logs, viruses by 4.7 logs and protozoan cysts by 3.6 logs.\nPlacebo\nThe placebo had the same configuration, appearance and external components as the Lifestraw Family except that (i) the chlorine tablet was removed from the upper vessel to prevent possible microbicidal action, (ii) the filtration membranes were replaced by some extra piping to imitate the weight and effluent flow rate of the real cartridge, and (iii) the 27-\u00b5m screen on the pre-filter was removed to minimise retention of microbes adhering to suspended solids.\nThree weeks of testing in the laboratory confirmed that the placebo removed no bacteria, viruses and protozoan cysts from test water.\nDespite the challenge in blinding household filters, we determined after piloting that blinding the intervention would be feasible in our study area because the water had low turbidity, ranging from <5 nephelometric turbidity units (NTU) for most of the year to 10 NTU during heavy rains.\nEnrolment, baseline survey, randomization and filter deployment\nAfter discussing the proposed study with community leaders and obtaining consent from the heads of households, a baseline survey was undertaken in April 2008 to collect information on demographics, socio-economic characteristics, and water, hygiene and sanitation practices.\nData collection tools were translated in Tshiluba, the local language, and piloted before use.\nFollowing the baseline survey, households were randomly assigned to one of the two groups using a random number generator.\nRandomisation was stratified by village and was conducted by the trial manager who played no part in the collection of the data.\nBoth the intervention and the placebo were distributed door-to-door by five trained field workers who were unaware of whether the device was an active filter or a placebo.\nHouseholders were trained on use and maintenance of the device according to the manufacturer's instructions.\nThey were advised to drink filtered water directly from the tap and not to store filtered water in order to prevent recontamination.\nThe start of follow-up period was delayed by two months due to initial technical problems with the filters.\nBlinding\nThe allocation sequence was concealed from both field investigators and the study population.\nIn order to blind the intervention among assessors, field workers were divided into two teams.\nThe team responsible for assessing health outcomes was neither involved in the distribution of the filters at the commencement of the trial nor in the assessment of the filter performance and use during follow-up.\nAny questions from the householders that were related to the filter were referred to and dealt with by the filter assessment team.\nOutcome Assessment\nDiarrhoea\nInvestigators interviewed the female head of household or primary care giver of young children once each month over a 12-month period.\nThey recorded any diarrhoea cases in the preceding seven days.\nDiarrhoea was defined as three or more loose stools passed within a 24-hour period.\nIn an effort to further obscure the outcome of interest from the target population, field assessors also inquired about and recorded presence of fever and cough within the past seven days.\nChildren with diarrhoea were given oral rehydration sachets and instructions on how to use them.\nWhen necessary, they were referred to the closest community health post to receive medical care free of charge.\nFever and cough were also treated among young children.\nFilter monitoring\nEach month, a random sample of 30 filters and 30 placebos (25% of the total number distributed) was monitored.\nAt each household visit, field workers noted the location and condition of the filter and recorded if the respondent was able to use and clean the filter correctly.\nFilter components found to be damaged were replaced.\nFlow rate was monitored by filling the top container with 2.5 L of water, opening the tap and measuring the time necessary to fill a 125 ml container with water.\nThe flow rate was expressed in ml per minute.\nWater quality\nInfluent and effluent paired water samples were collected for each of the selected devices.\nIf the respondent mentioned storing the water once filtered, a third sample was collected from the container designated as the treated water storage vessel.\nAll samples were collected in sterile 125 ml Nalgene sampling bottles and assessed for thermotolerant coliforms (TTC) within 4 h after collection.\nMicrobiological assessment was performed using the membrane filtration technique (APHA Standard Methods) on membrane lauryl sulphate medium (Oxoid Limited, Basingstoke, Hampshire, UK) using a DelAgua field incubator (Robens Institute, University of Surrey, Guilford, Surrey, UK).\nMicrobiological performance of the filters was expressed in terms of log reduction value (LRV) calculated as the log of the influent concentration divided by the effluent concentration (log10 influent/effluent).\nCompliance\nCross-sectional surveys were conducted among each household eight and fourteen months after distribution.\nParticipants were classified as current users if they reported using the filter \u2018today or yesterday\u2019 and if the field investigator found the filter hung for use with water in the top vessel of the device.\nConsistency of use was estimated by asking the respondent if he/she had drunk unfiltered water within the previous day.\nThe survey covered further aspects on use and acceptability.\nBlinding assessment\nImmediately following the conclusion of the follow-up period, we assessed the effectiveness of blinding among participants.\nBlinding indices were calculated using methods developed by James and colleagues and Bang and colleagues Female heads of household or primary care giver were asked to identify which device they had received.\nSurveys were targeted at the respondents to the health surveys because they would be most likely to be influenced by their belief in treatment assignment.\nData analysis\nThe analysis of the primary outcome was on an intention-to-treat basis.\nWe used Poisson regression with robust standard errors to estimate the effect of the intervention on the longitudinal prevalence of diarrhea and other health outcomes.\nWe used generalized estimating equations (GEE) to account for clustering at the household level.\nCategorical data were compared using a Chi square or a Fisher's exact test where appropriate.\nContinuous variables were compared with a Student t test.\nStatistical analyses of microbiological data were conducted after log10 transformation of TTC counts to normalize the distribution.\nData analysis was conducted in Stata (Stata Corporation, College Station, Texas, US).\nEthics\nThe study was reviewed and approved by the ethics committee at the London School of Hygiene and Tropical Medicine and the ethics committee at the School of Public Health in Kinshasa.\nWritten consent to participate in the research was obtained from community leaders and the head of each participating household.\nInvestigators explained that half of the study population would be receiving effective microbiological purifiers while the others would receive placebos and that householders should continue their existing water management practices since their device may not be protective against microbial contamination.\nAt the conclusion of the follow-up period, all placebo filters were replaced by effective filters.\nFollowing the completion of the study, the results were communicated to all study participants.\nResults\nParticipant flow\n259 households initially volunteered to participate in the study.\nNineteen households were excluded because they did not reside in the selected villages; they relied primarily on spring water for drinking, or subsequently elected not to participate.\nA total of 240 households were enrolled; 120 were assigned to receive the Lifestraw Family filter and 120 the placebo.\nOver the 12-month follow-up period, data were collected for 11,236 (81.8%) possible person-weeks of observation.\nData were missing for 2492 weeks (18.2%) due primarily to participants leaving the study area or being absent at the time of visit (Figure 1).\nOver the study period, twenty participants died, six of them were children under the age of five.\nThe number of deaths was 12 in the intervention group and 8 in the control group (p\u200a=\u200a0.27).\nBaseline surveys\nIntervention and control groups were similar in terms of demographic and socio-economic characteristics and hygiene and sanitation practices (Table 1).\nAlmost all households primarily used river water for drinking.\nHowever, intervention households were more likely to store their water in clay pots and access it by dipping a cup into the container compared with control households who often used jerrycans.\nOnly four households reported treating their water sometimes or rarely by boiling or adding bleach.\nOnly 37% of households had a latrine and 51% had soap present in the house at the time of visit (Table 1).\nDiarrhoea surveillance\nAt baseline, the prevalence of diarrhoea was similar in both groups (12.6% versus 10.6% for control and intervention groups, respectively).\nOver the 12-month follow-up period, participants of all ages who received the active filter experienced 15% fewer weeks with diarrhoea compared to those who received a placebo (mean, 2.66 versus 3.15, respectively).\nHowever, the confidence interval of the longitudinal prevalence ratio (LPR) adjusted for clustering within the household (LPR 0.85, 95% CI 0.61; 1.20) was wide and included 1.\nThe longitudinal prevalence ratio among children under five was 0.85 (95%CI 0.56; 1.28).\nFigure 2 shows the prevalence of diarrhoea between intervention and control over time.\nWe observed no difference in the mean longitudinal prevalence of fever (LPR 0.99; 95% CI 0.80; 1.22) or cough (LPR 0.99; 95% CI 0.81; 1.22) between the two groups.\nHealth outcome data are presented in Table 2.\nWater quality\nEach device was tested on average 3 times during follow-up. 580 (81%) of the total possible paired water samples were collected.\nMissing samples are due to householders being absent or not being in possession of their filter at the time of visit.\nSource drinking water was highly contaminated, with 75% of household samples showing contamination levels above 1000 TTC/100 ml (Figure 3).\nThe active filter achieved a LRV of 2.98 (95% CI 2.88, 3.08), removing about 99.8% of the indicator bacteria.\nOverall, 64% of water samples treated with filter were free of TTC and 27% had TTC levels between 1\u201310 TTC/100 ml.\nNone of the filters produced water with >100 TTC/100 ml consistently over the three visits.\nSamples from placebos were also contaminated, with 73% of the water samples containing between 100\u20131000 TTC/100 ml.\nHowever, unlike the results from laboratory testing that showed the placebo to be microbiologically ineffective, results from the field showed that the placebo actually removed more than 90% of the TTC from source water (LRV 1.05, 95% CI: 0.93, 1.16).\nFlow rate\nThe mean flow rate of the filters over the study period was 202 ml/min (95% CI 198, 206) or 12 L/hour.\nIt declined slightly over time (\u22121.5 ml/min per month, p<0.002).\nOperation and maintenance; acceptability\nOver half of the respondents (56%) correctly demonstrated how to clean the filters.\nThe pre-filter was cleaned at each use (40%) or once a day (41%), wheareas the cartrige was generally backwashed once a day (67%).\nOverall, 36 (12.4%) of the 290 active filters tested were found damaged during visits, mainly due to rodents chewing on the soft hoses (n\u200a=\u200a35).\nIntervention households reported liking the filter due to improved aesthetics (88%), taste (92%), odour (56%) and health (35%).\nReasons for dissatisfaction were slow flow rate (87%), small size of the top container (85%) and problems with rats (44%).\nCompliance\nEight months after distribution, 183 (76%) of the households were present at the time of visit and were still in possession of their filter.\n68% of respondents in the intervention group could be defined as current users against 48% in the placebo group (p<0.001).\nHowever, nearly all adults (83%) and young children (95%) also reported drinking untreated water in the previous day.\nFourteen months after distribution, the proportion of current users was slightly higher in both groups (76% versus 69% among intervention and control groups, respectively).\nAdditional details about use are included in Table 3.\nSubgroup analysis showed no evidence of an association between use and diarrhoea morbidity (Table 4).\nBlinding status\nTable 5 shows respondent guesses for each treatment assignment groups.\nJames' method, similar to the kappa statistics, produced a blinding index (BI) score of 0.42 (95%CI 0.38; 0.46).\nA score of 0 means that all respondents guessed correctly, 1 indicates that all respondents guessed incorrectly and 0.5 indicates random guessing.\nBang's method calculates the proportion of correct guesses beyond chance in each treatment group.\nBang's BI was 0.96 (95%CI 0.90; 0.99) for the intervention group and \u22120.63 (95%CI \u22120.73; \u22120.53) for the placebo-controlled group.\nBang's blinding index varies from \u22121 to 1.\n1 indicates complete lack of blinding, \u22121 opposite guess about treatment assignment and 0 random guessing.\nSubgroup analysis showed no evidence of an association between diarrhoea and respondents' guesses.\nDiscussion\nWe undertook the first double-blinded, placebo-controlled trial of household-based water filters in a low-income setting with water known to be contaminated with faecal pathogens.\nThis design sought to assess the impact of the intervention in the absence of respondents' bias that is common in open trials.\nDue to challenges of developing a placebo in such settings and to successfully blinding the intervention, we monitored placebo performance and conducted a post-intervention assessment of blinding among study participants.\nFilter performance and health impact were monitored for a full year to account for seasonal variations and minimise the potential for exaggerated health impact often associated with shorter-term trials.\nAfter adjusting for clustering, members of intervention households had 15% fewer weeks of diarrhoea than those of control households, but the confidence intervals indicated little statistical support (longitudinal prevalence ratio 0.85, 95%CI: 0.61 to 1.20).\nWith the exception of a recent study in the United States among an elderly population17, this finding is consistent with other placebo-controlled trials of household water treatment interventions which found no protective effect against diarrhoea.\nHowever, as we have observed elsewhere, those studies may have had insufficient power to identify a statistically significant impact on diarrhoea.\nOur sample size also was not sufficiently large to detect a statistically significant difference in diarrhoea of 15%.\nMoreover, the baseline prevalence of diarrhoea was lower than anticipated and the clustering effect due to repeated measurement and household randomisation was higher.\nPos-hoc sample size calculations indicated that we would have needed a study approximately ten times larger to achieve statistical significance.\nMoreover, the placebo was not microbiologically neutral, as it removed about 90% of faecal bacteria from the source water used by control households.\nThe reasons for this apparent effectiveness are not clear.\nField staff responsible for water quality testing were extensively trained and supervised throughout the study, thereby minimizing the risk of measurement errors.\nOne of the most plausible explanations is the formation of a biofilm resulting from adhesion of suspended solid particles and bacteria on the inner surface of the plastic pipe forming the placebo cartridge.\nThe effectiveness of the placebo rendered our trial a comparison between a 1-log filter and a 3-log filter.\nStudies have reported an association between 1 log removal of faecal bacteria from drinking water and a reduction in diarrhoeal disease.\nOur results may therefore understate the effectiveness of the active filter if it were compared to a true placebo.\nThese results suggest that in this setting with relatively high levels of microbial contamination in source water, a filter of superior microbiological performance may be more effective at preventing diarrhoea than one that removes only 90% of waterborne pathogens.\nThis finding, if validated in future studies, would support the need for high performance standards in water treatment devices in order to optimize health benefits.\nThe blinding of the intervention was not successful.\nIn both treatment groups, the vast majority of survey respondents believed that they had received the active filter, although this proportion was significantly lower in the placebo group.\nUnsuccessful blinding means that we cannot rule out the possibility that the observed effect on diarrhoea is unbiased.\nHowever, the interpretation of blinding indices is not always clear.\nThe fact that a large proportion of control households remained blinded throughout the trial suggests that respondents' bias may have at least been partly reduced.\nThe smaller effect size we observed here may be indicative of a less biased estimate compared with open trials.\nOur estimate is similar to the pooled estimate of effect of open trials of ceramic filters after adjustment for lack of blinding.\nThe fact that \u2018control\u2019 health conditions (fever and cough) remained unchanged by the intervention also suggests that blinding may have been effective, although the usefulness of this approach to detect the presence of respondents' bias has not been validated.\nIncluding a third arm with no intervention would have provided a better understanding of the role of bias in this study.\nUnder field conditions, the Lifestraw Family filters were effective in removing faecal bacteria from source water.\nTwo-thirds of filtered water samples were free of faecal coliforms while most of the remaining samples had low levels of contamination.\nThe fact that specific filters did not consistently produce contaminated water suggests that contamination may have occurred during collection of the sample, perhaps from the tap.\nThe flow rate was higher than that observed in laboratory conditions, possibly due to lower water turbidity at the study site (compared to lab testing at 15 NTU) or inconsistent use by householders.\nThe damage rate was high although the most common problems were due to rats eating the soft plastic components.\nEight months after distribution, two-thirds of the respondents met the study's definition of current users, although almost none of them drank filtered water exclusively.\nThis pattern of use was seen among both adults and children under five.\nParticipants drank unfiltered water when spending time outside their home, but also when they felt eager to drink and did not want to wait for filtration.\nYoung children did not have access to the filter when their parents were away from home.\nIn accordance with the manufacturer's instructions, householders were advised to use water directly from the filter and not to store treated water due to the risk of recontamination.\nConsistent with these instructions, almost none of the households stored filtered water for their children, though many lacked a storage container even if they had chosen to do so.\nThe manufacturer has advised that in future deployment of the filters, it will consider changes in instructions to encourage safe storage of treated water or provide a storage vessel for the filtered water to help increase exclusive consumption of treated water, especially by this vulnerable group of young children.\nHowever, there is also evidence that even occasional consumption of untreated water may eliminate the protective effect of water treatment and changes to the configuration of the filter may not be sufficient to increase exclusive use unless accompanied by fundamental changes in behaviour to increase compliance.\nOur study had certain additional limitations.\nThe study sites were not randomly selected, but were chosen based on eligibility criteria that included high levels of faecal contamination in source water and high prevalence of diarrhoea at baseline.\nAccordingly, these results are not necessarily generalizable to other populations in the Congo or beyond.\nSecond, the use of a seven-day recall period is known to produce less precise estimates compared with a 48-hour recall period.\nOur study provides little evidence of a protective effect of the filter against diarrhoea.\nNevertheless, an effect of 15%, which we observed but could not confirm here, would represent a substantial impact on diarrhoea, a major killer of young children.\nFuture studies with sufficient power to detect this effect size will be necessary to determine the magnitude of any effect against a neutral placebo and to confirm that the effect is not attributable to chance.\nOur study also demonstrates the need to monitor placebo performance and the challenge of blinding household-based water treatment interventions under adverse conditions.\nCONSORT diagram showing the flow of participants through the trial.\nPrevalence of diarrhoea over the course of the study among participants of all ages.\nPercentage of water samples by level of contamination (TTC/100 ml).\n\nBaseline characteristics of participating households.\n | Control | Intervention | Total\n | N | % | N | % | N | %\nDemographic and socio-economic |  |  |  |  |  | \nNumber of households | 120 | (50) | 120 | (50) | 240 | (100)\nNumber of persons | 598 | (52.3) | 546 | (47.7) | 1144 | (100)\nNumber of households with children <5 | 66 | (55) | 57 | (47.5) | 123 | (51.2)\nNumber of children <5 | 105 | (17.6) | 85 | (15.8) | 190 | (16.6)\nMean number of persons per household | 5.0 | 4.5 | 4.8\nMean number of rooms in the house | 2.2 | 2.3 | 2.3\nRespondent is female | 76 | (63.3) | 76 | (63.3) | 152 | (63.3)\nMean age of respondent | 37.5 | 40.8 | 39.1\nLevel of education |  |  |  |  |  | \nNo formal education | 47 | (39.2) | 38 | (31.7) | 85 | (35.4)\nPrimary | 44 | (60.3) | 45 | (54.9) | 89 | (57.4)\nSecondary | 29 | (39.7) | 36 | (43.9) | 65 | (41.9)\nHigher | 0 | (0) | 1 | (1.2) | 1 | (0.6)\nOwns |  |  |  |  |  | \nHouse | 113 | (94.2) | 116 | (96.7) | 229 | (95.4)\nLand | 115 | (95.8) | 117 | (97.5) | 232 | (96.7)\nLivestock | 59 | (49.2) | 64 | (53.8) | 123 | (51.5)\nRadio | 27 | (22.7) | 34 | (28.3) | 61 | (25.5)\nPhone | 10 | (8.3) | 16 | (13.3) | 26 | (10.8)\nBicycle | 18 | (15) | 16 | (13.3) | 34 | (14.2)\nHygiene and sanitation |  |  |  |  |  | \nUse soap to wash hands | 54 | (45) | 54 | (45) | 108 | (45)\nPresence of soap at the time of visit | 65 | (54.2) | 59 | (49.2) | 124 | (51.7)\nReceived hygiene advice in past 6 months | 4 | (3.4) | 10 | (8.4) | 14 | (5.9)\nPresence of latrine | 47 | (39.2) | 41 | (34.2) | 88 | (36.7)\nWater handling practices |  |  |  |  |  | \nPrimary source of drinking water |  |  |  |  |  | \nRiver | 120 | (100) | 117 | (97.5) | 237 | (98.7)\nRainwater | 44 | (36.7) | 46 | (38.3) | 90 | (37.5)\nSpring | 15 | (12.5) | 19 | (15.8) | 34 | (14.2)\nType of drinking water container |  |  |  |  |  | \nClay pot | 68 | (56.7) | 83 | (69.2) | 151 | (62.9)\nJerry can | 50 | (41.7) | 30 | (25) | 80 | (33.3)\nOther | 2 | (1.7) | 7 | (5.8) | 9 | (3.7)\nVessel opening |  |  |  |  |  | \nWide mouth | 71 | (59.2) | 92 | (76.7) | 163 | (67.9)\nNarrow mouth | 49 | (40.8) | 28 | (23.3) | 77 | (32.1)\nStorage vessels covered | 111 | (93.3) | 113 | (95.0) | 224 | (94.1)\nMeans of obtaining water |  |  |  |  |  | \nPour | 48 | (41.0) | 27 | (23.3) | 75 | (32.2)\nDip | 69 | (59.0) | 89 | (76.7) | 158 | (67.8)\nTreat water* | 3 | (2.5) | 1 | (0.8) | 4 | (1.7)\n\n*Treat water sometimes (n\u200a=\u200a1) or rarely (n\u200a=\u200a3). Treatment methods boil (n\u200a=\u200a2), bleach (n\u200a=\u200a1), water settle (n\u200a=\u200a1).\n\nLongitudinal prevalence of diarrhoea and other health conditions by age and treatment group.\n | Mean longitudinal prevalence | LPR (95% CI) | LPR* (95%CI)\n | Control | Intervention |  | \n | Weeks of illness | Person-weeks of observation | % Weeks ill | Weeks of illness | Person-weeks of observation | % Weeks ill |  | \nDiarrhoea |  |  |  |  |  |  |  | \n<5 | 96 | 1072 | 8.96 | 60 | 801 | 7.49 | 0.84 (0.61; 1.14) | 0.85 (0.56; 1.28)\n5\u201315 | 31 | 1880 | 1.65 | 29 | 1765 | 1.64 | 1.00 (0.60; 1.65) | 0.91 (0.49; 1.67)\n>15 | 59 | 2945 | 2.00 | 52 | 2752 | 1.89 | 0.94 (0.65; 1.36) | 0.95 (0.61; 1.57)\nAll ages** | 186 | 5907 | 3.15 | 142 | 5329 | 2.66 | 0.85 (0.68; 1.05) | 0.85 (0.61; 1.20)\nFever |  |  |  |  |  |  |  | \n<5 | 249 | 1072 | 23.23 | 187 | 801 | 23.35 | 1.00 (0.85; 1.19) | 1.02 (0.79; 1.30)\n5\u201315 | 99 | 1880 | 5.27 | 123 | 1765 | 6.97 | 1.32 (1.02; 1.71) | 1.28 (0.89; 1.85)\n>15 | 226 | 2945 | 7.67 | 188 | 2752 | 6.83 | 0.89 (0.74; 1.07) | 0.91 (0.68; 1.22)\nAll ages** | 576 | 5907 | 9.75 | 500 | 5329 | 9.38 | 0.96 (0.86; 1.08) | 0.99 (0.80; 1.22)\nCough |  |  |  |  |  |  |  | \n<5 | 196 | 1072 | 18.28 | 162 | 801 | 20.22 | 1.11 (0.92; 1.33) | 1.11 (0.85; 1.43)\n5\u201315 | 163 | 1880 | 8.67 | 142 | 1765 | 8.05 | 0.93 (0.75; 1.50) | 0.89 (0.63; 1.27)\n>15 | 192 | 2945 | 6.52 | 201 | 2752 | 7.30 | 1.12 (0.93; 1.35) | 1.07 (0.82; 1.39)\nAll ages** | 551 | 5907 | 9.33 | 505 | 5329 | 9.48 | 1.01 (0.90; 1.14) | 0.99 (0.81; 1.22)\n\n*Adjusted for clustering within household.\n**Age missing for 3 participants.\n\nDescription of use among study participants.\n | Control | Intervention | Total\n | n | % | n | % | n | %\nMONTH 8 |  |  |  |  |  | \nLast use (n\u200a=\u200a183)* |  |  |  |  |  | \nPrevious day | 44 | (48.3) | 63 | (68.5) | 107 | (58.5)\nPrevious week | 30 | (33.0) | 14 | (15.2) | 44 | (24.0)\n>1 week ago | 17 | (18.7) | 15 | (16.3) | 32 | (17.5)\nConsistency of use on previous day (n\u200a=\u200a107) |  |  |  |  |  | \nRespondent drank unfiltered water | 43 | (97.7) | 46 | (73.0) | 89 | (83.2)\nChildren (<5) drank unfiltered water | 31 | (93.9) | 39 | (95.1) | 70 | (94.6)\nFilter accessible to young children | 1 | (2.3) | 6 | (9.5) | 7 | (6.5)\nStore filtered water for young children | 4 | (12.9) | 8 | (19.5) | 12 | (16.7)\nAdditional details on use in previous day (n\u200a=\u200a107) |  |  |  |  |  | \nRespondent drank unfiltered water when |  |  |  |  |  | \nIn the field | 33 | (76.7) | 39 | (78.3) | 72 | (77.9)\nIn a hurry to drink | 30 | (69.8) | 33 | (71.7) | 63 | (70.8)\nAway from village | 16 | (37.2) | 15 | (32.6) | 31 | (34.8)\nOther | 3 | (7.0) | 12 | (26) | 15 | (16.8)\nChildren drank unfiltered water when |  |  |  |  |  | \nPerson operating the filter not present | 21 | (67.7) | 31 | (79.5) | 52 | (74.3)\nIn a hurry to drink | 11 | (35.5) | 23 | (59.0) | 34 | (48.6)\nAway from home | 10 | (32.3) | 13 | (33.3) | 23 | (32.9)\nOther | 5 | (16.1) | 7 | (17.9) | 12 | (17.1)\nDid not store filtered water for children: |  |  |  |  |  | \nNo container | 17 | (68) | 28 | (87.1) | 45 | (78.9)\nLock the door | 6 | (24) | 3 | (9.3) | 9 | (15.8)\nDon't want to always filter, too slow | 2 | (8) | 0 | (0) | 2 | (3.5)\nTold not to store water | 0 | (0) | 1 | (3.1) | 1 | (1.7)\nMONTH 14 |  |  |  |  |  | \nLast use (n\u200a=\u200a190)** |  |  |  |  |  | \nPrevious day | 63 | (69.2) | 75 | (75.8) | 138 | (72.6)\nPrevious week | 14 | (15.4) | 11 | (11.1) | 25 | (13.2)\n>1 week ago | 14 | (15.4) | 13 | (13.3) | 27 | (14.2)\n\n*197 (82%) households present at the time of visit; 183 (93%) of them were still in possession of the filter and ever used it.\n**203 (85%) households present at the time of visit; 192 (94%) of them were still in possession of the filter and ever used it + answer missing for 2 households.\n\nLongitudinal prevalence of diarrhoea stratified by reported last time of use.\n | Mean longitudinal prevalence of diarrhoea | LPR (95% CI)\n | Control | Intervention | \n | Weeks of illness | Person-weeks of observation | % weeks ill | Weeks of illness | person-weeks of observation | % weeks ill | \n8 months |  |  |  |  |  |  | \nUser | 71 | 2475 | 2.87 | 74 | 3155 | 2.35 | 0.82 (0.59; 1.13)\nNon-user | 68 | 2420 | 2.81 | 41 | 1319 | 3.11 | 1.11 (0.75; 1.62)\n14 months |  |  |  |  |  |  | \nUser | 102 | 3463 | 2.95 | 99 | 3894 | 2.54 | 0.86 (0.66; 1.10)\nNon-user | 49 | 1642 | 2.98 | 27 | 1025 | 2.63 | 0.88 (0.55; 1.40)\n\n\nBlinding status of respondents by group assignment at the end of the study.\n | Group assignment\nGuess | Placebo* | Lifestraw Family* | Total*\nPlacebo | 17 | (18.3) | 2 | (2.0) | 19 | (9.9)\nLifestraw Family | 74 | (79.6) | 97 | (98.0) | 171 | (89.1)\nDon't know | 2 | (2.1) | 0 | (0) | 2 | (1.0)\nTotal** | 93 | (100.0) | 99 | (100.0) | 192 | (100.0)\n\n*N (%) - number of respondents and percentage in each group.\n**192 (80%) households present at the time of interview and still in possession of the filter.",
        "label": "low",
        "id": "task4_RLD_test_636"
    },
    {
        "paper_doi": "10.1371/journal.pone.0168702",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Individually RCT\n\n\nParticipants: 101 adults aged 18 to 65 years, 97% male presenting or referred with uncomplicated P. falciparum or mixed P. falciparum/P. vivax infection diagnosed with microscopy and confirmed with qPCR, had mild or moderate G6PD deficiency.Setting: Cambodia\n\n\nInterventions: DHAP: daily for 3 days; tablets containing 40mg DHA and 320 mg piperaquine eachPQ: single dose 45mg at day 3\n\n\nOutcomes: The proportion of individuals infecting at least 1 mosquito out of 50, at 1 and 2 weeks post-treatment in the 2 arms.The effect of the 2 treatment regimens on the risk of gametocyte carriage as measured by microscopy and RT-PCR. This was done by comparing gametocyte prevalence at weekly timepoints post-treatment and the time to gametocyte clearance in the 2 arms.The number of infected mosquitos per treatment arm, the relationship of gametocytaemia to mosquito infectivity, and within-person changes in haemoglobin 4 days post-PQ treatment among volunteers with G6PD-deficiency\n\n\nNotes: Baseline infectivity of participants (day 0): DHAP 6/51 (12%); DHAP+PQ 1/51 (2%)\n\n",
        "objective": "To assess the effects of single dose or short\u2010course PQ (or an alternative 8AQ) alongside treatment for people with P. falciparum malaria.",
        "full_paper": "Background\nSingle low dose primaquine (SLD PQ, 0.25mg/kg) is recommended in combination with artemisinin-based combination therapy (ACT) as a gametocytocide to prevent Plasmodium falciparum transmission in areas threatened by artemisinin resistance.\nTo date, no randomized controlled trials have measured primaquine\u2019s effect on infectiousness to Anopheline mosquitoes in Southeast Asia.\nMethods\nCambodian adults with uncomplicated falciparum malaria were randomized to receive a single 45mg dose of primaquine (equivalent to three SLD PQ) or no primaquine after the third dose of dihydroartemisin-piperaquine (DHP) therapy.\nA membrane-feeding assay measured infectiousness to Anopheles dirus on days 0, 3, 7, and 14 of blood-stage therapy.\nGametocytemia was evaluated by microscopy and reverse-transcriptase PCR.\nResults\nPrior to trial halt for poor DHP treatment efficacy, 101 participants were randomized and 50 received primaquine.\nOverall microscopic gametocyte prevalence was low (9%), but gametocytemic subjects given primaquine were gametocyte-free by day 14, and significantly less likely to harbor gametocytes by day 7 compared to those treated with DHP-alone, who remained gametocytemic for a median of two weeks.\nOnly one infectious subject was randomized to the primaquine group, precluding assessment of transmission-blocking efficacy.\nHowever, he showed a two-fold reduction in oocyst density of infected mosquitoes less than 24 hours after primaquine dosing.\nIn the DHP-alone group, four subjects remained infectious through day 14, infecting roughly the same number of mosquitoes pre and post-treatment.\nOverall, microscopic gametocytemia was an excellent predictor of infectiousness, and performed better than submicroscopic gametocytemia post-treatment, with none of 474 mosquitoes infected post-treatment arising from submicroscopic gametocytes.\nConclusions\nIn a setting of established ACT resistance, a single dose of 45mg primaquine added to DHP rapidly and significantly reduced gametocytemia, while DHP-alone failed to reduce gametocytemia and prevent malaria transmission to mosquitoes.\nContinued efforts to make single dose primaquine widely available are needed to help achieve malaria elimination.\nIntroduction\nMalaria containment efforts launched in western Cambodia over the last 7\u20138 years to curb artemisinin-resistant malaria have contributed to a marked decline in Plasmodium falciparum cases in the region, but have failed to prevent the emergence of slow-clearing parasites throughout Southeast Asia.\nCambodia adopted dihydroartemisinin-piperaquine (DHP) nationally as its first-line ACT in 2012, though significant clinical failures were reported the following year.\nAs few therapeutic options remain in the Mekong Subregion, it is even more important to pursue transmission-blocking interventions to help achieve regional malaria elimination.\nIn 2010, the WHO recommended that single dose primaquine (45mg or 0.75 mg/kg, later revised to 0.25mg/kg in 2012) be used as a gametocytocide in combination with ACT to prevent transmission of P. falciparum to mosquitoes in areas threatened by artemisinin resistance.\nAt the time, there was good clinical evidence that primaquine added to ACTs substantially reduced gametocytes, the parasite stages responsible for transmission to mosquitoes.\nHowever, actual entomological evidence of transmission-blocking to mosquitoes was limited to small non-randomized case series.\nEarlier this year, a randomized controlled trial in Mali was the first to show that single low dose primaquine, given alongside DHP, was indeed efficacious for preventing malaria transmission to Anopheles gambiae mosquitoes in West Africa.\nIn 2012\u20132014, we similarly undertook an open label randomized trial to determine the transmission-blocking efficacy of primaquine added to DHP in western Cambodia.\nWe used membrane feeding assays to measure infectiousness to Anopheles dirus mosquitoes, the major malaria vector in the region.\nUnfortunately, DHP failure in the cohort was unexpectedly high indicating rapid progression of clinical resistance, and the trial had to be halted early, as previously reported.\nHere we report on the mosquito infectivity and gametocyte outcomes of the 101 participants randomized before the trial was halted.\nThe overall rate of infectiousness in the cohort was low and precluded an ability to establish definitive transmission-blocking efficacy, but we saw a significant reduction in gametocytes post-treatment in those given the 45mg dose of primaquine.\nIn contrast, subjects treated with DHP-alone without primaquine were slow to clear gametocytes, and individuals continued to infect mosquitoes two weeks post-treatment.\nMethods\nStudy design and participants\nThis study was carried out between 10 December 2012\u201324 February 2014 in Oddar Meanchey Province, northwestern Cambodia.\nMalaria transmission is low, heterogeneous, and seasonal with entomological inoculation rates generally below one/person/year.\nThe majority of clinical cases occur during the rainy season between May and December.\nThis was an open-label randomized clinical trial that enrolled adults aged 18\u201365 years presenting or referred to Anlong Veng District Hospital with uncomplicated P. falciparum or mixed P. falciparum/P. vivax (Pf/Pv) infection diagnosed with microscopy and confirmed with real-time polymerase chain reaction (PCR).\nTesting for glucose-6-phosphate dehydrogenase (G6PD) deficiency was carried out using both the fluorescent spot test (R&D Diagnostics Ltd, Greece) and quantitative testing of enzyme activity (Trinity Biotech, Ireland).\nParticipants with severe deficiency (WHO Class I or II) defined as 10% or less of the lower limit of normal activity were excluded, while those with mild or moderate G6PD deficiency were included in the trial.\nMore details of the study design have been previously presented.\nParticipants provided written informed consent after referral from various medical facilities in Oddar Meanchey Province.\nEthical approval for the study was obtained from the Walter Reed Army Institute of Research (14 September 2012), National Ethics Committee for Health Research in Cambodia (21 October 2011), and University of North Carolina (25 September 2012).\nThe trial was registered with ClinicalTrials.gov, number NCT01849640.\nProcedures\nAll participants received three doses of dihydroartemisinin-piperaquine (40mg and 320mg per tablet) as directly observed therapy over 3 days (0, 24, and 48h from baseline).\nAfter the third dose, subjects were randomly allocated 1:1 to receive 45mg of primaquine or no primaquine.\nThis dose typically occurred 52 hours from baseline (on day 2).\nParticipants were released to outpatient follow-up on day 3 (72 hours) or once afebrile with two consecutive negative smears, returning for weekly follow-up visits until day 42.\nGiemsa-stained thick and thin blood smears collected at baseline, every 8 hours during treatment, and at weekly follow-up visits were examined by two microscopists blinded to each other\u2019s results, with gametocytes counted per 2000 white blood cells.\nMolecular detection of gametocytes at baseline, day 7, and day 14 was carried out using reverse-transcriptase PCR (RT-PCR) of Pfs25.\nThis assay has a reported detection limit of 1\u20132 gametocytes/\u03bcL when applied to blood spots.\nHemoglobin was measured at baseline and daily until discharge on day 3.\nIt was also measured in G6PD-deficient subjects at their follow-up visit on days 7 and 14, but not routinely checked for G6PD-normal subjects after day 3.\nMosquito infectivity was measured in all study participants, regardless of gametocyte status, via membrane feeding experiments at baseline (day 0); and days 3 (72 hours from baseline), 7, and 14 post-treatment.\nFor each experiment, 2mL of fresh venous blood was fed to ~300 colony-reared female Anopheles dirus mosquitoes, with the goal of yielding approximately 200 engorged (fed) mosquitoes.\nFurther details on the membrane feeding assay have been previously published.\nNine days after feeding, mosquito midguts from 50 mosquitoes were examined for the presence of oocysts following dissection in 2% mercurochrome, with oocyst counts confirmed by two independent readers.\nAnother 50 mosquitoes at day 9 were saved for molecular detection to confirm the presence of oocysts and perform speciation of parasites.\nAt day 16 post-feeding, the remaining mosquitoes were collected for molecular detection of sporozoites.\nIn particular, 18s rRNA PCR performed on DNA extracted from individual mosquitoes was used to help confirm the presence of P. falciparum infection in mosquitoes fed on subjects with mixed Pf/Pv infection.\nOutcomes\nWe assessed the transmission-blocking efficacy of a 3-day dose of DHP with or without a single oral 45 mg dose of primaquine by comparing the proportion of individuals infecting at least one mosquito out of fifty, at one and two weeks post-treatment in the two arms.\nWe also assessed the effect of the two treatment regimens on the risk of gametocyte carriage as measured by microscopy and RT-PCR.\nThis was done by comparing gametocyte prevalence at weekly timepoints post-treatment and the time to gametocyte clearance in the two arms.\nWe also analyzed the number of infected mosquitos per treatment arm, the relationship of gametocytemia to mosquito infectivity, and within-person changes in hemoglobin four days post-primaquine treatment among volunteers with G6PD-deficiency.\nStatistical analysis\nThe study was not powered for transmission-blocking efficacy, but for the main primary objective of measuring the therapeutic efficacy of DHP with and without primaquine.\nA sample size of 150 evaluable subjects was chosen to yield a 95% CI of 89\u201397% around a true efficacy estimate of 94%.\nWe entered data into a Microsoft Access 2007 database, with 100% clinical source data verification by the study monitor.\nStatistical analyses were performed using Stata version 12.1.\nIn this analysis, we included all subjects that were randomized to primaquine on day 2.\nDifferences in the proportion of infectious individuals and gametocyte prevalence were compared using \u03c72 test or Fisher\u2019s exact test as appropriate.\nUnadjusted Kaplan-Meier survival analysis was used to measure the risk of persistence of gametocytes in those gametocytemic on day 2, with log-rank testing used for comparison and hazard ratio generated using a Cox proportional hazards model.\nOther comparisons used Fisher\u2019s exact test for categorical variables and Student\u2019s t test or Wilcoxon rank-sum test (for non-normally distributed data) for continuous variables.\nWe regarded a one-sided p value of less than 0.05 significant when evaluating the effect of primaquine on infectiousness and gametocyte carriage; otherwise a two-sided p-value was used.\nResults\nStudy population\nOut of 107 subjects who were enrolled and given dihydroartemisinin-piperaquine (DHP) therapy, six withdrew from the study before being randomized, leaving 101 who were randomized to primaquine (PQ) or no primaquine on Day 2; 50 received DHP+PQ and 51 DHP-only.\nThere were no differences in baseline demographics between primaquine and non-primaquine groups (Table 1).\nMost participants were male (98/101) and either farmers (83/101) or military personnel (11/101).\nRoughly sixty percent (60/101) were febrile at presentation, with a median reported duration of fever of 2 days (IQR 2).\nWhile none reported antimalarial use in the previous 28 days, 35/101 (35%) had detectable levels of piperaquine in the blood, which was not associated with baseline gametocyte carriage (p = 0.27).\nThe therapeutic efficacy of DHP was poor in both groups, leading to voluntary study halt prior to reaching target enrollment.\nInfectiousness to mosquitoes\nA total of 387 membrane feeding assays were successfully conducted in the 101 subjects pre and post-treatment (n = 101 day 0, n = 100 day 3, n = 96 day 7, n = 90 day 14).\nFeeding assays were performed a median of 1.0 hours (interquartile range 28 min to 2.2 hours) after venous blood draw.\nIn every experiment, 50 mosquito midguts were dissected 9 days after initial membrane feeding and examined for oocysts.\nPositive mosquito infection (at least 1/50 with midgut oocysts) was confirmed in all cases by real-time PCR of mosquitoes also saved at day 9 post-feeding.\nOnly 7/101 (6.9%) participants were infectious to mosquitoes pre-treatment\u20146 in the DHP-only arm and 1 in the DHP+PQ arm.\nOf the 6 infectious subjects in the DHP-only arm, 4 remained infectious through day 14 post-treatment, and an additional subject who was non-infectious at presentation became infectious on day 3, ~20hrs after primaquine dosing.\nIn contrast, in the DHP+PQ arm, the 1 infectious subject at presentation was rendered non-infectious by day 7, and no others became infectious (Fig 1, Table 2).\nThese small numbers precluded the ability to show a reduction in individual infectiousness due to primaquine.\nAlso, mosquito-feeding assays were not performed on day 2, just prior to PQ dosing, to more directly measure the transmission-blocking effect of primaquine separate from DHP.\nHowever, subjects in the non-primaquine arm continued to infect significant numbers of mosquitoes post-treatment (165/2400 or 6.9% at day 7 and 106/2100 or 5.0% at day 14), whereas no mosquito infections occurred in the primaquine arm one and two weeks post-treatment (Table 2).\nOf note, the one gametocytemic and infectious subject in the primaquine group (SN-060) remained infectious at the day 3 membrane feeding, which occurred just ~20hrs after primaquine dosing, but displayed a two-fold reduction in oocyst density in the infected mosquitoes (Fig 1, Table 3).\nAt this 72-hour timepoint, the gametocyte density of 772 gametocytes/\u03bcL was comparable to that at baseline (728 gametocytes/\u03bcL), having decreased from a peak of 967 gametocytes/\u03bcL the first day.\nThe majority of mosquitoes still became infected on day 3 (35/50), but the median oocyst density was 42 (IQR 26 to 78) at day 0 vs. 21 (IQR 5 to 36) at day 3.\nBy day 7, or 5 days post-primaquine, gametocytes were greatly reduced but still present at 71 gametocytes/\u03bcL, but these were not infective to mosquitoes.\nGametocyte prevalence and clearance\nAt baseline, 9/101 (9%) participants were gametocyte positive by microscopy, while 46/101 (46%) were gametocyte positive by Pfs25 RT-PCR.\nJust prior to primaquine dosing on day 2, the proportion of participants with microscopic gametocytemia increased to 12/101 (12%).\nBy one week post follow-up, 5 days after primaquine dosing, those given primaquine had a lower rate of gametocyte carriage, with 2.1% of subjects in the DHP + PQ group displaying patent gametocytes vs. 15% of those in the DHP-only arm (prevalence ratio 0.12, 95% CI 0.02\u20130.9, p = 0.03) (Table 4, Fig 2).\nIn the primaquine group, only 1/4 subjects with patent gametocytemia at day 2 remained gametocytemic at day 7, and all were gametocyte-free at day 14.\nIn contrast, the number of subjects with patent gametocytes declined gradually in the non-primaquine group, and gametocyte prevalence did not fall below 10% until 3 weeks post-treatment (Table 4, Fig 2).\nAccordingly, gametocyte clearance was faster in the primaquine group.\nAmong the 12 participants who were gametocytemic by microscopy at day 2, the median time to clearance was 1 day in the primaquine group vs. 12 days in the non-primaquine group (hazard ratio 7.3 (95% CI 1.3\u201342), logrank p = 0.01) (Fig 3).\nThe detection of submicroscopic gametocytes further corroborated these findings: gametocytes detected by RT-PCR dropped approximately 10-fold from baseline by day 7 due to the combined effects of DHP and primaquine (49% to 4.3%), but only dropped two-fold from baseline in the non-primaquine group (44% to 26%) (Table 4, S1 Fig).\nRelationship of gametocytemia to infectiousness\nWe previously showed a close relationship between microscopic gametocytemia and infectiousness to mosquitoes at baseline pre-treatment in this cohort.\nUsing all the data from serial membrane feedings conducted both pre and post-treatment (N = 387 or roughly 4 feeding assays per subject), microscopic gametocytemia was a good predictor of infectiousness: 19/36, or roughly half, of patent gametocyte episodes led to mosquito infection, while successful transmission was only observed in 2/351 instances when gametocytes were not visible by smear (p<0.001).\nThese two instances of transmission arising from submicroscopic gametocytes caused low-level mosquito infection and occurred pre-treatment (Table 3).\nPost-treatment, none of the 474 infected mosquitoes found out of 14,350 dissected arose from submicroscopic gametocytes.\nDespite this difference pre and post-treatment in the infectiousness of submicroscopic gametocytes, we observed no difference in the apparent threshold of microscopic gametocytes (~100 gametocytes/mL) that rendered persons infectious to mosquitoes pre and post-treatment in the absence of primaquine (Fig 4).\nG6PD status and hemolysis\nG6PD screening by both qualitative and quantitative tests led to the exclusion of 5 persons with severe G6PD deficiency (WHO Class I or II, less than 10% of normal G6PD activity) prior to enrollment.\nOf 101 participants enrolled in the study, 8 had mild or moderate G6PD deficiency, with activity ranging from 0.6 to 3.2 U/Hgb (median 1.2 U/Hgb).\nAmong these 8 subjects, a greater drop in hemoglobin at day 7 was seen among those treated with primaquine but was not statistically significant: median fractional reduction in hemoglobin was 24.8% (range 0.9% to 29.8%) in the six subjects given primaquine vs. 7.1% (-4.4% and 18.7%) in the two subjects not given primaquine (p = 0.14).\nThe largest drop was to 9.9 g/dL from a baseline Hgb of 14.1; all Hgb values were >11.0 g/dL by day 14 (S2 Fig).\nIn the G6PD normal subjects, hemoglobin was only measured through day 3, at which time there was no change in the fractional drop in hemoglobin between treatment arms (9.8% vs. 9.3% in the PQ and non-PQ groups, respectively, p = 0.79).\nDiscussion\nOur study is one of now three randomized control trials using a mosquito infection endpoint to provide evidence for the WHO\u2019s 2010 recommendation to add single dose primaquine to ACT as a transmission-blocking intervention in areas nearing elimination or threatened by artemisinin resistance.\nIt is the only trial conducted in Asia and used Anopheles dirus, one of the predominant outdoor biting malaria vectors residing in the forest and forest fringes throughout Southeast Asia.\nIt was also conducted in an area of ACT-resistant malaria, where containment has been a global priority.\nUnfortunately, despite even distribution of microscopic gametocytes between arms and a high rate of submicroscopic gametocytemia overall (~50%), the low rate of infectiousness in the cohort and its uneven distribution between arms led to only one infectious subject in the primaquine arm at baseline vs. six in the non-primaquine arm.\nThus, though there was no transmission observed by day 7 among those treated with primaquine, reduction in human to mosquito transmission could not be adequately assessed in the primaquine-treated arm.\nDespite this limitation, there was a statistically significant reduction in gametocyte carriage due to a 45mg dose of primaquine that paralleled the lack of infectious subjects in the primaquine arm post-treatment.\nThis study used the single 45mg dose (0.75 mg/kg) of primaquine recommended by WHO at the time it was conducted.\nThe WHO revised their recommendation to use a lower dose of 0.25mg/kg or 15mg in 2012.\nThe switch to a lower dose was motivated by pooled analysis of prior studies suggesting that the 0.25mg/kg dose would retain transmission-blocking efficacy while avoiding clinically significant hemolysis observed in G6PD-deficient persons at the higher dose, albeit rarely.\nLow dose primaquine offers a significant strategic advantage in most malaria endemic settings where G6PD screening prior to treatment remains impracticable.\nGiven the higher 0.75mg/g dose, we performed universal G6PD screening and excluded 5 volunteers with severe G6PD deficiency (<10% of normal activity).\nWhile the higher dose ultimately proved to be safe in 6 subjects with mild to moderate (WHO Class III) G6PD deficiency, prior to recovery at day 14, 3 of the 6 subjects experienced a >25% fractional drop in hemoglobin at day 7 compared to just 2/124, or 1.6% of G6PD-deficient volunteers given the lower 15mg dose in a recent study on the Thai-Myanmar border.\nPart of this difference may be attributable to the hemolytic effects of symptomatic malaria in our study, absent in the Thai study of healthy asymptomatic volunteers as part of a mass drug administration effort.\nStill, the results of the Thai-Myanmar study are encouraging.\nRoughly 1.2% (15/1226) of G6PD-normal volunteers in that study also experienced an asymptomatic >25% reduction in hemoglobin.\nAnother study from Tanzania found mean hemoglobin reductions of roughly 8% among both G6PD deficient and normal volunteers with smear-positive malaria treated with artemether-lumefantrine (AL) and low dose primaquine.\nTogether, this emerging evidence suggests the 025.mg/kg primaquine dose does not increase the risk for clinically significant hemolysis and can be safely given without prior G6PD screening.\nOngoing pharmacovigilance studies in Burkina Faso, Mali, and the Gambia are expected to augment this evidence base.\nFrom an efficacy standpoint, dose-finding transmission-blocking studies recently completed in Mali and Burkina Faso now support adequate transmission-blocking efficacy of the lower 0.25mg/kg dose recommended by the WHO in combination with DHP and AL.\nIt will be useful to replicate these studies in regions outside Africa with different vectors and newer ACT regimens.\nAs we have previously noted, microscopic gametocytemia was an excellent predictor of infectiousness in our cohort.\nOf 21 episodes of human to mosquito transmission in the cohort (7 pre-treatment and 14 post-treatment), all but two arose from subjects with microscopic gametocytemia.\nThe two that arose from submicroscopic gametocytes occurred pre-treatment and resulted in low-level mosquito infection (Table 3).\nOur molecular detection of gametocytes relied on a nested PCR assay that was qualitative, limiting a more quantitative analysis of the gametocyte infectivity relationship at low levels of gametocytes.\nWe also did not attempt to determine the limit of detection of our molecular procedures for oocyst and sporozoite detection.\nHowever, they showed good concordance with the results of midgut dissection, suggesting similar levels of detection by these two methods (S1 Table).\nThis is similar to a prior study which used phenol-chloroform DNA extraction from single mosquitoes experimentally infected with low intensity infection and applied a nested PCR also targeting 18srRNA.\nPerhaps the most unsettling finding here was how much post-treatment gametocytemia still remained in the ACT-alone \u201ccontrol\u201d arm in this setting of prevalent artemisinin and piperaquine resistance.\nIn areas where ACTs have not been compromised, artemisinin-based therapies rapidly clear asexual parasite stages and are thought be effective against immature gametocytes as well, resulting in low rates of post-treatment gametocytemia compared to non-artemisinin drugs.\nIn a recent meta-analysis of nearly 49,000 patients treated with ACTs, only 1.9% of patients without gametocytemia at enrollment developed patent gametocytemia within 28 days following treatment.\nIn contrast, microscopic gametocytes appeared in 4 patients (approximately 8%) in our ACT-only arm within the first week of therapy.\nTwo of these four patients later developed recrudescent parasitemia.\nWith regards to gametocyte clearance, the same meta-analysis found that over half with patent gametocytes at enrollment were gametocyte-free by day 7, while the median time to clearance in our DHP-only arm was 14 days.\nThese comparisons suggest that as artemisinin-resistant slow-clearing parasites proliferate, gametocyte carriage will increase, as has been noted in some, though not all studies.\nIn our cohort, gametocyte prevalence actually increased post-treatment in the non-primaquine arm, a finding that was not present in a large study in Myanmar in 2008\u20139 with >800 patients treated with various ACTs, nor in Indonesian patients treated with DHP in 2008\u201310.\nWe further demonstrate that gametocytes persisting post-treatment remained infectious to mosquitoes, contributing to the reservoir of multidrug resistant malaria (Table 3, S1 Fig).\nThe relative contribution of these previously treated patients to the infectious reservoir (including those receiving suboptimal therapy either in the private sector or through self-treatment) versus the larger pool of untreated asymptomatic persons with submicroscopic malaria remains unknown, yet critically important to shaping malaria elimination strategies.\nNot all ACTs are equal in their ability to reduce gametocytemia.\nWhile DHP offers advantages of once daily dosing and a long elimination half-life, its Achilles heel may be more post-treatment gametocyte carriage compared to other ACTs.\nTwo large studies both found that gametocyte carriage rates following DHP treatment were 2\u20133 fold higher than those following artesunate-mefloquine (ASMQ), the other ACT frequently used in Southeast Asia, a finding confirmed by meta-analysis.\nAn African study showed that infectivity to mosquitoes was greater in patients treated with DHP vs. artemether-lumefantrine.\nThis may be due to a lower dose of artemisinin in the combination pill compared to other ACTs or a partner drug effect.\nNo matter the reason, the recent policy recommendation to revert to ASMQ in western Cambodia due to escalating DHP failure may prove advantageous from a transmission standpoint.\nIncreasing the number of days that an artemisinin is given, perhaps in a directly observed therapy setting, would also likely have a greater inhibitory effect on developing gametocytes and thus transmission-blocking effect.\nThe weakness of DHP as an anti-gametocyte ACT is overcome when given with single dose primaquine (8,14,24).\nOur study shows that this is true even in the setting of high-grade DHP failure, at least with the higher 0.75mg/kg dosing.\nIn our study, a single dose of 45mg primaquine effectively killed pre-existing gametocytes and prevented the development of new gametocytes.\nSubmicroscopic gametocytemia also fell dramatically within 4 days of primaquine administration.\nThe submicroscopic gametocytemia that remains is much less infectious, due to sterilization of gametocytes before they are cleared, as well as a possible bias towards measurement of the more abundant female gametocytes, whereas transmission requires the presence of both male and female gametocytes.\nOur infectivity data bears this out, as all episodes of mosquito infection after administration of primaquine arose from microscopic gametocytemia.\nOverall, our findings, though limited by small numbers and a higher dose of primaquine than currently recommended, lend support to the WHO recommendation to give single dose primaquine to reduce P. falciparum transmission.\nThey highlight the importance of adding primaquine as a transmission-blocking intervention in areas where ACTs are failing, as we found that Cambodian patients treated with dihydroartemisinin-piperaquine cleared gametocytes slowly and still contributed substantially to the infectious reservoir after treatment.\nFortunately, most infectious subjects were readily identified by the presence of patent gametocytes.\nDespite adoption of primaquine into the official drug policy of many countries, a large gap remains between policy and real-world availability.\nOvercoming these practice barriers needs to be a priority if primaquine is to be deployed effectively for regional malaria elimination in the Mekong and to mitigate the global spread of multidrug resistant malaria.\nSchematic of gametocyte and mosquito infectivity status through treatment in 101 randomized participants.Participants in the primaquine (PQ) and non-primaquine arms are depicted in the same ordered configuration from Day 0 pre-treatment through Week 2 post-treatment. Subjects with patent gametocytes detected by microscopy are colored black, while subjects who infected at least one mosquito on membrane feeding are colored blue. Persons that were both gametocytemic and infectious are colored half black-half blue. Persons who missed follow-up are shown as missing.\nGametocyte prevalence during 42-day follow-up.Gametocyte prevalence for each regimen, as measured by microscopy. Dihydroartemisinin-piperaquine (DHP) was dosed on days 0\u20132. Primaquine (PQ) was dosed on day 2. Error bars indicate the upper and lower limits of the 95% CI. *Indicates a statistically significant difference between groups based on a one-tailed Fisher\u2019s exact test.\nKaplan-Meier survival curves of gametocyte clearance by treatment regimen.Twelve subjects that were gametocytemic at day 2, just prior to primaquine (PQ) dosing as measured by microscopy, are included. 95% confidence bands are shown.\nRelationship of microscopic gametocytemia to prevalence of mosquito infection.The results of 35 membrane feeding assays performed on gametocytemic blood from 14 subjects pre and post-treatment. Black circles denote assays performed pre-treatment (Day 0), while colored squares denote assays performed post-treatment (Days 3, 7, 14) on subjects in the DHP-only group (indigo) and subjects in the DHP+PQ group (green). Additionally, the only two mosquito infections observed to arise from submicroscopic gametocytemia (pre-treatment) are depicted on the y-axis. Raw data for infected mosquitoes is available in Table 3. Note pre-treatment data (black circles) were previously presented in Ref 18.\n\nBaseline characteristics of study subjects.\n | DHA-PPQ+ Primaquine (n = 50) | DHA-PPQNo Primaquine (n = 51) | p-value\nMale, no. (%) | 48 (96) | 50 (98) | 0.49\nWeight, kg, mean (SD) | 57 (8) | 58 (6) | 0.55\nAge, y, median (IQR) | 25 (10) | 25 (13) | 0.49\nOccupation, No. (%) |  |  | 0.86\n\u2003Farmer | 40 (80) | 43 (84) | \n\u2003Forest worker | 1 (2) | 0 | \n\u2003Military | 5 (10) | 6 (12) | \n\u2003Other | 4 (8) | 2 (4) | \nHistory of malaria in previous year, no. (%) | 22 (44) | 21 (41) | 0.47\nAntimalarial use in past 28 days | 0 | 0 | \u2014\nFever (\u226538.0\u00b0C) at presentation, no. (%) | 29 (58) | 31 (61) | 0.78\nDuration of fever, days, median (IQR) | 2.5 (1.5) | 2 (2) | 0.46\nHemoglobin, mg/dL, mean (SD) | 13.4 (1.8) | 13.6 (1.5) | 0.46\nParasite density, per \u03bcL, mean (95% CI) | 14,962 (10,185\u201321,978) | 17,218 (6,389\u201343,451) | 0.60\nGametocyte prevalence by microscopy, no. (%) | 4 (8) | 5 (10) | 0.75\nGametocyte prevalence by RT-PCR*, no. (%) | 24 (49) | 22 (44) | 0.62\nMixed Pf/Pv infection, no. (%) | 4 (8) | 4 (8) | 0.63\nG6PD activity**, no. (%) |  |  | 0.11\n\u2003Normal | 43 (86) | 49 (96) | \n\u2003Partial deficiency | 1 (2) | 0 | \n\u2003Deficiency | 6 (12) | 2 (4) | \n\n*denominators are 49 and 50 in the primaquine and no primaquine arms, respectively\n**as determined by weak-moderate fluorescence on fluorescent spot testing, indicating intermediate G6PD enzyme activity.\n\nInfectiousness to mosquitoes during follow-up.\n | DHP+ Primaquine | DHPNo Primaquine\nno. of infectious participants | no. of mosquitoes infected/dissected | median no. of oocysts/mosquito (IQR) | no. of infectious participants | no. of mosquitoes infected/dissected | median no. of oocysts/mosquito (IQR)\nDay 0 | 1/50 (2.0%) | 35/2500 (1.4%) | 42 (26 to 78) | 6/51 (12%) | 135/2550 (5.3%) | 94* (3 to 166)\nDay 2 | \u2014 | \u2014 | \u2014 | \u2014 | \u2014 | \u2014\nDay 3 | 1/50 (2.0%) | 35/2500 (1.4%) | 22 (5 to 36) | 5/51 (9.8%) | 168/2550 (6.6%) | 26 (6 to 68)\nDay 7 | 0/48 (0) | 0/2400 (0) | - | 4/48 (8.3%) | 165/2400 (6.9%) | 39 (15 to 93)\nDay 14 | 0/48 (0) | 0/2400 (0) | - | 4/42 (9.5%) | 106/2100 (5.0%) | 27 (6 to 57)\n\n*data restricted to 3 of the 6 participants as 1 subject (SN-119) had a mixed Pf/Pv infection and in 2 subjects (SN-063 and SN-086), oocysts were not visualized though multiple PCR-positive mosquito pools suggested these individuals were infectious.\n\nGametocytemia and mosquito infection among infectious patients.\nTreatment Arm | Subject ID | Day of followup | Gametocytes/\u03bcL | Pfs25 RT-PCR result | Mosquitoes infected | Oocysts/mosquito, median (range)\nDHP + PQ | SN-060 | 0 | 728 | + | 35/50 (70%) | 42 (1\u2013231)\n2 | 670 |  |  | \n3 | 772 |  | 35/50 (70%) | 21(1\u201356)\n7 | 71 | + | 0/50 | \n14 | 0 | - | 0/50 | \nDHP only | SN-002 | 0 | 705 | + | 13/50 (26%) | 1 (1\u201312)\n2 | 569 |  |  | \n3 | 427 |  | 22/50 (44%) | 59 (1\u2013147)\n7 | 247 | + | 37/50 (74%) | 33 (1\u2013120)\n14 | 161 | + | 6/50 (12%) | 5 (1\u201315)\nSN-010 | 0 | 16 | + | 0/50 | \n2 | 87 |  |  | \n3 | 71 |  | 22/50 (44%) | 3 (1\u20138)\n7 | 70 | + | 0/50 | \n14 | 0 | - | 0/50 | \nSN-063 (M*) | 0 | 690 | + | 26/30 (87%) | \n2 | 652 |  |  | \n3 | 860 |  | 26/30 (87%) | \n7 | 844 | + | 27/30 (90%) | \n14 | 250 | + | 19/30 (63%) | \nSN-070 | 0 | 0 | + | 4/50 (8%) | 1 (1\u20132)\n2 | 0 |  |  | \n3 | 0 |  | 0/50 | \n7 | 0 | - | 0/50 | \n14 | 0 | - | 0/50 | \nSN-086 | 0 | 0 | + | ?** | \n2 | 47 |  |  | \n3 | 77 |  | 32/50 (64%) | 6 (1\u201334)\n7 | 55 | + | 33/50 (66%) | 9 (1\u201327)\n14 | 54 | + | 40/50 (80%) | 12 (1\u201342)\nSN-108 | 0 | 342 | + | 49/50 (98%) | 149 (48\u2013398)\n2 | 412 |  |  | \n3 | 375 |  | 49/50 (98%) | 78 (7\u2013191)\n7 | 293 | + | 50/50 (100%) | 111(32\u2013147)\n14 | 80 | + | 49/50 (98%) | 58 (14\u2013132)\nSN-119 (M*) | 0 | 118 | + | 3/30 (10%) | \n2 | 145 |  |  | \n3 | 96 |  | 0/30 | \n7 | 42 | + | 0/30 | \n14 | 28 | - | 0/30 | \n\n*Note that for subjects with mixed Pf/Pv infection(M), the percentage of infected mosquitoes was determined by species-specific PCR of individual mosquitoes, and oocyst counts are not reported because of inability to distinguish falciparum vs. vivax oocysts.\n** Two of 5 and 5 of 5 pools of mosquitoes were real-time PCR positive at days 9 and 16 after feeding, respectively, but oocysts were not seen.\n\nGametocyte carriage during follow-up.\n | DHP+ Primaquine | DHPNo Primaquine | p-value*\nby microscopy | by RT-PCR | by microscopy | by RT-PCR | microscopy | RT-PCR\nDay 0 | 4/50 (8.0%) | 24/49 (49%) | 5/51 (9.8%) | 22/50 (44%) |  | \nDay 2 | 4/50 (8.0%) |  | 8/51 (16%) |  |  | \nDay 3 | 4/50 (8.0%) |  | 9/51 (18%) |  |  | \nDay 7 | 1/48 (2.1%) | 2/47 (4.3%) | 7/48 (15%) | 11/42 (26%) | 0.03 | 0.003\nDay 14 | 0/42 (0) | 0/42 (0) | 6/48 (13%) | 6/47 (13%) | 0.02 | 0.02\n\n*p-value compares gametocyte prevalence in the two arms post-treatment using one-tailed Fisher's exact test",
        "label": "low",
        "id": "task4_RLD_test_931"
    },
    {
        "paper_doi": "10.1371/journal.pntd.0002983",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: RCTLength of follow-up: 6 months\n\n\nParticipants: Infected children identified by screeningNumber analysed for primary outcome: 194Age range: 9 to 12 yearsInclusion criteria: children aged 9 to 12 years from 5 primary schools, with at least one type of STH infection.Exclusion criteria: deworming treatment within 6 months before the current trial.\n\n\nInterventions: Single dose vs placeboAlbendazole: 3 x 400 mg for 3 consecutive daysMatching placebo\n\n\nOutcomes: Physical fitness (10 m shuttle run and VO2 max)Physical strength (grip strength and standing broad jump test)HeightWeightTriceps and subscapular skinfold thicknessHaemoglobinNot included in review: parasitological examination.\n\n\nNotes: Location: Bulanghsam township bordering Myanmar, a sub-division of Menghai county in Xishuangbanna Dai autonomous prefecture, situated in Yunnan province, P.R. ChinaBurden: highSource of funding: Swiss Tropical and Public Health Institute in Basel, Switzerland and the National Institute of Parasitic Diseases, Chinese Center of Diseases Control and Prevention in Shanghai, P.R. China\n\n",
        "objective": "To summarize the effects of public health programmes to regularly treat all children with deworming drugs on child growth, haemoglobin, cognition, school attendance, school performance, physical fitness, and mortality.",
        "full_paper": "Background\nThere is considerable debate on the health impacts of soil-transmitted helminth infections.\nWe assessed effects of deworming on physical fitness and strength of children in an area in Yunnan, People's Republic of China, where soil-transmitted helminthiasis is highly endemic.\nMethodology\nThe double-blind, randomized, placebo-controlled trial was conducted between October 2011 and May 2012.\nChildren, aged 9\u201312 years, were treated with either triple-dose albendazole or placebo, and monitored for 6 months post-treatment.\nThe Kato-Katz and Baermann techniques were used for the diagnosis of soil-transmitted helminth infections.\nPhysical fitness was assessed with a 20-m shuttle run test, where the maximum aerobic capacity within 1 min of exhaustive exercise (VO2 max estimate) and the number of 20-m laps completed were recorded.\nPhysical strength was determined with grip strength and standing broad jump tests.\nBody height and weight, the sum of skinfolds, and hemoglobin levels were recorded as secondary outcomes.\nPrincipal Findings\nChildren receiving triple-dose albendazole scored slightly higher in the primary and secondary outcomes than placebo recipients, but the difference lacked statistical significance.\nTrichuris trichiura-infected children had 1.6 ml kg\u22121 min\u22121 (P\u200a=\u200a0.02) less increase in their VO2 max estimate and completed 4.6 (P\u200a=\u200a0.04) fewer 20-m laps than at baseline compared to non-infected peers.\nSimilar trends were detected in the VO2 max estimate and grip strength of children infected with hookworm and Ascaris lumbricoides, respectively.\nIn addition, the increase in the VO2 max estimate from baseline was consistently higher in children with low-intensity T. trichiura and hookworm infections than in their peers with high-intensity infections of all soil-transmitted helminths (range: 1.9\u20132.1 ml kg\u22121 min\u22121; all P<0.05).\nConclusions/Significance\nWe found no strong evidence for significant improvements in physical fitness and anthropometric indicators due to deworming over a 6-month follow-up period.\nHowever, the negative effect of T. trichiura infections on physical fitness warrants further investigation.\nAuthor Summary\nChildren from the developing world are often burdened with intestinal worms due to poor water supply, sanitation, and hygiene.\nHowever, the assessment of the burden due to intestinal worms is difficult, and thus, the benefits of deworming are unclear.\nIn this study, we determined the effect of deworming on the physical fitness and strength of 9- to 12-year-old children in Yunnan, China, where intestinal worms are common.\nChildren were treated with triple-dose albendazole or placebo and monitored over a 6-month period.\nStool samples were collected for the diagnosis of intestinal worm infections.\nPhysical fitness was estimated with a 20-m shuttle run test and physical strength was assessed with grip strength and standing broad jump tests.\nChildren receiving triple-dose albendazole scored slightly higher values in the primary and secondary outcomes than those children who were given placebo.\nHowever, the differences were not significant.\nWe also found that children infected with intestinal worms performed significantly worse in the physical fitness and strength tests than their non-infected counterparts.\nIn particular, the negative impact of whipworm infection on physical fitness warrants further investigation.\nIntroduction\nSoil-transmitted helminths, namely Ascaris lumbricoides, Trichuris trichiura, and the hookworms (Ancylostoma duodenale and Necator americanus), are the most common parasitic worm infections of humans.\nIndeed, more than 1 billion people are infected and approximately 5.4 billion people are at risk of infection.\nIn 2011, an estimated 875 million children, 70% of whom were school-aged, were at risk globally.\nImpoverished communities with poor hygiene and no access to clean water and improved sanitation are especially vulnerable.\nThe global burden of soil-transmitted helminthiasis is currently estimated at 5.2 million disability-adjusted life years (DALYs), mainly due to sub-clinical morbidities, but also anemia and reduced cognitive and physical development.\nInfections are largely chronic and usually asymptomatic, and hence the study and quantification of the morbidity associated with soil-transmitted helminth infections are difficult, and only few studies have ventured to do so.\nIn particular, no conclusive evidence has yet been established whether reduced physical fitness or strength are a consequence of soil-transmitted helminth infections.\nPhysical fitness has been positively correlated with academic performance through enhanced memory and attention, while physical strength is demanded in labor-intensive agriculture jobs, which often provide the main source of income in rural communities of the developing world.\nA lack in both attributes due to soil-transmitted helminthiasis could arguably prevent school-aged children living in impoverished conditions from realizing their full potential and perpetuate their entrapment in the vicious cycle of poverty and poor health.\nBased on the rationale that lowering infection intensity would help to control morbidity associated with chronic helminth infection, and that morbidity is infection intensity-dependent, the World Health Organization (WHO) advocates periodic deworming of at-risk populations (e.g., school-aged children and pregnant women) with single-dose albendazole (400 mg) or mebendazole (500 mg).\nSuch an approach indeed reduces infection intensity in the target population, but high-quality evidence on the health benefits of de-worming in children is scant.\nTwo randomized controlled trials have shown that physical fitness in school boys infected with soil-transmitted helminths improved 7 weeks to 4 months after treatment with single-dose albendazole.\nPhysical fitness was also negatively correlated with T. trichiura and hookworm infections in two cross-sectional studies but another cross-sectional study did not find any correlation between physical fitness and soil-transmitted helminth infections.\nHowever, it is important to note that in the latter study, both the prevalence and intensity of soil-transmitted helminth infections were very low.\nWe designed a randomized controlled trial to investigate the health benefits of deworming and thereby deepen our understanding of the burden caused by soil-transmitted helminth infection among school-aged children.\nThe study was conducted in a highly endemic area in the People's Republic of China (P.R. China) and assessed the effects of triple-dose albendazole on physical fitness and strength of initially soil-transmitted helminth-infected children.\nThe infection and fitness dynamics were then studied over a 6-month period post-treatment.\nChanges in anthropometric indicators and hemoglobin levels were also measured, and are reported as secondary outcomes.\nMethods\nEthics Statement\nThe study protocol was approved by the institutional research commission of the Swiss Tropical and Public Health Institute (Basel, Switzerland).\nThe ethics committee of Basel (EKBB, reference no. 144/11) and the Academic Board of the National Institute of Parasitic Diseases, Chinese Center for Disease Control and Prevention (Shanghai, P.R. China) provided ethical clearance.\nThe trial is registered with Current Controlled Trials (identifier: ISRCTN 25371788).\nThe village doctor, chief, and teachers of each village were briefed on the aims of the study.\nWith help from the teachers, the investigators further explained the procedures to the children and their parents/guardians.\nWritten informed consent was obtained from parents/guardians, whereas children assented orally.\nData were kept anonymous.\nAfter the 6-month final follow-up, all children attending the five schools were given triple-dose albendazole (3\u00d7400 mg) irrespective of their infection status, study participation, and treatment during the study.\nChildren diagnosed with Strongyloides stercoralis were offered a single dose of ivermectin (200 \u00b5g/kg).\nParticipants\nParticipants were recruited from five primary schools, where a 70% or higher prevalence of soil-transmitted helminth infections had been detected during a rapid appraisal.\nAll schools belonged to villages exclusively inhabited by the Bulang ethnic minority group, and were located in the mountainous Bulangshan township bordering Myanmar, a sub-division of Menghai county in Xishuangbanna Dai autonomous prefecture, situated in Yunnan province, P.R. China.\nThe five villages are: (i) Sandui (geographical coordinates: 21\u00b033\u203207\u2033N latitude, 100\u00b019\u203234\u2033E longitude, altitude: 1,566 m above sea level (asl)); (ii) Kongkan (21\u00b032\u203234\u2033N, 100\u00b020\u203225\u2033E, 1,195 m asl); (iii) Laozhai (21\u00b031\u203237\u2033N, 100\u00b018\u203201\u2033E, 1,399 m asl); (iv) Laonandong (21\u00b033\u203228\u2033N, 100\u00b021\u203245\u2033E, 1,188 m asl); and (v) Mannuo (21\u00b033\u203227\u2033N, 100\u00b023\u203253\u2033E, 1,352 m asl).\nPrior to the current trial, no survey or control activities targeting soil-transmitted helminthiasis had been implemented in the study villages.\nDetailed information on the study area has been published along with data on soil-transmitted helminth re-infection patterns among participants.\nMoreover, the epidemiology and control of soil-transmitted helminthiasis in comparable Bulang communities previously studied by our group have been described elsewhere.\nStudy Design\nThe study was designed as a double-blind, randomized, placebo-controlled trial with three follow-ups, and was carried out between October 2011 and May 2012.\nAssuming a prevalence of 70% with any soil-transmitted helminth infection and 50% loss to follow-up, the trial aimed to enroll 250 children at baseline to achieve a power of 80% at an alpha error of 5% for the detection of a 2.5 ml kg\u22121 min\u22121 difference in the maximum aerobic capacity within 1 min of exhaustive exercise (VO2 max estimate) between the intervention and placebo groups.\nInclusion criteria for the trial were: (i) provision of two stool samples at baseline; (ii) presence of at least one type of soil-transmitted helminth infection; (iii) no deworming treatment within 6 months before the current study; (iv) no known allergy to albendazole; (v) no major systemic illnesses as determined by a medical doctor; (vi) no concurrent participation in other clinical trials; (vii) residency in the study area for at least 1 year before enrolment; and (viii) participant should be between the age of 9 and 12 years.\nChildren aged 9\u201312 years who met the inclusion criteria were enrolled by field investigators for a baseline assessment involving parasitological examination, physical fitness and strength tests, and anthropometric and hemoglobin measurements.\nThe same measurements were repeated 1, 4, and 6 months after treatment, with the exception of anthropometric indicators that were only re-assessed at the 4- and 6-month follow-ups (Figure 1).\nThe treatment allocation sequence was generated by a statistician using block randomization with randomly varying block sizes of 2, 4, and 6.\nAlbendazole and placebo tablets were packaged by staff not involved in the field work into sealed envelopes marked with unique identifiers.\nFollowing the order of the class list provided by the teachers, each child was sequentially assigned a random number, which corresponded to a number on the sealed envelopes.\nBoth children and field investigators were blinded to the nature of the tablets.\nThe assigned triple-dose treatment (i.e., 3\u00d7400 mg albendazole (GlaxoSmithKline; London, United Kingdom) or 3\u00d7 shape- and color-matched placebo (Fagron; Barsb\u00fcttel, Germany)), was started on treatment day 1 with a single dose, with subsequent doses administered every day until treatment day 3.\nThe field investigators directly observed the consumption of each treatment by all children.\nField and Laboratory Procedures\nTwo stool samples were collected from each child on consecutive days.\nBoth the Kato-Katz (duplicate slides per sample) and Baermann techniques (one examination per sample) were used; Kato-Katz for the detection of eggs of A. lumbricoides, hookworm, and T. trichiura, and Baermann for larvae of S. stercoralis .\nAdditionally, stool samples were visually inspected for Taenia spp. proglottids.\nUsing the Kato-Katz technique, eggs of A. lumbricoides, hookworm, and T. trichiura were counted separately and the results of both slides averaged.\nThe mean was then multiplied by a factor of 24 to obtain the number of eggs per 1 g of stool (EPG).\nFor quality control purposes, the two Kato-Katz slides were examined independently and results compared.\nSlides were re-read if inconsistencies were detected.\nAn inconsistency was defined as a difference in the infection intensity (EPG) groupings based on WHO guidelines.\nPhysical fitness was assessed with a 20-m shuttle run test.\nThe running speed from the last completed 20-m lap and the total number of intervals completed were recorded.\nThe child's age and speed were then converted into a VO2 max estimate (to the nearest 0.1 ml kg\u22121 min\u22121) with an equation put forth by L\u00e9ger et al. .\nPhysical strength was assessed with a grip strength and a standing broad jump test.\nFor the grip strength test, the hand span (distance from the tip of the thumb to the tip of the little finger) of the child's dominant hand was measured (to the nearest 0.5 cm) and an electronic dynamometer (Yi Lian Medicine; Shanghai, P.R. China) adjusted accordingly to provide the optimal grip span.\nChildren were asked to stand straight yet relaxed, and grip the dynamometer with the dominant hand as hard as possible for 5 sec, with the arm fully extended and without other parts of the body touching it.\nEach child had two tries (with a 15-sec rest in between), but only the maximum reading was recorded, to the nearest 0.1 kg.\nFor the standing broad jump test, each child, standing behind a straight line, had two tries (with a 15-sec rest in between) to jump as far forward as possible with both legs.\nThe jumps were recorded to the nearest 1 cm and the longer jump considered for an individual.\nThe distance of the jump was measured from the starting line to the heel of the most back foot.\nFor the measurement of body height and weight, children were asked to take off their shoes and sweater before standing on a digital weighing scale (Model RCS-150; Nantong Xineng Ltd., Jiangsu, P.R. China) or stadiometer (Nantong Xineng Ltd., Jiangsu, P.R. China).\nBoth height and weight were recorded twice, to the nearest 0.1 cm or kg, respectively, and averaged.\nThe body mass index (BMI) was defined as (weight in kg)/(height in m)2; the BMI-for-age Z score (BAZ) and height-for-age Z score (HAZ) were used as indicators for wasting and stunting, respectively.\nThe thickness of skinfolds was measured at two sites, namely triceps and subscapular, with the Holtain skinfold caliper (Holtain Ltd.; Crymych, United Kingdom).\nMeasurements were performed in triplicate to the nearest 1 mm and averaged.\nThe sum of the mean skinfolds at both sites was used as an estimate for body fat.\nThe hemoglobin level was measured once, to the nearest 1 g l\u22121, with a HemoCue Hb 301 system (HemoCue AB.;\n\u00c4ngelholm, Sweden) using a drop of blood from the ear lobe.\nAnemia was defined according to WHO age- and sex-specific cut-offs.\nThe socioeconomic status of the participants at baseline was assessed through a questionnaire asking the children for the education level of their parents and the main source of their household income.\nThe full trial protocol is available as supporting information (Protocol S1).\nStatistical Analysis\nData were entered into Excel version 2008 (Microsoft Corp.; Redmond, United States of America), double-checked, and merged into a single database for statistical analysis with STATA version 10.0 (STATA Corp.; College Station, United States of America).\nThe randomization code was broken after data entry and a series of internal consistency checks were completed.\nA per-protocol analysis was carried out in an un-blinded manner.\nIn the primary analysis, physical fitness and strength scores, anthropometric measurements, and hemoglobin levels were expressed as means, and changes in the means between baseline and treatment follow-ups were compared between treatment groups in a multivariate linear regression model.\nIn a sub-analysis, changes in the means of physical fitness and strength indicators between baseline and follow-up were compared among children of distinct soil-transmitted helminth infection status regardless of treatment status.\nTo further explore the effect of infection intensity on these measurements, distinct groups of children were identified using principal component and cluster analysis, based on species-specific soil-transmitted helminth log transformed egg counts at baseline and at the 1- and 4-month follow-ups.\nChanges in the means of physical fitness and strength indicators between baseline and follow-up were compared among six biologically meaningful groups of varying soil-transmitted helminth infection intensity.\nResults\nBaseline Characteristics\nAs illustrated in Figure 1, an overall compliance of 92% was achieved with only nine children lost to follow-up over the 6-month trial period.\nThese nine children were sick and absent during the follow-ups.\nComplete datasets were available for 99 children in the albendazole group and 95 children in the placebo group.\nNo noteworthy difference in baseline socio-demographics or soil-transmitted helminth infection prevalence and intensity was observed between the albendazole and placebo groups (Table 1).\nChildren were also comparable in terms of physical fitness and strength at baseline.\nThe mean VO2 max estimate and number of 20-m laps completed were 44.9 ml kg\u22121 min\u22121 (standard deviation (SD): 2.8 ml kg\u22121 min\u22121) and 23.8 laps (SD: 8.7 laps), respectively, for the albendazole group, and 45.4 ml kg\u22121 min\u22121 (SD: 3.2 ml kg\u22121 min\u22121) and 25.2 laps (SD: 9.9 laps), respectively, for the placebo group.\nWith regard to physical strength, the mean grip strength and standing broad jump distance were 12.6 kg (SD: 4.1 kg) and 142 cm (SD: 14 cm), respectively, for the albendazole group, and 12.6 kg (SD: 3.5 kg) and 142 cm (SD: 14 cm), respectively, for the placebo group.\nHowever, when further stratified by sex, boys had higher physical fitness and strength than girls (statistical significance achieved for all indicators except grip strength) (Table 1).\nTreatment groups were also comparable in terms of anthropometric and hematologic characteristics at baseline.\nDespite the mean BMI for the albendazole and placebo group being 16.0 (SD: 1.3) and 16.1 (SD: 1.3) respectively, the baseline prevalence of wasting was only 3.6%.\nOn the other hand, stunting was present in 76.8% (mean height for the albendazole and placebo group was 127.4 cm (SD: 8.9 cm) and 126.3 cm (SD: 7.1 cm), respectively) of the cohort.\nBaseline prevalence of anemia was low at 4.6%, as the mean hemoglobin levels for the albendazole and placebo group were 161 g l\u22121 (SD: 24 g l\u22121) and 157 g l\u22121 (SD: 28 g l\u22121), respectively.\nEffects of Deworming on Primary and Secondary Outcomes\nChildren receiving triple-dose albendazole experienced a greater, but mostly not statistically significant, change in the means of their physical fitness scores than their peers from the placebo group at all three follow-ups (Table 2).\nVO2 max estimates increased by 1.0\u20132.3 ml kg\u22121 min\u22121 from baseline over the 6-month trial period for the albendazole group, while the increase for the placebo group ranged from 0.2\u20132.1 ml kg\u22121 min\u22121.\nLikewise for the number of 20-m laps completed, the range of increase was 3.4\u201311.9 laps and 1.4\u201311.7 laps for the albendazole and placebo group, respectively.\nWhen adjusted for village, and at the individual level for sex, age, height, and weight at follow-up, the difference in the increase of physical fitness between both groups was highest at the 1-month follow-up, where the increase from baseline in the albendazole group was 0.9 ml kg\u22121 min\u22121 (P\u200a=\u200a0.05) or 2.1 laps (P\u200a=\u200a0.14) higher than the placebo group.\nWith regard to physical strength, the grip strength increased 0.8\u20132.0 kg from baseline for the albendazole group, while the placebo group experienced an increase of 0.4\u20131.8 kg.\nThe difference in the increase between both groups was highest at the 1-month follow-up (0.3 kg higher in the albendazole group), but this difference was not statistically significant.\nThe largest change in standing broad jump distance from baseline was observed among the albendazole group at the 1-month follow-up (+2 cm), but the placebo group fared better at the 6-month follow-up (+2 cm).\nHowever, both of these increases were not statistically significant.\nIn terms of secondary outcomes, children in the albendazole group had a larger increase, from baseline, in the means of their body height and weight and sum of skinfolds than their counterparts from the placebo group (Table 3).\nThe range of increase for body height, weight, and sum of skin folds were 2.9\u20133.5 cm, 1.4\u20132.2 kg, and 1 mm, respectively, for the albendazole group, and 2.7\u20133.3 cm, 1.2\u20131.9 kg, and 1 mm for the placebo group.\nHowever, differences between both groups in the change from baseline were statistically non-significant at all follow-ups after adjusting for sex, age at follow-up, and village.\nA reduction in hemoglobin level was observed in both groups at the 1- and 6-month follow-ups, and the respective reduction in the albendazole group was 2 g l\u22121 (P\u200a=\u200a0.72) and 3 g l\u22121 (P\u200a=\u200a0.49) higher compared to the placebo group\nOn the other hand, the increase from baseline in the albendazole group was 3 g l\u22121 higher than the placebo group at the 4-month follow-up (P\u200a=\u200a0.65).\nEffects of Soil-Transmitted Helminth Infections on Primary Outcomes\nWhen the soil-transmitted helminth infection status was used as explanatory variable for the primary outcomes (Table 4), T. trichiura-infected children had 1.6 ml kg\u22121 min\u22121 less increase in their VO2 max estimate from baseline than their non-infected peers at the 1-month follow-up (P\u200a=\u200a0.012).\nSimilarly, hookworm-infected children had 1.1 ml kg\u22121 min\u22121 less increase in their VO2 max estimate from baseline than their non-infected peers at the 6-month follow-up (P\u200a=\u200a0.03).\nIn addition, the increase in the number of 20-m laps completed from baseline was 4.6 (P\u200a=\u200a0.04) and 6.0 (P\u200a=\u200a0.01) laps less for T. trichiura-infected children than their non-infected counterparts at the 1- and 4-month follow-ups, respectively.\nAs further illustrated in Figure 2, an increase from baseline (positive change) in the number of 20-m laps completed at the 4-month follow-up was more dependent on a reduction in T. trichiura infection intensity than diminished A. lumbricoides and hookworm infection intensity.\nIn terms of grip strength at the 1-month follow-up (Table 4), the increase from baseline among A. lumbricoides-infected children was 0.8 kg lower than among children not infected with this helminth species (P\u200a=\u200a0.05), but hookworm-infected children had 0.9 kg more increase from baseline than their non-infected peers (P\u200a=\u200a0.04).\nNo statistically significant change in standing broad jump distance due to soil-transmitted helminth infection status was observed at each of the three follow-ups.\nWhen the children were grouped according to their longitudinal infection intensity patterns, the six groups that emerged (Figure 3) had the following characteristics: group 1, high infection intensity of all species at all time-points; group 2, high infection intensity of all species except hookworm at all time-points; group 3, high intensity of A. lumbricoides re-infection by the 4-month follow-up, high infection intensity of T. trichiura at all time-points, and no or minimal hookworm re-infection at follow-ups; group 4, low intensity of A. lumbricoides re-infection by the 4-month follow-up, intermediate infection intensity of T. trichiura at all time-points, and no or minimal hookworm re-infection at follow-ups; group 5, intermediate intensity of A. lumbricoides re-infection by the 4-month follow-up, and no or minimal T. trichiura and hookworm re-infection at follow-ups; and group 6, infection intensity of all species were higher during the follow-ups compared to the pre-treatment baseline.\nWhen group 1 was used as the reference group in the multivariate linear regression models (Table 5), children from group 5 had consistently more increase in their VO2 max estimate from baseline than their peers from group 1 at all follow-ups (range: 1.9\u20132.1 ml kg\u22121 min\u22121; all P<0.05).\nA similar trend was observed for the number of 20-m laps completed and a statistically significant 5.7 more 20-m laps were completed by children from group 5 as compared to group 1 (P\u200a=\u200a0.04) and the baseline.\nIn terms of standing broad jump distance, children from group 4 had a 6 cm higher increase from baseline than children from group 1 at the 4-month follow-up (P\u200a=\u200a0.03).\nNo statistically significant change in grip strength dependent on soil-transmitted helminth infection intensity was observed at all follow-ups.\nDiscussion\nAs shown in our preceding work in Bulang communities, the prevalence and intensity of soil-transmitted helminth infections in villages inhabited by this ethnic minority group can be very high.\nFor example, in the current randomized controlled trial, we found baseline prevalence of T. trichiura, A. lumbricoides, and hookworm at 94.5%, 93.3%, and 61.3%, respectively.\nTherefore, an intensive de-worming regimen, consisting of triple-dose albendazole, was employed to allow children a fair chance of developing their physical fitness unaffected by intestinal helminth infections.\nRe-infection with A. lumbricoides occurred more rapidly than expected and the prevalence of A. lumbricoides reached 80% of the pre-treatment prevalence 4 months after treatment.\nDespite triple-dose albendazole treatment, a low cure rate of 19.6% was obtained against T. trichiura, corroborating previous conclusions that T. trichiura infection is particularly hard to cure with current anthelmintic drugs.\nSuch re-infection dynamics have complicated the evaluation of the potential health benefits of deworming and rendered the grouping of the children according to intervention near-irrelevant as the treated children might not have benefited from a meaningful helminth-free period for substantial catch-up growth.\nThis finding further suggests that in our study area, the current WHO recommendation of single-dose albendazole (400 mg) twice yearly might be insufficient for controlling soil-transmitted helminthiasis.\nIn a recently published trial conducted in India, where 1 million preschool-aged children, 1 to 6 years old at baseline, were treated with albendazole every 6 months for 5 years, no statistically significant difference in anthropometric measurements was detected between the albendazole and control groups.\nIn our study, even though a trend toward higher values was observed for the treated cohort, no statistically significant difference in most primary and secondary outcomes between the albendazole and placebo groups was detected during the 6-month follow-up period.\nHowever, we did find one statistically significant, and biologically important, difference in the VO2 max estimate at 1-month follow-up between the albendazole and placebo groups despite the relatively small sample size.\nIt is interesting to note the variation in hemoglobin levels for both groups throughout the follow-up period and this could probably be due to seasonal dietary changes or the presence of other infections.\nIn addition, we observed a general learning effect with the physical fitness and strength tests, especially the 20-m shuttle run test, amongst the children, but this was mitigated in the analysis by having a control group.\nIn the sub-group analysis, we found that soil-transmitted helminth-infected children had performed significantly worse in the physical fitness and strength tests than their non-infected peers.\nWhen we grouped children according to their infection status at each follow-up, we observed that T. trichiura-infected children performed worse in the 20-m shuttle run than their non-infected peers.\nThis confirmed the results from a cross-sectional study conducted by our group where T. trichiura-infected children were found to complete, on average, 6.1 20-m laps less and have a VO2 max estimate which was 1.9 ml kg\u22121 min\u22121 lower than their non-infected counterparts.\nTo survive in a host, adult T. trichiura worms anchor their whip-like anterior end into the wall of the large intestine and caecum by secreting pore-forming proteins.\nSuch an invasive mechanism causes inflammation and bleeding, resulting in abdominal pain in the short term, and anemia and rectal prolapse in the long term, especially when large numbers of worms are present.\nA significant change in physical fitness already at the 1-month follow-up could indicate that removing abdominal pain alone through the expulsion of T. trichiura might enhance the host's endurance in exhaustive exercises, such as the 20-m shuttle run.\nHookworm-infected children were also found to have a significantly lower increase from baseline in their VO2 max estimates than children not infected with hookworm at the 6-month follow-up.\nAlthough anemia is a known symptom of hookworm infection and would be a plausible cause for reduced VO2 max estimates, it was detected in only 10.7% of the hookworm-infected children and no significant association was found between any soil-transmitted helminth infection and hemoglobin level.\nThe migration of the hookworm larvae through the pulmonary blood vessels, where they bore into the alveoli, could offer an alternative explanation to this observation.\nAlthough the larvae of A. lumbricoides undergo a similar migratory process, no reduction in VO2 max estimates was observed in children infected with A. lumbricoides.\nIn terms of grip strength, the increase from baseline among A. lumbricoides-infected children was significantly lower, while hookworm-infected children had a higher increase from baseline, when compared to their non-infected peers.\nAs there is currently limited evidence on the association of soil-transmitted helminth infection and grip strength, these inconsistent findings warrant further investigation.\nWhen children were grouped according to infection intensity, we were able to take into consideration the degree of infection at baseline, 1-month, and 4-month follow-ups, and the extent of multiparasitism for each child.\nThese analyses revealed that individuals with a combination of no or minimal T. trichiura and hookworm re-infection achieved higher improvements during the follow-ups in the 20-m shuttle run, as compared to peers with high infection intensity of all species.\nIn addition, children with no or minimal A. lumbricoides and hookworm re-infection performed better in the standing broad jump than their counterparts with high infection intensity of all species.\nThese findings provide further evidence of the impact of soil-transmitted helminth infections on the physical fitness and strength of school-aged children.\nThe anthropometric and physical strength findings from this trial should be viewed in the light of the following limitations.\nA follow-up period of 6 months is probably too short for an accurate evaluation of anthropometric gains and physical strength increments from longer-term physical growth due to deworming.\nTaking into account that keeping controls untreated for a long period would be difficult due to ethical considerations, a 3- to 5-year prospective cohort study, where children are treated regularly to ensure that they are helminth-free, and the changes in anthropometric indicators and physical strength from baseline are monitored and compared with changes in soil-transmitted helminth infection intensity over time, could be a more appropriate study design.\nFinally, catch-up growth after anthelmintic treatment can only occur if the diet is sufficient.\nBased on the investigators' observations in the field, most of the children's diet consists mainly of white rice with little protein sources.\nDietary improvements, in addition to deworming, are therefore necessary in the current setting for catch-up growth to occur and perhaps to aid in the absorption of albendazole, and should be considered in future, more comprehensive studies.\nWe conclude that there is no strong evidence for significant improvements in physical fitness and anthropometric indicators due to deworming with triple-dose albendazole.\nThis might be partly explained by the rapid re-infection observed with A. lumbricoides and low cure rates with T. trichiura.\nHowever, negative impacts on the physical fitness and strength were observed in school-aged children infected with soil-transmitted helminths in sub-group analyses.\nIn particular, the clear effects of T. trichiura infection on physical fitness in this trial is intriguing as the public health burden of this helminth species is currently not as well defined as that of the other two species.\nThe fact that T. trichiura infection had the strongest negative impact on the physical fitness of the children but was hardly cured with triple-dose albendazole is another major concern.\nFinally, we also showed that the observed morbidities were infection intensity-dependent and in order to control them, regular deworming, coupled with dietary improvements and water, sanitation, and hygiene development, should be considered.\nProfile of randomized controlled trial conducted in south-west Yunnan province, P.R. China, from October 2011 to May 2012.\nThree-dimensional visualization of changes in 20-m laps due to differences in soil-transmitted helminth infection intensities in Chinese children.The study was carried out in south-west Yunnan province, P.R. China between October 2011 and May 2012 among 194 children aged 9\u201312 years. The changes in infection intensities between the 4-month follow-up and baseline for Ascaris lumbricoides, Trichuris trichiura, and hookworm are reflected along the X-, Y-, and Z-axis, respectively. Each circle indicates the change for a particular child harboring a certain mixture of soil-transmitted helminth infection intensities. Blue circles indicate positive change, red circles indicate negative change, and white circles indicate no change. A darker shade of color indicates a greater degree of change.\nBoxplots of six infection intensity groups identified by principal component and cluster analysis.The groups are based on infection intensities of the three soil-transmitted helminths at baseline (white), 1-month follow-up (light grey), and 4-month follow-up (dark grey), among 194 children from a randomized controlled trial conducted in south-west Yunnan province, P.R. China, from October 2011 to May 2012.\n\nBaseline characteristics of 194 children from south-west Yunnan province, P.R. China, who participated in a randomized controlled trial conducted from October 2011 to May 2012.\nCharacteristicsa | Triple-dose albendazole (n\u200a=\u200a99) | Placebo (n\u200a=\u200a95)\n | Male (n\u200a=\u200a46) | Female (n\u200a=\u200a53) | Male (n\u200a=\u200a48) | Female (n\u200a=\u200a47)\nAge [years] | 10.4 (1.1) | 10.5 (1.2) | 10.4 (1.0) | 10.2 (1.0)\nIlliterate parents | 32 (69.6%) | 27 (50.9%) | 28 (58.3%) | 22 (46.8%)\nFamily relying on farming for income | 46 (100%) | 53 (100%) | 48 (100%) | 47 (100%)\nSoil-transmitted helminth prevalence |  |  |  | \nAscaris lumbricoides | 46 (100%) | 48 (90.6%) | 43 (89.6%) | 44 (93.6%)\nTrichuris trichiura | 44 (95.7%) | 48 (90.6%) | 47 (97.9%) | 44 (93.6%)\nHookworm | 31 (67.4%) | 29 (54.7%) | 34 (70.8%) | 25 (53.2%)\nTaenia spp. | n.r. | n.r. | 1 (2.1%) | n.r.\nStrongyloides stercoralis | n.r. | 2 (3.8%) | 1 (2.1%) | 3 (6.4%)\nSoil-transmitted helminth infection intensity |  |  |  | \nAscaris lumbricoides [EPG] | 17,163 (7,548\u201359,106) | 14,652 (5,922\u201351,324) | 21,579 (4,653\u201343,425) | 21,432 (6,390\u201365,532)\nTrichuris trichiura [EPG] | 150 (60\u2013690) | 222 (72\u2013522) | 219 (90\u2013741) | 318 (108\u2013738)\nHookworm [EPG] | 72 (0\u2013204) | 30 (0\u2013132) | 48 (0\u2013126) | 12 (0\u2013198)\nPhysical fitness |  |  |  | \nVO2 max estimate [ml kg\u22121 min\u22121] | 45.5 (3.0) | 44.4 (2.6) | 45.7 (3.1) | 45.0 (3.3)\n20-m laps completed | 25.4 (8.9) | 22.4 (8.3) | 27.2 (9.3) | 23.2 (10.1)\nPhysical strength |  |  |  | \nGrip strength [kg] | 12.9 (3.7) | 12.5 (4.5) | 13.0 (3.5) | 12.1 (3.5)\nStanding broad jump distance [cm] | 144 (14) | 140 (14) | 146 (14) | 138 (13)\nAnthropometric indicators |  |  |  | \nBody height [cm] | 124.8 (6.6) | 129.7 (10.0) | 125.6 (7.4) | 127.0 (6.7)\nBody weight [kg] | 25.0 (3.8) | 27.3 (6.1) | 25.5 (4.2) | 26.2 (4.7)\nBody mass index [BMI; kg m\u22122] | 16.0 (1.1) | 16.0 (1.4) | 16.1 (1.1) | 16.1 (1.6)\nWastedb | 2 (4.4%) | 2 (3.8%) | n.r. | 3 (6.4%)\nStuntedc | 37 (80.4%) | 36 (67.9%) | 39 (81.3%) | 37 (78.7%)\nSum of skinfolds [mm] | 10 (2) | 12 (3) | 10 (2) | 12 (4)\nHematologic |  |  |  | \nHemoglobin level [g l\u22121] | 159 (19) | 162 (28) | 159 (24) | 155 (32)\nAnemicd | n.r. | 3 (5.7%) | 1 (2.1%) | 5 (10.6%)\n\nValues are number of children (%) or mean (standard deviation; SD). For infection intensity, data is presented as median (interquartile range).\nWasting is defined as \u2264\u22122 BAZ score.\nStunting is defined as \u2264\u22122 HAZ score.\nAnemia is defined according to WHO age-specific cut-offs: Hb<115 g l\u22121 for ages <12 years; Hb<120 g l\u22121 for ages \u226512 and <15 years.\nn.a.: not applicable; n.r.: not represented.\n\nEffects of deworming on changes in physical fitness and strength indicators (primary outcomes) at various follow-ups from baseline among 194 children from south-west Yunnan province, P.R. China, who participated in a randomized controlled trial conducted from October 2011 to May 2012.\n | 1-month follow-up | 4-month follow-up | 6-month follow-up\nCharacteristicsa | ALB (n\u200a=\u200a99) | PLB (n\u200a=\u200a95) | Difference in \u0394 from baselineb | ALB (n\u200a=\u200a99) | PLB (n\u200a=\u200a95) | Differences in \u0394 from baselineb | ALB (n\u200a=\u200a99) | PLB (n\u200a=\u200a95) | Differences in \u0394 from baselineb\nVO2 max estimate [ml kg\u22121 min\u22121] | 45.9 (1.0) | 45.5 (0.2) | 0.9 (0.0 to 1.8) | 45.7 (0.8) | 46.2 (0.8) | 0.2 (\u22120.8 to 1.1) | 47.2 (2.3) | 47.5 (2.1) | 0.2 (\u22120.7 to 1.2)\n20-m laps completed | 27.2 (3.4) | 26.6 (1.4) | 2.1 (\u22120.7 to 5.0) | 29.4 (5.6) | 30.2 (5.0) | 0.9 (\u22122.1 to 3.9) | 35.7 (11.9) | 36.9 (11.7) | 0.2 (\u22122.7 to 3.1)\nGrip strength [kg] | 13.4 (0.8) | 13.0 (0.4) | 0.3 (\u22120.3 to 0.9) | 14.5 (1.8) | 14.2 (1.6) | 0.1 (\u22120.5 to 0.8) | 14.6 (2.0) | 14.4 (1.8) | 0.1 (\u22120.6 to 0.9)\nStanding broad jump distance [cm] | 142 (0) | 141 (\u22121) | 2 (\u22122 to 5) | 146 (4) | 145 (3) | 1 (\u22122 to 4) | 146 (5) | 149 (7) | \u22122 (\u22125 to 2)\n\nValues are mean (\u0394 from baseline), unless otherwise stated.\nDifferences in the changes between follow-up and baseline among the intervention groups are adjusted for village, and at the individual level for sex, age at follow-up, and height and weight at baseline (for the 1-month follow-up) or follow-up (for the 4- and 6-month follow-ups). Values are calculated from a multivariate linear regression model, presented as coefficient (95% confidence interval) and highlighted in bold if statistical significance is achieved (P<0.05).\nALB: triple-dose albendazole; PLB: placebo.\n\nEffects of de-worming on changes in nutritional indicators (secondary outcomes) at various follow-ups from baseline among 194 children from south-west Yunnan province, P.R. China, who participated in the randomized controlled trial conducted from October 2011 to May 2012.\n | 1-month follow-up | 4-month follow-up | 6-month follow-up\nCharacteristicsa | ALB (n\u200a=\u200a99) | PLB (n\u200a=\u200a95) | Differences in \u0394 from baselineb | ALB (n\u200a=\u200a99) | PLB (n\u200a=\u200a95) | Differences in \u0394 from baselineb | ALB (n\u200a=\u200a99) | PLB (n\u200a=\u200a95) | Differences in \u0394 from baselineb\nBody height [cm] | n.d. | 130.3 (2.9) | 129.0 (2.7) | 0.2 (\u22120.1 to 0.4) | 130.9 (3.5) | 129.6 (3.3) | 0.2 (\u22120.1 to 0.4)\nBody weight [kg] |  | 27.6 (1.4) | 27.0 (1.2) | 0.2 (\u22120.1 to 0.4) | 28.4 (2.2) | 27.7 (1.9) | 0.2 (\u22120.1 to 0.6)\nStunted (%)c | n.d. | 63 (\u221210.1%) | 70 (\u22126.3%) | n.a.d | 66 (\u22127.0%) | 69 (\u22127.4%) | n.a.d\nSum of skinfolds [mm] |  | 12 (1) | 12 (1) | 0 (0 to 1) | 12 (1) | 12 (1) | 0 (0 to 1)\nHemoglobin level [g l\u22121] | 151 (\u221210) | 150 (\u22127) | \u22122 (\u221210 to 7) | 164 (3) | 159 (2) | 3 (\u22128 to 13) | 148 (\u221213) | 148 (\u22129) | \u22123 (\u221211 to 6)\n\nValues are number of children (% change from baseline) or mean (\u0394 from baseline), unless otherwise stated.\nDifferences in the changes between follow-up and baseline among the intervention groups are adjusted for village, and at the individual level for sex and age at follow-up. Values are calculated from a multivariate linear regression model, presented as coefficient (95% confidence interval) and highlighted in bold if statistical significance is achieved (P<0.05). For percentage stunted, the \u03c72 test is used and thus, no 95% confidence interval is presented.\nStunting is defined as \u2264\u22122 HAZ score.\n\nP-value calculated from \u03c72 test comparing % stunted between ALB and PLB for statistical significance.\nALB: triple-dose albendazole; PLB: placebo; n.d.: not determined.\n\nEffects of soil-transmitted helminth infection status on changes in physical fitness and strength indicators (primary outcomes) at various follow-ups from baseline among 194 children from south-west Yunnan province, P.R. China, who participated in the randomized controlled trial conducted from October 2011 to May 2012.\n | 1-month follow-up | 4-month follow-up | 6-month follow-up\nMultivariate linear regression modelsa | n | Coefficient (95% CI) | n | Coefficient (95% CI) | n | Coefficient (95% CI)\n(A) Change in VO2 max estimate [ml kg\u22121 min\u22121] |  |  |  | \nAscaris lumbricoides-infected | 97 | \u22120.1 (\u22121.2 to 1.0) | 167 | 0.7 (\u22120.8 to 2.2) | 175 | 1.2 (\u22120.4 to 2.8)\nTrichuris trichiura-infected | 166 | \u22121.6 (\u22123.0 to \u22120.3) | 170 | \u22121.5 (\u22123.1 to 0.1) | 175 | 0.5 (\u22121.1 to 2.1)\nHookworm-infected | 59 | \u22120.2 (\u22121.5 to 1.0) | 51 | 0.4 (\u22120.8 to 1.6) | 56 | \u22121.1 (\u22122.1 to \u22120.1)\n(B) Change in 20-m intervals completed |  |  |  | \nAscaris lumbricoides-infected | 97 | \u22120.2 (\u22123.7 to 3.4) | 167 | 3.1 (\u22121.4 to 7.5) | 175 | 3.4 (\u22121.7 to 8.5)\nTrichuris trichiura-infected | 166 | \u22124.6 (\u22128.9 to \u22120.3) | 170 | \u22126.0 (\u221210.7 to \u22121.2) | 175 | 1.2 (\u22123.9 to 6.3)\nHookworm-infected | 59 | \u22120.1 (\u22124.1 to 3.9) | 51 | 2.0 (\u22121.5 to 5.5) | 56 | \u22122.2 (\u22125.5 to 1.1)\n(C) Change in grip strength [kg] |  |  |  | \nAscaris lumbricoides-infected | 97 | \u22120.8 (\u22121.5 to 0.0) | 167 | \u22120.5 (\u22121.5 to 0.5) | 175 | \u22121.2 (\u22122.4 to 0.1)\nTrichuris trichiura-infected | 166 | 0.0 (\u22120.9 to 0.8) | 170 | 0.9 (\u22120.1 to 2.0) | 175 | 0.7 (\u22120.6 to 2.0)\nHookworm-infected | 59 | 0.9 (0.1 to 1.7) | 51 | 0.1 (\u22120.7 to 0.9) | 56 | 0.7 (\u22120.1 to 1.5)\n(D) Change in standing broad jump distance [cm] |  |  |  | \nAscaris lumbricoides-infected | 97 | \u22122 (\u22126 to 2) | 167 | \u22122 (\u22127 to 3) | 175 | 1 (\u22124 to 7)\nTrichuris trichiura-infected | 166 | 5 (0 to 10) | 170 | 5 (0 to 10) | 175 | 5 (\u22121 to 11)\nHookworm-infected | 59 | 0 (\u22124 to 5) | 51 | \u22123 (\u22127 to 1) | 56 | 1 (\u22123 to 5)\n\nFor each model, the outcome variable is highlighted with a grey bar and the explanatory variables (reference group is always not infected with the particular soil-transmitted helminth species) are presented below. All models have been adjusted for village, and at the individual level for sex, age at follow-up, and height and weight at baseline (for the 1-month follow-up) or follow-up (for the 4- and 6-month follow-ups). Values are presented as coefficient (95% confidence interval) and highlighted in bold if statistical significance is achieved (P<0.05).\n\nEffects of soil-transmitted helminth infection intensity on changes in physical fitness and strength indicators (primary outcomes) at various follow-ups from baseline among 194 children from south-west Yunnan province, P.R. China, who participated in a randomized controlled trial conducted from October 2011 to May 2012.\n | 1-month follow-up | 4-month follow-up | 6-month follow-up\nMultivariate linear regression modelsa | Coefficient (95% CI) | Coefficient (95% CI) | Coefficient (95% CI)\n(A) Change in VO2 max estimate [ml kg\u22121 min\u22121] | \nGroup 2 (n\u200a=\u200a36) | \u22120.2 (\u22121.6 to 1.2) | \u22120.2 (\u22121.7 to 1.3) | 0.7 (\u22120.7 to 2.1)\nGroup 3 (n\u200a=\u200a53) | 0.3 (\u22120.9 to 1.5) | \u22120.1 (\u22121.4 to 1.3) | 0.1 (\u22121.2 to 1.4)\nGroup 4 (n\u200a=\u200a27) | 1.2 (\u22120.3 to 2.7) | 0.2 (\u22121.5 to 1.9) | 0.7 (\u22120.8 to 2.2)\nGroup 5 (n\u200a=\u200a22) | 1.9 (0.3 to 3.5) | 2.1 (0.3 to 3.9) | 1.9 (0.2 to 3.6)\nGroup 6 (n\u200a=\u200a10) | 0.3 (\u22121.9 to 2.4) | 0.9 (\u22121.5 to 3.2) | 0.3 (\u22121.9 to 2.5)\n(B) Change in 20-m intervals completed |  | \nGroup 2 (n\u200a=\u200a36) | \u22121.4 (\u22125.8 to 3.0) | \u22122.2 (\u22126.7 to 2.4) | 0.5 (\u22124.0 to 5.0)\nGroup 3 (n\u200a=\u200a53) | \u22120.2 (\u22124.1 to 3.8) | \u22120.8 (\u22124.8 to 3.3) | \u22121.6 (\u22125.6 to 2.5)\nGroup 4 (n\u200a=\u200a27) | 2.5 (\u22122.3 to 7.3) | 0.3 (\u22124.7 to 5.3) | 2.7 (\u22122.3 to 7.7)\nGroup 5 (n\u200a=\u200a22) | 5.1 (0.0 to 10.3) | 5.7 (0.4 to 11.1) | 3.9 (\u22121.5 to 9.2)\nGroup 6 (n\u200a=\u200a10) | \u22121.7 (\u22128.5 to 5.1) | 1.4 (\u22125.6 to 8.5) | 0.3 (\u22126.7 to 7.3)\n(C) Change in grip strength [kg] |  | \nGroup 2 (n\u200a=\u200a36) | \u22120.6 (\u22121.5 to 0.3) | \u22120.7 (\u22121.7 to 0.3) | \u22121.2 (\u22122.4 to \u22120.1)\nGroup 3 (n\u200a=\u200a53) | \u22120.2 (\u22121.1 to 0.6) | \u22120.1 (\u22121.0 to 0.8) | \u22120.5 (\u22121.5 to 0.5)\nGroup 4 (n\u200a=\u200a27) | \u22120.4 (\u22121.4 to 0.7) | 0.3 (\u22120.8 to 1.4) | \u22120.1 (\u22121.4 to 1.1)\nGroup 5 (n\u200a=\u200a22) | \u22120.5 (\u22121.6 to 0.6) | \u22120.8 (\u22121.9 to 0.4) | \u22120.8 (\u22122.2 to 0.5)\nGroup 6 (n\u200a=\u200a10) | 0.0 (\u22121.5 to 1.4) | 0.0 (\u22121.6 to 1.5) | \u22120.1 (\u22122.3 to 1.2)\n(D) Change in standing broad jump distance [cm] | \nGroup 2 (n\u200a=\u200a36) | \u22121 (\u22126 to 4) | 1 (\u22124 to 6) | \u22122 (\u22127 to 3)\nGroup 3 (n\u200a=\u200a53) | 1 (\u22124 to 5) | 3 (\u22121 to 7) | \u22121 (\u22125 to 4)\nGroup 4 (n\u200a=\u200a27) | 4 (\u22122 to 9) | 6 (1 to 12) | 0 (\u22125 to 6)\nGroup 5 (n\u200a=\u200a22) | \u22123 (\u22129 to 3) | \u22121 (\u22127 to 4) | \u22125 (\u221211 to 1)\nGroup 6 (n\u200a=\u200a10) | \u22123 (\u221211 to 5) | 1 (\u22127 to 9) | 1 (\u22127 to 9)\n\nFor each model, the outcome variable is highlighted with a grey bar and the explanatory variables (reference group is always group 1) are presented below. All models have been adjusted for village, and at the individual level for sex, age at follow-up, and height and weight at baseline (for the 1-month follow-up) or follow-up (for the 4-month follow-up). Values are presented as coefficient (95% confidence interval) and highlighted in bold if statistical significance is achieved (P<0.05).",
        "label": "low",
        "id": "task4_RLD_test_746"
    },
    {
        "paper_doi": "10.1186/1471-2334-13-482",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: A randomized, open label, trial conducted at All India Institute of Medical Sciences, New Delhi, India.\n\n\nParticipants: 142 participantsInclusion criteria: positive for HIV by ELISA, ART-naive and presenting with concomitant TB, CD4 count < 200 cells/mm3 and normal renal and hepatic function.Exclusion criteria: positive in hepatitis B and C serologies, taking antiepileptic drugs, immunosuppressant, and other drugs that induce liver microsomal enzyme systems, and pregnant.\n\n\nInterventions: Zidovudine or Stavudine and Lamivudine/NVP once daily for the first 14 days and twice daily thereafter (200mg) or Zidovudine or Stavudine and Lamivudine/EFV (600 mg) daily.\n\n\nOutcomes: Virological response (< 400 copies/mL), discontinuation rate, treatment failure, mortality, and adverse events\n\n\nNotes: All participants gave signed informed consent to participate in this study. Funding was provided by the National AIDS Control Organization, Ministry of Health & Family Welfare, and the Government of India\n\n",
        "objective": "To determine which non\u2010nucleoside reverse transcriptase inhibitor, either EFV or NVP, is more effective in suppressing viral load when given in combination with two nucleoside reverse transcriptase inhibitors as part of initial antiretroviral therapy for HIV infection in adults and children.",
        "full_paper": "Background\nAdministration of rifampicin along with nevirapine reduces the plasma concentration of nevirapine in human immunodeficiency virus positive individuals with concomitant tuberculosis (HIV-TB patients).\nNevirapine is a much cheaper drug than its alternative efavirenz, and might be beneficial in resource constrained settings.\nMethods\nA randomised open label trial was conducted at All India Institute of Medical Sciences, New Delhi, India.\nDuring the regimen of an antiretroviral therapy (ART), naive HIV-TB patients were randomly assigned to receive either nevirapine or efavirenz based ART with concomitant rifampicin based anti-tubercular therapy (ATT).\nParticipants were followed for 24 months after starting ART.\nThe end points were virological, immunological and clinical responses, and progression of HIV disease marked by failure of ART.\nResults\nOf the 135 HIV-TB patients, who were receiving rifampicin based ATT, 68 were selected randomly to receive efavirenz based ART and 67 to receive nevirapine based ART.\nThe virological failure rates in the overall population, and the nevirapine and efavirenz groups were 14.1% (19/135); 14.9% (10/67) and 13.2% (9/68), respectively (p\u2009=\u20090.94).\nNo significant difference was found between the groups in the rate of clinical, immunological or virological failures.\nThe overall mortality was 17% with no significant difference between the two groups.\nExcept for the lead in period on day 14, the mean nevirapine concentration remained above 3 mg/L.\nNo association was found between plasma levels of nevirapine and incidence of unfavourable outcomes in this group.\nConclusions\nOutcome of ART in HIV-TB patients on rifampicin based ATT showed no significant difference, irrespective of whether efavirenz or nevirapine was used.\nTherefore, nevirapine based ART could be an alternative in the resource limited settings in patients with HIV and tuberculosis co-infection.\nTrial registration\nNCT No. 01805258.\nBackground\nAccording to the UNAIDS 2010 report on the global AIDS epidemic, there are 33.3 million people living with human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS) in the world.\nFurthermore, Tuberculosis (TB) is the most common opportunistic infection affecting around 40% of people with HIV/AIDS worldwide.\nAs per the recent report of National AIDS Control Organisation (NACO), the prevalence of HIV in India is 0.29% with a total burden of 2.27 million HIV-infected patients.\nHowever, there has been a decline in the incidence of new HIV/AIDS cases in India after the introduction of a free ART program by NACO since April, 2004.\nAnti-retroviral therapy (ART) based on non-nucleoside reverse transcriptase inhibitors (NNRTIs) is a widely used regimen, particularly in resource-limited countries.\nAccording to the current WHO HIV treatment guidelines and the NACO guidelines for India, efavirenz-based ART is a preferred first-line regimen in HIV/TB co-infected patients already receiving rifampicin-containing ATT regimen, because of the lower drug-drug interactions when compared with nevirapine or protease inhibitors (PIs).\nHowever, nevirapine is frequently used in India in HIV/AIDS patients as a component of first line regimens, which is also available as fixed drug combinations (with zidovudine plus lamivudine or stavudine plus lamivudine).\nThese drug combinations ensure good adherence as they are given as two tablets twice a day, are modestly priced, do not require food restrictions, and are safe during pregnancy.\nTo date, rifampicin, a potent cytochrome P450 enzyme inducer, which reduces the plasma concentrations of nevirapine, is still an important drug prescribed for the treatment of TB in many resource limited countries like India.\nThis drug is usually administered for 6 to 8 months as a part of ATT to reduce the relapse of TB in HIV-TB co-infected patients.\nAlthough, rifabutin has fewer problematic drug interactions, it is not available in many resource limited countries including India.\nThe concomitant use of ART and ATT is associated with a reduction in the mortality in patients co-infected with HIV and tuberculosis.\nNevertheless, treatment in this setting is complex because of high pill burden leading to poor adherence, drug-drug interactions and hepatotoxicity caused by both nevirapine and rifampicin.\nThere is also a rising concern that rifampicin, a potent inducer of cytochrome P450 enzyme, reduces the plasma concentrations of nevirapine leading to virological failure and occurrence of resistance mutations.\nHowever, recent studies have shown that even though nevirapine concentrations are lower when it is co-administered with rifampicin, the immunological and virological responses of nevirapine-containing ART have been found to be satisfactory.\nWhile multiple studies have shown that both nevirapine and efavirenz based regimens have equal efficacy in ART naive patients without TB, but there is not much information in the literature in setting of HIV-TB co-infection.\nThe present study is a comparative, randomised control trial, conducted prospectively to compare the safety and efficacy of nevirapine and efavirenz based ART in HIV-TB co-infected ART-naive patients, who were concomitantly receiving rifampicin based anti-tubercular regimen (ATT).\nIn this study, we have also measured the plasma nevirapine concentrations and correlated them with the immunological and virological responses to ART for a follow up period of more than two years.\nMethods\nThis was a randomised control study conducted at All India Institute of Medical Sciences (AIIMS), New Delhi, between September, 2007 to December, 2012.\nPatients, positive for HIV by ELISA, ART-na\u00efve and presented with concomitant TB, were enrolled as study participants.\nOnly patients having CD4 count <200 cells/mm3 and with normal renal and hepatic function were included.\nThe other inclusion criteria were age >18 years and absence of concomitant diabetes mellitus.\nPatients positive in hepatitis B and C serologies were excluded to avoid confusion between drug induced and viral hepatitis.\nBesides these, patients on anti-epileptic drugs, immunosuppressant and other drugs that induce liver microsomal enzyme systems were also excluded.\nAll female patients were screened with a urine pregnancy test and were excluded if pregnant.\nHIV infection was documented by licensed ELISA test kit (As per NACO guidelines).\nCD4/CD8 cell counts were determined by flow-cytometry (BD FACS CALIBUR).\nViral load testing was done using AMPLICOR HIV-1 Monitor Test, version 1\u2009\u00b7\u20095, manufactured by ROCHE Diagnostics and Abbott\u2019s RealTime HIV-1 Qualitative Assay performed on Abbott\u2019s automated high-throughput m2000 system.\nThe protocol was approved by the institutional research Ethics Committee of the All India Institute of Medical Sciences, New Delhi.\nAll participants gave signed informed consent to participate in this study.\nInitial evaluation\nAll patients underwent a detailed physical examination.\nTheir body mass index (BMI) was calculated.\nHaemoglobin, complete blood counts, erythrocyte sedimentation rate, fasting blood glucose, renal function tests, liver function tests, serum albumin, serum uric acid and routine urinalysis were done for all patients.\nIn addition, their CD4 counts and plasma HIV viral load were also determined.\nRandomisation and treatment\nIn this randomised open label trial, eligible ART naive HIV-TB patients were assigned to receive either nevirapine or efavirenz based ART.\nAll the ART naive patients attending the ART clinic at our centre were screened for tuberculosis by physical exam, sputum examination for AFB, chest radiographs and ultrasound abdomen as part of routine screening recommended by NACO and Revised National Tuberculosis Control Programme (RNTCP).\nART naive patients co-infected with tuberculosis were randomised into one of the trial arms using computer generated random number tables.\nATT was started for the patients according to the RNTCP guidelines for directly observed therapyshort-course (DOTS).\nAfter 2\u20138 weeks of ATT, ART was started, which consisted of zidovudine and lamivudine combined with either twice a day nevirapine or once a day efavirenz as per the respective randomisation.\nThose who had haemoglobin less than 8 g/dl were administered stavudine in place of zidovudine.\nThe administered doses were in accordance with the NACO guidelines.\nZidovudine was given in a dose of 300 mg twice a day, lamivudine 150 mg twice a day and stavudine 30 mg twice a day.\nNevirapine was administered at a dose of 200 mg once a day for the first 14 days (called the lead-in dose) as per NACO guideline, and then the dose was escalated to 200 mg twice a day.\nThe patients were advised to take the drug at 9 am for the first 14 days and at 9 am and 9 pm during the rest of the period of the follow up.\nEfavirenz was given in a dose of 600 mg per day.\nThe timing advised for efavirenz intake was daily at 9 pm after dinner.\nFollow up\nPatients were assessed at day 14 after the start of ART, then on day 28, and every 4 weeks thereafter through 96 weeks.\nA complete haemogram, and liver and kidney function tests were obtained on all of the visits.\nCD4 counts and HIV plasma viral loads were measured at baseline, 6, 12, 18, and 24 months after the start of ART.\nTrough nevirapine concentrations were assessed at day 14, 28, 42 and 180, 12 hours after the evening dose of nevirapine in all patients.\nThe method used for the measurement of nevirapine concentrations has been described earlier.\nDefinitions\nDisease progression or clinical failure was defined as a new or recurrent WHO stage-4 condition, after at least 6 months of ART.\nImmunological failure was defined as a decrease in CD4 count from the baseline values, for that either 50% decrease from the peak CD4 count during the treatment or persistent counts below 100 cells/mm3 after 24 weeks of the treatment was considered.\nVirological response was defined as HIV plasma viral loads <400 copies/ml after 24 weeks of ART.\nComposite unfavourable outcome was defined as when a patient failed to suppress the HIV plasma viral load to <400 copies/ml at the end of 24 weeks of the treatment, or failed to sustain a suppressed plasma viral load <400 copies/ml after 24 weeks, or had immunological failure at any time during the treatment as defined above, or had the disease progression as defined above, or expired during the treatment.\nCombined ART failure was defined as the development of clinical, immunological or virological failure at anytime during the treatment.\nTreatment success and failure of ATT were defined as per the WHO guidelines.\nOutcomes\nThe primary outcome of the study was the proportion of the subjects after 24 months who died or had a CD4 count below 200 cells/ml at 24th month.\nThe secondary outcome of the study was assessment of safety and tolerability of ART, measured by the proportion of the subjects with toxicities and the proportion of subjects changed/discontinued ART because of toxicities or treatment failure.\nThe overall outcome of ATT was assessed by both of the outcome.\nStatistical analysis\nData were recorded on a pre-designed data sheet and managed on an MS-Excel spreadsheet.\nAll entries were checked twice for any possible recording error.\nMean, frequency and median were calculated for all quantitative variables along with their respective standard deviations and Interquartile ranges.\nInitially, considering this a pilot study, sample size calculation was not planned and data were gathered as per the sample of convenience.\nAll analyses were done by Intention to treat analysis principle.\nAll continuous variables having normal distribution were analysed using Student\u2019s t-test.\nOrdinal variables and variables with non-normal distribution were analysed using Wilcoxon rank-sum test.\nThe categorical variables with dichotomous outcomes like ART failure and unfavourable outcomes were analysed using logistic regression model.\nGeneralised estimation equations were used to analyse the predictors of immunological response in terms of the increase in CD4 count.\nStatistical analyses were performed using software package STATA version 11\u2009\u00b7\u20090 [(intercooled version), Stata Corporation, Houston, Texas, USA].\nResults\nOf the total 135 HIV-TB patients enrolled, 67 were randomised into nevirapine (NVP) arm and 68 into efavirenz (EFV) arm of the study (Figure\u00a01).\nTheir baseline characteristics are summarised in Table\u00a01.\nThe two groups were not significantly different in any respect; however, the distribution of the type of TB bordered at a nearly significant level (p\u2009=\u20090\u2009\u00b7\u200906), owing to the greater number of disseminated/miliary TB cases in the nevirapine group.\nMajority of the patients (96\u2009\u00b7\u20093%, 130/135) were at WHO stage-4 of the HIV disease, and the rest were at stage-3.\nThere was a clear male predominance in both the groups in terms of gender distribution.\nMost of the patients were suffering from their first episode of TB, and therefore, were started on DOTS category I ATT.\nTable\u00a02 presents outcomes of ATT and ART of the participants.\nThe overall mortality was 17% (23/135) with no significant difference between the groups.\nThe mortality rates were 19\u2009\u00b7\u20094% (13/67) and 14\u2009\u00b7\u20097% (10/68) in the nevirapine and efavirenz groups, respectively (p\u2009=\u20090\u2009\u00b7\u200946).\nConsidering the type of TB, the difference in mortality between the groups was further reduced (p\u2009=\u20090\u2009\u00b7\u200950).\nOf the 23 patients who died, 16 (69\u2009\u00b7\u20096%) had a CD4 count <100 cells/mm3.\nAlso, most of the deaths (87%, 20/23), in both the groups, occurred at the early stage of the treatment, when the patients were receiving both ATT and ART.\nOutcome of TB treatments in both the groups was comparable (p\u2009=\u20090\u2009\u00b7\u200926).\nThe combined incidence of ART failure was 28\u2009\u00b7\u20094% (19/67) in the nevirapine arm and 30\u2009\u00b7\u20099% (21/68) in the efavirenz arm (p\u2009=\u20090\u2009\u00b7\u200975) with an overall ART failure of 29\u2009\u00b7\u20096% (40/135).\nRates of clinical, immunological and virological types of failures were compared separately, which reflected no significant difference between the groups for any category (Figures\u00a02, 3).\nComposite unfavourable outcome was defined as either death or any type of ART failure or both (Figure\u00a04).\nThe incidence of composite outcome in the entire study population was 43\u2009\u00b7\u20097% (59/135).\nNevirapine group had 44\u2009\u00b7\u20098% (30/67) composite unfavourable outcome, and efavirenz group had 42\u2009\u00b7\u20096% (29/68); p\u2009=\u20090\u2009\u00b7\u200998.\nNevirapine drug level was measured in the stored plasma collected from patients in the arm on day 14, 28, 42, and 180 after the initiation of ART.\nExcept for the lead in period on day 14, the mean nevirapine concentration remained above 3 \u03bcg/mL (Figure\u00a05).\nThe mean difference in the plasma nevirapine concentrations between those who had composite unfavourable outcome at 24th month as compared to those who had favourable outcome were 0.01 (P\u2009=\u20090.97) at day 14, 0.21(P\u2009=\u20090.11) at day 28, 0.34 (P\u2009=\u20090.13) at day 42 and 0.62 (P\u2009=\u20090.06) at day 180.\nNo association was found between plasma levels of nevirapine and incidence of unfavourable outcomes in this group at any time-point.\nThe rate of adverse events was similar in the two groups.\nSeven patients in the nevirapine arm and six in the efavirenz arm needed a change in the ART regimen after developing an adverse effect to one of the NRTI constituents, zidovudine or stavudine.\nNo regimen change was required with respect to the NNRTI component of ART.\nNo hospital admission or death was ascribed to any adverse event.\nGeneral health parameters like body mass index, blood haemoglobin level, and liver function tests showed similar favourable trends over time in both the groups (Table\u00a03).\nDiscussion\nThis open-label, randomised clinical trial demonstrated that, in HIV infected individuals with tuberculosis co-infection receiving rifampicin based ATT, there was no significant difference between twice daily nevirapine based ART and efavirenz based ART with respect to mortality, or immunological, virological and clinical response or side effect profile.\nThere are multiple observational and retrospective studies comparing the clinical efficacy of nevirapine and efavirenz based ART in HIV/TB co-infected patients; however, results of this study were found contradictory to those.\nThe South African study showed some difference between the two treatments, while the studies from Botswana and Thailand failed to demonstrate the difference.\nTo the best of our knowledge, there are only two randomised control trials that have made this head to head comparison.\nHowever, one of these studies used 400 mg once a day nevirapine dose instead of the standard regimen of 200 mg twice a day.\nThe N2R trial by Manosuthi et al. showed no significant difference in the virological outcomes in patients receiving efavirenz or nevirapine based regimens.\nOur results of comparable study of mortality, virological, clinical and immunological responses between the efavirenz and nevirapine groups are consistent with previous studies done in this patient sub-population.\nIn our study, there was a trend in increase in mortality in the nevirapine group as compared to that of the efavirenz group (19\u2009\u00b7\u20094% vs. 14\u2009\u00b7\u20097%); this difference, however, was not significant (P\u2009=\u20090\u2009\u00b7\u200946).\nWe also noted comparable rate of treatment failures in the two groups (30\u2009\u00b7\u20099% in EFV group vs. 28\u2009\u00b7\u20094% in NVP group, P\u2009=\u20090\u2009\u00b7\u200975).\nThe higher mortality observed in both the treatment groups as compared to other studies could be attributed to the long follow-up period of 2 years, and the fact that many of the patients enrolled in the trial had advanced tuberculosis (disseminated TB) and immunosuppression at the time of enrolment.\nThis is supported by the fact that the majority of the deaths (87%) occurred in the early stage of the treatment in both groups.\nConcerns have been raised about decreased plasma concentration and clinical efficacy of nevirapine in patients receiving rifampicin, due to induction of CYP450 enzyme.\nIn our study, the mean plasma concentration of nevirapine remained above 3 \u03bcg/ml except during the initial lead in period.\nAlso, no correlation was found between the plasma drug levels and unfavourable outcomes.\nThis is consistent with our previous study, where this lack of correlation between nevirapine blood levels and treatment outcomes in participants receiving rifampicin based ATT was demonstrated.\nOur study demonstrated a favourable response in terms of cure and treatment completion rates (74\u2009\u00b7\u20096% in NVP vs. 75% in EFV) of TB treatment in both the groups.\nDuring the 2 year follow-up, none of the patients who successfully completed their treatment relapsed.\nThis rate of successful treatment is considered favourable and consistent with the data from older studies showing cure rates between 59\u2009\u00b7\u20094% to 97\u2009\u00b7\u20091%.\nThough some studies have shown that thrice a week.\nATT might be sub-optimal in the first two months of the treatment in HIV-TB co-infections, we used the standard thrice a week DOTS regimens in accordance with the standard national guidelines available to us at the start of the trial.\nIt is possible that the differences in the ATT regimens accounted for the higher rate of failure in the nevirapine group in the study in South Africa by Boulle A. et al., which used once a day treatment.\nIt is still unclear at this time whether use of daily rifampicin will alter the efficacy of nevirapine in HIV-TB co-infected individuals.\nThis remains an interesting question for future research, since very limited data are available to us at this point of time.\nThe overall rates of adverse drug events were low in both the treatment groups.\nChange in ART regimen was needed in few cases for adverse effects of one of the NRTI constituents, zidovudine or stavudine.\nNo change in regimen was required with respect to the NNRTI component of ART.\nThis is in contrast to the studies by Manosuthi et al. and Van Leth et al. which demonstrated a higher rate of hepatotoxicity in patients receiving nevirapine requiring change of regimen.\nThis study however included patients with hepatitis B and hepatitis C co-infections.\nIt is possible that the exclusion of these patients in our study resulted in lower rates of hepatotoxicity.\nAlso, some cohort studies have reported low and comparable levels of hepatotoxicity in efavirenz and nevirapine when these drugs were administered in combination with rifampicin; however, the levels were even higher in patients without TB.\nThe most commonly observed adverse event was mild skin reactions that did not require any treatment.\nThe important factor in our study was that this was a randomised control study while most of other studies comparing the efficacy of nevirapine and efavirenz in HIV/TB patients were observational studies.\nIn addition, this study comprised a long follow-up period of 2 years which allowed us to study the outcome of TB in terms of failure and relapse.\nThe longer follow-up period also enabled us to measure directly, the effect of the two regimens on mortality rather than just measuring virological suppression.\nGeneralisability of the outcome measures is another strength, given the high disease severity in terms of degree of immunosuppression and severity of tuberculosis at baseline of our study population, which is typical for TB and HIV co-infection cohorts in resource constrained settings.\nBesides this, it also included correlation with nevirapine levels up to 6 months, and close monthly follow-up.\nThe study also had few limitations.\nGiven the number of patients who completed the study follow-up at the end of the trial, the power of the study was less than 50%, and therefore was underpowered to detect a difference of less than 20% between the two groups.\nMany of the participants in the trial received stavudine, which has been recently phased out from the ART program in India.\nAnother limitation was that the types of TB in the two groups differed significantly, and adjusting the types of TB did not alter the treatment outcomes significantly.\nConclusions\nIn conclusion, the efficacy and safety of nevirapine based ART seemed to be comparable to that of efavirenz containing regimens.\nOwing to its lower cost and easy availability, nevirapine based ART could be an alternative in the resource limited settings in patients with HIV and tuberculosis co-infection.\nScreening, enrolment and follow-up of study participants.\nCD4 cell count at different time points in nevirapine and efavirenz groups.NVP\u2009nevirapine, EFV\u2009efavirenz.\nViral load count at different time points in log scale in nevirapine and efavirenz group.NVP\u2009nevirapine, EFV\u2009efavirenz.\nKaplan Meier survival curve with cumulative probability of death or ART failure by 24 months.NVP\u2009nevirapine, EFV\u2009efavirenz.\nPlasma Nevirapine concentrations at different time points in nevirapine group.\n\nBaseline characteristics of the study participants\nVariable | Nevirapine group | Efavirenz group | P value\nn\u2009=\u200967 | n\u2009=\u200968\nAge, years: | \u00a0 | \u00a0 | \u00a0\nMean\u2009\u00b1\u2009SDa | 36.3 \u00b19.2 | 34.8 \u00b16.9 | 0.29\nGender, number (%): | \u00a0 | \u00a0 | \u00a0\nMale | 53 (79.1) | 59 (86.8) | 0.24\nFemale | 14 (20.9) | 09 (13.2)\nBMIb, kg/m2: | \u00a0 | \u00a0 | \u00a0\nMean\u2009\u00b1\u2009SD | 18.2 \u00b12.7 | 18.4 \u00b12.7 | 0.57\nHaemoglobin, g/dL: | \u00a0 | \u00a0 | \u00a0\nMean\u2009\u00b1\u2009SD | 10.0 \u00b11.9 | 10.4 \u00b11.8 | 0.27\nCD4 count, cells/mm3: | \u00a0 | \u00a0 | \u00a0\nMedian (Range) | 137 (20\u2013506) | 139 (7\u2013588) | 0.90\nlog10 viral load/ml: | \u00a0 | \u00a0 | \u00a0\nMedian (Range) | 5.52 (2.60\u20136.58) | 5.19 (2.76\u20136.76) | 0.43\nWHO Staging of HIV disease, number (%): | \u00a0 | \u00a0 | \u00a0\nStage-3 | 03 (4.5) | 02 (2.9) | 0.64\nStage-4 | 64 (95.5) | 66 (97.1) | \u00a0\nType of Tuberculosis, number (%): | \u00a0 | \u00a0 | \u00a0\nPTBc | 16 (23.9) | 19 (27.9) | 0.06\nEPTBd | 31 (46.3) | 40 (58.8)\nDisseminated/Miliary TB | 20 (29.8) | 09 (13.3)\nCategory of ATTe, number (%): | \u00a0 | \u00a0 | \u00a0\nCategory I | 52 (77.6) | 55 (80.9) | 0.64\nCategory II | 15 (22.4) | 13 (19.1) | \u00a0\nATT-ARTf gap, days: | \u00a0 | \u00a0 | \u00a0\nMedian (Range) | 27 (11\u201385) | 26 (04\u201393) | 0.92\n\naStandard Deviation, bBody Mass Index, cPulmonary Tuberculosis, dExtra Pulmonary Tuberculosis, eAntituberculosis treatment, fAntiretroviral therapy.\n\nOutcomes of antituberculosis and antiretroviral treatment\n\u00a0\nOutcomes of tuberculosis at completion of ATT\nOutcome | Nevirapine group | Efavirenz group | P value\nn\u2009=\u200967 | n\u2009=\u200968\nSuccessfully treated | 50 (74.6%) | 51 (75.0%) | 0.26\nLost to follow up | 06 (9.0%) | 06 (8.8%)\nDied | 11 (16.4%) | 09 (13.2%)\nOn ATT at end of study | 00 | 02 (2.9%)\nOutcomes of ART after 24 months\nOutcome | Nevirapine group | Efavirenz group | P value\nn\u2009=\u200967 | n\u2009=\u200968\nMortality: | \u00a0 | \u00a0 | \u00a0\nObserved | 13 (19.4%) | 10 (14.7%) | 0.46\n(% Adjusted for type of TB) | 19.1% | 14.9% | 0.50\nART failure: | 19 (28.4%) | 21 (30.9%) | 0.75\nClinical failure | 06 (9.0%) | 06 (8.8%) | 0.98\nImmunological failure | 06 (9.0%) | 08 (11.8%) | 0.58\nVirological failure | 10 (14.9%) | 09 (13.2%) | 0.94\nComposite unfavourable outcome | 30 (44.8%) | 29 (42.6%) | 0.98\n(Death and/or ART failure)\n\nATT\u2009Antituberculosis treatment, ART\u2009Antiretroviral therapy.\n\nBlood parameters and BMI at different time points\nVariable mean\u2009\u00b1\u2009S.D. | Nevirapine group (n\u2009=\u200967) | Efavirenz group (n\u2009=\u200968)\nBaseline | 6 months | 24 months | Baseline | 6 months | 24 months\nHaemoglobin, mg/dL | 10.0 \u00b1 1.9 | 11.8 \u00b1 2.3 | 12.5 \u00b1 2.3 | 10.4 \u00b1 1.8 | 12.5 \u00b1 1.3 | 12.7 \u00b1 2.0\nBMI, kg/m2 | 18.2 \u00b1 2.7 | 20.4 \u00b1 2.7 | 21.8 \u00b1 2.5 | 18.4 \u00b1 2.7 | 20.0 \u00b1 2.3 | 21.0 \u00b1 2.4\nBilirubin, mg/dL | 0.6 \u00b1 0.2 | 0.7 \u00b1 0.1 | 0.7 \u00b1 0.1 | 0.6 \u00b1 0.2 | 0.7 \u00b1 0.1 | 0.7 \u00b1 0.1\nSGOT, IU/L | 43.9 \u00b1 26.4 | 34.1 \u00b1 9.5 | 31.1 \u00b1 11.1 | 42.6 \u00b1 21.9 | 37.5 \u00b1 21.5 | 34.4 \u00b1 19.6\nSGPT, IU/L | 34.1 \u00b1 23.6 | 32.6 \u00b1 11.8 | 31.1 \u00b1 11.6 | 36.0 \u00b1 22.2 | 31.6 \u00b1 17.9 | 36.8 \u00b1 40.1\n\nBMI\u2009Body mass index, SGOT\u2009Serum glutamic oxaloacetic transaminase, SGPT\u2009Serum glutamic pyruvic transaminase, IU\u2009International unit.",
        "label": "unclear",
        "id": "task4_RLD_test_451"
    },
    {
        "paper_doi": "10.1016/s1473-3099(15)00530-7",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Trial design: randomizedTime period/duration of trial: 18 September 2011-14 July 2014Duration of follow-up: 6 months\n\n\nParticipants: Setting: presenting to the outpatient or emergency department of Patan Hospital or the Civil Hospital\nLocation: Lalitpur, Nepal\nAge: 2-45 years; median age 20 (IQR 14-23.5 years) in ceftriaxone group 19 (IQR 15-23 years) in gatifloxacin group\nGender: male = 180, female = 59\nHealth status of participants: not recorded\nInclusion criteria:body temperature at least 38.0 degC for >= 4 daysno focus of infection as assessed by physical examination and laboratory testsability to give written, informed consent (from parent or guardian if patient < 18 years)Exclusion criteria:pregnancydiabetes mellitussigns of severe typhoid (e.g., obtundation, shock, clinical jaundice, active gastrointestinal bleeding)history of hypersensitivity to either of the trial drugs, or had been given a fluoroquinolone, a third-generation cephalosporin, or macrolide within the previous week.Patients who had received chloramphenicol, amoxycillin, or co-trimoxazole within the previous week if the treating clinician reported a clinical response\n\n\nInterventions: Gatifloxacin 10 mg/kg/day, oral, 7 daysCeftriaxone 60 mg/kg/day to 2 g/day if 2-13 years old, IV, 7 days OR ceftriaxone 2 g/day if >= 14 years, IV, 7 days\n\n\nOutcomes: Primary endpoint was a composite of treatment failure as >= 1 of:prolonged fever clearance time > 7 days after treatment initiation (fever clearance time defined as the time from the first dose of the trial drug until the temperature fell to <= 37.5 degC and remained there for at least 2 days)need for rescue treatment as judged by treating physicianpositive blood culture on day 8 of treatment (microbiological failure)culture-confirmed or syndromic enteric fever relapse within 28 days of treatment initiationdevelopment of enteric fever-related complications (e.g. clinically significant bleed, fall in GCS, GI perforation, admission to hospital within 28 days of treatment initiation)Secondary endpoints were:fever clearance timetime to relapse until day 28 or at any time during follow-upconfirmed and syndromic relapse at day 28faecal carriage of S typhi or S paratyphi at 1 month, 3 months, or 6 months after randomization assessed in culture-positive participants only\n\n\nOrganism type and antimicrobial susceptibility: S typhi = 81 (43 in gatifloxacin group and 38 in ceftriaxone group); S paratyphi = 35 (19 in gatifloxacin group and 16 in ceftriaxone group). 14 of the 81 S typhi isolates were resistant to gatifloxacin (MIC > 1 ug/mL). None of the isolates were resistant to ceftriaxone.\n\n\nNotes: 725 patients were screened for the trial, with 246 enrolled and randomized. 479 were ineligible, because of being in receipt of antibiotics in the week before trial entry (n = 178), refusal of consent (n = 163), outside the age criteria (n = 53), could not arrange to be followed up (n = 64), pregnant or lactating (n = 6), or other reasons (n = 15). 7 patients were excluded after randomization because of refusing IV medication (n = 3), an alternative diagnosis was confirmed (n = 3), and dropped out before a single dose (n = 1) leaving 239 participants analysed by ITT. 120 participants assigned to receive gatifloxacin, and 62 of these were culture-positive; 119 were assigned to receive ceftriaxone and 54 of these were culture-positive. The trial was designed to randomly assign 300 patients to treatment. The trial was halted prematurely on clinical grounds following the recommendation of the Data and Safety Monitoring Board following the appearance of gatifloxacin-resistant strains and resulting treatment failures\n\n",
        "objective": "To evaluate the effectiveness of cephalosporins for treating enteric fever in children and adults compared to other antimicrobials.",
        "full_paper": "Summary\nBackground\nBecause treatment with third-generation cephalosporins is associated with slow clinical improvement and high relapse burden for enteric fever, whereas the fluoroquinolone gatifloxacin is associated with rapid fever clearance and low relapse burden, we postulated that gatifloxacin would be superior to the cephalosporin ceftriaxone in treating enteric fever.\nMethods\nWe did an open-label, randomised, controlled, superiority trial at two hospitals in the Kathmandu valley, Nepal.\nEligible participants were children (aged 2\u201313 years) and adult (aged 14\u201345 years) with criteria for suspected enteric fever (body temperature \u226538\u00b70\u00b0C for \u22654 days without a focus of infection).\nWe randomly assigned eligible patients (1:1) without stratification to 7 days of either oral gatifloxacin (10 mg/kg per day) or intravenous ceftriaxone (60 mg/kg up to 2 g per day for patients aged 2\u201313 years, or 2 g per day for patients aged \u226514 years).\nThe randomisation list was computer-generated using blocks of four and six.\nThe primary outcome was a composite of treatment failure, defined as the occurrence of at least one of the following: fever clearance time of more than 7 days after treatment initiation; the need for rescue treatment on day 8; microbiological failure (ie, blood cultures positive for Salmonella enterica serotype Typhi, or Paratyphi A, B, or C) on day 8; or relapse or disease-related complications within 28 days of treatment initiation.\nWe did the analyses in the modified intention-to-treat population, and subpopulations with either confirmed blood-culture positivity, or blood-culture negativity.\nThe trial was powered to detect an increase of 20% in the risk of failure.\nThis trial was registered at ClinicalTrials.gov, number NCT01421693, and is now closed.\nFindings\nBetween Sept 18, 2011, and July 14, 2014, we screened 725 patients for eligibility.\nOn July 14, 2014, the trial was stopped early by the data safety and monitoring board because S Typhi strains with high-level resistance to ciprofloxacin and gatifloxacin had emerged.\nAt this point, 239 were in the modified intention-to-treat population (120 assigned to gatifloxacin, 119 to ceftriaxone).\n18 (15%) patients who received gatifloxacin had treatment failure, compared with 19 (16%) who received ceftriaxone (hazard ratio [HR] 1\u00b704 [95% CI 0\u00b755\u20131\u00b798]; p=0\u00b791).\nIn the culture-confirmed population, 16 (26%) of 62 patients who received gatifloxacin failed treatment, compared with four (7%) of 54 who received ceftriaxone (HR 0\u00b724 [95% CI 0\u00b708\u20130\u00b773]; p=0\u00b701).\nTreatment failure was associated with the emergence of S Typhi exhibiting resistance against fluoroquinolones, requiring the trial to be stopped.\nBy contrast, in patients with a negative blood culture, only two (3%) of 58 who received gatifloxacin failed treatment versus 15 (23%) of 65 who received ceftriaxone (HR 7\u00b750 [95% CI 1\u00b771\u201332\u00b780]; p=0\u00b701).\nA similar number of non-serious adverse events occurred in each treatment group, and no serious events were reported.\nInterpretation\nOur results suggest that fluoroquinolones should no longer be used for treatment of enteric fever in Nepal.\nAdditionally, under our study conditions, ceftriaxone was suboptimum in a high proportion of patients with culture-negative enteric fever.\nSince antimicrobials, specifically fluoroquinolones, are one of the only routinely used control measures for enteric fever, the assessment of novel diagnostics, new treatment options, and use of existing vaccines and development of next-generation vaccines are now a high priority.\nFunding\nWellcome Trust and Li Ka Shing Foundation.\nIntroduction\nEnteric (typhoid) fever, a systemic infection caused by the Salmonella enterica serovars Typhi and Paratyphi A, B, and C, is a leading cause of febrile disease in many low-income countries.\n27 million new infections and more than 200\u2008000 deaths are estimated to be attributable to enteric fever worldwide each year.\nIn Kathmandu, the capital of Nepal and the setting of this study, the burden of enteric fever is particularly high, and is the leading cause of febrile bacterial disease in adults and children.\nResearch in context\nEvidence before this study\nWe searched MEDLINE, PubMed, and Scopus without date restrictions for English-language articles with the search terms \u201crandomized controlled trial\u201d (\u201cRCT\u201d or \u201crandomized* control* trial*\u201d) AND \u201ctyphoid fever\u201d, \u201centeric fever\u201d, AND \u201cceftriaxone\u201d.\nWe also noted relevant articles outlined in a Cochrane review, and a meta-analysis, of fluoroquinolones versus other antimicrobials in the treatment of enteric fever.\nWe identified 11 randomised trials that used intravenous ceftriaxone in one of their treatment groups.\nThe selected trials had small sample sizes (ranging from 15 to 43 patients) and the definitions of outcomes for the primary and secondary outcomes were not standardised.\nThree trials gave ceftriaxone for 7 days in one of their groups (for 73 patients), the same duration as in our trial, but the drug dose was variable.\nThe mean fever clearance times ranged from 3\u00b79 days to 5\u00b74 days, the number of clinical failures ranged from none to six, and the number of relapses ranged from one to four in these small trials.\nAdded value of this study\nOur data augment previous findings, predicting that ceftriaxone is safe and effective for the treatment of enteric fever and out-performs gatifloxacin, with only 7% of culture-positive patients failing treatment, and a median fever clearance time of 2\u00b778 days.\nHowever, our study, by contrast with the outlined studies, also investigated the clinical outcome in culture-negative patients with suspected enteric fever\u2014in this group, gatifloxacin out-performed ceftriaxone with median fever clearance times of 1\u00b712 days and 3\u00b703, respectively.\nFurthermore, our work is the first to describe the clinical implications of fluoroquinolone-resistant Salmonella enterica serovar Typhi.\nImplications of all the available evidence\nIn view of the emergence of fluoroquinolone-resistant S Typhi in this setting and the poor efficacy of ceftriaxone in the culture-negative group, we advocate better diagnostic testing for febrile diseases in low-income countries, and suggest that fluoroquinolones are no longer effective for treatment of enteric fever in Nepal.\nResistance and reduced susceptibility to antimicrobials are the major challenges to successful treatment of enteric fever.\nWe have previously reported a high prevalence of S Typhi and S Paratyphi A strains in Nepal that show resistance against the quinolone nalidixic acid (minimum inhibitory concentration [MIC] \u2265256 \u03bcg/mL) with a corresponding decreased susceptibility against fluoroquinolones such as ciprofloxacin (MIC \u22650\u00b7125 \u03bcg/mL).\nCeftriaxone, a parenteral, third-generation cephalosporin, is a common empirical therapy for febrile disease in endemic enteric fever locations, and is used for the treatment of enteric fever in south Asia and other regions where nalidixic acid-resistant strains predominate.\nFurthermore, ceftriaxone is also advocated for the treatment of travellers returning with enteric fever from areas of enteric fever endemicity.\nInvestigators for three randomised controlled trials have compared fluoroquinolones with ceftriaxone for treatment of enteric fever.\nTheir findings generally favoured the fluoroquinolones, but the studies were insufficiently powered (only 15 to 25 patients per treatment group) to reach significance and data for the prevalence of nalidixic acid-resistant strains were not reported.\nIntravenous therapy is expensive and difficult to give reliably (particularly to outpatients) in most countries where enteric fever is endemic; therefore, effective oral antimicrobials are more practical for treatment of this disease.\nPreviously, we have shown that even without reported resistance, the oral third-generation cephalosporin, cefixime, did poorly in Nepalese patients with enteric fever\u2014treatment failure was reported in 26 (37%) of 70 patients receiving cefixime versus three (3%) of 88 patients receiving gatifloxacin.\nConversely, we have also shown in Nepalese and Vietnamese children and adults with uncomplicated enteric fever that the fourth-generation 8-methoxy-fluoroquinolone gatifloxacin is safe and effective despite an increase in prevalence of S Typhi strains with reduced ciprofloxacin susceptibility.\nTherefore, because third-generation cephalosporins are generally associated with slow clinical improvement and high relapse burden, and 7 days of oral gatifloxacin is associated with rapid fever clearance (\u22644 days) and low relapse burden, we postulated that gatifloxacin is superior to ceftriaxone in treating enteric fever, and did a study to test this hypothesis.\nMethods\nStudy design and participants\nWe did an open-label, randomised, controlled, superiority trial at Patan Hospital and the Civil Services Hospital in the Kathmandu valley, Nepal.\nThe study protocol was reviewed and approved by the Ethics Committee of the Nepal Health Research Council and the Oxford Tropical Research Ethics Committee (UK).\nWe screened children aged 2\u201313 years and adults aged 14\u201345 years with suspected enteric fever.\nThe criteria for suspected enteric fever were body temperature at least 38\u00b70\u00b0C for 4 days or more without a focus of infection, as assessed by physical examination and laboratory tests, and as previously described.\nPatients were excluded if they were pregnant; had diabetes mellitus, signs of severe infection (eg, obtundation, shock, clinical jaundice, or active gastrointestinal bleeding), or a history of hypersensitivity to either of the trial drugs; or had been given a fluoroquinolone, a third-generation cephalosporin, or a macrolide within the previous week.\nPatients who had received chloramphenicol, amoxicillin, or co-trimoxazole could be included, provided the treating clinician reported no clinical response.\nWritten, informed consent to participate in the study was required from all patients.\nFor patients younger than 18 years, we obtained written, informed consent from their parent or an adult guardian.\nRandomisation and masking\nWe randomly assigned patients (1:1) without stratification to 7 days of treatment with either oral gatifloxacin (10 mg/kg) once per day or intravenous ceftriaxone (60 mg/kg up to a maximum of 2 g for patients aged 2\u201313 years or 2 g for patients aged \u226514 years) once per day.\nThe randomisation list was computer-generated with blocks of four and six (with equal probability) and maintained by a clinical trials pharmacist.\nWe concealed treatment allocations inside opaque sealed envelopes, which were numbered sequentially to correspond to patient enrolment numbers.\nEnvelopes were kept in a locked drawer and were opened in strictly numerical order by a study clinician (who had previously screened the patients and obtained consent).\nTreatment allocation was open-label; masking was not possible because of a difference in the administration route of the two drugs.\nProcedures\nGatifloxacin 400 mg tablets (Square Pharmaceuticals, Bangladesh) were weighed and cut at a dose of 10 mg/kg once per day.\nCeftriaxone (Powercef, 1000 mg injection vial, Wockhardt Ltd, India), was injected slowly over 10 min once per day.\nPatients received the first dose (on day 1) of the study drug in hospital to monitor for anaphylaxis.\nPatients receiving ceftriaxone were discharged with an intravenous cannula in situ and had a new cannula inserted on day 4 of treatment.\nHome treatment was monitored by trained community medical auxiliaries (CMAs), as described in previous studies.\nA CMA visited each patient assigned to treatment twice per day for at least 10 days or until the patient was asymptomatic.\nThe CMAs gave the drugs, and recorded drug doses, administration times, oral temperatures, symptoms, and potential adverse effects in a standard case-record form.\nWe measured complete blood count, serum creatinine, liver-function parameters (total bilirubin, aspartate aminotransferase, and alanine aminotransferase), and serum glucose at enrolment and on day 8 of treatment.\nWe did a finger-prick test for glucose each day on days 2\u20137 after randomisation, and measured random serum glucose on day 8, day 15, and at 1 month.\nWe took blood from all patients (3 mL from those aged <14 years; 8 mL from those aged \u226514 years) for bacterial culture at enrolment and on day 8 after randomisation if S Typhi or S Paratyphi were isolated at enrolment, or if their symptoms suggested a clinical relapse.\nWe inoculated blood samples from adults into media-containing tryptone soya broth and sodium polyanethol sulphonate, up to a total volume of 50 mL.\nWe used BactecPeds Plus culture bottles (Becton Dickinson, New Jersey, USA) for paediatric blood samples.\nCulture results were reported for up to 7 days; positive bottles were subcultured onto blood, chocolate, and MacConkey agar, and colonies presumptive of salmonella were identified using standard biochemical tests and serotype-specific antisera (Murex Biotech, Dartford, UK).\nWe measured antimicrobial sensitivities by the modified Kirby-Bauer disc diffusion method with zone size interpretation based on guidelines from the Clinical and Laboratory Standards Institute.\nAntimicrobial discs tested were ceftriaxone (30 \u03bcg), ciprofloxacin (5 \u03bcg), gatifloxacin (5 \u03bcg), and nalidixic acid (30 \u03bcg).\nMICs against these antimicrobials were measured by Etest, according to the manufacturer's recommendations (BioM\u00e9rieux, France).\nAll patients were requested to attend a clinic at Patan Hospital on day 8, day 15, and 1 month, 3 months, and 6 months after randomisation for clinical assessments and stool culture.\nOutcomes\nThe primary endpoint of this trial was a composite of treatment failure, defined as the occurrence of at least one of the following events: fever clearance time (ie, time from the first dose of a study drug until the temperature fell to \u226437\u00b75\u00b0C and remained there for at least 2 days) more than 7 days after treatment initiation; the need for rescue treatment as judged by the treating physician (the recommended rescue treatment was azithromycin; however, any treatment other than the assigned treatment was acceptable); blood-culture positivity for S Typhi or S Paratyphi on day 8 of treatment (microbiological failure); culture-confirmed or syndromic enteric fever relapse within 28 days of treatment initiation; or development of any enteric fever-related complication (eg, clinically significant bleeding, a fall in the patient's Glasgow Coma Score, perforation of the gastrointestinal tract, or admission to hospital) within 28 days after treatment initiation.\nTime to treatment failure was defined as the time from the first dose of treatment until the date of the earliest failure event; patients without a failure event were censored on day 28 or the date of their last follow-up.\nSecondary endpoints were fever clearance time only; time to relapse until day 28 or at any time during follow-up; confirmed and syndromic relapse until day 28; confirmed or syndromic relapse at any time; and faecal carriage of S Typhi or S Paratyphi at 1 month, 3 months, or 6 months after randomisation assessed in culture-positive patients only.\nWe calculated fever clearance times electronically using temperatures recorded twice per day.\nWe treated these times as interval-censored outcomes to show that fever clearance was known to have occurred at some unknown point in the interval from the last febrile temperature assessment until the first afebrile assessment.\nWe treated patients without fever clearance or relapse as right-censored.\nSafety and adverse events were assessed each day by the CMAs at the patient's home, by giving a symptom questionnaire and simple physical examinaton.\nAny patient who had unexpected symptoms was assessed by a study clinician in the hospital.\nEach patient was also seen by the study clinician in the hospital on the scheduled follow-up days and asked about any symptoms, and a physical examination was undertaken to assess for possible adverse events.\nStatistical analysis\nIn this study, we aimed to address the hypothesis that gatifloxacin was superior to ceftriaxone.\nOn the basis of our previous data, we predicted that about 7% of patients with a positive culture given gatifloxacin would have treatment failure.\nTo detect an increase in the risk of failure by 20% (from 7% to 27%) in the ceftriaxone group with 80% power at the two-sided 5% significance level, and allowing for a 10% loss to follow-up, we calculated that a sample size of 120 culture-positive patients (60 per treatment group) was needed.\nWe assumed a culture-positive rate of at least 40%, and designed the trial to randomly assign 300 patients to treatment.\nFor treatment failure, we based the comparison of the absolute risk of treatment failure until day 28 on Kaplan-Meier estimates and corresponding standard errors according to Greenwood's formula.\nWe used survival methods for the analysis of the time to treatment failure, fever clearance time, and time to relapse.\nFor the times to treatment failure and relapse, we used the Kaplan-Meier method to calculate the cumulative incidence of events and Cox regression models with treatment group as the only covariate used for comparison between treatment groups.\nFor the interval-censored fever clearance times, we used the non-parametric maximum likelihood estimator (NPMLE) to estimate their distribution, and parametric Weibull accelerated failure time models for the estimation of quantiles of the fever clearance time in each group and for the comparison between groups.\nWe based median (IQR) calculations of fever clearance times on models for each treatment group separately, and acceleration factors on models that included treatment as the only covariate.\nWe undertook all analyses for each of the three main analysis populations: a modified intention-to-treat (ITT) population (consisting of all randomised patients who received at least one dose of study treatment and did not have a confirmed alternative diagnosis), and the subpopulations with either confirmed blood-culture positivity, or blood-culture negativity.\nTreatment failure and fever clearance time were also assessed in predefined subgroups (age <16 years; age \u226516 years; female; male; recruited before or after April, 2013; MIC against ciprofloxacin <0\u00b712 \u03bcg/mL, 0\u00b712\u20132\u00b700 \u03bcg/mL, or >2\u00b700 \u03bcg/mL; MIC against gatifloxacin \u22641\u00b700 \u03bcg/mL or >1\u00b700 \u03bcg/mL; S Typhi infection; and S Paratyphi infection).\nWe tested for heterogeneity of treatment effects between subgroups with a Cox regression model (for analysis of treatment failure) or a Weibull accelerated failure time model (for fever clearance times) that included an interaction term between the treatment and the subgrouping variable.\nWe did all analyses using the statistical software R version 3.0.1, based on available data without imputation of missing data.\nThe safety of the trial was overseen by an independent data and safety monitoring board.\nThis trial was registered at ClinicalTrials.gov, number NCT01421693.\nRole of the funding source\nThe funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.\nThe corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.\nResults\nBetween Sept 18, 2011, and July 14, 2014, we screened 725 patients with suspected enteric fever for enrolment (figure 1).\nThe data and safety monitoring board reviewed the outcome data after 100 patients, and then 200 patients, were randomised.\nAt the 200-patient review, the board requested an additional review within 3 months of MICs against ciprofloxacin and gatifloxacin against all bacterial isolates.\nData from 109 culture-confirmed patients (and 233 patients in total) were analysed at this additional review.\nA comparison of treatment failures between treatment groups did not cross the predefined Haybittle-Peto stopping boundary of p less than 0\u00b7001 (overall or in any subgroup), but the emergence of S Typhi strains with MICs against ciprofloxacin that were greater than 16 \u03bcg/mL and against gatifloxacin that were greater than 1 \u03bcg/mL, and a significant difference (p=0\u00b70002 for susceptive vs resistant strains in the gatifloxacin group) in treatment response between patients with fluoroquinolone-resistant strains and those with susceptible strains, led the board to recommend stopping the trial on clinical grounds supported by data on the changing in-vitro susceptibility.\nThe trial was stopped on July 14, 2014.\nAt this point, we had recruited and randomly assigned 246 eligible patients to treatment (including 116 patients with microbiologically confirmed disease; figure 1).\nSeven patients were excluded from the ITT population\u2014four withdrew consent after randomisation but before receiving the first dose of study drug, and three had an alternative diagnosis.\nOn stopping the trial, 120 patients had received gatifloxacin and 119 patients had received ceftriaxone, totalling 239 analysed in the modified ITT population.\nS Typhi or S Paratyphi\nA was isolated from the blood of 62 patients in the gatifloxacin group and 54 patients in the ceftriaxone group (figure 1).\nAnalyses were not adjusted for early stopping of the trial.\nThe baseline characteristics of patients were balanced between the two treatment groups in the modified ITT population (table 1) except for a larger proportion of men in the gatifloxacin group.\nSimilar numbers of patients in each group had received a non-exclusion antimicrobial in the 2 weeks before randomisation.\nHowever, culture-negative patients were more likely to have had enteric fever previously and to report coughing, and had lower serum transaminase concentrations, than patients with blood culture-confirmed S Typhi or S Paratyphi A (appendix).\nMoreover, patients with S Typhi were more likely to report anorexia, nausea, and diarrhoea, and had a lower haematocrit compared with the other two patient groups.\nThe MICs against ciprofloxacin and the study drugs were also balanced between the treatment groups (table 2).\nThe first patient with a ciprofloxacin-resistant S Typhi culture (MIC >32 \u03bcg/mL) was enrolled on April 30, 2013.\nFrom that date, 118 additional patients were recruited, 55 of whom had positive blood cultures.\nAmong these, 14 (25%) of 55 patients with S Typhi infections with high MICs against ciprofloxacin (12 >32 \u03bcg/mL, two 24 \u03bcg/mL) were assigned to a study drug; all 14 strains were also highly resistant to gatifloxacin (MIC \u22651\u00b75 \u03bcg/mL).\nTreatment failure in the modified ITT population was similar between treatment groups: 18 (15%) of 120 patients who received gatifloxacin had treatment failure, compared with 19 (16%) of 119 who received ceftriaxone (hazard ratio [HR] of time to failure 1\u00b704 [95% CI 0\u00b755\u20131\u00b798]; p=0\u00b791 [table 3]).\nDetails for each event in the composite endpoint are in the appendix.\nHowever, there was significant heterogeneity in the primary outcome between the subpopulations of blood culture-confirmed and culture-negative patients (pinteraction<0\u00b70001; table 3, figure 2).\nIn the culture-confirmed population, 16 (26%) of 62 patients given gatifloxacin had treatment failure, compared with four (7%) of 54 patients given ceftriaxone (HR 0\u00b724 [95% CI 0\u00b708\u20130\u00b773, p=0\u00b701; table 3, absolute risks of failure in appendix).\nFor the four patients with treatment failure in the ceftriaxone group, MICs against ceftriaxone ranged from 0\u20130\u00b72 \u03bcg/mL, and were similar to MICs in patients without treatment failure.\nTreatment failure was associated with the emergence of S Typhi exhibiting resistance against fluoroquinolones.\nNone of the subgroup analyses for culture-positive patients showed significant treatment effect heterogeneity of the primary endpoint (table 3).\nBy contrast, in culture-negative patients, only two (3%) of 58 who received gatifloxacin failed treatment compared with 15 (23%) of 65 who received ceftriaxone (HR 7\u00b750 [95% CI 1\u00b771\u201332\u00b780]; p=0\u00b701 [table 3, absolute risks of failure in appendix]).\nThe most common cause of treatment failure in culture-negative patients treated with ceftriaxone was fever for more than 7 days (12 [80%] of 15 patients) and nine [60%] of 15 received rescue treatment (appendix).\nIn the modified ITT population, fever clearance times did not differ between the two treatment groups (table 4, figure 2).\nFurthermore, the incidence of microbiological failure or syndromic relapse at any time until 6 months did not differ between the two treatment groups by day 28 or by 6 months (appendix).\nWe noted significant heterogeneity (pinteraction<0\u00b70001) of the treatment effect for fever clearance times in the blood culture-positive and blood culture-negative subgroups (table 4, figure 2).\nIn culture-positive patients, median fever clearance times were longer in those treated with gatifloxacin than ceftriaxone (p=0\u00b7001) and outcomes with gatifloxacin were worse for patients with a raised MIC against ciprofloxacin and gatifloxacin (table 4).\nOccurrence of relapse did not differ between treatment groups in culture-positive patients (appendix).\nAt 1-month follow-up, only two patients had positive stool cultures (one for S Typhi and one for S Paratyphi A), both in the gatifloxacin group.\nWe noted no positive stool cultures at 3 months or 6 months in culture-positive patients (appendix).\nIn culture-negative patients, fever clearance times were shorter in those treated with gatifloxacin than ceftriaxone (p<0\u00b70001; table 4), but occurrences of relapse did not differ between treatment groups (appendix).\nOver the course of the trial, 122 adverse events occurred in the gatifloxacin group and 120 in the ceftriaxone group (appendix); no serious adverse events were reported.\nThe most common adverse events reported were vomiting (in 27 [23%] of 120 patients receiving gatifloxacin and 17 [14%] of 119 receiving ceftriaxone; p=0\u00b713), and cough (which did significantly differ between groups: in 15 [12%] patients receiving gatifloxacin and 29 [24%] patients receiving ceftriaxone; p=0\u00b702).\nNo adverse events were attributed to any of the study treatments.\nThe frequency of dysglycaemia and abnormal liver-function tests did not differ between the treatment groups (appendix), and none of the study participants died.\nDiscussion\nIn patients with clinically suspected enteric fever, we showed that outcomes did not differ between patients receiving gatifloxacin and those receiving ceftriaxone.\nHowever, patients with blood culture-confirmed enteric fever fared less well when given gatifloxacin, as suggested by an increased likelihood of treatment failure and protracted fever clearance times.\nThis finding was apparent only in the last year of recruitment into the trial as S Typhi strains with high-level resistance to ciprofloxacin and gatifloxacin (MICs >16 \u03bcg/mL and >1 \u03bcg/mL, respectively) emerged during the study, leading to the trial being stopped in July, 2014.\nThe data resulting from this trial contradict our hypothesis, because before the emergence of fluoroquinolone-resistant S Typhi in Nepal, we had shown in a series of clinical trials (done between 2004 and 2011) that gatifloxacin was both a safe and effective treatment for uncomplicated enteric fever.\nThis antimicrobial has provided good clinical outcomes despite the continuing isolation of S Typhi and S Paratyphi A organisms showing reduced susceptibility against ciprofloxacin (MICs from 0\u00b71 \u03bcg/mL\u20131 \u03bcg/mL).\nAlthough the treatment failure results with gatifloxacin in our study are based on data from a few patients, these findings are highly consistent with the results of larger numbers for the secondary endpoint of fever clearance, which lend support to the credibility of our results.\nGenerally, we still regard gatifloxacin as a safe drug for use in this setting since we have no evidence for an increased risk of dysglycaemia or hepatitis.\nHowever, our new data suggest that the efficacy of gatifloxacin (and older-generation fluoroquinolones) for the treatment of enteric fever in Nepal is now compromised by the emergence of high-level fluoroquinolone-resistant S Typhi.\nAs a result, we no longer advocate gatifloxacin as a treatment for enteric fever in Nepal.\nDespite being a two-centre study in Kathmandu, our findings raise substantial questions regarding the use of fluoroquinolones in south Asia and other endemic regions for treating enteric fever.\nGenomic data has shown that strains with reduced susceptibility to fluoroquinolones are now globally dominant.\nThe process of resistance development against the fluoroquinolones is a stepwise process; mutations are sequentially acquired in the target genes, thus determining higher MICs.\nData obtained from in-vitro experiments suggest that strains with fluoroquinolone-resistance-associated mutations might actually have a selective advantage over wild-type strains.\nFluoroquinolone resistance is clearly increasing, not only in Nepal, but also in neighbouring areas and other parts of the world where resources are low.\nTherefore in view of this combination of evidence, we predict a short window before the international emergence of S Typhi, and potentially S Paratyphi A, strains with high-level fluoroquinolone resistance, thus rendering this important group of antimicrobials globally ineffective for enteric fever.\nConversely, the outcomes for ceftriaxone-treated, culture-positive patients were good, with short fever clearance times and few relapses.\nThe optimum duration of ceftriaxone treatment is not clearly defined; in WHO guidelines, it is recommended for 10\u201314 days, but our data lend support to a 7-day treatment course for patients with uncomplicated enteric fever in an endemic setting.\nAlongside antimicrobial-resistant S Typhi, our clinical findings pose an additional clinical and public health challenge regarding ceftriaxone treatment.\nIn patients with a negative blood-culture result, the absolute risk of treatment failure was 0\u00b724 in the ceftriaxone group versus 0\u00b704 in the gatifloxacin group.\nFurthermore, the median fever clearance times in this group were 3\u00b703 days in the ceftriaxone group, versus 1\u00b712 days in the gatifloxacin group.\nThese contrasting outcomes for ceftriaxone in the two predefined patient populations were not predicted and previous similar data from other enteric fever trials are scarce.\nThe reason for this shortage of data is because the results of patients enrolled in enteric fever trials who did not have a positive blood culture were, until recently, not reported.\nFrom a small ceftriaxone study done in Vietnamese patients with enteric fever, the investigators reported that two of the six culture-negative patients given ceftriaxone failed treatment, but responded to rescue treatment with ofloxacin.\nOnly four randomised trials for enteric fever did an intention-to-treat analysis and incorporated an analysis for the blood culture-negative patients.\nIn these trials, culture-negative patients given gatifloxacin, ofloxacin, or azithromycin achieved similar (or better) outcomes than those reported in blood culture-confirmed patients with enteric fever.\nBetter clinical outcomes in patients with syndromic enteric fever but with a negative blood-culture result might be attributed to the low sensitivity of blood-culture tests (estimated to be 50\u201360%) and the possibility of fewer bacteria in the bloodstream, corresponding with less severe disease.\nWe do not know how many culture-negative patients actually had enteric fever in our study; some might have had alternative bacterial infections.\nOur previous study examined archived blood samples from 765 adults presenting at Patan Hospital, Nepal, with undifferentiated febrile illness in 2001.\n50 (7%) patients had Rickettsia typhi (murine typhus) DNA detected by PCR amplification.\nFurthermore, we investigated the infectious cause of culture-negative patients enrolled into one of our other enteric fever trials in Nepal and noted serological evidence of murine typhus in 21 (22%) of 96 blood culture-negative patients, with 12 (57%) of 21 testing positive for R typhi with PCR amplification.\nThus, we surmise that a substantial proportion of culture-negative patients with suspected enteric fever in Nepal might actually have other bacterial infections, included those caused by the rickettsiaceae.\nCeftriaxone is not regarded as an effective treatment for murine typhus and other rickettsial illnesses, whereas fluoroquinolones do seem to have clinical activity against these pathogens.\nHowever, no rapid diagnostic tests are available that can accurately differentiate between rickettsial infections, enteric fever, or indeed any bacteraemia, and inform patient management in a timely manner.\nWithout such a test, we suggest that a more pragmatic approach would be to combine ceftriaxone with doxycycline for patients without a positive blood culture who do not respond to ceftriaxone monotherapy.\nIn conclusion, the results of our trial underline two substantial problems for patients with enteric fever.\nFirst, the continued development of antimicrobial resistance in the pathogens causing the disease; second, the absence of point-of-care diagnostic tests for febrile diseases in low-income settings.\nWe have shown that high-level fluoroquinolone-resistant S Typhi is now likely to be endemic in Nepal and suggest that fluoroquinolones\u2014even the fourth-generation fluoroquinolone gatifloxacin\u2014cannot be recommended as empirical therapy for enteric fever in this setting.\nAzithromycin and ceftriaxone are alternative options, although sporadic cases of resistance have been reported, and data comparing in-vitro azithromycin susceptibility against clinical outcomes are poor.\nAdditionally, under our study conditions, ceftriaxone was suboptimum in a large proportion of culture-negative patients with suspected enteric fever, which further emphasises the need for diagnostic tests for enteric fever and other common febrile diseases.\nSince antimicrobials, specifically fluoroquinolones, are one of the only routinely used control measures for enteric fever, effective surveillance programmes, the assessment of novel diagnostics, new treatment options, and the use of existing vaccines and development of next-generation vaccines are now a greater priority than ever.\nTrial profileSalmonella enterica Typhi or Salmonella enterica Paratyphi A were isolated from the blood of patients with culture-confirmed enteric fever.\nTime to treatment failure and fever clearance timeTime to treatment failure shown in the (A) modified intention to treat, (B) culture-confirmed, and (C) culture-negative populations. Fever clearance times shown in the (D) modified intention to treat, (E) culture-confirmed, and (F) culture-negative populations. Fever clearance times were interval-censored; because numbers at risk are not well defined in this setting they are not shown for graphs D, E, or F. HR=hazard ratio. AF=acceleration factor.\n\nBaseline characteristics of the modified intention-to-treat population\n |  | Gatifloxacin (N=120) | Ceftriaxone (N=119)\n |  | N | n (%) or median (IQR) | N | n (%) or median (IQR)\nAge (years) | 120 | 19\u00b70 (15\u00b70\u221223\u00b70) | 119 | 20\u00b70 (14\u00b70\u221223\u00b75)\nSex |  |  |  | \n | Male | 120 | 99 (83%) | 119 | 81 (68%)\n | Female | 120 | 21 (18%) | 119 | 38 (32%)\nTemperature (\u00b0C) | 116 | 38\u00b78 (38\u00b73\u221239\u00b74) | 116 | 38\u00b78 (38\u00b73\u221239\u00b74)\nDays of illness before enrolment | 120 | 5\u00b70 (4\u00b70\u22126\u00b70) | 119 | 5\u00b70 (4\u00b70\u22127\u00b70)\nTreatment with antibiotics in past 2 weeks | 120 | 21 (18%) | 119 | 17 (14%)\nPrevious history of typhoid | 120 | 18 (15%) | 118 | 19 (16%)\nFamily history of typhoid | 120 | 18 (15%) | 119 | 17 (14%)\nTyphoid vaccination | 119 | 6 (5%) | 119 | 5 (4%)\nFever | 120 | 120 (100%) | 118 | 118 (100%)\nCough | 115 | 38 (33%) | 113 | 42 (37%)\nConstipation | 117 | 9 (8%) | 116 | 16 (14%)\nHeadache | 119 | 99 (83%) | 116 | 108 (93%)\nDiarrhoea | 117 | 25 (21%) | 116 | 28 (24%)\nVomiting | 116 | 32 (28%) | 116 | 30 (26%)\nAbdominal pain | 114 | 31 (27%) | 115 | 27 (23%)\nAnorexia | 118 | 88 (75%) | 116 | 80 (69%)\nNausea | 116 | 60 (52%) | 114 | 55 (48%)\nSplenomegaly | 117 | 0 | 114 | 2 (2%)\nHepatomegaly | 117 | 0 | 114 | 0\nRandom blood glucose (mmol/L) | 117 | 5\u00b738 (4\u00b777\u22126\u00b711) | 117 | 5\u00b738 (4\u00b794\u22125\u00b788)\nCreatinine (\u03bcmol/L) | 116 | 70\u00b772 (61\u00b788\u221279\u00b756) | 114 | 70\u00b772 (61\u00b788\u221279\u00b756)\nTotal bilirubin (\u03bcmol/L) | 117 | 13\u00b768 (10\u00b726\u221217\u00b710) | 117 | 11\u00b797 (10\u00b726\u221215\u00b739)\nLeucocyte cell count (\u00d7109/L) | 120 | 6\u00b79 (4\u00b78\u22127\u00b72) | 119 | 5\u00b78 (4\u00b77\u22127\u00b73)\nHaematocrit (%) | 119 | 39\u00b76 (37\u00b70\u221243\u00b70) | 116 | 38\u00b77 (35\u00b78\u221244\u00b70)\nPlatelet cell count (\u00d7109/L) | 120 | 170\u00b70 (150\u00b70\u2212210\u00b70) | 119 | 167\u00b70 (145\u00b75\u2212203\u00b70)\nAST (U/L) | 117 | 46\u00b70 (32\u00b70\u221266\u00b70) | 116 | 51\u00b75 (38\u00b78\u221280\u00b70)\nALT (U/L) | 117 | 46\u00b70 (30\u00b70\u221263\u00b70) | 117 | 45\u00b70 (33\u00b70\u221263\u00b70)\nCulture positive |  |  |  | \n | Salmonella Paratyphi A isolated | 120 | 19 (16%) | 119 | 16 (13%)\n | Salmonella Typhi isolated | 120 | 43 (36%) | 119 | 38 (32%)\nNo growth or culture negative | 120 | 58 (48%) | 119 | 65 (55%)\n\nN refers to the number of patients with non-missing data in each group. AST=serum aspartate aminotransferase. ALT=serum alanine aminotransferase.\n\nMinimum inhibitory concentration of organism in the culture-confirmed population at enrolment\n |  |  | All patients | Gatifloxacin | Ceftriaxone\nSalmonella Typhi | N=81 | N=43 | N=38\n | MIC against ciprofloxacin (\u03bcg/mL) | n=78 | n=41 | n=37\n |  | MIC 50 | 0\u00b738 | 0\u00b738 | 0\u00b738\n |  | MIC 90 | >32\u00b700 | >32\u00b700 | 13\u00b740\n |  | Range* | 0\u00b7008\u2013>32\u00b700 | 0\u00b7008\u2013>32\u00b700 | 0\u00b7016\u2013>32\u00b700\n | MIC against gatifloxacin (\u03bcg/mL) | n=78 | n=41 | n=37\n |  | MIC 50 | 0\u00b7125 | 0\u00b7125 | 0\u00b7125\n |  | MIC 90 | 2\u00b7000 | 2\u00b7000 | 1\u00b7250\n |  | Range* | 0\u00b7006\u22123\u00b7000 | 0\u00b7006\u22123\u00b7000 | 0\u00b7006\u22123\u00b7000\n | MIC against ceftriaxone (\u03bcg/mL) | n=78 | n=41 | n=37\n |  | MIC 50 | 0\u00b7094 | 0\u00b7094 | 0\u00b7125\n |  | MIC 90 | 0\u00b7190 | 0\u00b7190 | 0\u00b7190\n |  | Range* | 0\u00b7032\u22120\u00b7640 | 0\u00b7032\u22120\u00b7250 | 0\u00b7047\u22120\u00b7640\nSalmonella Paratyphi A | N=35 | N=19 | N=16\n | MIC against ciprofloxacin (\u03bcg/mL) | n=34 | n=18 | n=16\n |  | MIC 50 | 0\u00b7500 | 0\u00b7625 | 0\u00b7500\n |  | MIC 90 | 0\u00b7925 | 1\u00b7000 | 0\u00b7750\n |  | Range* | 0\u00b7380\u22121\u00b7000 | 0\u00b7380\u22121\u00b7000 | 0\u00b7380\u22121\u00b7000\n | MIC against gatifloxacin (\u03bcg/mL) | n=34 | n=18 | n=16\n |  | MIC 50 | 0\u00b7500 | 0\u00b7500 | 0\u00b7500\n |  | MIC 90 | 0\u00b7750 | 0\u00b7575 | 0\u00b7750\n |  | Range* | 0\u00b7380\u22120\u00b7750 | 0\u00b7380\u22120\u00b7750 | 0\u00b7380\u22120\u00b7750\n | MIC against ceftriaxone (\u03bcg/mL) | n=34 | n=18 | n=16\n |  | MIC 50 | 0\u00b7125 | 0\u00b7125 | 0\u00b7125\n |  | MIC 90 | 0\u00b7190 | 0\u00b7145 | 0\u00b7220\n |  | Range* | 0\u00b7064\u22120\u00b7500 | 0\u00b7094\u22120\u00b7190 | 0\u00b7064\u22120\u00b7500\n\nn refers to the number of patients with non-missing data in each group. MIC=minimum inhibitory concentration. MIC 50=minimum inhibitory concentration at the 50th percentile. MIC 90=minimum inhibitory concentration at the 90th percentile.\nRange from the minimum to the maximum noted MIC.\n\nTreatment failure (primary endpoint) overall and in predefined subgroups\n |  | Gatifloxacin (events/n [%]) | Ceftriaxone (events/n [%]) | Hazard ratio of time to failure (95% CI); p value | Heterogeneity test (pinteraction value)\nAll patients (modified intention-to-treat population) | 18/120 (15%) | 19/119 (16%) | 1\u00b704 (0\u00b755\u22121\u00b798); p=0\u00b791 | \nCulture-negative or culture-positive populations |  |  |  | <0\u00b70001\n | Culture negative | 2/58 (3%) | 15/65 (23%) | 7\u00b750 (1\u00b771\u221232\u00b780); p=0\u00b701 | \n | Culture positive | 16/62 (26%) | 4/54 (7%) | 0\u00b724 (0\u00b708\u22120\u00b773); p=0\u00b701 | \nPathogen (culture-confirmed population) |  |  |  | 0\u00b725\n | Salmonella Paratyphi A | 1/19 (5%) | 1/16 (6%) | 1\u00b713 (0\u00b707\u221218\u00b702); p=0\u00b793 | \n | Salmonella Typhi | 15/43 (35%) | 3/38 (8%) | 0\u00b718 (0\u00b705\u22120\u00b762); p=0\u00b701 | \nAge (modified intention-to-treat population) |  |  |  | 0\u00b725\n | <16 years | 6/32 (19%) | 4/36 (11%) | 0\u00b757 (0\u00b716\u22122\u00b700); p=0\u00b738 | \n | \u226516 years | 12/88 (14%) | 15/83 (18%) | 1\u00b731 (0\u00b761\u22122\u00b780); p=0\u00b748 | \nAge (culture-confirmed population) |  |  |  | 0\u00b776\n | <16 years | 6/21 (29%) | 1/16 (6%) | 0\u00b719 (0\u00b702\u22121\u00b762); p=0\u00b713 | \n | \u226516 years | 10/41 (24%) | 3/38 (8%) | 0\u00b727 (0\u00b707\u22120\u00b798); p=0\u00b7047 | \nSex (modified intention-to-treat population) |  |  |  | 0\u00b752\n | Female | 3/21 (14%) | 4/38 (11%) | 0\u00b769 (0\u00b715\u22123\u00b707); p=0\u00b762 | \n | Male | 15/99 (15%) | 15/81 (19%) | 1\u00b721 (0\u00b759\u22122\u00b747); p=0\u00b761 | \nSex (culture-confirmed population) |  |  |  | 0\u00b708\n | Female | 3/11 (27%) | 0/17 | 0 (0\u2013\u221e); p=1\u00b700 | \n | Male | 13/51 (25%) | 4/37 (11%) | 0\u00b737 (0\u00b712\u22121\u00b715); p=0\u00b709 | \nRecruitment date (modified intention-to-treat population) |  |  |  | 0\u00b715\n | Before April 1, 2013 | 7/62 (11%) | 11/59 (19%) | 1\u00b769 (0\u00b766\u22124\u00b736); p=0\u00b728 | \n | April 1, 2013, or later | 11/58 (19%) | 8/60 (13%) | 0\u00b765 (0\u00b726\u22121\u00b761); p=0\u00b735 | \nRecruitment date (culture-confirmed population) |  |  |  | 0\u00b770\n | Before April 1, 2013 | 6/33 (18%) | 1/28 (4%) | 0\u00b718 (0\u00b702\u22121\u00b746); p=0\u00b711 | \n | April 1, 2013, or later | 10/29 (34%) | 3/26 (12%) | 0\u00b728 (0\u00b708\u22121\u00b700); p=0\u00b705 | \nMIC against ciprofloxacin (culture-confirmed population) |  |  |  | 0\u00b715\n | <0\u00b712 \u03bcg/mL | 0/4 | 1/3 (33%) | \u221e (0\u2013\u221e); p=1\u00b700 | \n | 0\u00b712\u22122\u00b700 \u03bcg/mL | 8/45 (18%) | 2/46 (4%) | 0\u00b722 (0\u00b705\u22121\u00b705); p=0\u00b706 | \n | >2\u00b700 \u03bcg/mL* | 8/10 (80%) | 1/4 (25%) | 0\u00b717 (0\u00b702\u22121\u00b738); p=0\u00b710 | \nMIC against gatifloxacin (culture-confirmed population) |  |  |  | 0\u00b758\n | \u22641 \u03bcg/mL | 8/49 (16%) | 3/49 (6%) | 0\u00b734 (0\u00b709\u22121\u00b728); p=0\u00b711 | \n | >1 \u03bcg/mL | 8/10 (80%) | 1/4 (25%) | 0\u00b717 (0\u00b702\u22121\u00b738); p=0\u00b710 | \n\nMIC=minimum inhibitory concentration.\nAmong the 14 strains, two had a ciprofloxacin MIC of 24 \u03bcg/mL and 12 had a ciprofloxacin MIC>32 \u03bcg/mL.\n\nFever clearance time (secondary endpoint) overall and in predefined subgroups\n |  | Gatifloxacin | Ceftriaxone | Acceleration factor (95% CI), p value | Heterogeneity test (pinteraction value)\n |  | n | Median (IQR) days | n | Median (IQR) days |  | \nAll patients (modified intention-to-treat population) | 120 | 2\u00b743 (1\u00b709\u22124\u00b756) | 119 | 2\u00b793 (1\u00b744\u22125\u00b712) | 0\u00b789 (0\u00b772\u22121\u00b711); p=0\u00b731 | \nCulture-negative or culture-positive population |  |  |  |  |  | <0\u00b70001\n | Culture negative | 58 | 1\u00b712 (0\u00b739\u22122\u00b758) | 65 | 3\u00b703 (1\u00b731\u22125\u00b785) | 0\u00b744 (0\u00b730\u22120\u00b765); p<0\u00b70001 | \n | Culture positive | 62 | 4\u00b721 (2\u00b763\u22126\u00b710) | 54 | 2\u00b778 (1\u00b762\u22124\u00b726) | 1\u00b742 (1\u00b715\u22121\u00b776); p=0\u00b7001 | \nPathogen (culture-confirmed population) |  |  |  |  |  | 0\u00b757\n | Salmonella Paratyphi A | 19 | 3\u00b768 (2\u00b750\u22124\u00b798) | 16 | 2\u00b724 (1\u00b718\u22123\u00b769) | 1\u00b731 (0\u00b788\u22121\u00b794); p=0\u00b719 | \n | Salmonella Typhi | 43 | 4\u00b751 (2\u00b779\u22126\u00b758) | 38 | 3\u00b703 (1\u00b786\u22124\u00b746) | 1\u00b747 (1\u00b715\u22121\u00b788); p=0\u00b7002 | \nAge (modified intention-to-treat population) |  |  |  |  |  | 0\u00b729\n | <16 years | 32 | 3\u00b702 (1\u00b770\u22124\u00b775) | 36 | 2\u00b723 (0\u00b793\u22124\u00b746) | 1\u00b708 (0\u00b772\u22121\u00b760); p=0\u00b772 | \n | \u226516 years | 88 | 2\u00b722 (0\u00b792\u22124\u00b745) | 83 | 3\u00b725 (1\u00b770\u22125\u00b741) | 0\u00b783 (0\u00b764\u22121\u00b708); p=0\u00b716 | \nAge (culture-confirmed population) |  |  |  |  |  | 0\u00b746\n | <16 years | 21 | 3\u00b782 (2\u00b745\u22125\u00b743) | 16 | 3\u00b704 (1\u00b794\u22124\u00b732) | 1\u00b726 (0\u00b790\u22121\u00b776); p=0\u00b718 | \n | \u226516 years | 41 | 4\u00b743 (2\u00b777\u22126\u00b743) | 38 | 2\u00b766 (1\u00b747\u22124\u00b722) | 1\u00b751 (1\u00b715\u22121\u00b797); p=0\u00b7003 | \nSex (modified intention-to-treat population) |  |  |  |  |  | 0\u00b799\n | Female | 21 | 2\u00b741 (1\u00b713\u22124\u00b739) | 38 | 2\u00b778 (1\u00b734\u22124\u00b794) | 0\u00b789 (0\u00b756\u22121\u00b741); p=0\u00b761 | \n | Male | 99 | 2\u00b744 (1\u00b709\u22124\u00b759) | 81 | 2\u00b799 (1\u00b748\u22125\u00b720) | 0\u00b789 (0\u00b768\u22121\u00b714); p=0\u00b735 | \nSex (culture-confirmed population) |  |  |  |  |  | 0\u00b768\n | Female | 11 | 4\u00b727 (3\u00b712\u22125\u00b747) | 17 | 3\u00b706 (1\u00b793\u22124\u00b741) | 1\u00b725 (0\u00b786\u22121\u00b783); p=0\u00b725 | \n | Male | 51 | 4\u00b718 (2\u00b756\u22126\u00b716) | 37 | 2\u00b766 (1\u00b750\u22124\u00b717) | 1\u00b746 (1\u00b713\u22121\u00b789); p=0\u00b7004 | \nRecruitment date (modified intention-to-treat population) |  |  |  |  |  | 0\u00b709\n | Before April 1, 2013 | 62 | 2\u00b730 (1\u00b710\u22124\u00b709) | 59 | 2\u00b779 (1\u00b716\u22125\u00b755) | 0\u00b774 (0\u00b753\u22121\u00b703); p=0\u00b708 | \n | April 1, 2013, or later | 58 | 2\u00b760 (1\u00b712\u22125\u00b704) | 60 | 3\u00b705 (1\u00b776\u22124\u00b769) | 1\u00b709 (0\u00b782\u22121\u00b745); p=0\u00b756 | \nRecruitment date (culture-confirmed population) |  |  |  |  |  | 0\u00b712\n | Before April 1, 2013 | 33 | 3\u00b788 (2\u00b763\u22125\u00b726) | 28 | 2\u00b754 (1\u00b731\u22124\u00b727) | 1\u00b721 (0\u00b790\u22121\u00b763); p=0\u00b722 | \n | April 1, 2013, or later | 29 | 4\u00b768 (2\u00b782\u22126\u00b797) | 26 | 3\u00b700 (1\u00b796\u22124\u00b720) | 1\u00b768 (1\u00b726\u22122\u00b723); p=0\u00b70004 | \nMIC against ciprofloxacin (culture-confirmed population) |  |  |  |  |  | 0\u00b702\n | <0\u00b712 \u03bcg/mL | 4 | 2\u00b755 (1\u00b782\u22123\u00b732) | 3 | 4\u00b798 (4\u00b709\u22125\u00b782) | 0\u00b758 (0\u00b735\u22120\u00b794); p=0\u00b7028 | \n | 0\u00b712\u22122\u00b700 \u03bcg/mL | 45 | 3\u00b788 (2\u00b767\u22125\u00b721) | 46 | 2\u00b763 (1\u00b749\u22124\u00b712) | 1\u00b724 (0\u00b799\u22121\u00b756); p=0\u00b706 | \n | >2\u00b700 \u03bcg/mL* | 10 | 8\u00b720 (5\u00b799\u221210\u00b750) | 4 | 3\u00b766 (2\u00b784\u22124\u00b746) | 2\u00b736 (1\u00b758\u22123\u00b751); p=<0\u00b70001 | \nMIC against gatifloxacin (culture-confirmed population) |  |  |  |  |  | 0\u00b7049\n | \u22641\u00b700 \u03bcg/mL | 49 | 3\u00b776 (2\u00b756\u22125\u00b708) | 49 | 2\u00b775 (1\u00b758\u22124\u00b727) | 1\u00b717 (0\u00b794\u22121\u00b745); p=0\u00b715 | \n | >1\u00b700 \u03bcg/mL | 10 | 8\u00b720 (5\u00b799\u221210\u00b750) | 4 | 3\u00b766 (2\u00b784\u22124\u00b746) | 2\u00b736 (1\u00b758\u22123\u00b751); p<0\u00b70001 | \n\nPercentages not added to this table because the denominators for populations change and are not clearly specified. MIC=minimum inhibitory concentration.\nAmong the 14 strains, two had a ciprofloxacin MIC of 24 \u03bcg/mL and 12 had an MIC >32 \u03bcg/mL.",
        "label": "low",
        "id": "task4_RLD_test_191"
    },
    {
        "paper_doi": "10.1371/journal.pone.0007871",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: An open label RCT (non-inferiority).Follow-up: Children were kept at the health facility for the three day treatment period, and then returned on days 7, 14, 21, 28, 35, and 42, and any time symptoms occurred, for clinical assessment and blood smears. Haematological and biochemical assessments were carried out at enrolment, days 3, 28, and 42 and at clinician request.Adverse event monitoring: Monitoring and recording of adverse events was carried out throughout the trial. A 12-lead ECG was performed at enrolment and on days 2 and 7 to assess any QT/QTc interval prolongation.\n\n\nParticipants: Number of participants: 1553Inclusion criteria: Uncomplicated malaria, age 6 to 59 months, body weight > 5 kg, fever (axillary temperature >= 37.5 degC) or history of fever in the preceding 24 hrs, microscopically confirmed P. falciparum mono-infection, asexual parasite densities between 2,000 and 200,000/mL, informed consentExclusion criteria: severe malaria or other danger signs, acute malnutrition (weight for height < 70% of the median National Center for Health Statistics/WHO reference), any other concomitant illness or underlying disease, contra-indication to trial drugs, ongoing antimalarial prophylaxis, ECG abnormality requiring urgent management\n\n\nInterventions: 1. DHA-P, fixed dose combination, 40 mg/320 mg tablets and 20 mg/160 mg tablets (Eurartesim(r), Sigma-Tau)Daily dose of 2.25 mg/kg dihydroartemisinin and 18 mg/kg piperaquine, rounded up to the nearest half tablet2. Artemether-lumefantrine, fixed dose combination, 20 mg/120 mg (Coartem: Novartis)5 to 14 kg 1 tablet twice daily for 3 days15 to 24 kg 2 tablets twice daily for 3 days25 to 34 kg 3 tablets twice daily for 3 daysAll doses were supervised.\n\n\nOutcomes: Adequate clinical and parasitological response on days 14, 28, and 42, PCR-adjusted and PCR-unadjustedGametocyte presence and clearanceHb changes from baseline to day 28Adverse eventsNot included in this review:Fever clearance timeParasite clearance time\n\n\nNotes: Country: Burkina Faso, Kenya, Mozambique, Uganda, and ZambiaSetting: Four rural sites and one peri-urban site.Transmission: Malaria mesoendemic at all sites. Two sites had high transmission in one period of the year (Jun to Dec or Nov to May), three others had perennial malaria with two sites having two peak seasons and one with marked seasonality (Oct to Apr)Resistance: Documented resistance to chloroquine ranged from 35% in Burkina Faso to 81% in UgandaDates: Aug 2005 to Jul 2006Funding: Medicine for Malaria Venture and Sigma-Tau I.F.R. SpA (Rome)\n\n",
        "objective": "To evaluate the effectiveness and safety of DHA\u2010P compared to other ACTs for treating uncomplicated P. falciparum malaria in adults and children.",
        "full_paper": "Background\nArtemisinin combination therapies (ACTs) are currently the preferred option for treating uncomplicated malaria.\nDihydroartemisinin-piperaquine (DHA-PQP) is a promising fixed-dose ACT with limited information on its safety and efficacy in African children.\nMethodology/Principal Findings\nThe non-inferiority of DHA-PQP versus artemether-lumefantrine (AL) in children 6\u201359 months old with uncomplicated P. falciparum malaria was tested in five African countries (Burkina Faso, Kenya, Mozambique, Uganda and Zambia).\nPatients were randomised (2\u22361) to receive either DHA-PQP or AL.\nNon-inferiority was assessed using a margin of \u22125% for the lower limit of the one-sided 97.5% confidence interval on the treatment difference (DHA-PQP vs. AL) of the day 28 polymerase chain reaction (PCR) corrected cure rate.\nEfficacy analysis was performed in several populations, and two of them are presented here: intention-to-treat (ITT) and enlarged per-protocol (ePP).\n1553 children were randomised, 1039 receiving DHA-PQP and 514 AL.\nThe PCR-corrected day 28 cure rate was 90.4% (ITT) and 94.7% (ePP) in the DHA-PQP group, and 90.0% (ITT) and 95.3% (ePP) in the AL group.\nThe lower limits of the one-sided 97.5% CI of the difference between the two treatments were \u22122.80% and \u22122.96%, in the ITT and ePP populations, respectively.\nIn the ITT population, the Kaplan-Meier estimate of the proportion of new infections up to Day 42 was 13.55% (95% CI: 11.35%\u201315.76%) for DHA-PQP vs 24.00% (95% CI: 20.11%\u201327.88%) for AL (p<0.0001).\nConclusions/Significance\nDHA-PQP is as efficacious as AL in treating uncomplicated malaria in African children from different endemicity settings, and shows a comparable safety profile.\nThe occurrence of new infections within the 42-day follow up was significantly lower in the DHA-PQP group, indicating a longer post-treatment prophylactic effect.\nTrial Registration\nControlled-trials.com ISRCTN16263443\nIntroduction\nArtemisinin-based combination therapies (ACTs) are highly efficacious and fast acting antimalarial medicines.\nThe World Health Organization (WHO) recommends their use for treating uncomplicated malaria.\nIn Africa, their introduction on a wide scale began in 2003 and currently most African countries have adopted or are using ACTs as first or second line treatments, either artesunate-amodiaquine or artemether-lumefantrine (AL), available as co-formulations produced under GMP, though the former is also used as a co-blistered or non-co-formulated product.\nThe co-formulation of dihydroartemisinin (DHA), the active metabolite of artemisinin derivatives, with piperaquine (PQP), a bisquinoline structurally close to chloroquine, seems to be a promising combination and a good alternative to AL, whose optimal use in the public health system is challenged by the twice-daily dosing scheme and the need for co-administration with fatty food, necessary for improving the absorption of lumefantrine.\nDHA-PQP provides a simpler dosage scheme (a single daily dose over 3 days) than AL and is generally administered without specific food instructions, though recent data indicate that co-administration with fat (milk, biscuit, or other food) increases bio-availability of piperaquine and possibly efficacy.\nSeveral trials have assessed DHA-PQP safety, efficacy and effectiveness, mostly in Asia, reporting an efficacy of about 90% over 28\u201363 days.\nThere is little information on the safety and efficacy of DHA-PQP in African children, as only a few single-centre trials have been done in Africa.\nDHA-PQP is registered in several countries in Africa and South-East Asia and has been widely used in Vietnam or Cambodia, though these formulations are not manufactured according to internationally recognised GMP.\nIn 2005, a public-private partnership programme funded by the Medicines for Malaria Venture (MMV) and led by the Italian Company Sigma-Tau I.F.R. SpA (Rome) in collaboration with the University of Oxford was set up to fill the gaps needed for the international registration of DHA-PQP.\nThis included a phase III, randomized multicentre trial to test the non-inferiority of DHA-PQP compared with AL in treating uncomplicated malaria in African children.\nMethods\nThe protocol for this trial and supporting non-inferiority adapted CONSORT checklist are available and annexed as supporting information; see Protocol S1 and Checklist S1.\nEthical Considerations and Patient Safety\nThe study protocol was approved by the Institutional Review Board of the Institute of Tropical Medicine, Antwerp, the Ethical Committee of the Antwerp University Hospital, the University of Heidelberg Ethics Committee and by the National Ethics Review Committee or Institutional Review Board at each trial site.\nThe trial was conducted under the provisions of the Declaration of Helsinki and in accordance with Good Clinical Practices guidelines set up by the International Conference on Harmonization.\nA Study Steering Committee, a Data Monitoring Committee and a Clinical Development Committee were created prior to the beginning of the trial, and worked independently to harmonise and monitor the study.\nThe trial was registered prior to the enrolment of the first patient in the International Standard Randomized Controlled Trials Register, number ISRCTN 16263443, at http://www.controlled-trials.com/isrctn.\nStudy Design, Sites and Concealment of Patient Allocation\nBetween August 2005 and July 2006, a randomised, open-label, multicentre clinical trial was carried out in five African sites (Nanoro, Burkina Faso; Kilifi, Kenya; Manhi\u00e7a, Mozambique; Mbarara, Uganda; and Ndola, Zambia).\nCharacteristics of the five sites are summarized in Table 1.\n[ONLINE PUBLICATION ONLY]\nChildren 6\u201359 months old attending the health facilities with uncomplicated malaria were included in the study if they fulfilled the following inclusion criteria: body weight >5 kg; microscopically confirmed Plasmodium falciparum mono-infection with asexual parasite densities between 2,000 and 200,000/\u00b5l; fever (axillary temperature \u226537.5\u00b0C) or history of fever in the preceding 24 h.\nPatients were not recruited if they met at least one of the following exclusion criteria: severe malaria, or other danger signs; acute malnutrition (weight for height <70% of the median National Center for Health Statistics/WHO reference) or any other concomitant illness or underlying disease; contra-indication to receive the trial drugs or ongoing prophylaxis with drugs having antimalarial activity.\nPatients satisfying the inclusion/exclusion criteria were enrolled if the parent/guardian signed a detailed written informed consent.\nPatients were individually randomised according to a 2\u22361 (DHA-PQP\u2236AL) scheme, so as to have more patients in the DHA-PQP arm to provide better estimates for its cure rates and more cases for the integrated safety data base.\nA randomisation list stratified by country was generated by an independent off site contract research organisation (CRO), with each treatment allocation concealed in opaque sealed envelopes that were opened only after the patient's recruitment.\nBoth drugs were administered under direct supervision during 3 consecutive days, according to the patient's body weight.\nAL (Coartem\u2122, Novartis, Switzerland) was administered twice a day (at enrolment and at 8, 24, 36, 48 and 60 h) according to the following dosage: weight 5\u201314 kg: one tablet per dose; weight 15\u201324 kg: two tablets per dose; weight 25\u201334 kg: three tablets per dose.\nDHA-PQP (Eurartesim\u2122, Sigma-Tau, Italy) was given once daily, at the standard dosage of 2.25 mg/kg and 18 mg/kg per dose of DHA and PQP, respectively, rounded up to the nearest half tablet.\nTo facilitate the correct dosing of DHA-PQP, two formulations were used (DHA 20 mg + PQP 160 mg and DHA 40 mg + PQP 320 mg).\nIn case of vomiting, a full dose was repeated if this occurred within the first half an hour, or half a dose if it occurred between 30 minutes and 1 h.\nAL was administered concomitantly with milk (as recommended by the manufacturer) while for DHA-PQP no specific instructions regarding co-administration with food were given.\nFor infants, drugs were crushed, mixed with water and administered as slurry.\nBoth patient allocation to the different analysis populations and assessment of the primary end-point were made by staff blinded to the treatment assignment and before availability of the PCR results.\nTreatment Follow-Up, Clinical and Laboratory Procedures\nAll children were kept at the health facility for the 3-day dosing period.\nThe mother/guardian was asked to return with the child for scheduled visits on days 7, 14, 21, 28, 35 and 42 post-treatment, or if any symptoms occurred.\nField workers traced patients missing any visit.\nFor each visit, a physical examination was performed by the study clinicians, vital signs were recorded, and axillary temperature measured with an electronic thermometer.\nAdverse events and serious adverse events were recorded and monitored throughout the study.\nA 12-lead electrocardiogram (ECG) was performed at enrolment and repeated on days 2 and 7 to assess any QT/QTc interval prolongation.\nAny ECG abnormality detected at enrolment requiring urgent management was considered an exclusion criterion.\nAll ECG records were transmitted daily online to a central cardiologist (Paris, France) who interpreted them in a blinded manner, and feedback was sent to the sites as soon as available.\nThe QTc interval (ms) was evaluated after correcting for the heart rate with Bazett's or Fridericia's formulae and classified according to the following categories: Normal <430 ms; Borderline: 431\u2013450 ms; Prolonged >450 ms.\nThe study was supervised by monthly monitoring visits.\nRescue treatment for recurrent parasitaemia was according to local national guidelines.\nAll participants, with the exception of those in Kilifi, received a free insecticide-treated bed net at recruitment.\nCapillary or venous blood was taken at every visit.\nThick and thin blood films were prepared, dried and Giemsa-stained, and parasite density estimated by counting the number of asexual parasites in 200 white blood cells (WBC), assuming a standard WBC count of 8,000/\u00b5l.\nQuality control was performed in blind conditions on 20% of all the slides at a central laboratory.\nSamples for haematology (full blood count) and biochemistry (liver and renal function) were taken at enrolment, at days 3, 28 and 42, and at any other visit if judged necessary by the clinician.\nFor PCR analysis, three blood spots were collected on filter paper (Whatmann 3 MM) at enrolment and at any visit after day 7.\nEach filter paper was dried and individually stored in a plastic bag containing silica gel.\nAll filter papers were subsequently transferred to the Institute of Tropical Medicine (Antwerp, Belgium) where centralised genotyping was conducted.\nPurification of DNA was conducted as previously described.\nThree polymorphic genetic markers MSP1, MSP2 and GluRP were used to distinguish recrudescence from new infections.\nRecrudescence was defined as at least one identical allele for each of the three markers in the pre-treatment and post-treatment samples.\nNew infections were diagnosed when all alleles for at least one of the markers differed between the two samples.\nAll gels were re-read under blinded conditions by an independent expert (National Museum of Natural History, Paris, France).\nIn addition, 20% of the filter papers were re-analysed and assessed by an independent laboratory (Shoklo Malaria Research Unit, Mae Sot, Thailand).\nOutcome Classification\nThe primary endpoint was the PCR-corrected adequate clinical and parasitological response (ACPR) at day 28; secondary efficacy outcomes included PCR-corrected cure rates at days 14 and 42, PCR-uncorrected cure rates at days 14, 28 and 42; parasite and fever clearance times, presence and clearance of gametocytes, and haemoglobin (Hb) changes from baseline to day 28.\nAll standard safety outcomes such as incidence of adverse events, changes from baseline on haematology and clinical chemistry parameters, ECG findings and vital sign variation during the study were also evaluated.\nTreatment outcome was analysed in two ways.\nThe first, based on a pre-defined (in the protocol) procedure further developed with the Data Monitoring and the Clinical Development Committees, complemented the WHO definitions (see below) with a set of rules allowing the evaluation of each individually randomised patient, e.g. patients having taken not-allowed anti-malarial drugs or with halfway missing data such as blood parasitaemia (Table 2).\nSuch an approach was defined as primary because it was deemed in line with the requirements of the most stringent regulatory authorities.\nAll cases not strictly matching the WHO definitions and/or the described procedure were reviewed individually at the data review meetings in blind conditions.\nThe second approach, based on the standard definitions of early/late clinical and parasitological failure (World Health Organization), was used to allow comparison with previously published results.\nAccordingly, true treatment failure (TTF) was defined as the sum of the early and late (either LPF or LCF) treatment failures occurring until Day 13 (recrudescence by default) and the late treatment failures from Day 14 onwards classified as recrudescence by PCR analysis.\nStatistical Analysis\nSeveral populations were defined for the analysis and two of them (intention-to-treat, ITT, and enlarged per-protocol, ePP), despite not being the primary ones as defined by the protocol, are presented here on the basis of their comparability with the populations discussed in previous published studies, their clinical relevance and the fact that conclusions are similar on all considered populations.\nThe ITT included all randomised patients having taken at least one dose of the study treatments.\nThe ePP population included all randomised patients fulfilling the protocol eligibility criteria, having taken at least 80% of the study medication when not previously classified as early treatment failures, completing the day 28 assessment and having an evaluable PCR in case of recurrent parasitaemia.\nTable 2 provides details for patient classification in these two populations.\nPrimary efficacy analysis was based on a 97.5% (one-sided) confidence interval (CI) computed on the difference between the day 28 PCR-corrected cure rates (defined as in Table 2) of DHA-PQP and AL.\nTo prove non-inferiority, the lower limit of this CI was to be within \u22125%, the pre-established non-inferiority margin.\nThe Wald method (without continuity correction) was used to compute the CI, as this method was known to provide control of type I error around the nominal level for the 2\u22361 allocation, and also in the context of a hypothesis test of non-inferiority.\nPCR-corrected and uncorrected cure rates at the other time points were assessed similarly.\nTTF was estimated with the Kaplan-Meier method as suggested by WHO, in the ePP population.\nPatients withdrawing the study, with a new infection, or with a non-interpretable or missing PCR were censored at the withdrawal or PCR sampling time.\nSurvival analysis was applied also to the new infections.\nIn this case censoring was applied to recrudescences.\nCure rates were also stratified by country and age (age groups: \u226412 months; >12 months), though the study was not powered for proving efficacy within each country or age group.\nThe Breslow-Day test, or logistic regression when the former was not applicable, was used to assess homogeneity across countries and age groups.\nFor exploratory testing, categorical variables were compared using \u03c72 or Fisher's exact test, and continuous variables using the Student t-test for independent samples.\nRates of person-gametocyte-weeks for measuring gametocyte carriage and transmission potential were calculated as the number of weeks in which blood slides were positive for gametocytes divided by the total number of follow-up weeks and expressed per 1,000 person-weeks.\nAll safety variables were analysed in the ITT population.\nSample Size Calculation\nThis study was designed as a non-inferiority trial.\nAssuming 80% statistical power, a one-sided \u03b1 level of 2.5%, and adopting an unequal 2\u22361 randomisation ratio, 1,500 patients (1000 DHA-PQP, 500 AL) were needed to show that the difference of the day 28 PCR-corrected cure rates between DHA-PQP and AL was within \u22125%, assuming a response rate for AL of at least 92%.\nResults\nTrial Profile and Baseline Characteristics\nOverall, 2,001 patients were screened, and 1,553 recruited and randomised to receive the study drugs (1,039 DHA-PQP and 514 AL) (Figure 1).\nFive patients were excluded from all analyses: one child in each treatment group who did not receive any treatment and three children in the AL group who were recruited twice (only data for the first recruitment were retained).\nA total of 1,548 patients were considered for the ITT population and the safety analysis, and the ePP population consisted of 1,425 patients.\nThe attrition rate of the ePP population as compared to the ITT was approximately 8% and was due to lost-to-follow-up (\u223c2%) or major protocol violations (\u223c6%).\nThese proportions were equally distributed between treatments (data not shown).\nRandomisation generated comparable groups between countries and overall (Table 3).\nEfficacy Results\nDHA-PQP was as efficacious as AL.\nThe day 28 PCR-corrected cure rate was 90.4% (ITT) and 94.7% (ePP) in the DHA-PQP group, and 90.0% (ITT) and 95.3% (ePP) in the AL group (ITT: p\u200a=\u200a0.820; ePP: p\u200a=\u200a0.650).\nThe lower limits of the one-sided 97.5% CIs on the differences between the two treatments were \u22122.80% and \u22122.96%, in the ITT and ePP populations, respectively (Table 4 and Figure 2).\nThe analyses in the other populations and all sensitivity analyses confirmed the robustness of these results.\nThe day 42 PCR-corrected cure rates were lower than those at day 28 but similar for the two treatments for the ITT population (Table 4).\nHowever, the lower limit of the one-sided 97.5% CI on the cure rate difference for the ePP population was \u22125.29%, a value slightly below the pre-established non-inferiority margin (Table 4 and Figure 2).\nThe details of the classification of patients for the PCR-corrected response both in the ITT and ePP populations are presented in Table 4 for both the analyses at day 28 and day 42.\nThe percentage of recrudescent infections and new infections, as detected by PCR at or before day 28, was lower in the DHA-PQP group with respect to the AL group, while there were more withdrawals and/or treatment failures at or before day 14 in the DHA-PQP group compared with the AL group (Table 4).\nAt day 42, the findings were similar with slightly more recrudescent infections in the DHA-PQP group.\nThe day 28 PCR-corrected cure rates in infants (6\u201311 months-old) were similar to those in older children and above 90% in both treatment groups (ITT: DHA-PQP 90.70%, AL 92.65%, p\u200a=\u200a0.643, 97.5% CI>\u22129.92%).\nThe uncorrected cure rates were significantly higher in the DHA-PQP group, both at day 28 (ITT: DHA-PQP 87.7% vs. AL 76.7%, p<0.001, 97.5% CI>6.82%; ePP: DHA-PQP 91.99% vs. AL 81.03%, p<0.001, 97.5% CI>6.99%) and at day 42 (ITT: DHA-PQP 74.08% vs. AL 64.71%, p<0.001, 97.5% CI>4.45%; ePP: DHA-PQP 77.63% vs. AL 68.75%, p<0.001, 97.5% CI>3.90%).\nThis was mainly due to fewer late failures later classified as new infections in the DHA-PQP as compared to the AL arm.\nIn the ITT population, new infections until day 42 occurred significantly less in the DHA-PQP group (Kaplan-Meier estimate: 13.55%; 95% CI: 11.35%\u201315.76%) than in the AL group (Kaplan-Meier estimate: 24.00%; 95% CI: 20.11%\u201327.88%) (Figure 3).\nSimilar results were obtained in the ePP population (data not shown).\nWhen the day 28 PCR-corrected cure rates were analysed by country, the heterogeneity test was borderline significant at the 10% level only in the ePP population (ITT: p\u200a=\u200a0.324; ePP: p\u200a=\u200a0.082), suggesting some minor differences among sites (Figure 2).\nHowever, the CIs adjusted by country were almost identical to the unadjusted ones (data not shown).\nAt day 42, the heterogeneity test was not statistically significant in either the ITT or the ePP populations (ITT: p\u200a=\u200a0.582; ePP: p\u200a=\u200a0.703).\nHeterogeneity across countries was more marked for the uncorrected cure rates at day 28 (ITT and ePP: p<0.001), while at day 42 the heterogeneity test was borderline significant only in the ePP population (ITT: p\u200a=\u200a0.186; ePP: p\u200a=\u200a0.051).\nHowever, with the exception of Kenya, such differences on the uncorrected cure rates were of a quantitative type, i.e. rather in the size of the treatment effect across countries, not in its direction, always favouring DHA-PQP.\nWhen considering the WHO standard definition of TTF, the two treatment groups were similar at day 28 (Kaplan-Meier estimate in ePP: DHA-PQP 3.78% [95%CI: 2.57%\u20135.00%] vs AL 3.19% [95%CI: 1.54%\u20134.84%], p\u200a=\u200a0.528), while at day 42 TTF tended to be lower in the AL group (Kaplan-Meier estimate in ePP: DHA-PQP 6.86% [95%CI: 5.22%\u20138.50%] vs AL 4.52%, [95%CI: 2.52%\u20136.51%], p\u200a=\u200a0.119).\nParasite clearance was rapid in both treatment groups (Kaplan-Meier estimate of median time was 2 days in each group, in both populations).\nAbout 60% of patients had fever at baseline while at day 2 more than 97% of patients were afebrile in both treatment groups.\nGametocyte prevalence at recruitment was similar in both study arms (ITT: DHA-PQP 11.75%; AL 12.94%, p\u200a=\u200a0.501; ePP: DHA-PQP 11.55%; AL 13.36%, p\u200a=\u200a0.326).\nHowever, gametocyte carriage measured as rate of person-gametocyte-weeks was significantly higher in the DHA-PQP group than in the AL group, both for the ITT (DHA-PQP: 43.97/1,000; AL: 21.43/1,000; p\u200a=\u200a0.005) and the ePP (DHA-PQP: 42.65/1,000; AL: 21.23/1,000; p\u200a=\u200a0.006) populations.\nHaemoglobin changes from baseline to day 28 were comparable between treatment groups (data not shown) while the change from baseline to the last available data was significantly higher in the DHA-PQP group than in the AL group (ITT: 17.0\u00b118.18 g/L vs 14.27\u00b118.54 g/L, p\u200a=\u200a0.007; ePP: 17.19\u00b117.96 g/L vs 15.07\u00b118.56 g/L, p\u200a=\u200a0.044).\nSafety Results\nBoth DHA-PQP and AL were well tolerated with the majority of adverse events of mild or moderate severity, and consistent with symptoms attributable to malaria (Table 5).\nThere were no significant differences in the occurrence of events, including serious adverse events.\nGastrointestinal tolerability of both drugs was similar (DHA-PQP: 207/1038, 19.9%; AL: 92/510, 18.0%), with the majority of events being mild.\nCutaneous adverse events were infrequent, and mainly involved minor dermatitis or rash (DHA-PQP: 70/1038, 6.7%; AL: 29/510, 5.7%).\nThree patients developed urticaria (one (0.1%) in the DHA-PQP group and two (0.4%) in the AL group) and three more developed mild hypersensitivity (two (0.2%) in the DHA-PQP group and one (0.2%) in the AL group).\nNone of them required hospitalization.\nOccurrence of laboratory AEs, e.g. neutropenia (DHA-PQP: 18/1038, 1.7%; AL: 12/510, 2.4%) and altered liver enzymes (ALT) (DHA-PQP: 20/1038, 1.9%; AL: 19/510, 3.7%), was similar between the two treatment groups.\nECG was performed in more than 98% of patients at day 0 and day 2, always before the administration of the treatment (96% of patients had ECG also at day 7).\nIn the DHA-PQP group, the proportion of patients with borderline (29.1%) and prolonged (9.1%) QTc interval at day 2 corrected by the Bazett's method was higher than in the AL group (19.8% and 6.9%) (p<0.001).\nHowever, this was not confirmed when applying the Fridericia's correction as the corresponding proportions were 1.0% and 0.2% in the DHA-PQP group and 1.2% and 0.2% in the AL group (p\u200a=\u200a0.76).\nIn addition, a \u226560 ms increase of the QTc interval between day 0 and day 2 (Bazett's correction) was observed in just 2.7% (DHA-PQP) and 2.0% (AL) patients; only two patients per group showed a QTc at day 2 higher than 500 ms.\nWhen considering the occurrence of the AE \u201cElectrocardiogram QT prolonged\u201d, similar percentages (DHA-PQP: 26/1038, 2.5%; AL: 13/510, 2.6%) were observed in the two treatment groups (Table 5).\nNo other difference between groups was observed during the follow up (data not shown).\nTwo deaths (one per group) occurred during the study.\nIn Uganda, a 3 year-old girl died 24 h after commencing treatment with DHA-PQP.\nSepsis or severe malaria was considered by the investigating clinician as the most likely cause.\nIn Mozambique, an 18 month-old girl died 7 h after the first dose of AL.\nSevere malaria was considered the most likely cause of death, although other aetiologies such as sepsis, hypoglycaemia, heart conditions or bronco-aspiration could not be excluded.\nNo autopsy could be performed in these two children and a causal relationship with the treatment could not be ruled out.\nDiscussion\nThe fulfilment of the non-inferiority criterion on all analysis populations and the confirmation that in this study the comparator treatment had the expected efficacy proved that DHA-PQP is non inferior to AL in treating African children aged 6\u201359 months with uncomplicated malaria.\nThe two treatments had similar safety profiles.\nOur study confirms the results of previous trials in Asia and Africa that found DHA-PQP to be as effective as other ACTs, including AL.\nA recent study in Papua New Guinea (PNG) reported a significantly higher cure rate (adequate clinical and parasitological response) in children treated with AL as compared to DHA-PQP.\nThe reasons for such discordant results are unclear though the authors mention the cross-resistance between chloroquine and PQP.\nHowever, PQP, though structurally related to chloroquine, has been shown to be effective in vitro against chloroquine-resistant strains.\nIn addition, it has been suggested that the lower-than-expected DHA-PQP efficacy reported in PNG may be due to administration of the treatment without any food.\nIndeed, PQP is highly lipid-soluble and its oral bioavailability is enhanced when given with food, though an additional study in Vietnamese healthy volunteers reports no influence of food intake (standardised Vietnamese meal) on PQP pharmacokinetics.\nThe issue on whether to recommend the administration of DHA-PQP with a biscuit or a glass of milk remains unanswered.\nThough co-administration with food may improve the drug's bioavailability, it is unclear whether this will translate in a higher efficacy.\nIn our study, DHA-PQP was given without specific instructions regarding co-administration with food but its efficacy at day 42 was over 90%, similar to that reported in a study carried out in Uganda but lower than in two other African studies.\nMoreover, no clinically relevant heterogeneity was shown across the five African countries despite the high chloroquine resistance previously reported from most study sites.\nWhen taking into account all recurrent infections observed during the follow up period, i.e., without the PCR correction, the cure rates for DHA-PQP were significantly better than AL, indicating a better post-treatment prophylaxis (PTP) than AL and confirming that chloroquine resistance did not interfere with DHA-PQP efficacy.\nThe significantly higher Hb change from baseline to the last available data in the DHA-PQP group is in line with this observation.\nTherefore, the longer PQP's elimination half-life (about 20 days) as compared to lumefantrine (4\u201310 days), provides a longer PTP, prevents the emergence of new infections and improves the patient's haematological recovery, despite a significant chloroquine resistance background.\nWhile this is clearly an advantage for the individual, at the population level, it may increase the risk of selecting resistant parasites among the new infection and stress the need of matching the large scale deployment of DHA-PQP with the careful monitoring of resistance.\nOne hundred twenty nine infants aged 6\u201311 months treated with DHA-PQP responded as well to treatment as older children, though the study was not powered to confirm non-inferiority between the two treatment groups.\nInfants represent a special group as they are more at risk of malaria and of receiving inadequate doses of antimalarial treatments.\nIn Papua, Indonesia, the PQP plasma concentration at day 7 was the major determinant of the therapeutic response to DHA-PQP.\nThe best cut-off for the day 7 PQP concentration predicting any treatment failure was 30 ng/ml and children had a higher risk of having lower levels.\nSimilarly, in PNG, a trend toward a lower risk of treatment failure (PCR uncorrected) and plasma PQP levels at day 7 has been reported, suggesting that an increase of the weight-adjusted dosage in children may be required.\nIn our study, preliminary results on predictors of treatment failure seem to confirm the need of reviewing and possibly increasing the weight-adjusted dosage for children.\nPatients treated with DHA-PQP had a significantly higher rate of person-gametocyte-weeks compared with those having received AL.\nThis contrasts with a previous study in Papua, Indonesia, which showed no difference in gametocyte carriage between DHA-PQP and AL, but is in line with comparisons between DHA-PQP and mefloquine-artesunate, where a higher production of gametocytes in patients treated with DHA-PQP was observed.\nSuch an effect has been attributed to the lower dose of artemisinin derivative used in the DHA-PQP.\nGametocytaemia is a proxy measure of transmission potential and the increased gametocyte production related to DHA-PQP use may be a public health disadvantage that should be nevertheless balanced against a better PTP, particularly useful in areas of intense transmission.\nDihydroartemisinin-piperaquine was well tolerated, with few adverse events of clinical relevance.\nA higher frequency of abdominal pain and diarrhoea has previously been reported for DHA-PQP compared with mefloquine-artesunate but this disadvantage of DHA-PQP was not observed in this trial where the comparator treatment was AL.\nThe statistical significant difference in the QTc interval at day 2 was observed only when applying Bazett's but not Fridericia's correction.\nThis was not considered as clinical relevant because of the discrepant results obtained with the 2 methods and because both the proportions of patients with a QTc prolongation between day 0 and 2 higher than 60 ms or with an absolute QTc value greater than 500 ms were extremely low and balanced between groups.\nTherefore, considering that no cardiovascular AEs were reported, this study adds to the evidence that, at therapeutic doses, DHA-PQP and AL do not have any clinically significant cardiotoxicity.\nIt has also been previously reported that the only potentially serious adverse effects of artemisinin derivatives are rare type 1 hypersensitivity reactions.\nHowever, no evidence of moderate or severe adverse reactions of this kind was observed in the current study and this was despite the larger sample size compared with other published studies in which the number of patients recruited for each arm did not exceed a few hundred.\nIndeed, it is reassuring that no major safety problem has been observed in more than 1,000 children treated with DHA-PQP.\nNevertheless, such a sample size is unable to detect rare and unexpected serious adverse events and the development of a pharmacovigilance system should be a priority, not only for DHA-PQP but also for all other ACTs.\nAfrican countries should be encouraged, as the use of ACTs increases, to establish pharmacovigilance systems and drug developers and funding agencies should contribute to their development.\nIn conclusion, DHA-PQP is a safe, efficacious, tolerable and affordable new antimalarial treatment option in Africa.\nIts longer PTP period may be particularly useful in areas where transmission is intense, though it may exert an important drug pressure on the parasite populations, possibly selecting resistant strains.\nThe deployment of several ACTs as multiple first line treatments may overcome this problem.\nIndeed, assuming that different treatments are used in equal amounts in the host population, the use of multiple first line therapies would have two main benefits, i.e. the inability of the parasite to adapt to a variable environment and the reduced drug pressure as the rate at which a given treatment is used would be lower than if it was the only one available.\nDHA-PQP can definitely play an essential role in our effort to reduce the currently high malaria burden.\nTrial profile.\nPCR-corrected Adequate Clinical and Parasitological Response (ACPR) (ePP population) by country and by time point.\nKaplan Meier curve showing the cumulative proportion until day 42 of children with new infections (ITT population).\n\nBasic characteristics of the 5 African sites (online publication only).\n | Nanoro, Centre Muraz (Burkina Faso) | Kilifi, KEMRI (Kenya) | CISM, Manhi\u00e7a (Mozambique) | Epicentre, Mbarara (Uganda) | TDRC, Ndola (Zambia)\nCharacteristics of the area | Rural | Rural | Rural | Rural | Periurban\nMalaria endemicity | Mesoendemic | Mesoendemic | Mesoendemic | Mesoendemic | Mesoendemic\nSeasonality | High transmission between June and December | Perennial, with two peak seasons: Jul-Sep; Dec-Jan | Perennial with marked seasonality (Oct-April) | Perennial with two peaks: April and October | High transmission between November and May\nEntomological Inoculation rate (EIR) | 100 to 160 (2003) | 22 to 531 | 38 (2002)2 | Not available | Not available\nSite area under Demographic surveillance system (DSS) | No | Yes | Yes | No | No\nITNs coverage | <10% | Subsidised available | <10% | 11,4% | Approximately 30%\nFirst line treatment at the time of the study | Amodiaquine-artesunate or AL | SP, and then AL | Amodiaquine-SP | AL | AL\nDocumented resistance to chloroquine | 35% | 60% | 69%3 | 81%4 | 60%\nDates start patients' recruitment/end follow up | 16 Aug 2005/18 Jan 2006 | 22 Sep 2005/14 July 2006 | 14 Nov 2005/10 July 2006 | 17 Oct 2005/11 July 2006 | 16 Nov 2005/10 July 2006\n\nMbogo CM, Mwangangi JM, Nzovu J, Gu W, Yan G, Gunter JT, et al. Spatial and temporal heterogeneity of Anopheles mosquitoes and Plasmodium falciparum transmission along the Kenyan coast. Am J Trop Med Hyg. 2003 Jun; 68(6):734\u201342.\nAlonso PL, Sacarlal J, Aponte JJ, Leach A, Macete E, Milman J, et al. Efficacy of the RTS,S/AS02A vaccine against Plasmodium falciparum infection and disease in young African children: randomised controlled trial. Lancet. 2004 Oct 16; 364(9443):1411\u201320.\nAbacassamo F, Enosse S, Aponte JJ, Gomez-Olive FX, Quinto L, Mabunda S, et al. Efficacy of chloroquine, amodiaquine, sulphadoxine-pyrimethamine and combination therapy with artesunate in Mozambican children with uncomplicated malaria. Trop Med Int Health. 2004 Feb; 9(2):200\u20138.\nLegros D, Johnson K, Houpikian P, Makanga M, Kabakyenga JK, Talisuna AO, et al. Clinical efficacy of chloroquine or sulfadoxine-pyrimethamine in children under five from south-western Uganda with uncomplicated falciparum malaria. Trans R Soc Trop Med Hyg. 2002 Mar-Apr; 96(2):199\u2013201.\n\nDay 28 and Day 42 uncorrected ACPR (steps 1\u201311) and PCR-corrected ACPR (steps 1\u201316) in the different populations of analysis.\nStep | Event to be assessed | ePP | ITT\n1 | Withdrawal BEFORE OR AT D28: any reason except lost to follow-up (LFU) | Depending on reason, a patient can be: Excluded or Failure | Failure\n2 | Withdrawal BEFORE OR AT D28: LFU | Excluded | Failure\n3* | Withdrawal AFTER D28: any reason except LFU | Failure | Failure\n4* | Withdrawal AFTER D28: LFU | Failure | Failure\n5 | ETF, LCF, and LPF in accordance with the WHO criteria | Failure | Failure\n6** | Presence of major protocol violations | Excluded | No effect\n7** | Occurrence of adverse events highlighting recurrence of malaria | Failure | Failure\n8** | Presence of missing parasitaemia at two or more consecutive scheduled visits or presence of an isolated missing parasitaemia not preceded and followed by a negative parasitaemia | Failure | Failure\n9** | Administration of drugs with a known or suspected anti-malaria action as rescue treatment | Failure | Failure\n10** | Administration of drugs with a known or suspected anti-malaria action as non rescue treatment | Excluded | Failure\n11** | Administration of anti-malarial drugs for P. vivax, P. malariae, or P. ovale during the course of the study in patients not classified as ETF/LTF | Failure with new infection | Failure with new infection\n12 | PCR not done IN (DAY 4\u2013DAY 13) | Recrudescence | Recrudescence\n13 | PCR: non interpretable or missing or not done IN (DAY 14\u2013D28) | Excluded | Recrudescence\n14** | PCR: non interpretable or missing or not done AFTER D28 | Rule*** | Recrudescence\n15 | PCR\u200a=\u200anew infection or uncorrected ACPR\u200a=\u200aFailure with new infection | Success | Success\n16 | PCR\u200a=\u200arecrudescence | Recrudescence | Recrudescence\n\n*For the Day 42 endpoint.\n\n**All such cases were individually revised at the Blind Data Review meeting. Protocol violations were pre-defined.\n\n***Result \u201crecrudescence\u201d or \u201cnew infection\u201d was assigned according to the ratio between these outcomes in the patients with a valid PCR result at the corresponding time point and within each treatment group, separately considered.\n\n\nBaseline characteristics (ITT population).\nVariable | DHA-PQP (N\u200a=\u200a1038) | AL (N\u200a=\u200a510)\nGender M/F (%M/%F) | 525/513 (50.1/49.4) | 281/229 (55.1/44.9)\nAge in years (mean\u00b1SD) | 2.42\u00b11.14 | 2.43\u00b11.16\nWeight in kg (mean\u00b1SD) | 11.19\u00b12.55 | 11.28\u00b12.67\nFever (n (%)) | 624 (60.12) | 307 (60.20)\nTemperature in \u00b0C (mean\u00b1SD) | 37.88\u00b11.22 | 37.86\u00b11.18\nParasite density (geometric mean) | 24557 | 25884\nPresence of Gametocytes (n (%)) | 122 (11.75) | 66 (12.94)\nHb in g/L (mean\u00b1SD) | 89.23\u00b118.15 | 90.59\u00b118.20\nAnaemia (\u200a=\u200aHb<7 g/dL) (n (%)) | 141 (13.58) | 63 (12.35)\nLeucocytes in 10\u03029/L (mean\u00b1SD) | 9.62\u00b14.15 | 9.59\u00b13.94\nPlatelets in 10\u03029/L (mean\u00b1SD) | 182.84\u00b1108.70 | 181.59\u00b1106.74\nSplenomegaly (n (%)) | 41 (3.95) | 19 (3.73)\nHepatomegaly (n (%)) | 6 (0.58) | 3 (0.59)\nALAT in IU/L (mean\u00b1SD) | 34.08\u00b161.34 | 31.08\u00b136.23\nBilirubin in mg/dl (mean\u00b1SD) | 0.97\u00b11.04 | 0.94\u00b10.81\nCreatinine in U/L (mean\u00b1SD) | 40.96\u00b117.91 | 41.16\u00b119.17\n\n\nPCR-Corrected and Uncorrected Adequate Clinical and Parasitological Response (ACPR) by time point in ITT and ePP Population.\n | Day 28 | Day 42\n | DHA-PQP | AL | Lower Limit of one-sided 97.5% CI on difference | DHA-PQP | AL | Lower Limit of one-sided 97.5% CI on difference\nPCR-Corrected Cure Rate (n (%)) in ITT | 938 (90.37) | 459 (90.00) | \u22122.80 | 895 (86.22) | 442 (86.67) | \u22124.06\nUncorrected Cure Rate (n (%)) in ITT | 910 (87.67) | 391 (76.67) | 6.82 | 769 (74.08) | 330 (64.71) | 4.45\nTotal number of failures in ITT (PCR-uncorrected) | 128 (12.33) | 119 (23.33) |  | 269 (25.92) | 180 (35.29) | \nRecrudescences by PCR | 14 (1.35) | 11 (2.16) |  | 41 (3.95) | 17 (3.33) | \nRecrudescences due to informative withdrawals (including LFU) or failure before D14 (PCR not needed) | 65 (6.26) | 26 (5.10) |  | 65 (6.26) | 26 (5.10) | \nRecrudescences imputed (PCR missing, indet., not done) | 21 (2.02) | 14 (2.75) |  | 37 (3.56) | 25 (4.90) | \nNew Infection by PCR | 27 (2.60) | 64 (12.55) |  | 122 (11.75) | 105 (20.59) | \nNew Infection \u2260 from Plasmodium Falciparum | 1 (0.10) | 4 (0.78) |  | 4 (0.39) | 7 (1.37) | \nPCR-Corrected Cure Rate (n (%)) in ePP | 910 (94.69) | 442 (95.26) | \u22122.96 | 879 (91.47) | 436 (93.97) | \u22125.29\nUncorrected Cure Rate (n (%)) in ePP | 884 (91.99) | 376 (81.03) | 6.99 | 746 (77.63) | 319 (68.75) | 3.90\nTotal number of failures in ePP (PCR-uncorrected) | 77 (8.01) | 88 (18.97) |  | 215 (22.37) | 145 (31.25) | \nRecrudescences by PCR | 14 (1.46) | 11 (2.37) |  | 41 (4.27) | 16 (3.45) | \nRecrudescences due to informative withdrawals or failure before D14 (PCR not needed) | 37 (3.85) | 11 (2.37) |  | 37 (3.85) | 11 (2.37) | \nRecrudescences imputed (rule for missing PCR) | 0 | 0 |  | 4 (0.42) | 1 (0.22) | \nNew Infections imputed (rule for missing PCR) | 0 | 0 |  | 11 (1.14) | 8 (1.72) | \nNew Infection by PCR | 25 (2.60) | 62 (13.36) |  | 118 (12.28) | 102 (21.98) | \nNew Infection \u2260 from Plasmodium Falciparum | 1 (0.10) | 4 (0.86) |  | 4 (0.42) | 7 (1.51) | \n\nNote: In ITT, percentages are based on N\u200a=\u200a1038 (DHA-PQP) and N\u200a=\u200a510 (AL); in ePP, percentages are based on N\u200a=\u200a961 (DHA-PQP) and N\u200a=\u200a464 (AL).\n\nSummary of adverse events (ITT population).\nSafety/ITT Population | DHA-PQP (N\u200a=\u200a1038) | AL (N\u200a=\u200a510) | p-value\nAt least one AE (n,%) | 823 (79.29%) | 411 (80.59%) | 0.550\nNeutropenia | 18 (1.73%) | 12 (2.35%) | \nVomiting | 71 (6.84%) | 35 (6.86%) | \nGastrointestinal disorders (including vomiting) | 207 (19.94%) | 92 (18.04%) | \nSkin and subcutaneous tissue disorders | 70 (6.74) | 29 (5.69%) | \nAlanine aminotransferase increased | 20 (1.93%) | 19 (3.73%) | \nElectrocardiogram QT prolonged | 26 (2.50%) | 13 (2.55%) | \nAt least one related AE (n,%) | 737 (71.00%) | 368 (72.16%) | 0.637\nAt least one SAE (n,%) | 18 (1.73%) | 5 (0.98%) | 0.249\nAt least one related SAE (n,%) | 15 (1.45%) | 4 (0.78%) | 0.332\nAt least one AE which caused discontinuation (n,%) | 5 (0.48%) | 0 | 0.178\nAt Least one SAE which caused death (n,%) | 1 (0.10%) | 1 (0.20%) | 0.551\n\nAE\u200a=\u200aAdverse event; SAE\u200a=\u200aSerious adverse event; Related SAE\u200a=\u200aSerious adverse event for which the investigator classifies the relationship to the study drug as unlikely, possible, probable, definitely related or whose classification is missing.",
        "label": "low",
        "id": "task4_RLD_test_533"
    },
    {
        "paper_doi": "10.1371/journal.pone.0011880",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Trial design: An open-label (non-inferiority) RCTFollow-up: Particiopants were managed as outpatients unless local practice dictated otherwise (some centres used hospital stays of between 3 and 28 days). Outpatients were asked to return on days 1, 2, 3, 7, 14, 21, 28, 35, 42, 49, 56, and 63, and any time they felt unwell. Blood smears were performed at each visit.Adverse event monitoring: Blood and urine samples were taken for analysis on days 0, 28, 63 (if abnormal on day 28) and on the day of any recurrent parasitaemia. Twelve-lead ECGs were performed at days 0, 2, 7, 28 (if abnormal on day 7), 63 and on the day of any recurrent parasitaemia.\n\n\nParticipants: Number of participants: 1150Inclusion criteria: Age 3 months to 65 years (>=18 years in India), P. falciparum mono-infection (80 to 200,000 parasites/uL) or mixed infection, weight >= 5 kg, fever (>= 37.5 degC) or history of fever, informed consent.Exclusion criteria: Severe malaria, treatment with MQ in the 60 days before screening, treatment with DHA-P in the 3 months before screening, > 4% parasitised red blood cells, pregnancy or lactation.\n\n\nInterventions: 1. DHA-P, fixed dose combination, adult tablets 40 mg/320 mg, child tablets 20 mg/160 mg (Eurartesim(r): Sigma Tau)One dose daily for 3 days2.25 mg/kg DHA and 18 mg/kg piperaquine per doseDose rounded up to the nearest half tablet2. Artesunate plus mefloquine, loose dose combination, AS 50mg tablets, MQ 250 mg tablets (Mepha Ltd)AS 4mg/kg once daily for 3 daysMQ none on day 0, then 15 mg/kg once on day 1 and 10 mg/kg once on day 2All doses supervised.\n\n\nOutcomes: Cure rate at days 28, 42, and 63, PCR corrected and uncorrectedMean change in Hb day 0 to day 63Gametocyte carriagePerson-gametocyte-weeksAdverse eventsNot included in this review:Fever clearanceParasite clearance\n\n\nNotes: Country: Thailand (six sites), Laos (two centres), and India (three centres).Setting: Hospitals and research units.Transmission: Varied across trial regions. Trial regions in Thailand had unstable, low and seasonal malaria transmission; trial regions in Laos had seasonal transmission with a peak just after the heavy rainy months of July to August; trial regions in India included areas with perennial transmission, perennial transmission with a seasonal peak from June to September, and transmission active in post monsoon months.Resistance: All sites had notable CQ resistance (estimates of 28 day treatment failure at the Indian sites ranged from 32% to 67% between 2002 and 2007). The Thai sites also had multi-drug resistant P. falciparum.Dates: Jun 2005 to Feb 2007.Funding: Medicines for Malaria Venture, Sigma Tau, and Oxford University\n\n",
        "objective": "To evaluate the effectiveness and safety of DHA\u2010P compared to other ACTs for treating uncomplicated P. falciparum malaria in adults and children.",
        "full_paper": "Background\nThe artemisinin-based combination treatment (ACT) of dihydroartemisinin (DHA) and piperaquine (PQP) is a promising novel anti-malarial drug effective against multi-drug resistant falciparum malaria.\nThe aim of this study was to show non-inferiority of DHA/PQP vs. artesunate-mefloquine (AS+MQ) in Asia.\nMethods and Findings\nThis was an open-label, randomised, non-inferiority, 63-day follow-up study conducted in Thailand, Laos and India.\nPatients aged 3 months to 65 years with Plasmodium falciparum mono-infection or mixed infection were randomised with an allocation ratio of 2\u22361 to a fixed-dose DHA/PQP combination tablet (adults: 40 mg/160 mg; children: 20 mg/320 mg; n\u200a=\u200a769) or loose combination of AS+MQ (AS: 50 mg, MQ: 250 mg; n\u200a=\u200a381).\nThe cumulative doses of study treatment over the 3 days were of about 6.75 mg/kg of DHA and 54 mg/kg of PQP and about 12 mg/kg of AS and 25 mg/kg of MQ.\nDoses were rounded up to the nearest half tablet.\nThe primary endpoint was day-63 polymerase chain reaction (PCR) genotype-corrected cure rate.\nResults were 87.9% for DHA/PQP and 86.6% for AS+MQ in the intention-to-treat (ITT; 97.5% one-sided confidence interval, CI: >\u22122.87%), and 98.7% and 97.0%, respectively, in the per protocol population (97.5% CI: >\u22120.39%).\nNo country effect was observed.\nKaplan-Meier estimates of proportions of patients with new infections on day 63 (secondary endpoint) were significantly lower for DHA/PQP than AS+MQ: 22.7% versus 30.3% (p\u200a=\u200a0.0042; ITT).\nOverall gametocyte prevalence (days 7 to 63; secondary endpoint), measured as person-gametocyte-weeks, was significantly higher for DHA/PQP than AS+MQ (10.15% versus 4.88%; p\u200a=\u200a0.003; ITT).\nFifteen serious adverse events were reported, 12 (1.6%) in DHA/PQP and three (0.8%) in AS+MQ, among which six (0.8%) were considered related to DHA/PQP and three (0.8%) to AS+MQ.\nConclusions\nDHA/PQP was a highly efficacious drug for P. falciparum malaria in areas where multidrug parasites are prevalent.\nThe DHA/PQP combination can play an important role in the first-line treatment of uncomplicated falciparum malaria.\nTrial Registration\nControlled-Trials.com ISRCTN81306618\nIntroduction\nThe bisquinoline piperaquine (PQP) was first synthesised in the 1960s independently by teams in China and France.\nIn 1978, because of its greater potency and tolerability relative to chloroquine, PQP replaced chloroquine in the Chinese National Malaria Control Programme.\nSuch was the success of the programme, that the following decade saw the use of PQP decrease because of the decrease in patients suffering from malaria.\nIn 1990, Chinese scientists working to develop alternative anti-malarial treatments that might offer higher cure rates with better tolerability profiles included PQP phosphate in an artemisinin-based combination treatment (ACT), known as China-Vietnam 4 (CV4\u00ae), that also included dihydroartemisinin (DHA), trimethoprim and primaquine phosphate.\nDuring the development programme, doses of each component of the combination were modified resulting in a new formulation, CV8\u00ae, which was tested in Vietnam and administered in the Vietnamese Malaria Control Programme in 2000.\nThis programme was widely successful despite the presence of chloroquine resistance in Vietnam.\nHowever, concerns about the association of red cell haemolysis with primaquine in populations with glucose-6-phosphate dehydrogenase deficiency and the questionable anti-malarial potency of trimethoprim led to the removal of these two components of the drug combination in CV8\u00ae.\nThe remaining two components of the regimen, the artemisinin DHA and PQP (Artekin\u00ae), provide a combination that is relatively inexpensive and has been shown to be effective both in curing malaria and preventing re-infection.\nSince their introduction in the 1990s, ACTs have been found to be highly effective treatments for malaria.\nAcross Asia, Africa and South America, clinical and parasitological responses to the combination of DHA and PQP have generally exceeded the 95% value that the World Health Organization (WHO) recommend for anti-malarial treatments.\nHowever, in recent years there have been suggestions that the overall efficacy of ACTs in Thailand and Cambodia may be declining.\nEvidence of falling efficacy has been characterised by reductions in the proportions of patients clearing their parasitaemia by day 2 of treatment in the Thailand-Myanmar border region, and by reductions in polymerase chain reaction (PCR)-corrected cure rates at 28- and 42-day follow-up assessments in the Cambodia-Thailand border region.\nRelatively low cure rates were also reported from Papua New Guinea and, most recently, there has been evidence of reduced artesunate susceptibility in Western Cambodia.\nHistorically, anti-malarial drug resistance has spread westwards from Cambodia through South Asia to Africa.\nConsequently, the recent reports of potential resistance to artemisinins alone and ACTs are of great concern.\nIn order to assess the safety and efficacy of the treatment combination of DHA and PQP in Asia, we conducted a randomised trial in Thailand, Laos and India comparing DHA/PQP with another ACT, artesunate (AS) plus mefloquine (MQ).\nAll study sites were located in areas of notable chloroquine resistance.\nDihydroartemisinin plus PQP was administered as a single tablet (DHA/PQP) and AS plus MQ were administered as separate loose tablets (AS+MQ).\nThis study had the largest sample size to date of any study assessing DHA/PQP in South East Asia.\nMethods\nThe protocol and amendments for this trial and supporting CONSORT checklist are available as supporting information; see Protocol S1 to S8 and Checklist S1.\nStudy regions\nThis study was conducted in six centres in Thailand (Hospital for Tropical Diseases, Faculty of Tropical Medicine, Mahidol University, Bangkok; Suanphung Hospital, Ratchaburi; Proppra Hospital, Proppra District, Tak; Shoklo Malaria Research Unit, Mae Sod District, Tak; Mae Sod Hospital, Muang District, Tak; and Mae Ramat Hospital, Mae Ramat District, Tak), two centres in Laos (Phalanxay District Hospital, Savannakhet Province; Xepon District Hospital, Savannakhet Province) and three centres in India (Down Town Hospital, Guwahati, Assam; Goa Medical College and Hospital, Goa; and Wenlock District Government Hospital, Mangalore) over two malaria seasons, with patients screened from June 2005 to February 2007.\nThe study regions provided a range of transmission conditions and standard treatments.\nStudy regions in Thailand were located in malarious forest on the Thai-Myanmar border in regions of unstable, low and seasonal malaria transmission that used AS+MQ as a first-line treatment.\nNearly all infections with Plasmodium falciparum and P. vivax in the region are symptomatic and P. falciparum is multi-drug resistant, with high levels of resistance to chloroquine.\nMalaria transmission in the Phalanxay and Xepon districts in Laos was seasonal with a peak just after the heavy rainy months of July to August.\nPrevious studies in the Phalanxay district showed high levels of chloroquine resistance.\nThe first-line treatment was artemether/lumefantrine.\nIn India, study regions included areas of perennial transmission (Assam), perennial transmission with a seasonal peak from June to September (Goa), and transmission active in post-monsoon months (Mangalore).\nChloroquine resistance was present at all study sites, with site records showing 28-day treatment failure rates of 32% in Assam (in 2002), 54% in Goa (in 2007) and 67% in Mangalore City (in 2005).\nArtemisinin-based combination treatments are used as standard treatments in Goa, Mangalore and Assam (specifically AS plus sulphadoxine/pyrimethamine).\nPatients\nMales and females aged from 3 months to 65 years weighing at least 5 kg with fever or history of fever (\u226537.5\u00b0C) and microscopically confirmed mono-infection with P. falciparum (asexual forms parasitaemia \u226580 per \u00b5L \u2264200,000 per \u00b5L) or mixed infection were eligible for the study.\nLocal regulations were followed in India that required all recruited patients to be aged 18 years and over.\nKey exclusion criteria were: severe malaria, treatment with MQ in the 60 days prior to screening, treatment with DHA/PQP in the 3 months prior to screening and >4% parasitised red blood cells.\nPregnant or lactating women were not eligible for the study.\nStudy design\nThis was a randomised, Phase III, open-label study with two treatment arms: DHA/PQP and the active comparator AS+MQ.\nAt the time the study was designed, there was no single first line treatment in the countries involved in the study.\nHowever, AS+MQ was first line treatment in Thailand, was well characterized and was widely used in all the considered countries.\nThis made AS+MQ suitable as the control treatment against which the efficacy and safety of DHA/PQP would be tested.\nAs the study was not blinded, to limit bias, the following procedures were put in place: 1) Randomisation was conducted under blinded conditions: the blind to the investigator and patient in the randomisation process was maintained by the use of sealed envelopes.\n2) Evaluation of the PCR test results was blinded (centralised at the Prince Leopold Institute of Tropical Medicine, Antwerp, Belgium, with quality control at Shoklo Malaria Research Unit, Mae Sod, Thailand).\n3) The chairman and the statistician of the independent Data Monitoring Committee reviewed the most relevant safety data and participated in the final blinded data review meeting where all decisions about assessment of the primary outcome and patient allocation to the pre-defined populations were made under blinded conditions.\nThe randomisation list was generated by an external contract research organisation (MDS Pharma Services) using the plan procedure of SAS (Cary, NC, USA).\nPatients were allocated to receive either DHA/PQP or AS+MQ following a 2\u22361 randomisation schedule ratio.\nThe unbalanced ratio was chosen to increase the chance of detecting rare adverse reactions and to provide more precise estimates of cure rates in the DHA/PQP arm.\nDihydroartemisinin/PQP\n(Eurartesim\u2122, Sigma-Tau, Italy) was given once daily, on days 0, 1 and 2 of the study, at the standard dosage of 2.25 mg/kg and 18 mg/kg per dose of DHA and PQP, respectively, rounded up to the nearest half tablet.\nTo facilitate the correct dosing of DHA/PQP, two formulations were used (DHA 20 mg + PQP 160 mg and DHA 40 mg + PQP 320 mg).\nOver the 3 days, the cumulative doses were of about 6.75 mg/kg of DHA and 54 mg/kg of PQP.\nArtesunate and MQ (Mepha Ltd, Switzerland) were administered as separate tablets containing AS 50 mg and MQ 250 mg; AS was administered at 24-h intervals on days 0, 1 and 2 with a daily dose of 4 mg/kg and MQ was administered at 15 mg/kg on day 1 and 10 mg/kg on day 2 at 24-h intervals, but was not administered on day 0.\nAs part of their routine treatment, febrile patients who attended the study centres underwent a thick blood smear test for malaria before any anti-malarial treatment was administered.\nIf the smear was found to be positive for P. falciparum, patients were told about the study and offered the opportunity to participate.\nEligible patients who agreed to participate were given detailed explanations of the trial from study staff, including the information that they would be given one of two different anti-malaria treatments on a random basis.\nParticipating patients all provided written informed consent.\nPatients who declined to participate were provided with treatment for uncomplicated P. falciparum malaria, standard for the area in which they lived.\nParticipating patients were managed as outpatients and doses were given under medical supervision, although at some centres patients could be hospitalised according to local practice, which ranged from 3 days at Mae Ramat Hospital, Thailand and at centres in Laos and India to 28 days at Mahidol University, Thailand.\nIf not in hospital, patients were encouraged to return to the study centre for assessments on days 1, 2, 3, 7, 14, 21, 28, 35, 42, 49, 56 and 63, and they were also told that they could make unscheduled visits on any day on which they felt unwell.\nPatients were followed-up for 63 days.\nIn accordance with WHO (WHO 2003 and 2009) and standard practice in malaria clinical trials, the primary endpoint of the study was the PCR-corrected cure rate, based on the adequate clinical and parasitological response (ACPR), which was defined as the absence of parasitaemia irrespective of the patient's body temperature, with the patient not meeting any of the pre-defined criteria of early treatment failure or late clinical or parasitological failure (see below).\nIn more detail, the primary endpoint was defined using two methods.\nThe first method was based purely on the standard definitions of early/late clinical and parasitological failure as defined by the WHO.\nThis endpoint is referred to as the true treatment failure and is defined as the sum of early treatment failures and late recrudescences, which included late treatment failures that were assessed as recrudescences according to PCR analysis (100 minus the true treatment failure rate provides the WHO cure rate).\nThe second method, agreed with the Data Monitoring and the Clinical Development Committees, was based on a pre-defined procedure that expanded the WHO definitions with a set of rules allowing the evaluation of each individually randomised patient (for definitions see Table 1).\nThis approach was considered to be primary because it was deemed to be in line with the requirements of the most stringent regulatory authorities.\nAll cases not strictly matching the WHO definitions and/or the described procedure were reviewed individually at the final blinded data review meeting.\nDay 63 was chosen as the primary time-point because of the long half lives of piperaquine and mefloquine and because the risk of new infection is lower than in other parts of the world.\nSecondary endpoints included PCR-corrected adequate clinical and parasitological response on days 28 and 42, PCR-uncorrected adequate clinical and parasitological response (PCR-uncorrected), proportion of patients with early and late treatment failure, proportion of aparasitaemic patients, proportion of afebrile patients, number of new infections, gametocyte carriage and the safety profile of the two treatments including adverse events and 12-lead electrocardiograph parameters (QT interval corrected according to the methods of Bazett, QTc(B), and Fridericia, QTc(F)).\nProcedures\nThe presence of P. falciparum was verified at all study visits including screening, using thick and thin Giemsa-stained blood smears obtained from the patient to calculate parasite density, which was initially calculated by counting the number of asexual parasites per 500 leukocytes in the thick blood film, based on an assumed white cell count of 8,000 cells per \u00b5L.\nBlood smears were obtained from a finger prick applied directly to a microscope slide to create the blood film.\nParasite density per \u00b5L was calculated as: (number of parasites counted \u00d78,000)/(number of leukocytes counted).\nFor samples with higher levels of parasitaemia (>3 parasites/1000 red blood cells), parasite density was calculated from the thin film per 1000 red blood cells as: (number of P. falciparum trophozoites per 1000 red blood cells x haematocrit \u00d7125.6).\nGametocyte prevalence was also evaluated.\nBoth the thick and thin blood smear readings were done locally following the above described procedure, while the gametocyte assessments were carried out in accordance with standard practice at each individual site.\nAll technicians who read the slides had undergone appropriate training in malaria-related microscopy and had at least 5 years experience in reading blood smears.\nA process of quality control was used to monitor the values being provided by the local laboratories.\nOne in five of every samples was sent to a central, independent laboratory (Department of Parasitology and Medical Entomology, Muhimbili University of Health and Allied Sciences, Dar Es Salaam, Tanzania) for review.\nThree spots of blood were collected on 3MM filter paper (Whatman, UK) at the enrolment visit and at any visit after day 7 for PCR analysis.\nFilter papers were dried and individually stored in a plastic bag containing silica gel.\nAll filter papers were subsequently transferred to the Institute of Tropical Medicine (Antwerp, Belgium) where centralised genotyping was conducted; deoxyribonucleic acid was purified as described elsewhere.\nThree polymorphic genetic markers, MSP1, MSP2 and GluRP were used to distinguish recrudescence from new infections.\nRecrudescence was defined as at least one identical allele for each of the three markers in the pre-treatment and post-treatment samples.\nNew infections were diagnosed when all alleles for at least one of the markers differed between the two samples.\nAn independent expert read all gels under blinded conditions (National Museum of Natural History, Paris, France).\nFor quality control, 20% of the filter papers were re-analysed and read by an independent laboratory (Shoklo Malaria Research Unit, Mae Sot District, Tak, Thailand).\nTwelve-lead electrocardiograms were recorded at days 0, 2, 7, 28 (if abnormal on day 7), 63 and on the day of any recurrent parasitaemia that occurred using the CarTouch device with CarTouch version 1.4.1 software (Cardionics SA, Brussels, Belgium).\nThe electrocardiogram was viewed during recording and then transmitted via modem to MDS Pharma Services Central Telemedicine Department (Paris, France) for interpretation and reporting.\nThe data were analysed using both Bazett's (QTc(B)) and Fridericia's (QTc(F)) correction methods.\nElectrocardiogram results were tabulated as normal, borderline or prolonged according to gender and age normal ranges (adult males and children (1\u201312 years): normal: <430 msec; borderline: 430\u2013450 msec; prolonged: >450 msec.\nAdult females: normal: <450 msec; borderline: 450\u2013470 msec; prolonged: >470 msec).\nBlood samples were taken for standard laboratory assessments of haematology, biochemistry and urinalysis on days 0, 28, 63 (if abnormal on day 28) and on the day of any recurrent parasitaemia.\nEthical approval and informed consent\nStudy staff conducted all interviews with patients, and children's parents, in their native language and they explained their rights under International Conference on Harmonisation-Good Clinical Practice (ICH-GCP).\nWhen obtaining informed consent from patients the signature of a witness was obtained if patients were unable to write.\nConsideration was given to the ethical implications of the randomisation ratio assigning more patients to receive DHA/PQP than AS+MQ during the design of the study.\nAs the safety and tolerability profile of DHA and PQP has been well characterised at the doses proposed for the present study, it was considered acceptable to adopt this approach in the present study that would result in more patients receiving DHA/PQP.\nThe study protocol was approved by the following ethics committees: Institutional Ethics Committee, National Institute of Malaria Research (ICMR), Delhi, India; Goa Medical College and Hospitals Local Ethics Committee, Goa, India; Institutional Ethics Committee, Kasturba Medical College, Mangalore, India; Laos National Ethics Committee for Health Research (NECHR), National Institute of Public Health, Vientiane, Laos; Oxford Tropical Research Ethics Committee (OXTREC), University of Oxford, Headington, United Kingdom; The Ethical Review Committee for Research in Human Subjects, Ministry of Public Health, Nontaburi, Thailand; Tropical Medicine Ethics Committee (TMEC), Mahidol University, Rachathewi District, Bangkok, Thailand.\nThe Food and Drug Department, Government of Laos PDR, approved the use of DHA/PQP in that country.\nThe trial was conducted under the provisions of the Declaration of Helsinki (1964 and its subsequent amendments up to 2002) and in accordance with ICH-GCP.\nA Study Steering Committee, a Data Monitoring Committee and a Clinical Development Committee were created prior to the beginning of the trial, and worked independently to harmonise and monitor the study.\nThe trial was registered prior to the enrolment of the first patient in the International Standard Randomised Controlled Trials Register, number ISRCTN81306618, at http://www.controlled-trials.com/isrctn/trial/l/0/81306618.html.\nStatistical methods\nThe statistical analysis was conducted according to a pre-defined data analysis plan.\nThree populations were prospectively planned.\nThe intention-to-treat (ITT) population was defined as all randomised patients who took at least one dose of study treatment.\nThe per protocol population was defined as all randomised patients who were eligible according to the main (pre-defined) protocol inclusion and exclusion criteria, received at least 80% of the study medication, underwent the day 63 assessment, took no other anti-malarial drugs (excluding rescue therapies) and, in the presence of asexual parasite stages on thick or thin blood smears, had an evaluable PCR test.\nPatients who missed visits from day 0 to day 2 were excluded from the per protocol population.\nThe third population, referred to as modified ITT, was midway between the two described populations.\nAlthough this was pre-defined as the co-primary population (together with the per protocol), results in this paper are presented only for the two most extreme populations, i.e., ITT and per protocol populations, because these are standard and the findings were very similar in all populations analysed.\nThe analysis of the PCR-corrected and uncorrected ACPR (as defined in Table 1) was based on simple proportions and 97.5% (one-sided) confidence intervals (CIs) computed on the difference between these proportions of the test and reference treatments.\nIf the lower limit of this CI was greater than \u22125% DHA/PQP could be considered non-inferior to AS+MQ.\nThe primary time point was day 63 but this analysis was repeated also at days 42 and 28.\nVarious sensitivity analyses were carried-out on the primary endpoint for verifying the robustness of findings towards different assumptions.\nThese included an analysis where all patients with missing parasitaemia were treated as failures, an analysis where patients with new infections as detected by PCR were excluded and an analysis on two enlarged per protocol populations where all patients/all early failures \u201cnot receiving at least 80% of study treatment\u201d or \u201cfailing to attend a visit at days 0\u20132\u201d were not excluded.\nThe true treatment failures defined in accordance with the WHO handbook were analysed by means of the survival analysis (Kaplan-Meier estimates).\nAll sources of uncertainty (i.e. withdrawals, new infections, patients with PCR results either not available or indeterminate) were censored.\nSurvival analysis techniques were also applied to the analysis of time to parasite clearance and the estimation of the rate of new infections (in the latter analysis recrudescent infections were censored).\nFor descriptive purposes, the proportions of early, late and true treatment failures were also computed as simple rates for each treatment arm together with the relevant CIs.\nBy-country heterogeneity in cure rates was assessed by the Breslow-Day test or by logistic regression, when the former was not applicable.\nThe study was conducted over two malaria seasons with different centres working in the two study periods.\nTo evaluate how the 63-day PCR-corrected cure rates varied across the cohorts from the two seasons, a logistic regression model was fitted with cohort and treatment as the explanatory variables.\nThe treatment by cohort interaction was evaluated as a candidate to enter this model through a residual score test.\nAll tests for heterogeneity and interaction were evaluated at the 10% significance level.\nRates of person-fever-days and person-gametocyte-weeks were calculated as the number of weeks in which fever was present or blood slides were positive for gametocytes, respectively, divided by the number of follow-up weeks and expressed per 100 person-weeks.\nThe safety population, which coincides with the ITT population, was used for all safety assessments.\nAdverse events were coded using the MedDRA dictionary (MedDRA V8.1).\nProportions of patients experiencing at least one adverse event were compared between treatments using the Pearson Chi square test.\nTo determine the sample size of this study, the PCR-corrected cure rate at day 63 was estimated to be at least 92% for AS+MQ in the ITT population using a literature search.\nExpert opinion was used to define the non-inferiority margin, which was set at 5%.\nThe planned sample size of 700 in the DHA/PQP arm and 350 in the AS+MQ arm (1050 patients in total) provided 80% power to show non-inferiority with a non-inferiority margin of \u22125% (test minus reference) and a one-sided \u03b1 level of 2.5%.\nThe rate of patient attrition in the per protocol population was expected to be 20% compared with the ITT population, but the PCR-corrected cure rate was expected to be higher than in the ITT population (95% in the per protocol population), therefore the projected power of 80% was maintained also for the analysis on the per protocol population.\nWhen India was included in the study to increase the speed of recruitment, the total sample size was increased by 150 patients to ensure that 100 Indian patients were treated with DHA/PQP, in accordance with Indian requirements.\nConsequently, the power of the study exceeded 80%.\nResults\nPatient disposition and demographic characteristics\nOne thousand two hundred and thirty-nine patients were screened, with 769 patients randomised to DHA/PQP and 381 patients randomised to AS+MQ.\nFigure 1 illustrates the number of patients completing the study and included in the different study populations.\nSixty-one percent of patients were recruited in Thailand, 26% in Laos and 13% in India (Table 2).\nThere were no notable differences in demographic parameters between treatment arms overall or by country (Table 2).\nChildren \u22645 years of age comprised approximately 8% of the study population, with most of the study population over 18 years of age (approximately 75%).\nMedian [range] doses received by patients of the ITT population are presented by age class in Table 3.\nOne hundred and forty-four patients were excluded from the per protocol population (Table 2).\nThere were no notable differences between the treatment arms in the number of patients excluded from the per protocol population, the most frequent reasons for exclusion were lost to follow-up before or at day 63 and PCR missing or indeterminate before or at day 63 (DHA/PQP: 35 patients: AS+MQ: 19 patients).\nPrimary Endpoint\nThe analysis of the PCR-corrected cure rate at day 63 confirmed that DHA/PQP was non-inferior to AS+MQ.\nFor the ITT population, PCR-corrected cure rates were 87.9% for DHA/PQP and 86.6% for AS+MQ (97.5% CI: >\u22122.87%; Table 4).\nAs expected, better absolute but comparatively similar results were obtained for the per protocol population with PCR-corrected cure rates of 98.7% for DHA/PQP and 97.0% for AS+MQ (97.5% CI: >\u22120.39%).\nSensitivity analyses confirmed these results, with similar treatment differences between cure rates and relevant CIs compared with those described above.\nSecondary endpoints\nNon-inferiority was also proved for the uncorrected cure rates at day 63: the cure rates were 67.3% in the DHA/PQP arm and 59.6% in the AS+MQ arm (ITT population) with a CI >1.75%.\nThis treatment difference was also statistically significant (Chi-squared test: p\u200a=\u200a0.010; Table 4).\nSimilar results were obtained for the per protocol population (Table 4).\nSignificantly more patients who received AS+MQ experienced new infections compared with those receiving DHA/PQP.\nKaplan-Meier estimates for the proportion of patients with new infection at day 63 were 22.7% for DHA/PQP and 30.3% for AS+MQ (ITT population; p\u200a=\u200a0.0042).\nSimilar results were obtained for the other populations.\nNon-inferiority of DHA/PQP versus AS+MQ was also shown at days 28 and 42 in all study populations (Table 4).\nOn days 28 and 42, in the per protocol population, the PCR-corrected cure rates defined in Table 1 were significantly greater for DHA/PQP than AS+MQ (p\u200a=\u200a0.001 and p\u200a=\u200a0.031, respectively) while these differences were not statistically significant in the ITT population (Table 4).\nA logistic regression test to assess by-country heterogeneity in the PCR-corrected cure rates (evaluated according to Table 1) showed no significant differences between countries (ITT: p\u200a=\u200a0.688; per protocol: p\u200a=\u200a0.988;).\nSimilar results were obtained for a Breslow-Day test conducted to assess by-country heterogeneity in the uncorrected cure rate (ITT: p\u200a=\u200a0.896; per protocol: p\u200a=\u200a0.728).\nThe 95% (two-sided) CIs at each individual country level for the PCR-corrected cure rate at day 63 for the per protocol population are shown in Figure 2A.\nNo effect due to the cohorts from the two seasons (per protocol: p\u200a=\u200a0.760) was shown in this study (Figure 2B shows the relevant CIs) and no difference was observed by age class (per protocol: p\u200a=\u200a0.998).\nClassification by age was: \u22642 years; 2\u201312 (included) years; 12\u201318 (included) years and >18 years (relevant CIs are shown in Figure 2C).\nResults in the other populations for country, cohort (season) of enrolment and age groups were similar.\nWhen the PCR-corrected cure rates were assessed and analysed according to the WHO recommendations, i.e. as 100 minus the true treatment failure rate, the results were similar to those in the PCR-corrected cure rate as defined in Table 1, although cure rates were always slightly higher.\nThe Kaplan-Meier estimates of the cure rates were 97.6% (DHA/PQP) versus 96.5% (AS+MQ) in the ITT population and 98.2% (DHA/PQP) versus 96.8% (AS+MQ) in the per protocol population.\nThe simple rates of early, late and true treatment failures on days 63, 42 and 28 are shown in Table 5 together with the relevant CIs.\nKaplan-Meier estimates of median time to parasite clearance were 2 days for each treatment (ITT population), while the Kaplan-Meier estimate of the rate of aparasitaemic patients at day 3 (i.e. 24 h after completion of study treatment) was 97.6% for DHA/PQP and 97.6% for AS+MQ.\nSimilar results were obtained for the per protocol population.\nProportions of afebrile patients showed very similar profiles to aparasitaemic patients, with profiles being virtually superimposable.\nFever incidence measured in person-fever-days (per 100 person-days) was similar for DHA/PQP and AS+MQ: 1065/5315 (20.04%) versus 524/2636 (19.88%), p\u200a=\u200a0.929 (Table 6; ITT population).\nThe overall gametocyte prevalence throughout the study (from day 7 to day 63), measured as person-gametocyte-week rates, was significantly higher in the DHA/PQP arm in all populations (DHA/PQP: 76/749, 10.15%; AS+MQ: 18/369, 4.88%; p\u200a=\u200a0.003; Table 6, ITT population).\nHowever, the treatment difference for prevalence was not statistically significant from day 35 onwards (Table 6; ITT population).\nThere were no deaths during the study.\nIn the DHA/PQP group, there were 12 (1.56%) events judged by the investigator to be serious, which occurred in 12 patients.\nSix (0.78%) of these events were considered by the investigators to be related to DHA/PQP.\nThese were two cases of anaemia, one from day 7 to day 90, the other from day 7 to day 35; one viral infection (possibly Dengue fever) from day 15, fully recovered; and one Wolf Parkinson White (WPW) syndrome from day 2 to day 90.\nWolf Parkinson\nWhite syndrome is a congenital defect of accessory conduction pathways.\nElectrocardiograms for this patient were submitted to two cardiologists to provide expert opinions.\nBoth considered that the failure to diagnose WPW at baseline could be attributed to the increased heart rate caused by the patients' fever which hid the electrocardiographic characteristics of WPW.\nOther events considered serious and related were one convulsion on day 0, and one encephalitis on day 45 which resulted in a left-sided hemiplegia.\nThe patient was diagnosed with P. falciparum malaria on day 48, which was treated with i.v. artesunate and oral mefloquine.\nThe other six (0.8%) serious events were judged unrelated to DHA/PQP by the investigator: two cases of pyelonephritis, one case of aspiration pneumonia, and three cases of P. falciparum malaria.\nIn the AS+MQ group three (0.8%) events in three patients were judged to be serious and all were judged related to study treatment: one case of anaemia from day 8, one case of convulsion on day 1, and one case of encephalitis from day 16 to day 31.\nAdverse event profiles for DHA/PQP and AS+MQ were very similar in terms of type and frequency of events and were consistent with those expected in adult patients with acute malaria.\nMost patients in the study experienced adverse events: 69.4% (532/767) of patients in the DHA/PQP arm and 72.4% (276/381) of patients in the AS+MQ arm.\nThere was no statistically significant difference between the treatments in the incidence of adverse events (p\u200a=\u200a0.282, Chi-square test).\nThe most frequently reported adverse events were malaria symptoms, with headache the most commonly reported (Table 7).\nThe frequencies of individual adverse events were generally similar between treatments, although the frequencies of nausea, vomiting and dizziness appeared to be higher in the AS+MQ arm (Table 7).\nApproximately 3% of patients in each arm experienced at least one adverse event related to skin or subcutaneous tissue; the most frequent was pruritus (DHA/PQP: 17 patients [2.2%]; AS+MQ: 8 patients [2.1%]).\nOne patient in each arm experienced allergic dermatitis.\nAs would be expected in patients with malaria, anaemia and thrombocytopenia were common in each arm, approximate proportions were 30% for low red blood cells (normal range for >15 years of age: 3.5\u22125.5\u00d71012/L for women and 4\u22125.5\u00d71012/L for men), 50% for low haemoglobin (normal range for >15 years of age: 110\u2212150 g/L for women and 120\u2212160 g/L for men), and 67% for low platelets (normal range for all ages: 100\u2212300\u00d7109/L).\nAt the end of the study following treatment, these proportions had decreased to approximately, 20%, 40% and 17%, respectively.\nThere was no apparent difference between treatments.\nIncreases in mean haemoglobin levels were observed over the 63-day follow-up period.\nIn the ITT population, mean (standard deviation) haemoglobin values on day 0 were 118.7 (24.4) g/L for DHA/PQP and 119.8 (23.4) g/L for AS+MQ.\nAt day 63, mean (standard deviation) changes from day 0 were 12.8 (22.2) g/L for DHA/PQP and 14.21 (21.2) g/L for AS+MQ.\nSimilar results were observed for the per protocol population.\nOther than elevated liver parameters, as might be expected in this population, there were no relevant changes in biochemistry parameters.\nAt baseline, using the QTc(B) method, there was a statistically significant difference between treatments (Mantel-Haenszel Chi-square test, p\u200a=\u200a0.026), with a higher proportion of patients in the DHA/PQP group having borderline QTc(B) values (16.6% vs. 12.2%; p\u200a=\u200a0.066).\nNo statistically significant difference between treatments was observed at baseline for QTc(F) (2.9% for DHA/PQP vs. 1.6% in AS+MQ).\nOn day 2, there was a statistically significant difference between treatments (Mantel-Haenszel Chi-square test p<0.001), with a higher proportion of patients in the DHA/PQP group having borderline (21.4%; p\u200a=\u200a0.043) or prolonged (8.6%; p\u200a=\u200a0.007) QTc(B) intervals than in the AS+MQ group (16.3% and 4.2%, respectively).\nThis difference was also observed for the QTc(F) method: borderline, 13.0% for DHA/PQP vs. 5.3% for AS+MQ (p<0.001); prolonged, 4.7% for DHA/PQP vs. 5.3% for AS+MQ (p<0.001).\nBy day 7, there was no difference between treatments.\nThe proportion of patients with QTc(B) increase >60 msec from baseline to day 2 was 0.9% for DHA/PQP vs. 0.8% for AS+MQ (not significant), and 4.6% for DHA/PQP vs. 2.9% for AS+MQ when the same increase was assessed with QTc(F) (p<0.001).\nHowever, QTc and QT prolongation were reported as adverse events by 43 (5.6%) patients in the DHA/PQP group and 16 (4.20%) patients in the AS+MQ group; these were judged by the investigator to be related to study treatment for 28 (3.65%) patients in the DHA/PQP group and 13 (3.41%) patients in the AS+MQ group.\nMean QTc(F)values on day 0 were 387.70 msec for DHA/PQP and 385.54 msec for AS+MQ.\nMean increases from baseline to day 2 were 22.93 msec and 14.65 msec, respectively.\nThis difference between treatments was statistically significant (p <0.001).\nOn day 7, the mean increase from day 0 for DHA/PQP had fallen to 10.47 msec with the value for AS+MQ being 13.39 msec; this difference was not statistically significant (p\u200a=\u200a0.075).\nDiscussion\nIn this study conducted in Thailand, Laos and India, we have shown that both the DHA/PQP and AS+MQ treatment combinations are efficacious treatments of P. falciparum malaria.\nThe day-63 PCR-corrected cure rates (as defined in Table 1) were 87.9% for DHA/PQP and 86.6% for AS+MQ in the ITT population and 98.7% for DHA/PQP and 97.0% for AS+MQ in the per protocol population.\nIn terms of this primary outcome variable, DHA/PQP was non-inferior to AS+MQ.\nReview of the data by country found no differences between the primary outcome measures, further confirming that DHA/PQP was similarly active against the P. falciparum found in India, Laos and Thailand, relative to AS+MQ.\nThe PCR-corrected cure rates observed for DHA/PQP in this study were in line with the day-63 PCR-corrected cure rates noted in a previous study conducted in Thailand, day-42 rates observed in Myanmar (Burma) and Laos and day-28 cure rates observed in Cambodia.\nLikewise, PCR-corrected cure rates for AS+MQ were also similar to previous studies of AS+MQ conducted in Thailand, Laos and India, which generally showed rates ranging from approximately 95% to 100% in the per protocol population.\nPolymerase chain reaction-corrected cure rates for both DHA/PQP and AS+MQ exceeded 95% on days 28, 42 and 63.\nThe 95% threshold on day 28 is the level the WHO recommend for adoption of a new treatment.\nThis is the first study of the use of DHA/PQP in the Indian population, with 101 patients receiving the combination.\nAlthough this was only 13% of the patients in our study, we feel that it provides sufficient numbers to enable an initial assessment of the response to DHA/PQP in communities in Assam, Goa and Mangalore.\nIndia, Thailand and Laos have different backgrounds in terms of parasite transmission, resistance and seasonality of infection, yet a heterogeneity analysis showed no statistically significant difference in the relative responses to the two treatments between India and the other countries in the study.\nIndian communities that we sampled in this study had notable levels of chloroquine resistance, with historical site records showing treatment failure rates for chloroquine treatment ranging from 32% to 67% (28 days of follow-up).\nThe sites in Laos and Thailand were also in regions with high levels of chloroquine resistance.\nBased on the structural similarities of chloroquine and PQP, there was the potential for cross-resistance and a low response to DHA/PQP may have been expected.\nHowever, the DHA/PQP combination appeared to be unaffected by any potential cross-resistance and PCR-corrected cure rates were similar for India (96.5%), Thailand (96.6%) and Laos (98.0%).\nThe DHA/PQP combination exerted a significant post-treatment prophylactic effect in this study.\nThis is supported by, first, significant reductions in the incidence of new infections for DHA/PQP compared with AS+MQ and second, by higher uncorrected cure rates in the DHA/PQP arm than AS+MQ, as the uncorrected cure rate endpoint includes new infections as well as recurrence of infection.\nThis effect is thought to be modulated by levels of drug remaining in the blood because of the long half-life of PQP.\nThe post-treatment prophylaxis was significantly better for DHA/PQP reflecting the differential terminal half lives of PQP and MQ (4\u20135 weeks for PQP and 14 days for MQ).\nThe extended post-treatment prophylactic effect is of particular importance in countries with a high risk of new infection and can also reduce the risk of anaemia by allowing patients more time for haematological recovery between infections.\nThe DHA/PQP combination, like all ACTs, rapidly reduces parasite biomass in the patient through the brief yet potent activity of DHA (the artemisinin component).\nSubsequent removal of uncleared parasites is achieved by the less active but more slowly eliminated partner drug; in this case PQP with a half-life of 4\u20135 weeks.\nThe slightly shorter half-life of MQ may explain the difference observed in the post-treatment prophylactic effect of the two regimens.\nThe DHA/PQP regimen has previously been shown to exert a superior post-treatment prophylactic effect to another ACT, artemether-lumefantrine.\nOverall gametocyte prevalence was significantly higher in the DHA/PQP arm than AS+MQ.\nThe viability of gametocytes remaining post-treatment was not assessed in this study.\nOne potential point of interest for future studies would be to determine whether the viability of any remaining gametocytes is compromised by treatment with DHA/PQP.\nAlthough formal statistical comparison of the tolerability profile of the two combinations was not planned or conducted, some individual adverse events appeared to be reported more frequently in patients receiving AS+MQ than those receiving DHA/PQP.\nMefloquine has been linked with adverse events of gastrointestinal and central nervous system origin.\nIn this study, the frequency of reporting of nausea, vomiting and dizziness was 2\u20133 fold higher in the AS+MQ arm than the DHA/PQP arm.\nVomiting soon after dosing is an important determinant of treatment efficacy as attaining and maintaining effective systemic levels of anti-malarial drugs is essential to disease outcome.\nThis is of particular importance where there are powerful influences on adherence to dosing regimens.\nIncreases in QTc(F) interval were seen for both DHA/PQP and AS+MQ after the start of treatment, but a statistically significant increase from baseline was observed in QTc interval for DHA/PQP relative to AS+MQ on day 2.\nAn important element in assessing the clinical importance of QTc prolongation is the extent of change from baseline: an increase in QTc interval from baseline of >60 msec could be clinically significant and this is the recognised threshold for drug-induced arrhythmias in non-cardiovascular indications (ICH E14 guideline).\nThere was a statistically significant treatment difference for the proportion of patients with such an increase in QTc(F), while no treatment differences were observed for the same increase in QTc(B) and for the incidence of cardiac events.\nFinally, by day 7 there were no significant differences between treatments.\nBased on our results, it is difficult to attribute a particular clinical relevance to the QTc increase observed with DHA/PQP.\nIn fact, it is expected that malaria shortens the QT interval during the acute illness, leading to increases after the start of treatment.\nMost studies of antimalarial drugs report some transient prolongation of the QT interval in the days after the start of treatment.\nNo warnings were found in the literature about QTc prolongation with DHA/PQP.\nFor patients to access DHA/PQP via public sector healthcare systems, this fixed dose combination requires regulatory approval.\nOne requirement for registration is that the formulation must be manufactured according to GMP.\nThis study was conducted as one of a series of pharmacokinetic and ICH-GCP-compliant randomised, controlled trials at sites across Africa and South East Asia using DHA/PQP manufactured according to GMP.\nIn conclusion, the fixed dose DHA/PQP combination tablet in this study emerged as a highly efficacious treatment for P. falciparum malaria.\nThe effects were observed across the three Asian countries in which the study was conducted, with no country effect.\nThe combination of DHA/PQP provided greater protection against new infections than AS+MQ.\nResults from this study indicate that, although there may be evidence that suggests that the overall efficacy of ACTs may be falling, DHA/PQP can play an important role in, possibly, a new policy of multiple first-line treatments of uncomplicated falciparum malaria.\nFlow chart of patient disposition.\nNinety-five percent (two-sided) confidence intervals of PCR-corrected cure rates in the ITT population, by country, cohort of enrolment from the two seasons and age group.\n\nRules used to determine patient outcome for the ITT and per protocol populations for the day 63 uncorrected adequate clinical and parasitological response (ACPR) and polymerase chain reaction (PCR)-corrected ACPR.\nStep | Event to be assessed | ITT | Per Protocol\nDay-63 uncorrected adequate clinical and parasitological response\n1 | Informative withdrawal before or at day 63: any reason except lost to follow-up | Failure | Failure or excluded depending on reason\n2 | Non-informative withdrawal before or at day 63: lost to follow-up | Failure | Excluded\n31 | Presence of major protocol violation | No effect | Excluded\n4 | ETF2, LCF3, and LPF4 in accordance with the WHO criteria | Failure | Failure\n51 | Data collected on CRF (such as adverse events) raising the suspicion of recurrence of malaria | Failure | Failure\n61 | Presence of missing parasitaemia at two or more consecutive scheduled visits or presence of an isolated missing parasitaemia not preceded and followed by a negative parasitaemia | Failure | Failure\n71 | Administration of drugs with a known or suspected anti-malaria action as rescue treatment | Failure | Failure\n81 | Administration of drugs with a known or suspected anti-malaria action as non rescue treatment | Failure | Excluded\nPCR-corrected adequate clinical and parasitological response\n9 | PCR: non interpretable or missing or not done at or after day 4 | Failure | Excluded\n10 | PCR\u200a=\u200anew infection or uncorrected ACPR\u200a=\u200afailure for non-falciparum Plasmodia | Success | Success\n11 | PCR\u200a=\u200arecrudescence | Failure | Failure\n\nITT \u200a=\u200a intention to treat.\nCases in these categories were individually revised at the blind data review meeting. Protocol violations were pre-defined.\nETF \u200a=\u200a early treatment failure, defined as development of danger signs (recent convulsions, altered consciousness, lethargy, unable to drink or breast feed, recurrent vomiting, unable to stand/sit due to weakness) or severe malaria (unarousable coma, repeated convulsions, severe anaemia, respiratory distress, jaundice) on days 0, 1, 2 or 3, and the presence of parasitaemia; parasitaemia with a parasite count on day 2 greater than that on day 0 irrespective of body temperature; parasitaemia on day 3 with fever (temperature \u226537.5\u00b0C); or parasitaemia on day 3\u226525% of count on day 0.\nLCF \u200a=\u200a late clinical failure, defined as development of danger signs or severe malaria after day 3 in the presence of parasitaemia or presence of parasitaemia and temperature \u226537.5\u00b0C (or history of fever) on any day from day 4 to day 63, without previously meeting the criteria of early treatment failure.\nLPF \u200a=\u200a late parasitological failure, defined as reappearance of parasitaemia after initial clearance between day 7 and day 63 and temperature <37.5\u00b0C, without previously meeting the criteria of early treatment failure or late clinical failure.\n\nBaseline characteristics (ITT population).\nVariable | DHA/PQP | AS+MQ\n | N\u200a=\u200a767 | N\u200a=\u200a381\nSample size by country; n (%) |  | \nThailand | 466 (61) | 234 (61)\nLaos | 200 (26) | 98 (26)\nIndia | 101 (13) | 49 (13)\nGender; n : n |  | \nMale : Female | 582\u2236185 | 295\u223686\nAge; years |  | \nMean \u00b1 SD | 25.4\u00b113.3 | 25.8\u00b113.7\n\u22645 years, n (%) | 57 (7) | 32 (8)\n>5\u2013\u226412 years, n (%) | 68 (9) | 31 (8)\n>12\u2013\u226418 years, n (%) | 76 (10) | 31 (8)\n>18\u2013\u226464 years, n (%) | 566 (74) | 287 (75)\nWeight; kg |  | \nMean \u00b1 SD | 44.3\u00b115.1 | 44.6\u00b115.1\nRace; n (%) |  | \nAsian | 767 (100) | 381 (100)\nPresence of fever |  | \nn (%) | 509 (66) | 258 (68)\nTemperature in \u00b0C |  | \nMean \u00b1 SD | 37.9 (1.01) | 37.9 (1.02)\nParasite density |  | \nGeometric mean | 7923.8 | 9735.4\nHaemoglobin; g/L |  | \nNormal range (min - max)*: (105\u2013180) |  | \nMean \u00b1 SD | 118.2\u00b124.5 | 120.0\u00b123.2\nn Missing (%) | 4 (0.52) | 2 (0.52)\nn <100 g/L (%) | 162 (21.12) | 72 (18.90)\nn\u2265100 g/L (%) | 601 (78.36) | 307 (80.58)\nWhite cells; \u00d7109/L |  | \nNormal range (min - max)*: (3.6\u201311.0) |  | \nMean \u00b1 SD | 6.3\u00b12.6 | 6.3\u00b12.3\nPlatelets; \u00d7109/L |  | \nNormal range (min - max)*: (140\u2013500) |  | \nMean \u00b1 SD | 127.7\u00b170.6 | 124.8\u00b165.5\nALT; U/L |  | \nNormal range (min - max)*: (0\u201350) |  | \nMean \u00b1 SD | 31.1\u00b129.3 | 32.9\u00b141.5\nTotal bilirubin; mg/dl |  | \nNormal range (min - max)*: (0\u20131.5) |  | \nMean \u00b1 SD | 1.19\u00b10.85 | 1.20\u00b10.78\nCreatinine; \u00b5mol/L |  | \nNormal range (min - max)*: (0\u2013150.28) |  | \nMean \u00b1 SD | 75.5\u00b128.7 | 76.2\u00b130.7\nStudy populations; n (%) |  | \nSafety/ITT | 767 (99.7) | 381 (100.0)\nPer protocol | 668 (86.9) | 336 (88.2)\n\nDHA/PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine; SD \u200a=\u200a standard deviation; Hb \u200a=\u200a haemoglobin; ALT \u200a=\u200a alanine aminotransferase.\n*Since the normal laboratory reference ranges vary across centres, the minimum of the lower limits and the maximum of the upper limits are reported.\n\nDoses (mg/kg) received for each age class (ITT population).\nAge range (years) | Doses received over 3 days; median [range]\n | DHA | PQP | AS | MQ\n\u22645 | 6.7 [1.8\u20139.2] | 53.3 [14.5\u201373.8] | 12.5 [10.7\u201313.6] | 24.0 [21.7\u201327.8]\n>5\u2013\u226412 | 7.2 [5.0\u201310.0] | 57.3 [40.0\u201380.0] | 12.5 [10.7\u201313.2] | 25.0 [22.3\u201326.3]\n>12\u2013\u226418 | 7.9 [2.9\u20139.7] | 63.3 [22.9\u201377.8] | 12.0 [11.4\u201312.5] | 25.0 [24.2\u201331.3]\n>18\u2013\u226464 | 7.1 [2.1\u201310.0] | 56.5 [16.8\u201380.0] | 12.0 [3.9\u201312.5] | 25.0 [0.0\u201326.5]\n\nDHA \u200a=\u200a dihydroartemisinin; PQP \u200a=\u200a piperaquine; AS \u200a=\u200a artesunate; MQ \u200a=\u200a mefloquine.\n\nPolymerase chain reaction (PCR)-corrected and uncorrected cure rates (ITT and per protocol populations).\n | DHA/PQP | AS+MQ | Lower limit of the 97.5% | p-value2\n | % (n) | % (n) | (one-sided) CI1 | \nITT | N\u200a=\u200a767 | N\u200a=\u200a381 |  | \nDay 63 |  |  |  | \nPCR-corrected cure rate | 87.9 (674) | 86.6 (330) | \u22122.87 | 0.544\nUncorrected cure rate | 67.3 (516) | 59.6 (227) | 1.75 | 0.010\nDay 42 |  |  |  | \nPCR-corrected cure rate | 90.5 (694) | 88.2 (336) | \u22121.56 | 0.228\nUncorrected cure rate | 83.2 (638) | 77.4 (295) | 0.79 | 0.019\nDay 28 |  |  |  | \nPCR-corrected cure rate | 93.7 (719) | 91.9 (350) | \u22121.36 | 0.236\nUncorrected cure rate | 92.3 (708) | 88.2 (336) | 0.37 | 0.022\nPer protocol | N\u200a=\u200a668 | N\u200a=\u200a336 |  | \nDay 63 |  |  |  | \nPCR-corrected cure rate | 98.7 (659/668) | 97.0 (326/336) | \u22120.39 | 0.074\nUncorrected cure rate | 75.5 (504/668) | 66.4 (223/336) | 3.07 | 0.002\nDay 42 |  |  |  | \nPCR-corrected cure rate | 99.3 (663) | 97.6 (328) | \u22120.12 | 0.031\nUncorrected cure rate | 91.2 (609) | 85.4 (287) | 1.41 | 0.006\nDay 28 |  |  |  | \nPCR-corrected cure rate | 99.9 (667) | 97.9 (329) | 0.38 | 0.001\nUncorrected cure rate | 98.2 (656) | 93.8 (315) | 1.68 | <0.001\n\nDHA-PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine; CI \u200a=\u200a confidence interval.\nConfidence interval for the difference DHA/PQP minus AS+MQ.\nChi-squared test.\n\nEarly, late and true treatment failure rates1 as defined by WHO (ITT and per protocol populations).\n | DHA/PQP | AS+MQ | 95% CI\n | % (n) | % (n) | \nITT\nEarly treatment failures | 0.52 (4) | 0.26 (1) | \u22120.46, 0.98\nDay 63 |  |  | \nLate treatment failures | 13.17 (101) | 14.96 (57) | \u22126.10, 2.52\nTrue treatment failures | 2.09 (16) | 2.62 (10) | \u22122.44, 1.36\nDay 42 |  |  | \nLate treatment failures | 5.74 (44) | 9.71 (37) | \u22127.37, \u22120.58\nTrue treatment failures | 1.17 (9) | 2.10 (8) | \u22122.56, 0.70\nDay 28 |  |  | \nLate treatment failures | 1.17 (9) | 5.25 (20) | \u22126.44, \u22121.71\nTrue treatment failures | 0.65 (5) | 1.84 (7) | \u22122.65, 0.28\nPer protocol |  |  | \nEarly treatment failures | 0 | 0.30 (1) | \u22120.89, 0.28\nDay 63 |  |  | \nLate treatment failures | 12.87 (86) | 15.48 (52) | \u22127.23, 2.02\nTrue treatment failures | 1.65 (11) | 2.98 (10) | \u22123.39, 0.73\nDay 42 |  |  | \nLate treatment failures | 5.84 (39) | 9.82 (33) | \u22127.63, \u22120.34\nTrue treatment failures | 0.75 (5) | 2.38 (8) | \u22123.39, 0.12\nDay 28 |  |  | \nLate treatment failures | 0.90 (6) | 4.76 (16) | \u22126.25, \u22121.48\nTrue treatment failures | 0.15 (1) | 2.08 (7) | \u22123.49, \u22120.38\n\nThe rates reported in this table are simple rates.\n\nPrevalence of gametocytes and fever according to day of follow-up (ITT population).\n | DHA/PQP | AS+MQ | p-value1\n | N\u200a=\u200a767 | N\u200a=\u200a381 | \n | n/N (%) | n/N (%) | \nGametocyte prevalence2 |  |  | \nDay 7 | 59/749 (7.88) | 15/369 (4.07) | 0.016\nDay 14 | 30/742 (4.04) | 3/365 (0.82) | 0.003\nDay 21 | 16/733 (2.18) | 0/362 | 0.005\nDay 28 | 9/722 (1.25) | 0/353 | 0.035\nDay 35 | 1/715 (0.14) | 0/339 | 1.000\nDay 42 | 1/692 (0.14) | 0/328 | 1.000\nDay 49 | 2/666 (0.30) | 0/316 | 1.000\nDay 56 | 1/651 (0.15) | 1/312 (0.32) | 0.543\nDay 63 | 1/623 (0.16) | 2/301 (0.66) | 0.249\nOverall (from day 7 up to day 63) | 76/749 (10.15) | 18/369 (4.88) | 0.003\nPerson-gametocye-weeks3 (/100 person-weeks) | 130/6420 (2.02) | 23/3108 (0.74) | 0.014\nFever prevalence2 |  |  | \nDay 0 | 509/767 (66.36) | 258/381 (67.72) | 0.646\nDay 1 | 244/767 (31.81) | 129/381 (33.86) | 0.486\nDay 2 | 80/765 (10.46) | 44/379 (11.61) | 0.555\nDay 3 | 51/764 (6.68) | 21/379 (5.54) | 0.457\nDay 7 | 40/753 (5.31) | 16/373 (4.29) | 0.458\nOverall (from day 0 up to day 7) | 566/767 (73.79) | 298/381 (78.22) | 0.102\nPerson-fever-days4 (/100 person-days) | 1065/5315 (20.03) | 524/2636 (19.88) | 0.929\n\nDHA/PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine.\nPearson Chi-square or Fisher's exact test, as appropriate.\nCalculated, at a given time, as number of patients with gametocytes, or fever \u226537.5\u00b0C, at that time divided by the number of patients having reached that time.\nCalculated as number of weeks in which blood slides were positive for gametocytes during the whole study (up to a maximum duration of 70 days) divided by the number of all follow-up weeks and expressed per 100 person-weeks. Withdrawal patients were analysed up to the date of withdrawal recorded in the efficacy dataset even if they performed additional gametocyte assessments.\nCalculated as number of days in which temperature was greater or equal to 37.5 during the first study week divided by the number of all first follow-up weeks and expressed per 100 person-days. Withdrawal patients were analysed up to the date of withdrawal recorded in the efficacy dataset even if they performed additional assessments for temperature.\n\nMost frequently reported adverse events (>5% of either treatment arm; ITT population).\nEvent | DHA/PQP | AS+MQ | Chi-Square\n | N\u200a=\u200a767 | N\u200a=\u200a381 | p-value\n | n (%) | n (%) | \nHeadache | 138 (18.0) | 77 (20.2) | 0.3644\nMalaria1 | 111 (14.5) | 86 (22.6) | 0.0006\nP. falciparum infection | 103 (13.4) | 58 (15.2) | 0.4097\nPyrexia | 81 (10.6) | 43 (11.3) | 0.7092\nEosinophilia | 65 (8.5) | 38 (10.0) | 0.4026\nCough | 60 (7.8) | 37 (9.7) | 0.2786\nAnaemia | 55 (7.2) | 25 (6.6) | 0.7027\nMyalgia | 46 (6.0) | 22 (5.8) | 0.8801\nArthralgia | 42 (5.5) | 21 (5.5) | 0.9799\nProlonged QTc interval | 41 (5.4) | 16 (4.2) | 0.3999\nAbdominal pain | 40 (5.2) | 20 (5.3) | 0.9804\nAsthenia | 38 (5.0) | 29 (7.6) | 0.0705\nAnorexia | 38 (5.0) | 21 (5.5) | 0.6871\nNausea | 22 (2.9) | 26 (6.8) | 0.0016\nVomiting | 19 (2.5) | 24 (6.3) | 0.0013\nDizziness | 11 (1.4) | 24 (6.3) | <.0001\n\nDHA/PQP \u200a=\u200a dihydroartemisinin-piperaquine; AS+MQ \u200a=\u200a artesunate-mefloquine.\nReporting of malaria as an adverse event was not complete in this study. Some study centres chose not to report malaria as it was known that to enter the study all patients had to have Plasmodium falciparum infection.",
        "label": "unclear",
        "id": "task4_RLD_test_569"
    },
    {
        "paper_doi": "10.1186/1471-2458-11-438",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: Setting: cinema, the NetherlandsDesign: quasi-randomised controlled trialRecruitment: announcements in local newspapers, radio, and on the Internet. Other recruitment methods included posting flyers in mailboxes and handing out flyersAllocation to group: allocated according to evening available\n\n\nParticipants: 89 participants. Mean age of 50.44 (SD 12.35), 26.4% were male, 33% were overweight or obese, 50.5% had moderate educational level and 41.4% had high\n\n\nInterventions: Intervention: large poster with portion size and caloric guidelines for daily amounts (GDA) information on soft drinks (n = 48)Control: no label; different portion sizes for soft drinks were displayed indicating only the amount of millilitres that each cup contained (n = 41)\n\n\nOutcomes: Soft drink consumed (mL) during film was calculated by electronic weighing of leftovers\n\n\nNotes: Participants could choose between five portion sizes (200 mL, 250 mL, 400 mL, 500 mL and 750 mL cups). The study took place on two subsequent evenings during which participants could order free soft drinks. Authors were contacted to request information about the energy content of the soft drinks, but this information was not forthcoming. Information on study funding was not reporte\n\n",
        "objective": "To assess the impact of nutritional labelling for food and non\u2010alcoholic drinks on purchasing and consumption of healthier items. Our secondary objective was to explore possible effect moderators of nutritional labelling on purchasing and consumption.",
        "full_paper": "Background\nLarge soft drink sizes increase consumption, and thereby contribute to obesity.\nPortion size labelling may help consumers to select more appropriate food portions.\nThis study aimed to assess the effectiveness of portion size and caloric Guidelines for Daily Amounts (GDA) labelling on consumers' portion size choices and consumption of regular soft drinks.\nMethods\nA field experiment that took place on two subsequent evenings in a Dutch cinema.\nParticipants (n = 101) were asked to select one of five different portion sizes of a soft drink.\nConsumers were provided with either portion size and caloric GDA labelling (experimental condition) or with millilitre information (control condition).\nResults\nLabelling neither stimulated participants to choose small portion sizes (OR = .75, p = .61, CI: .25 - 2.25), nor did labelling dissuade participants to choose large portion sizes (OR = .51, p = .36, CI: .12 - 2.15).\nConclusions\nPortion size and caloric GDA labelling were found to have no effect on soft drink intake.\nFurther research among a larger group of participants combined with pricing strategies is required.\nThe results of this study are relevant for the current public health debate on food labelling.\nBackground\nThe mean portion size of soft drinks has increased in the past decades, and over time larger portion sizes have been added to the product lines.\nSoft drinks have been recognized as potentially important contributors to obesity and it has been demonstrated that serving larger soft drink portions results in increased beverage consumption.\nNext to the availability of larger portion sizes, 'portion distortion' might stimulate the consumption of increasingly larger amounts of soft drinks.\nNutrition labelling could help consumers to make healthy choices, and many different formats of nutrition labels, varying in design and complexity, are currently being used.\nGuidelines for Daily Amounts (GDA labelling) is one example and gives consumers standards against which they can evaluate the number of calories that a food or drink serving provides.\nIn the UK, GDA labelling was introduced by many manufacturers and retailers in 1998, whereas in continental Europe, GDA's are gradually gaining acceptance.\nOther labelling formats that have been implemented internationally are for instance the Multiple Traffic Light system, the Heart Symbol, and the Choices logo.\nPortion size labelling could both be a promising and feasible intervention to help consumers to select appropriate portion sizes.\nEspecially in Europe, portion size labelling is currently not a widespread practice, and a standard format does not yet exist.\nHowever, a pilot study on the most effective format for portion size labelling indicated that providing consumers with a reference portion size was the most promising format.\nAll in all, portion size information combined with caloric GDA labelling may help consumers choose appropriate portion sizes and moderate the effects of portion distortion in a complex food environment that provides several and large portion sizes.\nThe few experimental studies that have explored the effectiveness of portion size labelling on consumption provide inconclusive results.\nAlso, previous studies have shown that once food is served, people find it difficult to regulate their intake.\nIt is therefore important to assess the impact of labelling both on portion size choices as well as on consumption.\nFurthermore, it is important to assess the impact of labelling in more realistic settings than the laboratory.\nThe aim of the present study was to assess the impact of portion size and caloric GDA labelling on consumers' portion size choices and consumption of regular soft drinks.\nMethods\nBrief Overview\nThe study, that took place on two subsequent evenings, employed an experimental between subject design with an experimental condition with portion size and caloric GDA labelling (second evening) and a control condition (first evening).\nIn both conditions, participants could choose between five portion sizes (200, 250, 400, 500 and 750 millilitre cups).\nThese portion sizes were selected as being representative of the portion sizes currently available in the Netherlands.\nThe experimental manipulation consisted of information displayed near the bar where participants ordered their drinks.\nIn the experimental condition, portion sizes were presented on a display with both the number of portions each cup represented and the caloric GDA information (see Figure 1).\nPortion sizes were based on guidelines from the Netherlands Nutrition Centre (an institution funded by the Dutch government that provides information and education about healthy nutrition) that defines one portion of soft drink as 225 millilitres.\nHowever, as 225 millilitre cups were not the market standard, the 250-millilitre cup was designated as the reference portion.\nThe smallest size was labelled as 0.8 portions, and the largest size was labelled as three portions.\nIn the control condition, different portion sizes for soft drinks were displayed indicating only the amount of millilitres that each cup contained.\nThe comprehensibility of both the portion size and the caloric GDA information was pretested with satisfying results.\nRecruitment Procedures\nParticipants were recruited through announcements in local newspapers, radio, and on the internet.\nOther recruitment methods included posting flyers in mailboxes and handing out flyers.\nPotential participants were told that a marketing study was conducted into consumers' attitudes towards cinemas.\nParticipants, unknowing of the study conditions, could choose the evening that was most convenient for them to participate.\nThe true purpose of the study was not revealed until the conclusion of the experiment.\nParticipants were considered eligible if they were between 21 and 65 years of age.\nParticipants received a gift voucher worth \u20ac10, -.\nParticipants\nThere were 101 participants in the study.\nAfter excluding participants who had not ordered a soft drink (n = 12), the experimental condition consisted of 48 participants and the control condition consisted of 41 participants.\nOverall, participants' mean age was 50.44 (12.35), 26.4% were male and 33% were overweight or obese.\nSee Table 1 for further details.\nStudy Procedures and Data Collection\nA cinema was chosen as the location for the study because it is a setting in which a diverse range of people can be found.\nUpon arrival at the cinema, participants received the first questionnaire and were assigned a unique number and asked to write it on each questionnaire.\nParticipants self-completed the first questionnaire in the lobby, before the beginning of the film.\nThis questionnaire consisted of spurious questions about the participants' previous cinematic experiences and mood.\nIn addition, this questionnaire contained one control question measuring thirst using a visual analogue scale ranging from 0 (not at all thirsty) to 10 (very thirsty).\nSubsequently, participants were invited to the bar for a free regular soft drink.\nAfter participants received their drinks, they were invited to watch the film.\nWhen the movie was finished, participants were asked to fill out the second questionnaire.\nThe second questionnaire consisted of items that were to be used as control variables in the data analyses.\nThe questionnaire began by asking participants what they thought was the true purpose of the study.\nSubsequently, they were asked a number of questions regarding their soft drink consumption (i.e. general consumption frequency, and whether they made a habit of drinking diet or regular soft drinks).\nAdditionally, participants were asked if they had seen the displays situated above the bar.\nA number of 5-point Likert items regarding the participants' opinions of the display material and their self-reported impact of the labelling were also included.\nTo measure the participants' self-reported impact of the labelling, they were asked to rate on a 5-point Likert scale whether labelling had affected their portion size choices and soft drink consumption.\nFurthermore, participants were asked whether labelling had made them aware of appropriate soft drink portion sizes\nAdditionally, the dietary restraint and external disinhibition scales derived from the Dutch Eating Behaviour Questionnaire (DEBQ,) were included in the second questionnaire.\nDietary restraint (i.e. the deliberate restriction of energy intake with the intent to decrease or maintain weight) was measured with a scale consisting of ten 5-point Likert items (e.g. 'Do you try to eat only a little when you want to eat a lot?') with \u03b1 = .91.\nExternal disinhibition (i.e. overeating in response to external food-related cues such as sight and smell of attractive food) was measured with ten 5-point Likert items (e.g. 'If food smells yummy, do you eat a lot of it?') with \u03b1 = .83.\nThe questionnaire also contained questions on gender, age, height, and body weight.\nWhen the participants had completed the second questionnaire, they were asked to mark their participant number on their cup, and to hand in their cups and questionnaires to the research assistants.\nIf soft drink remained in the cups, this amount was weighed afterwards.\nThis study was approved by the VU Medical Centre's Institutional Review Board.\nWritten informed consent was obtained from all subjects.\nData analysis\nLogistic regression and chi square analyses were run in order to assess the impact of labelling on participants' portion size choices.\nSince we considered it relevant to assess 1) whether labelling had an effect on selecting reference portion sizes of soft drinks, and 2) whether labelling had an effect on selecting one of the two largest soft drink sizes, the data were dichotomized and coded in two different ways.\nFirst, the portion size choices were dichotomized in order to assess whether labelling stimulated participants to choose the reference portion size or smaller (i.e. 250 or 200 millilitres).\nTherefore, participants' portion size choices were either coded as the reference portion size or smaller, or as being larger than the reference portion size.\nSecond, portion size choices were dichotomized in order to assess the effect of labelling on discouraging participants from choosing one of the two largest portion sizes (i.e. 500 or 750 millilitres).\nData were either dichotomized as choosing one of the two largest portion sizes, or not choosing one of the two largest portion sizes.\nTo assess the impact of portion size and caloric GDA labelling on soft drink consumption and to assess the self-reported impact of labelling, General Linear Model procedures were used.\nThe dependent variable was either the amount of soft drink consumed, or the self-reported impact of labelling on 1) size choice, 2) soft drink consumption or 3) portion size awareness.\nBecause we randomized the study conditions instead of the individual participants, we could not rule out differences in background characteristics that are likely to be related to choice and consumption behaviour of soft drinks.\nTherefore, both the logistic analyses and the General Linear Models were adjusted for these variables (i.e. age, gender, BMI, external disinhibition, dietary restraint, thirst, and a preference for diet versus regular soft drinks).\nResults\nOn the whole, 59.8% of the participants had noticed the displays (68.8% in the experimental and 48.7% in the control condition, \u03c72 (1) = 3.59, p = .06).\nImpact of Labelling on Choice and Consumption Behaviour\nOverall, 37.5% chose the reference amount or smaller.\nA chi square analysis did not show a significant difference between both conditions, see Figure 2.\nThe logistic regression analyses indicated that portion size labelling did not increase the likelihood of choosing the reference portion size or smaller (OR = .75, p = .61, CI: .25 - 2.25).\nFurthermore, portion size labelling did not dissuade participants to choose one of the two largest portion sizes (OR = .51, p = .36, CI: .12 - 2.15).\nFinally, no significant effects of labelling were found on soft drink consumption (experimental condition: Mean = 376.30, SD = 125.40, control condition: Mean = 382.14 SD = 147.60), F (1, 71) = .39, p = .50.\nSelf-reported Impact of Labelling\nWith respect to the participants' self-reported impact of labelling, results showed no differences between both conditions on portion size choices, F (1, 46) = 2.31, p = .14.\nHowever, a significant interaction effect was found between labelling and gender, F (1, 46) = 6.66, p = .01.\nSpecifically, for women the self-reported impact on choice behaviour was slightly higher in the experimental condition (Mean = 2.76, SD = 1.48) than in the control condition (Mean = 2.20, SD = 1.58).\nWhereas, for men the self-reported impact was lower in the experimental condition (Mean = 1.50, SD = .71) compared to the control condition (Mean = 3.20, SD = 1.48).\nFinally, no significant results were found for the self-reported impact of labelling on consumption F (1, 47) = .15, p = .70 or on portion size awareness, F (1, 47) = .17, p = .68.\nDiscussion\nThis study was one of the first experimental studies that are known to us, that assessed the impact of portion size and caloric GDA labelling on consumers' regular soft drink portion size choices, their intake of soft drinks, and their self-reported awareness of portion sizes.\nThe study results did not demonstrate significant effects of portion size labelling on increasing the likelihood of selecting one of the reference sizes or decreasing the likelihood of selecting the largest sizes.\nWith respect to the latter however, it is relevant to note that the OR of selecting one of the largest sizes was lower in the experimental condition than in the control condition.\nA lack of power might explain that this result was not significant.\nTherefore, we conclude that portion size labelling did not have an effect on selecting reference portion sizes of soft drink, and that further research is needed to assess the impact of labelling on selecting large portion sizes.\nWith respect to the self-reported impact of portion size labelling on portion size choices, it seemed that labelling had a neutral effect on women, but a detrimental impact on men.\nAlthough this gender difference was not found for participants' actual consumption, this finding is partly in line with other studies showing that women generally attach greater importance to healthy eating than men and report more health information seeking behaviour.\nIt is therefore recommended to further study gender differences in consumers' responses to labelling.\nAn important factor that might explain that that we found no effect for GDA labelling is that a large majority of the participants indicated that they never or seldomly drank regular soft drinks.\nConsequently, this could make portion size and caloric GDA labelling less relevant for them.\nIn order to assess whether labelling was more effective among participants who reported drinking soft drinks regularly, the logistic analyses were also run among this subgroup of participants.\nDue to a lack of power these results could not be tested for significance, but the OR's did not indicate that portion size labelling had a beneficial impact on portion size choices (results not shown).\nIn this study we were interested in the effect of portion size labelling on portion size choices, as opposed to the replacement of regular products by diet products.\nDiet soft drinks were therefore unavailable and, as a result, participants who only drank diet soft drinks might have refused the free regular soft drink.\nAnother consequence is that we could not test the potential effect of portion size labelling on the selection of diet soft drinks instead of regular soft drinks.\nIn addition, participants did not have to purchase their drinks, obviating the cost of the drink from affecting portion size choice.\nIt is unclear how pricing would affect the impact of labelling.\nOn the one hand, free soft drinks might have stimulated participants to select larger portion sizes than they would normally have if they had to pay.\nOn the other hand, point of purchase settings employ value size pricing to stimulate consumers to choose large portion sizes too.\nNevertheless, it would be interesting to assess the impact of portion size and GDA labelling combined with proportional pricing (i.e. eliminating beneficial pricing for large portion sizes by keeping the price per millilitre consistent).\nAlso, about 40% of the participants in both conditions had not noticed the displays.\nWe chose to include all participants in the analyses, regardless of whether they had seen the displays.\nThe reason for this was that the results from these analyses would be more generalizable to real world settings in which people often oversee nutrition labels.\nIt is nevertheless worth mentioning that when the logistic regression analyses were run solely on participants who had seen the displays, comparable OR's were found (results not shown).\nAnother issue is that with respect to the participants' BMI, in this study we had to rely on self-reported data that might have suffered from a social desirability bias and under-reporting.\nWe expect that the amount of underreporting was approximately the same for both conditions, but random measurement errors resulting from the self-reported data might still have caused some residual confounding.\nLast, some researchers have suggested that multiple exposures (i.e. seeing the labels more often) may be required in order for labelling to become effective.\nFurther research on portion size and caloric GDA labelling among a larger number of participants is necessary to draw more definitive conclusions.\nIt is suggested to conduct studies that provide participants with multiple exposures to labelling and studies in which labelling is combined with pricing strategies.\nFuture studies might benefit from more objective methods to define the participants' BMI.\nLast, it is recommended to gain more insight into gender differences related to labelling.\nConclusions\nPortion size and caloric GDA labelling were found to have no effect on regular soft drink portion size choices and intake.\nFurther research with multiple exposures combined with pricing strategies among a larger number of people who have a habit of drinking regular soft drinks is recommended.\nDisplay material in the experimental condition.\nCup size choices (in %) in both study conditions1. 1\u03c72 (4) = 3.58, p = .47.\n\nParticipant characteristics\n |  | Total sample (n = 89) | Experimental condition (n = 48) | Control condition (n = 41)\n |  | Mean (SD) or % | Mean (SD) or % | Mean (SD) or %\n\nAge |  | 50.44 (12.35) | 50.12 (12.17) | 50.82 (12.71)\nSex (female) |  | 73.6 | 68.8 | 79.5\nThirst |  | 6.36 (2.87) | 6.28 (2.73) | 6.46 (3.06)\nDietary restraint |  | 2.92 (.74) | 3.00 (.68) | 2.82 (.79)\nExternal disinhibition |  | 2.81 (.52) | 2.84 (.54) | 2.80 (.50)\nEducational level\n | Low | 8 | 8.3 | 7.7\n | Moderate | 50.5 | 45.9 | 56.4\n | High | 41.4 | 45.8 | 35.9\nWeight status1\n | Underweight2 | 3.2 | 2.2 | 2.6\n | Healthy weight3 | 63.8 | 68.9 | 53.8\n | Overweight4 | 27.7 | 26.7 | 33.3\n | Obese5 | 5.3 | 2.2 | 10.3\nSoft drink consumption frequency\n | Never | 47.1 | 56.3 | 35.9\n | Seldom | 32.2 | 27.1 | 38.5\n | Sometimes | 13.8 | 12.5 | 15.4\n | Often | 5.7 | 2.1 | 10.3\n | Very often | 1.1 | 2.1 | 0\nHabitually drinks regular soft drink (when drinking soft drink) | 55.3 | 54.3 | 56.4\nInferred that study was about soft drink consumption and health. | 13.4 | 15.6 | 10.8\nHad seen display | 59.8 | 68.8 | 48.7\n\n1In the Netherlands, 35% of the population are considered to be overweight and 11% are obese\n2 BMI < 18.50\n3 BMI 18.50-24.99\n4 BMI 25.00-29.99\n5 BMI \u2265 30.00\nNote: No significant differences were found with respect to age, sex, BMI, dietary restraint, external disinhibition, thirst, and educational level between participants in the experimental and in the control condition.",
        "label": "low",
        "id": "task4_RLD_test_136"
    },
    {
        "paper_doi": "10.1186/1475-2875-12-81",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Cluster-randomised controlled trial\n\n\nParticipants: Country: KenyaSetting (coverage): 3 districts (9 sublocations allocated to intervention, 9 sublocations allocated to control)Outlets: Retail outlets (specialised drug shops and general shops)Age group: Children under 5\n\n\nInterventions: Intervention: Tibamal (subsidised ACT: AL) plus supportive interventionsComparison: No subsidised ACT (control)Supportive interventions: Training of retail outlet staff, job aids, community awareness activities (e.g. workshops, posters and paintings on shops; these activities were designed to make the community aware of malaria, the availability of Tibamal, and the importance of adherence to the medication).\n\n\nOutcomes: AL uptake (provider behaviour), availability of older antimalarials, AL price, AL stocking, provider knowledge, and provider dispensing practices. Generally, outlets that received subsidised AL plus training and job aids performed better than those receiving one or none of these intervention components.\n\n\nNotes: At the time of the study, AL had a retail price of around US$ 6.16 (500 Kenyan shillings) compared with an average of around US$ 0.37 for common, older antimalarials such as SP and AQ. The outlets were instructed to sell the packs at a retail price of US$ 0.25, which was printed on the drug packaging, providing a 150% retailer mark-up (exceeding that of AQ and SP, which generally had retail markups of 50% to 100%). Generally, outlets that received training and job aids performed better than those receiving one or none of these intervention components\n\n",
        "objective": "To assess the effect of programmes that include ACT price subsidies for private retailers on ACT use, availability, price and market share.",
        "full_paper": "Background\nMany patients with suspected malaria in sub-Saharan Africa seek treatment from private providers, but this sector suffers from sub-standard medicine dispensing practices.\nTo improve the quality of care received for presumptive malaria from the highly accessed private retail sector in western Kenya, subsidized pre-packaged artemether-lumefantrine (AL) was provided to private retailers, together with a one day training for retail staff on malaria diagnosis and treatment, job aids and community engagement activities.\nMethods\nThe intervention was assessed using a cluster-randomized, controlled design.\nProvider and mystery-shopper cross-sectional surveys were conducted at baseline and eight months post-intervention to assess provider practices.\nData were analysed based on cluster-level summaries, comparing control and intervention arms.\nResults\nOn average, 564 retail outlets were interviewed per year.\nAt follow-up, 43% of respondents reported that at least one staff member had attended the training in the intervention arm.\nThe intervention significantly increased the percentage of providers knowing the first line treatment for uncomplicated malaria by 24.2% points (confidence interval (CI): 14.8%, 33.6%; adjusted p=0.0001); the percentage of outlets stocking AL by 31.7% points (CI: 22.0%, 41.3%; adjusted p=0.0001); and the percentage of providers prescribing AL for presumptive malaria by 23.6% points (CI: 18.7%, 28.6%; adjusted p=0.0001).\nGenerally outlets that received training and job aids performed better than those receiving one or none of these intervention components.\nConclusion\nOverall, subsidizing ACT and retailer training can significantly increase the percentage of outlets stocking and selling AL for the presumptive treatment of malaria, but further research is needed on strategies to improve the provision of counselling advice to retail customers.\nBackground\nArtemisinin-based combination therapy (ACT) has been incorporated into national policies for the treatment of uncomplicated malaria in all Plasmodium falciparum-endemic countries in Africa.\nMany patients with suspected malaria in sub-Saharan Africa seek treatment from private providers, but this sector has been associated with substandard practices, such as misdiagnoses, poor counselling and incorrect dosing.\nIn addition, private sector retail prices for ACT have been high compared to those for less effective, but more widely used monotherapy, such as amodiaquine or sulphadoxine-pyrimethamine (SP).\nBarriers in the private and public sectors to accessing ACT treatment have contributed to the finding that less than 22% of children with fever were treated with an ACT across six African countries.\nIt is, therefore, important to investigate ways to improve the quality of malaria case-management in the private sector.\nAttention has focused on a range of strategies, including shopkeeper training, pre-packaging of drugs, community education, and most recently, ACT subsidies.\nIn Kenya, previous pilot studies have shown that interventions that included training of general shopkeepers could improve malaria case-management.\nHowever, these interventions were carried out when anti-malarial monotherapies such as SP and amodiaquine were still effective and were the first line treatment for uncomplicated malaria.\nThe introduction of AL as first line treatment posed major challenges to the shopkeeper training strategies since AL was less well-known than the monotherapies, and required specific knowledge and dispensing practices.\nExisting policies to train shopkeepers on older monotherapies were no longer appropriate or likely to have a significant impact on coverage of appropriate treatment.\nHere data are presented from provider and mystery-shopper surveys conducted in western Kenya, as part of a cluster-randomized, controlled trial of an ACT subsidy programme, which included shopkeeper training and community awareness activities, to evaluate its impact on private sector ACT availability, provider knowledge, and provider dispensing practices.\nMethods\nStudy sites\nThe study was conducted in three districts in western Kenya (Busia, Butere-Mumias and Teso) where the first-line treatment for uncomplicated malaria was the ACT artemether-lumefantrine (AL).\nThese areas were selected because of the presence of a relatively active retail market, and their high malaria endemicity.\nAcross the three districts, the percentage of the population living below the poverty line averaged 60%.\nAt the time of the study, Butere-Mumias had 51 government health facilities, Busia 39 and Teso 21, consisting of dispensaries, health centres and one district hospital per district (Noor, unpublished observations).\nAll government health facilities in Kenya are supposed to supply AL free to patients, although stock-outs and unofficial fees are common.\nThe three main types of retail outlets supplying anti-malarials in Kenya were registered and unregistered pharmacies (together termed specialized drug stores), and general stores.\nRegistered pharmacies are generally rare in rural areas.\nArtemether-lumefantrine was a prescription-only medicine, officially available at registered health facilities and pharmacies only, although in practice many prescription-only drugs were dispensed without a prescription in pharmacies and other retail outlets.\nAt the time of the study, AL had a retail price of around 6.16 US dollars (USD) (500 Kenya Shillings [KSH]) source of exchange rate: http://www.exchangerate.com/.\nOn 1st November 2008, one USD was equivalent to 81.23 KSH), compared with an average of around 0.37 USD for common, older anti-malarials such as SP and amodiaquine.\nMalaria diagnosis was predominantly presumptive, based on the presence of fever, in both public and private health facilities.\nStudy design\nThis study forms part of a cluster-randomized, controlled trial, evaluating the impact of the ACT subsidy intervention on effective malaria treatment of children under five.\nData were collected both before (at baseline) and after (at follow-up) the roll out of the intervention and consisted of four main data collection activities: household surveys, provider surveys, mystery-shopper surveys, and focus group discussions.\nCommunity randomization of rural sublocations is described in detail elsewhere.\nNine sublocations were allocated to the intervention arm, and nine to the control arm, with a buffer zone of two sublocations between selected sublocations.\nDue to the public information campaign around the subsidized drugs in the intervention arm, blinding was not possible for shopkeepers or data collectors.\nThe intervention\nThe intervention was implemented by the Division of Malaria Control (DOMC) of the Kenyan Ministry of Public Health and Sanitation, Population Services International (PSI), and the Pharmacy and Poisons Board (PPB).\nThe three main intervention components were provision of subsidized packs of paediatric AL to retail outlets, training of retail outlet staff, and community awareness activities.\nNo interventions were implemented in the control arm.\nIn both intervention and control arms the policy of provision of free AL at government facilities continued unchanged.\nAt baseline, public sector AL stock-outs were common, with only one third of facilities serving the study areas stocking both children\u2019s packs (six- and 12-tablet packs) of AL, but by follow-up this figure had almost doubled to 65% [19; Kangwana, unpublished observation].\nIn 2006/7 the government had carried out AL awareness campaigns across the country, so both arms had previously received some general information on the current malaria treatment policy.\nThe intervention targeted retail outlets serving intervention sublocations, which were identified through an outlet census.\nOutlets were included in the census if they were located in or on the borders of the intervention sublocations and identified by key informants as serving their populations.\nAn initial list of retail outlets was sourced from local public health officers, and updated with input from local chiefs and subchiefs.\nThe list was further amended after visiting the study areas with village elders, through the snowball technique where each shop visited was asked about the presence of other outlets in their area, and by discussions with community members.\nEnumerated outlets were invited for training if they had been functioning for a minimum period of six months and were identified as selling anti-malarials and/ or antipyretics during the past year.\nOutlet staff attended a one-day, malaria-related training between August and October 2008 covering clinical diagnosis, treatment, adverse drug reactions, and patient referral.\nTraining materials were developed by the implementation team, building on those used previously for shopkeeper training in Kenya.\nFrom November 2008, subsidized AL was provided to trained retail outlets in packs of six tablets (for children aged three to 35\u00a0months) and 12 tablets (for children aged 36 to 59\u00a0months).\nThe AL was branded as Tibamal, a pretested name derived from the Kiswahili words \u201cTiba ya Malaria\u201d meaning malaria cure, and came with patient instructions suitable for those with low literacy levels.\nThe PPB granted special dispensation for AL to be dispensed over the counter in the intervention arm.\nPSI sales staff delivered the treatment directly to trained outlets on a monthly basis, at a wholesaler price of 0.10 USD per pack (both packs were the same price).\nThe outlets were instructed to sell the packs at a retail price of 0.25 USD, which was printed on the drug packaging, providing a 150% retailer mark-up, exceeding that of amodiaquine and SP, which generally had retail mark-ups of 50% to 100%.\nTrained outlets were supplied with job aids to support dispensing, consisting of a referral flow chart and dosing guidelines.\nThe main community awareness activities began in March 2009, focusing on malaria symptoms, Tibamal availability, and patient adherence.\nThey consisted of community leader workshops, community events; small group discussions and outreach activities.\nTibamal was also advertised through posters and paintings on shops selling the treatment, and through distribution of branded headscarves, tee shirts, and pens.\nA follow-up, supervisory visit was made by the implementation team three months after initial provision of Tibamal supplies to monitor outlet practices.\nData collection\nData on retailer performance were collected through provider surveys and mystery-shopper surveys, which aimed to determine whether private sector retailers could deliver AL to appropriate standards of quality for the treatment of fever in children under five.\nThe provider survey assessed the anti-malarials stocked, and knowledge and stated practices of the provider related to malaria treatment, such as knowledge of the first-line anti-malarial, malaria symptoms, and advice to provide when dispensing AL.\nThe mystery-shopper survey assessed patient-provider interactions, to provide information on actual rather than self-reported provider practice.\nBoth data collection activities were conducted at baseline (July-August 2008) and follow-up (July-August 2009).\nThe mystery-shopper survey was conducted first, to avoid raising awareness by the shopkeepers of the presence of the survey team.\nThe sampling frame for both surveys was based on the retail census carried out in May 2008 and May 2009, described above.\nAll outlets that had been functioning for at least six months and had sold an anti-malarial or antipyretic within the past year were included in survey data collection.\nFor the provider survey, written consent was obtained from shopkeepers at the time of the interview.\nFor the mystery-shopper survey, written consent was sought in advance from all shopkeepers interviewed during the retail census.\nAt the time of the census, staff were informed about the nature of the mystery-shopper survey, but details on whether their shop was selected and the date of the survey visit were not revealed.\nThe gap between the census and the mystery-shopper survey was just over a month.\nProvider survey data were collected using a pretested structured questionnaire administered to the shopkeeper present at the time of the visit who was most responsible for selling medication.\nIn the mystery-shopper survey, fieldworkers disguised themselves as local residents seeking treatment for a child.\nAll fieldworkers were trained to present the following scenario to the shopkeeper: a four-year-old child (weighing 15 kg), under their care who has been suffering from a recurring fever for three days, especially at night.\nThe child had no other symptoms, and no medication had been given so far.\nThe fieldworker discreetly recorded the details of the interaction on a questionnaire away from the outlet, once the interview was complete.\nAll interviews were carried out in each district\u2019s local dialect.\nData analysis\nBased on baseline values of the primary indicator \u2018proportion of outlets stocking any unexpired AL\u2019 the study sample size of nine clusters per arm and an average of 26 outlets per cluster was sufficient to detect a 20% point difference at a 0.05 significance level and 80% power.\nData were analysed in STATA version 11 (College Station, TX, USA) by a two-stage process, with baseline and follow-up data analysed separately.\nIn the first stage a summary cluster measure was obtained for each cluster.\nThe second stage involved comparing the sets of cluster-specific measures in control and intervention arms at follow-up using the unpaired t-test.\nA crude analysis was carried out on the cluster summaries using the simple two-tailed t-test to obtain the mean percentage difference between the intervention and control arms, 95% confidence intervals (CIs) and standard deviations (SDs) for each outcome.\nIn addition, an adjusted analysis was carried out at follow-up using an individual level logistic regression run on the pooled data set (control and intervention arms).\nTo control for potential confounders the covariates considered were outlet type (specialized drug store or general store), distance of shop to nearest road, whether any staff had clinically related training, and district.\nAll covariates significant at a p-value of <0.2 were retained in the regression model.\nBaseline values for the outcome in question were also included as covariates if a difference of 5% points or more was observed between the arms at baseline.\nThe intervention status of the cluster was not included in the logistic regression model.\nRather, the regression model provided the predicted outcome in the absence of the intervention effect.\nMean predicted and observed outcomes were obtained per cluster and residuals were obtained by subtracting the predicted outcomes from those observed in each cluster.\nThe t-test was used on these residuals to assess the difference in the residuals, which represents the intervention effect adjusted for the covariates included in the logistic regression model.\nThe t-test was used for both crude and adjusted analyses, as it has been shown to be highly robust even for small numbers of clusters.\nFor indicators concerning the \u2018percentage of outlets providing correct dispensing advice for AL\u2019 in the mystery-shopper survey, the Fisher\u2019s exact test was used to calculate p values, due to the small number of observations in the denominator in the control arm.\nDistances from retail outlets to nearest roads were calculated using the Euclidean tool in ArcGIS (ESRI, Redlands, CA, USA) spatial analysis tool.\nSince the intervention was delivered under operational rather than ideal conditions, outlets in the intervention arm varied in their level of exposure to the intervention.\nSpecifically, not all outlets in the intervention arm were trained, and not all outlets received job aids.\nThe analysis was therefore conducted on both an intention to treat and per-protocol analysis basis.\nIn the intention to treat analysis outlets were analysed in the clusters to which they were randomized, regardless of whether they received the intervention.\nIn the per-protocol analysis only outlets in the intervention cluster that received the intervention were included.\nTwo levels of intensity of intervention were considered: outlets that received the Tibamal training, termed \u2018trained\u2019, and those that received the Tibamal training and had a Tibamal job aid in the outlet at the time of the interview, termed \u2018trained with job aid\u2019.\nNo hypothesis testing was carried out to evaluate the statistical significance in differences observed between these subsamples and the population of all outlets in the intervention arm as the groups represent overlapping categories (some outlets were present in two or three of these groups).\nIn both the provider and mystery-shopper surveys, at baseline and also at follow-up, less than 10% of outlets were not interviewed either because the respondent refused to be interviewed or the outlet was closed during visits.\nEthical approval\nEthical approval was obtained from the Kenya Medical Research Institute (KEMRI) Ethical Review Committee (#1361), the Kenya Pharmacy and Poisons Board Ethical Committee for Clinical Trials (# PPB/ECCT/08/07), and the London School of Hygiene and Tropical Medicine Ethical Review Committee (# 5288).\nThe study is registered with Current Controlled Trials (# ISRCTN59275137).\nResults\nOutlet characteristics\nThe number of outlets meeting the census selection criteria was 295 in the control arm and 225 in the intervention arm at baseline, and 369 and 351 respectively at follow-up.\nFor the provider survey, a total of 468 outlets were successfully interviewed at baseline and 639 at follow-up; and for the mystery-shopper survey, 499 at baseline and 653 at follow-up (Table\u00a01).\nThe changes in number of outlets surveyed at baseline compared to follow-up were partly due to the fluidity of the market, where over time some shops would close temporarily or permanently, and new ones open.\nIn addition, there is a possibility that fieldworkers became better at identifying outlets at follow-up.\nIn both surveys, general stores constituted over 70% of shops evaluated at baseline and follow-up, with specialized drug shops making up almost all the remainder (Table\u00a01).\nThe mean number of staff serving customers was just under two per outlet at baseline and follow-up.\nThe percentage of outlets with staff with any clinical training, who usually or occasionally served customers, ranged from 15.7% at follow-up in the control arm to 23.5% at baseline in the intervention arm (Table\u00a01).\nWhen broken down by outlet type, this averaged 63.6% and 75.7% in specialized drug stores and 9.6% and 4.5% in general stores, across both arms at baseline and follow-up, respectively.\nLess than 4% of outlets had a child below 16\u00a0years of age usually or occasionally serving customers.\nThe mean distance of retail outlets to the nearest road (any road excluding footpaths) was 188 and 327\u00a0m in the control and intervention arms respectively at baseline, and 203 and 231 respectively at follow-up (Table\u00a01).\nWithin the subgroup of outlets that had received Tibamal training (trained) at follow-up, 31.1% were specialized drug stores, and in outlets that had received training and had a Tibamal job aid (trained with job aid) 33.9% were specialized drug stores.\nOutlet characteristics within the trained and trained with job aid subgroups were similar to those for all outlets in the intervention arm at follow-up (Table\u00a01), although specialized drug stores were slightly more common among the trained and trained with job aid outlets than among all intervention outlets.\nThe characteristics of outlets and staff surveyed in the mystery-shopper survey were similar to those in the provider survey (data not shown).\nExposure to the intervention\nIn the provider survey 71.3% of respondents at baseline and 77.0% at follow-up had heard of AL, across the arms.\nAt follow-up, 13.9% and 91.6% of respondents were aware of Tibamal, in the control and intervention arms, respectively.\nAlso at follow-up, 43.1% of respondents in the intervention arm reported having at least one member of staff that had attended the Tibamal training, compared to only 1.0% in the control arm.\nOf outlets in the intervention arm, 67 (22.1%) were in possession of a Tibamal job aid and 62 (20.5%) had at least one trained staff member and a Tibamal job aid.\nNo outlet in the control arm had a job aid.\nAnti-malarial availability and storage\nDuring the provider survey, unexpired AL was found in less than 3% of outlets at baseline, across both arms.\nAL stocking had increased by follow-up to 36.8% in the intervention arm, but remained low at 5.2% in the control arm (Table\u00a02).\nThe difference between the arms at follow-up was significant (adjusted p=0.0001, difference in means: 31.7%; 95%CI: 22.0, 41.3).\nTibamal constituted around 95% of all AL available in the intervention arm at follow-up, with no outlets stocking Tibamal in the control arm.\nLess than 3% of outlets were observed with stocks of any other ACT in both arms and at both time points.\nAt baseline 52.8% of outlets in the control arm stocked a non-ACT anti-malarial (SP, amodiaquine, artemisinin monotherapy, quinine or chloroquine) and 63.7% in the intervention arm.\nThe availability of non-ACT fell by 13 percentage points and 24 percentage points in the control and intervention arms respectively, from baseline to follow-up, though the difference in availability between the arms at follow-up was not significant (adjusted p value=0.5187, difference in means: 0.4%; 95%CI: -7.6, 8.5) (Table\u00a02).\nThe availability of artemisinin monotherapy remained below 5% at baseline and follow-up, and across the arms.\nAmong the subgroup of trained outlets 65.7% stocked AL at follow-up, and among the subgroup of trained with job aid outlets, 76.8%, 29 and 40 percentage points higher than among all intervention arm outlets.\nAvailability of non-ACT was also higher in the two subgroups when compared to all outlets in the intervention arm at follow-up.\nNine intervention outlets stocked Tibamal even though none of the staff were reported to have attended the training.\nIt was not clear whether this was because the staff member who had attended training was not present on the day of the survey and other staff were not aware they had attended; whether there was some \u201cleakage\u201d of the subsidized drug to untrained shops; or whether the trained staff were no longer working in the outlet.\nOutlets stocking AL were assessed to see if the treatment was stored appropriately.\nThe definition of \u2018appropriately\u2019 was all AL packs kept off the floor, out of direct sunlight, in a dry area and with packaging intact.\nAt follow-up, 78.8% and 81.9% of outlets were observed to be storing all AL stocks appropriately in the control and intervention arms, respectively.\nThe percentages were similar in trained outlets and trained with job aid outlets and in all intervention outlets.\nExpired AL stock was found in less than 1% of outlets in both arms and time points.\nProvider knowledge\nData from the provider survey showed that at baseline 37.8% of respondents in the control arm and 34.3% in the intervention arm were able to identify AL as the first-line treatment for uncomplicated malaria.\nAt follow-up, this had improved in both arms, but was significantly greater in the intervention arm compared to the control arm (adjusted p value = 0.0001, difference in means: 24.2%; 95%CI: 14.8, 33.6) (Table\u00a03).\nFever, as a symptom of uncomplicated malaria was mentioned by 67.3% of respondents at baseline, across both arms, and increased to 74.3% and 84.0% in the control and intervention arms respectively.\nRespondents were asked where they would recommend a four year-old child suspected to be suffering from complicated malaria to seek treatment first.\nOver 70% of respondents correctly said they would refer the child directly to a health facility at baseline, in both arms.\nThis increased to more than 80% at follow-up, across both arms.\nFor all the knowledge indicators described above, respondents from the subgroup of trained outlets performed better than the population of all outlets in the intervention arm at follow-up.\nOutlets that were trained with job aid performed better than trained outlets with the greatest difference of 10.7 percentage points (82.7% trained: 93.4% trained with job aid) observed in the percentage knowing the first-line anti-malarial (Table\u00a03).\nIn the provider survey, respondents were asked about the advice they would give to a caregiver purchasing any brand of AL for their four year-old child.\nRespondents were asked to advise on AL administration, what to do if the child vomits, what to do if the child does not improve, and foods to administer with the medication (Table\u00a03).\nAcross all the advice points there was significantly higher knowledge (p<0.005) in the intervention arm, compared to the control arm, at follow-up.\nThe greatest difference was observed in advice to give \u2018if the child does not get better\u2019; at baseline 47.7% of respondents knew the correct advice, and at follow-up 39.8% and 66.0% in the control and intervention arms respectively (adjusted p value=0.0001; difference in means: 26.2%; 95%CI: 15.0, 37.4).\nThe least improvement was observed in \u2018what to do if the child vomits after taking the medication\u2019, where no respondents knew the correct response at baseline, and at follow-up this increased to only 1.5% and 9.4% in the control and intervention arms respectively (adjusted p value: 0.0027; difference in means 7.9%; 95%CI: 2.9, 12.9).\nAcross these advice indicators, the subgroup of trained outlets performed better than the population of all outlets in the intervention arm, with the greatest difference observed in \u2018what to do if the child does not get better\u2019 where there was a 22.8 percentage point difference (66.0% all outlets; 88.8% trained outlets).\nThe least improvement was observed in \u2018what to do if the child vomits\u2019 where there was a 7.5 percentage point difference (9.4% all outlets; 16.9% trained outlets).\nTrained outlets with job aids performed better than trained outlets, with the greatest difference in the correct advice on foods to give the child (45.1% trained outlets; 61.1% trained with job aid), and the least difference in what to do if the child vomits (16.9% trained outlets; 18.2% trained outlets with job aids) (Table\u00a03).\nProvider behaviour\nAt baseline anti-malarials were dispensed to 25.2% of mystery shoppers in the control arm and 40.7% in the intervention arm, and this was little changed at endline, with 20.2% in the control arm and 40.3% in the intervention arm (Table\u00a04).\nThis remained fairly similar at follow-up.\nThe reasons for the higher frequency of anti-malarial dispensing in the intervention arm at both time points were not clear.\nLess than 1% of mystery shoppers were sold AL at baseline, across both arms.\nAt follow-up, sales of AL remained low at 1.8% in the control arm, but were significantly higher at 25.4% in the intervention arm (adjusted p=<0.0001; difference in means: 23.6%; 95%CI: 18.7, 28.6).\nIf only those intervention outlets are considered which had a staff member who had attended Tibamal training, and had Tibamal in stock during the provider survey, 60% dispensed Tibamal to the mystery shoppers.\nOf the 40% (39 outlets) that did not dispense Tibamal, 16 (mainly general stores) referred the mystery shopper to a specialized drug store, 13 (mainly drug stores) dispensed another anti-malarial (mainly SP), 6 dispensed an antipyretic only, 3 referred to a general shop, and 1 referred to a health facility.\nAn average of 24.5% of respondents across both arms asked their clients whether the child had at least one danger sign at baseline.\nThis improved slightly in the intervention arm at follow-up, however the difference between the arms was not statistically significant (adjusted p value=0.2667; difference in means 4.9%; 95% CI: -7.9, 17.7).\nFor both of these provider-behaviour indicators trained outlets performed better than all outlets in the intervention arm, with the greatest difference observed in outlets dispensing AL (25.4% all outlets; 43.3% trained outlets).\nSimilarly, trained outlets with job aids performed better than trained outlets, with the greatest difference also observed in dispensing of AL (43.3% trained outlets; 62.7% trained with job aid).\nLess than 1% of providers gave correct dispensing advice for AL on any of the counselling points at baseline, across both arms.\nAt follow-up there were improvements in advice in the intervention arm, but the differences between the arms were not significant for any counselling indicators related to AL dispensing.\nThe greatest difference was observed in what to do if the child does not get better, which remained 0% in the control arm compared with 35.3% in the intervention arm (difference in means: 35.3%; 95% CI: 23.6, 47.1).\nTrained outlets performed better than all outlets in the intervention arm on advice on administration of AL, and what to do if the child does not get better.\nLittle difference was observed between the groups in what to do if the child vomits, and foods to give the child.\nTrained outlets with job aids performed better than trained outlets in all counselling points apart from on advice on foods to give the child.\nCost to consumers of artemether-lumefantrine\nIn the mystery-shopper survey, at baseline there were only two doses of AL sold, at a cost of 2.46 and 2.22 USD.\nAt follow-up, the 12 tab Tibamal was sold at a median price of 0.25 USD (IQR: 20\u201320), which was the recommended retail price.\nOf those not paying the recommended price, two paid 0.37 USD, another two paid 0.49 USD because of buying two packs of the six-tab to meet the required dose, and one paid 0.74 USD.\nDiscussion\nBefore discussing the results in more detail a number of limitations in the provider and mystery-shopper surveys should be highlighted.\nShops that had undergone Tibamal training were identified by asking the respondent if any of the staff had attended the training, but it is possible that some responses were inaccurate due to recall bias or lack of awareness of training of fellow staff members.\nIn addition, given the prior consent process, it is also possible that retailers were suspicious and therefore altered their practice and the advice they gave to the mystery shopper.\nHowever, if this were the case then the data would display \u2018best practice\u2019 of providers, while still showing considerable room for improvement, especially in areas such as the provision of appropriate counselling.\nIn the scenario, the mystery shoppers waited for the retailer to recommend treatment and paid whatever price they were asked to.\nOther data (Kangwana and colleagues, unpublished observations from Tibamal focus group discussions and provider survey reports) suggest that in practice the consumer often asks for a specific treatment instead of the provider recommending it, so real-life interactions may be somewhat different.\nIt is also possible that there was some contamination of the control arm outlets, which could have heard some of the communication activities.\nHowever, results indicated that such exposure was low, with only 1% of control arm respondents saying that they had attended the Tibamal training, 14% having heard of Tibamal, and no outlets stocking Tibamal.\nThe intervention was able to significantly improve the percentage of outlets stocking AL, and more than 90% of the AL available at follow-up in the intervention arm was Tibamal.\nThis indicates a willingness of shopkeepers to take part in the intervention and make the treatment available in their outlets.\nThe intervention had an effect on most provider knowledge indicators.\nSignificantly more providers in the intervention arm compared to the control arm knew AL was the first-line treatment for uncomplicated malaria and knew fever as a symptom of uncomplicated malaria.\nProviders were also more knowledgeable on correct dispensing practices for AL, than those in the control arm.\nHowever, although the difference observed between the arms was significant, there remains a need for renewed effort to improve these components of knowledge and prescription.\nFor example, for the scenario of \u2018what to do if the child does not get better\u2019, only 66% of providers could give the correct response.\nIn addition, findings from the mystery-shopper survey revealed that knowledge was not always transferred into practice, since no significant improvements were observed in any of the four appropriate counselling indicators.\nThe mystery-shopper survey revealed that the intervention not only encouraged shopkeepers to stock AL but also significantly encouraged them to dispense AL to clinically diagnosed cases of uncomplicated malaria.\nThis is in line with findings from the household surveys, which showed that a significantly greater percentage of febrile children in the intervention arm were treated with AL compared to the control arm at follow-up.\nThe findings in this paper on changes in provider behaviour also serve to strengthen the claim that the changes observed in the household survey were very likely due to the Tibamal intervention.\nEncouragingly, the mystery shopper data showed providers adhering to the recommended retail price of Tibamal.\nFindings from the household survey also revealed that >95% of caregivers purchased Tibamal at its recommended retail price.\nPrinting of the recommended price on Tibamal packaging as well as making consumers aware of the subsidy through the community activities may have contributed to providers not inflating Tibamal prices.\nAlso, the way in which the pricing was structured meant that even with the subsidy, the retailer mark-up would be greater than for other more commonly prescribed anti-malarials, which may have facilitated both stocking of Tibamal by retailers and appropriate pricing.\nThe intervention did not significantly increase the percentage of providers that would refer complicated cases of malaria directly to a health facility.\nThe data show that even without the intervention, around 80% of shopkeepers stated that they would automatically refer these cases to a health facility, but since immediate specialized care is required for complicated cases, one would hope to see a referral rate of close to 100%.\nThe intervention was also not able to significantly increase the percentage of providers asking for at least one danger sign, an important way of identifying severe cases.\nIt could be that providers instead tended to rely on the consumer to provide all the necessary information without being probed.\nIdentifying and implementing ways to improve enquiries about danger signs is therefore important, in addition to providing caregivers with the knowledge of when to bypass the retail sector and go straight to a health facility.\nAlthough the intervention improved the share of AL among mystery-shopper purchases it had no significant impact on decreasing the availability of non-ACT in outlets.\nThere was a decline in non-ACT availability at follow-up, but this was observed in both arms and was thought to be as a result of a government directive to halt the production and supply of less effective monotherapy at the time of the survey.\nAvailability of artemisinin monotherapy was not a concern, with less than 5% of outlets stocking this treatment, probably due to low demand as a result of its high cost compared to other anti-malarial monotherapy.\nOverall, it seemed that the greater exposure shopkeepers had to all components in the intervention, the better they tended to perform.\nOutlets that had received training seemed to perform better than all outlets in the intervention arm, and outlets that had received both training and job aids seemed to perform better than outlets just exposed to training.\nThis shows the importance of ensuring that implementation of the intervention is as \u2018ideal\u2019 as possible; had higher coverage of training and job aids been achieved, even more substantial improvements in provider behaviour and treatment coverage would likely have been achieved.\nThese findings are of particular importance given the current roll out of similar ACT subsidies under the Affordable Medicines Facility - malaria (AMFm) on a national scale in Kenya and seven other countries, also accompanied by training and communications activities.\nIn Kenya AMFm training of retail providers was limited to registered pharmacists who were the only retailers officially allowed to stock the AMFm subsidized product, although in practice it was widely available in unregistered pharmacies.\nBy demonstrating variation in performance in relation to intervention intensity, and highlighting areas of particularly weak provision, these results can be used to identify potential strategies to enhance provision of subsidized ACT under AMFm and other similar subsidy mechanisms.\nThe authors of this paper are not aware of any other studies that explore the effect of an intervention including an ACT subsidy on the performance of private sector retailers in the treatment of presumptive malaria.\nSeveral studies in sub-Saharan Africa have however evaluated the effect of other interventions to improve the quality of care received from the retail sector in the treatment of presumptive malaria, mostly with anti-malarial monotherapy.\nThe majority of the interventions included training of either providers or users, and combined this with other supporting activities such as provision of job aids, follow-up monitoring and provision of prepackaged anti-malarial treatment with pictorials to guide administration.\nThe outcomes of the interventions varied between studies but the majority showed positive outcomes.\nOverall the interventions were able to improve provider knowledge on signs and symptoms of malaria, and the proportion of providers giving correct treatment and dose.\nSome interventions increased provision of correct advice on administration, and asking for danger signs.\nInterventions were also able to improve the availability of anti-malarials in outlets.\nStudies differed in their design, data analysis techniques used and the type of outcome measures used.\nFew studies carried out hypothesis testing on their outcome results so it is difficult to interpret the importance of any observed differences, and some studies had limitations, for example very small sample sizes or no appropriate comparison group.\nAll these factors make it difficult to make a direct comparison of outcomes in these studies with the data presented here.\nExploration of the context in which the survey took place, together with interviews with participants in the intervention through focus group discussions (Kedenge and colleagues, unpublished observations from the Tibamal focus group discussion report), have provided some insight into the findings and how the results could have been further improved.\nAt follow-up, only 43% of outlets in the intervention arm were identified as trained.\nSeveral possible reasons were given including that some businesses trained at baseline had closed due to lack of capital; some businesses relocated to other areas or had changed their type of business, for example, from a general store to a bicycle spare part shop.\nAlso, recently opened but untrained shops had not had the opportunity to be trained.\nThis highlights the need to hold regular courses to ensure that a well-trained cadre of shopkeepers is maintained.\nIn addition some providers thought that the training should have been longer, refresher courses should have been included and the training could have better catered for illiterate shopkeepers.\nFurthermore in response to the low percentage of patients being counselled, some shopkeepers were said to work in a very busy environment, having to attend to more than one customer at a time, making it difficult to discuss these issues with caregivers in any detail.\nAs this is unlikely to change, this highlights the merits of also providing this information directly to consumers through communication activities.\nCare should be taken when extrapolating or generalising the findings of this pilot to other areas.\nSince Tibamal was being supplied directly to outlets, there is the possibility that monthly contact with PSI staff distributing AL may have had the effect of influencing providers to work at their best.\nOther factors affecting the generalizability of these findings have been discussed by Kangwana et al..\nFinally, the World Health Organization now recommends that all suspected cases of malaria should be parasitologically diagnosed before treatment, where diagnostics are available.\nTherefore further research is required into understanding how diagnostics will change provider practices, and assessing strategies to ensure diagnostics are used appropriately and supplied at an affordable price.\nConclusion\nThe private sector remains a preferred source of anti-malarial treatment in much of sub-Saharan Africa, including Kenya.\nIt is, therefore, important to support and enable providers to provide appropriate and rational management of malaria.\nThis study has shown how an intervention comprising of subsidized pre-packaged AL, retailer training and community awareness activities can improve certain aspects of provider knowledge and practices of presumptive malaria treatment.\nThe intervention significantly increased provider awareness of the government\u2019s recommended first-line treatment for uncomplicated malaria, and along with granting of over-the-counter status to AL, significantly increased its availability.\nThe intervention also significantly increased the percentage of providers recommending and dispensing AL, and encouraged providers to pass on the subsidy to customers, making the treatment more affordable.\nFurther research is needed to improve aspects of care where the intervention had minimal impact, particularly provider knowledge and behaviour around danger signs of malaria, and counselling practices.\n\nCharacteristics of staff interviewed in the provider survey (cluster summaries from the 9 intervention and 9 control clusters)\n\u00a0 | Baseline | Follow up\n\u00a0 | Control% (SD) n | Intervention% (SD) n | Control% (SD) n | All Intervention% (SD) n | Trained Intervention% (SD) n | Trained & Job aid Intervention% (SD) n\nNo. of outlets | 263 | 205 | 319 | 320 | 154 | 62\nOutlet type: | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nSpecialized drug stores | 19.6 (8.4) 49 | 26.2 (16.3) 53 | 17.8 (7.3) 56 | 22.9 (8.7) 74 | 31.1 (5.5) 49 | 33.9 (28.6) 21\nGeneral stores | 80.4 (8.4) 214 | 73.8 (16.3) 152 | 81.9 (7.6) 262 | 77.1 (8.7) 246 | 68.8 (5.5) 105 | 66.1 (28.6) 41\nOther | 0 (0) 0 | 0 (0) 0 | 0.3 (1.0) 1 | 0 (0) 0 | 0 (0) 0 | 0 (0) 0\nOutlets with at least one member of staff usually serving customers with the following characteristics: | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nPrimary education complete and above | 79.2 (9.1) 211 | 76.8 (17.0) 159 | 74.8 (11.5) 237 | 68.7 (13.1) 222 | 72.6 (18.2) 111 | 70.4 (23.9) 42\nAny clinical training1 | 21.3 (8.1) 55 | 23.5 (18.8) 51 | 15.7 (6.7) 50 | 18.9 (8.4) 63 | 24.3 (6.3) 40 | 27.4 (28.2) 16\nBelow 16\u00a0years | 3.5 (5.8) 8 | 3.3 (3.5) 7 | 3.4 (3.5) 10 | 2.1 (2.7) 6 | 2.1 (4.2) 4 | 2.7 (5.5) 2\nDistance to nearest road: | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0 | \u00a0\nMetres (SD) | 187.7 (123.8) | 326.6 (286.9) | 201.6 (121.5) | 231.4 (98.8) | 256.3 (156.7) | 300.6 (213.6)\n\nn=numerator (denominator for all indicators is shown under No. of outlets).\n1 Any clinical related training consists of: pharmacy, nurse and medical doctor related training; Pharmacy related training includes pharmacy studied to a certificate or diploma level; Nurse related training includes studying nursing to a certificate level (nurse aid) and diploma level; Medical doctor training includes clinical officer who studied medicine to a diploma level.\nNote: In the mystery shopper survey, a total of 499 outlets were successfully interviewed at baseline, 284 and 215 in control and intervention arms respectively, and 653 outlets at follow-up, 336 in the control and 317 in the intervention arm.\n\nAvailability of anti-malarials (provider survey) (cluster summaries from the 9 intervention and 9 control clusters)\nAvailability of anti-malarials: | Control% (SD) n/d | Intervention% (SD) n/d | Difference in means (95%CI) | P value Unadjusted Adjusted\nAny AL (unexpired): | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u2003\u2002Baseline | 0.5 (1.2) 2/263 | 1.5 (3.2) 3/205 | \u00a0 | \u00a0\n\u2003\u2003\u2002Follow up | 5.2 (4.0) 18/319 | 36.8 (13.1) 117/320 | 31.7 (22.0, 41.3) | 0.0001 0.00011\n\u2003\u2003\u2002Tibamal trained | - | 65.7 (16.1) 103/154 | \u00a0 | \u00a0\n\u2003\u2003\u2002Tibamal trained & job aid | - | 76.8 (32.4) 50/62 | \u00a0 | \u00a0\nTibamal (unexpired): | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u2003\u2002Baseline | - | - | \u00a0 | \u00a0\n\u2003\u2003\u2002Follow up | 0 (0) 0/319 | 35.1 (12.3) 111/320 | 35.1 (26.4, 43.8) | 0.0001 0.00012\n\u2003\u2003\u2002Tibamal trained | - | 65.3 (15.6) 102/154 | \u00a0 | \u00a0\n\u2003\u2003\u2002Tibamal trained & job aid | - | 76.8 (32.4) 50/62 | \u00a0 | \u00a0\nOther ACT: | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u2003\u2002Baseline | 0.8 (1.6) 2/263 | 0 (0) 0/205 | \u00a0 | \u00a0\n\u2003\u2003\u2002Follow up | 2.1 (2.8) 7/319 | 0.4 (1.1) 1/320 | \u22121.7 (\u22123.9, 0.4) | 0.1058 0.05802\n\u2003\u2003\u2002Tibamal trained | - | 0.5 (1.6) 1/154 | \u00a0 | \u00a0\n\u2003\u2003\u2002Tibamal trained & job aid | - | 1.1 (3.3) 1/62 | \u00a0 | \u00a0\nNon-ACT therapy: | \u00a0 | \u00a0 | \u00a0 | \u00a0\n\u2003\u2003\u2002Baseline | 52.8 (12.0) 140/263 | 63.7 (10.4) 130/205 | \u00a0 | \u00a0\n\u2003\u2003\u2002Follow up | 39.6 (10.0) 128/319 | 40.0 (5.5) 127/320 | 0.4 (\u22127.6, 8.5) | 0.9082 0.51873\n\u2003\u2003\u2002Tibamal trained | - | 54.3 (13.5) 80/154 | \u00a0 | \u00a0\n\u2003\u2003\u2002Tibamal trained & job aid | - | 53.8 (28.1) 35/62 | \u00a0 | \u00a0\n\nn numerator, d denominator, SD standard deviation, CI confidence interval.\n1 adjusted for distance to nearest road; 2 adjusted for outlet type and district; 3 adjusted for outlet type, distance to nearest road and clinical training.\n\nProvider knowledge (provider survey) (cluster summaries from the 9 intervention and 9 control clusters)\n\u00a0 | Control% (SD) n/d | Intervention% (SD) n/d | Difference in means (95%CI) | P value Unadjusted Adjusted\nPercentage knowing the first line anti-malarial for uncomplicated malaria: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 37.8 (9.0) 103/263 | 34.3 (16.6) 70/205 | \u00a0 | \u00a0\nFollow up | 46.9 (7.9) 151/319 | 71.1 (10.9) 227/320 | 24.2 (14.8, 33.6) | 0.0001 0.00011\nTibamal trained | - | 82.7 (9.8) 129/154 | \u00a0 | \u00a0\nTibamal trained & job aid | - | 93.4 (8.2) 58/62 | \u00a0 | \u00a0\nPercentage knowing fever as a symptom of uncomplicated malaria: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 66.4 (10.7) 179/263 | 68.1 (11.6) 142/205 | \u00a0 | \u00a0\nFollow up | 74.3 (8.0) 238/319 | 84.0 (6.5) 268/320 | 9.7 (2.4,17.0) | 0.0124 0.01372\nTibamal trained | - | 95.2 (7.7) 136/154 | \u00a0 | \u00a0\nTibamal trained & job aid | - | 96.9 (8.8) 53/62 | \u00a0 | \u00a0\nPercentage that would refer a suspected case of complicated malaria in a four year old child to a health facility: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 74.1 (12.0) 194/263 | 75.9 (7.3) 155/205 | \u00a0 | \u00a0\nFollow up | 83.7 (6.4) 266/319 | 86.9 (4.0) 278/320 | 3.2 (\u22122.1,8.6) | 0.2148 0.06941\nTibamal trained | - | 89.6 (7.8) 138/154 | \u00a0 | \u00a0\nTibamal trained & job aid | - | 90.9 (16.2) 58/62 | \u00a0 | \u00a0\nPercentage knowing correct dispensing practices of AL: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nAL administration | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 0.7 (1.4) 2/263 | 0 (0) 0/205 | \u00a0 | \u00a0\nFollow up | 1.2 (1.5) 4/319 | 13.0 (11.3) 44/320 | 11.7 (3.7, 19.8) | 0.0070 0.00263\nTibamal trained | - | 21.4 (18.3) 36/154 | \u00a0 | \u00a0\nTibamal trained & job aid | - | 30.7 (22.6) 16/62 | \u00a0 | \u00a0\nWhat to do if the child vomits after taking the medication | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 0 (0) 0/263 | 0 (0) 0/205 | \u00a0 | \u00a0\nFollow up | 1.5 (2.4) 4/319 | 9.4 (6.7) 30/320 | 7.9 (2.9, 12.9) | 0.0043 0.00274\nTibamal trained | - | 16.9 (10.8) 27/154 | \u00a0 | \u00a0\nTibamal trained & job aid | - | 18.2 (18.8) 15/62 | \u00a0 | \u00a0\nWhat to do if the child does not get better | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 46.4 (10.0) 120/263 | 49.0 (12.6) 99/205 | \u00a0 | \u00a0\nFollow up | 39.8 (13.6) 127/319 | 66.0 (8.3) 209/320 | 26.2 (15.0, 37.4) | 0.0001 0.00014\nTibamal trained | - | 88.8 (7.1) 136/154 | \u00a0 | \u00a0\nTibamal trained & job aid | - | 97.9 (4.2) 60/62 | \u00a0 | \u00a0\nFood to give the child with AL | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 8.3 (5.7) 22/263 | 8.6 (9.8) 18/205 | \u00a0 | \u00a0\nFollow up | 6.3 (5.9) 20/319 | 27.4 (11.2) 86/320 | 21.1 (12.1, 30.0) | 0.0001 0.00014\nTibamal trained | - | 45.1 (5.0) 71/154 | \u00a0 | \u00a0\nTibamal trained & job aid | - | 61.1 (23.1) 35/62 | \u00a0 | \u00a0\n\nn numerator, d denominator, SD standard deviation, CI confidence interval.\n1adjusted for outlet type and district; 2adjusted for outlet type; 3 adjusted for district and clinical training; 4adjusted for outlet type and district.\n\nProvider practices (mystery shopper survey) (cluster summaries from the 9 intervention and 9 control clusters)\n\u00a0 | Control% (SD) n/d | Intervention% (SD) n/d | Difference in means (95%CI) | P value Unadjusted Adjusted\nPercentage of outlets dispensing any anti-malarial: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 25.2 (8.9) (74/284) | 40.7 (9.3) (88/215) | \u00a0 | \u00a0\nFollow up | 20.2 (3.7) (68/336) | 40.3 (6.5) (127/317) | 20.1 (14.8, 25.4) | 0.0001 0.02601\nTibamal trained | \u00a0 | 56.6 (23.0) (88/144) | \u00a0 | \u00a0\nTibamal trained & job aid | \u00a0 | 70.8 (30.0) (37/58) | \u00a0 | \u00a0\nPercentage of outlets dispensing any AL: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 0.5 (1.6) (2/284) | 0.0 (0) (0/215) | \u00a0 | \u00a0\nFollow up | 1.8 (1.3) (6/336) | 25.4 (6.9) (77/317) | 23.6 (18.7, 28.6) | 0.0001 0.00012\nTibamal trained | \u00a0 | 43.3 (18.9) (67/144) | \u00a0 | \u00a0\nTibamal trained & job aid | \u00a0 | 62.7 (31.0) (31/58) | \u00a0 | \u00a0\nPercentage of outlets asking about at least one danger sign: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 26.5 (14.1) (21/284) | 21.6 (12.4) (14/215) | \u00a0 | \u00a0\nFollow up | 27.4 (10.7) (90/336) | 33.3 (14.2) (97/317) | 4.9 (\u22127.9, 17.7) | 0.4295 0.26673\nTibamal trained | \u00a0 | 46.3 (25.5) (58/144) | \u00a0 | \u00a0\nTibamal trained & job aid | \u00a0 | 51.3 (30.0) (28/58) | \u00a0 | \u00a0\nPercentage of outlets providing the correct dispensing advice for AL: | \u00a0 | \u00a0 | \u00a0 | \u00a0\nAL administration | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 0.0 (0) (0/2) | 0.0 (0) (0/0) | \u00a0 | \u00a0\nFollow up | 0.0 (0) (0/6) | 31.3 (16.3) (24/77) | 31.3 (16.8, 45.8) | 0.175\nTibamal trained | \u00a0 | 34.2 (18.0) (23/67) | \u00a0 | \u00a0\nTibamal trained & job aid | \u00a0 | 45.2 (30.5) (12/31) | \u00a0 | \u00a0\nWhat to do if the child vomits after taking the medication | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 0.0 (0) (0/2) | 0.0 (0) (0/0) | \u00a0 | \u00a0\nFollow up | 0.0 (0) (0/6) | 2.5 (3.9) (2/77) | 2.5 (\u22121.9, 6.8) | 1.000\nTibamal trained | \u00a0 | 2.8 (5.7) (2/67) | \u00a0 | \u00a0\nTibamal trained & job aid | \u00a0 | 4.4 (8.8) (2/31) | \u00a0 | \u00a0\nWhat to do if the child does not get better | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 0.5 (1.6) (1/2) | 0.0 (0) (0/0) | \u00a0 | \u00a0\nFollow up | 0.0 (0) (0/6) | 35.3 (13.1) (26/77) | 35.3 (23.6, 47.1) | 0.170\nTibamal trained | \u00a0 | 37.7 (12.6) (24/67) | \u00a0 | \u00a0\nTibamal trained & job aid | \u00a0 | 38.9 (24.1) (14/31) | \u00a0 | \u00a0\nFood to give the child with AL | \u00a0 | \u00a0 | \u00a0 | \u00a0\nBaseline | 0.0 (0) (0/2) | 0.0 (0) (0/0) | \u00a0 | \u00a0\nFollow up | 0.0 (0) (0/6) | 13.7 (14.4) (9/77) | 13.7 (0.8, 26.6) | 1.000\nTibamal trained | \u00a0 | 13.3 (14.2) (8/67) | \u00a0 | \u00a0\nTibamal trained & job aid | \u00a0 | 11.9 (14.8) (5/31) | \u00a0 | \u00a0\n\n1adjusted for outlet type and clinical training; 2adjusted for outlet type; 3adjusted for district and outlet type.\nn numerator, d denominator, SD standard deviation, CI confidence interval.\nNote: In the mystery shopper survey, a total of 499 outlets were successfully interviewed at baseline, 284 and 215 in control and intervention arms respectively, and 653 outlets at follow-up, 336 in the control and 317 in the intervention arm.",
        "label": "low",
        "id": "task4_RLD_test_347"
    },
    {
        "paper_doi": "10.1186/1475-2875-11-153",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: A phase III, randomized, open label, multi-centre, comparative study.Follow-up: Limited physical exam on Days 1-2, 3, 7, 14, 12, 28, 35, and 42 and if clinically indicated. ECG 2 to 4 hours after drug administration and on follow-up Days 7, 28, and 42. Thick and thin smears eight hourly in Days 0 to 2, then Days 3, 7, 14, 21, 28, 35, and 42. Haematology and blood chemistry at Days 3, 7, and 28. Blood spot for PCR at Day 42 or failure. Urinalysis on Day 3. HCG for women on Days 0 and 28.Adverse event monitoring: \"adverse events collected each time\" and \"Safety was assessed through direct questioning, physical examinations, ECG abnormalities (prolongation QT-interval), and significant change from baseline clinical laboratory parameters [17]. Adverse events were followed up until the event had resolved.\"\n\n\nParticipants: Number: 401 randomized (153 P. falciparum only, 90 mixed, 158 P. vivax only).Inclusion criteria: adult, absence of severe malnutrition, axillary temperature > 37.5degC or a history of fever within the preceding 24 hours, asexual P. falciparum density 1000 to 200,000/mL, P. vivax and other malaria density >= 250/mL, able to take oral treatment, informed consent, uncomplicated P. falciparum or P. vivax mono-infection, or mixed infection.Exclusion criteria: severe vomiting, history or evidence of 'clinically systematic significant disorders', other febrile conditions, hypersensitivity or adverse reactions to antimalarials, history of use of any other antimalarial agent within four weeks of the start of the trial and confirmed by urine test, and pregnancy or lactating.\n\n\nInterventions: 1. Artemesinin-naphthoquine, fixed-dose combination, 250 mg/100 mg tablets (Arco, Kunming Pharmaceutical Corporation, China):4 tablets as a single dose2. DHA-P, fixed-dose combination, 40 mg/320 mg tablets (Duo-Cotecxin: Holey-Cotec Pharmaceutical Co. Ltd, China):Daily doses of dihydroartemisinin 2 to 4 mg/kg and piperaquine 16 to 32 mg/kgweight <= 60 kg 3 tablets each day for 3 daysweight > 60 kg 4 tablets each day for 3 daysAll doses supervised.\n\n\nOutcomes: ACPR at Day 42, PCR-adjusted and PCR-unadjustedGametocyte carriageAdverse eventsFever clearanceParasite clearance\n\n\nNotes: Country: IndonesiaSetting: Three Armed Forces hospitals in Jayapura (Marthen Indeys/Army, Soedibjo Sardadi/Navy, and Bhayangkara/Police Hospitals) and one public hospital in Maumere (St. Gabriel Hospital)Transmission: Not reportedResistance: Widely reported resistance of P. falciparum to chloroquine, sulphadoxine-pyrimethamine, and quinineDates: 2007 to 2008Funding: Kunming Pharmaceutical Corporatio\n\n",
        "objective": "To evaluate the efficacy and safety of the artemisinin\u2010naphthoquine combination for treating adults and children with uncomplicated P. falciparum malaria.",
        "full_paper": "Background\nA practical and simple regimen for all malaria species is needed towards malaria elimination in Indonesia.\nIt is worth to compare the efficacy and safety of a single dose of artemisinin-naphthoquine (AN) with a three-day regimen of dihydroartemisinin-piperaquine (DHP), the existing programme drug, in adults with uncomplicated symptomatic malaria.\nMethods\nThis is a phase III, randomized, open label using sealed envelopes, multi-centre, comparative study between a single dose of AN and a three-day dose of DHP in Jayapura and Maumere.\nThe modified WHO inclusion and exclusion criteria for efficacy study were used in this trial.\nA total of 401 eligible adult malaria subjects were hospitalized for three days and randomly treated with AN four tablets single dose on day 0 or DHP three to four tablets single daily dose for three days, and followed for 42 days for physical examination, thick and thin smears microscopy, and other necessary tests.\nThe efficacy of drug was assessed by polymerase chain reaction (PCR) uncorrected and corrected.\nResults\nThere were 153 Plasmodium falciparum, 158 Plasmodium vivax and 90 P. falciparum/P. vivax malaria.\nMean of fever clearance times were similar, 13.0\u2009\u00b1\u200910.3 hours in AN and 11.3\u2009\u00b1\u20097.3 hours in DHP groups.\nThe mean of parasite clearance times were longer in AN compared with DHP (28.0\u2009\u00b1\u200911.7 hours vs 25.5\u2009\u00b1\u200912.2 hours, p\u2009=\u20090.04).\nThere were only 12 PCR-corrected P. falciparum late treatment failures: seven in AN and five in DHP groups.\nThe PCR uncorrected and corrected on day \u221242 of adequate clinical and parasitological responses for treatment of any malaria were 93.7% (95% Cl: 90.3\u201397.2) and 96.3% (95% Cl: 93.6\u201399.0) in AN, 96.3% (95% Cl: 93.5\u201399.0) and 97.3% (95% Cl: 95.0\u201399.6) in DHP groups.\nFew and mild adverse events were reported.\nAll the abnormal haematology and blood chemistry values had no clinical abnormality.\nConclusion\nAN and DHP are confirmed very effective, safe and tolerate for treatment of any malaria.\nBoth drugs are promising for multiple first-line therapy policies in Indonesia.\nBackground\nMalaria remains a major public health problem.\nThe World Health Organization (WHO) estimated the number of reported cases from Indonesia were 2.5 million in 2006.\nMost cases were recorded from Papua and East Nusa Tenggara provinces, where about 300,000 and 70,000 clinical malaria cases were reported annually from its provinces.\nTo accelerate malaria control, one of the four key recommended interventions is to give appropriate anti-malarial drugs with artemisinin-based combination therapy (ACT) for patients with confirmed malaria.\nArtemisinin was discovered by Chinese scientists in the 1970s.\nArtemisinin derivatives are the most rapidly acting and efficacious anti-malarial drugs.\nThese drugs show rapid absorption and activity against many stages of Plasmodium, including trophozoites and early sexual forms (gametocytes).\nTheir short elimination half-life (<3.7 hours) protects them from resistance and their potent and broad specificity of action reduces gametocyte carriage and infectivity, and reduces the transmission of Plasmodium falciparum.\nTolerability of these drugs is also very good.\nThis makes artemisinin derivatives the ideal partner drugs.\nWHO recommends the use of artemisinins only in combination with another anti-malarial drug to prevent the occurrence of drug resistance and to address the issue of its relatively short half-life.\nACT is the best therapeutic option for treating drug-resistant malaria and retarding the development of resistance presently.\nSince 2004, the Indonesian Malaria Control Programme has chosen a non-fixed dose artesunate-amodiaquine (AA) as the programme drug for treatment of uncomplicated P. falciparum, which had widely reported resistant to chloroquine, sulphadoxine-pyrimethamine or quinine.\nAA is also effective for Plasmodium vivax, safe for all age groups and relatively cheap.\nThough combination of AA is palatable and given by single daily dose for three days, the compliance is poor because of the number of pills to be swallowed.\nThis ACT resulted a problem in its wide-scale implementation.\nMoreover, there was reported in vitro amodiaquine cross-resistance with chloroquine.\nConsequently, the efficacy of AA was varied (78\u201396%).\nTo overcome these problems, other ACT is needed for treatment of any malaria in Indonesia.\nFixed-dose combination of artemether-lumefantrine (AL) has been registered recently in Indonesia.\nThis ACT was safe and very effective with a cure rate of 95% for P. falciparum malaria, but its efficacy was modest (43%) for P. vivax.\nAL is not a practical regimen because it should be administered twice daily for three days and given with fatty food.\nMoreover, the cost is expensive, >10 US$ per treatment course.\nA better ACT that would be simple to use, effective and affordable for all types of malaria is needed to eliminate malaria in Indonesia.\nDihydroartemisinin-piperaquine (DHP) is a fixed-dose ACT and given single daily dose for three days.\nClinical trials for treatment of uncomplicated malaria proved more superior compared with the existing fixed-dose AL, and ACT programme AA in Papua.\nThis ACT is very effective with cure rates of \u226595% for all malaria and safe.\nDHP has been used widely as the first-line ACT for more than two years in Papua.\nIn addition, the cost of DHP is similar to AA per treatment course.\nDHP trials in other areas are needed to collect unexpected adverse events.\nArtemisinin-naphthoquine (AN) is a new fixed-dose ACT.\nNaphthoquine phosphate is an anti-malarial drug synthesized by Chinese Academy of Military Medical Science in late 1980s.\nThough naphthoquine has a similar structure to chloroquine, cross-resistance has not yet been reported.\nExisting pharmacological data indicate that naphthoquine is effective against erythrocytic phase of Plasmodium, even in chloroquine resistance cases.\nNaphthoquine has a longer half-life (276 hours), compared to chloroquine and mefloquine.\nIt has been shown to be effective against P. falciparum at doses of 12 mg/kg used alone or at 400 mg in combination with artemisinin for adult patients.\nNaphthoquine appears to be an ideal partner drug for artemisinin.\nAN is administered with a single dose of therapy and has few side effects.\nOf limited data, AN was reported safe and effective for both P. falciparum and P. vivax .\nIn a study in Chinese adult subjects, one dose of a fixed-dose artemisinin (1,000 mg)-naphthoquine (400 mg) for treatment of uncomplicated malaria showed a cure rate of 98% at day 28 in P. falciparum, and 90% at day 56 in P. vivax .\nMore studies should be done to confirm its efficacy, safety, and tolerability in an enlarge scale and another population.\nIndonesia has been committed to eliminate malaria by year 2030.\nA clinical trial of AN was conducted to improve compliance and find a practical and simple ACT for all malaria species.\nBased on the previous Indonesian ACT trials, DHP is the best ACT for treatment of uncomplicated malaria in multi-drug resistant areas.\nTherefore, a clinical trial of AN was compared to DHP for its efficacy and safety in Jayapura and Maumere.\nMethods\nTime and study location\nThe trial was carried out in 2007\u20132008 at four hospitals, three Armed Forces hospitals in Jayapura (Marthen Indeys/Army, Soedibjo Sardadi/Navy and Bhayangkara/Police Hospitals), and one public hospital in Maumere (St Gabriel Hospital).\nStudy design\nThe study was a phase III, randomized, open label, multi-centre, comparative of the efficacy, safety and tolerability of a single dose AN vs DHP in adults with uncomplicated symptomatic P. falciparum, P. vivax or mixed P. falciparum/P. vivax malaria.\nThis trial was approved in writing by the Ethics Committee of National Institute of Health Research and Development, Ministry of Health (No. LB.03.02/2/449/2007), and the Bureau of Food and Drug Control (No.PO.01.01.3.1.1682), Republic of Indonesia.\nSample size\nAssuming the failure rates of AN (P1) and DHP (P2) were 0.1% and 5.0% based on un-published African AN studies prior a trial and the WHO recommendation for choosing ACT programme.\nThe study estimated risk ratio of failure rate was 2% with \u03b1 (type I error) of 0.05 and power (1-\u03b2) of 80%.\nA minimum sample size was 166 subjects per treatment group, and adjusted 20% for follow-up losses and withdrawals.\nA total 401 subjects were recruited in this trial.\nThe formula for calculating the sample size as the following\nN = The sample size of each of the treatment groups; P1 = The failure rate in the Arco\u2122 group (0.1%); P2 = The failure rate in the Duo-Cotecxin\u2122 group (5.0%); R = The risk ratio of treatment failure (P1/P2\u2009=\u20090.02).\nProcedures\nThe population of this study was adult males and females aged 15\u201369 years, body weight 35\u201375 kg and presenting with acute, symptomatic, uncomplicated P. falciparum and/or P. vivax malaria.\nThey were recruited according to the modified WHO inclusion criteria (absence of severe malnutrition, axillary temperature of \u226537.5\u00b0C or history of fever in the last 24 hours, asexual P. falciparum density 1,000\u2013200,000/\u03bcl, P. vivax and other malaria density \u2265250/\u03bcl, and ability to swallow oral medication), and exclusion criteria [severe vomiting, history or evidence of clinically systematic significant disorders, other febrile conditions, hypersensitivity or adverse reactions to anti-malarials, history of use of any other anti-malarial agent within four weeks prior to start of the study and confirmed by urine test (Dill Glazko and Lignin tests), and pregnancy or lactating] for therapeutic efficacy study.\nSubject informed consent was requested prior the study.\nSubjects who withdrew early were not replaced.\nEligible subjects were blindly, randomly assigned equally to one of the two treatment groups using sealed envelopes.\nAN (Arco\u2122, Kunming Pharmaceutical Corporation with Chinese quality standards, one tablet contained 250 mg of artemisinin and 100 mg of naphthoquine) was administered four tablets single dose only.\nDHP (Duo-Cotecxin\u2122, Holey-Cotec Pharmaceutical Co.LTd, China, one tablet contained 40 mg of dihydroartemisinin and 320 mg of piperaquine) was administered three (body weight of \u226460 kg) to four tablets (body weight of >60 kg) single daily dose for three days based on dosage of dihydroartemisinin 2\u20134 mg/kg bw or piperaquine 16\u201332 mg/kg bw.\nSubjects were observed for one hour to ensure that the medications were not vomited.\nAll subjects were hospitalized for three days or until fever and parasite had cleared for at least 24 hours, returned to study site for follow-up at all scheduled visits to day 42, and they had additional primaquine for radical treatment on day 42.\nSubjects with treatment failure were withdrawn from the study and given a rescue malaria treatment.\nThey had no study investigations performed thereafter.\nClinical and laboratory assessments\nAll eligible subjects had medical history and detail demography completed at enrolment.\nA full physical examination, electrocardiography (ECG) and laboratory tests (malaria microscopy, haematology, blood chemistry, PCR genotyping and urinalysis) were performed at baseline (day 0, prior to dose).\nLimited physical examination was performed during hospitalization (days 1\u20132) and on follow-up days (3, 7 14, 21, 28, 35, and 42) and if clinically indicated as well as adverse events collected each time.\nA 12-lead resting ECG was obtained approximately 2 to 4 hours after study drug administration on day 0\u20132, and follow up days 7, 28 and 42.\nThick and thin smears were examined at screening, days 0\u20132 hospitalization: eight hourly, days 3, 7, 14, 21, 28, 35, and 42; haematology and blood chemistry at days 0, 3, 7, and 28; and blood spot for PCR at days 0 and 42 or failure.\nUrinalysis was assessed on days 0 and 3.\nHCG was tested for women of potential pregnancy at screening and day 28.\nMicroscopy results were blind cross-checked by certified microscopists, and treatment failures were corrected by PCR.\nPCR was performed for speciation of plasmodium and genotype of P. falciparum.\nThere were 3 loci genotype of P. falciparum tested (MSP1, MSP2 and GLURP).\nThe primary (P) and nested (N) primers are as following : MSP1 (P1: 5\u2032CAC ATG AAA GTT ATC AAG AAC TTG TC3\u2032, P2: 3\u2032GTA CGT CTA ATT CAT TTG CAC G5\u2032; N1: 5\u2032GCA GTA TTG ACA GGT TAT GG3\u2032, N2: 3\u2032GAT TGA AAG GTA TTT GAC5\u2032); MSP2 (P1: 5\u2032GAA GTT AAT TAA AAC ATT GTC3\u2032, P2: 3\u2032GAG GGA TGT TGC TGC TCC ACA G5\u2032, N1: 5\u2032CTA GAA CCA TGC ATA TGT CC3\u2032, N2: 3\u2032GAG TAT AAG GAG AAG TAT G5\u2032) and GLURP (P1: 5\u2032ACA TGC AAG TGT TGA TCC3\u2032, P2: 3\u2032GAT GGT TTG GGA GTA ACG5\u2032, N1: 5\u2032TGA ATT CGA AGA TGT TCA CAC TGA AC3\u2032, N2: 3\u2032TGT AGG TAC CAC GGG TTC TTG TGG5\u2032).\nTo date, there are no recommended markers to distinguish recrudescence, relapse and new infection of P. vivax malaria from P. vivax treatment failures.\nEfficacy assessment\nThe efficacy of AN and DHP was assessed in intent-to-treat (ITT), modified ITT, and evaluable or per-protocol (PP) population.\nITT population included all randomized subjects who had received any amount of study medication.\nModified ITT population only included correctly randomized subjects in analysis, and excluded wrongly randomized and those lost to follow-up.\nPP analysis consisted only of the efficacy evaluable (EE) subjects defined according to the 2003 WHO criteria and constituted as PP population that did not include subjects who failed to comply with per-protocol.\nThe efficacy was a proportion of subjects with PCR-corrected adequate clinical and parasitological response (ACPR) at day 42.\nACPR was defined as the absence of asexual parasitaemia on day 42 irrespective of the temperature and not meeting any of criteria of early treatment failure (ETF) or late clinical or parasitological failure (LCF or LPF), or as subjects with clearance of asexual parasitaemia within 42 days of initiation of study treatment.\nSubjects classified as failures by clinical and parasitological criteria were considered ACPR if the PCR analysis showed a new infection (all the alleles in parasites from the failure-treatment sample were different from those in the admission sample, for one or more loci tested) rather than a recrudescence .\nRecrudescence was defined as reappearance of asexual parasites of the same isolate as initial infection with or without clinical signs, after initial clearance of parasites from the peripheral blood with positive blood smear and PCR confirmation of the same isolate (presence of at least one matching alleles).\nThe early and late failures were classified according to the 2003 WHO guidelines.\nThe total treatment failure was defined as the sum of early and late treatment failures.\nSafety assessment\nThe safety population was defined as all randomized subjects who had received any amount of study medication.\nSafety was assessed through direct questioning, physical examinations, ECG abnormalities (prolongation QT- interval), and significant change from baseline clinical laboratory parameters.\nAdverse events were followed up until the event had resolved.\nData analysis\nData were double entered and validated including microscopic validation and PCR corrected treatment failure data using EpiData 3.02, and analysed using SPSS for Windows version 15.\nThe Mann\u2013Whitney U-test was used for non-parametric comparisons, and Student\u2019s t-test for parametric comparisons.\nProportions were examined using \u03c72 with Yates\u2019 correction or by Fisher\u2019s exact-test.\nThe efficacy was assessed by survival analysis in which the cumulative risk of failure was calculated by the Kaplan Meier product limit formula.\nResults\nBaseline characteristics of study subjects\nIn Jayapura, over 3,000 clinical malaria cases had been screened, only 301 could be enrolled for this trial, and 151 cases were treated with AN and 150 treated with DHP.\nIn Maumere, of a total 154 screened clinical malaria, only 100 cases could participate, 50 were treated with AN and the other 50 treated with DHP.\nThe Armed Forces Hospitals contributed to 75% sample size, so were mostly male subjects.\nThere were seven subjects had weight >75 kgs (four in AN and three in DHP), 56% with axillary temperature of \u226537.5\u00b0C.\nThe characteristics of study subjects in both treatment groups were not different (Table 1).\nThe clinical symptoms of subjects in AN and DHP groups were also not different.\nFever, nausea, headache and rigors were the common symptoms documented in this study.\nOther classical symptoms are shown on Figure 1.\nThe initial laboratory findings of haematology, blood chemistry, and parasitology showed no differences between AN and DHP groups (Table 2).\nThough some of the study subjects had abnormal values, only a few had significant clinical abnormalities.\nThe distribution of malaria subjects with anaemia (Hb\u2009<\u200911 g/dL), thrombocytopaenia (platelet\u2009<\u2009150,000/ul) or leucocytosis (>10,000/ul) were almost similar between the treatment groups (52.2% vs 47.8%, 71.6% vs 73%, 6% vs 5%).\nOverall, only pallor was documented as a significant clinical abnormality with abnormal values of haematology.\nSome study subjects had higher or lower values of blood chemistry, such as alanine aminotransferase (ALT) (23.4%), aspartate aminotransferase (AST) (25.9%), bilirubin (31.4%), albumin (44.6%), urea (5.7%), sodium (23.1%), potassium (25.9%), creatinine (28.2%), and chloride (32.6%).\nOnly jaundice was documented as a significant clinical abnormality with abnormal values of bilirubin.\nThere were 153 Plasmodium falciparum (Pf), 158 Plasmodium vivax (Pv) and 90 P.falciparum/P.vivax malaria (Figure 2).\nGametocyte carriages were detected 34.4% (52 of 151) in P. falciparum, 94.3% (148 of 157) in P. vivax and 82.2% (74 of 90) in mixed P.falciparum/P. vivax, respectively.\nThe range of gametocyte densities was 1\u20132,697/ul (Table 2).\nAnalysed population\nThe ITT (401), modified ITT (384) and PP (378) population in each treatment group were microscopically cross-checked.\nThere were three of 401 study subjects not eligible after cross-checking.\nAll protocol violation cases were in AN group (two P. falciparum cases having asexual parasitaemia <1,000/ul, and one P. vivax case had taken an anti-malarial drug/chloroquine prior to the study).\nWhile all three withdrew consent, the cases were in DHP group (one parasitaemic P. falciparum case on day 0, one withdrawn P. vivax case by family of subject on day 0, and the one P. vivax case felt discomfort and dizzy by day 4).\nDuring the follow up, 17 cases were documented lost to follow up, seven (two P. falciparum cases on days 14 and 35; three P. vivax cases on days 3, 21, and 22; and two mixed P. falciparum/P. vivax cases on days 7 and 42) in AN group, and 10 [five P. falciparum cases on days 3, 7, 15, 20, and 21; four P. vivax cases on days 3 (two), 21, and 22; and one mixed P. falciparum/P. vivax case on day 3)] in DHP group ( 2 and Table 3).\nTherapeutic efficacy\nBoth study drugs had rapid fever clearance.\nOver 90% of study subjects became afebrile by the first 16 hours after the first dose of treatment, and all cleared in 56 hours (Figure 3).\nThe mean of fever clearance times (FCTs) were 13.0\u2009\u00b1\u200910.3 hours in AN and 11.3\u2009\u00b1\u20097.3 hours in DHP groups, and did not significantly differ.\nAlmost all hospitalized study subjects (>80%) were asymptomatic when discharged.\nDuring study follow-up, only mild reported symptoms (headache, dizzy, cough, abdominal pain, myalgia, sleeping disturbance, and fatigue) resolved with or without a simple symptomatic treatment.\nMost subjects (>90%) had cleared asexual parasitaemia by day 1\u201316 hours (Figure 4).\nAN had longer mean of parasite clearance time (PCT: 28.0\u2009\u00b1\u200911.7 hours) compared with DHP groups (25.5\u2009\u00b1\u200912.2 hours) (p\u2009=\u20090.04).\nOverall, mean of PCT of these ACTs was 26.7 (8\u201372) hours.\nThere was 68.8% gametocyte carriages prior ACT treatment.\nHowever, the proportion of gametocyte carriages reduced by day of follow-up.\nIt became 28.7% and 25.8% in the first 24 hours post-treatment, and 18.6% and 17% by day 3 in AN and DHP groups, respectively.\nGametocyte carriage was still detected in 1.1% subjects on completion of the study on day 42.\nOf 401 randomized study subjects, there were 19 TFs, 12 in AN and seven in DHP groups.\nNo ETF was reported.\nThere were six documented as LCFs and 13 as LPFs (Table 3).\nAll the TF cases were from Jayapura, four P. falciparum, three P. vivax and 12 P. falciparum/P. vivax malaria.\nOnly three of 19 TFs occurred by day \u226428, two in the AN and one in the DHP group.\nPCR speciation of the 19 paired samples of TF cases showed three LPFs diagnosed as P. vivax (one in AN group by day 32 and two in DHP by day 35) detected as P. falciparum, and classified as protocol violation cases.\nThere were another two LPFs by days 35 and 42 diagnosed as P. falciparum, one LPF by day 35 as mixed P. falciparum/P. vivax, and another one LCF by day 42 as mixed P. falciparum/P. vivax detected as P. falciparum new infections in AN group,\nOf the four TFs detected as new infections, two TFs had been treated with quinine plus doxycycline and classified as protocol violations at day 35.\nThe PCR-corrected treatment outcomes are shown in Table 4.\nThe overall uncorrected and corrected PCR therapeutic efficacies of both ACT for any malaria were between 89% to 95% in ITT and modified ITT population, and 94% to 97% in PP population.\nDHP had slightly higher uncorrected and corrected PCR at day 42 ACPR (96.3% and 97.3%) compared to AN (93.7% and 96.3%) for treatment of any malaria (Tables 3 and 4).\nAll TFs\u2019 PCR-corrected differences were detected in >28 day.\nTherefore, the day 28 uncorrected and corrected ACPR were similar, 95.5% (192 of 201) and 93.0% (186 of 200) in AN and DHP ITT populations; 98.0% (192 of 196) and 97.9% (186 of 190) in AN and DHP modified ITT population; 99.0% (192 of 194) and 99.5% (186 of 187) in AN and DHP PP population.\nFigures 5, 6 and 7 show the survival curves of PCR-corrected cumulative risk of failures of AN and DHP for treatment of any malaria in ITT, modified ITT and PP population.\nThe hazard ratio of risk failures were no different between treatment groups in the three populations.\nSafety\nThere were no serious adverse events reported in malaria subjects treated with AN and DHP during the study.\nOnly few (<10%) and mild symptoms as adverse events were documented.\nThe common reported adverse events were headache, dizzy, and cough.\nThere were also no clinically significant effects on myocardial electrophysiology identified through a series of ECG examinations.\nIn both study groups, haemoglobin, haematocrit, red blood cell (RBC) and platelet counts were gradually increased to normal limit by day 28.\nMeans of haematocrit and RBC were slightly decreased on day 3, and 7, and became normal by day 28.\nIn contrast, means of platelet on days 3 and 7 were increased significantly and then slightly decreased to normal value by day 28.\nThe means of haematocrit, Hb, RBC, WBC and platelet at point of investigation were not statistical significant different between treatment groups.\nOnly few leucocytosis (5.5%) were found in this trial.\nInterestingly, there were 21.4% eosinophilia (eosinophil >3%) on day 0, and increased to 38.5% by day 28.\nAll the abnormal haematological values had no significant clinical abnormality.\nThere were mild liver impairment with or without mild renal impairment on day 0 however bilirubin, albumin, ALT, AST, creatinine, urea and electrolytes were gradually improved to normal limit by day 28 in malaria subjects treated with AN and DHP.\nThe means of these blood chemistry parameters were not different between treatment groups.\nDiscussion\nACT is recommended to be given for 3 days when given with slowly eliminated partner drug.\nIn a three-day regimen, the artemisinin component is present in body during only two asexual parasite life-cycles, except for Plasmodium malariae.\nIn each asexual cycle, artemisinin and its derivatives reduce parasite numbers by a factor of approximately 10,000.\nThis anti-malarial drug is a potent and rapidly acting blood schizontocide, and gametocytocide.\nIt is still rational to study one-day ACT with a new good partner drug to improve the compliance and have a simple and practical regimen.\nMoreover, the risk of the development of de novo resistance is increased by the greater time dividing asexual parasites are exposed to drugs.\nThis makes the long-term risk of resistance developing a concern for single dose ACT.\nTherefore, this study will be useful to confirm the findings of other previous studies with different types of one-day ACT (artesunate-amodiaquine, artesunate-sulphadoxine/pyrimethamine, or artesunate-mefloquine) and in different geographical settings.\nNaphthoquine is a tetra-aminoquinoline, synthetic blood schizontocide anti-malarial drug.\nThough this anti-malarial drug has a similar structure to chloroquine, cross-resistance with chloroquine has not yet been reported.\nOf the limited clinical studies in China, naphthoquine in combination with artemisinin given in a single dose, was effective and safe with cure rates of 97.5% for treatment of P. falciparum malaria by day 28, and 90.0% for the treatment of P. vivax malaria by day 56.\nThis preliminary findings is valuable data and a good start for clarification whether the three-day regimen ACT is mandatory.\nOf the existing forms of ACT which had been studied in Indonesia, a three-day dihydroartemisinin-piperaquine (DHP) is the best alternative for the Indonesian ACT programme.\nDHP is safe and effective for both P. falciparum and P. vivax malaria.\nThe cure rates of DHP for the treatment of P. falciparum and P. vivax were reported 95.2% and 92.7%.\nWhile the cure rates of AA, the first ACT programme were only 84% and 53.5%.\nIn addition, DHP had post treatment prophylactic effect in high transmission area.\nDHP is a good comparator for a trial of a new ACT.\nCurrently, multiple first-line therapies (MFT) policies against malaria have been introduced.\nIt was developed in the context of an evolutionary-epidemiological modelling framework for malaria resistance.\nThe benefits of using MFT against malaria yields a better clinical outcome, delay the emergence and slow the fixation of resistant strains, and allow a larger fraction of the population to be treated without trading off future treatment of cases that may be untreatable because of high resistance levels.\nDespite DHP, other forms of ACT should be identified to be chosen for MFT policies, including AN.\nIn this trial, one-day single dose of fixed dose regimen of AN and three-day regimen of DHP did not result ETF for treatment of any malaria.\nAll 19 TFs were as LTFs and reported from Jayapura only.\nOf the 19 TFs, 63.2% were related with clinical symptoms and classified as LCFs.\nThough 84.2% TFs were occurred by day >28, only 21.1% TFs (4 of 19) confirmed as new infections which were detected by days 35 and 42.\nThis findings support Papua as a highly multidrug resistance area.\nA significant bigger number of samples from Jayapura (75% of a total sample) probably gave a chance detected more TF cases.\nMoreover, a relatively small number of new infections in a moderate-high transmission Papua by days 35 and 42 could be because of long half-life of both study drugs AN and DHP which known have post treatment prophylactic effect.\nACT with a long half-life partner drug is a good choice for malaria in high transmission area, however there will be also an increased risk of selecting drug-resistant isolates.\nTherefore, monitoring drug efficacy should be routinely maintained to detect the spread of drug resistance.\nIn this trial, both drugs showed very effective for treatment of any malaria.\nAN and DHP had day-42 PCR corrected ACPR of \u226590% in ITT and modified population, and >95% in PP population.\nMoreover, the day-28 and day-42 ACPRs of AN and DHP were not statistically different (p\u2009>\u20090.05).\nThese findings are consistent with previous efficacy studies of AN for treatment of uncomplicated P.falciparum malaria in China, Myanmar and Papua New Guinea, and DHP for treatment of uncomplicated P.falciparum and P.vivax malaria in Indonesia and for treatment of uncomplicated P.falciparum malaria in other Asian countries [24-28].\nSimilarly to other forms of ACT, AN and DHP cleared fever and asexual parasites rapidly with means FCT 13.0 and 11.3 hours, and PCT 28.0 and 25.5 hours.\nThese findings were similar to the previous studies.\nBoth forms of ACT also resulted 66.5% haemoglobin recovery by day 28.\nThe anaemic study subjects with Hb\u2009<\u200911 g% were significantly decreased from 23.6% and 21.6% prior treatment and became 5.9% and 9.9% by day 28 in AN and DHP groups.\nThough many factors influence the recovery of haemoglobin, a long half-life of partner drugs naphthoquine and piperaquine have an important role to prevent re-infections or relapses, which might cause anaemia.\nBoth AN and DHP were well-tolerated, with no significant ECG changes identified.\nMost of the adverse events were mild and related with symptoms attributable to malaria.\nAll symptoms were recovered with or without simple treatment.\nThere were no significant differences in tolerability between the two study drugs.\nIn this clinical trial, eosinophilia was found in one third of study subjects treated with AN or DHP.\nSeveral studies also reported that pseudoeosinophilia was associated with malaria infection detected by Sysmex XE-2100 haematology analyzer due to the presence of haemozoin-containing neutrophils.\nEosinophilia may also represent a normal late response to malaria infection.\nHowever, all eosinophilia subjects had no significant clinically abnormality.\nThere were also slightly elevations of aminotransferase (ALT and AST) in study subjects treated with AN and DHP which gradually decreased to normal limit.\nThese findings were reported similar to Chinese studies.\nThis study has shown the efficacy and safety a single dose of AN and a three-day dose of DHP for treatment of any malaria in adult subjects.\nFurther analysis of the efficacy and safety specifically in P. falciparum and P. vivax, and others malaria should be performed and published to show the detail of study findings.\nIn addition, there was a study reported a high cure rate of twice daily one day of AN (100%) versus AL (98.4%) in African children with P. falciparum malaria.\nThe dosage used in that study was based on the children weight groups.\nDosage for children is crucial and will be safe determined by per kg body weight.\nCost of drug is also important factor for choosing programme malaria drug.\nA single dose ACT, such as AN actually should be cheaper compare with three-day regimen ACT.\nMoreover, there are study limitations to extend follow up day to D42 whereas D63 is recommended for AN with a long half lives, and the small sample size and consequent loss of power.\nConclusions\nBoth fixed-dose forms of ACT are confirmed very effective, safe and tolerate for treatment of any malaria in adults, and meet with the recent WHO recommendation for replacing ineffective drugs.\nTheir longer post-treatment prophylactic effect is useful in areas where transmission is intense.\nArtemisinin-naphthoquine and dihydroartemisinin-piperaquine are promising forms of ACT for MFT policy.\nPharmacokinetic and therapeutic efficacy study in children, and cost-effectiveness study should be carried out for the safety and effectiveness of large-scale use.\nProportions of malaria symptom and sign on enrolment in artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups were no significant difference (p\u2009>\u20090.05).\nClinical trial profile of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups.\nProportions of afebrile malaria subject treated with artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups were similar by point of observation (p\u2009>\u20090.05).\nProportions of aparasitaemic (asexual) malaria subject treated with artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) groups were significantly difference by Day 0\u201316 hr (21.8% vs 55.6%, p\u2009<\u20090.0001), Day 1\u20130 hr (57.5% vs 71.3%, p\u2009=\u20090.01), and Day 1\u20138 hr (79.8% vs 88.5%, p\u2009=\u20090.03).\nCumulative risk of failure of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) in ITT population infected with any malaria. The TF by day 28 and 42 were 4.5% (9 of 201) and 10% (20 of 201) in AN group, and 7% (14 of 200) and 10% (20 of 200) in DHP group. The hazard ratio of failure between AN and DHP groups was 0.98 (95% CI: 0.53\u20131.82) with p\u2009=\u20090.95.\nCumulative risk of failure of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) in modified ITT population infected with any malaria. The TF by day 28 and 42 were 2% (4 of 196) and 6.7% (13 of 194) in AN group, and 2.1% (4 of 190) and 5.3% (10 of 190) in DHP group. The hazard ratio of failure between AN and DHP groups was 1.28 (95% CI: 0.56 \u2013 2.92) with p\u2009=\u20090.56.\nCumulative risk of failure of artemisinin-naphthoquine (AN) versus dihydroartemisinin-piperaquine (DHP) in PP population infected with any malaria. The TF by day 28 and 42 were 1% (2 of 194) and 3.7% (7 of 188) in AN group, and 0.5% (1 of 187) and 2.7% (5 of 185) in DHP group. The Hazard ratio of failure between AN and DHP groups was 1.38 (95% CI: 0.44 \u2013 4.35) with p\u2009=\u20090.58.\n\nCharacteristics of study subjects by the study drug on enrolment\nCharacteristic | Artemisinin- naphthoquine(AN) | Dihydroartemisinin- piperaquine(DHP) | P | Overall \nNumber of subjects | 201 | 200 | \u00a0 | 401\nAge:mean \u00b1 SD (range) years | 27.6\u2009\u00b1\u200910.8 (15\u201369) | 26.2\u2009\u00b1\u20099.2 (15\u201367) | 0.18 | 26.9\u2009\u00b1\u200910.0 (15\u201369)\nBody weight: mean \u00b1 SD (range) kgs | 58.7\u2009\u00b1\u20098.5 (36\u201381) | 58.5\u2009\u00b1\u20098.5 (37\u201385) | 0.86 | 58.6\u2009\u00b1\u20098.5 (36\u201385)\nAxillary temperature: mean \u00b1 SD (range) \u00b0C | 37.9\u2009\u00b1\u20091.0 (36\u201340.3) | 37.9\u2009\u00b1\u20091.1 (35\u201340.2) | 0.84 | 37.9\u2009\u00b1\u20091.1 (35\u201340.3)\nBlood Pressure Systole: mean \u00b1 SD (range) mmHg | 113\u2009\u00b1\u200913 (90\u2013150) | 114\u2009\u00b1\u200913 (80\u2013170) | 0.44 | 113\u2009\u00b1\u200913 (80\u2013170)\nDiastole: mean \u00b1 SD (range) mmHg | 72\u2009\u00b1\u20099 (50\u201392) | 73\u2009\u00b1\u20099 (40\u2013120) | 0.39 | 72\u2009\u00b1\u20099 (40\u2013120)\nHeart rate: mean \u00b1 SD (range) per min | 86\u2009\u00b1\u200913 (52\u2013126) | 87\u2009\u00b1\u200913 (56\u2013126) | 0.59 | 86\u2009\u00b1\u200913 (52\u2013126)\nRespiration rate: mean \u00b1 SD (range) per min | 20\u2009\u00b1\u20093 (16\u201336) | 20\u2009\u00b1\u20093 (16\u201332) | 0.51 | 20\u2009\u00b1\u20093.0 (16\u201336)\nSex: male:female (%) | 181:20 (90:10) | 170:30 (85:15) | 0.17 | 351:50 (87.5:12.5)\nHistory of fever in the last 24 hours (%) | 195 (97.0) | 199 (99.5) | 0.12 | 394 (98.3)\nStudy subject with T.axillary \u226537.5\u00b0C (%) | 114 (56.7) | 109 (54.5) | 0.73 | 223 (55.6)\nHistory of experience malaria in the last year (%) | 160 (79.6) | 159 (79.5) | 1.00 | 319 (79.6)\nFrequency of malaria in the last year: mean \u00b1 SD (range) times | 2.6\u2009\u00b1\u20092.3 (1\u201320) | 2.4\u2009\u00b1\u20091.6 (1\u201312) | 0.53 | 2.7\u2009\u00b1\u20093 (1\u201320)\nHistory of taken antimalarial drugs in the last 4 weeks (%) | 55 (27.4) | 57 (28.5) | 0.89 | 112 (27.9)\n\n\nHaematology, blood chemistry and parasitology findings of study subjects on enrolment by the study drug\nParameter | Artemisinin- naphthoquine (AN) | Dihydroartemisinin- piperaquine (DHP) | P | Overall\nNumber of subjects | 201 | 200 | \u00a0 | 401\nHaematocrit: mean \u00b1 SD (range)% | 36.1\u2009\u00b1\u20097.1 (15.5\u201368.9) | 36.6\u2009\u00b1\u20096.6 (19.1\u201363.4) | 0.46 | 36.4\u2009\u00b1\u20096.8 (15.5\u201368.9)\nHaemoglobin: mean \u00b1 SD (range) g/dL | 12.5\u2009\u00b1\u20092.3 (6.2\u201323.1) | 12.6\u2009\u00b1\u20092.1 (6.6\u201321.1) | 0.67 | 12.6\u2009\u00b1\u20092.2 (6.2\u201323.1)\nRed Blood Cell: mean \u00b1 SD (range) per uL | 4.5\u2009\u00b1\u20090.8 (1.5\u20137.7) | 4.4\u2009\u00b1\u20090.8 (2.2\u20137.3) | 0.70 | 4.4\u2009\u00b1\u20090.8 (1.5\u20137.7)\nPlatelet: mean \u00b1 SD (range) 103/mm3 | 116.4\u2009\u00b1\u200968.7 (1.8\u2013381.0) | 116.5\u2009\u00b1\u200965.9 (1 .0\u2013601.0) | 0.98 | 116\u2009\u00b1\u200967.2 (1.0\u2013601.0)\nWhite Blood Cell: mean \u00b1 SD (range) per uL | 6.4\u2009\u00b1\u20092.2 (2.0\u201314.0) | 7.1\u2009\u00b1\u20096.0 (2.2\u201365.0) | 0.11 | 6.8\u2009\u00b1\u20094.5 (2.0\u201365.0)\nBilirubin: mean \u00b1 SD (range) mg/dL | 1.0\u2009\u00b1\u20090.6 (0.1\u20135.2) | 1.0\u2009\u00b1\u20090.5 (0.1\u20133.8) | 0.67 | 1.0\u2009\u00b1\u20090.5 (0.1\u20135.2)\nAlbumin: mean \u00b1 SD (range) g/dL | 4.1\u2009\u00b1\u20091.0 (2.0\u20137.4) | 4.2\u2009\u00b1\u20091.0 (2.2\u20136.2) | 0.25 | 4.2\u2009\u00b1\u20091.0 (2.1\u20137.4)\nALT (SGPT): mean \u00b1 SD (range) IU/L | 28.8\u2009\u00b1\u200913.0 (5\u201373) | 28.3\u2009\u00b1\u200914.2 (6.9\u201398.0) | 0.71 | 28.5\u2009\u00b1\u200913.9 (5.0\u201398.0)\nAST (SGOT): mean \u00b1 SD (range) IU/L | 29.4\u2009\u00b1\u200914.0 (2.0\u201387.0) | 28.5\u2009\u00b1\u200911.8 (2.6\u201391.0) | 0.49 | 28.9\u2009\u00b1\u200912.9 (2.0\u201391.0)\nCreatinine: mean \u00b1 SD (range) mg/dL | 1.0\u2009\u00b1\u20090.4 (0.2\u20132.5) | 0.9\u2009\u00b1\u20090.3 (0.2\u20132.0) | 0.15 | 0.9\u2009\u00b1\u20090.4 (0.2\u20132.5)\nUrea: mean \u00b1 SD (range) mg/dL | 27.5\u2009\u00b1\u200911.7 (2.7\u201380.0) | 27.6\u2009\u00b1\u200911.7 (8.6\u2013110.0) | 0.94 | 27.6\u2009\u00b1\u200911.7 (2.7\u2013110.0)\nDensity of asexual parasites: geometric mean (range) per uL | 6310 (304\u2013113550) | 6972 (372\u2013140084) | 0.21 | 6634 (304\u2013140084)\nGametocyte carriages (%) | 134 (67.3) | 140 (70.3) | 0.54 | 274 (68.8)\nDensity of gametocytes: geometric mean (range) per uL | 40 (1\u20132697) | 35 (2\u20132208) | 0.55 | 38 (1\u20132697)\n\n\nPCR uncorrected efficacy of AN vs DHP in all malaria, Indonesia\nOutcome | Artemisinin- naphthoquine (AN) | Dihydroartemisinin-piperaquine (DHP) | P | Overall\nACPR/Treatment Success (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 179 [89.1 (84.7\u201393.4)] | 180 [90.0 (85.8\u201394.2)] | 0.76 | 359 [89.5 (86.1\u201392.3)]\nModified ITT | 179 [92.3 (88.5\u201396.0)] | 180 [94.7 (91.6\u201397.9)] | 0.33 | 359 [93.5 (90.5\u201395.7)]\nPer Protocol | 179 [93.7 (90.3\u201397.2)] | 180 [96.3 (93.5\u201399.0)] | 0.26 | 359 [95.0 (92.3\u201396.9)]\nLate Clinical Failure-LCF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 5 [2.5 (0.3\u20134.6)] | 1 [0.5 (0.5\u20131.5)] | 0.10 | 6 [1.5 (0.6\u20133.2)]\nModified ITT | 5 [2.6 (0.3\u20134.8)] | 1 [0.5 (0.5\u20131.6)] | 0.10 | 6 [1.6 (0.6\u20133.4)]\nPer Protocol | 5 [2.6 (0.4\u20134.9)] | 1 [0.5 (0.5\u20131.6)] | 0.10 | 6 [1.6 (0.6\u20133.4)]\nLate Parasitological Failure-LPF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 7 [3.5 (0.9\u20136.0)] | 6 [3.0 (0.6\u20135.4)] | 0.78 | 13 [3.2 (1.7\u20135.5)]\nModified ITT | 7 [3.6 (1.0\u20136.2)] | 6 [3.2 (0.7\u20135.6)] | 0.81 | 13 [3.4 (1.8\u20135.7)]\nPer Protocol | 7 [3.7 (1.0\u20136.3)] | 6 [3.2 (0.7\u20135.7)] | 0.81 | 13 [3.4 (1.8\u20135.8)]\n\u201cOther Failures\u201d (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 10 [5.0 (2.0\u20138.0)] | 13 [6.5 (3.1\u20139.9)] | 0.51 | 23 [5.7 (3.7\u20138.5)]\nModified ITT | 3 [1.5 (0.2\u20133.3)] | 3 [1.6 (0.2\u20133.4)] | 0.98 | 6 [1.6 (0.6\u20133.4)]\nPer Protocol | 0 | 0 | \u00a0 | 0\n\nAll protocol violations, withdrawn consents, and lost to follow up were classified as \u201cother failures\u201d in ITT analysis.\nAll protocol violations and withdrawn consents were classified as \u201cother failures\u201d in modified ITT analysis.\n\nPCR corrected efficacy of AN vs Duocotecxin in all malaria, Indonesia\nOutcome | Artemisinin- naphthoquine (AN) | Dihydroartemisinin-piperaquine (DHP) | P | Overall\nACPR/Treatment Success (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 181 [90.0 (85.9\u201394.2)] | 180 [90.0 (85.8\u201394.2)] | 0.99 | 361 [90.0 (86.7\u201392.8)]\nModified ITT | 181 [93.3 (89.8\u201396.8)] | 180 [94.7 (91.6\u201397.9)] | 0.55 | 361 [94.0 (91.1\u201396.2)]\nPer Protocol | 181 [96.3 (93.6\u201399.0)] | 180 [97.3 (95.0\u201399.6)] | 0.58 | 361 [96.8 (94.4\u201398.3)]\nLate Clinical Failure-LCF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 4 [2.0 (0.1\u20133.9)] | 1 [0.5 (0.5\u20131.5)] | 0.18 | 5 [1.2 (0.4\u20132.9)]\nModified ITT | 4 [2.1 (0.1\u20134.1)] | 1 [0.5 (0.5\u20131.6)] | 0.18 | 5 [1.3 (0.4\u20133.0)]\n3. Per Protocol | 4 [2.1 (0.1\u20134.2)] | 1 [0.5 (0.5\u20131.6)] | 0.18 | 5 [1.3 (0.4\u20133.1)]\nLate Parasitological Failure-LPF (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 3 [1.5 (0.2\u20133.2)] | 4 [2.0 (0.1\u20133.9)] | 0.70 | 7 [1.7 (0.7\u20133.6)]\nModified ITT | 3 [1.6 (0.2\u20133.3)] | 4 [2.1 (0.1\u20134.1)] | 0.68 | 7 [1.8 (0.7\u20133.7)]\nPer Protocol | 3 [1.6 (0.2\u20133.4)] | 4 [2.2 (0.1\u20134.3)] | 0.69 | 7 [1.9 (0.8\u20133.8)]\n\u201cOther Failures\u201d (%, 95% CL) | \u00a0 | \u00a0 | \u00a0 | \u00a0\nITT | 13 [6.5 (3.1\u20139.9)] | 15 [7.5 (3.8\u201311.1)] | 0.68 | 28 [7.0 (4.7\u20139.9)]\nModified ITT | 6 [3.1 (0.7\u20135.5)] | 5[2.6 (0.4\u20134.9)] | \u00a0 | 11 [2.9 (1.4\u20135.1)]\nPer Protocol | 0 | 0 | 0.79 | 0\n\nAll protocol violations, withdrawn consents, and lost to follow up were classified as \u201cother failures\u201d in ITT analysis.\nAll protocol violations and withdrawn consents were classified as \u201cother failures\u201d in modified ITT analysis. PCR corrected was carried out only for P.falciparum recurrences.",
        "label": "unclear",
        "id": "task4_RLD_test_1045"
    },
    {
        "paper_doi": "10.1371/journal.pone.0046548",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 120 households, 599 individuals, 121 children < 2Inclusion criteria: mothers who disclosed their HIV status, had a child 6-12 months old, and permanently resided in the catchment area\n\n\nInterventions: Filter (LifeStraw(r) Family); two 5 L storage vessels (61 households, 299 individuals, 61 children < 2)Primary drinking supply (59 households, 300 individuals, 60 children < 2)\n\n\nOutcomes: Use of filterMicrobiological water qualityLongitudinal diarrhoeal prevalenceWeight-for-age Z-scores\n\n\nNotes: Location: two peri-urban neighbourhoods, Chongwe district, ZambiaLength: 12 monthPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "Background\nUnsafe drinking water presents a particular threat to people living with HIV/AIDS (PLHIV) due to the increased risk of opportunistic infections, diarrhea-associated malabsorption of essential nutrients, and increased exposure to untreated water for children of HIV-positive mothers who use replacement feeding to reduce the risk of HIV transmission.\nThis population may particularly benefit from an intervention to improve water quality in the home.\nMethods and Findings\nWe conducted a 12-month randomized, controlled field trial in Zambia among 120 households with children <2 years (100 with HIV-positive mothers and 20 with HIV-negative mothers to reduce stigma of participation) to assess a high-performance water filter and jerry cans for safe storage.\nHouseholds were followed up monthly to assess use, drinking water quality (thermotolerant coliforms (TTC), an indicator of fecal contamination) and reported diarrhea (7-day recall) among children <2 years and all members of the household.\nBecause previous attempts to blind the filter have been unsuccessful, we also assessed weight-for-age Z-scores (WAZ) as an objective measure of diarrhea impact.\nFilter use was high, with 96% (596/620) of household visits meeting the criteria for users.\nThe quality of water stored in intervention households was significantly better than in control households (3 vs. 181 TTC/100 mL, respectively, p<0.001).\nThe intervention was associated with reductions in the longitudinal prevalence of reported diarrhea of 53% among children <2 years (LPR\u200a=\u200a0.47, 95% CI: 0.30\u20130.73, p\u200a=\u200a0.001) and 54% among all household members (LPR\u200a=\u200a0.46, 95% CI: 0.30\u20130.70, p<0.001).\nWhile reduced WAZ was associated with reported diarrhea (\u22120.26; 95% CI: \u22120.37 to \u22120.14, p<0.001), there was no difference in WAZ between intervention and control groups.\nConclusion\nIn this population living with HIV/AIDS, a water filter combined with safe storage was used correctly and consistently, was highly effective in improving drinking water quality, and was protective against diarrhea.\nTrial Registration\nClinicaltrials.gov NCT01116908\nIntroduction\nUnsafe drinking water is a major cause of diarrheal death and disease, especially for young children in low-income countries and people living with HIV/AIDS (PLHIV).\nThe 33 million PLHIV worldwide - including almost 1 million living in Zambia - are especially vulnerable to diarrheal disease caused by opportunistic infections from waterborne pathogens, such as Cryptosporidium spp..\nDiarrheal disease may lead to intestinal malabsorption so that PLHIV on antiretrovirals (ARVs) are not acquiring their essential nutrients and therapeutic dosages of medications.\nFurthermore, diarrheal disease and unsafe drinking water may be particularly debilitating for children born to HIV-positive mothers.\nYoung children born to HIV-positive mothers are at greater risk of mortality, morbidity, and malnutrition, which may be aggravated by enteric infection.\nSafe water is critical for HIV-positive mothers who choose to replacement feed in order to prevent transmission of the virus via breast milk; \u201csafe water and sanitation\u201d is the first condition for replacement feeding in the new World Health Organization (WHO) guidelines.\nCurrent WHO guidelines for infant feeding for HIV-positive women recommend that virtually all women breastfeed their children for up to 2 years while either the mother or child is on ARVs; the risks of diarrheal disease and malnutrition outweigh the risks of HIV transmission in the majority of low-income settings.\nEven for mothers who choose to breastfeed, infants may be exposed to waterborne pathogens; exclusive breastfeeding is less common among HIV-positive mothers and water treatment has been found to reduce diarrhea even among breastfed children.\nFinally, young children who do contract the HIV virus will be more susceptible to water-related pathogens because of a weakened immune system and may particularly benefit from improved environmental conditions.\nOur previous research in Zambia found that children <2 years born to HIV-positive mothers are particularly at risk of diarrheal disease.\nIn our cross-sectional study, 26% of children <2 years had diarrhea in the past week and bacterial contamination of drinking water was found in 70% of households.\nChildren were more likely to have diarrhea if they had been given water in the past two days, suggesting that diarrheal disease may be at least partially attributable to unsafe drinking water.\nAdditionally, diarrhea in children was significantly associated with mother's diarrhea, which is of particular concern in HIV-affected areas; mothers with HIV may be more likely to have diarrhea and consequently more likely to pass diarrhea onto their children.\nTherefore, for children born to HIV-positive mothers in low-income settings, water quality interventions may be particularly critical.\nImproving household drinking water quality through household water treatment and safe storage (HWTS) has been shown to have the potential to significantly reduce diarrheal disease.\nInternational organizations including USAID, the World Bank, and WHO have recently called for an integration of water and sanitation activities in HIV/AIDS programs, and the number of programs including HWTS for PLHIV is increasing.\nHowever, despite these programs, there is relatively little evidence demonstrating the health impact or examining use of HWTS interventions for PLHIV.\nOnly one study has assessed the health impact of HWTS for PLHIV in a low-income setting in the form of a randomized, controlled trial.\nThis trial in Uganda found that PLHIV with a household chlorination technology had 25% fewer diarrhea episodes and 33% fewer days with diarrhea compared to the control group, though diarrhea reductions were not significant for children under five.\nOther observational studies of household chlorination interventions have found significant associations with diarrhea reductions in Nigeria among adults with HIV/AIDS and in Kenya among infants born to HIV-positive mothers.\nHowever, these studies and the majority of HWTS programs for PLHIV have been in the form of chlorination products, which do not inactivate or remove the full array of waterborne pathogens (such as Cryptosporidium spp.) unless combined with other treatment mechanisms.\nFurthermore, there are questions about whether HWTS interventions are used correctly and consistently over an extended period of time; this study is primarily designed to examine HWTS use, which is vital to the success of HWTS programs.\nWe undertook a randomized controlled trial to assess a gravity water filter combined with local jerry cans for safe storage.\nSpecifically, we examined 1) the use of the HWTS, both for children <2 years and all household members, 2) the microbiological performance of the HWTS intervention, measured as thermotolerant (fecal) coliforms (TTC), a well-established WHO indicator organism for fecal contamination, and 3) the impact of the intervention on the longitudinal prevalence of diarrhea among children <2 years and all household members, measured both as reported by the primary caretaker and by the weight-for-age z-score (WAZ) of children <2 years \u2014a potential measure for reported diarrhea.\nMethods\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nStudy Design and Sample Size\nA randomized, controlled trial was designed to assess use, microbiological performance, and health impact of a household filtration intervention over 1 year.\nThis study followed an open (non-blinded) design because previous attempts to blind the same intervention (LifeStraw Family filter) in the Congo were unsuccessful; the \u201cplacebo\u201d provided to control households removed approximately 1 log (90%) of fecal contamination, potentially due to the formation of a biofilm, and the authors concluded that blinding this filter is not likely to be possible.\nWe estimated a sample size of 50 households per arm (100 total) would allow us to estimate use with a precision of at least \u00b115% with 20% loss to follow-up, assuming at least 70% use.\nAdditionally, 10 HIV-negative mothers and their households were included in each arm (20 total, an additional 20% of households).\nThis figure represents a balance between the need to reduce potential stigma of participation and the cost and inconvenience to additional participants.\nBecause recruitment occurred over an eight month period, the length of possible follow-up depended on the time of enrollment, up to 12 months.\nWith this sample size, we had 80% power to detect a 40% reduction in diarrhea prevalence.\nStudy Location\nFrom our previous work, Chongwe District, Zambia was identified for this study based on the lack of piped water supply systems, inadequate water quality, and presence of active health clinics.\nThe project sites included two neighboring compounds, Kasisi and Ngwerere in Chongwe District, both approximately 30 min\u20131 hour from central Lusaka, Zambia.\nNeither Kasisi nor Ngwerere were serviced by municipal piped water systems at the time of this study.\nParticipant Eligibility and Enrollment\nChildren <2 years born to eligible HIV-positive mothers were targeted by recruiting and enrolling their mothers.\nWomen were eligible to participate in the trial if they (i) had a child aged 6 to 12 months at the beginning of the trial, (ii) reported that they were HIV-positive (or HIV-negative) confirmed with antenatal clinic records and willing to disclose their status to our study team, and (iii) resided in a household located within the catchment areas of the Ngwerere or Kasisi health clinics in Chongwe district, Zambia and did not plan to move in the next 12 months.\nHealth clinic staff identified potentially eligible women consecutively through under-five clinics and ART programs at their respective health clinics and referred them to our field team.\nHIV status of the children <2 years was recorded as reported by the mother.\nIntervention\nEach intervention household received one LifeStraw Family filter and two 5-L safe storage containers.\nThe LifeStraw Family is a novel HWTS filtration technology developed by Vestergaard-Frandsen SA that uses ultrafiltration in the form of a hollow-fiber cartridge to remove pathogens from drinking water.\nTo operate, untreated (influent) water is poured into a 2.5 L container, flows down a 1 m long tube designed to provide head pressure, and through the ulrafiltration cartridge where is it dispensed via tap (effluent).\nIn addition to the filter, we provided two locally-procured 5-L jerry cans (Merco Ltd, Ndola, Zambia) for safely storing water following treatment.\nHouseholds that were allocated to the intervention group received the filter and training on use and maintenance by our fieldworkers, who were previously trained by the filter manufacturer.\nHouseholds allocated to the control group were instructed to continue usual practices throughout the study and were allocated filters and storage containers with subsequent training at the end of the study in August 2011.\nBaseline survey and randomization\nAt enrollment, baseline data were collected on demographics, sanitation facilities, hygiene practices, water sources and treatment practices, and feeding practices for children <2 years.\nFor each household, baseline water samples were collected from drinking water sources and stored drinking water in the home.\nHouseholds were randomly allocated using a computer random number generator to either a) the intervention group receiving the LifeStraw Family filter and storage containers, or b) the control group.\nThe randomization was stratified by maternal HIV-status and catchment area (either Ngwerere or Kasisi) in blocks of 8 maximum.\nThe randomization was conducted by the trial manager (RP) who was not involved in the enrollment of participants, and fieldworkers were not involved in the randomization.\nParticipants were recruited from April\u2013December 2010, and followed for 7\u201312 months depending on time of recruitment.\nHouseholds were considered to have completed the trial that continued until July 2011, regardless of the time of recruitment; total possible follow-up visits were calculated based on the time from enrollment until July 2011.\nHouseholds were visited monthly; visits were unannounced and the field team made a repeat visit if the mother was not at home.\nAlthough we cannot rule out the potential of courtesy bias assessments of compliance, we took steps to minimize this by making all visits unannounced and sampling water quality, an objective measure.\nOutcome Measures\nUse\nHouseholds were followed monthly to obtain information on filter use and acceptability.\nHouseholds were classified as \u201creported users\u201d if 1) the filter was observed in household at the time of visit, 2) the storage vessel contained water reported to be treated at the time of visit, and 3) the respondent reported using the filter on the day of or day prior to the day of visit.\nHouseholds were classified as \u201cconfirmed users\u201d if, in addition to these three criteria, there was at least a 1 log10 TTC improvement in their stored household water over their unfiltered water, or stored water quality was <10 TTC/100 mL.\n\u201cExclusive use\u201d was defined as not drinking any unfiltered water in the day of or day prior to the interview as reported by the mother.\nThe acceptability of the technology was evaluated through monthly household surveys.\nWater Quality\nWater samples were collected during monthly visits.\nFor the stored drinking water, the respondent was asked if there was any drinking water in the house and samples were collected from the vessel that the householder identified for drinking.\nFor control households, only stored drinking water was collected.\nFor intervention households, water samples were collected of i) unfiltered water stored in the home (influent water), ii) filtered water immediately after filtration (effluent water), and iii) stored water that the household reported to be filtered, if available.\nSamples (125-mL) were collected in sterile Whirl-Pak\u2122 Bags (Nasco International, Fort Atikinson, WI, USA) containing a tablet of sodium thiosulfate to neutralize any disinfectant, placed on ice, and processed within 4 hours of collection to assess levels of TTC/100 mL at the University Teaching Hospital, Zambia.\nMicrobiological assessment was performed using a membrane filtration method with membrane lauryl suphate medium using using a DelAgua field incubator (Robens Institute, University of Surrey, Guildford, Surry, UK) in accordance with the Standard Methods .\nAfter piloting the assay procedures, we elected to use full 100 ml samples for filtered and filtered & stored samples (intervention households) and 10-fold diluted samples for unfiltered samples (intervention and control households) to minimize the number of samples that yielded plates with colonies that were too numerous to count (TNTC).\nWhere plates were TNTC, we ascribed a value of 500 TTC to such plates; this is a conservative estimate of the upper detection limit considering up to 1500 TTC were counted per plate.\nBaseline samples were also tested for free and total chlorine residuals using a Hach color-wheel test kit (Hach Company, Loveland, CO, USA).\nDiarrhea Longitudinal Prevalence\nAt all monthly visits, the mother was asked whether each household member experienced any diarrhea in the past 7 days.\nDiarrhea was measured as longitudinal prevalence (the proportion of weeks with diarrhea divided by the number of weeks under observation).\nDiarrhea was defined according to the WHO definition of 3 or more loose stools within a 24-hour period.\nMothers who reported diarrhea were also asked whether the episode extended for 14 days or longer in order to assess persistent diarrhea.\nWeight-for-Age Z-scores (WAZ)\nChildren <2 years were weighed during monthly visits on baby scales (Seca Model 384, Chasmors, London, UK) according to standard protocol.\nDuring weight measurements, children were only wearing a minimum of light clothing without shoes.\nChildren were weighed a minimum of twice during every visit to verify the weight measurement; if the two measurements were not equal (particularly from child movement), the child was weighed a third time and the confirmed weight was recorded.\nDate of birth was verified on the child's health card to calculate WAZ.\nData management and analysis\nData were double-entered into EpiData 3.1 and analyzed using Stata 12.\nThe analysis plan was finalized before the data were examined.\nWAZ scores were calculated using the WHO growth reference data.\nSocioeconomic status was measured using an asset index created by combining data on household possessions and characteristics based on asset questionnaires used in the Zambia Demographic and Health Survey.\nData were analyzed on an intention-to-treat basis in order to estimate the effectiveness of supplying households with the intervention, regardless of filter use.\nThe data from households with HIV-negative mothers were included in all analyses unless stated otherwise.\nTo assess acceptability and filter use, we tabulated data for all visits combined, and separately for the \u2018final\u2019 visit, defined as the final follow-up visit for households that completed the trial.\nTo assess the effect of the intervention on water quality, TTC counts during follow-up were compared using random effects linear regression to account for repeated observations within households.\nTTC counts were normalized with log10 transformations; a value of 1 was added to all TTC levels before transformation to account for samples with TTC values of zero, log10(TTC level+1).\nMicrobiological filter performance was calculated as the difference of the log of the influent concentration and log of the effluent concentration.\nAll water quality analyses assumed that intervention households were drinking unfiltered water if stored filtered water was not available at the time of visit.\nThe effect of the intervention on diarrhea longitudinal prevalence was examined using binomial regression with a log link function and robust standard errors, with generalized estimating equations (GEE) to account for correlation of repeated measures within individuals.\nIn the analysis of diarrhea for all household members, we accounted for clustering at the household level, since this adequately accounted for within-individual correlation.\nThe effect of the intervention on WAZ was assessed using random effects linear regression to account for repeated observations within individuals.\nIn a secondary analysis we controlled for WAZ at baseline.\nTo examine the relationship between WAZ and diarrhea, we used random effects linear regression to account for repeated measures and adjusted for baseline WAZ.\nTo assess the relationship between water quality and diarrhea longitudinal prevalence, we used binomial regression with a log link function and robust standard errors with GEE to account for correlation of repeated measures.\nWater quality results were transformed to log10(TTC level+1), to account for samples with TTC values of zero.\nAdjusted analyses controlled for age and trial arm, since both were strongly associated with diarrhea.\nPredicted probabilities of diarrhea from the unadjusted and adjusted models were calculated at fixed values of log10 TTC and plotted.\nWe used fractional polynomials to examine the shape of the relationship of water quality (log10 TTC) with log diarrhea prevalence, using a set of defined powers (\u22122, \u22121, \u22120.5, 0.5, 1, 2 and ln(x)) and a maximum of two power terms in the model.\nModels were adjusted for intervention arm.\nThe differences in model deviances were compared; the linear model was used if the improvement in fit was not statistically significant at p<0.05.\nThe relationship between water quality and WAZ was assessed with random effects linear regression accounting for repeated measures and adjusted for baseline WAZ; adjusting for baseline WAZ accounts for genetic variability and events prior to the intervention.\nTo examine the effect of the intervention on mortality, we used a Cox Proportional hazard model to estimate mortality rates.\nEthics\nThis study was approved by the Biomedical Research Ethics Committee of the University of Zambia and the London School of Hygiene and Tropical Medicine Ethics Committee, and registered with clinicaltrials.gov (NCT01116908).\nParticipants were provided with verbal and printed details of the study in the local language; informed, written consent was obtained from all participating mothers for their respective households.\nMeasures were taken to ensure confidentiality for all participants.\nIf members of participating households were found to be in need of health care during the household visits, they were referred to health clinics.\nAt the conclusion of the study, the results were disseminated to all participants in community meetings, and all control households received the intervention.\nBesides the intervention, households were not given incentives to participate.\nResults\nStudy Population\n141 mothers were screened; 17 (12%) were ineligible and 4 (3%) refused to participate (Figure 1).\nOf the 120 households enrolled, 59 (49%) were allocated to the control group and 61 (51%) households were allocated to the intervention arm.\nOne household in the control arm had twins; a total of 121 children <2 years were included.\n91/120 (76%) households were enrolled for 12 months, the remaining were enrolled for 7\u201311 months; 101/120 (84%) households completed follow-up.\nHousehold loss-to-follow up was 16%, primarily due to participants moving out of the study area, and did not vary significantly by trial arm (p\u200a=\u200a0.47).\nThere were 3/61 (5%) deaths in children <2 years in the intervention arm and 6/60 (10%) in the control (p\u200a=\u200a0.28).\nAmong children <2 years, data were collected for 82% (1138/1382) of possible person-weeks of diarrhea.\nBaseline characteristics were distributed evenly between the trial arms, with the exception of mother's marital status, sex of child <2 years, and reported diarrhea (Table 1).\nOnly 12% (14/121) of children <2 years were reported to be HIV-positive, 50% (61/121) were negative, and 38% (46/121) had not been tested by the end of our study.\nFilter use\nMost households used the filters throughout the study (Table 2).\nHouseholds were classified as reported users in 96% (596/620) of all household visits and as confirmed users in 87% (540/620) visits.\nIf we were to restrict our definition of confirmed user to only those that had at least 1 log10 removal of TTC, 82% (507/620) of intervention households would still be considered confirmed users.\nAmong households that did not meet the criteria of confirmed users, 4% (24/620) visits had stored water of somewhat better water quality compared to unfiltered water (<1 log10) and therefore may have been actually using the filter.\nIn instances when households did not have stored filtered water at the time of visit (3% of all visits, 16/622) the mother reported that she did not have time to filter the water.\nOnly 3/61 (<5%) of filters had to be replaced during the study; 1 clogged and 2 were eaten by rats along the filter tubing\nMothers reported exclusively using the filters in 95% (591/624) of all visits.\nFor children <2 years, exclusive use was reported in 93% (171/184) of all visits.\nReasons for not using the filter exclusively were that the mother or children were away from home, such as visiting relatives or at church.\nAlmost all households (>99%, 623/625 visits) reported using the storage containers provided to store filtered water.\nResults at the final visit were similar to those at all visits (Table 2).\nWater Quality\nUnfiltered water had a geometric mean of 190 TTC/100 mL (95% CI: 147\u2013245 TTC/100 mL), with 60.3% (720/1194) of samples over 100 TTC/100 mL (Figure 2). 3.3% of unfiltered intervention group water samples and 4.5% of unfiltered control group water samples yielded plates that were TNTC; no filtered samples and filtered and stored samples resulted in TNTC plates.\nUnfiltered water did not differ significantly between the intervention and control groups (geometric mean 199 vs. 181 TTC/100 mL, respectively, p\u200a=\u200a0.61).\nIn intervention households, water quality was significantly better in filtered water (geometric mean of 1.2 TTC/100 mL; 95% CI: 1.1\u20131.2 TTC/100 mL) and stored filtered water (geometric mean of 2.7 TTC/100 mL; 95% CI: 2.3\u20133.0 TTC/100 mL) compared with unfiltered water (Figure 2).\nThe quality of stored drinking water was significantly better in intervention households than control households (geometric mean 3 vs. 181 TTC/100 mL, respectively, p<0.001).\nIn intervention households, the geometric mean removal from influent (unfiltered) to effluent was 2.2 log10 TTC/100 mL (95% CI: 2.1\u20132.3 log10 TTC/100 mL), corresponding to a 99.4% (95% CI: 99.3\u201399.5%) reduction.\nReported Diarrhea\nDiarrhea longitudinal prevalence in children <2 years was 13.6% (72/530) in the control arm and 6.6% (40/608) in the intervention arm, representing a 53% reduction (longitudinal prevalence ratio, LPR\u200a=\u200a0.47, 95% CI: 0.30\u20130.73, p\u200a=\u200a0.001) (Table 3 and Figure 3).\nWhen restricted to children of HIV-positive mothers, the intervention was associated with a 50% reduction in diarrhea (LPR\u200a=\u200a0.50, 95% CI: 0.31\u20130.80, p\u200a=\u200a0.004).\nFor all household members, diarrhea longitudinal prevalence was 3.5% (101/2906) in the control group and 1.6% (50/3168) in the intervention (LPR\u200a=\u200a0.46, 95% CI: 0.30\u20130.70, p<0.001).\nDiarrhea was classified as persistent (\u226514 days) in 26.2% (39/149) of reported weeks with diarrhea for all household members and 27.0% (30/111) of reported weeks with diarrhea for children <2 years (Table 3).\nMost persistent diarrhea occurred in children <2 years (76.9%, 30/39), and the 5 people who had more than one visit with persistent diarrhea were all children <2 years.\nThe intervention resulted in reductions in persistent diarrhea for children <2 years (LPR\u200a=\u200a0.63, 95% CI: 0.28\u20131.40, p\u200a=\u200a0.26) and all household members (LPR\u200a=\u200a0.75, 95% CI: 0.37\u20131.53, p\u200a=\u200a0.43) though results were not statistically significant.\nWeight-for-age z-scores (WAZ) in children <2 years\nThere was no evidence of a difference between the intervention and control groups in mean WAZ scores (\u22121.21 vs. \u22121.24, respectively, p\u200a=\u200a0.92).\nAdjusting for baseline WAZ did not change this conclusion (\u22121.18 vs. \u22121.31, respectively, p\u200a=\u200a0.85).\nChildren with concurrent diarrhea had lower average WAZ scores compared to children without diarrhea (\u22121.46 vs. \u22121.20, respectively, p<0.001).\nAfter adjusting for WAZ at baseline, mean WAZ scores among children <2 years with diarrhea were 0.26 lower than in children without diarrhea (95% CI: \u22120.37 to \u22120.14, p<0.001).\nWater Quality, Diarrhea, and WAZ\nThere was a suggestion of a positive trend between diarrhea prevalence and household fecal water contamination (Figure 4).\nThe results of the fractional polynomial models showed that the linear model adequately described the relationship between log diarrhea prevalence and log10 TTC.\nThis relationship was significant for all household members (age-adjusted LPR for the increase in prevalence with log10 TTC\u200a=\u200a1.29, 95% CI: 1.14\u20131.45, p<0.001), and for children <2 years (age-adjusted LPR for log10 TTC\u200a=\u200a1.20, 95% CI: 1.05\u20131.39, p\u200a=\u200a0.01).\nThough adjusting for trial arm attenuated the association between water quality and diarrhea, there was still weak evidence of an effect (age- and arm-adjusted LPR for log10 TTC\u200a=\u200a1.15, 95% CI: 0.99\u20131.33, p\u200a=\u200a0.07 for all household members; age- and arm-adjusted LPR for log10 TTC\u200a=\u200a1.09, 95% CI: 0.92\u20131.28, p\u200a=\u200a0.33 for children <2 years).\nIn contrast, there was no evidence of an association of water quality and WAZ (mean change in WAZ for log10 TTC\u200a=\u200a0.00, 95% CI: \u22120.05 to 0.04, p\u200a=\u200a0.93); adjusting for trial arm did not change this conclusion.\nMortality of children <2 years\nDuring the study, there were 9 deaths, all in children <2 years; 3/61 (5%) in the intervention arm and 6/60 (10%) in the control.\nThe cause of death was recorded as reported by the primary caregiver.\nIn the intervention arm, only one death was gastrointestinal (reported as diarrhea/vomiting); other deaths were from respiratory illness and consuming rat poison.\nIn the control arm, deaths were potentially all gastrointestinal-related (diarrhea/vomiting, diarrhea/malnutrition [3 children], diarrhea/coughing, and malnutrition).\nAll but one were children born to HIV-positive mothers, and two children were known to be HIV-positive.\nThere was no evidence of an impact of the intervention on all-cause mortality among children <2 years (RR\u200a=\u200a0.56; 95% CI: 0.13\u20132.37, p\u200a=\u200a0.43), though the study was not designed to detect a difference in mortality as an outcome.\nDiscussion\nTo our knowledge, this study is the first randomized controlled trial to examine a HWTS intervention among HIV-positive mothers with young children.\nOur findings suggest that the intervention was used correctly and consistently, was highly effective in improving drinking water quality, and was protective against diarrhea.\nFilter use was particularly high in our study; households were using the filters in 96% of visits and use was further confirmed with water quality testing in 87% of visits.\nSome of the households that did not meet the water quality testing criterion for confirmed use may have been actually using the filter, but recontamination during storage prevented the criterion from being met.\nIt is possible that repeated surveying contributed to increased use of the intervention; some studies have lower uptake of HWTS when delivered programmatically rather than in research-driven efficacy trials such as this.\nHowever, there is some evidence that use is particularly high for filtration compared to other HWTS technologies.\nPrevious studies of LifeStraw filters reported 68% use 8 months after distribution (Boisson et al 2010) and 83% use 2 months after distribution LifeStraw Family and LifeStraw personal filters combined).\nFurthermore, it is possible that use may be particularly high among HIV-positive mothers with young children because of increased concern and awareness of health; chlorination use has been found to be high among similar populations.\nA previous field trial of the LifeStraw Family filter in the Congo also reported high rates of use (76%).\nHowever, nearly all householders in that study (83% of adults and 95% of children <5 years) reported also drinking from other untreated sources, compared with only 5% of mothers and 7% of children <2 years in our trial.\nThe large difference in exclusive use may be attributable to the fact that in the Congo trial households were advised to only use water directly from the filter and were not provided with safe storage containers, implying that safe storage containers may be essential to ensure exclusive use of HWTS.\nAt the same time, there is little evidence that the practice of storing water after it is filtered adversely impacted drinking water quality in the home.\nDiarrhea reductions in our study exceeded the 35\u201344% commonly found by HWTS.\nDiarrhea reductions may have been particularly high among our population because of the increased risk of water-related pathogens in households with PLHIV and the performance of the intervention in removing the full array of microbial pathogens.\nFurthermore, use and exclusive use was high among our population, and there is an increased health impact among high-frequency HWTS users.\nHowever, the intervention did not result in significant reductions in persistent diarrhea among children <2 years or all household members.\nPrevious research has found that household water treatment may be more effective in reducing shorter episodes of diarrhea compared to persistent diarrhea.\nWater quality showed a positive trend with reported diarrhea, both for children <2 years and all household members.\nInterventions that improve water quality are known to reduce diarrheal disease, though the relationship between drinking water quality bacterial indicators and general diarrheal disease is not well established.\nAn observational study in Tanzania found a relationship between health and fecal contamination on hands but not in stored drinking water, though a previous trial of a household ceramic filter in Colombia found a significant relationship between water quality and diarrhea.\nIn our study, the suggestion of positive trend between diarrhea and water quality supports our finding that the water quality intervention resulted in a reduction in diarrheal disease; presumably participants would be unable to base reported diarrhea on actual TTC levels in their water considering they were not aware of exact TTC levels.\nThough we did not find an impact of the intervention on WAZ, we did detect a significant association between WAZ and reported diarrhea.\nThe lack of difference in WAZ between our trial arms despite the reduction in reported diarrhea and the association between WAZ and reported diarrhea merits further discussion.\nIt is possible that reported diarrhea data may be of questionable reliability; open trial designs of self-reported outcomes are subject to bias.\nWe cannot entirely rule out or assess the effects of biased self-reporting of diarrhea.\nHowever, the relationship between diarrhea and water quality is well-established and is the basis for international drinking water quality standards.\nThe fact that we observed this same relationship here suggests that our results are not solely attributable to bias self-report.\nMoreover, we found no association between WAZ and water quality; given that the intervention may only influence WAZ via water quality, the intervention may not be appropriate to improve WAZ.\nFurthermore, diarrhea and WAZ may be associated primarily due to persistent diarrhea.\nWe did not find a significant reduction in persistent diarrhea in children <2 years (p\u200a=\u200a0.26) and a previous trial in Guatemala found that a HWTS intervention mostly prevented short episodes.\nTherefore, the diarrhea experienced by our intervention arm may have been more persistent compared to the intervention group.\nThis is supported by a stronger relationship between diarrhea and WAZ in the intervention arm than in the control arm (p\u200a=\u200a0.003 for interaction); persistent diarrhea is known to impair growth.\nThough we cannot entirely discount the possibility of reporting bias, WAZ may not be an appropriate measure for diarrhea in HWTS trials, though further investigation is needed.\nThere are some limitations to our study.\nFirst, the reliance on self-reported data for diarrhea disease in a non-blinded HWTS intervention trial has previously been criticized.\nHowever, the suggestion of positive trend between water quality and diarrhea suggests that most of the self-reported diarrhea may be verifiable.\nSecond, baseline diarrhea prevalence was not evenly distributed between our trial arms for all household members, though this would only result in a conservative estimate of the intervention effect and baseline diarrhea may not be predictive of diarrhea during the intervention period.\nThird, because we recruited from health clinics, we were not capturing the most vulnerable population that does not have access to health facilities or is too sick to access these services.\nFinally, our study was conducted in Chongwe District, Zambia and may not be generalizable to other locations with different water quality and practices.\nDespite these limitations, our findings indicate that HWTS may be particularly beneficial among HIV-positive mothers with young children.\nThough our study was not designed to examine mortality of children <2 years, our study results and previous research suggest that HWTS may have the potential to reduce mortality in young children.\nThe effect of HWTS on mortality of young children needs to be further explored in the form of a full randomized, controlled trial.\nRecruitment flow diagram.\nWater quality testing results.Unfiltered water is for all households; filtered and stored filtered is only for the intervention arm.\nLongitudinal prevalence of diarrhea in children <2 years and all household members.Data for July 2010 are grouped with August 2010, due to follow-up visits commencing the final week of July.\nWater quality and diarrhea in children <2 years.Water quality is of stored drinking water (stored filtered water for intervention households and unfiltered water for control households). If intervention households did not have stored filtered water available, it was assumed they were drinking unfiltered water. Both analyses are adjusted for age; adjusting for trial arm is examined separately due to the partial collinearity between trial arm and water quality. Predicted probabilities of diarrhea are from unadjusted and adjusted binomial regression models with log link functions and robust standard errors with GEE to account for repeated measures. Error bars represent 95% confidence intervals. Unadjusted model coefficients: ln(diarrhea prevalence)\u200a=\u200a\u22121.25+0.186(log10TTC)+\u22120.0991(child's age). Adjusted model coefficients: ln(diarrhea prevalence)\u200a=\u200a\u22120.868+0.0825(log10 water quality)+\u22120.0990(child's age)+\u22120.506 (trial arm).\n\nSelected baseline characteristics of intervention and control households.\n |  | Intervention | Control\nDemographics | Number of households | 61 (51%) | 59 (49%)\n | Number of households in Ngwerere | 27 (44%) | 27 (46%)\n | Number of households in Kasisi | 34 (56%) | 32 (54%)\n | Number of people | 299 | 300\n | Number of children 6\u201312 months | 61 | 60\n | Median persons per household (range) | 5 (2\u201310) | 5 (2\u201310)\n | Median mother's age (range) | 28 (17\u201344) | 30 (18\u201341)\n | Mother is married or living with partner | 54 (89%) | 41 (69%)\n | Mother has some education | 49 (80%) | 48 (81%)\n | Mother is HIV-positive | 51 (84%) | 49 (83%)\n | Mother on antiretroviral therapy | 17 (28%) | 16 (27%)\nSocioeconomic quintiles | Lowest | 3 (5%) | 14 (24%)\n | Low | 4 (7%) | 4 (7%)\n | Middle | 20 (33%) | 13 (22%)\n | High | 21 (34%) | 16 (27%)\n | Highest | 13 (21%) | 12 (20%)\nWater Source | Piped into home or yard | 3 (5%) | 3 (5%)\n | Public standpipe | 12 (20%) | 10 (17%)\n | Borehole | 7 (11%) | 11 (19%)\n | Protected dug well | 3 (5%) | 7 (12%)\n | Unprotected dug well | 35 (57%) | 27 (46%)\n | Surface Water | 1 (2%) | 1 (2%)\nWater, Sanitation & Hygiene practices | Report usually treating water | 12 (20%) | 13 (22%)\n | Report usually chlorinating | 12 (20%) | 11 (19%)\n | Report usually boiling | 0 (0%) | 3 (5%)\n | Had treated water at time of visit | 6 (10%) | 7 (12%)\n | Water storage container covered | 54 (89%) | 49 (83%)\n | Use cup used to draw water from storage container | 10 (16%) | 17 (29%)\n | Improved sanitation facility | 15 (25%) | 16 (28%)\n | Soap present in household | 27 (44%) | 32 (54%)\nWater Quality | Household TTC Geometric Mean (95%CI)1 | 272 (157\u2013470) | 317 (179\u2013564)\n | Source TTC: Geometric mean (95% CI)1 | 117 (72\u2013190) | 193 (114\u2013328)\n | Household free chlorine \u22650.2 mg/L1 | 1 (2%) | 0 (0%)\n | Source free chlorine \u22650.2 mg/L1 | 0 (0%) | 0 (0%)\nDiarrhea in all household members | Diarrhea in the past 7 days1 | 44 (15%) | 27 (9%)\n | Persistent diarrhea in the past 7 days1 | 10 (3%) | 5 (2%)\n | Persistent diarrhea in the past 7 days1 | 10 (3%) | 5 (2%)\nChildren <2 years | Median age (SD) in months at recruitment | 7.5 (1.9) | 6.9 (1.9)\n | Male | 34 (56%) | 22 (37%)\n | Diarrhea in past 7 days | 18 (30%) | 17 (28%)\n | Persistent diarrhea in past 7 days | 4 (7%) | 4 (7%)\n | HIV-positive, if known | 2 (3%) | 1 (2%)\n | Mean (SD) Weight-for-age z-score | \u22120.81 (1.42) | \u22120.97 (1.59)\n | Currently breastfed | 49 (80%) | 46 (77%)\nStudy Follow up Time | Mean follow up (min-max) in months | 11.2 (7\u201312) | 11.3 (8\u201312)\n\nData are missing for 1 household on stored water TTC, 2 households on source water TTC, and 3 households on stored water chlorine residual. Three individuals are missing data on reported diarrhea and 5 individuals missing data on persistent diarrhea.\n\nFilter use and acceptability among intervention households.\n | Final Visit | All Visits\n | N\u200a=\u200a53 | % | N\u200a=\u200a627 | %\nFilter Use |  |  |  | \nReported user1 | 51/53 | 96% | 596/620 | 96%\nConfirmed user2 | 49/53 | 92% | 540/620 | 87%\nExclusive use by mother today/yesterday3 | 49/53 | 92% | 591/624 | 95%\nExclusive use by child <2 years today/yesterday3 | 48/50 | 96% | 171/184 | 93%\nFilter present in household | 53/53 | 100% | 625/626 | >99%\nFiltered water for drinking today or yesterday | 53/53 | 100% | 606/624 | 97%\nCurrently have filtered water stored4 | 51/53 | 96% | 606/622 | 97%\nAlways used filter in past week | 53/53 | 100% | 620/623 | >99%\nStored filtered \u22651 log10 TTC lower than unfiltered water, or <10 TTC/100 mL | 49/51 | 96% | 557/604 | 92%\nMedian volume of filtered water used per day (range)5 | 5 L (5 L) |  | 5 L (2.5\u201320 L) | \nMother is responsible for filter | 53/53 | 100% | 617/626 | 99%\nWhat people like best about the filter |  |  |  | \nProvides safe water | 40/53 | 75% | 337/618 | 55%\nImproves water taste | 7/53 | 13% | 129/618 | 21%\nProvides good water | 5/53 | 9% | 143/618 | 23%\nEasy to Use | 1/53 | 2% | 11/618 | 2%\nWhat people like least about the filter |  |  |  | \nNothing \u2013 everything is ok | 53/53 | 100% | 615/621 | 99%\nFlow rate is too slow | 0/53 | 0% | 3/621 | <1%\nFilter is broken/has a problem | 0/53 | 0% | 2/621 | <1%\nDoesn't provide enough water | 0/53 | 0% | 1/621 | <1%\nFilter Maintenance6 |  |  |  | \nBackwashed today or yesterday | 52/53 | 98% | 601/624 | 96%\nCleaned pre-filter today or yesterday | 52/53 | 98% | 603/624 | 97%\nWater Storage |  |  |  | \nUsing storage container provided | 53/53 | 100% | 623/625 | >99%\nStorage container capped | 52/53 | 98% | 623/624 | >99%\nOnly store filtered water in supplied containers | 51/53 | 96% | 610/624 | 98%\n\nHouseholds were classified as \u201creported users\u201d if 1) the filter was observed at the time of visit, 2) the storage vessel contained water reported to be treated, and 3) the respondent reported using the filter today or yesterday.\nHouseholds were classified as \u201cconfirmed users\u201d if in addition to the criteria for reported users, there was at least a 1 log10 TTC improvement in stored household water over unfiltered water, or stored water quality was <10 TTC/100 ml.\nExclusive use was defined as not drinking any unfiltered water today or yesterday. For all households that did not report exclusive use, the reason for drinking unfiltered water was that they were away from home. For children <2 years, 3 children in intervention arm died so there are data missing at the final visit. Exclusive use for children <2 years data were only collected in the last quarter of the study period.\nMothers that didn't have filtered water reported that they did not have time to filter.\n5 L is 1 container provided; all households reported 1 container (2 households missing data).\nHouseholds were instructed to backwash and clean the pre-filters daily, as recommended by the manufacturer.\n\nLongitudinal prevalence of diarrhea in intervention and control groups.\n | % Weeks with diarrhea of total possible person-weeks of diarrhea | LPR1 (95% CI) | P\n | Intervention | Control |  | \nDiarrhea |  |  |  | \n<2 years | 6.6% (40/608) | 13.6% (72/530) | 0.47 (0.30\u20130.73) | 0.001\n<2 years, HIV-exposed2 | 7.1% (36/509) | 13.8% (58/419) | 0.50 (0.31\u20130.80) | 0.004\n<5 years | 4.3% (42/967) | 8.9% (79/891) | 0.51 (0.32\u20130.80) | 0.003\nAll household | 1.6% (50/3168) | 3.5% (101/2906) | 0.46 (0.30\u20130.70) | <0.001\nPersistent diarrhea (\u226514 d) |  |  | \n<2 years | 2.1% (13/608) | 3.2% (17/529) | 0.63 (0.28\u20131.40) | 0.26\n<2 years, HIV-exposed2 | 2.2% (11/509) | 3.3% (14/419) | 0.61 (0.25\u20131.49) | 0.28\n<5 years | 1.5% (14/967) | 1.9% (17/890) | 0.77 (0.35\u20131.70) | 0.51\nAll household | 0.6% (18/3168) | 0.7% (21/2904) | 0.75 (0.37\u20131.53) | 0.43\n\nLPR\u200a=\u200aLongitudinal Prevalence Ratio.\nAccounting for repeated measures (children <2 years) and clustering within household (all household data).\nChild is considered HIV-exposed if their mother is HIV-positive.",
        "label": "low",
        "id": "task4_RLD_test_679"
    },
    {
        "paper_doi": "10.1371/journal.pone.0012613",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 190 children under 5, 1144 people, 240 householdsInclusion criteria: unimproved water sources that tested over 1000 thermotolerant coliforms (TTC)/100 ml, reported low use of household water treatment, were easily accessible all year round and were motivated to take part in the project\n\n\nInterventions: LifeStraw(r) Family filter\n\n\nOutcomes: Incidence of diarrhoea among young children in the preceding seven days (recorded monthly); cough and fever also recordedFilter and water quality monitoringCompliance\n\n\nNotes: Location: rural eastern province of Kasai, Democratic Republic of CongoLength: 12 monthsPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "Background\nHousehold water treatment can improve the microbiological quality of drinking water and may prevent diarrheal diseases.\nHowever, current methods of treating water at home have certain shortcomings, and there is evidence of bias in the reported health impact of the intervention in open trial designs.\nMethods and Findings\nWe undertook a randomised, double-blinded, placebo-controlled trial among 240 households (1,144 persons) in rural Democratic Republic of Congo to assess the field performance, use and effectiveness of a novel filtration device in preventing diarrhea.\nHouseholds were followed up monthly for 12 months.\nFilters and placebos were monitored for longevity and for microbiological performance by comparing thermotolerant coliform (TTC) levels in influent and effluent water samples.\nMean longitudinal prevalence of diarrhea was estimated among participants of all ages.\nCompliance was assessed through self-reported use and presence of water in the top vessel of the device at the time of visit.\nOver the 12-month follow-up period, data were collected for 11,236 person-weeks of observation (81.8% total possible).\nAfter adjusting for clustering within the household, the longitudinal prevalence ratio of diarrhoea was 0.85 (95% confidence interval: 0.61\u20131.20).\nThe filters achieved a 2.98 log reduction in TTC levels while, for reasons that are unclear, the placebos achieved a 1.05 log reduction (p<0.0001).\nAfter 8 months, 68% of intervention households met the study's definition of current users, though most (73% of adults and 95% of children) also reported drinking untreated water the previous day.\nThe filter maintained a constant flow rate over time, though 12.4% of filters were damaged during the course of the study.\nConclusions\nWhile the filter was effective in improving water quality, our results provide little evidence that it was protective against diarrhea.\nThe moderate reduction observed nevertheless supports the need for larger studies that measure impact against a neutral placebo.\nTrial Registration\nCurrent Controlled Trials ISRCTN03844341\nIntroduction\nDiarrhoea is responsible for 1.8 million deaths annually, mostly among children under five in developing countries.\nMuch of this disease burden is attributable to unsafe water, poor hygiene and sanitation.\nAn estimated 884 million people worldwide lack access to improved water sources; hundreds of millions more rely on improved sources that are not consistently safe for drinking.\nEven water that is safe at the point of distribution often becomes contaminated during collection, transport and storage within the home due to poor hygiene conditions and practices.\nWhile safe, reliable, piped-in water is an essential goal, treating water at the household or other point of consumption provides a means by which vulnerable populations can improve the quality of their own drinking water.\nThe practice is widespread, with hundreds of millions reporting that they usually treat their water at home before drinking it.\nThere is also evidence that household water treatment is protective against diarrhoea though research suggests that placebo effect and reporting bias play a role in the estimate of effect reported in open trials.\nPlacebo-controlled trials of chlorine-based interventions have been conducted, but apart from a recent study in Ghana, none have assessed the neutrality of the placebo or the effectiveness of the blinding, and other issues have been raised about their methodological quality.\nFilters are more difficult to blind among populations relying on unimproved water.\nIf the water is turbid, a placebo that contains no filter medium is readily identified by comparing its effluent with the effective filter.\nHowever, a placebo that removes turbidity to ensure blinding will probably also remove pathogens that tend to adhere to the suspended solids; it may also create adsorption sites or promote biofilm adhesion that will also render the \u201cplacebo\u201d at least somewhat effective in removing pathogens.\nTo date, the only placebo-controlled trials of household-based filters have been conducted in the United States with municipally treated water that is low in turbidity but also met WHO water quality standards.\nThus, these results cannot be generalised to settings with turbid and contaminated water.\nSeveral water treatment methods have been promoted in low-income settings, including disinfection, disinfection/flocculation, ceramic filtration, solar disinfection and boiling.\nEach has limitations in terms of microbiological effectiveness, cost, acceptability, environmental impact, and sustainability among target populations.\nMoreover, except for boiling, none of these interventions have achieved scale except in limited settings.\nThis has led to calls for alternative technologies that are effective against the full array of microbial pathogens, that can be deployed and used at a large scale with minimum programmatic support, and that will be embraced by the target population.\nThe Lifestraw Family\u00ae is a newly developed household-based gravity filter that employs hollow-fibre membranes to remove waterborne pathogens by ultrafiltration.\nIndependent laboratory testing has shown the device to meet the US Environmental Protection Agency (USEPA) standards for bacteria, viruses and protozoan cysts.\nThe device is designed to treat a minimum of 18,000 L of water and assumed to last for about three years.\nThe manufacturer, Vestergaard-Frandsen SA of Lausanne, Switzerland, plans to sell the filter in large volumes for about US$20.\nMethods\nStudy design; sample size calculation\nThe study was designed as a randomised, double-blinded, placebo-controlled trial.\nOur primary outcome was longitudinal prevalence of diarrhea defined as the number of weeks with diarrhea divided by the total number of weeks under observation.\nThe study was powered to detect a 30% reduction in the mean longitudinal diarrhoea between the two groups.\nThis was a conservative estimate in comparison with the pooled risk reduction of 63% calculated from six previous studies of household filters.\nThe calculation assumed 80% power, \u03b1\u200a=\u200a0.05, a baseline longitudinal prevalence of diarrhoea of 5%, and a coefficient of variation of 2.\nIn order to account for potential lost to follow-up (10%) as well as clustering of diarrhoea within household and intermittent surveillance (7-day period prevalence measured repeatedly once a month over the 12-month follow-up period) (10%), we estimated that we needed at least 600 individuals in each arm.\nAssuming a mean of 5 persons per household, the number of households to be recruited was approximately 120 per arm, or 240 households in total.\nThe protocol for this trial (Protocol S1) and CONSORT checklist (Checklist S1) are available as supporting information.\nSetting and participant eligibility\nThe study was conducted from April 2008 to July 2009 in the rural health zone of Bibanga, 80 km from the city of Mbuji-Mayi in the eastern province of Kasai, the Democratic Republic of Congo (DRC).\nDespite abundant water resources, more than three quarters of the population in rural areas in the DRC rely on unimproved water sources for drinking, mainly surface water and unprotected springs.\nWith the assistance of the Presbyterian Church of Kinshasa, which has been supporting community health programmes in this area for many years, and staff at the health zone level, we identified possible study sites.\nSelected communities relied on unimproved water sources that tested over 1000 thermotolerant coliforms (TTC)/100 ml, reported low use of household water treatment, were easily accessible all year round from the reference hospital of Bibanga where the field team was established, and were motivated to take part in the project.\nIn order to meet the sample size requirements, the study was conducted in two neighbouring villages.\nIntervention\nEach intervention household received a Lifestraw Family filter and each control household received a placebo.\nThe Lifestraw Family is a gravity-fed microbiological water purifier.\nWater is poured into a 2.5 L plastic vessel, passes through a 27-\u00b5m pre-filter, and flows down a 1 m long plastic pipe before passing through the filtration cartridge comprised of hollow-fibres with a 20-nm pore size.\nThe top vessel contains a slow eluding chlorine tablet designed to prevent biofilm formation and increase the life of the cartridge.\nTreated water is accessed from the side of the cartridge via a tap.\nThe device is cleaned daily by rinsing the pre-filter and backwashing the cartridge using a squeeze-pump and outlet valve mounted on the bottom of the cartridge.\nThe device is designed to treat at least 18,000 L of water with a flow rate of approximately 150 ml per minute or 9 L per hour.\nIn the laboratory, the filter was found to meet the USEPA standards for microbiological water purifiers by reducing bacteria by 6.9 logs, viruses by 4.7 logs and protozoan cysts by 3.6 logs.\nPlacebo\nThe placebo had the same configuration, appearance and external components as the Lifestraw Family except that (i) the chlorine tablet was removed from the upper vessel to prevent possible microbicidal action, (ii) the filtration membranes were replaced by some extra piping to imitate the weight and effluent flow rate of the real cartridge, and (iii) the 27-\u00b5m screen on the pre-filter was removed to minimise retention of microbes adhering to suspended solids.\nThree weeks of testing in the laboratory confirmed that the placebo removed no bacteria, viruses and protozoan cysts from test water.\nDespite the challenge in blinding household filters, we determined after piloting that blinding the intervention would be feasible in our study area because the water had low turbidity, ranging from <5 nephelometric turbidity units (NTU) for most of the year to 10 NTU during heavy rains.\nEnrolment, baseline survey, randomization and filter deployment\nAfter discussing the proposed study with community leaders and obtaining consent from the heads of households, a baseline survey was undertaken in April 2008 to collect information on demographics, socio-economic characteristics, and water, hygiene and sanitation practices.\nData collection tools were translated in Tshiluba, the local language, and piloted before use.\nFollowing the baseline survey, households were randomly assigned to one of the two groups using a random number generator.\nRandomisation was stratified by village and was conducted by the trial manager who played no part in the collection of the data.\nBoth the intervention and the placebo were distributed door-to-door by five trained field workers who were unaware of whether the device was an active filter or a placebo.\nHouseholders were trained on use and maintenance of the device according to the manufacturer's instructions.\nThey were advised to drink filtered water directly from the tap and not to store filtered water in order to prevent recontamination.\nThe start of follow-up period was delayed by two months due to initial technical problems with the filters.\nBlinding\nThe allocation sequence was concealed from both field investigators and the study population.\nIn order to blind the intervention among assessors, field workers were divided into two teams.\nThe team responsible for assessing health outcomes was neither involved in the distribution of the filters at the commencement of the trial nor in the assessment of the filter performance and use during follow-up.\nAny questions from the householders that were related to the filter were referred to and dealt with by the filter assessment team.\nOutcome Assessment\nDiarrhoea\nInvestigators interviewed the female head of household or primary care giver of young children once each month over a 12-month period.\nThey recorded any diarrhoea cases in the preceding seven days.\nDiarrhoea was defined as three or more loose stools passed within a 24-hour period.\nIn an effort to further obscure the outcome of interest from the target population, field assessors also inquired about and recorded presence of fever and cough within the past seven days.\nChildren with diarrhoea were given oral rehydration sachets and instructions on how to use them.\nWhen necessary, they were referred to the closest community health post to receive medical care free of charge.\nFever and cough were also treated among young children.\nFilter monitoring\nEach month, a random sample of 30 filters and 30 placebos (25% of the total number distributed) was monitored.\nAt each household visit, field workers noted the location and condition of the filter and recorded if the respondent was able to use and clean the filter correctly.\nFilter components found to be damaged were replaced.\nFlow rate was monitored by filling the top container with 2.5 L of water, opening the tap and measuring the time necessary to fill a 125 ml container with water.\nThe flow rate was expressed in ml per minute.\nWater quality\nInfluent and effluent paired water samples were collected for each of the selected devices.\nIf the respondent mentioned storing the water once filtered, a third sample was collected from the container designated as the treated water storage vessel.\nAll samples were collected in sterile 125 ml Nalgene sampling bottles and assessed for thermotolerant coliforms (TTC) within 4 h after collection.\nMicrobiological assessment was performed using the membrane filtration technique (APHA Standard Methods) on membrane lauryl sulphate medium (Oxoid Limited, Basingstoke, Hampshire, UK) using a DelAgua field incubator (Robens Institute, University of Surrey, Guilford, Surrey, UK).\nMicrobiological performance of the filters was expressed in terms of log reduction value (LRV) calculated as the log of the influent concentration divided by the effluent concentration (log10 influent/effluent).\nCompliance\nCross-sectional surveys were conducted among each household eight and fourteen months after distribution.\nParticipants were classified as current users if they reported using the filter \u2018today or yesterday\u2019 and if the field investigator found the filter hung for use with water in the top vessel of the device.\nConsistency of use was estimated by asking the respondent if he/she had drunk unfiltered water within the previous day.\nThe survey covered further aspects on use and acceptability.\nBlinding assessment\nImmediately following the conclusion of the follow-up period, we assessed the effectiveness of blinding among participants.\nBlinding indices were calculated using methods developed by James and colleagues and Bang and colleagues Female heads of household or primary care giver were asked to identify which device they had received.\nSurveys were targeted at the respondents to the health surveys because they would be most likely to be influenced by their belief in treatment assignment.\nData analysis\nThe analysis of the primary outcome was on an intention-to-treat basis.\nWe used Poisson regression with robust standard errors to estimate the effect of the intervention on the longitudinal prevalence of diarrhea and other health outcomes.\nWe used generalized estimating equations (GEE) to account for clustering at the household level.\nCategorical data were compared using a Chi square or a Fisher's exact test where appropriate.\nContinuous variables were compared with a Student t test.\nStatistical analyses of microbiological data were conducted after log10 transformation of TTC counts to normalize the distribution.\nData analysis was conducted in Stata (Stata Corporation, College Station, Texas, US).\nEthics\nThe study was reviewed and approved by the ethics committee at the London School of Hygiene and Tropical Medicine and the ethics committee at the School of Public Health in Kinshasa.\nWritten consent to participate in the research was obtained from community leaders and the head of each participating household.\nInvestigators explained that half of the study population would be receiving effective microbiological purifiers while the others would receive placebos and that householders should continue their existing water management practices since their device may not be protective against microbial contamination.\nAt the conclusion of the follow-up period, all placebo filters were replaced by effective filters.\nFollowing the completion of the study, the results were communicated to all study participants.\nResults\nParticipant flow\n259 households initially volunteered to participate in the study.\nNineteen households were excluded because they did not reside in the selected villages; they relied primarily on spring water for drinking, or subsequently elected not to participate.\nA total of 240 households were enrolled; 120 were assigned to receive the Lifestraw Family filter and 120 the placebo.\nOver the 12-month follow-up period, data were collected for 11,236 (81.8%) possible person-weeks of observation.\nData were missing for 2492 weeks (18.2%) due primarily to participants leaving the study area or being absent at the time of visit (Figure 1).\nOver the study period, twenty participants died, six of them were children under the age of five.\nThe number of deaths was 12 in the intervention group and 8 in the control group (p\u200a=\u200a0.27).\nBaseline surveys\nIntervention and control groups were similar in terms of demographic and socio-economic characteristics and hygiene and sanitation practices (Table 1).\nAlmost all households primarily used river water for drinking.\nHowever, intervention households were more likely to store their water in clay pots and access it by dipping a cup into the container compared with control households who often used jerrycans.\nOnly four households reported treating their water sometimes or rarely by boiling or adding bleach.\nOnly 37% of households had a latrine and 51% had soap present in the house at the time of visit (Table 1).\nDiarrhoea surveillance\nAt baseline, the prevalence of diarrhoea was similar in both groups (12.6% versus 10.6% for control and intervention groups, respectively).\nOver the 12-month follow-up period, participants of all ages who received the active filter experienced 15% fewer weeks with diarrhoea compared to those who received a placebo (mean, 2.66 versus 3.15, respectively).\nHowever, the confidence interval of the longitudinal prevalence ratio (LPR) adjusted for clustering within the household (LPR 0.85, 95% CI 0.61; 1.20) was wide and included 1.\nThe longitudinal prevalence ratio among children under five was 0.85 (95%CI 0.56; 1.28).\nFigure 2 shows the prevalence of diarrhoea between intervention and control over time.\nWe observed no difference in the mean longitudinal prevalence of fever (LPR 0.99; 95% CI 0.80; 1.22) or cough (LPR 0.99; 95% CI 0.81; 1.22) between the two groups.\nHealth outcome data are presented in Table 2.\nWater quality\nEach device was tested on average 3 times during follow-up. 580 (81%) of the total possible paired water samples were collected.\nMissing samples are due to householders being absent or not being in possession of their filter at the time of visit.\nSource drinking water was highly contaminated, with 75% of household samples showing contamination levels above 1000 TTC/100 ml (Figure 3).\nThe active filter achieved a LRV of 2.98 (95% CI 2.88, 3.08), removing about 99.8% of the indicator bacteria.\nOverall, 64% of water samples treated with filter were free of TTC and 27% had TTC levels between 1\u201310 TTC/100 ml.\nNone of the filters produced water with >100 TTC/100 ml consistently over the three visits.\nSamples from placebos were also contaminated, with 73% of the water samples containing between 100\u20131000 TTC/100 ml.\nHowever, unlike the results from laboratory testing that showed the placebo to be microbiologically ineffective, results from the field showed that the placebo actually removed more than 90% of the TTC from source water (LRV 1.05, 95% CI: 0.93, 1.16).\nFlow rate\nThe mean flow rate of the filters over the study period was 202 ml/min (95% CI 198, 206) or 12 L/hour.\nIt declined slightly over time (\u22121.5 ml/min per month, p<0.002).\nOperation and maintenance; acceptability\nOver half of the respondents (56%) correctly demonstrated how to clean the filters.\nThe pre-filter was cleaned at each use (40%) or once a day (41%), wheareas the cartrige was generally backwashed once a day (67%).\nOverall, 36 (12.4%) of the 290 active filters tested were found damaged during visits, mainly due to rodents chewing on the soft hoses (n\u200a=\u200a35).\nIntervention households reported liking the filter due to improved aesthetics (88%), taste (92%), odour (56%) and health (35%).\nReasons for dissatisfaction were slow flow rate (87%), small size of the top container (85%) and problems with rats (44%).\nCompliance\nEight months after distribution, 183 (76%) of the households were present at the time of visit and were still in possession of their filter.\n68% of respondents in the intervention group could be defined as current users against 48% in the placebo group (p<0.001).\nHowever, nearly all adults (83%) and young children (95%) also reported drinking untreated water in the previous day.\nFourteen months after distribution, the proportion of current users was slightly higher in both groups (76% versus 69% among intervention and control groups, respectively).\nAdditional details about use are included in Table 3.\nSubgroup analysis showed no evidence of an association between use and diarrhoea morbidity (Table 4).\nBlinding status\nTable 5 shows respondent guesses for each treatment assignment groups.\nJames' method, similar to the kappa statistics, produced a blinding index (BI) score of 0.42 (95%CI 0.38; 0.46).\nA score of 0 means that all respondents guessed correctly, 1 indicates that all respondents guessed incorrectly and 0.5 indicates random guessing.\nBang's method calculates the proportion of correct guesses beyond chance in each treatment group.\nBang's BI was 0.96 (95%CI 0.90; 0.99) for the intervention group and \u22120.63 (95%CI \u22120.73; \u22120.53) for the placebo-controlled group.\nBang's blinding index varies from \u22121 to 1.\n1 indicates complete lack of blinding, \u22121 opposite guess about treatment assignment and 0 random guessing.\nSubgroup analysis showed no evidence of an association between diarrhoea and respondents' guesses.\nDiscussion\nWe undertook the first double-blinded, placebo-controlled trial of household-based water filters in a low-income setting with water known to be contaminated with faecal pathogens.\nThis design sought to assess the impact of the intervention in the absence of respondents' bias that is common in open trials.\nDue to challenges of developing a placebo in such settings and to successfully blinding the intervention, we monitored placebo performance and conducted a post-intervention assessment of blinding among study participants.\nFilter performance and health impact were monitored for a full year to account for seasonal variations and minimise the potential for exaggerated health impact often associated with shorter-term trials.\nAfter adjusting for clustering, members of intervention households had 15% fewer weeks of diarrhoea than those of control households, but the confidence intervals indicated little statistical support (longitudinal prevalence ratio 0.85, 95%CI: 0.61 to 1.20).\nWith the exception of a recent study in the United States among an elderly population17, this finding is consistent with other placebo-controlled trials of household water treatment interventions which found no protective effect against diarrhoea.\nHowever, as we have observed elsewhere, those studies may have had insufficient power to identify a statistically significant impact on diarrhoea.\nOur sample size also was not sufficiently large to detect a statistically significant difference in diarrhoea of 15%.\nMoreover, the baseline prevalence of diarrhoea was lower than anticipated and the clustering effect due to repeated measurement and household randomisation was higher.\nPos-hoc sample size calculations indicated that we would have needed a study approximately ten times larger to achieve statistical significance.\nMoreover, the placebo was not microbiologically neutral, as it removed about 90% of faecal bacteria from the source water used by control households.\nThe reasons for this apparent effectiveness are not clear.\nField staff responsible for water quality testing were extensively trained and supervised throughout the study, thereby minimizing the risk of measurement errors.\nOne of the most plausible explanations is the formation of a biofilm resulting from adhesion of suspended solid particles and bacteria on the inner surface of the plastic pipe forming the placebo cartridge.\nThe effectiveness of the placebo rendered our trial a comparison between a 1-log filter and a 3-log filter.\nStudies have reported an association between 1 log removal of faecal bacteria from drinking water and a reduction in diarrhoeal disease.\nOur results may therefore understate the effectiveness of the active filter if it were compared to a true placebo.\nThese results suggest that in this setting with relatively high levels of microbial contamination in source water, a filter of superior microbiological performance may be more effective at preventing diarrhoea than one that removes only 90% of waterborne pathogens.\nThis finding, if validated in future studies, would support the need for high performance standards in water treatment devices in order to optimize health benefits.\nThe blinding of the intervention was not successful.\nIn both treatment groups, the vast majority of survey respondents believed that they had received the active filter, although this proportion was significantly lower in the placebo group.\nUnsuccessful blinding means that we cannot rule out the possibility that the observed effect on diarrhoea is unbiased.\nHowever, the interpretation of blinding indices is not always clear.\nThe fact that a large proportion of control households remained blinded throughout the trial suggests that respondents' bias may have at least been partly reduced.\nThe smaller effect size we observed here may be indicative of a less biased estimate compared with open trials.\nOur estimate is similar to the pooled estimate of effect of open trials of ceramic filters after adjustment for lack of blinding.\nThe fact that \u2018control\u2019 health conditions (fever and cough) remained unchanged by the intervention also suggests that blinding may have been effective, although the usefulness of this approach to detect the presence of respondents' bias has not been validated.\nIncluding a third arm with no intervention would have provided a better understanding of the role of bias in this study.\nUnder field conditions, the Lifestraw Family filters were effective in removing faecal bacteria from source water.\nTwo-thirds of filtered water samples were free of faecal coliforms while most of the remaining samples had low levels of contamination.\nThe fact that specific filters did not consistently produce contaminated water suggests that contamination may have occurred during collection of the sample, perhaps from the tap.\nThe flow rate was higher than that observed in laboratory conditions, possibly due to lower water turbidity at the study site (compared to lab testing at 15 NTU) or inconsistent use by householders.\nThe damage rate was high although the most common problems were due to rats eating the soft plastic components.\nEight months after distribution, two-thirds of the respondents met the study's definition of current users, although almost none of them drank filtered water exclusively.\nThis pattern of use was seen among both adults and children under five.\nParticipants drank unfiltered water when spending time outside their home, but also when they felt eager to drink and did not want to wait for filtration.\nYoung children did not have access to the filter when their parents were away from home.\nIn accordance with the manufacturer's instructions, householders were advised to use water directly from the filter and not to store treated water due to the risk of recontamination.\nConsistent with these instructions, almost none of the households stored filtered water for their children, though many lacked a storage container even if they had chosen to do so.\nThe manufacturer has advised that in future deployment of the filters, it will consider changes in instructions to encourage safe storage of treated water or provide a storage vessel for the filtered water to help increase exclusive consumption of treated water, especially by this vulnerable group of young children.\nHowever, there is also evidence that even occasional consumption of untreated water may eliminate the protective effect of water treatment and changes to the configuration of the filter may not be sufficient to increase exclusive use unless accompanied by fundamental changes in behaviour to increase compliance.\nOur study had certain additional limitations.\nThe study sites were not randomly selected, but were chosen based on eligibility criteria that included high levels of faecal contamination in source water and high prevalence of diarrhoea at baseline.\nAccordingly, these results are not necessarily generalizable to other populations in the Congo or beyond.\nSecond, the use of a seven-day recall period is known to produce less precise estimates compared with a 48-hour recall period.\nOur study provides little evidence of a protective effect of the filter against diarrhoea.\nNevertheless, an effect of 15%, which we observed but could not confirm here, would represent a substantial impact on diarrhoea, a major killer of young children.\nFuture studies with sufficient power to detect this effect size will be necessary to determine the magnitude of any effect against a neutral placebo and to confirm that the effect is not attributable to chance.\nOur study also demonstrates the need to monitor placebo performance and the challenge of blinding household-based water treatment interventions under adverse conditions.\nCONSORT diagram showing the flow of participants through the trial.\nPrevalence of diarrhoea over the course of the study among participants of all ages.\nPercentage of water samples by level of contamination (TTC/100 ml).\n\nBaseline characteristics of participating households.\n | Control | Intervention | Total\n | N | % | N | % | N | %\nDemographic and socio-economic |  |  |  |  |  | \nNumber of households | 120 | (50) | 120 | (50) | 240 | (100)\nNumber of persons | 598 | (52.3) | 546 | (47.7) | 1144 | (100)\nNumber of households with children <5 | 66 | (55) | 57 | (47.5) | 123 | (51.2)\nNumber of children <5 | 105 | (17.6) | 85 | (15.8) | 190 | (16.6)\nMean number of persons per household | 5.0 | 4.5 | 4.8\nMean number of rooms in the house | 2.2 | 2.3 | 2.3\nRespondent is female | 76 | (63.3) | 76 | (63.3) | 152 | (63.3)\nMean age of respondent | 37.5 | 40.8 | 39.1\nLevel of education |  |  |  |  |  | \nNo formal education | 47 | (39.2) | 38 | (31.7) | 85 | (35.4)\nPrimary | 44 | (60.3) | 45 | (54.9) | 89 | (57.4)\nSecondary | 29 | (39.7) | 36 | (43.9) | 65 | (41.9)\nHigher | 0 | (0) | 1 | (1.2) | 1 | (0.6)\nOwns |  |  |  |  |  | \nHouse | 113 | (94.2) | 116 | (96.7) | 229 | (95.4)\nLand | 115 | (95.8) | 117 | (97.5) | 232 | (96.7)\nLivestock | 59 | (49.2) | 64 | (53.8) | 123 | (51.5)\nRadio | 27 | (22.7) | 34 | (28.3) | 61 | (25.5)\nPhone | 10 | (8.3) | 16 | (13.3) | 26 | (10.8)\nBicycle | 18 | (15) | 16 | (13.3) | 34 | (14.2)\nHygiene and sanitation |  |  |  |  |  | \nUse soap to wash hands | 54 | (45) | 54 | (45) | 108 | (45)\nPresence of soap at the time of visit | 65 | (54.2) | 59 | (49.2) | 124 | (51.7)\nReceived hygiene advice in past 6 months | 4 | (3.4) | 10 | (8.4) | 14 | (5.9)\nPresence of latrine | 47 | (39.2) | 41 | (34.2) | 88 | (36.7)\nWater handling practices |  |  |  |  |  | \nPrimary source of drinking water |  |  |  |  |  | \nRiver | 120 | (100) | 117 | (97.5) | 237 | (98.7)\nRainwater | 44 | (36.7) | 46 | (38.3) | 90 | (37.5)\nSpring | 15 | (12.5) | 19 | (15.8) | 34 | (14.2)\nType of drinking water container |  |  |  |  |  | \nClay pot | 68 | (56.7) | 83 | (69.2) | 151 | (62.9)\nJerry can | 50 | (41.7) | 30 | (25) | 80 | (33.3)\nOther | 2 | (1.7) | 7 | (5.8) | 9 | (3.7)\nVessel opening |  |  |  |  |  | \nWide mouth | 71 | (59.2) | 92 | (76.7) | 163 | (67.9)\nNarrow mouth | 49 | (40.8) | 28 | (23.3) | 77 | (32.1)\nStorage vessels covered | 111 | (93.3) | 113 | (95.0) | 224 | (94.1)\nMeans of obtaining water |  |  |  |  |  | \nPour | 48 | (41.0) | 27 | (23.3) | 75 | (32.2)\nDip | 69 | (59.0) | 89 | (76.7) | 158 | (67.8)\nTreat water* | 3 | (2.5) | 1 | (0.8) | 4 | (1.7)\n\n*Treat water sometimes (n\u200a=\u200a1) or rarely (n\u200a=\u200a3). Treatment methods boil (n\u200a=\u200a2), bleach (n\u200a=\u200a1), water settle (n\u200a=\u200a1).\n\nLongitudinal prevalence of diarrhoea and other health conditions by age and treatment group.\n | Mean longitudinal prevalence | LPR (95% CI) | LPR* (95%CI)\n | Control | Intervention |  | \n | Weeks of illness | Person-weeks of observation | % Weeks ill | Weeks of illness | Person-weeks of observation | % Weeks ill |  | \nDiarrhoea |  |  |  |  |  |  |  | \n<5 | 96 | 1072 | 8.96 | 60 | 801 | 7.49 | 0.84 (0.61; 1.14) | 0.85 (0.56; 1.28)\n5\u201315 | 31 | 1880 | 1.65 | 29 | 1765 | 1.64 | 1.00 (0.60; 1.65) | 0.91 (0.49; 1.67)\n>15 | 59 | 2945 | 2.00 | 52 | 2752 | 1.89 | 0.94 (0.65; 1.36) | 0.95 (0.61; 1.57)\nAll ages** | 186 | 5907 | 3.15 | 142 | 5329 | 2.66 | 0.85 (0.68; 1.05) | 0.85 (0.61; 1.20)\nFever |  |  |  |  |  |  |  | \n<5 | 249 | 1072 | 23.23 | 187 | 801 | 23.35 | 1.00 (0.85; 1.19) | 1.02 (0.79; 1.30)\n5\u201315 | 99 | 1880 | 5.27 | 123 | 1765 | 6.97 | 1.32 (1.02; 1.71) | 1.28 (0.89; 1.85)\n>15 | 226 | 2945 | 7.67 | 188 | 2752 | 6.83 | 0.89 (0.74; 1.07) | 0.91 (0.68; 1.22)\nAll ages** | 576 | 5907 | 9.75 | 500 | 5329 | 9.38 | 0.96 (0.86; 1.08) | 0.99 (0.80; 1.22)\nCough |  |  |  |  |  |  |  | \n<5 | 196 | 1072 | 18.28 | 162 | 801 | 20.22 | 1.11 (0.92; 1.33) | 1.11 (0.85; 1.43)\n5\u201315 | 163 | 1880 | 8.67 | 142 | 1765 | 8.05 | 0.93 (0.75; 1.50) | 0.89 (0.63; 1.27)\n>15 | 192 | 2945 | 6.52 | 201 | 2752 | 7.30 | 1.12 (0.93; 1.35) | 1.07 (0.82; 1.39)\nAll ages** | 551 | 5907 | 9.33 | 505 | 5329 | 9.48 | 1.01 (0.90; 1.14) | 0.99 (0.81; 1.22)\n\n*Adjusted for clustering within household.\n**Age missing for 3 participants.\n\nDescription of use among study participants.\n | Control | Intervention | Total\n | n | % | n | % | n | %\nMONTH 8 |  |  |  |  |  | \nLast use (n\u200a=\u200a183)* |  |  |  |  |  | \nPrevious day | 44 | (48.3) | 63 | (68.5) | 107 | (58.5)\nPrevious week | 30 | (33.0) | 14 | (15.2) | 44 | (24.0)\n>1 week ago | 17 | (18.7) | 15 | (16.3) | 32 | (17.5)\nConsistency of use on previous day (n\u200a=\u200a107) |  |  |  |  |  | \nRespondent drank unfiltered water | 43 | (97.7) | 46 | (73.0) | 89 | (83.2)\nChildren (<5) drank unfiltered water | 31 | (93.9) | 39 | (95.1) | 70 | (94.6)\nFilter accessible to young children | 1 | (2.3) | 6 | (9.5) | 7 | (6.5)\nStore filtered water for young children | 4 | (12.9) | 8 | (19.5) | 12 | (16.7)\nAdditional details on use in previous day (n\u200a=\u200a107) |  |  |  |  |  | \nRespondent drank unfiltered water when |  |  |  |  |  | \nIn the field | 33 | (76.7) | 39 | (78.3) | 72 | (77.9)\nIn a hurry to drink | 30 | (69.8) | 33 | (71.7) | 63 | (70.8)\nAway from village | 16 | (37.2) | 15 | (32.6) | 31 | (34.8)\nOther | 3 | (7.0) | 12 | (26) | 15 | (16.8)\nChildren drank unfiltered water when |  |  |  |  |  | \nPerson operating the filter not present | 21 | (67.7) | 31 | (79.5) | 52 | (74.3)\nIn a hurry to drink | 11 | (35.5) | 23 | (59.0) | 34 | (48.6)\nAway from home | 10 | (32.3) | 13 | (33.3) | 23 | (32.9)\nOther | 5 | (16.1) | 7 | (17.9) | 12 | (17.1)\nDid not store filtered water for children: |  |  |  |  |  | \nNo container | 17 | (68) | 28 | (87.1) | 45 | (78.9)\nLock the door | 6 | (24) | 3 | (9.3) | 9 | (15.8)\nDon't want to always filter, too slow | 2 | (8) | 0 | (0) | 2 | (3.5)\nTold not to store water | 0 | (0) | 1 | (3.1) | 1 | (1.7)\nMONTH 14 |  |  |  |  |  | \nLast use (n\u200a=\u200a190)** |  |  |  |  |  | \nPrevious day | 63 | (69.2) | 75 | (75.8) | 138 | (72.6)\nPrevious week | 14 | (15.4) | 11 | (11.1) | 25 | (13.2)\n>1 week ago | 14 | (15.4) | 13 | (13.3) | 27 | (14.2)\n\n*197 (82%) households present at the time of visit; 183 (93%) of them were still in possession of the filter and ever used it.\n**203 (85%) households present at the time of visit; 192 (94%) of them were still in possession of the filter and ever used it + answer missing for 2 households.\n\nLongitudinal prevalence of diarrhoea stratified by reported last time of use.\n | Mean longitudinal prevalence of diarrhoea | LPR (95% CI)\n | Control | Intervention | \n | Weeks of illness | Person-weeks of observation | % weeks ill | Weeks of illness | person-weeks of observation | % weeks ill | \n8 months |  |  |  |  |  |  | \nUser | 71 | 2475 | 2.87 | 74 | 3155 | 2.35 | 0.82 (0.59; 1.13)\nNon-user | 68 | 2420 | 2.81 | 41 | 1319 | 3.11 | 1.11 (0.75; 1.62)\n14 months |  |  |  |  |  |  | \nUser | 102 | 3463 | 2.95 | 99 | 3894 | 2.54 | 0.86 (0.66; 1.10)\nNon-user | 49 | 1642 | 2.98 | 27 | 1025 | 2.63 | 0.88 (0.55; 1.40)\n\n\nBlinding status of respondents by group assignment at the end of the study.\n | Group assignment\nGuess | Placebo* | Lifestraw Family* | Total*\nPlacebo | 17 | (18.3) | 2 | (2.0) | 19 | (9.9)\nLifestraw Family | 74 | (79.6) | 97 | (98.0) | 171 | (89.1)\nDon't know | 2 | (2.1) | 0 | (0) | 2 | (1.0)\nTotal** | 93 | (100.0) | 99 | (100.0) | 192 | (100.0)\n\n*N (%) - number of respondents and percentage in each group.\n**192 (80%) households present at the time of interview and still in possession of the filter.",
        "label": "low",
        "id": "task4_RLD_test_640"
    },
    {
        "paper_doi": "10.1111/iwj.12836",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallelEthics and informed consent: yesSample size estimate: noFollow-up period: 30 days postoperativelyITT analysis: yes, number randomised: 129 groin incisions (100 participants), number analysed: 129 incisionsFunding: \"funded by our own department, without any financial or scientific involvement or support from KCI, ACELITY Company\"Preregistration: no\n\n\nParticipants: Location: Germany\nIntervention group: n = 58 incisions,control group: n = 71 incisionsMean age: intervention group = 71 (range 54 to 89),control group = 66.5 (range 41 to 86)\nInclusion criteria: vascular procedures with access to the common femoral artery with at least 1 of the known main risk factors of wound healing: age > 50 years, diabetes mellitus, renal insufficiency, malnutrition, obesity, and chronic obstructive pulmonary disease.\nExclusion criteria: not stated\n\n\nInterventions: Aim/s: to investigate the effectiveness of ciNPT compared with conventional therapy with regard to the incidence of groin WHC on postoperative days 5 to 7 and 30 and the incidence of surgery revisions 30 days postoperatively after various vascular surgeries.\n Group 1 (NPWT) intervention: ciNPT applied for postoperative days 5 to 7Group 2 (control) intervention: a conventional adhesive plaster that was changed daily\nStudy date/s: 1 February to 30 October 2015\n\n\nOutcomes: Wound complications including SSIValidity of measure/s: Szilagyi classificationTime points: the first evaluation took place on postoperative days 5 to 7 during the hospital stay, while the second evaluation was conducted on postoperative day 30 in the outpatient clinic.\n\n\nNotes: \n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "Abstract\nGroin wound infections in patients undergoing vascular procedures often cause a lengthy process of wound healing.\nSeveral clinical studies and case reports show a reduction of surgical site infections (SSIs) in various wound types after using closed incision negative pressure therapy (ciNPT).\nThe aim of this prospective, randomised, single\u2010institution study was to investigate the effectiveness of ciNPT (PREVENA\u2122 Therapy) compared to conventional therapy on groin incisions after vascular surgery.\nFrom 1 February to 30 October 2015, 100 patients with 129 groin incisions were analysed.\nPatients were randomised and treated with either ciNPT (n = 58 groins) or the control dressing (n = 71 groins).\nciNPT was applied intraoperatively and removed on days 5\u20137 postoperatively.\nThe control group received a conventional adhesive plaster.\nWound evaluation based on the Szilagyi classification took place postoperatively on days 5\u20137 and 30.\nCompared to the control group, the ciNPT group showed a significant reduction in wound complications (P < 0\u00b70005) after both wound evaluation periods and in revision surgeries (P = 0\u00b7022) until 30 days postoperatively.\nSubgroup analysis revealed that ciNPT had a significant effect on almost all examined risk factors for wound healing.\nciNPT significantly reduced the incidence of incision complications and revision procedures after vascular surgery.\nIntroduction\nThe healing process of postoperative groin wounds, often rich in complications, frequently leads to a long course with high treatment costs1.\nDue to the anatomical proximity to the lymph nodes and urogenital organs, as well as its function as a leading access in vascular procedures, the groin is prone to infections.\nAdditional wound\u2010healing complications (WHC) in the groin include wound dehiscence, lymphatic leaks with lymphatic fistula and lymphocele, seroma, haematoma, skin necrosis and delayed healing 1, 2, 3, 4, 5, 6, 7.\nSurgical site infections (SSIs) are present in 2\u201322% of all surgical procedures and contribute more than 20% of the costs of all complicated wounds 8, 9.\nIn accordance with international data, the incidence of SSIs after vascular surgery in the groin are 3\u201344%, and deep groin infections with prosthetic material involvement are described in up to 6% of cases 4, 10.\nThe relationship between SSIs and morbidity correlates with extended hospital stay, severe limb ischaemia, extremity loss, massive haemorrhage, systemic sepsis and septic embolisation 1, 4, 5.\nDe spite increasing knowledge of systemic wound\u2010healing factors and many surgical techniques (e.g., sloping groin cut, implantation of obturator and lateral femoral bypasses, use of antibiotic\u2010coated prosthesis, rotation flaps, and fibrin glue), only systemic antibiotic therapy has yielded acceptable results 3, 4, 5, 6, 11.\nAdditionally, negative pressure wound therapy (NPWT; V.A.C.\u00ae\nTherapy, KCI, an ACELITY Company, San Antonio, TX) has been used to manage groin incisions.\nSince the development of NPWT by Morykwas and Argenta in the United States and Fleischmann in Germany in the second half of the 1990s, this method has been used as a supporting therapy for wound healing 12, 13.\nIn subsequent years, several case reports and clinical studies described the effectiveness of NPWT in the management of the following wounds: complex open wounds intended for secondary closure, infected wounds as a supplement to surgical debridement and antibiotic therapy, degloving injuries, sternal wound dehiscence, wounds after open traumatic injuries and high\u2010energy trauma wounds.\nNPWT has also been used as a method to bolster transplanted skin grafts 7, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25.\nThis successful use led to the idea of applying the NPWT dressing on primarily closed wounds to facilitate incision healing.\nUnder the designation closed incision negative pressure therapy (ciNPT), this new technique has resulted in many significant clinical results 11, 13, 25, 26.\nSince 2010, multiple studies and case reports comparing standard\u2010of\u2010care dressings to ciNPT have reported a decrease in SSIs in a wide spectrum of traumatic, orthopaedic, abdominal, sternal and plastic surgery incisions 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37.\nThe reason for this success may be due to the reported mechanisms of action of the ciNPT, which protects the incision from external wound contamination, strengthens the cohesiveness of the wound edges, removes fluids and infectious materials from the wound, decreases the lateral tension around the incision and facilitates oxygen saturation and blood microcirculation within the incision area 11, 38, 39.\nciNPT as delivered by the PREVENA\u2122 Incision Management Therapy System (KCI, an ACELITY Company) consists of a vacuum unit with a battery and a preset negative pressure of \u2212125 mmHg.\nIts integrated individual components includes a mobile therapy unit with a replaceable exudate collection canister, a polyester fabric interface layer with 0\u00b7019% silver for the control of bioburden within the dressing, a polyurethane foam bolster and a polyurethane film with acrylic adhesive.\nA polyurethane shell encapsulates the foam bolster and interface layer, providing a closed system.\nUntil now, four clinical studies have reported on the use of the ciNPT in the groin after vascular surgeries 4, 9, 10, 40; however, these studies lacked a prospective, randomised study design and a subgroup analysis of risk factors and perioperative parameters.\nThe aim of this study was to investigate the effectiveness of ciNPT compared to conventional therapy with regards to the incidence of groin WHC on postoperative days 5\u20137 and 30 and the incidence of surgery revisions 30 days postoperatively after various vascular surgeries.\nAdditionally, subgroup analyses of the main wound\u2010healing risk factors and perioperative risk factors were evaluated to assess the effect of ciNPT on specific patients at risk of postoperative WHC in the groin.\nFurthermore, a logistic regression and a receiver operating characteristic (ROC) analysis were conducted to forecast the risk of postoperative WHC within the main patient risk factors and perioperative risk factors.\nMethods\nThis prospective, randomised, monocentric study design was approved by the ethics committee of Justus Liebig University Giessen.\nThe study was conducted independently and was fully funded by our own department, without any financial or scientific involvement or support from KCI, ACELITY Company.\nFrom 1 February to 30 October 2015, 100 patients with 129 groin incisions were evaluated.\nInclusion criteria were as follows: vascular procedures with access to the common femoral artery with at least one of the known main risk factors of wound healing: age >50 years, diabetes mellitus, renal insufficiency, malnutrition, obesity and chronic obstructive pulmonary disease (COPD).\nAll patients received at least a 5\u2010cm longitudinal incision in the groin.\nThe total number of groins was divided into either the ciNPT group or the control group.\nAll patients received perioperative antibiotic prophylaxis and a preoperative hair shave and sterile skin disinfection with the antiseptic kodan\u00aeTinktur forte (Sch\u00fclke & Mayr GmbH 22840 Norderstedt, Germany) in the surgery area.\nAfter placing a drain, subcutaneous tissue was re\u2010approximated with Vicryl 2.0 sutures (Johnson & Johnson Medical GmbH, Ethicon, Norderstedt, Germany), and the skin was secured with a skin\u2010clamping device (WECK Visistat 35 W. Teleflex Medical GmbH. Kernen, Germany).\nAfterwards, ciNPT (utilizing PREVENA\u2122 Therapy) was applied on the incision (Figure 1).\nOn postoperative days 5\u20137, ciNPT was removed, and a conventional adhesive plaster (Cosmopor E Steril Hartmannn, Heidenheim, Germany) was used.\nThe control group received a conventional adhesive plaster that was changed daily.\nWound evaluations were determined at two\u2010time points.\nThe first evaluation took place on postoperative days 5\u20137 during the hospital stay, and the second evaluation was conducted on postoperative day 30 in the outpatient clinic.\nGroin incisions were graded using the Szilagyi classification 41.\nGrade I describes superficial infections that remain restricted on the skin.\nGrade II contains an infiltration of the subcutaneous layer without participation of the arterial graft.\nGrade III describes an infection involving the arterial graft.\nAlthough this classification system mainly describes only tissue and prosthetic infections, it does consider the anatomical tissue layers for the evaluation of all kinds of groin wound complications.\nIn this study, patients with cutaneous wound dehiscence, skin necrosis and single local infection signs were classified as grade I.\nWound dehiscence in the subcutaneous layer, haematoma, lymphatic fistula, lymphocele, seroma, single local infection signs and systemic infection parameters [leukocytes >13 109/dl, C\u2010reactive protein (CRP) > 100 mg/l] were classified as grade II.\nAll classical local infection signs (pain, swelling, redness and hyperaemia, warmth, dysfunction), systemic infection parameters and arterial graft infections were classified as grade III.\nSubgroup analysis included the main wound\u2010healing risk factors and perioperative risk factors.\nAll risk factors were examined with regards to the incidence of groin incision complications on postoperative days 5\u20137 and 30 and surgery revisions until postoperative day 30.\nThe main risk factors were defined as: age > 50 years, diabetes mellitus with haemoglobin A1c (HbA1c) > 6\u00b75% and 48 mmol/mol glucose, renal insufficiency with glomerular filtration rate < 89 ml/min (stage 2) and creatinine >1\u00b73 mg/dl, overweight with BMI > 25 kg/m2, malnutrition with albumin <35 g/l, protein <65 g/l and transferrin <2 g/l and COPD with the Global Initiative For Chronic Obstructive Lung Disease (GOLD) grade 1 FEV1 \u2265 80%.\nThe perioperative risk factors were defined as wound length > 8 cm, hospital stay >8 days, operative time > 142 minutes, perioperative blood transfusion with haemoglobin <8 mg/dl and previous vascular interventions (Digital Subtraction Angiography or Percutaneous Transluminal Angioplasty).\nStatistical analysis was performed with the Student's test, Levene's test and Fisher's exact test.\nFor the subgroup analyses, the following tests were performed: Fisher's exact test and Pearson Chi Square test.\nFurther analytical methods within the study included logistic regression, ROC analysis and calculation of the correlation coefficients.\nStatistical significance was determined by a P\u2010value <0\u00b705.\nResults\nThe study included 100 patients with 129 groin wounds.\nThe patients included 28 females and 72 males, with a median age of 68\u00b75 \u00b1 9\u00b76.\nOf the 129 groin wounds, 29 were a result of bilateral surgery.\nThe most frequently reported comorbidities were peripheral artery disease (62%) and abdominal aortic aneurysm (21%) (Table 1).\nPrior surgeries included femoral endarterectomy (29%), endovascular aneurysm repair (EVAR) (24%) and femoral popliteal bypass (21%) (Table 2).\nFor the main analysis, there were 35 (27\u00b71%) groin WHCs, with 5 (8\u00b76%) in the ciNPT group (n = 58) and 30 (42\u00b73%) in the control group (n = 71).\nThe first postoperative wound examination on postoperative days 5\u20137 in the ciNPT group showed no WHC in Szilagyi grades I\u2013III wounds, while the control group had 5 (7%) in Szilagyi grade I and 10 (14\u00b71%) in Szilagyi grade II.\nDuring the second examination on postoperative day 30 in the ciNPT group, four (6\u00b79%) WHCs were noted in Szilagyi grade I and one (1\u00b77%) in Szilagyi grade II.\nThe control group showed 3 (4\u00b72%) WHCs in Szilagyi grade I, 10 (14\u00b71%) in grade II and 2 (2\u00b78%) in grade III (Table 3; Figures 2 and 3).\nBoth WHCs in Szilagyi grade III appeared after implantation of a femoro\u2010femoral cross\u2010over bypass.\nBecause of the infection, the prosthesis was removed immediately.\nThe overall incidence of postoperative wound complications (P < 0\u00b70005) and the incidence on postoperative days 5\u20137 (P < 0\u00b70005) and 30 (P = 0\u00b7023) were statistically significant and favoured the ciNPT group (Table 3).\nWhen comparing the incidence of revision surgeries, there was only 1 (1\u00b77%) case in the ciNPT group versus 10 (14\u00b71%) cases in the control group.\nThe comparison of both groups showed a significant advantage for ciNPT (P = 0\u00b7022) (Table 3).\nThe most frequently occurring WHC in the ciNPT group was superficial wound dehiscence.\nIn the control group, haematoma and local infection were the leading WHCs (Table 4).\nFor both groups, local infections with and without revision surgery were treated with antibiotics.\nOne patient died on day 1 postoperatively in the ciNPT group.\nThe cause of death was unrelated to ciNPT.\nFurther subgroup analyses were based on the perioperative risk factors of wound length > 8 cm, hospital stay >8 days, operating time > 42 minutes, previous interventions and perioperative blood transfusions with regards to the incidence of groin WHCs.\nIn patients who had a wound length > 8 cm, the ciNPT group had significantly fewer total WHCs as compared to the control group (P = 0\u00b7003).\nThese results were similar for ciNPT patients in the following subgroups: operation time > 142 minutes (P = 0\u00b70005), hospital stay >8 days (P = 0\u00b7001) and perioperative blood transfusion (P = 0\u00b7004).\nThe significant effect of ciNPT with regards to groin WHCs was also observed on postoperative days 5\u20137 in the following subgroups: wound length > 8 cm (P = 0\u00b7015), operation time > 142 minutes (P = 0\u00b7002), hospital stay >8 days (P = 0\u00b7001) and perioperative blood transfusion (P = 0\u00b7023).\nHowever, only patients in the hospital stay >8 days (P = 0\u00b7014) or operation time > 142 minutes (P = 0\u00b7020) subgroups showed a positive effect of ciNPT on day 30.\nWith respect to revision surgeries, only ciNPT patients who had a hospital stay >8 days had significantly fewer revision surgeries compared to control patients [1 (2\u00b77%) versus 10 (20\u00b78%), respectively; P = 0\u00b7012] (Table 5).\nIn the logistic regression, all single risk factors were examined against the aim variable postoperative WHCs.\nThe biggest predictor for the development of postoperative WHCs could be shown only in the following perioperative risk factors: wound length (P = 0\u00b7003, OR = 4\u00b7800) and operation time (P = 0\u00b7046, OR = 2\u00b7571).\nWhen the all risk factors were investigated, the wound length (P = 0\u00b7015, OR = 7\u00b7503) showed the greatest potential for the prediction of postoperative WHCs.\nIn the ROC analysis, accurate forecasting of a postoperative WHC could be indicated but only for the perioperative risk factors with an area under the curve (AUC) of 0\u00b7662 (P = 0\u00b7016) (Figure 4).\nA closer consideration of wound length and operation time as the biggest predictors demonstrated that the classification achievement of the index perioperative risk factors in the logistic regression in a new ROC analysis is due primarily to the potential of wound length (AUC 0\u00b7664, P = 0\u00b7007) and operation time (AUC 0\u00b7690, P = 0\u00b7005) (Figure 4).\nA complementary calculation of the correlation coefficients of all risk factors demonstrated that wound length with operation time (0\u00b7386) and overweight status (\u22120\u00b7200) had the best correlation.\nOther correlating risk factors included renal insufficiency with age (0\u00b7342) and diabetes mellitus with excessive weight (0\u00b7326).\nThe performed statistical analyses of the subgroups of patients were based on the main wound\u2010healing risk factors of age (>50 years), diabetes mellitus, renal insufficiency, malnutrition, overweight and COPD with regards to the incidence of groin WHCs.\nIn patients whose age was >50 years, the ciNPT group had significantly fewer total WHCs as compared to the control group (P < 0\u00b70005).\nThese results were similar for the ciNPT patients in the following subgroups: diabetes mellitus (P < 0\u00b70005), renal insufficiency (P < 0\u00b70005), malnutrition (P = 0\u00b7043) and overweight status (P < 0\u00b70005).\nOn postoperative days 5\u20137, patients in all wound\u2010healing risk factor subgroups, except malnutrition (P = 0\u00b7081) and COPD (P = 0\u00b7206), showed a significant result for ciNPT.\nOn postoperative day 30, significance was found only for ciNPT patients in the age subgroup (P = 0\u00b7040).\nWith regards to the incidence of revision surgeries, only ciNPT patients in the age (>50 years) subgroup had fewer revision surgeries compared to the control patients [1 (3\u00b72%) versus 6 (23\u00b71%), respectively; P = 0\u00b7029] (Table 5).\nDiscussion\nAgainst the background of extended hospital stays and higher treatment costs, preventive measures to reduce the incidence of postoperative complications play an important role.\nIn order to optimise the advancement of preventive procedures, the effectiveness of ciNPT on groin wounds after vascular surgeries was evaluated.\nThe effect of ciNPT has been demonstrated in clinical studies and case reports in a variety wound types 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37.\nIn spite of many publications, clinical studies involving groin wounds are rare.\nTo date, there exist only four clinical studies in which ciNPT were examined on groin wounds after vascular surgeries 4, 9, 10, 40.\nOur results point out that ciNPT had a significant effect on the reduction of the incidence of wound complications and revision surgeries as well as the effect on almost all mean and perioperative risk factors.\nThe results of this study are comparable to the Matatov et al. study 4 given that both studies are similar in design and sample size.\nIn our study and the Matatov et al. evaluation, there were significant reductions in the overall incidence of WHCs (P = 0\u00b70011 and P < 0\u00b70005, respectively).\nAll three wound infections (6%) in the Matatov study were classified as Szilagyi grade I, while in this study, there were four incision complications classified as Szilagyi grade I and only one grade II.\nSimilarly, the results of this study were consistent with the results of Weir 9 (with only two incision complications) and Karl and Woeste 10 (with no complications) after application of ciNPT.\nThis observation supports the reduction of incision complications in deep tissue layers with possible revision surgeries when ciNPT is utilised.\nThis observation was confirmed in this study by the significant decrease of revision surgeries within the first 30 days postoperatively.\nThe statistical significance (P < 0\u00b70005) in the Szilagyi grade II complications after two evaluation periods with a wound proportion of 1:20 brings out the effectiveness of ciNPT (Table 3).\nThe detailed view of the separate types of wound complications in this study shows a particular reduction of subcutaneous haematoma with a clear statistical significance (P = 0\u00b7020) and clarifies the positive impact of the ciNPT suction effect (Table 4).\nAn 8\u00b76% difference with no WHCs on days 5\u20137 postoperatively and five WHCs on day 30 postoperatively could suggest that ciNPT loses its effectiveness after being removed.\nThe removal of ciNPT may compromise the sterile wound conditions and can lead to potential wound contamination during subsequent wound dressing changes, which may be interpreted as a casual explanation for higher wound complication rates on day 30 postoperatively.\nThis view is based on evidence on the number of local wound infections in the control group (Table 4).\nciNPT functions as a barrier against potential wound contamination and serves as an effective method for the removal of wound exudate.\nTo prolong the beneficial wound management effects of ciNPT over days 5\u20137 postoperatively, a longer application time of ciNPT should be considered.\nAlternative options to this could be a strictly sterile execution of the routine changes of the wound dressing and covering the incision wound with a dry gauze with an adhesive dressing attached on it as a contamination barrier.\nCertainly, the question about the application period of these measures should be discussed prior to removing surgical staples in order for the procedure to serve as a guideline.\nDespite significant results in the subgroup analysis, a loss of effectiveness was shown in almost all risk factors after days 5\u20137 postoperatively.\nA possible reason for this may be found in the loss of effectiveness of ciNPT over a longer time period as described above.\nIf special pathological influences on the individual risk factors exist and therefore limit the effectiveness of the ciNPT, then they must be cleared by secondary examinations.\nIn the study by Matatov et al., 4 subgroup analysis and isolated view on two evaluation periods regarding the incidence of incision complications were unfortunately not carried out.\nIn this subgroup analysis, ciNPT had a significant effect on clinical use with regards to all main and perioperative risk factors (Table 5).\nThis is strengthened by the results of our logistic regression analysis, the ROC analysis and the correlation calculation.\nIn particular, the perioperative risk factors, wound length and operation time (in the logistic regression analysis and in the ROC analysis) were the strongest predictors for postoperative WHCs (Figure 4).\nThe calculation of the correlations especially emphasised a stronger relationship between the wound length and the operation time.\nTherefore, the wound length and operation time could be determined in all three investigations as the strongest predictors with regards to the cause for postoperative WHCs in the groin.\nBy these data, the option of a specific indication for the application of the ciNPT is given and an arbitrary execution of this therapy can be prevented.\nAs a practical assistance in the specific use of the ciNPT, a scoring system to justify its indication appears to be useful.\nTherefore, we constructed a scoring system based on the significant study data in which the risk factors with the highest significance were assigned a value of 2 points and the risk factors with a lower significance were assigned 1 point (Table 6).\nThe limit for the indication of ciNPT was based on the average score of the point values of the scoring system.\nThe average score for all patients treated with the ciNPT was 7\u00b75 points.\nThe average score for patients with the ciNPT without WHCs was 8\u00b74 points.\nIn accordance with these average scores, the lower limit for the indication of ciNPT was set at 8 points.\nWith the use of this scoring system, 12% of the patients in the ciNPT group showed a WHC compared to 88% without a WHC.\nAccording to these data, it becomes evident that by using this scoring system for the ciNPT, postoperative groin WHCs may be prevented.\nThrough a firm consideration of the significant risk factors in this study within the preoperative phase, ciNPT can be specifically used as an efficient incision management measure to prevent postoperative inguinal WHCs.\nA similar conclusion was presented by Willy et al. in the international multidisciplinary consensus recommendations, where the experts examined 100 publications and recommended the consideration of ciNPT use for patients with risk factors and high\u2010risk procedures 42.\nGiven the fact that the study data regarding the application of ciNPT in groin wounds after vascular surgery are based solely on the clinical examinations with the PREVENA\u2122\nIncision Management Therapy System, the question arises whether other ciNPT systems can achieve a similar positive effect on groin wounds.\nAn almost equivalent ciNPT system is the PICO\u2122 Single Use Negative Pressure Wound Therapy System (Smith & Nephew, London, UK), which in comparison with the PREVENA\u2122\nIncision Management Therapy System eliminates the wound secretions by evaporation and uses less negative pressure (\u221280 mmHg).\nTo clarify whether both ciNPT systems show related positive clinical results in groin wounds, further studies are needed.\nConclusions\nIn comparison to conventional adhesive plaster, the use of ciNPT demonstrates a statistically significant reduction of postoperative WHCs in the groin on postoperative days 5\u20137 and 30 and revision surgeries until day 30 postoperatively in patients after several vascular surgeries.\nThe results of the subgroup analysis show a significant effect of ciNPT on almost all examined risk factors, through which a specific preventative use may be possible for patients with a corresponding risk profile.\n(A) Components of ciNPT; B) ciNPT after aortobifemoral bypass.\nWound complications of study patients based on Szilagyi classification. (A) Szilagyi I: Skin necrosis, superficial wound dehiscence and local infection; (B) Szilagyi II: Deep wound dehiscence and fat necrosis; (C) Szilagyi III: Prosthetic graft infection.\nWound results after removing ciNPT on (A) 5\u20137 days and (B) 30 days postoperatively.\nROC curve of (A) all perioperative risk factors, (B) perioperative risk factor operation time and (C) perioperative risk factor wound length.\n\nPatient characteristics\n | ciNPT group | Control group | P\u2010value\nNumber of patients | 43 | 57 | \nNumber of groin incisions | 58 | 71 | \nGender |  |  | \nMale | 29 (67%) | 43 (75%) | 0\u00b75\nFemale | 14 (33%) | 14 (25%) | 0\u00b75\nMean age [years] | 71 (range 54\u201389) | 66\u00b75 (range 41\u201386) | 0\u00b7020\nMean BMI [kg/m2] | 26\u00b77 (range 19\u00b71\u201337\u00b73) | 27\u00b78 (range 18\u00b74\u201337\u00b72) | 0\u00b7205\nHypertension | 38 (88%) | 53 (93%) | 0\u00b7325\nCoronary artery disease | 22 (51%) | 13 (23%) | 0\u00b7003\nDiabetes mellitus | 22 (51%) | 29 (51%) | 1\nRenal insufficiency | 27 (63%) | 30 (53%) | 0\u00b7415\nDialysis | 0 (0%) | 2 (35%) | 0\u00b7322\nMalnutrition | 13 (30%) | 22 (39%) | 0\u00b7406\nCOPD | 9 (21%) | 8 (14%) | 0\u00b7791\nSmoker | 23 (53%) | 22 (39%) | 0\u00b7159\nPreoperative anaemia | 19 (44%) | 30 (53%) | 0\u00b7426\nPostoperative anaemia | 19 (44%) | 27 (51%) | 0\u00b7840\nPostoperative leucocytosis | 22 (51%) | 33 (58%) | 0\u00b7547\nPeripheral artery disease |  |  | \nFontaine classification grade II | 13 (30%) | 26 (46%) | 0\u00b7149\nFontaine classification grade III | 5 (12%) | 2 (4%) | 0\u00b7234\nFontaine classification grade IV | 6 (14%) | 10 (18%) | 0\u00b7785\nInfrarenal abdominal aortic aneurysm | 14 (33%) | 7 (12%) | 0\u00b7024\nThoracic aortic aneurysm | 1 (2%) | 4 (7%) | 0\u00b7387\nThoracic abdominal aortic aneurysm | 3 (7%) | 5 (9%) | 1\nInfrarenal aortic stenosis | 0 (0%) | 1(2%) | 1\nArtery occlusion (thrombosis/embolism) | 0 (0%) | 3 (5%) | 0\u00b7257\nVisceral artery aneurysm | 0 (0%) | 1 (2%) | 1\nLeriche syndrome | 1 (2%) | 1 (2%) | 1\n\nBMI, body mass index; COPD, chronic obstructive pulmonary disease.\n\nPerioperative characteristics\n | ciNPT group | Control group | P\u2010value\nMean operative time [minutes] | 140 (range 40\u2013436) | 146 (range 32\u2013402) | 0\u00b7706\nMean hospital stay [days] | 12\u00b78 (range 5\u201343) | 13\u00b70 (range 5\u201344) | 0\u00b7909\nMean wound length [cm] | 7\u00b77 (range 5\u201315) | 8\u00b76 (range 5\u201315) | 0\u00b7017\nPerioperative blood transfusion | 9 (21%) | 13 (23%) | 1\nProcedure types |  |  | \nEVAR/TEVAR | 19 (44\u00b72%) | 17 (30%) | 0\u00b7148\nRevascularisation | 26 (61%) | 41 (72%) | 0\u00b7284\nBilateral procedures | 19 (44\u00b72%) | 14 (26%) | 0\u00b7053\nProsthetic material used |  |  | \nPTFE | 4 (9\u00b73%) | 6 (10\u00b75%) | 1\nDacron | 2 (4\u00b77%) | 4 (7%) | 0\u00b7697\nDacron patch | 10 (23\u00b73%) | 18 (31\u00b76%) | 0\u00b7380\nVein | 6 (14%) | 7 (12\u00b73%) | 1\n\nEVAR, endovascular aortic repair; PTFE, polytetrafluoroethylene; TEVAR, thoracic endovascular aortic repair.\n\nIncidence of wound\u2010healing disturbances with reference to the total number of groin incisions, wound evaluation on 5\u20137 and 30 day postoperatively and revision surgery on 30 day postoperatively based on Szilagyi classification\n | Total number | 5\u20137 day postoperatively | 30 day postoperatively | Revision surgery on 30 day postoperatively\nSzilagyi classification | ciNPT group n = 58 | Control group n = 71 | P\u2010value | ciNPT group n = 58 | Control group n = 71 | P\u2010value | ciNPT group n = 58 | Control group n = 71 | P\u2010value | ciNPT group n = 58 | Control group n = 71 | P\u2010value\nSzilagyi grade I | 4 (6\u00b79%) | 8 (11\u00b73%) | 0\u00b7545 | 0 (0%) | 5 (7%) | 0\u00b7064 | 4 (6\u00b79%) | 3 (4\u00b72%) | 0\u00b7070 | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501\nSzilagyi grade II | 1 (1\u00b77%) | 20 (28\u00b72%) | <0\u00b70005 | 0 (0%) | 10 (14\u00b71%) | 0\u00b7005 | 1 (1\u00b77%) | 10 (14\u00b71%) | 0\u00b7022 | 1 (1\u00b77%) | 6 (8\u00b75%) | 0\u00b7128\nSzilagyi grade III | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501 | 0 (0%) | 0 (0%) | 1 | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501 | 0 (0%) | 2 (2\u00b78%) | 0\u00b7501\nTotal number | 5 (8\u00b76%) | 30 (42\u00b73%) | <0\u00b70005 | 0 (0%) | 15 (21\u00b71%) | <0\u00b70005 | 5 (8\u00b76%) | 15 (21\u00b71%) | 0\u00b7023 | 1 (1\u00b77%) | 10 (14\u00b71%) | 0\u00b7022\n\n\nTypes of wound complications within the three grades of Szilagyi classification\n | ciNPT group | Control group | P\u2010value\nSuperficial wound dehiscence | 3 (7%) | 4 (7%) | 1\nSkin necrosis | 1 (2\u00b73%) | 3 (5%) | 0\u00b7632\nDeep wound dehiscence with fat necrosis | 1 (2\u00b73%) | 4 (7%) | 0\u00b7387\nHaematoma | 0 (0%) | 8 (14%) | 0\u00b7020\nSeroma | 0 (0%) | 1 (1\u00b78%) | 1\nLymphatic fistula | 1 (2\u00b73%) | 3 (5\u00b73%) | 0\u00b7632\nArterial graft infection | 0 (0%) | 2 (4%) | 0\u00b7322\nLocal infection | 1 (2\u00b73%) | 10 (17\u00b75%) | 0\u00b7022\n\n\nAnalyses on subgroups of patients based on main wound\u2010healing risk factors and perioperative risk factors with regards to WHCs and revision surgeries\n | Analysis Intervals | \n | Total number of WHCs | Number of WHCs at postoperative days 5\u20137 | Number of WHCs at postoperative day 30 | Patients requiring revision surgery on 30 day postoperatively\nPatient subgroups | ciNPT group | Control group | P\u2010value | ciNPT group | Control group | P\u2010value | ciNPT group | Control group | P\u2010value | ciNPT group | Control group | P\u2010value\nAge (>50 years) | n = 31, 2 (6\u00b75%) | n = 26, 18 (69\u00b72%) | <0\u00b70005 | n = 31, 0 (0%) | n = 26, 11 (42\u00b73%) | <0\u00b70005 | n = 31, 2 (6\u00b75%) | n = 26, 7 (26\u00b79%) | 0\u00b7040 | n = 31, 1 (3\u00b72%) | n = 26, 6 (23\u00b71%) | 0\u00b7029\nDiabetes mellitus | n = 22, 2 (9\u00b71%) | n = 29, 17 (58\u00b76%) | <0\u00b70005 | n = 22, 0 (0%) | n = 29, 8 (27\u00b76%) | 0\u00b7007 | n = 22, 2 (9\u00b71%) | n = 29, 9 (31%) | 0\u00b7059 | n = 22, 1 (4\u00b75%) | n = 29, 7 (24\u00b71%) | 0\u00b7061\nRenal insufficiency | n = 27, 2 (7\u00b74%) | n = 30, 15 (50%) | <0\u00b70005 | n = 27, 0 (0%) | n = 30, 9 (30%) | 0\u00b7002 | n = 27, 2 (7\u00b74%) | n = 30, 6 (20%) | 0\u00b7163 | n = 27, 1 (3\u00b77%) | n = 30, 3 (10%) | 0\u00b7347\nMalnutrition | n = 13, 2 (15\u00b74%) | n = 22, 11 (50%) | 0\u00b7043 | n = 13, 0 (0%) | n = 22, 5 (22\u00b77%) | 0\u00b7081 | n = 13, 2 (15\u00b74%) | n = 22, 6 (27\u00b73%) | 0\u00b7355 | n = 13, 0 (0%) | n = 22, 4 (18\u00b72%) | 0\u00b7140\nOverweight | n = 32, 5 (15\u00b76%) | n = 41, 23 (56\u00b71%) | <0\u00b70005 | n = 32, 0 (0%) | n = 41, 14 (34\u00b71%) | <0\u00b70005 | n = 32, 5 (15\u00b76%) | n = 41, 9 (22%) | 0\u00b7354 | n = 32, 1 (3\u00b71%) | n = 41, 5 (12\u00b72%) | 0\u00b7167\nCOPD | n = 9, 1 (11\u00b71%) | n = 8, 4 (50%) | 0\u00b7111 | n = 9, 0 (0%) | n = 8, 2 (25%) | 0\u00b7206 | n = 9, 1 (11\u00b71%) | n = 8, 2 (25%) | 0\u00b7453 | n = 9, 1 (11\u00b71%) | n = 8, 1 (12\u00b75%) | 0\u00b7735\nWound length (>8 centimetre) | n = 25, 4 (16%) | n = 49, 25 (51%) | 0\u00b7003 | n = 25, 0 (0%) | n = 49, 13 (26\u00b75%) | 0\u00b7015 | n = 25, 4 (16%) | n = 49, 12 (24\u00b75%) | 0\u00b7197 | n = 25, 1 (4%) | n = 49, 9 (18\u00b74%) | 0\u00b7083\nHospital stay (> 8 days) | n = 37, 3 (8,1%) | n = 48, 28 (58\u00b73%) | 0\u00b7001 | n = 37, 0 (0%) | n = 48, 14 (29\u00b72%) | 0\u00b7001 | n = 37, 3 (8\u00b71%) | n = 48, 14 (29\u00b72%) | 0\u00b7014 | n = 37, 1 (2\u00b77%) | n = 48, 10 (20\u00b78%) | 0\u00b7012\nOperation time (> 142 minutes) | n = 21, 2 (9\u00b75%) | n = 28, 21 (75%) | <0\u00b70005 | n = 21, 0 (0%) | n = 28, 10 (35\u00b77%) | 0\u00b7002 | n = 21, 2 (9\u00b75%) | n = 28, 11 (39\u00b73%) | 0\u00b7020 | n = 21, 1 (4\u00b77%) | n = 28, 7 (25%) | 0\u00b7062\nPreviousinterventions | n = 9, 1 (11\u00b71%) | n = 18, 7 (38\u00b79%) | 0\u00b7149 | n = 9, 0 (0%) | n = 18, 4 (22\u00b72%) | 0\u00b7174 | n = 9, 1 (11\u00b71%) | n = 18, 3 (16\u00b77%) | 0\u00b7593 | n = 9, 0 (0%) | n = 18, 3 (16\u00b77%) | 0\u00b7279\nPerioperative blood transfusion | n = 9, 1 (11\u00b71%) | n = 13, 10 (77%) | 0\u00b7004 | n = 9, 0 (0%) | n = 13, 6 (46\u00b71%) | 0\u00b7023 | n = 9, 1 (11\u00b71%) | n = 13, 4 (31%) | 0\u00b7230 | n = 9, 1 (11\u00b71%) | n = 13, 2 (15\u00b74%) | 0\u00b7642\n\nCOPD, chronic obstructive pulmonary disease.\n\nScoring system for ciNPT based on the significant risk factors for groin WHCs\nRisk factors | Points\nPatient age (P < 0\u00b70005) | 2\nDiabetes mellitus (P < 0\u00b70005) | 2\nRenal insufficiency (P < 0\u00b70005) | 2\nOverweight (P < 0\u00b70005) | 2\nOperation time (P < 0\u00b70005) | 2\nMalnutrition (P = 0\u00b7043) | 1\nWound length (P = 0\u00b7003) | 1\nPerioperative blood transfusion (P = 0\u00b7004) | 1\n",
        "label": "unclear",
        "id": "task4_RLD_test_713"
    },
    {
        "paper_doi": "10.1111/iwj.12436",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Study design: randomised controlled trialStudy grouping: parallelEthics and informed consent: yesSample size estimate: noFollow-up period: unknownITT analysis: yes, number randomised: 20, number analysed: unclearFunding: unclear. MHB gave scientific presentations for KCI.Preregistration: no\n\n\nParticipants: Location: Nuremberg, Germany\nIntervention group: n = 10,control group: n = 10Mean age: intervention group = 52.3 (16.3),control group = 57.8 (15.2)\nInclusion criteria: patients with spinal fractures who were scheduled for internal fixation\nExclusion criteria: not reported\n\n\nInterventions: Aim/s: to evaluate the different aspects of wound healing in spinal fractures treated with internal fixationGroup 1 (NPWT) intervention: the iNPWT group was treated with a PICO system (Smith & Nephew, UK). The PICO system was left on the wound for 5 days including the day of surgery. In addition to daily clinical examination, all wounds/seroma were analysed by ultrasonography on day 5 and day 10 after surgery.Group 2 (control) intervention: standard department wound dressing consisting of dry wound coverage (compresses attached to the skin) was used.\nStudy date/s: not reported\n\n\nOutcomes: SeromaValidity of measure/s: ultrasound was used as a standardised imaging modality to detect seromas in the wound area.Time points: day 5 and day 10 after surgery\n\n\nNotes: Investigator contacted for additional details\n\n",
        "objective": "To assess the effects of NPWT for preventing SSI in wounds healing through primary closure, and to assess the cost\u2010effectiveness of NPWT in wounds healing through primary closure.",
        "full_paper": "Abstract\nTo evaluate the clinical use and economic aspects of negative pressure wound therapy (NPWT) after dorsal stabilisation of spinal fractures.\nThis study is a prospective randomised evaluation of NPWT in patients with large surgical wounds after surgical stabilisation of spinal fractures by internal fixation.\nPatients were randomised to either standard wound dressing treatment (group A) or NPWT (group B).\nThe wound area was examined by ultrasound to measure seroma volumes in both groups on the 5th and 10th day after surgery.\nFurthermore, data on economic aspects such as nursing time for wound care and material used for wound dressing were evaluated.\nA total of 20 patients (10 in each group) were enrolled.\nThroughout the whole study, mean seroma volume was significantly higher in group A than that in group B (day 5: 1\u00b79 ml versus 0 ml; P = 0\u00b70007; day 10: 1\u00b76 ml versus 0\u00b75 ml; P <0\u00b7024).\nFurthermore, patients of group A required more wound care time (group A: 31 \u00b1 10 minutes; group B 13\u00b78 \u00b1 6 minutes; P = 0\u00b70005) and more number of compresses (total number; group A 35 \u00b1 15; group B 11 \u00b1 3; P = 0\u00b70376).\nNPWT reduced the development of postoperative seroma, reduced nursing time and reduced material required for wound care.\nIntroduction\nIn recent years, negative pressure wound therapy (NPWT) became a widely used therapy for many different indications in the treatment of wounds 1, 2, 3, 4.\nRecently, NPWT has also been used in the treatment of closed surgical wounds.\nThe incisional NPWT (iNPWT) has shown beneficial effects when administered after severe trauma 5, 6, 7, 8.\nThe indications and the evidence for the efficacy of iNPWT have increased recently 9, 10, 11.\nHowever, only a few prospective randomised studies have been published on those indications mostly dealing with hip arthroplasty 11, 12.\nThe mode of action of iNPWT is still not completely understood.\nThis is thought to be mediated by increased oxygen delivery to the tissue as result of enhanced tissue perfusion and promotion of angiogenesis 5, 9, 13.\nThe purpose of this study was to evaluate the different aspects of wound healing in spinal fractures treated by internal fixation.\nWe compared a standard wound dressing group with an iNPWT group in the context of postoperative seroma formation in the wound area, the total time of secretion, the total time needed for the wound care (dressing changes) and the material needed for dressing changes.\nMaterials and methods\nA total of 20 patients with spinal fractures were scheduled for internal fixation.\nThey were randomised into two groups.\nGroup A (10 patients) received the standard wound dressing of our department, consisting of a dry wound coverage (compresses attached to the skin).\nGroup B (10 patients) was treated with iNPWT over the sutured wound area.\nThe surgical intervention was identical in both the groups.\nAn open reduction technique with internal fixation system was performed for all the patients, which was obtained from the same manufacturer (Synthes, West Chester, PA).\nAll patients received two Redon\u00ae drains, one on each side of the spinal column subcutaneously.\nPostoperative physiotherapy and mobilisation were also identical for both the groups.\nThe iNPWT group (group B) was treated with a PICO\u2122 system (Smith & Nephew plc, London, UK).\nThe PICO\u2122 system was left on the wound for 5 days including the day of surgery.\nIn addition to daily clinical examination, all wounds/seroma were analysed by ultrasonography on the 5th and 10th day after surgery.\nBefore surgery, plasmatic coagulation was assessed in all patients using the Quick prothrombin time test.\nPostoperatively, the immediate amount of wound secretion in the Redon drain canisters was quantified.\nIn addition, the length of the incision was measured.\nThe duration of secretion from the wounds was monitored and the total number of dressing changes and the time required to perform the dressing changes were also assessed.\nThe material used for dressing changes was also quantified (compresses and gloves).\nStatistical significance was calculated with the Prism v6.0 GraphPad Software, Inc. (La Jolla, CA).\nFor Gaussian distributed data, the Student's t\u2010test was used.\nFor non\u2010Gaussian distributed data, the Mann\u2013Whitney test was used.\nInformed consent was obtained from each patient.\nThe study was approved by the local ethics committee (Re\u2010No.139_12 B) and conforms to the principles of the Declaration of Helsinki.\nResults\nIn this study, 10 patients (mean age 57\u00b780 \u00b1 15\u00b724 years) were randomised to group A and 10 patients (mean 52\u00b730 \u00b1 16\u00b732 years) to group B. Both groups displayed normal coagulation times according to the Quick prothrombin time test (group A: 94\u00b720 \u00b1 9\u00b752%; group B: 96\u00b700 \u00b1 7\u00b780%; P = 0\u00b773).\nThere was no significant difference in the postoperative wound size between both groups (group A: 17\u00b725 \u00b1 5\u00b770 cm; group B: 14\u00b760 \u00b1 4\u00b738 cm; P = 0\u00b726).\nFurthermore, both groups displayed almost equal volumes of wound secretion in the Redon\u00ae drain canisters after 2 days (group A: 621\u00b75 \u00b1 286\u00b75 ml; group B: 454\u00b70 \u00b1 229\u00b76 ml; P = 0\u00b716).\nThe seroma volume underneath the surgical wound was significantly lower at day 5 and day 10 in the iNPWT group (day 5: group A: 1\u00b79 \u00b1 2\u00b77 ml versus group B: 0 \u00b1 0 ml (P = 0\u00b70007); day 10: group A: 1\u00b76 \u00b1 2\u00b76 ml versus group B: 0\u00b75 \u00b1 1\u00b70 ml; P = 0\u00b7024).\nThe patients treated with iNPWT required fewer dressing changes: 48 dressing changes in group B patients, equating to 4\u00b78 per patient and 79 dressing changes in group A patients equating to 7\u00b79 per patient (Figure 1).\nGroup B patients also had lesser number of days of wound secretion (Figure 2) and required lesser time for wound care (Figure 3) and lesser material for dressing changes (Figures 4 and 5).\nDiscussion\nSince the development of NPWT, the indications for the use of NPWT have been mainly acute and chronic wounds 1, 2, 3, 14, 15, 16, 17, 18; however, the indications have increased over time 2, 19.\nNPWT exerts a positive effect on wound healing resulting in a reduction in wound healing complications.\nFurthermore, NPWT treatment excels because of its ease of application and a low risk of side effects 20, 21.\nRecent studies using NPWT showed a reduction of seromas in wounds after hip surgery.\nThe beneficial effect of NPWT was found after elective total hip arthoplasty and after arthroplasty of femoral neck fractures 11, 12.\nA recently published review confirmed the reduction of wound complications in high\u2010risk wounds 9.\nIn our study, we evaluated for the first time the possible effects of iNPWT in spinal fractures treated with open reduction and internal fixation.\nBesides the anatomical area of application, it is the first prospective randomised study in orthopaedics using the PICO\u2122 system (Figure 6) (Smith & Nephew plc).\nThis study provides compelling evidence that iNPWT may be useful to treat large surgical incision wounds after surgical treatment of spinal fractures.\nTo the best of our knowledge, significant reduction of wound complications in fractures of the spine treated with open reduction and internal fixation by using iNPWT has not been previously reported.\nIn addition, iNPWT treatment reduced the time and dressing material needed for postoperative wound care.\nThe present study used ultrasound as a standardised imaging modality to detect seromas in the wound area.\nThe high sensitivity of this imaging modality allowed for the detection of seroma volumes directly underneath the surgical incision.\nThis method was previously described for the evaluation of iNPWT after surgical interventions of the hip and femoral neck fracture 11, 12.\nIn line with these studies, we found a significant reduction of wound draining days in the iNPWT group compared with that in the control group.\nIn addition to reduced wound secretion, Stannard et al. showed that iNPTW treatment reduced haematoma size in patients suffering from high\u2010energy trauma injuries 5, 22.\nGiven the fact that haemotomas might favour as nutrient\u2010rich environments for bacterial replication and that persisting wound drainage might facilitate the entrance of bacteria into the wounds, the reduction of wound secretion and the haematoma size might reduce the risk of a wound infection.\nHaematomas are thought to serve as rich nutrient sources for infection 23.\nIn our opinion, a prolonged secretion is an important risk factor of early postoperative infections as well.\nHowever, to the best of our knowledge, we are not aware of a study that has addressed this issue.\nHence, further investigations into these mechanisms are warranted.\nFurthermore, it is not fully understood how iNPWT leads to a reduced seroma and haematoma formation in the wounded tissue.\nHorch et al. suggested that NPWT results in a significantly increased tissue perfusion and oxygenation 13.\nIn addition to altering tissue perfusion, treatment with iNPWT might reduce wound edge tension and thereby promote healing 24.\nOur results in this study are consistent with the findings of other studies regarding the reduction of wound healing complications after the treatment with iNPWT.\nWe and others have demonstrated that wound treatment with iNPWT after elective total hip arthroplasty for the treatment of osteoarthritis of the hip and the treatment with iNPWT after surgical treatment of femoral neck fractures reduced the wound healing complications 11, 12.\nIn addition, patients who suffered from severe soft tissue damage after trauma displayed a faster recovery when the iNPWT device was placed on the wounded tissue at early time points 13, 25.\nIn our study, we attached the iNPWT device immediately after surgical wound closure to the skin over the wound.\nIn a previous study, we observed that the time dedicated to the wound care and the consumption of wound care material were significantly shorter than those in the control group.\nIn agreement with our previous study 11, we found that iNPWT treatment reduced the time needed for wound care and the consumption of wound care material.\nThe 48 dressing changes in the iNPWT group also included the removal of the redon drains and the removal of the device.\nTo avoid a bias, these dressing changes were also documented; the number would have been even smaller if this study\u2010related dressing changes were not taken into account.\nLimitations of the present study are the relatively small number of enrolled patients and the use of an iNPWT device from a single manufacturer.\nThe difference between the different manufacturers seems to be relatively small and the principles of NPWT seem to be transferable between the different manufacturers.\nHowever, this has to be confirmed by comparative studies.\nConclusion\nIn summary, our results support the use of iNPWT after spinal surgery.\nApart from its economic benefits, iNPWT promotes wound healing and might prevent wound infections, which are the dreaded complications of spine surgery.\nDistribution of number of dressing changes (P < 0\u00b70001).\nDistribution of number of days of wound secretion.\nDistribution of wound care time.\nDistribution of number of used gloves for dressing changes.\nDistribution of number of used compresses for dressing changes.\nApplication of a PICO system to the wound.",
        "label": "unclear",
        "id": "task4_RLD_test_707"
    },
    {
        "paper_doi": "10.1186/1471-2458-11-475",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Study design: cluster randomised controlled trialStudy duration: recruitment occurred May to July 2006. The parent meeting was delivered eight times during July and August 2006. Follow-up was 3 monthsStudy arms: intervention - educational session plus leaflet; control - leaflet only\n\n\nParticipants: Setting:Country and income level: England (high; 3 of 33 wards, selected to represent high-, medium-, and low-deprived geographical areas)Degree of regional development: not describedParticipants:ParentsInclusion and exclusion criteria for participation in study: Clusters - primary healthcare centres with at least two medical practitioners selected by highest deprivation score. Childcare centres selected by largest size. Parents - English literate, with a child eligible for the first or second dose MMR vaccination (first dose was given at 13 months and the second dose between four and five and half years of age, so target age range was six months to five years.)Categorisation (mothers, fathers, parents, expectant parents, guardians): parentsNumber randomised to intervention: healthcare centres N = 3; childcare centres N = 3; parents N = 71 (N = 68 completed baseline survey)Number randomised to control: healthcare centres N = 3; childcare centres N = 3; parents N = 71 (N = 67 completed baseline survey)Total number randomised: healthcare centres N = 6; childcare centres N = 6; parents N = 142Number lost to follow-up, withdrew from intervention: N = 23 did not receive intervention; N = 13 lost to follow-up for final time pointNumber lost to follow-up, withdrew from control: N = 7 lost to follow-up for final time pointAge range: intervention 34.07 years +- 5.43; control 34.06 years +- 5.52Gender: intervention female N = 67; male N = 4. control female N = 67; male N = 4Ethnicity: intervention - White British 68 (95.8%); other 3 ( 4.2%). control - White British 68 (95.8%); other 3 ( 4.2%)Level of education or literacy: intervention - left school at 16 years: 24 (33.8%); left school at 18 years: 10 (14.1%); achieved degree or higher: 37 (52.1%). control - left school at 16 years 25: (35.2%); left school at 18 years 10: (14.1%); achieved degree or higher: 36 (50.7%)Socioeconomic status: not describedChildrenAge range and categorisation (infants, preschool-aged children, school aged children): preschool-aged childrenMean age +- SD of first (youngest) child eligible, months: intervention 25.73 +- 14.66. control 19.77 +- 11.69Gender: not stated\n\n\nInterventions: Intervention purpose: to provide parents with the opportunity to discuss MMR with other parents who are making an MMR decision; to provide information about MMR from a variety of perspectives; to introduce and practice one approach to supporting parents to ask questions about MMR of their healthcare practitionerDeliverer: co-facilitated by a researcher and a parent facilitator. Parent facilitators were all female and recruited from local communitiesFormat or delivery mode: parent meeting including three components: provision of balanced information, a group discussion, and a coaching exercise. Mean of 6 participants per meetingContent of communication: three parts. Group discussion provided opportunity to discuss any issues about MMR with other parents who were also making an MMR decision; Q&A session provided opportunity to ask questions of the immunisation nurse specialist; Coaching exercise introduced and practiced one approach to supporting parents to ask questions about MMR in the primary care consultation (role playing exercise using question prompt sheet)Vaccine or vaccines delivered or described: MMRDirection of communication: interactiveGroup or individual: groupWhere the intervention took place: non-healthcare venues (e.g. community centres) close to participating healthcare centres and childcare organisationsTraining required for intervention: parent facilitators received one half-day trainingTheoretical basis for intervention: in line with fundamental tenets of health promotion, i.e. based on an 'engagement' model of communication where a key goal is empowermentCost of intervention: not describedIntervention quality: content and delivery of parent meeting informed by interviews with parents, systematic review of parents' decision-support needs, and two focus groups with parentsFidelity and integrity: session was co-facilitated by a researcher at each meeting, which potentially kept the sessions consistent. 23/71 parents in intervention group did not attend an intervention meeting, so intervention was not delivered as intended to all those randomised to intervention groupDetails of control, usual, or routine care: MMR leaflet onlyDetails of co-interventions in all groups: all participants received 'MMR - your questions answered' leaflet. Authors noted this leaflet was meant to be equivalent to usual care, but turned out to be more comprehensive than usual care (parents reported they were not normally given leaflets that were meant to be mandatory).\n\n\nOutcomes: Primary outcomes measured:Vaccination statusDefinition of immunisation status used by authors: receipt of MMR ('Since the study started, have you taken your child to have the combined MMR vaccine?')Description of outcome assessment tool: postal questionnaireTiming of outcome assessment: three months post-interventionKnowledgeDefinition: knowledge about MMRDescription of outcome assessment tool: postal questionnaire, 11 questions (score 0 to 11)Timing of outcome assessment: one week and three months post-interventionAttitudes or beliefsDefinition: attitude towards MMR, beliefs about vaccine necessity, concerns about MMRDescription of outcome assessment tool: one question about attitude (rate 1 to 7); four items assessing necessity beliefs (score 4 to 20); four items assessing concern beliefs (score 4 to 20)Timing of outcome assessment: one week and three months post-interventionIntention to vaccinateDefinition: intention to vaccinate child with MMRDescription of outcome assessment tool: postal questionnaire, 3 items on a 7-point scale, averaged over the three itemsTiming of outcome assessment: one week and three months post-interventionAdverse effects (anxiety)Definition: anxiety from interventionDescription of outcome assessment: Short form State-Trait Anxiety Inventory (STAI) tool with 6 items scored 1 to 4. Total score multiplied by 20/6. A normal score is 34 to 36.Timing of outcome assessment: one week and three months post-interventionSecondary outcomes measured: none\n\n\nNotes: Sample size: \"To achieve 80% power to detect a standardised effect size of 0.67 on the primary outcome of decisional conflict, using a two-sided t-test with significance level of 0.05 and an estimated intracluster correlation coefficient (ICC) of 0.05 (giving a design effect of 1.5 based on an average of 11 parents per cluster) required a sample size of 108 parents (54 in each group). Predicting 25% attrition, 73 parents were required in each group. Parent numbers were not balanced across the clusters. Based on our previous research, we estimated recruiting 12 parents per week over three months.\"Contact with author: authors contacted to clarify sequence generation process and to identify any published protocol or trial record. No protocol available, and authors could not recall randomisation process but confirmed it was conducted by a statistician.Other: authors conducted qualitative research alongside, determining that the intervention was feasible to deliver in non-healthcare, community venues, and it was acceptable to parents, with the majority expressing positive view\n\n",
        "objective": "To assess the effects of face\u2010to\u2010face interventions for informing or educating parents about early childhood vaccination on vaccination status and parental knowledge, attitudes and intention to vaccinate.",
        "full_paper": "Background\nIn the UK public concern about the safety of the combined measles, mumps and rubella [MMR] vaccine continues to impact on MMR coverage.\nWhilst the sharp decline in uptake has begun to level out, first and second dose uptake rates remain short of that required for population immunity.\nFurthermore, international research consistently shows that some parents lack confidence in making a decision about MMR vaccination for their children.\nTogether, this work suggests that effective interventions are required to support parents to make informed decisions about MMR.\nThis trial assessed the impact of a parent-centred, multi-component intervention (balanced information, group discussion, coaching exercise) on informed parental decision-making for MMR.\nMethods\nThis was a two arm, cluster randomised trial.\nOne hundred and forty two UK parents of children eligible for MMR vaccination were recruited from six primary healthcare centres and six childcare organisations.\nThe intervention arm received an MMR information leaflet and participated in the intervention (parent meeting).\nThe control arm received the leaflet only.\nThe primary outcome was decisional conflict.\nSecondary outcomes were actual and intended MMR choice, knowledge, attitude, concern and necessity beliefs about MMR and anxiety.\nResults\nDecisional conflict decreased for both arms to a level where an 'effective' MMR decision could be made one-week (effect estimate = -0.54, p < 0.001) and three-months (effect estimate = -0.60, p < 0.001) post-intervention.\nThere was no significant difference between arms (effect estimate = 0.07, p = 0.215).\nHeightened decisional conflict was evident for parents making the MMR decision for their first child (effect estimate = -0.25, p = 0.003), who were concerned (effect estimate = 0.07, p < 0.001), had less positive attitudes (effect estimate = -0.20, p < 0.001) yet stronger intentions (effect estimate = 0.09, p = 0.006).\nSignificantly more parents in the intervention arm reported vaccinating their child (93% versus 73%, p = 0.04).\nConclusions\nWhilst both the leaflet and the parent meeting reduced parents' decisional conflict, the parent meeting appeared to enable parents to act upon their decision leading to vaccination uptake.\nBackground\nIn the UK public concern about the safety of the combined measles, mumps and rubella [MMR] vaccine as a consequence of Wakefield et al's now discredited research continues to impact on MMR coverage even though the sharp decline in uptake has begun to reverse.\nCurrent first and second dose MMR uptake rates in England are 84% and 77% respectively, short of the 95% required for population immunity.\nAs a consequence, there is a pool of unimmunised children susceptible to the diseases, reflected in persistent localised measles outbreaks with an epidemic predicted in the near future.\nIn some European countries and in the USA, childhood immunisation is mandatory yet MMR vaccine refusal has increased, similarly leading to measles outbreaks.\nRepeated assertions by the Department of Health for England and Wales and the U.S Centers for Disease Control and Prevention that the MMR vaccine is safe have had limited effect on allaying parents' concerns in some sections of the community.\nMore than ten years after the publication of Andrew Wakefield's now discredited findings, there is some evidence that parent trust in MMR has improved, yet significant numbers continue to lack confidence in making an MMR decision and many criticise what is perceived to be the poor quality of information provided.\nA recent systematic review identified the decision support needs of parents making child health decisions (including immunisation); these related to three themes: (i) a need for timely, consistent up-to-date evidence based information tailored to the individual child, delivered in a variety of formats from trustworthy sources; (ii) a need to talk with others facing the same decision to share experiences; and (iii) a need to be in control of their level of preferred involvement in the decision-making process.\nThis suggests that interventions, informed by parents' expressed needs of what they would find helpful to make informed decisions about MMR are required.\nHowever, in spite of a large literature describing the factors that influence parents' decisions whether to vaccinate their child with MMR, evaluations of decision support interventions in this context are limited.\nInformed by a systematic review and an interview study with parents we developed and evaluated an evidence-based, parent-centred, multi-component intervention to support informed parental decision-making for the MMR vaccine.\nThis parent-centred approach is consistent with Western health policy in which a clinician-patient partnership is emphasised.\nIt is also in line with the fundamental tenets of health promotion that is based on an 'engagement' model of communication where a key goal is empowerment.\nThe intervention was designed to supplement routine UK primary care service for childhood vaccination whereby parents are invited to take their child, free of charge, for all immunisations on the National Health Service Routine Childhood Immunisation Schedule.\nChildhood vaccination is not mandatory in the UK.\nIn this paper we report the findings of a cluster randomised controlled trial to evaluate a parent-centred, multi-component intervention to support informed decision-making for MMR.\nThe effectiveness data are presented here.\nAcceptability of the intervention is reported elsewhere.\nMethods\nSetting and Participants\nThe study was approved by the Local Research Ethics Committee (06/Q1107/25) on 18 May 2006.\nIt was located in three of 33 wards (electoral district) in Leeds, England (approx. 770,000 population).\nUsing the Index of Multiple Deprivation these wards were selected to represent low, medium and high deprived geographical areas [mean scores: low = 6.89; medium = 29.22; high = 55.07].\nPrimary healthcare centres employing at least two medical practitioners were purposively selected based on their low income scheme index [LISI, 39] scores.\nFor example, in the high deprived ward, we approached centres with the most deprived practice population first (i.e. highest LISI score).\nChildcare organisations in the same wards were approached on the basis of size, the largest first.\nEleven (of 15) healthcare centres and six (of eight) childcare organisations were invited to participate.\nSix primary care centres and six childcare organisations agreed.\nWithin these providers the target sample was parents who were English literate with a child eligible for the first or second dose MMR vaccination.\nAt the time of the study, in the UK, the first dose was given at 13 months and the second dose between four and five and half years of age.\nThe target age range was, therefore, six months to five years.\nLetters were sent to eligible parents on providers' registers.\nParents replied to the research team and were telephoned for screening and recruitment.\nRecruitment occurred May to July 2006.\nDesign and Intervention\nThe design was a cluster randomised controlled trial design with two arms: intervention and control.\nThis design was chosen to reduce the potential risk of contamination between arms.\nThe six healthcare centres and six childcare organisations were matched in pairs based on their ward (three pairs of healthcare centres, three pairs of childcare organisations).\nOne of each pair was randomly allocated to the intervention, the other to the control arm.\nA researcher not involved in the study and blind to the identity of clusters performed the randomisation using a sealed envelope procedure.\nThe study researcher (RP) was blind to arm assignment when screening and recruiting parents, and sending out the baseline questionnaire.\nStatisticians (WH, RW) saw blinded data.\nParents were blind to arm assignment at recruitment and in completing the baseline questionnaire.\nParents allocated to the intervention arm were invited to attend one two-hour parent meeting, co-facilitated by a researcher (CJ, FMC, RP) and a parent.\nThree parent facilitators (all women) were recruited from local communities.\nThey received one half-day training.\nIn advance of the meeting parents were sent an information leaflet ('MMR your questions answered',).\nThe content and delivery of the parent meeting (see Table 1) was informed by interviews with 69 parents and a systematic review of parents' decision support needs.\nThis was refined in two focus groups with local parents.\nThe meeting included three components: provision of balanced information, a group discussion and a coaching exercise.\nParents in the control arm were sent the same MMR leaflet.\nMeasures\nParent characteristics (e.g. age, ethnicity) were collected by telephone at recruitment.\nPrimary and secondary outcomes were collected by postal questionnaire prior to randomisation (T1), one week post-intervention (T2) and three months post-intervention (T3).\nThe questionnaire was developed in collaboration with an expert in the field of health decision-making (HB) and piloted with five parents, though no changes were made.\nImpact of the parent meeting and MMR leaflet\nThe primary outcome measure was decisional conflict as measured by the Decisional Conflict Scale.\nThis generic measure is a 16-item scale to assess people's perceptions about the quality of their decision-making process; it has five sub-sections for being informed, clear about their values, degree of support, uncertainty with the choice, effectiveness of their decision.\nIt has demonstrated test-retest reliability, construct and predictive validity in the patient health decision-making context.\nScores range from 1 (no decisional conflict) to 5 (extremely high decisional conflict).\nScores lower than two are associated with 'implementing decisions', higher scores are interpreted as 'decision delay or feeling unsure about implementation'.\nHigh Cronbach alpha coefficients of 0.95 (T1), 0.92 (T2) and 0.94 (T3) were achieved.\nSecondary outcomes were self-reported measures of the decision (actual and intended actions), attitude towards MMR and beliefs about the MMR options, knowledge and anxiety.\nThe MMR decision was measured at three months post-intervention using a self-report item 'Since this study started, have you taken your child to have the combined MMR vaccine?'\nIn addition, a measure of intended choice was developed using three items measured on a 7-point scale e.g.\n'I intend to give my child the combined MMR vaccine at the recommended ages' (definitely do not-definitely do).\nThese three items were measured at all three time-points.\nResponses were averaged over the three items.\nCronbach alpha coefficients of 0.84 (T1), 0.79 (T2) and 0.90 (T3) were obtained.\nKnowledge about MMR and the measles disease was measured using multiple choice items developed for the purposes of this study using Department of Health for England and Wales literature.\nThe measure was not validated.\nThe number of questions answered correctly were summed to produce a total knowledge score (maximum 11).\nAttitude towards MMR was measured on a 7-point scale.\nParents responded to the statement 'For me to give my child the combined MMR vaccine at the recommended ages would be' on three semantic differential evaluative endpoints (1 to 7); e.g. extremely bad/extremely good.\nResponses were averaged over the three items.\nCronbach alpha coefficients were 0.78 (T1), 0.73 (T2) and 0.80 (T3).\nThese intended choice and attitude items have demonstrated validity and reliability and were informed by guidelines on measuring health cognitions.\nParents' beliefs about the MMR options were assessed using a modified version of the Beliefs about Flu Vaccination Questionnaire.\nThe measure was not validated for use in this context.\nFour items assessed parents' beliefs about the necessity of MMR e.g.\n'Without the combined MMR vaccine, my child could get very ill from measles, mumps or rubella' and four items assessed parents' concerns about MMR e.g.\n'Giving my child the combined MMR vaccine worries me'.\nAll items were scored on a 5-point scale (strongly disagree-strongly agree).\nItems for each sub-scale were summed.\nTotal scores for the two scales range from 4 to 20 with higher scores representing stronger beliefs in the necessity for, and concerns about, MMR.\nCronbach alpha coefficients for the necessity sub-scale were 0.70 (T1), 0.63 (T2) and 0.70 (T3).\nReliability was not improved by eliminating any items.\nCronbach alpha coefficients for the concerns sub-scale were 0.77 (T1), 0.75 (T2) and 0.78 (T3).\nAnxiety was measured to ensure that the parent meeting and MMR leaflet did not evoke anxiety in parents.\nWe used the short form STAI.\nSix items were used e.g. 'I feel calm', 'I am tense' and were scored on a 4-point scale (not at all-very much).\nThe positive items (e.g. calm) were reverse scored and all six items were summed.\nThe total score was multiplied by 20/6.\nA normal score is 34 to 36.\nHigh Cronbach alpha coefficients of 0.81 (T1), 0.86 (T2) and 0.84 (T3) were obtained.\nSample size\nTo achieve 80% power to detect a standardised effect size of 0.67 on the primary outcome of decisional conflict, using a two-sided t-test with significance level of 0.05 and an estimated ICC of 0.05 (giving a design effect of 1.5 based on an average of 11 parents per cluster) required a sample size of 108 parents (54 in each group).\nPredicting 25% attrition 73 parents were required in each group.\nParent numbers were not balanced across the clusters.\nBased on our previous research we estimated recruiting 12 parents per week over three months.\nAnalysis\nAn intention to treat analysis was conducted.\nThis trial design was clustered within centres (healthcare centres, childcare organisations) and had repeated measures.\nThe number of clusters and parents within each cluster were small in respect to multilevel modelling.\nTo explore the potential effectiveness of the intervention on the primary outcome (decisional conflict) longitudinal analysis was used.\nThis accounted for the multilevel structure of the data, with outcome measures collected at different time points within parent data.\nWe were interested in exploring how decisional conflict changed over time with respect to covariates of interest, namely arm, focal MMR decision, parent characteristics (age, ethnicity, marital status, education, relationship to child, if have older child) and intended choice, knowledge, attitude, beliefs and anxiety at recruitment.\nA normal model was used, using MLwiN 2.10 beta 5 to perform these analyses.\nDue to missing values, owing to non-completion of some questionnaire items, complete case analysis corresponded to only 65% of the data.\nOf these 92 parents, 44 (48%) were in the intervention arm and 48 (52%) were in the control arm.\nMissing values appeared to be at random.\nMultiple imputation was undertaken in Stata 10.0 to account for this.\nFive imputed datasets were generated using the results from linear regression analyses.\nPrior to undertaking multiple imputations seven parents were excluded (n = 3 intervention; n = 4 control) as they had not completed any study questionnaires, providing only parent characteristics data at recruitment.\nThe best fit model for the complete case data was fitted to each imputed dataset and compared with the best fit model for those data.\nAll imputed datasets agreed on the importance of the significant variables in the complete case model and results were found to be similar, thus indicating that minimal bias was introduced due to missing values.\nAggregated results from the five imputed datasets using 135 participants (142 minus 7) are presented.\nConfidence intervals were calculated using the widest values to allow for errors generated through the imputation.\nTwo sided significance tests and an alpha level of 0.05 were used throughout.\nRepeated measures ANOVAs were computed for the secondary outcome measures using multilevel modelling using the aggregated results from the five imputed datasets.\nChi squared analysis explored MMR uptake by arm.\nResults\nIntervention delivery\nThe parent meeting was delivered eight times during July and August 2006 in non-healthcare venues (e.g. community centres) close to participating healthcare centres and childcare organisations.\nFour daytime and four evening meetings were organised.\nCr\u00e8che facilities were provided at three daytime meetings.\nForty one parents attended a parent meeting, 23 did not.\nParents attending the meeting did not differ from those not attending the meeting in their characteristics or in their decisional conflict levels at baseline.\nThe mean number of parents attending was 6 (range: 2 to 10).\nOne meeting had less than four parents attending.\nClusters and participants\nParticipant flow through the study is presented in Figure 1.\nThe two arms were equivalent on all but one cluster characteristic.\nMean list size for the childcare organisations was larger for the control arm (Table 2).\nOf 1447 eligible parents invited, 150 (10%) consented to participate.\nEight parents did not meet the inclusion criteria (did not have an 'actual' MMR decision to make at that time).\nRecruitment of 142 parents fell short of the 146 target allowing for 25% attrition, but the required sample size of 108 parents was achieved because of a less than anticipated drop out.\nThe two arms were equivalent on all parent characteristics (Table 2).\nThere was a difference in the age of the first (youngest) eligible child and therefore in the MMR decision parents were making.\nOne third of parents in the intervention arm were making a first dose decision compared with almost two thirds in the control arm.\nDose decision was therefore modelled in the analysis.\nImpact of the parent meeting and MMR leaflet\nWas the parent meeting associated with a reduction in decisional conflict?\nMean decisional conflict by arm over time is presented in Figure 2.\nAt recruitment parents in both arms reported levels of decisional conflict above two indicating that they had sufficient conflict about the choice to interfere with making the MMR decision effectively.\nAt one week post-intervention mean decisional conflict had decreased for both arms to below two; and remained below two at three months post-intervention.\nTime was significantly associated with decisional conflict.\nThere was no significant association between arm and decisional conflict at any time point (see Table 3).\nIn short, post-intervention, parents could implement an effective decision irrespective of arm allocation.\nThe greatest reduction in decisional conflict occurred at one-week post-intervention.\nFocal MMR decision (first or second dose) was not significantly associated with decisional conflict i.e. perceived decisional conflict about this choice reduced over time for parents making first or second dose decisions.\nWas the parent meeting associated with MMR decision?\nSixty six parents provided self-report data about their MMR choice.\nOf these 66, 29 (44%) were in the intervention arm and 37 (56%) were in the control arm.\nThe remaining parents did not have children who were invited for vaccination within the study time period.\nNinety three percent of parents in the intervention arm reported taking their child for the vaccination compared to 73% in the control arm.\nThis difference was statistically significant (\u03c72 (1, N = 66) = 4.43, 95% confidence interval 3.1% to 37.2%, p = 0.04).\nWas the parent meeting associated with changes in parents' intended choice, knowledge, attitudes, beliefs or anxiety?\nTime by group mean scores, 95% confidence intervals and significance levels for secondary outcomes are presented in Table 4.\nSmall changes in the predicted direction were evident for the intervention arm for knowledge, intended choice, attitudes, and beliefs.\nHowever repeated measures ANOVAs revealed no significant time by arm effects.\nMean anxiety remained below or within the normal range suggesting that neither the parent meeting nor the MMR leaflet evoked anxiety in parents.\nWhich parent characteristics and cognitions were associated with changes in decisional conflict?\nBaseline parent characteristics and outcome measures (irrespective of arm allocation) associated with changes in decisional conflict were: whether the parent had an older child; intended choice, attitude and concern beliefs (see Table 3).\nIf a parent had previously made an MMR decision for an older child, decisional conflict decreased over time by 0.25 compared to a parent who had not previously made an MMR decision.\nParents' concerns about the potiential adverse consequences of MMR at recruitment were significantly associated with changes in decisional conflict over time.\nThe more concerned parents were, decisional conflict increased by 0.07.\nFor each additional point increase in attitude i.e. the more positive parents were about MMR, decisional conflict decreased by 0.20.\nFor each additional point increase in intended choice, i.e. the stronger parents' intentions were to vaccinate, decisional conflict increased by 0.09.\nDiscussion\nIn response to a continuing lack of confidence amongst many UK parents facing a decision about MMR and informed by our earlier work we developed and evaluated a parent-centred, multi-component intervention, delivered in community-based (non-healthcare) venues.\nWe believe this to be the first study to evaluate a multi-component intervention to support informed decision making for MMR.\nSome study limitations should be acknowledged.\nParent numbers were not balanced across the clusters thus reducing statistical power.\nHowever it seems unlikely that this would have changed the non-significant time by arm effect for decisional conflict.\nOnly a small number of parents provided self-report data about their choice thus the study may have been under powered on this secondary outcome.\nThe study was based in one city and only 10% of parents invited to take part did so.\nDue to the Data Protection Act we cannot determine if they differ to non-responders.\nHowever the immunisation policy in Leeds mirrors UK policy and the sample was consistent with other MMR research that identifies parents who find it difficult to make this decision.\nFinally, whilst complete case analysis was undertaken on just 65% of the data we believe that minimal bias was introduced.\nThe intervention was feasible to deliver in non-healthcare, community venues and it was acceptable to parents, with the majority expressing positive views.\nParents were equally positive about the MMR leaflet.\nOur measure of decisional conflict showed a statistically significant decrease over time for both the intervention and control arms to a level where an informed decision for MMR could be made.\nThe positive effect of the MMR leaflet on decisional conflict observed in the control arm was unexpected.\nThe inclusion of the leaflet was to reduce possible bias from a 'Hawthorne' effect.\nFurther, we considered the provision of a leaflet reflected usual vaccination practice.\nHowever, parents reported during meetings and in questionnaires that, contrary to stated local vaccination policy, leaflets were not routinely provided.\nFor some parents this may have been because their child was not invited for MMR vaccination at the time of the study.\nNevertheless, parents reported that this leaflet was more helpful in addressing their concerns about MMR compared to the usual Department of Health information at that time.\nConsequently instead of comparing our intervention with a control (usual care) we were comparing two different decision support interventions, a decision support leaflet versus participation in a parent meeting and a decision support leaflet.\nWe were, therefore, unable to identify the independent effects on perceptions about the decision process of the parent meeting from usual care as originally intended.\nThis study did find that parents in the intervention arm were significantly more likely to report taking their child for the MMR vaccination than parents in the control arm.\nThis suggests that providing information in a well-designed leaflet may be insufficient to lead to subsequent changes in the final choice (i.e. taking the child for immunisation).\nEnabling parents to act on their informed decisions may require a more pro-active approach than increasing knowledge and enabling clarification of their values.\nIn this study, it seems likely that the parent meetings provided sufficient decision support to enable parents to act on their decision, possibly illustrating a greater values-choice outcome.\nParents making 'proxy' decisions about MMR on behalf of a child may require additional decision support where vaccination/non-vaccination consequences of regret and blame may be common and where media-induced controversy has adversely affected public trust in government and medical authorities.\nIn this vaccination context, the more proactive of the two decision support interventions was associated with more children receiving the MMR vaccination.\nWe suggest that the concerns parents felt about this choice were met more fully by both interventions in this study than those parents who receive standard invitation letters for, and advice about, MMR in the UK.\nInterestingly, for both groups the observed improvements in informed decision-making occurred for parents making first and second dose decisions.\nPrevious research would suggest that the first dose decision (for a first child) is the most anxiety-provoking and that parents may, therefore, experience greater uncertainty.\nOur study suggests that parents may benefit from concerted decision support for both doses.\nWe also found that parents who had not previously made an MMR decision for an older child, those who were less positive in their attitude and more concerned about MMR had higher decisional conflict.\nThese parents could usefully be targeted for decision support.\nOur finding that parents with strong intentions to vaccinate had higher decisional conflict suggests targeting parents who are approaching vaccination in the near future rather than those for whom vaccination and its potential consequences are more remote.\nThe majority of decision support research focuses on preference sensitive decisions for which the 'best' decision is considered to be unclear and dependent on personal values.\nImmunisation is considered to be an 'effective' health decision for which the weight of the scientific evidence would typically lead a health professional to recommend a particular course of action.\nThis perhaps explains why governments and health professionals have historically adopted the 'knowledge deficit model' approach of simply providing information and reassurance to parents.\nIrrespective of how childhood immunisation 'delivery systems' are organised, there is evidence from international literature that parents' decision support needs are generally the same.\nMoreover, the below target MMR uptake rates in many countries suggests that reliance on a passive information-giving approach has limited effect.\nComprehensive decision support, as provided in our parent meeting offers a potential solution.\nConclusions\nWhilst both the leaflet and the parent meeting reduced parents' decisional conflict, the parent meeting appeared to enable parents to act upon their decision leading to vaccination uptake.\nWe are now testing the effectiveness of a web based MMR decision aid for parents that could be made easily accessible, for example in public places where parents frequent such as schools, libraries, community centres as well as in the waiting rooms of healthcare centres.\nParticipant flow.\nMean decisional conflict by arm over time. Intervention/Control Note. Scores lower than two are associated with 'implementing decisions', higher scores are interpreted with decision delay or feeling unsure about implementation.\n\nOverview of parent meeting\nTime | Facilitator | Content | Aims\n15 minutes | Parent facilitator and Researcher | WELCOMEIntroductionsOutline aims of meetingGo through programmeAgree ground rules for meeting | Of meetingTo provide parents with the opportunity to discuss MMR with other parents who are making an MMR decisionTo provide information about MMR from a variety of perspectivesTo introduce and practice one approach to supporting parents to ask questions about MMR of their healthcare practitioner\n30 minutes | Parent facilitator | GROUP DISCUSSIONAim of sessionReminder of ground rulesIntroductory question - would anyone like to tell us what you hoped to get out of this parent meeting today?Discussion | Of sessionTo provide parents with the opportunity to discuss any issues about MMR with other parents who are also making an MMR decision\n30 minutes | Immunisation Nurse Specialist | QUESTION AND ANSWER SESSIONAim of sessionParents ask questions of the immunisation nurse specialistParents are alerted to resources that they can take away | Of sessionTo provide parents with the opportunity to ask questions of the immunisation nurse specialist\n35 minutes | Researcher | COACHING EXERCISEAim of sessionBrief input on the importance of raising questions about MMR with a healthcare practitionerIntroduce and discuss the question prompt sheetRole play exercise using the question prompt sheetBrief discussion on usefulness of the question prompt sheet and role play | Of sessionTo introduce and practice one approach to supporting parents to ask questions about MMR in the primary care consultation\n10 minutes | Parent facilitator and Researcher | CLOSE OF MEETINGThank parents and provide overview of next stage of research study | \n\n\nBaseline characteristics of clusters and parents by arm\nCharacteristics | Intervention arm | Control arm\nPrimary healthcare centres, n | 3 | 3\n\nChildcare organisations, n | 3 | 3\n\nMean healthcare centre parent list size | 216 | 210\n\nMean childcare organisation parent list size | 19 | 30\n\nMean Low Income Scheme Index scorea | 10 | 11\n\n |  | \n\nParents, n | 71 | 71\n\nMean age \u00b1 SD, yrs | 34.07 \u00b1 5.43 | 34.06 \u00b1 5.52\n\nEthnicity, n (%) |  | \nWhite British | 68 (95.8%) | 68 (95.8%)\nOther | 3 ( 4.2%) | 3 ( 4.2%)\n\nMarital status, n (%) |  | \nSingle or living with partner | 27 (38.0%) | 13 (18.3%)\nMarried or re-married | 40 (56.4%) | 57 (80.2%)\nSeparated/Divorced/Widowed | 4 .(5.6%) | 1 (1.5%)\n\nRelationship to eligible child, n (%) |  | \nMother | 67 (94.4%) | 67 (94.4%)\nFather | 4 (5.6%) | 4 (5.6%)\n\nEducation |  | \nLeft school at 16 years | 24 (33.8%) | 25 (35.2%)\nLeft school at 18 years | 10 (14.1%) | 10 (14.1%)\nAchieved Degree or higher | 37 (52.1%) | 36 (50.7%)\n\nHave older child |  | \nYes | 36 (50.7%) | 36 (50.7%)\nNo | 35 (49.3%) | 35 (49.3%)\n\nFirst (youngest) child eligible, n (%) First dose MMR decision | 23 (32.4%) | 44 (62.0%)\nSecond dose MMR decision | 48 (67.6%) | 27 (38.0%)\nMean age \u00b1 SD of first (youngest) child eligible, months | 25.73 \u00b1 14.66 | 19.77 \u00b1 11.69\n\nSecond youngest child eligible, n (%) |  | \nFirst dose MMR decision | 1 ( 4%) | 0 (0.00%)\nSecond dose MMR decision | 24 (96%) | 22 (100.0%)\n\nMean age \u00b1 SD of second youngest child eligible, months | 50.56 \u00b1 17.13 | 49.32 \u00b1 21.41\n\nNote. N = 12 clusters, N = 142 parents. aLow Income Scheme Index score is based on the percentage of prescribed items exempt from a prescription charge due to low income of the patient.\n\nCoefficients for the longitudinal multilevel model of decisional conflict on potential covariates\nModel variables | Effect Estimate | 95% CI | p-value\nTime-1 week post-intervention | -0.54 | -0.67 to -0.41 | <0.001\n\nTime-3 months post-intervention | -0.60 | -0.73 to -0.47 | <0.001\n\nArm-intervention | 0.07 | -0.11 to 0.25 | 0.215\n\nMMR decision-2nd dose | -0.05 | -0.24 to 0.14 | 0.310\n\nOlder child | -0.25 | -0.42 to -0.07 | 0.003\n\nIntended choice | 0.09 | 0.02 to 0.17 | 0.006\n\nAttitude | -0.20 | -0.30 to -0.10 | <0.001\n\nConcern beliefs | 0.07 | 0.04 to 0.10 | <0.001\n\nNote. N = 135. Results are presented per one point increase of decisional conflict.\n\nDescriptive data for secondary outcomes\n |  | Intervention | Control | \nOutcome | Time point | Mean | 95% CI | Mean | 95% CI | p-valuea\n\nKnowledgeb | T1 | 6.38 | 6.00- 6.77 | 5.97 | 5.53- 6.41 | 0.253\n\n | T2 | 8.22 | 7.84- 8.60 | 7.83 | 7.52- 8.15 | \n\n | T3 | 7.30 | 6.93- 7.66 | 7.08 | 6.80- 7.36 | \n\nIntended choicec | T1 | 5.93 | 5.54- 6.32 | 5.10 | 4.60- 5.60 | 0.605\n\n | T2 | 6.03 | 5.67- 6.39 | 5.44 | 4.96- 5.92 | \n\n | T3 | 6.34 | 6.00- 6.68 | 5.58 | 5.06- 6.09 | \n\nAttituded | T1 | 4.98 | 4.66- 5.30 | 4.52 | 4.17- 4.88 | 0.786\n\n | T2 | 5.19 | 4.89- 5.48 | 4.77 | 4.44- 5.10 | \n\n | T3 | 5.26 | 4.96- 5.55 | 4.68 | 4.33- 5.04 | \n\nNecessity beliefse | T1 | 16.88 | 16.22-17.53 | 16.73 | 15.99-17.47 | 0.578\n\n | T2 | 17.63 | 16.97-18.28 | 17.18 | 16.49-17.87 | \n\n | T3 | 17.43 | 16.87-18.00 | 17.21 | 16.46-17.96 | \n\nConcern beliefse | T1 | 9.83 | 8.92-10.73 | 11.00 | 10.04-11.96 | 0.939\n\n | T2 | 8.92 | 8.15- 9.70 | 10.15 | 9.07-11.22 | \n\n | T3 | 8.66 | 7.81- 9.50 | 10.06 | 9.04-11.09 | \n\nAnxietyf | T1 | 32.50 | 29.63-35.37 | 34.20 | 31.52-36.89 | 0.219\n\n | T2 | 30.89 | 28.00-33.77 | 33.78 | 30.43-37.13 | \n\n | T3 | 31.46 | 28.49-34.43 | 33.39 | 30.09-36.68 | \n\nNote. N = 135. aSignificance value for time by arm interaction. Values range from b0 (no knowledge) to 11 (good knowledge); c1(definitely do not intend) to 7 (definitely do intend); d1 (extremely negative attitude) to 7 (extremely positive attitude); e4 (not at all necessary/concerned) to 20 (very necessary/concerned); f20 (low anxiety) to 80 (high anxiety).",
        "label": "high",
        "id": "task4_RLD_test_37"
    },
    {
        "paper_doi": "10.1371/journal.pntd.0001044",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Trial design: randomized, non-blinded study.Participants: adult participants with chronic strongyloidiasis.Length of follow-up: participants were followed up 2 weeks after treatment initiation, then 1 month, 3 months, 6 months, 9 months, and 1 year post-treatment.Monitoring and diagnostics: infection with Strongyloides stercoralis was ascertained using the direct smear, formol-ether concentration method, and modified Koga agar plate culture method.\n\n\nParticipants: Number of participants: 90 adult participants with chronic Strongyloides infection were recruited. There were 10 participants with HIV co-infections, with 3 HIV-positive participants randomized to albendazole, 2 HIV-positive participants randomized to single dose ivermectin, and 5 HIV-positive participants randomized to double dose ivermectin.Inclusion criteria: adult participants with characteristic rhabditiform larvae of S. stercoralis present on faecal microscopy.Exclusion criteria: history of allergic reaction to either study medication, treatment within the month prior to the trial with any drug known to have anti-Strongyloides activity, pregnancy, or lactation, and any suggestion of disseminated strongyloidiasis.\n\n\nInterventions: Intervention: Group 1: ivermectin delivered as a single dose of 200 ug/kg; Intervention. Group 2: 2 doses of ivermectin (200 ug/kg) delivered 2 weeks apart. For the purpose of this analysis we considered both groups that received ivermectin together.Control: participants received 7 days of albendazole (800 mg per day).\n\n\nOutcomes: Outcomes included in this review: incidence of adverse events defined as \"symptoms or signs that developed after the study drug administration and had not been reported prior to the administration of the first dose of the antihelmintic.\"Other trial outcomes: treatment cure (defined as clinical improvement (if symptomatic before treatment) and the absence of rhabditiform larvae in the stool at day 14 of treatment and throughout the follow-up period) and treatment failure (defined as the presence of larvae two weeks after initiation of treatment or the reappearance of larvae during follow-up).\n\n\nNotes: Location: Siriraj Hospital, ThailandParticipant helminth status: ascertainedParticipant ART status: it was unspecified if any individuals were receiving ART treatment at the start or during the trial.Author contact: we requested additional data regarding the incidence of adverse events in the HIV-positive participants specifically from the trial authors, who provided this information\n\n",
        "objective": "To evaluate the effects of deworming drugs (antihelminthic therapy) on markers of HIV disease progression, anaemia, and adverse events in children and adults.",
        "full_paper": "Background\nStrongyloidiasis, caused by an intestinal helminth Strongyloides stercoralis, is common throughout the tropics.\nIt remains an important health problem due to autoinfection, which may result in hyperinfection and disseminated infection in immunosuppressed patients, especially patients receiving chemotherapy or corticosteroid treatment.\nIvermectin and albendazole are effective against strongyloidiasis.\nHowever, the efficacy and the most effective dosing regimen are to be determined.\nMethods\nA prospective, randomized, open study was conducted in which a 7-day course of oral albendazole 800 mg daily was compared with a single dose (200 microgram/kilogram body weight), or double doses, given 2 weeks apart, of ivermectin in Thai patients with chronic strongyloidiasis.\nPatients were followed-up with 2 weeks after initiation of treatment, then 1 month, 3 months, 6 months, 9 months, and 1 year after treatment.\nCombination of direct microscopic examination of fecal smear, formol-ether concentration method, and modified Koga agar plate culture were used to detect strongyloides larvae in two consecutive fecal samples in each follow-up visit.\nThe primary endpoint was clearance of strongyloides larvae from feces after treatment and at one year follow-up.\nResults\nNinety patients were included in the analysis (30, 31 and 29 patients in albendazole, single dose, and double doses ivermectin group, respectively).\nAll except one patient in this study had at least one concomitant disease.\nDiabetes mellitus, systemic lupus erythrematosus, nephrotic syndrome, hematologic malignancy, solid tumor and human immunodeficiency virus infection were common concomitant diseases in these patients.\nThe median (range) duration of follow-up were 19 (2\u201376) weeks in albendazole group, 39 (2\u201374) weeks in single dose ivermectin group, and 26 (2\u201374) weeks in double doses ivermectin group.\nParasitological cure rate were 63.3%, 96.8% and 93.1% in albendazole, single dose oral ivermectin, and double doses of oral ivermectin respectively (P\u200a=\u200a0.006) in modified intention to treat analysis.\nNo serious adverse event associated with treatment was found in any of the groups.\nConclusion/Significance\nThis study confirms that both a single, and a double dose of oral ivermectin taken two weeks apart, is more effective than a 7-day course of high dose albendazole for patients with chronic infection due to S. stercoralis.\nDouble dose of ivermectin, taken two weeks apart, might be more effective than a single dose in patients with concomitant illness.\nTrial Registration\nClinicalTrials.gov NCT00765024\nAuthor Summary\nStrongyloidiasis, caused by an intestinal helminth Strongyloides stercoralis, is common throughout the tropics.\nWe conducted a prospective, clinical study to compare the efficacy and safety of a 7-day course of oral albendazole with a single dose of oral ivermectin, or double doses, given 2 weeks apart, of ivermectin in Thai patients who developed this infection.\nPatients were regularly followed-up after initiation of treatment, until one year after treatment.\nNinety patients were studied (30, 31 and 29 patients in albendazole, single dose, and double doses ivermectin group, respectively).\nThe average duration of follow-up were 19 (range 2\u201376) weeks in albendazole group, 39 ( range 2\u201374) weeks in single dose ivermectin group, and 26 ( range 2\u201374) weeks in double doses ivermectin group.\nParasitological cure rate were 63.3%, 96.8% and 93.1% in albendazole, single dose oral ivermectin, and double doses of oral ivermectin respectively.\nNo serious adverse event associated with treatment was found in any of the groups.\nTherefore this study confirms that both a single, and a double dose of oral ivermectin taken two weeks apart, is more effective than a 7-day course of high dose albendazole for patients with chronic infection due to S. stercoralis.\nIntroduction\nInfection with the intestinal helminth Strongyloides stercoralis remains a common problem throughout the tropics, including Thailand.\nIt is estimated that 30 to 100 million people are infected worldwide.\nMost infected individuals are asymptomatic or developed minimally symptomatic chronic infection through autoinfection.\nPotentially fatal disseminated infections, due to an acceleration of the autoinfection cycle, are seen in immunocompromised patients, such as those with concurrent human T-lymphotropic virus-1 (HTLV-1) infection, or those on corticosteroid therapy.\nOther recognized risk factors for disseminated strongyloidiasis include malignancies especially lymphoma, organ transplantation and diabetes mellitus.\nGastrointestinal symptoms associated with strongyloidiasis include diarrhea, abdominal discomfort, nausea/vomiting and anorexia.\nThe diagnosis of strongyloidiasis should be suspected if there are clinical signs and symptoms, or eosinophilia.\nDefinitive diagnosis of strongyloidiasis is usually made on the basis of detection of larvae in the stool.\nThe combination of diagnostic approaches such as repeated direct microscopic examination of fecal smear, fecal concentration methods such as formol-ether concentration (FEC), and modified Koga agar plate culture have been used to improve the likelihood of detecting this parasite.\nIn the past, the treatment of choice for strongyloidiasis has been thiabendazole, but this drug has unpleasant side effects and is no longer available.\nAlbendazole, another broad-spectrum antihelmintic agent, was previously shown to be effective against S. stercoralis .\nMore recent reports suggest ivermectin, a macrolide-like agent developed primarily for the treatment of onchocerciasis, is as effective as thiabendazole and superior to albendazole against intestinal strongyloidiasis.\nAlthough a single dose of ivermectin 200 microgram/kilogram body weight (\u00b5g/kg) was shown to be effective in uncomplicated chronic strongyloidiasis, repeated treatment at two or three week intervals was thought to be necessary to eliminate larvae generated by autoinfection.\nA preparation of oral ivermectin licensed for human use has recently become available in Thailand.\nHowever, albendazole remains the most widely used antiparasitic drugs for the treatment of this infection in this country.\nThe purpose of the present study was to assess the safety and efficacy of a single dose of ivermectin (200 \u00b5g/kg), or two doses of ivermectin given 2 weeks apart, and a 7-day course of high dose albendazole for the treatment of chronic strongyloidiasis in adult patients who were at high risk of hyperinfection or disseminated infection.\nMaterials and Methods\nStudy design and ethics\nThis was a prospective open-label, randomised, controlled study conducted between July 2008 and April 2010 at Siriraj Hospital, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand.\nThe study was approved by the Ethical Committee on Research Involving Human Subjects, Siriraj Hospital, Faculty of Medicine, Mahidol University, Thailand.\nAll patients were informed about the purpose of the trial and gave written informed consent before enrollment.\nThe study enrollment was stopped in December 2009 after 100 eligible patients had been recruited.\nPatients\nAdult patients (>18 years) were recruited from Siriraj Hospital if characteristic rhabditiform larvae of S. stercoralis were present on fecal microscopy.\nExclusion criteria included a history of allergic reaction to either study medication, treatment within the month prior to the study with any drug known to have anti-strongyloides activity, pregnancy or lactation and any suggestion of disseminated strongyloidiasis.\nTreatment\nComputer generated, simple, random allocation sequences were prepared for 3 study groups by the investigator team.\nThese were sealed in an opaque envelope and numbered.\nThe investigator (YS) assigned study participants to their respective treatment group after opening the sealed envelope.\nOnce an eligible patient was identified and informed consent was obtained, the patient was randomly allocated to one of the following group (1\u22361\u22361 ratio):\nIvermectin-I group: a single oral dose of 200 \u00b5g/kg (Vermectin\u00ae, Atlantic Laboratories Co, Ltd., Thailand).\nIvermectin-II group: two oral doses of 200 \u00b5g/kg of ivermectin (Vermectin\u00ae, Atlantic Laboratories Co, Ltd., Thailand) given 2 weeks apart.\nAlbendazole group: oral albendazole (Albatel\u00ae, TO Chemical, Thailand) 400 mg twice daily for 7 days.\nStudy Procedures\nBaseline evaluation included history, detailed physical examination, and laboratory investigations such as complete blood count (CBC), urinalysis, and biochemistry.\nPatients were requested to collect two consecutive fecal samples at every hospital visit.\nThe coprodiagnosis for the detection of S. stercoralis larvae using direct smear, formol-ether concentration method, and modified Koga agar plate culture method was performed for each patient at the Infectious Diseases and Tropical Medicine Laboratory, Division of Infectious Diseases and Tropical Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Thailand.\nPatients were required to make seven hospital visits to complete the study: at baseline evaluation and initiation of treatment, at 2 weeks after initiation of treatment, then at 1 month, 3 months, 6 months, 9 months, and 1 year after treatment.\nPatients who completed 1 year of follow-up were invited for further follow-up visits every 3 months or at their convenience.\nOutcome measures\nEfficacy\nThe efficacy of treatment was analyzed on a modified intention to treat, and a per-protocol basis.\nModified intention to treat analysis was based on the number of patients who were randomized and received treatment.\nPer-protocol analysis was based on the number of patients who completed the treatment and were followed up as planned.\nAnalyses of adverse events were on the modified intention to treat basis.\nThe outcome was evaluated according to the following definitions; \u201cCure\u201d was defined as clinical improvement (if symptomatic before treatment) and the absence of rhabditiform larvae in the stool at day14 of treatment and throughout the follow-up period.\n\u201cFailure\u201d was defined as the presence of larvae two weeks after initiation of treatment or the reappearance of larvae during follow-up.\n\u201cAdverse events\u201d were defined as symptoms or signs that developed after the study drug administration and had not been reported prior to the administration of the first dose of the antihelmintic.\nSample Size and Statistical Methods\nAssuming the therapeutic efficacy of albendazole to be 60% and of both regimens of ivermectin treatment to be 90%, with alpha error 0.5, and 80% power, it was calculated that 22 patients would be needed in each study group.\nThe lost to follow up rate was assumed to be 20%.\nTherefore 27 patients would be needed in each study group.\nDemographic data, results of investigations, stool examination at baseline and follow up visits were recorded in the database.\nAll statistical analyses were performed using SPSS, version 17.5.\nPearson chi square or Fisher's exact tests were used to compare rates and proportions, as appropriate.\nMann-Whitney U tests were used to analyze continuous variables that were not normally distributed.\nIndependent sample t- tests were used to compare normally distributed variables, taking a probability of less than 5% as the level of significance.\nKaplan-Meier plot and Cox proportional hazard model were constructed to identify independent risk factors for treatment failure.\nResults\nOne hundred and fifty one patients had detectable rhabditiform larvae of S. stercoralis on fecal microscopy during the study period.\nOne hundred patients were enrolled (36, 32, and 32 patients in albendazole, ivermectin-I, and ivermectin-II groups respectively).\nTen patients were excluded from analysis because they did not receive or complete the study treatment (3 in albendazole group, 2 in ivermectin-II group), or they were lost to follow-up immediately after treatment (3 in albendazole group, 1 each in ivermectin-I and ivermectin-II respectively).\nOverall, 90 patients were eligible for the modified intention to treat analysis.\nDetail of the total number of enrollment, randomization, follow-up and inclusion in the final analysis comparing among the three treatment groups is shown in Figure 1.\nThe demographic data, concomitant diseases, baseline clinical and laboratory investigations are shown in Table 1 and 2.\nAll except one patient had an associated medical problem, including concurrent other parasitoses.\nThese patients also had abnormal serum aspartate aminotranferase (AST) and alanine aminotransferase (ALT) levels prior to entering the study due to their underlying conditions.\nThe intensity of initial infection of the three study groups was similar, i.e. S. stercoralis larvae were found from the direct fecal examination in 24 (80%), 25 (80.6%), and 28 (96.6%) in albendazole, Ivermectin-I, and ivermectin\u2013II groups, respectively (P\u200a=\u200a0.123).\nLarvae were also detected from modified Koga agar plate culture in 22/26 (84.6%) patients in the albendazole group, in 22/26 (84.6%) patients in the ivermectin-I group, and in 24/29 (82.8%) patients in the ivermectin-II group, respectively (P\u200a=\u200a0.976).\nDiarrhea was detected in half of the patients and it was relieved after treatment in most patients.\nAbnormal bowel movement at second week of follow-up was reported in 4 patients in the albendazole group, 2 patients in the ivermectin-I group, and 3 patients in the ivermectin-II group, respectively (P\u200a=\u200a0.641).\nS. stercoralis larvae were detected in one patient in the ivermectin-I group at third month of follow-up.\nIn ivermectin-II group, S. stercoralis larvae were detected at second week prior to the second dose of ivermectin in two patients.\nNo patients had reinfection/relapse after the second dose of ivermectin treatment.\nIn albendazole treated patients, S. stercoralis larvae were detected at second week of follow-up in 2 patients, at first month of follow-up in 2 patients, between 3\u20136 months of follow-up in 3 patients, and between 6\u201312 months of follow-up in 4 patients.\nAll of the relapses/ reinfections found during follow-up were clinically inapparent.\nParasitologically, parasite elimination was documented in 19 (63.3%) albendazole treated patients, in 30 (96.8%) single-dose ivermectin treated patients, and 27 (93.1%) two-dose ivermectin treated patients (P\u200a=\u200a0.006) (Table 3).\nCox regression analysis showed that albendazole treated group had 14.7 times (95%CI 1.8\u2013111.9), and 5.7 times (95%CI 1.3\u201325.7) higher risk for reinfection/ relapse of strongyloidiasis than ivermectin-I and ivermectin-II group, respectively.\nKaplan- Meier Plot compares the parasitological cure rate between these study groups is shown in figure 2.\nNo hyperinfection syndrome or disseminated infection was found among these patients during the study period.\nS. stercoralis larvae were detected after treatment using FEC in 8 patients, and by modified Koga agar plate culture only in 6 of them.\nAll patients with relapse/reinfection were retreated with two doses of ivermectin in two weeks apart.\nOverall albendazole and ivermectin were well tolerated.\nTransient elevation of AST, and ALT levels was detected in one patient in ivermectin-II group.\nThe AST and ALT levels returned to normal 2 weeks after the second dose of ivermectin treatment.\nSevere nausea and vomiting was reported in one patient in the albendazole group.\nFifteen patients died after enrollment (5 patients in each treatment group).\nCauses of death were not related to the study drugs, and were considered to be due to an underlying disease or its complications (solid tumor in 5, hematologic malignancies in 3, diabetes mellitus, or systemic lupus erythrematosus (SLE), or hypertension with complications such as myocardial infarction or sepsis in 7 patients).\nThe median duration from enrollment to death was 2 weeks (range 2\u201314 weeks) in the albendazole group, 5 weeks (range 2\u201338 weeks) in the ivermectin-I group, and 2 weeks (range 1\u201327 weeks) in the ivermectin-II group, respectively.\nDiscussion\nStrongyloidiasis remains a significant health problem in many developing countries, mainly due to the potential for lethal disseminated disease.\nGastrointestinal symptoms associated with strongyloidiasis found in this study included diarrhea, abdominal discomfort, nausea/vomiting and anorexia.\nChronic infection with S. stercoralis was clinically inapparent in half of the patients at enrollment, and in all of relapses/ reinfections found during follow up.\nPeripheral eosinophilia (>500 eosinophils/\u00b5L.) was detected in half of the patients at enrollment.\nS. stercoralis larvae were detected after treatment using FEC in 8 patients, and by modified Koga agar plate culture in 6 patients.\nThis information confirmed that fecal examination, including culture and/or serology, every 3\u20136 months of follow-up should be recommended for early detection and treatment of latent infection to prevent hyperinfection or disseminated disease in these patients.\nResults of this study corroborate the results from previous randomised controlled studies on the higher efficacy of ivermectin compared to various dosage regimens of albendazole for treating chronic strongyloidiasis.\nA summary of results from these previous controlled trials of ivermectin treatment for chronic strongyloidiasis is shown in Table 4.\nAlthough these studies were conducted in different geographical areas and population groups, i.e. in children and adults, they were considered to be within a community-based setting, such as schools or primary care clinic.\nThe duration of follow-up varied from 3 weeks to 12 months.\nThe present study was conducted in a tertiary hospital.\nThe majority of patients had known risk factors for disseminated strongyloidiasis, and approximately one-third of them received corticosteroid or chemotherapy.\nResults of this study confirmed that ivermectin was also effective in this population who were at high risk of severe infection.\nAlbendazole remains an option of treatment for chronic strongyloidiasis in many countries in South East Asia, where oral ivermectin is not widely available.\nCure rates of a regimen consisting of albendazole 400 mg daily for three to five days varied from 38\u201387% in those without underlying diseases.\nIn this study, the cure rate was found to be 63% when a 7-day course of high dose albendazole was used.\nThe efficacy of albendazole varied widely even when the same dose and duration of treatment was used.\nDifferences in duration of follow-up examinations could be one explanation, and re-infection from the environment may also be a factor when the efficacy is monitored for an extended period in endemic areas.\nThe study which reported the highest cure rate (87%) was conducted in Okinawa, Japan, where the chance of re-infection from the environment was less likely to occur compared to other studies conducted in endemic areas.\nTwo patients in the ivermectin-II group had detectable S. stercoralis larvae in the second week prior to the second dose of ivermectin treatment.\nOne patient in ivermectin-I group also had detectable S. stercoralis larvae 3 months after treatment.\nThis observation supports the recommendation that repeated doses of ivermectin should be the preferred treatment in patients with chronic strongyloidiasis who have an underlying or concomitant illness.\nThe limitation of this study was the high loss to follow-up rates over time.\nHigh mortality associated with the concomitant illnesses was an unavoidable cause of concern in this study.\nThe median duration of follow up was 19 weeks in albendazole group, 39 weeks in ivermectin-I, and 26 weeks in ivermectin-II group.\nThe non-significant shorter duration of follow up found in albendazole treatment group was due to the significant higher rate of treatment failure compared to ivermectin.\nHowever, the study still had sufficient power to detect a difference between albendazole and ivermectin treatments.\nThis study, however, was too small to detect any but the most severe and common side- effects of both albendazole and ivermectin.\nOnly one of albendazole treated patients and one treated with ivermectin had transient changes in transaminases, a well-recognized and reversible adverse event.\nIn conclusion, this clinical study confirms that both a single and a double dose of oral ivermectin taken at a two-week interval is more effective than a 7-day course of high dose of albendazole for patients with chronic infection due to S. stercoralis.\nTotal number of enrollment, randomization, follow-up, and inclusion in the final analysis comparing among three treatment groups.\nKaplan-Meier Plot comparing the parasitological cure among the albendazole, ivermectin-I, and ivermectin-II treatment groups over one year follow-up period.\n\nDemographic and baseline clinical features of the three study groups.\nCharacteristics | Albendazol(N\u200a=\u200a30) | Ivermectin-I(N\u200a=\u200a31) | Ivermectin \u2013II(N\u200a=\u200a29) | P-value\nMale: female | 21\u22369 | 21\u223610 | 19\u223610 | 0.934\nMedian (range) age, yr | 54 (23\u201381) | 51 (29\u201377) | 52 (25\u201378) | 0.273\nMedian (range) weight, kg | 59 (37\u201380) | 57 (37\u201385) | 59 (44\u201373) | 0.75\nConcomitant illnesses, n (%) |  |  |  | 0.607\n- None | 1 | 0 | 0 | \n- Diabetes mellitus | 8 | 6 | 5 | \n- NS/SLE | 4 | 5 | 4 | \n- AIDS/HIV infection | 3 | 2 | 5 | \n- Hematological malignancy | 2 | 3 | 4 | \n- Solid tumor | 3 | 1 | 6 | \n- Rheumatologic diseases | 2 | 1 | 1 | \n- Chronic kidney disease | 3 | 2 | 2 | \n- Alcohol drinker | 3 | 1 | 1 | \n- Others | 6 | 12 | 4 | \nImmunosuppressive drug, n (%) | 10 (33.3) | 11(35.5) | 11 (37.9) | 0.934\nConcomitant parasitoses* |  |  |  | 0.380\n- Hookworm infection | 2 | 1 | 0 | \n- Opisthorchiasis | 2 | 1 | 2 | \n- Enterobious infection | 0 | 1 | 0 | \n- Entamoeba histolytica infection | 0 | 1 | 0 | \n- Isospora belli infection | 0 | 0 | 1 | \n\n*Diagnosis obtained by ova or cyst found from fecal examination.\n\nComparison of symptoms related to chronic strongyloidiasis and baseline laboratory results.\nParameters | Albendazol(N\u200a=\u200a30) | Ivermectin-I(N\u200a=\u200a31) | Ivermectin \u2013II(N\u200a=\u200a29) | P-value\nSymptoms associated with strongyloidiasis, n (%) |  |  |  | \n- Diarrhea | 14 (46.7) | 11 (35.5) | 16 (51.6) | 0.609\n- Abdominal pain | 3 (10) | 4 (12.9) | 6 (20.7) | 0.483\n- Nausea/vomiting | 4 (13.3) | 4 (12.9) | 6 (20.7) | 0.650\nLaboratory test, mean (SD) |  |  |  | \n- Hematocrit, % (35\u201345) | 32.7 (7) | 35.4 (6) | 32 (8) | 0.132\n- Eosinophil count, \u00d7106/L (<500) | 967(1,239) | 1,203(2,714) | 554(1,781) | 0.366\n- Total eosinophil >500/\u00b5L, n (%) | 14 (46.7) | 18 (58.1) | 13 (44.8) | 0.535\n- AST, U/L (0\u201337) | 45(43) | 45(60) | 38(34) | 0.805\n- ALT, U/L (0\u201340) | 38(31) | 42(38) | 38(47) | 0.924\n- Creatinine, mg/dL (0.8\u20131.2) | 1.2(0.5) | 1.1(0.9) | 0.8(0.8) | 0.933\n\n\nOutcome of treatment among the three study groups.\nParameters | Albendazol(N\u200a=\u200a30) | Ivermectin-I(N\u200a=\u200a31) | Ivermectin-II(N\u200a=\u200a29) | P-value\nDuration of follow up |  |  |  | \n- Median (range), weeks | 19(2\u201376) | 39(2\u201374) | 26 (2\u201374) | 0.248\nOutcome: Parasitological responses |  |  |  | 0.006\n- Elimination, n (%) | 19 (63.3) | 30 (96.8) | 27 (93.1) | \n- Failure | 11 (36.7) | 1 (3.2) | 2 (6.9) | \n- Persistence at 2 week | 2 | 0 | 2 | \n- Relapse /reinfection | 9 | 1 | 0 | \n\n\nSummary of published controlled trials of oral ivermectin treatment for chronic strongyloidiasis.\nComparative Drug, Dosage Regimen | Durationof follow-up | N | Cure,N (%) | Author,year, [Ref]\n1. Albendazole 400 mg/d - 3days2. Ivermectin 150\u2013200 \u00b5g/kg, single dose | 30 days | 2429 | 9 (38)24 (83) | Datry A,1994 \n1. Thiabendazole 50 mg/kg/day - 3 days2. Ivermectin 200 \u00b5g/kg, single dose3. Ivermectin 200 \u00b5g/kg, - 2 days | 7 days, then1, 3, 6months | 191618 | 18(94.7)16(100)18(100) | Gann PH,1994 \n1. Albendazole 400 mg/d -3 days2. Ivermectin 200 \u00b5g/kg, single dose | 3 weeks | 149152 | 67(45)126(82.9) | Marti H,1996 [18]\n1. Pyrvinium pamoate 5 mg/kg/d -3 days2. Albendazole 400 mg/d - 3days3. Ivermectin 6 mg 2 doses- 2 weeks apart | 2 weeks,then 6, 12months | 608467 | 14 (23.3)65 (77.4)65 (97) | Toma H,2000 \n1. Albendazole 400 mg/d - 5 days2. Ivermectin 150\u2013200 \u00b5g/kg, single dose | 30 days | 3378 | 26(78.8)77(98.7) | Nontasut P,2005 \n",
        "label": "high",
        "id": "task4_RLD_test_968"
    },
    {
        "paper_doi": "10.1186/1475-2875-7-237",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Trial design: An open label RCTFollow-up: Malaria film and Hb level on days 0, 1, 2, 3, 7, 14, and 28, plus QT-NASBA for detection of sub-microscopic gametocytaemiaAdverse event monitoring: Adverse events were recorded at each visit in the case record form. An adverse event defined as any unfavourable and unintended sign.\n\n\nParticipants: Number of participants: 146Inclusion criteria: Age 6 months to 12 years, axillary temp > 37.5 degC or history of fever, P. falciparum mono-infection 1000 to 200,000/uL, informed consentExclusion criteria: Severe malaria, any other underlying illness\n\n\nInterventions: 1. DHA-P, fixed dose combination, 20 mg/160 mg tablets (Sigma-Tau)4 to 7 kg  1/2 tablet once daily for 3 days7 to 13 kg 1 tablet once daily for 3 days13 to 24 kg 2 tablets once daily for 3 days24 to 35 kg 4 tablets once daily for 3 days2. Artemether-lumefantrine, fixed dose combination, 20/120 mg tablets (Novartis)5 to 14 kg 1 tablet twice daily for 3 days15 to 24 kg 2 tablets twice daily for 3 days25 to 34 kg 3 tablets twice daily for 3 daysAll doses supervised and given with a glass of milk.\n\n\nOutcomes: Recurrent parasitaemia at day 28, PCR-adjusted and PCR-unadjustedGametocyte prevalence during follow-upMean Hb at day 28Adverse eventsNot included in this review:Fever clearanceParasite clearance\n\n\nNotes: Country: KenyaSetting: Health centreTransmission: High transmissionResistance: Not reportedDates: Apr 2007 to Jul 2007Funding: The Knowledge and Innovation Fund, Koninklijk Instituut voor de Tropen/Royal Tropical Institute. DHA-P provided free of charge by Sigma-Tau\n\n",
        "objective": "To evaluate the effectiveness and safety of DHA\u2010P compared to other ACTs for treating uncomplicated P. falciparum malaria in adults and children.",
        "full_paper": "Background\nMany countries have implemented artemisinin-based combination therapy (ACT) for the first-line treatment of malaria.\nAlthough many studies have been performed on efficacy and tolerability of the combination arthemeter-lumefantrine (AL) or dihydroartemisinin-piperaquine (DP), less is known of the effect of these drugs on gametocyte development, which is an important issue in malaria control.\nMethods and results\nIn this two-arm randomized controlled trial, 146 children were treated with either AL or DP.\nBoth groups received directly observed therapy and were followed for 28 days after treatment.\nBlood samples were analysed with microscopy and NASBA.\nIn comparison with microscopy NASBA detected much more gametocyte positive individuals.\nMoreover, NASBA showed a significant difference in gametocyte clearance in favour of AL compared to DP.\nThe decline of parasitaemia was slower and persistence or development of gametocytes was significantly higher and longer at day 3, 7 and 14 in the DP group but after 28 days no difference could be observed between both treatment arms.\nConclusion\nAlthough practical considerations could favour the use of one drug over another, the effect on gametocytogenesis should also be taken into account and studied further using molecular tools like NASBA.\nThis also applies when a new drug is introduced.\nTrial registration\nCurrent controlled trials ISRCTN36463274\nBackground\nIn response to widespread resistance of Plasmodium falciparum parasites to the commonly used drugs chloroquine (CQ) and sulphadoxine-pyrimethamine (SP), many African countries recently adopted artemisinin-based combination therapy (ACT) as first-line treatment for uncomplicated malaria.\nThe combination artemether-lumefantrine (AL) proved to be highly effective and well-tolerated in several studies in Africa.\nDisadvantages of this drug combination are the twice-daily dosing and the fact that it should be administered with a fat-rich meal or at least a cup of soya milk.\nIn Uganda, in an area of intense malaria transmission, recurrence of parasitaemia within 28 days occurred in 29% of AL treated patients, in 8.9% adjusted by genotyping, indicating recrudescence.\nAnother ACT, dihydroartemisinin combined with piperaquine (DP), which was originally developed in China, is increasingly used in Southeast Asia.\nPiperaquine is an orally active bisquinoline with a half-life elimination time of 2.5\u20134 weeks.\nThe drug is structurally related to CQ, but still active against highly CQ-resistant P. falciparum strains.\nThis relatively inexpensive drug was well tolerated and highly effective in Southeast Asia as was the case in two studies in Africa.\nConsequently, both AL and DP are considered to be amongst the most promising artimisinin-based drugs.\nArtemisinin-based drugs also act on gametocytes and thus on transmission, at least in low transmission areas.\nIn high transmission areas of Africa, not much information is yet available on gametocytaemia after ACT treatments and on possible influence on transmission.\nIn a comparative study of AL and DP in Uganda, the appearance of gametocytes in those who did not have gametocytes at the start of treatment was lower from day 15 to day 42 of follow up in those treated with DP than in those treated with AL.\nA limitation of this study was the fact that gametocytaemia was assessed by microscopic examination only.\nIt has recently been shown that sub-microscopic gametocyte densities may significantly contribute to the transmission of malaria.\nAdequate assessment of gametocytaemia is important.\nQuantitative nucleic sequence based amplification technique (QT-NASBA) has been shown to be much more sensitive for the detection of gametocytes than microscopy.\nIn this study, the post-treatment prevalence of gametocytes in children in Mbita, western Kenya was assessed, after treatment with AL and DP.\nMicroscopy and QT-NASBA for the quantification of gametocytaemia were compared and effectiveness of both drugs regarding clinical symptoms, clearance of parasites and tolerability was assessed.\nMaterials and methods\nStudy site and population\nThe study was conducted in Mbita, western Kenya, at the shores of Lake Victoria during the high malaria transmission season of April-July 2007.\nMbita, is an area with highly variable transmission that depends on the local environmental circumstances that can support mosquito conditions.\nThe EIR is calculated to be 6 infectious bites per person per month.\nChildren (6 months-12 years of age) visiting the out-patient clinic of the health centre and diagnosed with uncomplicated malaria were included after informed consent from parents or guardians.\nInclusion criteria were: uncomplicated P. falciparum malaria with initial parasitaemia between 1,000 and < 200,000 parasites/\u03bcl blood, axillary temperature \u2265 37.5\u00b0C (measured with a digital thermometer) or a history of fever.\nChildren with severe malaria, mixed infection or other underlying illness were excluded from the study.\nIn total 146 children were recruited for the present study.\nEthical approval for this study was obtained from appropriate local authorities and The Kenya Medical Research Institute (KEMRI, Nairobi, Kenya) Ethical Steering Committee (SSC protocol No 948).\nThe trial was registered as an International Standard Randomized Controlled Trial at current controlled trials (ISRCTN36463274).\nStudy design and treatment\nFollowing diagnosis (at day 0), the patients were randomly allocated to one of the two treatment groups following a computer generated randomization list.\nOne group was assigned DP (Sigma-Tau, Italy) once per day for three days.\nOne tablet of the study drug contained 20 mg of dihydroartemisinin and 160 mg of piperaquine (paediatric formulation).\nTreatment was according to body weight as follows: children between 4\u20137 kg received half a tablet per dose, those between 7\u201313 kg 1 tablet, 13\u201324 kg 2 tablets per dose and children between 24\u201335 kg 4 tablets.\nThe other group was assigned to AL (Novartis Pharma, Switzerland).\nEach tablet contained 20 mg artemether and 120 mg lumefantrine.\nPatients received treatment according to bodyweight; i.e. children between 5\u201314 kg received one tablet per dose, those between 15\u201324 kg two tablets and those between 25\u201334 kg received three tablets per dose.\nDoses were given twice daily.\nAll treatments were given with a glass of milk under direct supervision at the clinic or, for the 2nd dose of AL, at home.\nOutcomes: efficacy\nEfficacy was assessed using the WHO in vivo test with a follow-up period of 28 days.\nAt enrollment (day 0) a full clinical examination was performed; information was recorded on a case record form.\nAt initial diagnosis (day 0) and during follow-up (day 1, 2, 3, 7, 14, and 28), finger prick blood samples were collected for microscopy, measurement of haemoglobin level and molecular analysis.\nHaemoglobin was measured with Hemocue 201+ analyser and cuvettes (HemoCue diagnostics B.V. Waarle, The Netherlands).\nResponse to treatment was measured and defined according to WHO guidelines.\nPatients showing complications or treatment failure were treated with appropriate supportive therapy.\nChildren developing danger signs or severe malaria on day 1 or 2 of the study were withdrawn from the study, referred to the hospital, and given alternative treatment.\nAdverse events were recorded on the case record forms.\nAn AE was defined as an unfavourable and unintended symptom, sign or disease.\nA serious adverse event (SAE) was defined as a symptom or sign that is temporally associated with the drugs administered to the patient that is life threatening or results in hospitalization, permanent and significant disability or death.\nSAE's were immediately reported to the ethical committee of KEMRI and the drug safety department of Sigma-Tau, Italy.\nOutcomes: parasite clearance and gametocyte dynamics\nParasite clearance and gametocyte dynamics were assessed microscopically as well as with quantitative nucleic acid sequence based amplification assay, QT-NASBA.\nMicroscopy\nGiemsa-stained thick and thin smears were prepared according to WHO guidelines.\nTwo independent experienced microscopists, who were blinded to the treatment and clinical status of the patient, examined the smears for the presence of parasites and identified the observed parasite species.\nParasitaemia was determined by counting the number of parasites against 200 leukocytes for the asexual stages (assuming that there are 8,000 leukocytes/\u03bcl blood).\nThe presence of gametocytes was examined against 500 leukocytes.\nQT-NASBA\nFinger prick blood (50 ul) for NASBA analysis was collected on Whatman 903 filter paper (Whatman international Ltd. Maidston, United Kingdom) and air-dried at room temperature.\nNucleic acid extraction was performed as previously described by Boom et al .\nReal-time 18S rRNA QT-NASBA was applied to study asexual parasite clearance below microscopical threshold.\nIn order to quantify the number of parasites in blood, a 10 fold serial dilution of 106 to 10 in vitro cultured parasites/ml was used as reference and processed and analysed with NASBA.\nFurthermore, to assess prevalence of gametocytes below the detection limit of microscopy, QT-NASBA targeting Pfs25 mRNA as described by Schneider et al  was used on blot spots collected during follow-up.\nGenotyping\nIn order to discriminate between re-infection (RI) and recrudescence (RE), merozoite surface protein 1 and 2 (msp1 and msp2) and glutamate rich protein (GLURP) genotyping was performed as described by Snounou on blood spots obtained at primary (day 0) and secondary infection (time point of re-occurrence).\nBlood spots were collected on Whatman 903 filter paper (Whatman international Ltd. Maidston, United Kingdom) and air-dried at room temperature for PCR analysis.\nDNA was isolated as described by Boom et al .\nMolecular analysis was performed at Royal Tropical Institute, Amsterdam and was done blinded from the treatment that was given to the patients.\nSample size and statistical analysis\nThe aim of the study was to compare gametocytaemia after AL and DP and to compare assessment of gametocytaemia by microscopical examination versus QT-NASBA All data were entered in excel and analysed with SPSS for windows (version 12.0).\nParasite densities were analysed after natural log-transformation.\nWhere appropriate, proportions were compared with the \u03c72-test and means were compared with the one-way ANOVA or Student t-test.\nA simplified trapezoid area under the curve (AUC) analysis using gametocyte data from days 0, 3, 7, 14 and 28, as a surrogate for the infectiousness of the participants in the different treatment groups, was performed.\nResults\nPatient recruitment\nIn total 1882 cases suspected of uncomplicated malaria were screened for eligibility into the study during an 8-week recruitment period in April and May 2007.\n1,736 children were excluded because they did not meet the inclusion criteria (Figure 1).\n146 patients fulfilling the inclusion criteria entered the study; 73 were randomly allocated to the DP arm and 73 to the AL arm.\nBoth study groups were comparable at baseline for their demographical and clinical characteristics and parasite densities (Table 1).\nOn completion of follow up (day 28) data of 134 patients (92%) were available for analysis.\nTwelve patients did not reach the study endpoint.\nSeven patients were lost during follow up, one was unable to take oral medication, one developed severe anaemia, one did not receive the proper drugs, one withdrew from the study and one patient died.\nTreatment outcome\nThere were no early treatment failures during the first three days of follow up.\nOnly one patient in the AL arm had a recurrent parasitaemia (43,880 parasites/\u03bcl) at day 28 of follow up.\nGenotyping analysis revealed that this patient had a reinfection with P. falciparum.\nAll other 133 patients who completed follow-up had an adequate clinical and parasitological response.\nAfter one day of treatment, over 90% of the patients had no microscopically detectable asexual parasites.\nIn the AL group no parasites could be detected with microscopy in any of the patients at day two.\nOne patient was still microscopically positive at day two in the DP group with 40 parasites/\u03bcl, but this patient was also microscopically negative at day 3.\nThe parasite reduction ratios at 48 hours reproduction cycle (parasite count on admission/parasite count at 48 hours) was 8.96 * 105 at 48 hours for the AL treatment and 2.06 * 104 at 48 hours for the DP treatment.\nNASBA was also applied to monitor parasite dynamics below sub-microscopical level.\nHumidity in some of the filter papers degraded the RNA in the blood spots of some of the samples.\nThis led to several extraction failures.\nIn order to have a clear picture of parasite dynamics only those series with a full range of follow-up samples, i.e. 56 DP and 54 AL treated patients, were analysed.\nBoth treatment arms showed a steep decline in parasitaemia from the day of enrollment (day 0) to day 1; 62% reduction after DP treatment and 89% reduction in the AL arm.\nAt day 2, the level of parasitaemia was reduced to 1.2% in the AL group and 2.75% in the DP group.\nHb convalescence, fever clearance and adverse events\nAt baseline Hb levels in both treatment groups were comparable (Table 1).\nAt day 28 all groups had a significant increase of Hb however no significant difference between the treatments on the Hb convalescence was found.\nFinal mean Hb levels were 7.15 mmol/l \u00b1 1.07 for the DP treatment group and 6.79 mmol/l \u00b1 1.24 for the AL group.\nA possible influence of anaemia on gametocyte carriage at enrollment was not observed in the present study (p > 0.05).\nFever clearance was defined as the time from receiving the assigned treatment to the time a normal body temperature was recorded (\u2264 37.5\u00b0C), in study cases who presented with fever.\nFever clearance was rapid in both study groups.\nOn day 1, 10 cases (13.7%) in the DP group presented with fever and six cases (8.2%) were observed in the AL group.\nIn the DP group fever was observed on day 2 in four cases (5.5%) and three cases (4.1%) on day 3.\nIn the AL group cases with fever also presented on day 2 (12 study subjects, 16.4%) and on day 3 (two individuals, 2.7%).\nFever was not observed during follow-up after day 7, with the exception of the child that presented with a P. falciparum reinfection on day 28 in the AL group and the child that developed broncho-pneumonia (case presented below).\nFurthermore the presence of fever at recruitment was no predictor for gametocyte carriage (p > 0.05)\nMost adverse events were mild, self limiting and consistent with symptoms of malaria.\nThere was no significant difference between the two study groups (Table 2).\nOne patient died.\nThe child (63 months) had been ill for two weeks prior to presentation at the clinic.\nPlasmodium falciparum infection with parasitaemia of 20,120 parasites/\u03bcl was diagnosed.\nThere was a fever (38.6\u00b0C), but there were no other complaints and no signs of severe anaemia (Hb: 6.4 mmol/L).\nOn day 3, there were no signs of illness.\nOn day 7, the child presented with fever (38.6\u00b0C), cough and complaints of anorexia.\nThere was no history of significant illness or allergies.\nAfter examination (microscopy was negative for malaria parasites), broncho-pneumonia was diagnosed and the child was treated with oral phenoxymethylpenicillin for five days and Paracetamol syrup.\nOn day 14, the child did not attend the follow-up visit, the parents reported that the child died a day before in a local health post.\nThe event was assessed as unrelated to the study drug.\nAutopsy was not performed.\nGametocyte dynamics\nThe presence of gametocytes in clinical samples was assessed by microscopy and NASBA and is presented in Table 3.\nAt the start of the study, three patients in the DP arm (4.5%) and six patients in the AL arm (9.0%) carried microscopically detectable gametocytes.\nMicroscopical follow-up of the presence of gametocytes during the whole study period revealed that in total 39 samples (distributed over 13 patients) in the DP arm and 18 samples (distributed over nine patients) in the AL arm carried gametocytes (not significantly different).\nIt was observed that the microscopical detection of gametocytes in blood slides during the study was subjected to fluctuations (for example a case positive on day 0, negative on day 1 and subsequently again positive at day 3), which is probably due to the fact that gametocytes circulate at low levels.\nHowever, on day 7, three patients in the DP group and 1 in the AL group showed gametocyte positive slides for the first time.\nOn day 28, in none of the cases gametocytes were observed by microscopy.\nThere was no difference between children older than 60 months and younger as regards carriage of gametocytes and density.\nNASBA analysis on 56 DP treated subjects and 54 AL treated subjects detected strikingly more gametocyte carriers at the start of the study compared to microscopy; i.e. 22 study subjects in the DP arm (39.3%) and 21 in the AL (38.9%) were harbouring gametocytes before.\nThe pfs25 NASBA revealed that 34 cases (60.7%) were gametocyte positive in the DP group on day 3, of which 13 cases were newly identified compared to day 0.\nIn contrast, a significantly lower number of patients (20; 37%) were gametocyte positive in the AL treatment group on day 3, of which six new cases.\nThis trend was also observed on day 7: the DP arm had 33 (58.9%) gametocyte positive samples, whereas the AL treated group had a significantly lower number of gametocyte positive samples (11, 20.3%).\nHowever on day 14 (AL: 12 positive [22.2%], six new; DP: 17 [30.4%], five new) and day 28 (AL: 5 positive [9.3%], no new cases; DP: 8 [14.3%], no new cases) of follow-up no significant difference in gametocyte carriage was observed between both treatment groups.\nThe AUC (day 0\u201328) of the two treatment groups was calculated to be 20.0 infectious persons/day for the DP treatment arm and 10.5 infectious persons/day for the AL treatment arm.\nStratifying the data for age under and above 60 months showed no difference in either of the groups.\nDiscussion\nSeveral studies have analysed the efficacy and tolerability of AL and DP and all show very good results.\nIn the present study, the two drugs showed to be similar with respect to effectiveness and tolerability compared to other studies.\nHowever, most of these studies have a follow up of 42 days which makes a direct comparison of the results difficult.\nNo adverse events other than those related to malaria itself were observed in the current study, which is in line with other reports.\nIn the present study, all children experienced haemoglobin convalescence without difference between the two treatment arms, in contrast to the study of Kamya et al, who found a greater increase in the DP treated patients.\nThe difference in follow-up time and the numbers of patients included in both studies may be responsible for this difference.\nFurther studies with comparable study length should be done to give an answer to these discrepancies.\nThe effects on gametocytaemia and possibly malaria transmission deserve further study.\nWhereas asexual parasites were cleared in three days after the initiation of the two treatment schedules, gametocytaemia appeared different when assessed by microscopy as well as with NASBA.\nGametocytes were present in low numbers throughout follow-up in both study groups.\nArtemisinin derivatives have in general a negative effect on gametocyte development and survival and thus influence malaria transmission, at least in low transmission areas.\nIn this study, the actual infectiousness of the remaining gametocyte populations in both treatment arms was not assessed; the presence of gametocytes does not necessarily mean that they actually contribute to transmission.\nSeveral studies have shown that gametocytes persist in a large population of previously infected and treated children.\nA large proportion of these carriers has a parasite load below microscopical detection limit, a load that can be detected with molecular assays like NASBA.\nPatients with submicroscopic parasite densities may still be infectious to mosquitoes and may contribute to transmission, as confirmed with membrane feeding experiments.\nStudies that include reduction of transmission as a component of efficacy of drugs, thus need to incorporate highly sensitive molecular assays to reliably assess gametocyte densities.\nThe present study showed a limited effect of DP on gametocyte development in comparison with AL when a sensitive tool like NASBA is used for gametocyte detection, which could limit the usefulness of DP to areas with low transmission but this finding should be further investigated in larger studies in different study sites with different transmission intensities.\nIt is not clear if plasma concentrations of dihydroartemisinin in the blood could play a role.\nDihyrodartemisinin is the major and the active metabolite of artemether.\nSo far, no studies have been performed that compare the plasma levels of dihydroartemisinin when given as such or after administration of artemether and subsequent metabolisation.\nThis should be further investigated together with effect on gametocytogenesis, which should incorporate a sensitive detection tool for gametocytes such as NASBA.\nThe effect that drugs can have on gametocyte clearance as measured with NASBA could have some implications for the introduction drugs and especially the introduction of new drugs.\nThis study showed that with sensitive detection tools a difference in parasite clearance can be observed but these results should be confirmed in larger studies and in other study areas with different malaria transmission intensities.\nTransmission intensity varies significantly in the different African countries and within a country high and low transmission areas can often be identified.\nMalaria endemic countries generally have a national malaria drug policy for the whole country.\nAlthough this is logical from a practical and logistical point of view, it may not be the best approach for effective malaria control.\nIt could, therefore, be more effective if a country develops specific drug policies to suit regional instead of national requirements.\nSchematic representation and flowchart of the study.\n\nBaseline characteristics of patients included in the study at the time of enrollment in the study\nCharacteristic | DHA-PQP (n = 73) | ALN (n = 73)\nSex ratio male:female | 33:40 | 40:33\nAge (in months), median (IQR) | 60 (44) | 52 (44)\nBody weight, mean kg (range) | 17.62 (6\u201337) | 17.32 (6\u201342)\nTemperature, mean \u00b0C \u00b1 SD | 38.1 \u00b1 0.99 | 37.8 \u00b1 0.73\nHaemoglobin mmol/L \u00b1 SD | 6.33 \u00b1 1.29 | 6.28 \u00b1 1.27\nParasites/\u03bcl geometric mean (range) as determined by microscopy | 12145 (1000\u201372640) | 13379 (1080\u201372000)\n\n\nSummary of adverse events recorded during the study\nAdverse event | DHA-PQP | ALN | P-value\nHeadache | 43 (58.9%) | 37 (50.7%) | 0.318\nAbdominal pain | 25 (32.4%) | 26 (35.6%) | 0.862\nWeakness | 19 (26.0%) | 30 (41.1%) | 0.035\nAnorexia | 8 (10.9%) | 10 (13.7%) | 0.439\nDiarrhea | 9 (12.3%) | 7 (9.6%) | 0.785\nCough | 16 (21.9%) | 17 (23.3%) | 0.843\nVomiting | 11 (15.1%) | 9 (12.3%) | 0.806\nPruritis | 4 (5.5%) | 3 (4.1%) | 0.698\n\n\nOccurrence of gametocytes as detected by microscopy or NASBA in the different study groups at the start of the study and during subsequent follow-up.\nGametocyte positive samples | Day 0 | Day 1 | Day 2 | Day 3 | Day 7 | Day 14 | Day 28\nDP group microscopy (n = 67) |  |  |  |  |  |  | \nTotal number of positive cases | 3 | 7 | 7 | 10 | 7 | 5 | 0\nNumber of new cases observed |  | 5 | 2 | 0 | 3 | 0 | 0\n\nDP group Nasba (n = 56) |  |  |  |  |  |  | \nTotal number of positive cases | 22 |  |  | 34a | 33a | 17 | 8\nNumber of new cases observed |  |  |  | 13 | 8 | 5 | 0\n\nAL group microscopy (n = 67) |  |  |  |  |  |  | \nTotal number of positive cases | 6 | 3 | 3 | 3 | 2 | 1 | 0\nNumber of new cases observed |  | 1 | 1 | 0 | 1 | 0 | 0\n\nAL group Nasba (n = 54) |  |  |  |  |  |  | \nTotal number of positive cases | 21 |  |  | 20 | 11 | 12 | 5\nNumber of new cases observed |  |  |  | 6 | 4 | 6 | 0\n\na Number of gametocyte carriers detected with NASBA is significantly higher in the DP treated group compared to the AL treated group.",
        "label": "unclear",
        "id": "task4_RLD_test_555"
    },
    {
        "paper_doi": "10.1371/journal.pone.0057899",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Cluster-RCTMethod to adjust for clustering: primary outcome of BMI was not adjusted for clusteringCluster unit: householdAverage cluster size: 4ICCs: not reportedLength of follow-up: 21 months\n\n\nParticipants: All children living in endemic areaNumber analysed for primary outcome: 906 households containing 3230 participantsAge range/mean age: Children aged 19 years and lessInclusion criteria: all household in members except those < 2 years old or pregnantExclusion criteria: none stated\n\n\nInterventions: Multiple dose vs placeboAlbendazole: 400 mg for 3 consecutive days every 3 monthsMatching placebo: every 3 months\n\n\nOutcomes: WeightaHeightaBMIbAdverse eventsaWeight and height in children aged 16 and less.bBMI measured in children aged 19 years and less.Not included in review: Malaria-like symptoms questionnaire, finger prick blood test for malaria, skin prick tests, symptoms of asthma and atopic dermatitis, stool sample for Tichuris and hookworms.\n\n\nNotes: Location: Ende district of Flores Island, IndonesiaBurden: intermediateSource of funding: The Royal Netherlands Academy of Arts and Science (KNAW), European, Prof. Dr. P.C. Flu Foundation.Weight and height data were provided by the authors of the recent Campbell review (Welch 2016)\n\n",
        "objective": "To summarize the effects of public health programmes to regularly treat all children with deworming drugs on child growth, haemoglobin, cognition, school attendance, school performance, physical fitness, and mortality.",
        "full_paper": "Background\nHelminth infections are proposed to have immunomodulatory activities affecting health outcomes either detrimentally or beneficially.\nWe evaluated the effects of albendazole treatment, every three months for 21 months, on STH, malarial parasitemia and allergy.\nMethods and Findings\nA household-based cluster-randomized, double-blind, placebo-controlled trial was conducted in an area in Indonesia endemic for STH.\nUsing computer-aided block randomization, 481 households (2022 subjects) and 473 households (1982 subjects) were assigned to receive placebo and albendazole, respectively, every three months.\nThe treatment code was concealed from trial investigators and participants.\nMalarial parasitemia and malaria-like symptoms were assessed in participants older than four years of age while skin prick test (SPT) to allergens as well as reported symptoms of allergy in children aged 5\u201315 years.\nThe general impact of treatment on STH prevalence and body mass index (BMI) was evaluated.\nPrimary outcomes were prevalence of malarial parasitemia and SPT to any allergen.\nAnalysis was by intention to treat.\nAt 9 and 21 months post-treatment 80.8% and 80.1% of the study subjects were retained, respectively.\nThe intensive treatment regiment resulted in a reduction in the prevalence of STH by 48% in albendazole and 9% in placebo group.\nAlbendazole treatment led to a transient increase in malarial parasitemia at 6 months post treatment (OR 4.16(1.35\u201312.80)) and no statistically significant increase in SPT reactivity (OR 1.18(0.74\u20131.86) at 9 months or 1.37 (0.93\u20132.01) 21 months).\nNo effect of anthelminthic treatment was found on BMI, reported malaria-like- and allergy symptoms.\nNo adverse effects were reported.\nConclusions\nThe study indicates that intensive community treatment of 3 monthly albendazole administration for 21 months over two years leads to a reduction in STH.\nThis degree of reduction appears safe without any increased risk of malaria or allergies.\nTrial Registration\nControlled-Trials.com ISRCTN83830814\nIntroduction\nSoil transmitted helminths (STH) (hookworms, Ascaris lumbricoides and Trichuris trichiura) establish chronic infections in a large proportion of the world population.\nMajor intervention programs using mass drug administration (MDA) to control STH have been launched.\nHowever, STH infections seem to persist in the targeted populations raising concern over the development of drug resistance.\nIt is therefore important to conduct well-designed studies that allow evidence-based decisions to be made to maximize effective STH control toward elimination.\nWhile there is no doubt that STH are associated with morbidities in billions of people worldwide, there is also increasing awareness that helminth infections might, like bacterial commensals, play an important role in shaping human health.\nHelminths may contribute to immunologic and physiologic homeostasis.\nThe immune system is thought to have evolved to operate optimally in the face of helminth-induced immune regulation, and that any disturbance of this long evolutionary co-existence between humans and helminth parasites might be associated with the emergence of pathological conditions possibly involving outcomes of exposure to other pathogens or the development of inflammatory diseases.\nIn many parts of the world helminths and malarial parasites are co-endemic raising the question of what impact helminth infections may have on the plasmodial parasites that cause malaria.\nThe results have been conflicting in this regard.\nIn some studies a positive association has been reported between helminths and malarial parasitemia while in others, this has been refuted or in yet others a negative association has been shown between helminths and the severity of the clinical outcomes of malaria (reviewed by Nacher).\nAn increase in the prevalence of allergies has been reported worldwide, in particular in the urban areas of low- to middle-income countries.\nAlthough majority of cross-sectional studies have reported inverse associations between helminth infections and allergies, two randomized trials with albendazole, have shown conflicting results.\nOne in Ecuador, based on school randomization, reported no change in either SPT reactivity to allergens or allergic symptoms after one year of albendazole treatment while another in Vietnam, in which the randomization unit was individual schoolchildren, showed increased SPT reactivity after one year of albendazole treatment, but consistent with the Ecuadorean study, clinical allergy did not change significantly.\nIt has been suggested that anthelminthic treatment of longer duration might be needed to reveal the modulatory effect of helminths.\nIn the light of global deworming initiatives, it is important to assess the effectiveness of and to monitor the risks associated with anthelminthic treatment regiments.\nThere is as yet no report of a household-based cluster-randomized double-blind placebo-controlled trial of repeated anthelminthic administration in a community that would be expected to more effectively reduce transmission of STH by decreasing household cross-contamination.\nIn an area where STH and malaria are co-endemic on Flores Island, Indonesia, we conducted a household cluster-randomized trial of three-monthly albendazole treatment over a two year study period in a whole community to assess benefits and risks associated with this anthelmintic treatment.\nSpecifically we assessed its impact on STH, malarial parasitemia and allergy.\nMethods\nStudy population and design\nThis trial was conducted in two villages in the Ende District of Flores Island, Indonesia (Appendix S1, p2) as described in detail elsewhere.\nThe treatment was based on household and given to all household members except those less than two years old or pregnant (the Indonesian national program guideline).\nDirectly observed treatment was given three monthly during the trial period (June 2008 to July 2010, with treatment starting in Sept 2008).\nThe primary outcomes were prevalence of malarial parasitemia and SPT reactivity to allergens.\nAdditional outcomes were treatment effect on STH and BMI as well as malaria-like and allergy symptoms.\nWe measured malaria outcomes in Nangapanda only.\nMalaria was not endemic in Anaranda.\nArtemisinin-combination therapy (ACT) treatment and treated bed net distribution were not implemented during our study period.\nAllergy outcomes were measured, in both villages, in school-age children (5\u201315 years old) as this group is particularly at risk of developing allergy and asthma and is the target population of global deworming programs.\nThe study was approved by the Ethical Committee of the Medical Faculty, University of Indonesia (ethical clearance ref: 194/PT02.FK/Etik/2006) and filed by the Committee of Medical Ethics of the Leiden University Medical Center.\nThe trial was registered as clinical trial (Ref: ISRCTN83830814).\nPrior to the study, written informed consent was obtained from participants or from parents/guardians of children.\nThe study is reported in accordance with the CONSORT guidelines for cluster-randomized studies.\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nRandomization and masking\nThe population was randomized by IA using computer aided block randomization at household level utilising Random Allocation software to receive albendazole (single dose of 400 mg) or a matching placebo (both tablets from PT Indofarma Pharmaceutical, Bandung, Indonesia).\nThe treatment code was concealed from trial investigators and participants.\nThe un-blinding of treatment codes occurred after all laboratory results had been entered into the database (August 2011).\nProcedures\nTrained community workers measured fever, administered monthly malaria-like symptoms questionnaire which was based on WHO definitions and took finger-prick blood for the three-monthly malarial parasitemia survey.\nSubjects with fever (\u226537.5\u00b0C) or additional malaria-like symptoms (headache, fatigue and nausea) at the time of visits were referred to the local primary health centre (puskesmas).\nThick and thin Giemsa-stained blood smears were read at University of Indonesia.\nAt baseline, 9 months and 21 months after the first round of treatment blood was collected for PCR-based detection of Plasmodium spp. (Appendix S1, p2), a method that is more sensitive than microscopy.\nRegarding allergy outcomes, skin prick tests (SPT) with allergens were performed on school-age children in Nangapanda and Anaranda and clinical symptoms of allergy were recorded.\nHouse dust mite (Dermatophagoides pteronyssinus and D. farinae; kindly provided by Paul van Rijn from HAL Allergy Laboratories, Leiden, The Netherlands) and cockroach (Blatella germanica; Lofarma, Milan, Italy) were used for SPT which was considered positive with 3 mm cut off.\nThe SPT was performed by one investigator.\nIgE with specificity for aeroallergens (D. pteronyssinus and B. germanica) was measured in plasma using an ImmunoCAP 250 system (Phadia, Uppsala, Sweden) following the manufacturer's instructions.\nAll measurements were conducted in one laboratory in the Netherlands.\nSymptoms of asthma and atopic dermatitis were recorded using a modified visually-assisted version of the International Study of Asthma and Allergy in Childhood (ISAAC) questionnaire as reported before.\nYearly stool samples were collected on a voluntary basis.\nTrichuris was detected by microscopy and a multiplex real-time PCR was used for detection of hookworms (Ancylostoma duodenale, Necator americanus), Ascaris lumbricoides, and Strongyloides stercoralis DNA as detailed before (Appendix S1, p2).\nVery few subjects were infected with S. stercoralis and therefore this infection was not included in analyses.\nBody weight and height were measured using the National Heart Lung and Blood Institute practical guidelines (scale and microtoise from SECA GmBH & Co, Hamburg, Germany).\nPower calculation\nSample size estimation was based on the expected change in primary outcomes taking into account a power of 90% and a significance level of <0.05 with a loss to follow-up of 20%.\nBased on previous observations we expected to find a decrease in malarial parasitemia prevalence and an increase in SPT reactivity after anthelminthic treatment.\nBased on a prevalence of about 10% and a risk ratio (RR) of 0.60 we aimed to include 2412 people in the malaria assessments.\nIn a pilot study we found SPT to D. pteronyssinus to be around 15%, and expected that due to treatment the prevalence would increase.\nIn order to find a RR of 1.5 we aimed to include at least 1418 children.\nStatistical analyses\nFor children \u226419 years, BMI age-standardized z-scores were calculated according to WHO references.\nThe IgE data were log-transformed to obtain normally distributed variable.\nTo assess treatment effects generalized linear mixed models were used which provide a flexible and powerful tool to derive valid inference while capturing the data correlations induced by clustering within households and repeated evaluations in time of the same subject.\nParameter estimates for treatment effects at 9 and 21 months and 95% confidence intervals are reported.\nThe reported p-values are obtained using likelihood ratio tests by comparing the model with and without the treatment effect.\nUnless stated otherwise all outcomes were adjusted for area (the two study villages in Ende District: Nangapanda or Anaranda) as covariate in the model.\nFor the continuous outcomes (linear mixed-effects models were used with three random effects, namely to model clustering within households a random household specific intercept was used and to model correlation within subjects a random subject specific intercept and slope were used.\nFor the binary outcomes a logistic model was used with random household effects and random subject effects.\nAll models were fitted using the lme4 package (Appendix S1, p6-7).\nFor each fever and additional malaria-like symptoms, total number of events and person months are computed for each treatment arm.\nHazard ratios for effect of treatment were calculated with Cox regression with robust SE to allow for within-household clustering (STATA 11).\nResults\nAt baseline, 954 households with 4004 subjects were registered.\nRandomization of households resulted in 1982 people assigned to albendazole treatment and 2022 people to placebo (473 and 481 houses respectively).\nAt baseline 87\u00b73% of the individuals were infected with one or more helminth species.\nThe baseline characteristics were similar between the treatment arms (table 1).\nThe overall trial profile is shown in figure 1, and figure S1A, S1B, S1C (p13\u201315) in Appendix S1 separately for malaria, allergy and helminth outcomes.\nThe analysis was intention-to-treat and involved all participants as assigned randomly at the start of the trial.\nDuring the study, in the albendazole arm 61 people moved to a house that was assigned to placebo while in the placebo arm 62 people moved to a house that was assigned to albendazole.\nThe 44 subjects who died during the trial, included 4 people below the age of 20, 3 between 20 and 40 and the rest above 40 years of age, and were equally distributed between the treatment arms.\nAt 9 months post-treatment full compliance was 77.8% for albendazole treatment and 78.0% for placebo.\nThis was 63.1% and 62.5% respectively at 21 months.\nThis intensive treatment with albendazole resulted in a reduction but not elimination of STH.\nThere was a decrease both after 9 (OR (95% CI)\u200a=\u200a0.07 (0.04\u20130.11) and 21 months (0.05 (0.03\u20130.08)) of treatment (p<0.0001).\nAlbendazole had the largest effect on hookworm followed by Ascaris while the effect on Trichuris was less pronounced (figure 2A and table S1 in Appendix S1 p8).\nTreatment also led to statistically significant reduction in the intensity of hookworm and Ascaris infection as determined by PCR (figure 2B).\nThe fact that the stool sampling was on a voluntary basis could have created a selection bias.\nAnalyzing baseline characteristics of subjects providing stool samples and those who did not at 9 months follow up, showed no differences in helminth prevalence, age and sex.\nAlthough at 21 months post treatment, sex and helminth prevalence were not different, age was slightly but significantly higher in subjects who provided stool samples mean age in years (SD)\u200a=\u200a29.9 (20.4) vs 24.3 (17.5), p\u200a=\u200a0.006)\nThe overall percentage of subjects with malarial parasitemia, irrespective of treatment arm, decreased over the trial period (table 2).\nHowever, when the data were modelled to assess the effect of albendazole treatment over time, there was a significant (P\u200a=\u200a0.0064) increase, which might result from the transient four-fold increased risk of malarial parasitemia (OR 4.16 (1.35\u201312.80)) (table 3) at 6 months after initiation of treatment (after 2 doses of albendazole).\nThe effect of anthelminthic treatment was assessed in those younger than 15 years of age who would be the prime target of the global deworming programs.\nThe transient increase in parasitemia was only seen in the older (>15 years) age group (figure 3).\nMalarial parasites were also assessed by PCR, at 9 and 21 months after initiation of treatment and revealed that albendazole had no effect when all Plasmodium species were considered together, but when analyzed separately there was a significant increase in the percentage of subjects positive for P. falciparum at 9 months post-treatment (table 4).\nThere was no difference in the incidence of fever and additional malaria-like symptoms between the two treatment arms (table S2 in Appendix S1 p10).\nThe proportion of subjects with SPT reactivity was 353/1364 (25.9%) at baseline.\nAlbendazole treatment had no statistically significant effect on SPT to any allergen (table 5), but it was noted that there was an incremental increase in the risk of being SPT positive to any allergen at 9 months and 21 months post initiation of treatment.\nMoreover, additional analysis on allergens separately, showed a significantly higher SPT to cockroach at 21 months after treatment (1.63 (1.07\u20132.50)) (table 6).\nThe levels of IgE to allergens showed that albendazole treatment had no effect (table 6).\nNo effect of treatment was seen on symptoms of asthma or atopic dermatitis (table S3 in Appendix S1 p11).\nNo significant change in BMI was observed in children or in adults (table S4 in Appendix S1 p12).\nMoreover, there was no adverse effect of treatment reported.\nDiscussion\nThis household-based clustered-randomized, double-blind, placebo-controlled trial shows that administering a total of seven single doses of albendazole, at three-monthly intervals, to a population living in an area of Indonesia where STH are highly prevalent, leads to decreased prevalence of helminth infections which although statistically significant, can be taken as an incomplete reduction.\nThe results show a transient increase in malarial parasitemia in the albendazole- compared with the placebo-treated arm in the first six months after initiation of treatment.\nAlbendazole treatment had no statistically significant effect on the designated co-primary outcome, skin prick test reactivity to allergens.\nThe clinical data collected of fever and additional malaria-like symptoms, were not affected by the deworming.\nClinical signs of asthma and atopic dermatitis were also not affected by albendazole treatment.\nThe prevalence of infection was high (>60%), which reflects the situation in many areas that are being targeted by the global deworming programs.\nUsing a three-monthly treatment regimen which represents an extreme scenario for helminth control strategy, percentage of subjects positive for STH was reduced by 39% compared to placebo.\nIt should be noted that in our study the sensitive PCR method has been used.\nThe reduction in the proportion of subjects infected with hookworm and Ascaris was more pronounced than for Trichuris infections, confirming the findings using a single dose of albendazole.\nSubjects who provided stool samples at 21 months were slightly but significantly older than those who did not.\nGiven that hookworm infection is more prevalent in older subjects, this may have contributed to the poor deworming achieved by albendazole.\nThe reduction achieved in worm loads, did not have any beneficial effect on BMI.\nObservational studies have reported that helminth infections affect growth; however randomized trials have not been consistent.\nIn this regard, our study would support the outcome of a recent Cochrane review of no beneficial effect of deworming programs on nutritional indicators even though it can be argued that in our study the suboptimal reduction in the STH would not allow any beneficial effect of anthelmintic in terms of BMI to be seen in the community.\nImportantly, the fact that the effect of such an intensive deworming strategy in a community is incomplete, needs to be considered in the agenda for the control and elimination of helminth diseases of humans.\nMost studies on the effect of helminth infections on malarial parasitemia and clinical malaria episodes have used cross-sectional designs and have been inconclusive.\nLongitudinal studies of anthelminthic treatment have also reported conflicting results.\nA small study conducted in Madagascar has reported an increase in malarial parasitemia in levimasole treated subjects, older than 5 years of age, while in Nigeria, albendazole treatment of pre school children was associated with lower P. falciparum infection and anemia, however, the lost to follow up in this study was very high.\nThe question whether albendazole treatment during pregnancy could affect health outcomes in the offspring, was addressed in a recent report from Uganda.\nIt was found that the incidence of malaria up to one years of age was not different in the offspring of mothers born to those treated with albendazole or placebo.\nOur study, reports the results of a community wide randomized-controlled trial that used three-monthly malarial parasitemia data obtained by microscopy.\nA significantly higher percentage of subjects positive for malarial parasites in the albendazole compared to the placebo arm was seen but this seemed to be a transient effect and restricted to individuals older than 15 years of age, an age group that is not the main target of the current deworming programs.\nThe question arises as to why this effect was only seen in those >15 years of age.\nThis could be due to the fact that Ascaris infection is lower in older age and therefore more easily cleared.\nIt has been suggested that Ascaris is the spesies of helminth that has the most effect on malarial parasitemia and diseases.\nTherefore by removing Ascaris in older age, we might be seeing a more profound effect on malarial parasitemia.\nUsing PCR, which enables detection of sub-microscopic infections at species level, it was also concluded that albendazole did not affect overall malarial parasitemia.\nWhen malaria species were analyzed separately, the percentage of subjects infected with P. falciparum but not with P. vivax increased significantly in the first 9 months post-treatment in the albendazole-treated arm, which is contrary to our hypothesis that anthelminthic treatment would reduce prevalence of malarial parasitemia.\nIt was expected that by decreasing STH, the immune hyporesponsiveness would be reversed and this would be associated with stronger immune effector responses to malaria parasites.\nOne of the possible explanations for the enhanced malarial parasitemia would be that with a reduction in STH, there is increased nutrient availability for other co infections and their growth.\nIt has been suggested that there are different malaria outcomes with different species of helminths; Ascaris being associated with protection regarding parasitemia and severity of malaria while hookworm with higher incidence of malaria.\nOur study was not powered to conduct a stratified analysis, and with the overall gradual decrease in malaria in the study area during our study, the numbers of subjects positive for malaria parasites are too few for an ad hoc analysis.\nThe findings concerning allergy outcomes, although not significant, are in line with our hypothesis that anthelminthic treatment would increase SPT reactivity.\nThe risk of SPT reactivity increased incrementally with longer treatment and raises the question whether even longer deworming periods are needed for more pronounced effects on allergic outcomes.\nIn support of this, a recent study reported that 15\u201317 years of ivermectin treatment for onchocerciasis control in Ecuador led to a significant increase in SPT reactivity to allergens.\nIn the same country, one year of anthelmintic treatment in schoolchildren did not lead to any change in SPT.\nThe question whether different species of helminths might affect allergic outcomes to a different degree, remains unanswered.\nIt is interesting that a one year anthelmintic treatment in Vietnam where hookworm infection was the prominent species, as in our study, resulted in a significant increase in SPT positivity in schoolchildren.\nThis is in contrast to what was seen in Ecuador where Ascaris lumbricoides was the most prevalent species.\nOne common feature of the anthelmintic trials seems to be that clinical symptoms of allergy do not change with deworming.\nHowever, an important trial in pregnant women in Uganda has shown an increased risk of infantile eczema in infants of mothers treated with anthelminthics compared to those that received placebo.\nThis could indicate that exposure to worms in early life, might affect allergic outcomes more profoundly than when helminths are removed later in life.\nOne of the limitations of this trial is the overall decrease in malarial parasitemia during the two year study period, most probably caused by actively referring subjects with malaria-like symptoms to puskesmas.\nTherefore further studies in areas highly endemic for malaria are needed.\nTreatment in the trial did result in a significant reduction in percentage of subjects infected with STH, but this reduction was incomplete.\nIt is therefore possible that the community was insufficiently dewormed.\nHowever, our primary aim was to measure the possible effect of deworming programmes on malaria or allergy.\nWe conclude that despite transient increase in malarial parasitemia as a result of albendazole treatment, there were no clinically relevant changes to outcome measures 21 months after treatment was initiated.\nIn conclusion, an extremely intensive anthelminthic treatment in a community where STH are highly endemic, does not lead to elimination but reduces both prevalence and intensity of helminths.\nSuch MDA regiment appears safe and does not lead to any significant change with respect to malaria infections or allergies.\nHowever, it is worrying that such vigorous community treatment does not have a more pronounced effect on STH burden.\nBetter integrated control strategies would be needed to deworm and subsequently assess whether the risk for malaria infections or allergies change.\nTrial Profile.HH: Household. Lost to follow up implies that the participants have no data from this time point onward. Temporarily absent implies that the participants have no data at this time point but have data available at other time point.\nA) Percentage of helminth infected subjects in placebo and albendazole treatment arms.The presence of hookworms (by PCR), Ascaris lumbricoides (by PCR) and Trichuris trichiura (by microscopy) or any of these helminth infections in subjects who provided stool samples at baseline, 9 and 21 months post treatment (numbers are given in table S1A in Appendix S1). B) Effect of albendazole treatment on reduction in the intensity as well as percentage of subjects positive for hookworm and \nAscaris\n infection as determined by PCR. Negative is when no helminth specific DNA was found. Positive Ct- values were grouped into three categories: Ct<30.0, 30.0\u2264Ct<35.0 and \u226535.0 representing a high, moderate and low DNA load, respectively.\nEffect of albendazole treatment on malarial parasitemia based on two age categories.Malarial parasitemia A) \u226415 and B) >15 years of age. The risk of malarial parasitemia after albendazole treatment compared to placebo is shown as odds ratio with 95% CI. The reference line is set at 1, indicating that symbols at the right of this line represent an increased risk, while symbols at the left of the line would predict decreased risk of malarial parasitemia. Note: at 9 month time point in those >15 years of age, the OR is \u221e.\n\nBaseline characteristics.\n |  | N | Placebo | N | Albendazole\nAge (mean in years, SD) | 2022 | 25.7 (18.7) | 1982 | 25.8 (18.7)\nSex (female, n, %) | 2022 | 1090 (53.9) | 1982 | 1042 (52.6)\nArea (rural, n, %) | 2022 | 260 (12.9) | 1982 | 253 (12.8)\nBMI >19 years old (mean, SD) | 575 | 22.3 (4.0) | 582 | 21.8 (3.6)\nZ score of BMI \u2264 19 years old (mean, SD) | 427 | \u22121.20 (1.2) | 386 | \u22121.37 (1.3)\nParasite infection (n, %) |  |  |  | \nHelminth (any spp) | 655 | 571 (87.2) | 609 | 533 (87.5)\nHookworm1 | 683 | 509 (74.5) | 629 | 486 (77.3)\nN. americanus1 | 683 | 503 (73.7) | 629 | 481 (76.5)\nA. duodenale1 | 683 | 44 (6.4) | 629 | 41 (6.5)\nA. lumbricoides1 | 683 | 238 (34.9) | 629 | 209 (33.2)\nS. stercoralis1 | 683 | 7 (1.0) | 629 | 18 (2.9)\nT. trichiura2 | 953 | 258 (27.1) | 852 | 237 (27.8)\nMalarial parasitemia (any spp)2 | 1225 | 60 (4.9) | 1187 | 52 (4.4)\nP. falciparum | 1225 | 32 (2.6) | 1187 | 28 (2.4)\nP. vivax | 1225 | 26 (2.1) | 1187 | 18 (1.5)\nP. malariae | 1225 | 2 (0.2) | 1187 | 7 (0.6)\nMalarial parasitemia (any spp)1 | 772 | 195 (25.3) | 739 | 200 (27.1)\nP. falciparum | 772 | 106 (13.7) | 739 | 112 (15.2)\nP. vivax | 772 | 102 (13.2) | 739 | 93 (12.6)\nP. malariae | 772 | 10 (1.3) | 739 | 18 (2.4)\nSkin prick reactivity (n, %) |  |  |  | \nAny allergen | 711 | 190 (26.7) | 653 | 163 (25.0)\nHouse dust mite | 711 | 88 (12.4) | 653 | 75 (11.5)\nCockroach | 711 | 163 (22.9) | 653 | 140 (21.4)\nSpecific IgE, kU/L (median, IQR) |  |  |  | \nHouse dust mite | 452 | 0.8 (0.3\u20132.6) | 431 | 0.8 (0.2\u20132.4)\nCockroach | 452 | 1.5 (0.4\u20135.7) | 431 | 1.9 (0.5\u20135.0)\n\ndiagnosed by PCR; 2diagnosed by microscopy.\nThe number of positives (n) of the total population examined (N).\n\nEffect of three-monthly albendazole treatment on malaria outcomes: Percentage of subjects with malarial parasitemia.\n | P. falciparum | P. vivax | P. malariae\n | Placebo | Albendazole | Placebo | Albendazole | Placebo | Albendazole\n | n/N (%) | n/N (%) | n/N (%) | n/N (%) | n/N (%) | n/N (%)\nMalarial parasitemia by microscopy\n0 month | 32/1225 (2.6) | 28/1187 (2.4) | 26/1225 (2.1) | 18/1187 (1.5) | 2/1225 (0.2) | 7/1187 (0.6)\n3 months | 41/897 (4.6) | 46/910 (5.1) | 17/897 (1.9) | 22/910 (2.4) | 1/897 (0.1) | 6/910 (0.7)\n6 months | 8/815 (1.0) | 20/794 (2.5) | 4/815 (0.5) | 9/794 (1.1) | 0 | 0\n9 months | 14/947 (1.5) | 7/950 (0.7) | 4/947 (0.4) | 5/950 (0.5) | 1/947 (0.1) | 1/950 (0.1)\n12 months | 9/834 (1.1) | 9/813 (1.1) | 4/834 (0.5) | 2/813 (0.2) | 0 | 0\n15 months | 14/773 (1.8) | 13/772 (1.7) | 3/773 (0.4) | 4/772 (0.5) | 1/773 (0.1) | 3/772 (0.4)\n18 months | 3/815 (0.4) | 10/803 (1.2) | 1/815 (0.1) | 1/803 (0.1) | 1/815 (0.1) | 1/803 (0.1)\n21 months | 6/824 (0.7) | 11/824 (1.3) | 6/824 (0.7) | 0 | 3/824 (0.4) | 1/824 (0.1)\nMalarial parasitemia by PCR\n0 month | 106/772 (13.7) | 112/739 (15.2) | 102/772 (13.2) | 93/739 (12.6) | 10/772 (1.3) | 18/739 (2.4)\n9 months | 35/656 (5.3) | 56/627 (8.9) | 56/656 (8.5) | 50/627 (8.0) | 7/656 (1.1) | 9/627 (1.4)\n21 months | 21/584 (3.6) | 31/553 (5.6) | 24/584 (4.1) | 27/553 (4.9) | 10/584 (1.7) | 5/553 (0.9)\n\nThe number of positives (n) of the total population examined (N).\n\nEffect of three-monthly albendazole treatment on malaria outcomes: Malarial parasitemia by microscopy\n | Placebo | Albendazole | OR (95%CI) *\n | n/N (%) | n/N (%) | \nMalarial parasitemia (any spp) |  |  | \n3 months | 59/897 (6.6) | 72/910 (7.9) | 1.54 (0.75\u20133.16)\n6 months | 12/815 (1.5) | 29/794 (3.7) | 4.16 (1.35\u201312.80)\n9 months | 19/947 (2.0) | 13/950 (1.4) | 0.57 (0.16\u20132.04)\n12 months | 13/834 (1.6) | 10/813 (1.2) | 0.62 (0.12\u20133.15)\n15 months | 18/773 (2.3) | 20/772 (2.6) | 1.17 (0.18\u20137.65)\n18 months | 5/815 (0.6) | 12/803 (1.5) | 1.84 (0.12\u201329.03)\n21 months | 15/824 (1.8) | 12/824 (1.5) | 0.26 (0.01\u20136.59)\n\nThe number of positives (n) of the total population examined (N). *Odds ratio and 95% confidence interval are based on mixed effects logistic regression models. OR's and 95% CI are shown for the separate time points on malarial parasitemia. The p-value is generated from the modeled data for the combined effect of albendazole treatment over time, which is significant (P\u200a=\u200a0.0064) and might result from the effect of 6 months post treatment time point.\n\nEffect of three-monthly albendazole treatment on malaria outcomes: Malarial parasitemia by PCR.\n | Placebo | Albendazole | OR (95% CI)\n | n/N (%) | n/N (%) | \nMalaria (any spp) |  |  | \n 9 months | 95/656 (14.5) | 103/627 (16.4) | 1.13 (0.77\u20131.64)\n 21 months | 53/584 (9.1) | 59/553 (10.7) | 1.09 (0.68\u20131.76)\nP. falciparum |  |  | \n 9 months | 35/656 (5.3) | 56/627 (8.9) | 2.82 (1.29\u20136.15)\n 21 months | 21/584 (3.6) | 31/553 (5.6) | 1.63 (0.63\u20134.22)\nP. vivax |  |  | \n 9 months | 56/656 (8.5) | 50/627 (8.0) | 0.84 (0.41\u20131.71)\n 21 months | 24/584 (4.1) | 27/553 (4.9) | 1.40 (0.56\u20133.52)\nP. malariae |  |  | \n 9 months | 7/656 (1.1) | 9/627 (1.4) | 0.34 (0.04\u20132.79)\n 21 months | 10/584 (1.7) | 5/553 (0.9) | 0.04 (0.00\u20130.39)\n\nThe number of positives (n) of the total population examined (N). Odds ratio and 95% confidence interval based on logistic mixed models. The statistically significant results are given in bold. The p-values are generated from the modeled data for the combined effect of albendazole treatment over time for each of the species separately, which were significant for P. falciparum (P\u200a=\u200a0.029) and P. malariae (P\u200a=\u200a0.016).\n\nEffect of three-monthly albendazole treatment on allergy outcomes: Skin prick test to any allergens.\n | Placebo | Albendazole | OR (95%CI) *\n | n/N (%) | n/N (%) | \nSPT to any allergen |  |  | \n9 months | 80/462 (17.3) | 82/454 (18.1) | 1.18 (0.74\u20131.86)\n21 months | 145/455 (31.9) | 161/439 (36.7) | 1.37 (0.93\u20132.01)\n\nThe number of positives (n) of the total population examined (N). *Odds ratio and 95% confidence interval are based on mixed effects logistic regression models. OR's and 95% CI are shown for the separate time points on SPT to any allergen. The p-value is generated from the modeled data for the effect of albendazole treatment overtime and no significant effects were found (P>0.05).\n\nEffect of three-monthly albendazole treatment on allergy outcomes: Skin prick test and specific IgE to aeroallergens.\n | Placebo | Albendazole | \nSkin prick test reactivity* | n/N (%) | n/N (%) | OR (95% CI)\nHouse dust mite |  |  | \n 9 months | 36/462 (7.8) | 35/454 (7.7) | 1.31 (0.52\u20133.27)\n 21 months | 77/455 (16.9) | 76/439 (17.3) | 1.37 (0.62\u20133.02)\nCockroach |  |  | \n 9 months | 60/462 (13.0) | 65/454 (14.3) | 1.27 (0.75\u20132.15)\n 21 months | 112/455 (24.6) | 139/439 (31.7) | 1.63 (1.07\u20132.50)\nSpecific IgE** | N (Median, IQR) | N (Median, IQR) | \u03b2 (95% CI)\nHouse dust mite |  |  | \n 9 months | 391 (0.46, 0.16\u20132.35) | 381 (0.46, 0.14\u20131.98) | 1.01 (0.91\u20131.12)\n 21 months | 339 (0.82, 0.27\u20133.29) | 334 (0.65, 0.20\u20132.69) | 0.93 (0.81\u20131.06)\nCockroach |  |  | \n 9 months | 391 (1.47, 0.30\u20135.01) | 381 (1.55, 0.44\u20134.40) | 1.04 (0.93\u20131.16)\n 21 months | 339 (1.83, 0.47\u20135.44) | 334 (1.64, 0.42\u20134.82) | 0.98 (0.85\u20131.14)\n\nThe number of positives (n) of the total population examined (N). *Odds ratio and 95% confidence interval based on logistic mixed models; **\u03b2 (beta) and 95% confidence interval based on generalized linear mixed models from the log-transformed IgE. The values shown are back-transformed. The p-values are generated from the modeled data for the effect of albendazole treatment overtime and no significant effects were found (P>0.05).",
        "label": "low",
        "id": "task4_RLD_test_740"
    },
    {
        "paper_doi": "10.1371/journal.pone.0001023",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Individually RCTDates of trial: June to Sept 2006.\n\n\nParticipants: 108 children with fever > 37.5degC or history of fever in last 48 hours and P. falciparum mono-infection 500 to 100,000/mL.Age three to 15 years.Both sexes.Site: Mynuzi health centre, North-Eastern Tanzania, a hyperendemic area with rainy seasons in March to June and October to DecemberExclusion criteria: Hb < 8, inability to take drugs orally, known hypersensitivity to meds, reported anti-malarial treatment in last 2 weeks, evidence of chronic disease or acute infection other than malaria, domicile outside trial area, signs of severe malaria, eligible for other malaria studies.\n\n\nInterventions: AS+SP: AS: 4 mg/kg once daily for 3 days; SP: S 25 mg/kg and P: 1.125 mg/kg.AS+SP+PQ: As above for AS and SP plus PQ base 0.75 mg/kg on the third day.\n\n\nOutcomes: Proportion of people with gametocytes (by microscopy) days 1, 4, 8, 15, 29 and 43 (reported as 0, 3, 7, 14, 28, and 42).Proportion with gametocytes (by PCR), same time points.Gametocyte density by PCR.AUC for gametocyte presence.Adverse events.Adequate clinical and parasitological response.Haemoglobin.\n\n\nNotes: Hb outcome assessed with respect to G6PD variant\n\n",
        "objective": "To assess the effects of single dose or short\u2010course PQ (or an alternative 8AQ) alongside treatment for people with P. falciparum malaria.",
        "full_paper": "Background\nP. falciparum gametocytes may persist after treatment with sulphadoxine-pyrimethamine (SP) plus artesunate (AS) and contribute considerably to malaria transmission.\nWe determined the efficacy of SP+AS plus a single dose of primaquine (PQ, 0.75 mg/kg) on clearing gametocytaemia measured by molecular methods.\nMethodology\nThe study was conducted in Mnyuzi, an area of hyperendemic malaria in north-eastern Tanzania.\nChildren aged 3\u201315 years with uncomplicated P. falciparum malaria with an asexual parasite density between 500\u2013100,000 parasites/\u00b5L were randomized to receive treatment with either SP+AS or SP+AS+PQ.\nP. falciparum gametocyte prevalence and density during the 42-day follow-up period were determined by real-time nucleic acid sequence-based amplification (QT-NASBA).\nHaemoglobin levels (Hb) were determined to address concerns about haemolysis in G6PD-deficient individuals.\nResults\n108 individuals were randomized.\nPfs25 QT-NASBA gametocyte prevalence was 88\u201391% at enrolment and decreased afterwards for both treatment arms.\nGametocyte prevalence and density were significantly lower in children treated with SP+AS+PQ.\nOn day 14 after treatment 3.9% (2/51) of the SP+AS+PQ treated children harboured gametocytes compared to 62.7% (32/51) of those treated with SP+AS (p<0.001).\nHb levels were reduced in the week following treatment with SP+AS+PQ and this reduction was related to G6PD deficiency.\nThe Hb levels of all patients recovered to pre-treatment levels or greater within one month after treatment.\nConclusions\nPQ clears submicroscopic gametocytes after treatment with SP+AS and the persisting gametocytes circulated at densities that are unlikely to contribute to malaria transmission.\nFor individuals without severe anaemia, addition of a single dose of PQ to an efficacious antimalarial drug combination is a safe approach to reduce malaria transmission following treatment.\nTrial Registration\nControlled-Trials.com ISRCTN61534963\nIntroduction\nThe majority of anti-malarial drug treatments target the asexual blood stages of Plasmodium falciparum that are responsible for clinical disease and death.\nSexual stage parasites, gametocytes, can also be present in infected individuals and are responsible for the transmission of the parasite to mosquitoes.\nDrugs specifically targeting these sexual stage parasites may affect the spread of malaria in the human population.\nAnti-gametocyticidal drugs are used by several countries to prevent onward transmission from clinical malaria cases and have also been evaluated by mass drug administration to reduce malaria transmission in communities.\nArtemisinin-based combination therapies (ACT) are advocated as first-line antimalarial treatment because of their high treatment efficacy and beneficial effects on malaria transmission.\nAlthough ACT efficiently reduces microscopic levels of gametocytes, submicroscopic gametocytes (detected by molecular analysis) may persist after treatment and allow post-treatment malaria transmission.\nThe implementation of ACT may have a beneficial influence on malaria transmission in the general population but ACT may not be sufficient to completely prevent post treatment malaria transmission.\nPrimaquine (PQ) may be of added value in attempts to block malaria transmission as part of a mass drug administration.\nPQ is an 8-aminoquinolone that is widely used for the treatment of P. vivax malaria and actively clears mature P. falciparum gametocytes.\nAlthough there is no consensus about which drug is the most potent gametocytocidal drug, artesunate (AS) may predominantly inhibit gametocyte development while PQ may accelerate gametocyte clearance.\nIn combination with sulphadoxine-pyrimethamine (SP) and AS, PQ was found to be safe and highly efficacious in clearing asexual parasites and P. falciparum gametocytes detected by microscopy.\nThe efficacy of this combination on submicroscopic gametocytaemia is unknown and, in general, information on PQ use in Africa is scarce.\nPrior to the wide-scale introduction of ACT, a single dose of PQ following first line antimalarial treatment was recommended by the World Health Organisation to reduce malaria transmission in low endemic areas.\nAlthough several countries adopted this recommendation, there is concern for negative haemolytic side effects in individuals who are glucose-6-phosphate-dehydrogenase (G6PD) deficient.\nHere, we determine the safety and efficacy of SP+AS plus a single dose of PQ on clearing submicroscopic levels of P. falciparum gametocytaemia in an area of hyperendemic malaria in north eastern Tanzania.\nPossible haemolytic effects of PQ were determined in relation to G6PD status.\nMethods\nThe protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.\nThe trial was registered at Current Controlled Trials; ISRCTN61534963; http://www.controlled-trials.com/ISRCTN61534963/.\nRegistration was done after patient recruitment started due to communication problems.\nParticipants\nThis study was conducted in the period July through September 2006 in Mnyuzi, a rural village in the Tanga Region, north eastern Tanzania.\nMalaria transmission intensity is high with an estimated entomological inoculation rate (EIR) of 91 infectious bites per person per year.\nThe rainfall pattern is bimodal, with a long rainy season between March and June, and a short rainy season between October and December.\nThe study protocol was approved by the ethics committees of Kilimanjaro Christian Medical Centre, the Tanzanian National Institute for Medical Research (NIMR/HQ/R.8a Vol. XIII/446) and the London School of Hygiene and Tropical Medicine (#4097).\nParticipants were recruited from among children consulting the Mnyuzi health centre and who were resident within a 10 kilometres radius.\nInformed consent was obtained form the child's parents or guardians prior to inclusion.\nChildren aged 3\u201315 years with a temperature >37.5C\u00b0 or a history of fever within the last 48 hours and with P. falciparum mono-infection at a density between 500\u2013100,000 parasites/\u00b5L were eligible for recruitment.\nExclusion criteria were: a haemoglobin (Hb) concentration measured by Hemocue\u00ae below 8g/dL, inability to take drugs orally, known hypersensitivity to any of the drugs given, reported treatment with antimalarial chemotherapy in the past 2 weeks, evidence of chronic disease or acute infection other than malaria, domicile outside the study area, signs of severe malaria and eligibility for other malaria studies conducted in the region.\nInterventions\nParticipants enrolled were randomized to one of the two treatment regimes:\nSulphadoxine (25 mg/kg) and pyrimethamine (1.25 mg/kg) as a single dose (SP; Fansidar\u00ae, Roche, Switzerland) plus artesunate (AS), 4 mg/kg once daily for three days (Arsumax\u00ae, Sanofi-aventis, France) plus placebo once on the third day (Organon, The Netherlands);\nSP plus AS plus primaquine (base; department of clinical pharmacology, Radboud University Nijmegen Medical Centre, the Netherlands) as a single dose on the third day (0.75 mg/kg).\nPrimaquine capsules were produced following regulations of the European Pharmacopeia.\nTreatment was administered by staff at the recruitment clinic.\nEach child was observed for 30 minutes after treatment, a replacement dose was given in case of vomiting.\nNone of the children had repeated vomiting.\nParacetamol (10 mg/kg) was given until symptoms had subsided.\nIn case of parasitological treatment failure, rescue treatment with mefloquine was administered (Lariam\u00ae, Roche, Switzerland; 15 mg/kg on first day and 10 mg/kg on second day).\nAll staff engaged in the trial were blinded as to the treatment group of each child, apart from the study physician who administered medication.\nObjectives\nOur primary objectives were to determine the effect of SP+AS and SP+AS+PQ on submicroscopic P. falciparum gametocyte prevalence and density and to determine the safety of a single dose of PQ in glucose-6-phosphate-dehydrogenase (G6PD) deficient children.\nOutcomes\nThe primary outcomes were gametocyte prevalence and density by real-time nucleic acid sequence-based amplification (QT-NASBA).\nThe secondary outcome was haemoglobin concentration following treatment.\nOther outcomes that we evaluated were microscopic gametocyte prevalence, treatment efficacy and the occurrence of side effects.\nParticipants were encouraged to attend the recruiting clinic at day 1, 2, 3, 7, 14, 28 and 42 after enrolment and at any time the child became unwell.\nOn each day of follow-up, tympanic temperature was measured by electric thermometer and a finger prick blood sample was used for haemoglobin (Hb) measurement using a Hemocue photometer (Angelholm, Sweden), a microscopic slide, a 50 \u00b5L-blood sample for real-time nucleic acid sequence-based amplification (QT-NASBA) and a filter paper sample.\nThe presence of symptoms suggestive of anaemia (fatigue, weakness, dizziness, headache, heart palpitations) or allergic drug reactions (rash) was assessed verbally during every follow-up visit.\nField assistants visited the homes of children who failed to show up to collect additional samples.\nBlood smears were stained for 10 minutes with 10% Giemsa and screened for asexual parasites and gametocytes at enrolment and on day 3, 7, 14, 28 and 42 after treatment.\nAll slides were double-read by experienced microscopists and were declared negative if no parasites were observed in 100 microscopic fields.\nReadings were compared for validation and slides giving discordant results were read by a third reader.\nThe majority result was taken as final in the case of positive versus negative results and geometric mean of the two closest values for density discordants .\nAsexual parasites and gametocytes were counted against 200 and 500 white blood cells, respectively and converted to parasites/\u00b5L by assuming a density of 8000 white blood cells/\u00b5L blood.\nP. falciparum parasite detection by QT-NASBA was performed as described previously.\nBriefly, nucleic acids were extracted from 50 \u00b5L-blood samples with initial RNA extraction carried out in the field following the original Guanidine isothiocyanate (GuSCN) RNA extraction method until the nucleic acids were bound to silica dioxide particles.\nAt this point, samples were stored at \u221220\u00b0C prior to completion of the extraction and QT-NASBA analysis.\nQT-NASBA was performed on a NucliSens EasyQ analyser (bioM\u00e9rieux, Boxtel, the Netherlands) for Pfs25 mRNA.\nThe Pfs25 QT-NASBA is gametocyte-specific and has a detection limit of 10\u2013100 gametocytes/mL.\nNuclisens Basic kits (bioM\u00e9rieux, Boxtel, the Netherlands) were used for amplification according to the manufacturers instructions.\nA standard dilution series of in vitro cultured mature NF54 gametocytes was included in each run to ascertain gametocyte density.\nMolecular Genotyping\nDNA extraction and PCR genotype analysis\nDNA extraction from bloodspots on filter paper was carried out by the chelex-100 method as described by Wooden et al. with some modifications described in Pearce et al.\n.\nIn brief, all samples were extracted in a 96-well plate format.\nThe bloodspot was first soaked in phosphate-buffered saline (PBS) with 0.5% saponin overnight and was then washed twice in 1 ml of PBS.\nThe samples were then boiled for 8 min in 100 \u00b5L of H2O and 50 \u00b5L of 20% chelex suspension in distilled water (pH 9.5) and centrifugated at 5000 rpm for 10 minutes.\n1 \u00b5L of supernatant was used in the PCR reactions.\nGenotyping for MSP-1, MSP-2 and glucose-6-phosphate-dehydrogenase (G6PD) deficiency\nTo differentiate between recrudescent parasites i.e. those persisting from the initial infection and parasites from a new infection, a nested PCR amplification of the polymorphic regions of P. falciparum genes msp1 (K1, MAD20 and RO33) and msp2 (IC1 and FC27) was performed as described by Snounou et al .\nThis PCR was performed for follow-up samples with microscopically confirmed parasitaemia.\nThe PCR products (10 \u00b5L) were run in electrophoresis on 2\u20132.5% metaphor agarose gels in 1xTBE buffer, stained with ethidium bromide and then visualized in UV trans-illumination.\nThe procedure of Cattamanchi et al. was followed in that indeterminate samples for which a majority of novel bands appeared for the post-treatment infection were scored as new infections.\nG6PD deficiency was determined by screening human DNA for single nucleotide polymorphisms in the G6PD gene (G202A, A376G) by a simple high throughput method using PCR, sequence specific oligonucleotide probes (SSOPs) and ELISA-based technology.\nIndividuals with no G202A mutation were classified G6PD B, heterozygotes for the G202A mutation were classified G6PD A and homozygote or hemizygote (males) for the G202A mutation were classified G6PD A-.\nSample size\nThe primary endpoint used for sample size calculation was Pfs25 QT-NASBA gametocyte prevalence after treatment.\nAssuming a gametocyte prevalence in the SP+AS group of 50% on day 14 after treatment, a sample size of 50 individuals per group would allow over a power of 80% to detect a reduction in gametocyte prevalence to 20% in the SP+AS+PQ group, allowing for 5% drop-out and using a significance level of 0.05.\nThis sample size also allowed us to detect a two-fold reduction in gametocyte prevalence during the entire period of follow-up in longitudinal data analyses, assuming an average Pfs25 QT-NASBA gametocyte prevalence of 58% in SP+AS treated children, and a maximum correlation between observations of the same individual of 0.30.\nRandomization\nThe randomization sequence was generated in Stata 8.0 (Stata Corporation, Texas, USA) using restricted randomization with a block size of 20.\nTreatment allocation was determined by opening pre-prepared randomization envelopes in sequence by the study physician.\nThe same physician was involved in participant selection and clinical evaluation.\nParasite carriage by microscopy and Pfs25 QT-NASBA and haemoglobin concentrations were determined by technicians who were unaware of the treatment allocated to study participants.\nStatistical methods\nTherapeutic outcome was classified as early parasitological treatment failure (ETF), late treatment failure (LTF), re-infection or adequate clinical and parasitological response (ACPR).\nHaemoglobin concentrations during follow-up were expressed as a percentage of the enrolment concentration.\nTo quantify the effect of treatment on gametocyte densities, we determined the area under the curve (AUC) of Pfs25 QT-NASBA gametocyte density versus time.\nThis measure incorporates both the magnitude and the duration of gametocyte carriage and was described by M\u00e9ndez et al.\n.\nThe AUC from days 0\u201342 was calculated as: AUC\u200a=\u200a[(3\u22120)\u00d7(g0+g3)/2+(7\u22123)\u00d7(g3+g7)/2+(14\u22127)\u00d7(g7+g14)/2+(28\u221214)\u00d7(g14+g28)/2+(42\u221228)\u00d7(g28+g42)/2]/42; where gd represents Pfs25 QT-NASBA gametocyte density on day d. Gametocyte negative samples were included as zeroes.\nThe measure was scaled by 42 so that it represents AUC per day and this was transformed by log10.\nMicroscopic and QT-NASBA parasite densities were analysed after log10-transformation.\nBecause we were interested in clearance of gametocytes of the original infection, slides from individuals in which PCR analysis defined a new infection were excluded from analyses on post-treatment gametocyte prevalence and density.\nProportions were compared using the chi-squared statistic for a 2-by-2 contingency table.\nNormally-distributed continuous variables were compared using the Student t-test.\nVariables that were not normally distributed were compared using the Wilcoxon rank-sum test.\nMultiple linear regression models were used in case of continuous variables to adjust for potential confounding factors such as asexual parasite and gametocyte density at enrolment.\nMultiple logistic regression models with Generalized Estimating Equations (GEE) were used to test the influence of treatment on dichotomous variables with multiple observations per participant, such as gametocyte prevalence during follow-up.\nEstimates were adjusted for potential confounding factors and a random effect was included in the models to allow for correlations within individuals.\nRegression coefficients (\u03b2) were calculated for continuous dependent variables and odds ratios (OR) for dichotomous dependent variables, both with 95% confidence intervals (95% CI).\nStatistical analyses were performed using SPSS for Windows 12.0 (SPSS Inc., Chicago, USA) and Stata 8.0 (Stata Corporation, Texas, USA).\nResults\nRecruitment and Participant Flow\nA total of 108 children were randomised over the two treatment arms (figure 1).\nThis number exceeds the original sample size (100) because of a high number of patients appearing on the last day of enrolment.\nTwo children (1.9%) were lost for evaluation during the 42-day follow-up period, one in each treatment arm.\nParasite density, gametocyte prevalence, haemoglobin concentrations, fever prevalence and the proportion of G6PD-deficient children at enrolment were not different between the treatment regimens (table 1).\nAdequate clinical and parasitological response (ACPR) on day 42 after treatment was observed in 71.7% (38/53) of the children treated with SP+AS and in 67.9% (36/53) of those treated with SP+AS+PQ (\u03c72\u200a=\u200a1.70; p\u200a=\u200a0.64)(table 2).\nOutcomes and Estimation: Gametocyte carriage after treatment\nMicroscopic gametocyte prevalence at enrolment was 26.4% (14/53) for children treated with SP+AS and 18.9% (10/53) for children treated with SP+AS+PQ (table 1).\nMicroscopic gametocyte prevalence decreased after treatment in both treatment arms (figure 2A) and was significantly lower in individuals treated with SP+AS+PQ compared to those treated with SP+AS when the entire period of follow-up was considered, after adjustment for enrolment gametocyte prevalence (GEE: OR\u200a=\u200a0.17 (95% CI\u200a=\u200a0.049\u20130.57), p\u200a=\u200a0.004).\nNo microscopic gametocyte carriage was seen on day 7 and 14 after treatment with SP+AS+PQ.\nEnrolment gametocyte prevalence defined by Pfs25 QT-NASBA was 88.2% (45/51) for SP+AS treated individuals compared to 90.6% (48/53) for SP+AS+PQ treated children (table 1).\nPredictably, children with microscopically confirmed gametocytes at enrolment had a significantly higher median Pfs25 QT-NASBA gametocyte density (46.6 gametocytes/\u00b5L; IQR 8.9\u2013346.5) compared to those gametocyte-free by microscopy (7.5 gametocytes/\u00b5L; IQR 3.1\u201367.5)(Wilcoxon Rank-Sum test z\u200a=\u200a\u22122,01, p\u200a=\u200a0.04).\nPfs25 QT-NASBA gametocyte density was not initially defined as outcome measure, but densities were compared between the treatment arms post-hoc.\nAt the time of enrolment Pfs25 QT-NASBA gametocyte density was 28.8 gametocytes/\u00b5L (IQR 6.9\u2013109.9 gametocytes/\u00b5L) for SP+AS treated children compared to 17.5 gametocytes/\u00b5L (IQR 1.1\u201376.9) for SP+AS+PQ treated children (table 1).\nDuring follow-up Pfs25 QT-NASBA gametocyte prevalence decreased for both SP+AS and SP+AS+PQ treated children (figure 2B and table 3).\nAfter adjustment for enrolment Pfs25 QT-NASBA gametocyte density, Pfs25 QT-NASBA gametocyte prevalence was lower in individuals treated with SP+AS+PQ during the entire period of follow-up (GEE: OR\u200a=\u200a0.27 (95% CI\u200a=\u200a0.18\u20130.40), p<0.001).\nThe Pfs25 QT-NASBA gametocyte density in gametocyte positive samples was consistently lower for SP+AS+PQ treated children during follow-up (table 3).\nThe average duration of gametocyte carriage was summarised in the area under the curve of Pfs25 QT-NASBA gametocyte density versus time (table 4).\nThe area under the curve was significantly lower for SP+AS+PQ treated children (t-test t\u200a=\u200a3.28, p <0.001), as were the number of sampling times when gametocytes were detected (GEE: OR\u200a=\u200a0.27 (95% CI\u200a=\u200a0.18\u20130.40), p<0.001) and the geometric mean gametocyte density in gametocyte positive samples (GEE: \u03b2\u200a=\u200a\u22120.40 (95% CI\u200a=\u200a\u22120.74\u2212\u22120.07), p\u200a=\u200a0.02).\nMost treatment failures or re-infections appeared after day 14 (table 2).\nIt is therefore appropriate to focus on the first two weeks after treatment to determine the effect of treatment in the absence of treatment failure or re-infection.\nOn day 14 after SP+AS treatment, 32 (62.7%) individuals were gametocyte positive with a mean gametocyte density of 4.5 gametocytes per \u00b5L (IQR 1.8\u201337.1)(table 3).\nOnly two individuals (3.9%) were positive by Pfs25 QT-NASBA on day 14 after treatment with SP+AS+PQ with low gametocyte densities of 0.067 and 0.094 gametocytes per \u00b5L.\nHaemoglobin concentrations after treatment\nTo address concerns about haemolysis associated with PQ use in G6PD deficient individuals, Hb concentration was assessed at enrolment and during follow-up.\nMedian Hb at enrolment was 10.5 g/dL (IQR 9.4\u201311.8) for SP+AS treated children and 10.8 g/dL (9.8\u201311.8) for SP+AS+PQ treated children (table 1).\nWhile Hb relative to enrolment concentration gradually increased in the weeks after treatment with SP+AS, it first decreased for SP+AS+PQ treatment and remained lower up to day 14 (figure 3).\nThe relative decrease was most pronounced on day 7 after SP+AS+PQ treatment when Hb concentration was 5.2% lower (95% CI \u22128.4\u2013\u22121.8) than on enrolment (paired t-test: t\u200a=\u200a2.86; p\u200a=\u200a0.006).\nWhen only children with the G6PD B variant were considered, the relative decrease in Hb concentration on day 7 after treatment with SP+AS+PQ was no longer statistically significant (paired t-test: t\u200a=\u200a1.57; p\u200a=\u200a0.12).\nThe reduction in Hb shortly after SP+AS+PQ treatment was most pronounced in children with the G6PD A- variant (figure 4) although numbers were too small to allow statistical comparisons.\nNone of the children developed clinical symptoms related to anaemia or an Hb below 5g/dL.\nThe Hb concentrations on day 28 and day 42 after treatment were equal to or greater than that at enrolment for all G6PD categories (figure 4).\nEight children experienced \u226520% reduction in Hb concentration on day 7 relative to that at enrolment, compared to none in the SP+AS treatment arm.\nThe range of Hb concentration in these eight individuals was 5.4\u201310.4 g/dL on day 7 after SP+AS+PQ treatment.\nTwo of the children with \u226520% reduction in Hb concentration on day 7 had the G6PD A- variant (25%), one had the G6PD A variant (12.5%) and the other five had G6PD B variant (62.5%).\nNone of the children reported symptoms suggestive of anaemia or allergic drug reactions during follow-up.\nDiscussion\nInterpretation\nThis study shows that a single dose of primaquine (PQ) is of significant additive value in clearing gametocytes in an area of high malaria endemicity in Tanzania.\nOnly 3.9% of children treated with SP+AS+PQ had gametocytes on day 14 after successful treatment compared to 62.7% for SP+AS treated children.\nGametocytes persisting on day 14 after SP+AS+PQ treatment circulated at densities well below the theoretical threshold for mosquito infection.\nGametocyte carriage after treatment with SP+AS persisted for more than one month and more than two-thirds of the treated individuals harboured gametocytes on day 14.\nThese data confirm previous findings.\nThe addition of a single dose of PQ to this regime significantly reduced gametocyte carriage.\nWhen microscopy was used as tool to detect gametocytes, no gametocytes were observed seven days after the initiation of treatment with SP+AS+PQ.\nOn day 28 after treatment, gametocytes were detected in only one individual.\nSubmicroscopic gametocyte prevalence decreased to 3.9% (2/51) on day 14 after initiation of treatment but increased at subsequent follow up time points.\nThis increase is most likely due to recrudescence of infection or new infections that were undetected by microscopy.\nTreatment failure rates were similar to those reported in recent drug efficacy studies in east Africa and re-infection rates were high in this area of intense malaria transmission.\nAs a consequence, 28.3% of the SP+AS treated children and 32.1% of the SP+AS+PQ treated children experienced microscopically confirmed treatment failure or re-infection during follow-up.\nTrue treatment failure and re-infection rates may be higher as not all asexual parasites will be detected by microscopy.\nA high persistence of submicroscopic asexual parasites after apparently successful drug treatment has been described, as well as the acquisition of new sub-patent infections.\nGametocytes that newly developed from new or persisting asexual infections are therefore likely to be responsible for the increase in gametocyte prevalence after day 14 that has been reported previously.\nFocusing on the first two weeks after initiation of treatment, PQ seems to decrease the number of gametocytes rapidly to a level where onward transmission may be arrested completely.\nAlthough analyses on gametocyte densities should be considered as post-hoc analyses and we did not directly determine post-treatment malaria transmission, it is clear that the transmission potential is reduced in children treated with SP+AS+PQ.\nGametocyte prevalence and density were lower in children treated with SP+AS+PQ and gametocytes persisting on day 14 after treatment circulated at densities below 0.1 gametocyte/\u00b5L.\nWhen assuming an average mosquito blood meal size of 2\u20133 \u00b5L, these concentrations are unlikely to result in mosquito infection, although infectiousness of very low gametocyte densities can not be excluded.\nGeneralisability\nThe rationale for using PQ in P. falciparum infections is to reduce post-treatment infectivity (as routinely practiced in several countries in Asia and the Americas) and as a quarantine treatment to reduce the spread of (drug resistant) parasites.\nThe aim of the current study was not to determine if PQ should be routinely added to current ACT regimens.\nThe objective was to explore if PQ is a valuable component in mass drug administration studies that aim to reduce malaria transmission.\nMass drug administration (MDA) studies are typically conducted in areas of low malaria transmission intensity and also include asymptomatic parasite carriers.\nThe likelihood of re-infection is lower in these areas and the effect of PQ can be expected to be larger because enrolment gametocyte densities are lower.\nIn MDA studies, drugs are administered to asymptomatic individuals, raising safety concerns over the haemolytic effect of PQ on individuals with G6PD mutations.\nDue to the potential protective effect of G6PD-deficiency from malaria, G6PD deficiency is probably selected in malaria endemic regions, in a similar manner as for other haemoglobinopathies.\nIn our population, 6.5% (7/107) of the children had the A- variant of G6PD deficiency.\nWe observed that the addition of PQ resulted in a statistically significant but transient reduction in haemoglobin levels, as was previously reported for PQ administered at a curative dose (0.5 mg/kg, 14 days) to individuals with the African Variant (A-) of G6PD-deficiency.\nThe observed reduction in Hb concentration indicates a small but genuine risk of PQ use in individuals who are anaemic prior to drug administration.\nThe risk to the individual patient has to be weighted against the potential benefit of a reduced malaria transmission.\nIn the current study, we excluded individuals with an Hb<8 g/dL.\nThe risk of anaemia is likely to be increased in individuals with lower pre-treatment Hb concentrations, which is important since Hb is not routinely determined prior to treatment.\nIn case of mass-administration of SP+AS+PQ, anaemic individuals should preferably be identified prior to treatment and excluded from PQ treatment.\nThis can be done by Hemocue\u00ae which allows a rapid and reliable assessment of anaemia in the field.\nAreas of low and seasonal malaria transmission are those that are most likely to benefit from MDA.\nIn these areas, the prevalence of severe anaemia prior to the malaria season is likely to be low.\nIn these circumstances we consider the addition of a single dose of PQ to SP+AS to be a safe approach to reduce post-treatment malaria transmission.\nOverall evidence\nThis is the first study that determines the effect of PQ on submicroscopic gametocyte densities and our findings are in line with previous studies that where PQ efficiently cleared microscopic gametocyte concentrations.\nThe addition of a single dose of PQ had no beneficial influence on the clearance of asexual parasites, as was described previously.\nProfile of the study\nGametocyte prevalence by microscopy (A) and Pfs25 QT-NASBA (B).Gametocyte prevalence for SP+AS (closed diamonds, solid line) and SP+AS+PQ (open triangles, broken lines) treated children. Bars indicate the 95% confidence intervals around the proportions. * indicates a statistically significant difference between the two treatment arms.\nHaemoglobin concentration following treatment.Concentrations are expressed relative to that at enrolment for SP+AS (closed diamonds, solid line) and SP+AS+PQ (open triangles, broken lines). Bars indicate the 95% confidence intervals around the proportions. * indicates a statistically significant difference between the two treatment arms.\nRelative haemoglobin concentration after treatment with SP+AS+PQ for different G6PD genotypes.Haemoglobin concentration relative to enrolment for children without the G202A mutation (G6PD genotype B; black diamonds, n\u200a=\u200a39), heterozygotes (G6PD genotype A; white triangles, n\u200a=\u200a9) and homozygotes or hemizygotes (G6PD genotype A-: grey diamonds, n\u200a=\u200a4). Each individual measurement is shown; lines indicate the median value.\n\nCharacteristics of the study population at enrolment\n | SP+AS | SP+AS+PQ\nN | 54 | 54\nAge, median (IQR) | 5 (3\u20138) | 5.5 (3\u201310)\nSex, % male (n/N) | 46.3 (25/54) | 55.6 (30/54)\nTemperature, % fever (>37.5\u00b0C) (n/N) | 30.2 (16/53) | 35.8 (19/53)\nAsexual parasite density GM/\u00b5L (IQR) | 4,759 (959\u201322,248) | 7,379 (2,334\u201325,304)\nMicroscopic gametocyte prevalence, % (n/N) | 26.4 (14/53) | 18.9 (10/53)\nPfs25 QT-NASBA gametocyte prevalence, % (n/N) | 88.2 (45/51) | 90.6 (48/53)\nPfs25 QT-NASBA gametocyte density, GM/\u00b5L (IQR) * | 28.8 (6.9\u2013109.9) | 17.5 (1.1\u201376.9)\nG6PD, % (n/N)\nB | 68.5 (37/54) | 75.5 (40/53)\nA | 25.9 (14/54) | 17.0 (9/53)\nA- | 5.6 (3/54) | 7.5 (4/53)\nHaemoglobin concentration, g/dL, median (IQR) | 10.5 (9.4\u201311.8) | 10.8 (9.8\u201311.8)\n\nIQR\u200a=\u200ainterquartile range; GM\u200a=\u200ageometric mean; G6PD: B\u200a=\u200ano G202A mutation; A\u200a=\u200aheterozygotes, single G202A mutation; A-\u200a=\u200ahomozygote or hemizygote (males), only G202A mutations. *For gametocyte carriers only.\n\nTreatment outcome for the different treatment regimens on day 14, 28 and 42.\n | SP+AS | SP+AS+PQ\nNumber evaluated | 53 | 53\nDay 14 Treatment outcome, % (n)\nACPR | 98.1 (52) | 100.0 (53)\nETF | 0 (0) | 0\nLTF | 0 (0) | 0\nRe-infection | 1.9 (1) | 0\nIndeterminate | 0 (0) | 0\nDay 28 Treatment outcome, % (n)\nACPR | 77.4 (41) | 83.0 (44)\nETF | 0 (0) | 0 (0)\nLTF | 3.8 (2) | 5.7 (3)\nRe-infection | 15.1 (8) | 9.4 (5)\nIndeterminate | 3.8 (2) | 1.9 (1)\nDay 42 Treatment outcome, % (n)\nACPR | 71.7 (38) | 67.9 (36)\nETF | 0 (0) | 0 (0)\nLTF | 3.8 (2) | 9.4 (5)\nRe-infection | 17.0 (9) | 13.2 (7)\nIndeterminate | 7.5 (4) | 9.4 (5)\n\nACPR\u200a=\u200aadequate clinical and parasitological response, ETF\u200a=\u200aearly treatment failure, LTF\u200a=\u200alate treatment failure. Indeterminate\u200a=\u200aindeterminate due to PCR-failure (n\u200a=\u200a1) or missing filter paper DNA samples (n\u200a=\u200a8).\n\n\nPfs25 QT-NASBA gametocyte prevalence and density after treatment with SP+AS and SP+AS+PQ.\n | SP+AS | SP+AS+PQ\n | Gametocyte prevalence, % (n/N) | Gametocyte density/\u00b5L, GM (IQR) | Gametocyte prevalence, % (n/N) | Gametocyte density/\u00b5L, GM (IQR)\nDay 0 | 88.2% (45/51) | 28.8 (6.9\u2013109.9) | 90.6% (48/53) | 17.5 (1.1\u201376.9)\nDay 3 | 80.8% (42/52) | 25.5 (4.2\u2013132.9) | 53.8% (28/52) | 2.9 (0.6\u201336.6)\nDay 7 | 71.7% (38/53) | 8.0 (1.8\u201359.1) | 15.7% (8/51) | 0.3 (0.02\u20132.4)\nDay 14 | 62.7% (32/51) | 4.5 (1.8\u201337.1) | 3.9% (2/51) | 0.07&0.09\nDay 28 | 41.9% (18/43) | 10.9 (1.0\u201370.0) | 10.4% (4/46) | 3.8 (0.2\u20132.4)\nDay 42 | 28.2% (11/39) | 17.4 (1.2\u2013191.4) | 16.3% (5/40) | 7.9 (0.5\u201354.3)\n\nGM\u200a=\u200ageometric mean Pfs25 QT-NASBA gametocyte density per microlitre for gametocyte carriers only; IQR\u200a=\u200ainterquartile range\n\nEffect of treatment on gametocyte carriage during follow-up\n | SP+AS | SP+AS+PQ | p-value\nMean AUC of gametocyte density/\u00b5L versus time, (IQR) | 11.1 (2.2\u201353.8) | 1.5 (0.3\u20138.8) | <0.001*\nNumber of sampling times when gametocytes were detected, % (n/N) | 64.4 (186/289) | 32.4 (95/293) | <0.001\u2020\nGM gametocyte density/\u00b5L on days when gametocytes were detected, (IQR) | 15.8 (4.1\u201385.8) | 5.8 (0.8\u201355.1) | 0.02\u2020\n\nAUC\u200a=\u200aarea under the curve; GM\u200a=\u200ageometric mean; IQR\u200a=\u200ainterquartile range\nAdjusted for gametocyte density at enrolment; \u2020 adjusted for correlations between observations from the same individual",
        "label": "low",
        "id": "task4_RLD_test_946"
    },
    {
        "paper_doi": "10.1186/1475-2875-13-324",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Cluster-RCTUnit of randomization: cluster of housesICC is not reported.Trial duration: 14 months from July 2009 to August 2010\n\n\nParticipants: Adults or children living in endemic areas.\n\n\nInterventions: 15% DEET lotion versus placebo lotionCo-interventions: LLINsTreatment arms:- DEET 15% + LLINs arm - 10 clusters, 468 households and 2224 participants- Placebo + LLINs arm - 10 clusters, 469 households and 2202 participants\n\n\nOutcomes: - Participants with clinical malaria confirmed through blood smears or rapid diagnostic tests (P. falciparum); and- Adherence to regular usage of the intervention.\n\n\nNotes: Trial was conducted in rural communities of the Ulanga district, Kilombero Valley, Tanzania.Trial registration number: ISRCTN92202008Funded by Population Services International\n\n",
        "objective": "To assess the impact of topical repellents, insecticide\u2010treated clothing, and spatial repellents on malaria transmission.",
        "full_paper": "Background\nLong-lasting insecticidal nets (LLINs) have limited effect on malaria transmitted outside of sleeping hours.\nTopical repellents have demonstrated reduction in the incidence of malaria transmitted in the early evening.\nThis study assessed whether 15% DEET topical repellent used in combination with LLINs can prevent greater malaria transmission than placebo and LLINs, in rural Tanzania.\nMethods\nA cluster-randomized, placebo-controlled trial was conducted between July 2009 and August 2010 in a rural Tanzanian village.\nSample size calculation determined that 10 clusters of 47 households with five people/household were needed to observe a 24% treatment effect at the two-tailed 5% significance level, with 90% power, assuming a baseline malaria incidence of one case/person/year.\nTen clusters each were randomly assigned to repellent and control groups by lottery.\nA total of 4,426 individuals older than six months were enrolled.\nAll households in the village were provided with an LLIN per sleeping space.\nRepellent and placebo lotion was replaced monthly.\nThe main outcome was rapid diagnostic test (RDT)-confirmed malaria measured by passive case detection (PCD).\nIncidence rate ratios were estimated from a Poisson model, with adjustment for potential confounders, determined a priori.\nAccording-to-protocol approach was used for all primary analyses.\nResults\nThe placebo group comprised 1972.3 person-years with 68.29 (95% C.I 37.05-99.53) malaria cases/1,000 person-years.\nThe repellent group comprised 1,952.8 person-years with 60.45 (95% C.I 48.30-72.60) cases/1,000 person-years, demonstrating a non-significant 11.44% reduction in malaria incidence rate in this group, (Wilcoxon rank sum z\u2009=\u20090.529, p\u2009=\u20090.596).\nPrincipal components analysis (PCA) of the socio-economic status (SES) of the two groups demonstrated that the control group had a higher SES (Pearson\u2019s chi square\u2009=\u200913.38, p\u2009=\u20090.004).\nConclusions\nLack of an intervention effect was likely a result of lack of statistical power, poor capture of malaria events or bias caused by imbalance in the SES of the two groups.\nLow malaria transmission during the study period could have masked the intervention effect and a larger study size was needed to increase discriminatory power.\nAlternatively, topical repellents may have no impact on malaria transmission in this scenario.\nDesign and implementation of repellent intervention studies is discussed.\nTrial registration\nThe trial was registered ISRCTN92202008 - http://www.controlled-trials.com/ISRCTN92202008\nElectronic supplementary material\nThe online version of this article (doi:10.1186/1475-2875-13-324) contains supplementary material, which is available to authorized users.\nBackground\nIn the past decade, considerable financial and political resources have been mobilized for malaria control.\nThis has in turn led to extensive coverage and use of existing control tools, like long-lasting insecticidal nets LLINs and indoor-residual spraying (IRS).\nImplementation of these highly effective vector control tools has resulted in substantial decrease in malaria transmission, morbidity and mortality.\nDespite both extensive coverage and use, the sole use of these tools have not and will not be able to eliminate malaria in all malaria endemic regions.\nBecause LLINs and IRS target mainly indoor biting and indoor resting vectors their implementation may select for outdoor resting and biting vector populations that often become dominant, so that even though there is a diminished malaria transmission as a result of extensive LLINs and IRS use, there is likely to be a larger proportion of this residual transmission occurring outdoors compared to indoors.\nIncreased urbanization and rural electrification programmes have also had an impact on malaria transmission dynamics.\nAs a result of this, individuals stay up later in the evenings than they usually would in a situation where electricity was not available, and are, therefore, exposed to potentially infective mosquito bites for longer.\nWith the renewed push for malaria elimination, it is evident that new tools need to be developed to augment existing vector control tools to achieve this goal.\nTopical repellents provide excellent personal protection and could potentially be used to complement LLINs for additional protection from residual transmission.\nSeveral studies demonstrated that topical repellents offer additional protection from malaria transmission either when used alone, or in combination with LLINs, in areas with high early evening and outdoor malaria transmission.\nThis study assessed the potential additional benefit of using topical repellents in combination with LLINs compared to using only LLINs on early evening malaria transmission in a rural community in Kilombero valley, south-west Tanzania.\nThis community mainly relies on subsistence farming of rice, which provides for a large breeding site for both malaria vectors and nuisance biting mosquitoes.\nIt is customary that the community in the study area cook outdoors in the early evenings, a situation that is likely to expose them to mosquito bites and potential malaria transmission.\nRural development is also rapidly taking place in this study area.\nAs a result, many members of the community usually gather in the early evening and stay late into the night at local entertainment spots that are springing up in the study area owing to rural electrification programmes, thereby increasing the potential of malaria transmission at these times.\nA recent report estimates a malaria incidence rate of 0.67 cases/person/year confirmed by rapid diagnostic test (RDT) from passive case detection at a local clinic between December 2012 and July 2013 (Jabari Mohammed Namamba, pers. comm.).\nIn the past two decades, extensive malaria intervention programmes have taken place in this area, and it is therefore expected that the community be highly sensitized on malaria transmission and control methods.\nThere is high LLIN use in the study area.\nRepellent awareness and knowledge as assessed using a Knowledge, Attitude and Practice (KAP) baseline questionnaire at the inception of the clinical trial determined that this community did not use topical repellents as a mosquito control tool.\nAwareness and availability were reported as the major reasons for not using topical repellents [Sangoro O, Sarah M, Ann HK, Sarah M: Feasibility of repellent use in a context of increasing outdoor transmission: A Qualitative study in rural Tanzania, submitted to Malaria Journal for publication].\nThe major malaria vector in the study area is Anopheles arabiensis, which has been shown to exhibit elastic feeding behaviour depending on the availability and location of the host and is known to exhibit early evening biting.\nThe dominance of this vector in this area is also likely to be the result of extensive LLIN use in the study area.\nA field study conducted in the study area to determine the efficacy of this repellent (15% DEET) against An. arabiensis demonstrated >80% protection from bites over four hours of mosquito collection.\nTherefore, 15% DEET was considered appropriate to provide protection against early evening biting.\nThis study area was chosen because there are no studies that have been conducted to assess the additional benefits of topical repellents to LLINs in malaria control in East Africa, although this technology has been shown to work elsewhere in sub-Saharan Africa.\nAlso the vectors present in the area, An. arabiensis, exhibit early evening biting, a trait that made the use of repellents in the early evening ideal in this area.\nTherefore, even though extensive employment of current control tools will lower malaria transmission in this area, its is likely that residual transmission will continue occur at times when the effectiveness of these tools is diminished, like outdoors in the early evenings and mornings, and will require supplementary tools that target this scenario.\nTherefore, it was hypothesized that combined use of LLINs and topical repellents in this community would have a greater impact on malaria transmission in the early evening compared to sole use of LLINs.\nMethods\nStudy area\nThe study was carried out in Mbingu village, Ulanga district, situated 55kms west of Ifakara town at 8.195\u00b0S and 36.259\u00b0E.\nAt the time of the study inception, (July 2009), the village was estimated to have 7,609 inhabitants.\nThere is moderate malaria transmission in the study area, with peak transmission occurring in the months of May and June after the long rains.\nThe village experiences an annual rainfall of approximately 1,200-1,800\u00a0mm and an annual temperature range of between 20\u00b0C and 32.6\u00b0C.\nThe village borders an extensive field cleared for rice irrigation, which provides an ideal breeding site for malaria vectors.\nSample size rationale\nThe only available data from the study area were community reported fever incidence rate estimates of 3.2 cases/person/year for children under the age of five years.\nAssuming fever rates in children under five years are higher than the rest of the population, and that not all fevers reported are caused by malaria, a rate of one malaria case/person/year was used to calculate the sample size needed for this study.\nAvailable reports also indicated that 30% of mosquito bites occurs in the early evening.\nTherefore, assuming that mosquitoes have an equal probability of carrying sporozoites regardless of time of night, it was assumed there was a potential 30% malaria transmission occurring in the early evenings.\nExpecting that repellents would reduce 80% of this potential 30% early evening transmission, as observed from the field study, it was reasoned that repellents would reduce the overall transmission of malaria from one case/person/year to 0.76 cases/person/year.\nUsing the methods of Hayes et al. for sample size calculation for cluster randomized trials, it was estimated that to observe this treatment effect (24%), with 90% power at the two-tailed 5% significance level, 10 clusters of 47 households with five members each was required per treatment group.\nA coefficient of variation (k) of 0.20 was used based on published recommendations as the inter-cluster variation could not be estimated.\nHousehold recruitment\nHouseholds were recruited into the study in two phases.\nIn phase one, the study investigators and field team visited the study village for reconnaissance and introduction to the community leaders and members in December 2008.\nA week later, the study team returned to the study village and aided by community leaders, identified the centre of the village.\nHere, the field team spun a ballpoint pen and visited all the households that the writing end of the pen pointed to with the intention of recruiting all consenting households into the study.\nAfter all households in this direction had been exhausted, the field team went back to the village centre and spun the pen to choose the next direction in which to visit the households.\nIf the pen pointed in the direction where the households were already visited, then, the pen was spun again until a new direction was identified.\nThis progression was repeated until approximately, 1,000 households had been visited and recruited.\nThe village had 2,000 households and, therefore, by visiting and potentially enrolling at least 50% of the households, the study team were confident that they had captured a representative sample of households in the study area.\nEnrolment of households into the study\nDuring the household recruitment visits, each household head was informed of the purpose of the visit.\nThey were educated on the objectives, risks and benefits of the study to their household and the community.\nThey were encouraged to ask questions and after all their concerns had been addressed, they were asked if they were willing to participate in the study.\nIf willing, each household head was asked to sign a written informed consent form, confirming their participation and that of all household members.\nAs data was being collected at the household level, only the household head was asked for informed consent.\nIt was assumed that once that household head gave consent then all household members would likely comply with repellent use following instructions of the household head as the authority in each household.\nA structured questionnaire on the socio-economic status (SES) of the household and knowledge, attitude and practice (KAP) in relation to malaria and repellents was then administered [Sangoro O, Sarah M, Ann HK, Sarah M: Feasibility of repellent use in a context of increasing outdoor transmission: A Qualitative study in rural Tanzania, submitted to Malaria Journal for publication].\nThe GPS coordinate of the household enrolled was then recorded using a handheld GPS receiver (Garmin eTrex Legend\u00ae H).\nThese coordinates were then plotted using Arc GIS software (Arc GIS 9.0, ESRI, UK), to generate a map of all the households enrolled in the study area.\nSecond phase of household recruitment, household enrolment and cluster generation\nIn phase two, the map generated during the first phase of recruitment was used to delineate 20 clusters of households each while ensuring a buffer zone of 200 metres between clusters to prevent diversion of mosquitoes from the intervention group to the control group.\nAs a result of creation of this buffer area, some households that had been recruited in the first phase fell within this 200 metre buffer area.\nThese households were excluded from the study during this second phase of recruitment.\nTherefore, even though about 1,000 households were recruited in the first phase, more households needed to be recruited in the second phase as a result of loss of households within the buffer area.\nThese households were excluded because they would have potentially confounded the outcome of the study in case of diversion of mosquitoes.\nAll households within the buffer area were issued with an LLIN per sleeping space to protect them from potentially greater than normal bites from diverted mosquitoes.\nIn practice, the second phase of recruitment proceeded as follows: The field team visited the 20 clusters, using the household considered to be at the centre of these clusters (identified from the Arc GIS map), as the starting point.\nThe household head of the central household in the cluster was informed of the purpose of the visit.\nIf the household had been enrolled during the first phase of household recruitment, then the field team issued an LLIN for every sleeping space, stapled a unique identifier number on the door frame and moved to the next nearest household.\nIf the households had not been enrolled, the household head was informed of the objectives, risks and benefits of the study, enrolled on written informed consent, provided with a unique household identifier and LLINs for each sleeping space, and a SES and KAP questionnaire administered.\nThis progression was repeated until 47 households close together were enrolled to form a single cluster.\nAll 47 households in each of the 20 clusters were enrolled in this manner.\nThe newly enrolled households that did not appear on the map generated in the first phase of recruitment were plotted and the map updated to produce the final map of households recruited into the study (Figure\u00a01).\nClusters were used as the unit of randomization for three reasons: 1) since the intervention would be applied to a community, if proven to be effective, 2) to limit contamination of treatments between households, and 3) to avoid diversion of mosquitoes from individuals who used repellents to those who did not use repellent within the same household of from households using repellents to households that used the placebo, thereby putting non-repellent using individuals and households at a potentially higher risk of contracting malaria.\nEligibility criteria\nAll households were eligible to be recruited into the trial and no household was excluded on the basis of household structure, asset or livestock ownership.\nAll individuals older than six months of age were eligible to be recruited into the trial.\nThis age cut-off was used because re evaluation of DEET insect repellent estimated the margin of exposure (MOE) in children less than six months to be less than 100.\nMargin of exposure is defined as the ratio of dose of DEET used daily to the no observed effect level dose recommended by regulation agencies, which usually consider doses, which result in MOEs of less than 100, unacceptable.\nBased on this risk assessment, use of DEET was not recommended for children under six months.\nRandomization of clusters to treatments\nAll the 20 clusters in the map (Figure\u00a01) were assigned numbers 1 to 20, starting from the left hand side to the right.\nThe cluster numbers were then written down on small pieces of paper, which were placed in a bowl.\nThe principal investigator (PI) and project leader (PL) then drew the pieces of paper from the bowl one at a time.\nTwo three digit numbers (258 and 305) were used to classify clusters in to two groups.\nThe first cluster number to be drawn was assigned treatment 258 and the second cluster number assigned treatment 305.\nThis progression was repeated until all the clusters had been assigned to one of the two groups.\nBlinding\nThe repellent and placebo lotion smelt and felt the same and were placed in identical tubes, distinguishable only by the two three-digit numbers known only to the independent code keeper (SC Johnson and Sons).\nHowever, the PI and PL had previously conducted efficacy test of these two treatments, and could identify the repellent and placebo from the results of this study.\nTherefore, it was only the field team, study statistician and study participants who were blinded in this study.\nBlinding was broken after analysis.\nRepellent issuance, application and compliance\nIn June 2009, the field team visited all households enrolled in the study to distribute treatments to study participants.\nThe treatments, (15% DEET and placebo), both formulated as a pourable lotion that is applied by hand, were supplied by SC Johnson, Racine, USA, and packaged in 100\u00a0ml plastic tubes.\nDuring this visit, the field team informed the household members on how to apply the treatments provided on exposed areas of the body.\nThey also advised the participants not to apply the treatments on open wounds, eyes, mouth and areas with mucous membranes.\nThe repellent lotion was applied at an approximate rate of 0.002\u00a0mg DEET/cm2, the quantity of repellent that prevented >80% mosquito bites for 4\u00a0hours in a controlled environment and in the study area.\nEven though a repellent with a higher concentration would have provided greater protection, the Tanzania National Institute of Medical Research ethical approval board did not allow the use of a repellent that had more than 15% DEET due to safety concerns, despite the initial request of the PI to use 30% DEET and submission of detailed experimental justification and dossier of safety data justifying the use of a higher concentration.\nThe participants were issued measuring caps, with amounts of repellent required for adults (7mls) and children below 12\u00a0years (3mls) marked on the cap.\nEach tube held 100mls of repellent.\nTherefore, two tubes were considered enough to last an adult one month, i.e. if they applied the recommended dosage of 7 mls per day, while one tube was enough to last a child\u2009<\u200912\u00a0years for one month, if they used 3mls per day.\nChildren\u2009>\u200912\u00a0years were advised to use up to 7mls a day, and were therefore issued with 2 tubes for the month.\nAll the tubes issued per cluster and households were identical, and it is possible that the household members shared a single tube of repellent until it ran out.\nAs all households member were issued with enough treatment to last them month, either 15% DEET repellent lotion or placebo, and dosages for adults and children had been marked out, it was assumed that sharing of repellents within the household would have no effect on the outcome as long as there was daily compliance to the recommended dose by the participants.\nThe amounts recommended were adjusted to accommodate for individuals with greater than average body mass as it was determined from semi-field and field experiments that an average sized volunteer required 6 mls.\nThis amount was, therefore, adjusted upwards by an extra millilitre.\nThe community members were instructed to apply the repellent at dusk (1800\u00a0hrs) and to reapply it if they felt any mosquito bites or remained active for more than four hours after sunset.\nCompliance to lotion use (both repellent and placebo) was assessed by the field team visiting the enrolled households at the beginning of each subsequent month (monthly monitoring surveys) to issue new tubes of repellent and placebo lotion.\nTherefore compliance was assessed on a monthly basis using a short structured questionnaire, where the household head or an adult household member, was asked if all household members had used the repellents and reasons for non-compliance where relevant.\nHowever, as self-reported data are unreliable, the number of repellent/placebo tubes issued every month was also recorded as a secondary measure of compliance, to determine if there was a difference in the number of tubes issued in each month per treatment group.\nData on use of LLINs the previous night, malaria infection, recalled febrile illness and visit to the health centre during that month was also collected.\nIf, during these monthly monitoring surveys, the household head or any other adult household member was not available to answer the questionnaire on compliance, the field team visited that particular household daily for seven consecutive days.\nIf still no household member able to take the monitoring survey was available during these repeated visits, then that household, and all it members, was excluded from the calculation of person-time for that month.\nIn addition to the compliance, malaria and recalled febrile illness data collected during each month of the study period, an after study questionnaire was administered at the close of the study to assess the participants\u2019 knowledge, attitudes and practice in relation to repellents.\nThese results are reported elsewhere [Sangoro O, Sarah M, Ann HK, Sarah M: Feasibility of repellent use in a context of increasing outdoor transmission: A Qualitative study in rural Tanzania, submitted to Malaria Journal for publication].\nClinical data collection\nA single government health facility in the study area was recruited into the study.\nAt this facility, health services were provided for free by the project if the participants showed their project identification card with a household unique identification number on it.\nCommunity members that were not enrolled into the study were issued with a different kind of identification card to also allow them free consultation and treatment at the recruited health facility.\nThis was done to discourage community members attending the health facility under the guise of being a study participant and, therefore, contaminating the study by recording malaria status of community members not enrolled in the study as participants.\nIt was assumed that since services were provided for free at this facility, it would attract most community members seeking health services.\nA clinical officer (CO) and a nurse were employed by the project at this health facility.\nA ledger with the household unique identifier and names of each household member was drawn up and placed at this health facility.\nWhen a study participant visited the health facility with febrile illness, the CO checked against their name and household unique ID in the health facility ledger.\nThis way household and health facility data could be reconciled using the household unique identifier.\nFebrile participants were tested for malaria using rapid diagnostic test (RDT) (ICT Malaria cassette tests HRPII/pf test kit).\nA proportion of participants also had diagnosis by thick film microscopy to confirm the accuracy of the RDTs for diagnosis under field conditions.\nThe result of the RDT and the date of diagnosis were marked against the Household ID on the health facility ledger.\nThose found positive for malaria parasites were given artemether-lumefantrine (ALu), the first-line drug for treatment of malaria in Tanzania.\nOnly participants that were RDT or slide positive for malaria parasites were treated.\nThis was to avoid treating non-malaria patients with ALu, which might have affected malaria incidence rate in the village.\nThe RDT\u2019s were labelled with the patient\u2019s unique identifier, date and status (+ve or \u2013ve) and stored for verification.\nThese were later checked against the clinical trial database to ensure that no cases had been incorrectly entered into the database by the clinic staff.\nData management\nData from the structured questionnaires on SES of households and KAP in relation to malaria and repellents administered at baseline; follow-up data on compliance and recalled febrile illness administered throughout the study period; and the after study KAP survey, were double entered into a computer using an Epi \u2013Info\u2122 template with a drop down lists of values that corresponded to the format of the questionnaires.\nData was then exported to Microsoft Access 2008 (Microsoft Corporation), to check for lack/excesses of data, inconsistencies and outliers.\nAll data from the above mentioned questionnaires were linked using the household unique identifier.\nThe household unique identifier was made up of the household number, cluster number and treatment number.\nStatistical analysis\nData was collected and presented at household and cluster level as the study aimed at assessing the effectiveness of the repellents at the community level.\nIndividual level data was not collected.\nSocio-economic status (SES)\nAll data cleaning and analysis was performed using STATA 11.2 software (StataCorp LP, College Station, Texas, USA).\nBaseline household-level socio-economic indicators were collected using a structured questionnaire.\nAll variables representing asset ownership, household construction materials, source of fuel and light and the education level of the household head were examined individually before being combined using principal component analysis (PCA) to generate the socio-economic index of each household,, and are presented in here: (Additional file 1: Stata output showing Eigen scores of each variable used in calculation of socio economic status of households).\nThe households were grouped into quintiles of the socio-economic index generated and ranked from the poorest to the least poor.\nThis data was cross tabulated with treatment group using Pearson\u2019s chi-square (\u03c72) to assess whether there was a significant difference in the socio-economic status of the households in the two treatment groups (not accounting for the clustered design due to the exploratory nature of this analysis.\nThe number of treatment tubes issued was analysed by linear regression against month, treatment and an interaction of month and treatment to determine if there was a significant difference in the number of tubes issued in each month and per treatment group.\nClinical data\nClinical data was adjusted for covariates identified a priori to be confounders and analysed using the according-to-protocol approach, where person-time at risk was excluded when a participant reported or was observed to be non-compliant to the lotion (placebo or repellent) and for those with malaria for three weeks after they were diagnosed.\nThe total number of cases in each treatment group was divided by the sum of person years at risk to give the incidence rates in person years at risk.\nRate ratio and rate differences were then estimated.\nFor comparison, a secondary analysis using the intention-to-treat approach, where malaria incidence rates in the clusters were compared using all person-time at risk regardless of whether they complied with the study protocol but also adjusted for covariates identified a priori as confounders.\nSuch an approach would be expected to underestimate the treatment effect.\nIt was not possible to effectively blind the PI and PL as they had carried out both the semi field and field efficacy evaluations of these treatments and could identify the intervention and placebo.\nThe clinical data was, therefore, re-blinded by an independent statistician (ET), who was not aware of the intervention and placebo codes.\nPerson-time at risk estimation for according-to-protocol analysis\nThe study was conducted for 14\u00a0months from July 2009 to August 2010.\nTo calculate the person-time at risk, a closed cohort was assumed, so that the number of household members above six months recorded at baseline for each household was assumed to be constant throughout the study period.\nMonitoring surveys were conducted for each month of the study to establish compliance.\nIn a case where all individuals were susceptible to malaria infection and complied with the study protocol by applying the treatment issued on a nightly basis, each individual in the household was assumed to contribute one-person month at risk to the study.\nIn a case where the household head or an able household member was not available to take the monthly monitoring surveys, it was assumed that all members of that household did not comply with lotion (repellent or placebo) use for that month and one-person month at risk for each member of that household was excluded from the person time at risk of the study.\nIn a case where a household member contracted malaria, that individual was excluded from calculation of person time at risk for three weeks.\nPerson time at risk of each household was estimated according to one of the following three possible scenarios:\nPerson-time for all household members was calculated according to the appropriate scenario above.\nMalaria incidence rates and regression analysis of the intervention effect\nUsing data on the total number of confirmed malaria cases and person-time for each household, we used a two-stage approach to estimate intervention effects (recommended by Hayes et al. for studies with fewer than 15 clusters/group).\nIn the first stage, cluster-specific incidence rates were calculated using random effects Poisson regression modelling with adjustment for confounding variables.\nSpecifically, the outcome of total number of confirmed cases of malaria/household was regressed on the set of confounding variables (age categories of the household, education of the household head, and quintile of SES), with an offset for person-time at risk per household and a random intercept for cluster to account for the clustered study design.\nAs per Hayes et al., treatment was not included as a factor in the model.\nIn the second-stage, residuals, calculated from the regression model were aggregated by clusters.\nThe covariate-adjusted treatment effect was then estimated by comparing the residuals in the intervention relative to the control group using the Wilcoxon rank sum test, because the data were not normal.\nKnowledge attitude and practice (KAP) of community members in relation to malaria and repellent\nBaseline data on knowledge of malaria and malaria prevention practices and knowledge and practice in relation to repellents were analysed using descriptive statistics in STATA 11.2 to assess whether there was an imbalance between the treatment arms.\nData that recorded attitude with regards to repellents, perceived effectiveness and willingness to continue use and pay were also analysed and these results are presented elsewhere [Sangoro O, Sarah M, Ann HK, Sarah M: Feasibility of repellent use in a context of increasing outdoor transmission: A Qualitative study in rural Tanzania, submitted to Malaria Journal for publication].\nEthical and safety considerations\nDuring recruitment, the household head was asked for written informed consent for themselves and all household members.\nIf consent was obtained, all members of the household were recruited into the study.\nStudy participants were free to withdraw from the trial at any time.\nAll households in the village were issued with an LLIN for every sleeping space to ensure equity.\nAll individuals from the study village were allowed free consultation, treatment and drugs (ALu) from the village dispensary at project cost.\nParticipant confidentiality was maintained by using generated unique identifiers instead of individual names during analysis.\nParticipants were educated on correct repellent use and application.\nChildren under 6\u00a0months were excluded from the trial.\nAn illustrated label giving instructions in the native language (Swahili) on safe repellent use was provided on each tube.\nDEET repellent used in this study has undergone extensive toxicological tests and has been endorsed as safe for human use.\nThe concentration of DEET (15%), used in this trial was approved by the Tanzanian Pesticides Research Institute, the Tanzanian Bureau of Standards and is available in Tanzanian shops.\nGuardians to children\u2009<\u2009six months were reminded to put their children under an LLIN early to prevent them contracting malaria.\nA clinical officer (CO) was employed at the village dispensary by the project to perform RDTs and to investigate and treat any adverse effects arising from repellent use.\nEthical approval for the study was obtained from Ifakara Health Institute (IHI) (IHRDC IRB A46), Tanzanian National Institute of Medical Research (NIMR/HQ/R8a/VOL IX/780) and the London School of Hygiene and Tropical Medicine Ethical Review Board (LSHTM ERB 5174).\nIHI provided study monitoring.\nResults\nTrial profile and baseline data\nThe trial profile is summarized in Figure\u00a02.\nIn the intervention group 2,224 individuals were enrolled and 2,202 in the placebo group.\nLoss-to-follow up was higher in the placebo group: n\u2009=\u200934 versus n\u2009=\u200916, and no individuals withdrew from the trial.\nSimilar numbers of person-years were analysed: 1952.81 in the intervention group and 1972.38 in the control group of the trial.\nBaseline household level socio-economic data on education and gender of household head, age-groups of all study participants, household construction material, source of cooking fuel and lighting and asset ownership were examined individually and are presented in Table\u00a01.\nThe gender of the household heads was comparable between the two treatment groups, with 55.33% (n\u2009=\u2009514) females and 44.67 (n\u2009=\u2009415) males.\nMost of the household heads had received some form of formal education, 82.81% (n\u2009=\u2009702) while only 17.18% (n\u2009=\u2009161) had no formal education.\nOf all participants recruited in the study, 17.55% (n\u2009=\u2009771) were children under five years of age, 34.37% (n\u2009=\u20091,510) were between five to 18\u00a0years of age and 48.08% (n\u2009=\u20092,112) were above 18\u00a0years of age and age-category distribution was similar in the two treatment groups.\nThe predominant source of energy used by the households was wood fire, 89.96% (n\u2009=\u2009883), while the predominant source of lighting used was the traditional lamp, 93.76% (n\u2009=\u2009871).\nAssessment of household construction materials demonstrated that most households in the study area had floors made from mud, 82.78% (n\u2009=\u2009769), while tin and thatch were used equally as roofing materials, 49.35% (n\u2009=\u2009457).\nAlso, most households in the study area had walls made from bricks, 79.87% (n\u2009=\u2009742).\nSocio-economic indices generated from PCA suggested an imbalance between the two treatment groups, with the control group demonstrating a higher SES than the intervention group, (Pearson\u2019s \u03c72\u2009=\u200917.5519, p\u2009=\u20090.002), (Table\u00a02).\nThe use of repellents as a mosquito control tool was low in the study area, with only 1% (n\u2009=\u20096) of those interviewed reporting to have ever used repellents.\nResults on KAP of repellents are presented in details elsewhere [Sangoro O, Sarah M, Ann HK, Sarah M: Feasibility of repellent use in a context of increasing outdoor transmission: A Qualitative study in rural Tanzania, submitted to Malaria Journal for publication].\nThe average number of tubes issued per household was 6.73 (95% C.I. 6.51 \u2013 6.95) and 6.92 (95% C.I. 6.68 \u2013 7.16) in the intervention and control group respectively and there was no significant difference per treatment group, 1.68 (95% C.I. 0.32 \u2013 84.25, P\u2009=\u20090.803) from linear regression analysis.\nLikewise there was no significant difference on the number of treatment tubes issued per month throughout the study period.\nClinical outcomes\nAccording-to-protocol analysis\nWhen data was analysed as per protocol there was a non-significant difference in cluster and household malaria incidence rates among repellent users and non-users (Table\u00a03).\nIn the cluster-level analysis (data averaged over cluster specific rates), the malaria incidence rates differed by 11.48%; with 68.29 (95% C.I. 37.05-99.53) cases/ 1,000 person-years in the control group and 60.45 (95% C.I 48.30-72.60) cases/1,000 person-years (95% C.I. 44.55 \u2013 81.73) in intervention group, (Wilcoxon rank sum z\u2009=\u20090.529, p =0.5967).\nFor household-level malaria incidence rates (data averaged separately over household specific rates), the incidence rates differed by 28.88%: with 84.54 (95% C.I. 61.04-108.05), cases/1,000 person-years in the control group and 60.12 (95% C.I. 45.08-75.15) cases/1,000 person-years in the intervention group, (Wilcoxon rank sum z\u2009=\u2009-1.267, p\u2009=\u20090.2051).\nThese result should however be interpreted with caution as there is still an ongoing debate on whether it is correct to estimate incidence rate ratios using regression models on less than 10 clusters.\nCluster aggregated rates were reported because it measured the overall effect of the intervention at the population level and this was the major objective of the study.\nAge was a significant risk factor with risk decreasing with increase in age.\nSES did not influence the risk of malaria in the model.\nIntention-to-treat analysis\nCluster-level analysis of malaria rates in the two treatment arms demonstrated a non-significant, 14.62% difference in malaria rates with 53.21 cases/1,000 person-years (95% C.I. 30.98 \u2013 104.16) in the control group and 45.43 cases/1,000 person-years (95% C.I 36.02 \u2013 59.79) in the intervention group, (Wilcoxon rank sum z\u2009=\u20090.227, p\u2009=\u20090.8206), (Table\u00a03).\nHousehold-level analysis of malaria incidence rates demonstrated a 30.71% difference in malaria incidence rates, with 68.21 cases/1,000 person-years (95% C.I. 49.59 to 86.84) in the control group and 47.26 cases/1,000 person-years (95% C.I. 35.49 \u2013 59.04), in the intervention group, (Wilcoxon rank sum z\u2009=\u2009- 1.268, p\u2009=\u20090.2047).\nAge was a significant risk factor: malaria risk decreased with increase in age although SES did not influence the risk of malaria in the model.\nDiscussion\nThis randomized controlled trial demonstrated that 15% DEET topical repellents have no effect on malaria incidence transmitted in the early evening.\nAlthough there was a consistent decrease in malaria risk among repellent users in both the cluster and household malaria rates, as seen from the results above, this reduction was not significant.\nThis finding is consistent with a study carried out in southern Lao PDR using an identical 15% DEET repellent.\nIt should be noted that, findings from other studies using a higher concentration of 20% DEET with Permethrin in soap that gave over 12\u00a0hours of complete protection from mosquito bites and Para-menthane 3\u20138 diol repellents with close to 100% efficacy for over six hours did demonstrate a significant protective effect in Pakistan, Bolivia and Ghana and this could be one of the potential explanations for the observation of a treatment effect in these studies.\nIt can be argued that in the Lao-PDR study, 15% DEET provided\u2009~\u2009100% protection against mosquito bites.\nHowever, the number of major malaria vectors, Anopheles minimus and Anopheles maculatus, caught in entomological collections in the Lao-PDR study was very low and that the effect observed, was probably that of 15% DEET against Stegomyia and Culex mosquitoes which made up the bulk of the collections.\nTherefore, as Anophelines are known to show less response to repellents compared to Stegomyia and Culex mosquitoes, the repellent effect observed in the Lao-PDR study was greater than at higher densities with a greater proportion of Anophelines as tested in Tanzania.\nPower\nThere are several factors that are likely to have masked any treatment effect in this study, the most likely being the lack of power to discriminate a statistically significant difference between study arms.\nThe lack of power in the study was likely caused by four factors:\nFirst, rapid scale-up of LLINs to achieve universal coverage has been actively taking place in Tanzania.\nThis had led to a substantial decline in malaria in the country and by extension the study area.\nAs a result, the incidence of malaria in the village was likely lower than the incidence assumed for calculation of sample size for this study.\nThis likely led to an underestimation of the sample size required to observe a difference between the two treatment groups.\nSecondly, during the study period, Tanzania experienced a drought that likely further reduced malaria transmission, and as a result, there were too few malaria episodes in the study area to accurately discriminate any reduction in malaria attributable to the repellent, highlighting the need for such studies to be carried out for more than one transmission season to avoid such problems.\nThird, most of the participants recruited in to the study come from a farming community.\nTherefore, during the planting and harvesting seasons, these participants relocated to their farmhouses.\nAs a result it was difficult to establish compliance during these periods and those participants were excluded from the study.\nThis lowered the study sample size further and with it the power to detect a treatment effect.\nLastly was the likely overestimation of the assumed malaria incidence in the study area that was used for sample size calculations.\nMalaria incidence in this study was estimated from reported fever rates in children less than 5\u00a0years of age in the study area.\nTherefore, even though scale up of LLINs and the drought experienced during the study might have lowered the malaria incidence in the study area, it is also likely malaria rates used for estimation of sample size might have been overestimated and hence undermined the study power to observe a difference between the treatment groups.\nCompliance\nCompliance in this study was measured by self-reporting of use every evening by the household head or a household member that was able to engage with the field workers during the monitoring surveys.\nHowever self-reporting is an unreliable measure of compliance, as it have been shown to overestimate compliance.\nAs a result, the ATP analysis used to measure malaria incidence is likely to underestimate the actual malaria incidence in the intervention and control arms, as a larger value of person-time will be used than that of individuals that actually complied to the study reducing discriminatory power.\nHowever, if the randomisation between the two treatment groups was done correctly then the overestimation of compliance and its resultant effect of the study outcome, is likely to be similar in both treatment groups, ruling out the likelihood of overestimation of the treatment effect.\nThis underlines the importance of correctly estimating the compliance in studies of personal protection in order to avoid confounding the outcomes of such studies.\nActive versus passive case detection\nDue to logistical reasons, this study recruited a single government health facility for collection of clinical data by passive case detection.\nAs a result, the study is likely to have lost malaria cases to the other health facility present in the area.\nAnecdotally, some participants complained that they went to the other health facility because the study facility always told them that they did not have malaria even though they knew they had malaria, so they did not trust the diagnosis.\nAlso some individuals might have opted to use traditional medicine, treat diseases at home or buy drugs directly from the numerous drug stores in the study area if they felt sick.\nAll these are potential malaria cases that the study might have lost, lowering both the sample size and estimates of malaria incidence in the area.\nIt would have been advantageous to collect data from both health facilities or carry out active case detection.\nSince malaria was still most common in children under five years in the study site as seen elsewhere, targeted active case detection in under fives may have gathered more reliable and realistic data on the true impact of repellents in this scenario.\nPerforming supplementary testing of blood spots from all participants attending the health facility with polymerase chain reaction (PCR) diagnosis of subclinical malaria parasitaemia may have also yielded more accurate estimation of transmission prevention by repellents.\nSources of bias\nBias was introduced into the study by an imbalance in socio-economic status between the two study groups.\nThe control group demonstrated a higher socio economic status than the control arm.\nThis study however, did not demonstrate a statistically significant association between SES and malaria incidence.\nHowever, it is well known that improved housing, whose representative covariates had been adjusted for during analysis, is protective against malaria.\nA plausible explanation for this is that the participants in this study came from a single village or from villages located closely together.\nAs result they were exposed to the same levels of malaria transmission regardless of their socio-economic status.\nAs socio-economic status is positively associated with seeking treatment at a medical facility, it is likely that participants with higher SES sought treatment at the health facility in the study area at a higher rate compared to participants in the lower SES.\nTherefore as malaria data was only collected from a single health facility, it is likely that more cases of malaria were observed in participants with higher SES relative to participants from lower SES.\nAnother reason is that no association was seen may be because studies using material ownership as a proxy for measuring SES, to evaluate the relationship between SES and malaria incidence have yielded inconsistent results, at the household level.\nThe study participants were blinded up to some point after allocation of treatments, because of the identical packaging labelled with a three-digit code.\nHowever, after a while, field workers reported that study participants in the placebo group complained that they wanted to swap treatment.\nParticipants could differentiate the intervention from the placebo, as mosquitoes would still bite them after applying the \u2018treatment\u2019 while those in the treatment group bragged to their neighbours that they got the good lotion that was effective.\nThis is a source of bias and could have caused treatment contamination between clusters.\nThis problem would have been better overcome with clusters that were geographically isolated, for instance randomization on a village scale, so that individuals were less likely to be able to compare their treatment allocation.\nSome participants may have sold or given their repellent to relatives in other clusters.\nAnother potential confounder may have been diversion of mosquitoes from the intervention group to the placebo group.\nHowever, this was controlled by allowing for a buffer area of 200 metres between clusters.\nDiversion in repellent studies has usually been recorded over short distances, one metre.\nHowever, distances of 15\u201320 metres are recommended as the limit for short range attraction of host seeking mosquitoes and, therefore, distances of 200 metres between clusters were thought to be adequate to prevent diversion.\nTreatments were also issued at the household level to prevent intra and inter-household diversion within the cluster.\nIt has been later observed in the study area that mosquito diversion between households does occur and could have confounded data if compliance with the intervention was low by diverting mosquitoes from complying to non-complying households or individuals.\nThe community was highly knowledgeable about malaria transmission, prevention and control.\nThis is likely a result of the malaria intervention programmes that have taken place in the study village for over two decades.\nThe community awareness about topical repellents as a mosquito control tool was poor at the study inception.\nHowever, after the study, the community was highly aware of repellents and community members were willing to take up this intervention against malaria if available.\nThis finding demonstrates the feasibility of topical repellents as a potential tool to supplement LLINs to prevent early evening transmission.\nIn a separate study [Sangoro O, Sarah M, Ann HK, Sarah M: Feasibility of repellent use in a context of increasing outdoor transmission: A Qualitative study in rural Tanzania, submitted to Malaria Journal for publication], the community members reported bite avoidance as the major reason for using repellents in the early evenings.\nA posteriori analysis of data for children under six months was carried out to check whether this age group experienced high malaria transmission because of mosquitoes diverted to them as it was recommended that they not use the repellent.\nThis might also have affected the incidence of malaria in the treatment groups if there was uneven distribution of this age category between these groups.\nHowever, it was observed that there were only three children and a single case of malaria in this age category, and it can be confidently concluded that this age group did not have any influence on the outcomes observed.\nNet usage was also analysed to determine whether there was a difference between the two treatment groups, which would have confounded the outcome.\nIt was observed that reported net usage the previous night was 100% in both treatment groups.\nThese results are presented in detail elsewhere [Sangoro O, Sarah M, Ann HK, Sarah M: Feasibility of repellent use in a context of increasing outdoor transmission: A Qualitative study in rural Tanzania, submitted to Malaria Journal for publication].\nRecommendations\nIt was observed that estimation of a sample size with sufficient power was a major shortcoming of this study.\nTherefore, it is advisable to establish baseline disease incidence rates if a similar study is to be implemented in the future to avoid under powering the study.\nThis can be established from health facility records.\nHowever these records may not necessarily be accurate and the more appropriate measure may be to conduct a small cross-sectional or longitudinal survey of the community disease prevalence or incidence and then power accordingly.\nAnother important factor when testing personal protection tools is accurate establishment of compliance.\nBetter methods of establishing compliance are needed.\nThis can be done through frequent follow-up and spot checks or use of indirect methods, such as mosquito saliva antigens, that are a proxy of individual exposure to mosquito bites.\nAlso, development of new tools that require reduced compliance such as long lasting spatial repellents would likely offer greater protection because people often forget to comply daily with a topical repellent unless they feel mosquito bites.\nFinally, in a time when malaria is becoming more scant due to successful control, active case detection using RDT for clinical diagnosis followed up by PCR for malaria parasites is most likely the most appropriate means of measuring the impact of additional malaria control tools used in combination with LLINs.\nConclusion\nFindings of this trial could not demonstrate if 15% DEET topical repellents had any impact on incidence of malaria transmission in the early evening because the study lacked sufficient statistical power and had several important sources of bias.\nA better-designed study with sufficient power and fewer sources of bias and ideally a higher concentration of repellent is required to fully understand if topical mosquito repellents are a feasible malaria control tool in the early evenings in Eastern Africa, particularly as repellents have reduced malaria elsewhere in sub-Saharan Africa.\nThe acceptability of this intervention is an encouraging finding toward exploring supplementary malaria control tools.\n\nMap of households recruited into the trial in the study village.\n\n\nTrial Profile.\n\n\n\nBaseline household characteristics by treatment group\n\n | Intervention arm n (%) | Control arm n (%) | Totals n (%)\nNo. of households | 469 (50.05) | 468 (49.95) | 937 (100)\nNo. of participants | 2224 (50.05) | 2202 (49.95) | 4426 (100)\nGender of household head |  |  | \nMale | 215 (46.24) | 200 (43.10) | 415 (44.67)\nFemale | 250 (53.76) | 264 (56.90) | 514 (55.33)\nEducation of household head |  |  | \nNo education | 83 (17.74) | 78 (16.63) | 161 (17.18)\nEducated | 385 (82.26) | 391 (83.37) | 702 (82.82)\nAge group distribution of all participant/household |  |  | \nUnder 5\u2019s | 412 (18.50) | 359 (16.57) | 771 (17.55)\n5-18 years | 721 (32.38) | 789 (36.43) | 1510 (34.37)\nAbove 18\u00a0years | 1094 (49.12) | 1018 (47.00) | 2112 (48.08)\nSource of energy |  |  | \nWood fire | 431 (92.89) | 402 (86.83) | 883 (89.86)\nOther sources | 33 (7.11) | 61 (13.17) | 94 (10.14)\nSource of lighting |  |  | \nTraditional lamp | 445 (95.70) | 426 (91.81) | 871(93.76)\nOther source | 20 (4.30) | 38 (8.19) | 58 (6.24)\nFlooring material |  |  | \nMud | 404 (86.88) | 365 (78.66) | 769 (82.78)\nCement | 61 (13.12) | 99 (21.34) | 160 (17.22)\nRoofing materials |  |  | \nThatch | 256 (55.41) | 201 (43.32) | 457 (49.35)\nTin | 203 (43.94) | 254 (54.74) | 457 (49.35)\nOther | 3 (0.65) | 9 (1.94) | 12 (1.30)\nWall materials |  |  | \nMud | 121 (26.08) | 66 (14.19) | 187 (20.13)\nBricks | 343 (73.92) | 399 (85.81) | 742 (79.87)\nAssets ownership |  |  | \nMotorbike |  |  | \nYes | 72 (15.48) | 52 (11.18) | 124 (13.33)\nNo | 393 (84.52) | 413 (88.82) | 806 (86.67)\nBicycle |  |  | \nYes | 246 (52.90) | 198 (42.58) | 513 (55.16)\nNo | 219 (47.10) | 267 (57.42) | 417 (44.84)\nStove |  |  | \nYes | 344 (73.98) | 314 (67.53) | 658 (70.75)\nNo | 121 (26.02) | 151 (32.47) | 272 (29.25)\nMobile phone |  |  | \nYes | 197 (42.37) | 211 (45.38) | 408 (43.87)\nNo | 268 (57.63) | 254 (54.62) | 522 (56.13)\nRadio |  |  | \nYes | 140 (30.11) | 156 (33.55) | 296 (31.83)\nNo | 325 (69.89) | 309 (66.45) | 634 (68.17)\n\n\n\nRanking of households using Socio-economic scores generated for PCA analysis by treatment group\n\n | Intervention arm n (%) | Control arm n (%) | Total n (%) | Pearson\u2019s Chi2 | P value\nSES generated from PCA |  |  |  |  | \nPoorest | 39 (8.33) | 28 (5.97) | 67 (7.15) |  | \nPoor | 164 (35.04) | 121 (25.80) | 285 (30.42) | 17.5519 | 0.002\nMedian | 165 (35.26) | 174 (37.10) | 339 (36.18) |  | \nLess poor | 77 (16.45) | 107 (22.81) | 184 (19.64) |  | \nLeast poor | 23 (4.91) | 39 (8.32) | 62 (6.62) |  | \n\n\n\nEstimated incidence rates by treatment arm and estimated intervention effects\n\n | Intervention arm | Control arm | % Reduction in rates | Wilcoxon rank-sum on residuals (p-value)\nMalaria cases | 115 | 137 |  | \nATP analysis |  |  |  | \nIndividuals randomized | 2208 | 2168 |  | \nHouseholds randomized | 463 | 462\nTotal person-years | 1952.81 | 1972.38 |  | \nAverage Household rates/1000 person-years | 60.12 (95% C.I 45.08-75.15) | 84.54 `(95% C.I 61.04 108.05) | 24.42% | -1.267 (0.2051)\nS.D. | 164.42 | 257.07 |  | \nAverage cluster rates/1000 person-years | 60.45 (95% C.I 48.30 72.60) | 68.29 (95% C.I 37.05-99.53) | 8% | 0.529 (0.596)\nS.D. | 16.98 | 43.66 |  | \nITT analysis |  |  |  | \nIndividuals randomized | 2224 | 2202 |  | \nHouseholds randomized | 468 | 469\nTotal person-years | 2580.44 | 2554.92 |  | \nHousehold rates/1000 person-years | 47.26 (95% C.I. 35.49-59.04) | 68.21 (95% C.I. 49.59-86.84) | 20.95% | -1.268 (0.2047)\nS.D. | 129.60 | 205.23 |  | \nCluster rates/1000 person months | 45.43 (95% C.I 36.02\u201359.79) | 53.21 (95% C.I. 30.98\u2013104.16) | 7.78% | 0.227 (0.8206)\nS.D. | 11.32 | 34.90 |  | \n",
        "label": "high",
        "id": "task4_RLD_test_308"
    },
    {
        "paper_doi": "10.1371/journal.pmed.1001497",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 2986 children under 5, 12,454 people, 2163 householdsInclusion criteria: households were eligible if there was at least one child under 5, and they lived permanently in the study area\n\n\nInterventions: Sodium dichloroisocyanurate (NaDCC) disinfection tablets\n\n\nOutcomes: Longitudinal prevalence of diarrhoea among children under 5Diarrhoea among participants of all agesWeight-for-age z-score, school absenteeism, health care expenditures; adherence; water quality\n\n\nNotes: Location: informal settlements of Orissa, IndiaLength: 12 monthsPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "Sophie Boisson and colleagues conducted a double-blind, randomized placebo-controlled trial in Orissa, a state in southeast India, to evaluate the effect of household water treatment in preventing diarrheal illnesses in children aged under five years of age.\nPlease see later in the article for the Editors' Summary\nBackground\nBoiling, disinfecting, and filtering water within the home can improve the microbiological quality of drinking water among the hundreds of millions of people who rely on unsafe water supplies.\nHowever, the impact of these interventions on diarrhoea is unclear.\nMost studies using open trial designs have reported a protective effect on diarrhoea while blinded studies of household water treatment in low-income settings have found no such effect.\nHowever, none of those studies were powered to detect an impact among children under five and participants were followed-up over short periods of time.\nThe aim of this study was to measure the effect of in-home water disinfection on diarrhoea among children under five.\nMethods and Findings\nWe conducted a double-blind randomised controlled trial between November 2010 and December 2011.\nThe study included 2,163 households and 2,986 children under five in rural and urban communities of Orissa, India.\nThe intervention consisted of an intensive promotion campaign and free distribution of sodium dichloroisocyanurate (NaDCC) tablets during bi-monthly households visits.\nAn independent evaluation team visited households monthly for one year to collect health data and water samples.\nThe primary outcome was the longitudinal prevalence of diarrhoea (3-day point prevalence) among children aged under five.\nWeight-for-age was also measured at each visit to assess its potential as a proxy marker for diarrhoea.\nAdherence was monitored each month through caregiver's reports and the presence of residual free chlorine in the child's drinking water at the time of visit.\nOn 20% of the total household visits, children's drinking water was assayed for thermotolerant coliforms (TTC), an indicator of faecal contamination.\nThe primary analysis was on an intention-to-treat basis.\nBinomial regression with a log link function and robust standard errors was used to compare prevalence of diarrhoea between arms.\nWe used generalised estimating equations to account for clustering at the household level.\nThe impact of the intervention on weight-for-age z scores (WAZ) was analysed using random effect linear regression.\nOver the follow-up period, 84,391 child-days of observations were recorded, representing 88% of total possible child-days of observation.\nThe longitudinal prevalence of diarrhoea among intervention children was 1.69% compared to 1.74% among controls.\nAfter adjusting for clustering within household, the prevalence ratio of the intervention to control was 0.95 (95% CI 0.79\u20131.13).\nThe mean WAZ was similar among children of the intervention and control groups (\u22121.586 versus \u22121.589, respectively).\nAmong intervention households, 51% reported their child's drinking water to be treated with the tablets at the time of visit, though only 32% of water samples tested positive for residual chlorine.\nFaecal contamination of drinking water was lower among intervention households than controls (geometric mean TTC count of 50 [95% CI 44\u201357] per 100 ml compared to 122 [95% CI 107\u2013139] per 100 ml among controls [p<0.001] [n\u200a=\u200a4,546]).\nConclusions\nOur study was designed to overcome the shortcomings of previous double-blinded trials of household water treatment in low-income settings.\nThe sample size was larger, the follow-up period longer, both urban and rural populations were included, and adherence and water quality were monitored extensively over time.\nThese results provide no evidence that the intervention was protective against diarrhoea.\nLow compliance and modest reduction in water contamination may have contributed to the lack of effect.\nHowever, our findings are consistent with other blinded studies of similar interventions and raise additional questions about the actual health impact of household water treatment under these conditions.\nTrial Registration\nClinicalTrials.gov NCT01202383\nPlease see later in the article for the Editors' Summary\nEditors' Summary\nBackground\nMillennium Development Goal 7 calls for halving the proportion of the global population without sustainable access to safe drinking water between 1990 and 2015.\nAlthough this target was met in 2010, according to latest figures, 768 million people world-wide still rely on unimproved drinking water sources.\nAccess to clean drinking water is integral to good health and a key strategy in reducing diarrhoeal illness: Currently, 1.3 million children aged less than five years die of diarrhoeal illnesses every year with a sixth of such deaths occurring in one country\u2014India.\nAlthough India has recently made substantial progress in improving water supplies throughout the country, currently almost 90% of the rural population does not have a water connection to their house and drinking water supplies throughout the country are extensively contaminated with human waste.\nA strategy internationally referred to as Household Water Treatment and Safe Storage (HWTS), which involves people boiling, chlorinating, and filtering water at home, has been recommended by the World Health Organization and UNICEF to improve water quality at the point of delivery.\nWhy Was This Study Done?\nThe WHO and UNICEF strategy to promote HWTS is based on previous studies from low-income settings that found that such interventions could reduce diarrhoeal illnesses by between 30%\u201340%.\nHowever, these studies had several limitations including reporting bias, short follow up periods, and small sample sizes; and importantly, in blinded studies (in which both the study participants and researchers are unaware of which participants are receiving the intervention or the control) have found no evidence that HWTS is protective against diarrhoeal illnesses.\nSo the researchers conducted a blinded study (a double-blind, randomized placebo-controlled trial) in Orissa, a state in southeast India, to address those shortcomings and evaluate the effect of household water treatment in preventing diarrhoeal illnesses in children under five years of age.\nWhat Did the Researchers Do and Find?\nThe researchers conducted their study in 11 informal settlements (where the inhabitants do not benefit from public water or sewers) in the state's capital city and also in 20 rural villages.\n2,163 households were randomized to receive the intervention\u2014the promotion and free distribution of sodium dichloroisocyanurate (chlorine) disinfection tablets with instruction on how to use them\u2014or placebo tablets that were similar in appearance and had the same effervescent base as the chlorine tablets.\nTrained field workers visited households every month for 12 months (between December 2010 and December 2011) to record whether any child had experienced diarrhoea in the previous three days (as reported by the primary care giver).\nThe researchers tested compliance with the intervention by asking participants if they had treated the water and also by testing for chlorine in the water.\nUsing these methods, the researchers found that over the 12-month follow-up period, the longitudinal prevalence of diarrhoea among children in the intervention group was 1.69% compared to 1.74% in the control group, a non-significant finding (a finding that could have happened by chance).\nThere was also no difference in diarrhoea prevalence among other household members in the two groups and no difference in weight for age z scores (a measurement of growth) between children in the two groups.\nThe researchers also found that although just over half (51%) of households in the intervention group reported treating their water, on testing, only 32% of water samples tested positive for chlorine.\nFinally, the researchers found that water quality (as measured by thermotolerant coliforms, TTCs) was better in the intervention group than the control group.\nWhat Do These Findings Mean?\nThese findings suggest that treating water with chlorine tablets has no effect in reducing the prevalence of diarrhoea in both children aged under five years and in other household members in Orissa, India.\nHowever, poor compliance was a major issue with only a third of households in the intervention group confirmed as treating their water with chlorine tablets.\nFurthermore, these findings are limited in that the prevalence of diarrhoea was lower than expected, which may have also reduced the power of detecting a potential effect of the intervention.\nNevertheless, this study raises questions about the health impact of household water treatment and highlights the key challenge of poor compliance with public health interventions.\nAdditional Information\nPlease access these websites via the online version of this summary at http://dx.doi.org/10.1371/journal.pmed.1001497.\nThe website of the World Health Organization has a section dedicated to household water treatment and safe storage, including a network to promote the use of HWTS and a toolkit to measure HWTS\nThe Water Institute hosts the communications portal for the International Network on Household Water Treatment and Safe Storage\nIntroduction\nDiarrhoea is responsible for an estimated 1.3 million deaths among children under five each year, mostly in developing countries.\nWith over 287,000 deaths attributable to diarrhoeal diseases per year, India ranks first among countries contributing to this worldwide disease burden.\nIndia has made considerable progress in recent years in improving water supplies in both rural and urban settings.\nNevertheless, only 11% of the rural population is served by a household water connection.\nSurveys of microbial water quality throughout India have shown extensive faecal contamination of drinking water supplies.\nIn Hyderabad, for example, 50% of water samples drawn from pre-monsoon, monsoon, and post-monsoon period were positive for faecal coliforms.\nIn Madhya Pradesh, 33% of boreholes were faecally contaminated.\nEven water that is safe at the point of distribution is subject to frequent and substantial contamination during collection, transport, and storage.\nHousehold water treatment and safe storage (HWTS), including boiling, chlorinating, and filtering water at home, can improve water quality at the point of delivery and prevent post-collection contamination.\nSystematic reviews of water quality interventions have shown HWTS to be effective in improving drinking water quality and in preventing diarrhoea.\nBased on this evidence, the WHO and UNICEF recommend HWTS for populations relying on unsafe water supplies as part of a comprehensive strategy to prevent diarrhoeal disease, particularly among young children.\nThe evidence supporting a health impact from HWTS in low-income settings is from studies employing open trial designs.\nWhile open trials have found HWTS interventions to reduce diarrhoea within the range of 30%\u201340%, none of the blinded trials to date in low-income settings have found the intervention to be protective against diarrhoea.\nHowever, previous studies presented certain limitations including good ambient drinking water quality, short follow-up periods, and small sample sizes.\nThis disparity in results and the limited evidence from previous blinded trials has led researchers to call for more rigorous studies to quantify the contribution, if any, of HWTS in reducing diarrhoea.\nWe undertook this study to evaluate the effect of household water treatment in preventing diarrhoea prevalence among children under five.\nMethods\nEthics\nThe protocol was approved by the Ethics Committee of the London School of Hygiene and Tropical Medicine, and the Ethics Committee of the Indian Institute of Health Management Research.\nWritten consent was obtained from all heads of households after being given full details about the study.\nChildren reported to have diarrhoea at the time of visit were given oral rehydration sachets.\nWhen required, they were referred to the local community health worker or health centre.\nFollowing the conclusion of the study, results were shared with participating households.\nEach household was given a supply of oral rehydration solution and a 1-y supply of active tablets.\nStudy Setting\nThe study was conducted in Orissa, India among 11 informal settlements in the capital city of Bhubaneswar and 20 rural villages in the district of Dhenkanal.\nPartly because they are considered squatters on public land, residents of the informal settlements are provided no water, sewer, or other public services.\nDrinking water is procured mainly from hand-dug wells, many of which are open and unprotected or from boreholes and tap stand.\nDhenkanal district is located about 100 km northeast of Bhubaneswar and is inhabited mostly by agricultural labourers and workers at local steel plants.\nHouseholds rely on poorly protected open hand-dug wells, or from yard or public taps connected to a distribution system drawing from wells.\nOpen defecation is common among the study population.\nEligibility and Enrolment\nIn November 2010, a community meeting was held in each site to explain the objectives of the study.\nHouseholds were eligible to participate in the study if they had at least one child under 5 y of age and lived permanently in the selected area.\nEligibility was verified with children's immunization records.\nParticipating households were explicitly encouraged to continue their existing water treatment practices rather than rely on a tablet that may be a placebo.\nA baseline questionnaire was administered to each enrolled household to collect information on demographics, socio-economic characteristics, and water, hygiene and sanitation conditions and practices.\nIntervention\nThe intervention was implemented by Population Services International (PSI), a leading promoter of chlorine-based HWTS products worldwide.\nThe intervention consisted of the promotion and free distribution of sodium dichloroisocyanurate (NaDCC) disinfection tablets (Medentech, Ltd.).\nNaDCC tablets have long been used for the emergency treatment of water and more recently for the routine treatment of drinking water; they have been reported to offer some advantages over sodium hypochlorite in terms of safety, shelf life, up-front cost, and convenience.\nThis product was selected for the intervention because chlorine is widely used for water treatment and because a previous study reported that blinding of NaDCC tablets was feasible.\nPrior to the commencement of the study, extensive piloting was conducted to determine the optimal dosing of the tablets based on chlorine demand of the water in the study area.\nWhile most study households were using a 13 l aluminium container to store their drinking water, it was necessary to use a 67 mg NaDCC tablet (normally designed to treat 20 l of water) in order to achieve the WHO standard of a minimum of 0.2 ml/l of residual free chlorine (RFC) after 24 h.\nIn accordance with the manufacturer's instructions, households were advised to add the tablet to their water storage container, stir or agitate it, and wait for 30 min prior to consumption of the treated water.\nA team of 20 trained inter personal communicators (IPCs) employed by PSI visited each household fortnightly and gave them free of charge a box containing 30 tablets along with instructions on use.\nDuring each visit, IPCs also provided information on the adverse health effects associated with consuming contaminated drinking water and the importance of treating drinking water.\nGames and interactive pictures were used to engage household member in the discussions.\nCommunity-level activities such as street plays, game shows, wall paintings, and distribution of fliers, posters, and calendars were also conducted throughout the study.\nRandomisation and Blinding\nFollowing baseline, households were randomly assigned to one of the two study arms.\nRandomisation was stratified by community to ensure an equal number of intervention and control households in each of the 31 rural and urban communities.\nThe randomisation list was generated using Stata 10 and was conducted by a member of staff who was neither involved in the delivery of the intervention nor in the data collection.\nThe allocation sequence was concealed from the implementing team and the evaluation team.\nThe placebo tablet was similar in appearance to the active tablet.\nIt had the same effervescent base but did not contain the NaDCC disinfecting agent.\nThe active and placebo tablets were packaged in identical boxes of three strips containing ten tablets each.\nEach box was pre-labelled with the allocated household identification number prior to distribution.\nThe labeling of the boxes was conducted by members of staff who were neither involved in the implementation nor data collection or analysis.\nBoxes of active and placebo tablets were labelled with a distinct batch code generated by the manufacturer and only known to the member of staff responsible for supervising the labeling of the boxes with the household identification numbers.\nOutcome Assessment\nThe primary outcome was longitudinal prevalence (LP) of diarrhoea among children under 5 y of age.\nLongitudinal prevalence was recorded as the number of days with diarrhoea over the total number of days under observation.\nThis outcome was chosen because it is reported to be more closely associated with mortality.\nDiarrhoea among participants of all ages was recorded as a secondary outcome.\nTrained fieldworkers visited households every month for 12 mo (between late December 2010 and December 2011).\nAt each visit, daily point prevalence over the previous 3 d (today, yesterday, and the day before yesterday) was recorded for each child <5 y based on reports from the primary care giver.\nFollow-up resulted in 12 visits and 36 possible days of observation.\nDiarrhoea was defined using the WHO definition of three or more loose stools passed in one day.\nWeight-for-age z-score (WAZ), school absenteeism, and health care expenditure for diarrhoea were included as secondary outcomes.\nWAZ was measured as a potential proxy marker for recent diarrhoea.\nWAZ was measured among all children under five at baseline and each of the monthly visits.\nWeight was measured using a portable digital scale SECA 385 with an increment of 20 g for weight below 20 kg and 50 g for weight between 20\u201350 kg.\nField workers were trained and followed standardised procedures.\nTo monitor accuracy of the measurements, 10% of weight measurements were repeated over the course of the study.\nWAZ were calculated using WHO growth reference data.\nSchool absenteeism was assessed among primary school-aged children.\nInformation on school asbenteeism was collected through caregivers' reports, roll calls, and review of attendance records at the schools.\nAt each monthly visit, the mother or primary caregiver was asked if her child had been absent on any given day of school during the previous 5 d of school.\nRoll calls in the classroom were conducted in a total of nine visits over the course of the study.\nSchools to be visited for roll calls had to include at least ten children to be visited in order to facilitate logistics.\nSchool registers were reviewed for the entire follow-up period.\nAdherence\nAdherence to the intervention was assessed by asking householders if they had treated their children's drinking water and by testing it for RFC.\nIf the caregiver replied that the child's drinking water was treated, she was asked which method was used.\nThis open-ended question was designed to monitor not only reported use of tablets, but also of other water treatment methods.\nRFC concentrations were measured by colorimetric method using DPD1 reagent (Palintest Limited, Tyne & Wear) and a colour comparator.\nThe scale of the comparator allowed for readings by 0.1 mg/l from 0.1 to 1.0 plus eight readings between 0.5 and 6.0 mg/l.\nChlorine testing was done immediately after sample collection but off site and by a separate fieldworker to preserve blinding.\nHouseholders who reported their water to be treated with a tablet at the time of visit were defined as reported users while those with detectible RFC in the sample were defined as confirmed users.\nWater Quality\nEach month, 20% of households were randomly selected for testing of children's drinking water for the presence of thermotolerant coliforms (TTC), an indicator of faecal contamination.\nSamples were collected directly from the vessel containing water being consumed by the child <5 y using sterile 125 ml Whirl-Pak bags (Nasco) containing a tablet of sodium thiosulfate to neutralise any halogen disinfectant and were placed on ice for transport to the laboratory.\nSamples were processed within 4 h using the membrane filtration technique on a 0.45-micron membrane (Millipore Corporation), cultured on membrane lauryl sulphate medium (Oxoid Limited), and incubated at 44\u00b0C.\nFor quality control, 10% of samples were processed in duplicate and a negative control was processed with each batch.\nBlinding Assessment\nOn the last follow-up visit the mother or primary caregiver of the child was asked to guess whether they had been receiving the active tablet or placebo.\nSuccess of blinding was measured using both the James' and Bang's blinding indices.\nStatistical Analysis\nThe sample size was calculated assuming a 15% reduction in the longitudinal prevalence of diarrhoea among children of the intervention arm.\nWe assumed 80% power, significance level 2-sided \u03b1\u200a=\u200a0.05, a baseline diarrhoea prevalence of 4%, a standard deviation of 4.4%, a design effect of 1.1 for clustering of diarrhoea cases at the household level, and 15% loss-to follow-up.\nThe sample size was increased to account for intermittent sampling and a 3-d recall period.\nCalculations yielded a sample size of 1,500 children <5 y per arm.\nAssuming 1.5 children <5 y per household, the number of household to be recruited was 1,000 per arm.\nThe primary measure of efficacy of the intervention was estimated using a intention-to-treat analysis.\nBy intention-to-treat, we mean that participants were analysed in the treatment groups to which they were randomised, irrespective of the intervention they actually received.\nHowever, the analysis does not include missing outcome data resulting from lost-to-follow-up or participants who remained in the study until the end of the follow-up period, but who were absent on some of the visits.\nAll statistical analyses were conducted in Stata 10.\nWe estimated the prevalence ratio of diarrhoea using binomial regression with a log link function and robust standard errors.\nWe used generalised estimating equations (GEE) with an exchangeable correlation matrix to account for clustering at the household level.\nGEE has been suggested as an appropriate method to account for clustering since it relies on fewer assumptions than random effect models for binary outcomes.\nThe effect of the intervention on WAZ was assessed using random effect linear regression adjusting for baseline WAZ measurement.\nIn a deviation from the study protocol, we explored the within-child effect of diarrhoea on WAZ using fixed effect linear regression with number of days with diarrhoea in the previous 3-d window as a categorical variable.\nThe fixed effect model was deemed more appropriate to control for any confounding due to differences between individual children.\nStatistical analyses of microbiological data were conducted after log10 transformation of TTC counts to account for the skewed distribution.\nGeometric mean TTC counts were reported in each treatment arm.\nMeans of the log-transformed values were compared between groups using non-parametric tests.\nSubgroup analyses were conducted to assess the effect of the intervention among the treated.\nWe compared prevalence of diarrhoea and faecal contamination levels among self-reported users of the control and intervention groups.\nResults\nParticipants\nOverall, 2,163 households and 12,084 individuals were enrolled in the study.\nThis number includes 2,744 children under 5 y.\nAn additional 242 children were born during the 12-mo follow-up period, and 128 household members who were missed during the census at baseline were included from the first round of follow-up.\nA total of 46 individuals died over the course of the study, three were children <5 y (two due to accidents, one cause of death not specified) (Figure 1).\nOver the follow-up period, data were obtained on 84,391 d of observation for children <5 y and 389,825 d of observation for participants of all ages, representing 88% of total possible days of observation.\nObservations were missing due to household members moving out of the study area or being absent at the time of visit.\nIn per-visit analysis of missing data, including participants who remained in the study but missed visits, the proportion of missed visits was neither associated with treatment arm (p\u200a=\u200a0.94) nor water treatment practices at baseline (p\u200a=\u200a0.69) and age (p\u200a=\u200a0.83).\nHowever, it was higher among male than female (12.5% versus 11.2%, p<0.01) and among those who had diarrhoea at baseline (17.2% versus 11.5%, p<0.001).\nBaseline Characteristics\nBaseline characteristics were well balanced between groups (Table 1).\nOverall, 44% of households reported treating their drinking water.\nOf those, boiling was the most common method (68%) followed by straining water through a cloth (11%).\nWater was typically stored in wide neck aluminum containers (81%).\nHalf of household reported dipping a cup or using a long ladle to serve water from the container.\nDiarrhoea and Weight-for-Age\nOver the 12-mo follow-up period, the longitudinal prevalence of diarrhoea among children of the intervention group was 1.69% compared to 1.74% among the control group.\nAfter adjusting for clustering within the household, the longitudinal prevalence ratio was 0.95 (95% CI 0.79\u20131.13) (Table 2).\nAmong participants of all ages, the longitudinal prevalence ratio was 0.99 (95% CI 0.84\u20131.15).\nPrevalence of diarrhoea declined over time in both treatment arms ranging from 5.20% at baseline to 0.92% on the last round of follow-up among children under five (Figure 2) and from 1.90% to 0.31% among participants of all ages.\nWhen time was included in the model, we found no evidence of an interaction between time and treatment arm (p\u200a=\u200a0.67).\nThe effect of the intervention did not change over the course of the follow-up period.\nWe found a design effect of 1.1 for clustering of diarrhoea among children <5 y within the same household.\nEach additional day of diarrhoea resulted in a decrease of 0.0249 in WAZ (95% CI \u22120.0365 to \u22120.0133).\nHowever, the intervention had no effect on WAZ (mean WAZ of \u22121.589 among control children versus \u22121.586 among intervention).\nThe regression coefficient after adjusting for baseline WAZ was 0.0003 (\u22120.0347; 0.0354).\nAdherence\nChildren's drinking water was tested for RFC on 88% of 25,956 total possible household visits.\nWhile 51% of intervention households reported their child's drinking water to be treated with the tablets at the time of visit, only 32% of the water samples tested positive for RFC.\nReported use was higher among households who received the placebo (62%) compared to the active tablet (51%).\nOnly 2% of control group samples had RFC (Table 3).\nIn both groups, reported use of tablets increased over the 12-mo study period (Figure 3).\nPresence of RFC among intervention households also increased over time, ranging from 14% one month after the start of promotional efforts to 47% on the last month of follow-up.\nWhile the use of tablets increased over time, reported use of other water treatment methods declined.\nOverall, 20% of the 1,080 intervention households never had residual chlorine in their child's water during follow-up visits and 76% had chlorine on less than half of the total visits.\nOf the 3,863 samples that tested positive for chlorine, 70% (2,727) had RFC concentrations that fell within the CDC Safe Water Program recommended thresholds for acceptable taste and odour (\u22642.0 mg/l).\nHowever, 30% of the samples had concentrations above that level, including 11% (444) equal to or exceeding the WHO recommended maximum concentration of 5 mg/l.\nWater Quality\nA total of 4,546 samples of child's water were tested for thermotolerant coliforms (TTC/100 ml), representing 18% of the total number of household visits.\nFaecal contamination of drinking water was lower among intervention households than controls (geometric mean TTC count of 50 [95% CI 44\u201357] per 100 ml compared to 122 [95% CI 107\u2013139] per 100 ml among controls [p<0.001]).\nReported users in the intervention arm had lower TTC counts (geometric mean was 24, 95% CI 20\u201329) compared to those in the control arm (geometric mean 138, 95% CI 116\u2013162 [p<0.001]).\nOverall, 37% (417/1,126) of samples of reported users in the intervention arm had no detectible thermotolerant coliforms at the time of visit against 20% (268/1,340) in the control arm (Figure 4).\nOver the study period, a total of 621 water samples were processed in duplicate for quality controls.\nThe intraclass correlation coefficient calculated on the log-transformed values of TTC counts indicated good reliability between duplicate measurement, with an ICC of 0.82, 95% CI (0.69\u20130.96).\nDiarrhoea and Adherence\nThe prevalence of diarrhoea was significantly lower among children from families confirmed to be using the tablet than those who did not, irrespective of treatment arm (1.23% versus 1.78%, respectively) with a prevalence ratio of 0.72 (95% CI 0.57\u20130.91).\nHowever, this analysis may exaggerate the effect since users and non-users may differ with respect to certain characteristics affecting diarrhoea.\nThis is confirmed by the observation that even in the control arm, reported users had less diarrhoea than non-users.\nA comparison between users across treatment is likely to be less biased.\nWe estimated the average effect of treatment among reported users since laboratory confirmation of placebo tablet use was not possible to obtain from the control group.\nIn the intervention arm, reported users had residual chlorine in their water in 60% (3,440/5,784) of samples taken (versus only 3% [175/7,036] among reported non-users).\nLongitudinal prevalence of diarrhoea among children of households reporting use of the tablet in the intervention arm was the same compared to those of the control (1.47% versus 1.43%, respectively).\nThe risk ratio was 1.02 (95% CI 0.80\u20131.30).\nA similar pattern was observed among members of all ages (Table 2).\nDiarrhoea and Water Quality\nWe explored the relationship between risk of diarrhoea and levels of faecal contamination in the child's drinking water.\nChildren under five drinking water with TTC levels >1,000 per 100 ml did not have an increased risk of diarrhoea compared to TTC levels <1,000 TTC/100 ml (LPR 1.12, 95% CI 0.84\u20131.49).\nA similar result was observed among participants of all ages.\nSchool Absenteeism\nAt baseline, 1,059 children (5\u201310 y) were reported to attend primary school (grade 1 to 5) at 104 different schools.\nOverall, 34 schools had 10 or more pupils.\nOf those, it was possible to visit schools and conduct roll calls from 25 schools.\nSchool absenteeism information through roll calls and school registers was obtained for 611 students.\nOverall, 30% of primary care givers reported that their child missed at least 1 d of school during the previous 5 d of school (Table 4).\nOn roll-call days, 21% of children were absent at the time of the school visit; this estimate was higher than the 14% school absenteeism figure obtained from the school registers.\nPrevalence of school absenteeism assessed by caregiver's report and review of school records was similar between treatment groups.\nBlinding\nWhen asked to guess whether they had the active tablet or the placebo, 71% (669/936) of intervention households and 71% (684/961) of control households guessed they had the active tablet (Table 5).\nWhen the 25% who responded \u201cdon't know\u201d were encouraged to guess, the proportions were 90% (852/936) and 91% (864/961), respectively.\nThe blinding index calculated with James' method was 0.62 (0.61\u20130.63).\nThe blinding index calculated from Bang's method was 0.69 (0.66\u20130.72) among participants who received Aquatabs and \u22120.68 (\u22120.70 to \u22120.65) among those who received the placebo.\nAt the end of the study, when asked about taste and smell of their drinking water after adding the tablet, 51% (483) of the 947 intervention households interviewed reported that the smell of the water was worse than untreated water compared to 23% (225/961) of control households.\nIn addition, 22% (209) of intervention households complained about the taste of the water compared to 7% (72) among controls.\nDiscussion\nWe conducted a large double-blind randomised controlled trial to assess the impact of household water treatment on diarrhoea among children <5 y in rural and urban India.\nOur findings provide no evidence that the intervention was effective in preventing diarrhoea, either among children <5 y or among all members of the study population.\nNeither was there evidence of an impact of the intervention on WAZ.\nThe study was designed to address some of the limitations of previous double blinded trials.\nFirst, the study was powered to measure impact among children under five, who are most vulnerable to diarrhoea.\nThe sample population was ten times the size of previous blinded trials of household water treatment conducted in low-income settings.\nSecond, both urban and rural populations were included in order to increase generalisability of our findings.\nThird, the 1-y follow-up period was designed to account for seasonability and for reduced compliance over time as previously reported.\nFourth, we measured WAZ, a potential proxy for self-reported diarrhoea.\nLastly, adherence and water quality were also monitored extensively every month.\nOur findings are consistent with other blinded studies of water quality interventions that found no impact on diarrhoea.\nHowever, there are alternative explanations for the observed lack of impact.\nAdherence to the intervention was low.\nDespite free distribution of the tablets and an intensive promotion campaign, only a third of intervention households met the definition of confirmed users in any month during the follow-up period; three-quarters had chlorine on less than half of the total visits.\nSystematic reviews and modelling studies of water quality interventions have shown that the protective effect from HWTS interventions is reduced when adherence is low.\nHowever, other non-blinded studies of HWTS with comparable levels of adherence (around 30%) have nevertheless reported lower diarrhoea rates among intervention participants.\nAs with many other open trials reporting on subjective outcomes such as diarrhoea, those estimates may have been overestimated due to reporting bias.\nComparison was made between self-reported users of the intervention group with those of the control.\nWhile householders exaggerated reported use, confirmed users nevertheless accounted for most (60%) of reported users.\nAlthough reported users of the intervention group had significantly less contaminated water than those of the placebo group, they did not have a lower prevalence of diarrhoea.\nThis result speaks against low compliance being the only explanation for the lack of impact.\nThe low level of uptake was unanticipated.\nA 5-wk pilot of the same intervention among a comparable study population conducted immediately before the trial resulted in 68% compliance and greater than 2 log reduction in thermotolerant coliform counts.\nA number of chlorine-based interventions have achieved compliance in excess of 80% and a previous trial of NaDCC tablets in Bangladesh reported nearly full compliance.\nUptake did increase over time, and it is possible that the low uptake at the start was due in part to challenges in scaling up the promotional campaign.\nWhile the distribution of the tablets to all study households started the first month of follow-up, the community mobilisation activities were not fully rolled out until the second quarter of follow up.\nThese results highlight the challenges of acheiving high levels of uptake of the intervention despite an intensive campaign.\nEvaluations of other HWTS strategies in programmatic settings have also reported low levels of adoption, while research-driven studies have found higher levels of reported use.\nFurther research is needed to better understand how to achieve consistent and sustained adoption of these interventions on a programmatic basis and over the long term.\nAnother potential explanation for the lack of health effect was the comparatively modest improvement in water quality among intervention households.\nWith a mean of 50 TTC/100 ml, even water sampled from intervention households would be classified as \u201cmoderate risk\u201d using WHO nomenclature.\nOther studies of household water treatment have reported higher baseline levels of contamination and larger reduction in faecal contamination of drinking water from the intervention.\nAlthough we cannot rule out the possibility that this modest improvement in water quality reduced the potential for health impact among our study population, subgroup analysis found no evidence of a dose-response relationship between water quality and diarrhoea.\nOther studies have also reported weak associations between levels of indicators of faecal contamination and risk of diarrhoea.\nOur study had certain limitations.\nThe prevalence of diarrhoea among children <5 y was lower than expected.\nAs found in many other studies of water, sanitation, and hygiene (WASH) interventions, diarrhoea dropped significantly over time in both control and intervention groups.\nThe lower prevalence may have also reduced power for detecting a potential effect.\nThe amount of missing data was not unusual compared with similar trials.\nThe propotion of missing visits was similar between treatment arms, but certain groups (female and those reporting no diarrhoea at baseline) were over-represented.\nAdjusting for covariates predicting missingness in the model did not change the effect estimate.\nGiven the modest percentage of missing data in the study, any resulting bias is likely to be small.\nThe finding that children who experienced more diarrhoea in the previous 3 d had lower WAZ scores does provide support for this measure as a potential proxy marker for diarrhoea.\nMore objective indicators are needed, especially for environmental health interventions that are difficult or impossible to blind.\nImprovement in water quality alone may not be sufficient to prevent diarrhoea in settings with multiple sources of exposure to faecal pathogens.\nIn this population, sanitation coverage and practices of handwashing with soap were low, indicating that other transmission routes may have played a more important role.\nSystematic reviews have reported subtantial reductions in reported diarrhoea from HWTS interventions alone, often with no additive effect from multiple intervention strategies.\nHowever, mathematical models have suggested that the protective effect of water quality interventions against diarrheoa is largely influenced by the level of hygiene and sanitation in the community.\nIn conclusion, our study sought to measure the impact of household water treatment among children under five in low income settings in the absence of reporting bias.\nThe study was designed to address some of the shortcomings of previous trials.\nOur findings are consistent with other blinded studies of water quality interventions that found no impact on diarrhoea.\nBoth intention-to-treat analysis and analysis among reported users found no evidence of an impact on diarrhoea among children <5 y or all ages.\nAlthough we cannot rule out the possibility that this was due to low compliance and only a moderate impact of the intervention on water quality, our results raise additional questions about the protective effect of household water treatment under these conditions and underscore the need for promoters of household water treatment to demonstrate health impact in the absence of bias.\nTrial profile.\nPrevalence of diarrhoea among children <5 y over time.\nCompliance assessed by presence of residual chlorine in child's drinking water and self-reported use among intervention and control households over time.\nFaecal contamination levels in child's water samples by self-reported use (n = 4,344).\n\nBaseline characteristics of study households (n\u200a=\u200a2,163).\nCharacteristics | Control | Intervention\n | n | Percent | n | Percent\nDemographic and socio-economic |  |  |  | \nNumber of households | 1,083 | 50.1 | 1,080 | 49.9\nUrban | 340 | 31.4 | 338 | 31.3\nRural | 743 | 68.6 | 742 | 68.7\nMean (SD) number of persons per household | 5.5 (2.2) |  | 5.7 (2.3) | \nEducation head of household |  |  |  | \nIlliterate | 206 | 19.0 | 188 | 17.4\nLiterate no formal schooling | 69 | 6.4 | 90 | 8.3\nSome primary | 139 | 12.9 | 160 | 14.8\nCompleted primary | 160 | 14.8 | 146 | 13.5\nSome secondary | 387 | 35.8 | 400 | 37.0\nCompleted +2 y | 65 | 6.0 | 48 | 4.4\nCompleted +3 y (university) | 56 | 5.2 | 48 | 4.4\nMean (SD) number of rooms for sleeping | 1.7 (1.0) |  | 1.7 (1.0) | \nOwn |  |  |  | \nElectricity | 824 | 76.1 | 821 | 76.1\nTV | 554 | 51.2 | 549 | 50.9\nRefrigerator | 125 | 11.5 | 124 | 11.5\nBicycle | 762 | 70.4 | 748 | 69.3\nMotorbike | 250 | 23.1 | 247 | 22.9\nLand | 540 | 49.8 | 552 | 51.1\nLivestock | 452 | 41.8 | 464 | 43.0\nType of construction |  |  |  | \nPucca | 429 | 39.7 | 398 | 36.9\nSemi-pucca | 285 | 26.3 | 286 | 26.5\nKuchha | 268 | 34.0 | 396 | 36.7\nDrinking water source - rainy season |  |  |  | \nUnprotected dug well | 667 | 61.7 | 672 | 62.2\nTubewell | 187 | 17.3 | 175 | 16.2\nTap | 147 | 13.6 | 138 | 12.8\nSurface water | 53 | 4.9 | 58 | 5.4\nDrink from same water source during dry season | 946 | 87.3 | 951 | 88.1\nType of container |  |  |  | \nWide neck | 881 | 81.5 | 884 | 82.1\nNarrow neck | 36 | 3.3 | 28 | 2.6\nBoth types | 164 | 15.2 | 165 | 15.3\nServing child's water |  |  |  | \nDip cup | 542 | 50.4 | 508 | 47.3\nPour | 488 | 45.4 | 523 | 48.7\nUse tap | 34 | 3.2 | 28 | 2.6\nTreat water | 482 | 44.6 | 470 | 43.6\nTreatment method |  |  |  | \nBoil | 319 | 66.2 | 331 | 70.3\nStrain | 122 | 25.3 | 106 | 22.5\nChlorine | 11 | 2.3 | 9 | 1.9\nOther | 30 | 5.0 | 24 | 3.8\nToilet facilities | 422 | 39.0 | 410 | 38.0\nUse soap to wash hands | 372 | 34.4 | 330 | 30.6\nHand washing station | 659 | 60.9 | 658 | 60.9\nWater present | 596 | 90.4 | 595 | 90.6\nSoap present | 317 | 48.1 | 314 | 47.8\nBucket present | 522 | 79.5 | 533 | 81.4\nFaeces present in courtyard | 475 | 43.9 | 453 | 41.9\n\nSD, standard deviation.\n\nLongitudinal prevalence of diarrhoea among children under five and individuals of all ages by treatment arm and stratified by reported use.\nAnalysis | Control | Intervention | LPR Adjusteda\n | Days with Diarrhoea | Days of Observation | LP (%) | Days with Diarrhoea | Days of Observation | LP (%) | \nIntention-to-treat analysis |  |  |  |  |  |  | \n<5 | 733 | 42,060 | 1.74 | 715 | 42,331 | 1.69 | 0.95 (0.79\u20131.13)\nAll ages | 1,172 | 192,686 | 0.61 | 1,163 | 197,139 | 0.59 | 0.99 (0.84\u20131.15)\nSubgroup analysis stratified by reported use |  |  |  |  |  |  | \n<5 |  |  |  |  |  |  | \nUser | 360 | 25,157 | 1.43 | 310 | 21,122 | 1.47 | 1.02 (0.80\u20131.30)\nNon-user | 352 | 16,367 | 2.15 | 391 | 20,568 | 1.90 | 0.88 (0.70\u20131.12)\nAll ages |  |  |  |  |  |  | \nUser | 574 | 114,361 | 0.50 | 507 | 95,256 | 0.53 | 1.08 (0.88\u20131.32)\nNon-user | 551 | 72,028 | 0.76 | 630 | 95,249 | 0.66 | 0.90 (0.73\u20131.10)\n\nLPR Adjusted for clustering within household.\n\nAdherence measured by presence of residual free chlorine (n\u200a=\u200a22,804) and self-report (n\u200a=\u200a22,976).\nAdherence | Control | Intervention\n | n | Total | Percent | n | Total | Percent\nRFC | 223 | 11,407 | 2 | 3,630 | 11,397 | 32\nSelf-report | 7,071 | 11,485 | 62 | 5,829 | 11,491 | 51\n\n\nSchool absenteeism among school-aged children assessed via mother's report, classroom roll calls and school records.\nAbsenteeism | Control | Intervention | p-Value\n | n | Total | Percent | n | Total | Percent | \nReporteda | 1,412 | 4,641 | 30 | 1,406 | 4,600 | 31 | p\u200a=\u200a0.36\nRoll-callsb | 437 | 2,253 | 19 | 474 | 2,047 | 23 | p\u200a=\u200a0.02\nSchool recordsc | 6,896 | 48,014 | 14 | 6,253 | 43,932 | 14 | p\u200a=\u200a0.78\n\nReported: numbers and proportions of children who missed at least 1 d of school in the past 5 d of school (n\u200a=\u200a1,059 children 5\u201310 y enrolled in primary school standard 1\u20135 at baseline followed up for 12 mo).\nRoll call: number of days absent over total number of days of observation (n\u200a=\u200a611 children followed up for 9 mo).\nSchool records: number of days absent over total number of school days (n\u200a=\u200a611 children followed up for 12 mo).\n\nBlinding status of respondents by group assignment at the end of the study (n\u200a=\u200a1,897).\nGuess | Assignment\n | Placebo | Chlorine | Total\n | n | Percent | n | Percent | n | Percent\nChlorine | 684 | 71.2 | 669 | 71.5 | 1353 | 71.3\nPlacebo | 33 | 3.4 | 24 | 2.5 | 57 | 3.0\nDon't know | 244 | 25.4 | 243 | 26.0 | 487 | 25.7\nTotal | 961 | 100 | 936 | 100 | 1897 | 100\n",
        "label": "low",
        "id": "task4_RLD_test_644"
    },
    {
        "paper_doi": "10.1016/j.eclinm.2018.06.004",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Design: parallel group randomized trialDuration of study: July to September 2017Follow-up: 21-30 days\n\n\nParticipants: Country: TanzaniaSetting: schoolAge: 6-12 years (mean 10.1 years)Sex: 46% girlsNumber included in study: 186Inclusion criteria: 2 stool samples positive for hookworm eggs in the stool (>= 100 epg or >= 2 Kato-Katz thick smears slides with > 1 hookworm egg)Exclusion criteria: had menarche (for females); presence of severe anaemia (haemoglobin 8.0 g/dL considered severe anaemia); had any known or reported history of chronic illness such as cancer, diabetes, chronic heart, liver, or renal disease; received any recent anthelminthic treatment (within past 4 weeks); had any known allergy to mebendazole or albendazoleLost at follow-up: 1 (0.5%)Number positive for A lumbricoides: 98Number included in review: 98\n\n\nInterventions: Treatment strategy: screening and treat the positiveGroup 1: mebendazole 100 mg twice a day for 3 consecutive days + placebo (n = 47)Group 2: mebendazole 500 mg single dose + placebo for 3 consecutive days (n = 51)\n\n\nOutcomes: Outcomes included:Ascaris prevalence pre- and post-treatment, cure rates, pre- and post-treatment GM epg, ERR, adverse eventsOutcomes not included in review: anthelmintic efficacy for Trichuris and hookworm\n\n\nNotes: Diagnostic technique: Kato-KatzFunding support: PAT\n\n",
        "objective": "To compare the efficacy and safety of anthelmintic drugs (albendazole, mebendazole, ivermectin) for treating people with Ascaris infection.",
        "full_paper": "Background\nSingle-dose mebendazole is widely used in preventive chemotherapy against the soil-transmitted helminths (STHs) Ascaris lumbricoides, hookworm and Trichuris trichiura, yet it shows limited efficacy against hookworm and T. trichiura infections.\nThe use of adapted treatment regimens might provide a strategy to control and eliminate STH infections in STH-persistent settings.\nWe evaluated the safety and efficacy of the multiple dose mebendazole regimen (3\u202fdays 100\u202fmg bid) versus a single dose of 500\u202fmg mebendazole in a setting with high STH prevalence and high drug pressure.\nMethods\nThis randomised, double-blind clinical trial took place in a primary school on Pemba Island, Tanzania, in school-aged children (6\u201312\u202fyears).\nUsing a computer random number generator (block size 10), hookworm-positive children were randomly assigned (1:1) to either a single or multiple dose regimen of mebendazole by an independent statistician.\nTwo stool samples were collected at baseline and follow-up (18 to 22\u202fdays after treatment) for Kato-Katz analysis.\nThe primary outcome was cure rate (CR) against hookworm.\nSecondary outcomes were egg reduction rate (ERR) against hookworm, CRs and ERRs against A. lumbricoides and T. trichiura, and tolerability assessed 3, 24 and 48\u202fh post-treatment.\nParticipants, investigators, caregivers, outcome assessors and the trial statistician were blinded.\nThis trial is registered with ClinicalTrials.gov, number NCT03245398.\nFindings\n93 children were assigned to each treatment arm.\n185 children completed treatment and provided follow-up stool samples.\nCR against hookworm was significantly higher in the multiple dose (98%) than in the single dose arm (13%, OR 389.1, 95% CI 95.2 to 2885.7%, p\u202f<\u202f0.001). 34 and 42 children reported mild adverse events in the single and multiple dose arms, respectively.\nThe most common events were abdominal pain, headache and diarrhoea.\nInterpretation\nThe poor performance of single dose mebendazole against hookworm infections was confirmed, but the multiple dose treatment regimen of mebendazole showed high efficacy.\nHence, multiple dose mebendazole might provide a treatment strategy in given epidemiological situations to boost control and elimination of STH infections.\nFunding\nPATH.\nResearch in context\nEvidence before this study\nWe searched in PubMed for all articles published before June 1, 2017 which mentioned both \u201chookworm\u201d and \u201cmebendazole\u201d in the abstract, with no language restrictions.\nAlthough several studies have investigated the effect of either a single or a multiple dose of mebendazole, we only identified one open-label clinical trial, which compared the effect of both the single and the multiple dose mebendazole regimen 16 years ago, prior to commencement of large-scale administration of anthelminthic drugs.\nAdded value of this study\nThis is the first double-blind randomised clinical trial comparing the effect of a single dose (500 mg) to a multiple dose (100 mg twice a day during three consecutive days) of mebendazole against hookworm infections in Pemba, Tanzania, a setting with high drug pressure and persistent high hookworm prevalence.\nThe results of this study clearly showed that the multiple mebendazole dose is more effective than the single dose.\nBoth regimens were safe with only mild adverse events being reported.\nImplications of all the available evidence\nCurrently, the main control strategy against hookworm and other soil-transmitted helminths is preventive chemotherapy, which is based on the administration a single dose of either mebendazole or albendazole.\nOur study confirms that the curative effect of a single dose mebendazole is not sufficient for treating hookworm infections and that alternative, more effective treatments, as a multiple dose mebendazole regimens might be considered, in particular in persistent hotspot settings.\nAlt-text: Unlabelled Box\nIntroduction\nAn estimated 472 million people are infected with hookworms (Ancylostoma duodenale and Necator americanus).\nHookworm disease burden is mainly associated with anaemia, which can cause both physical and intellectual growth retardation among preschool- and school-aged children.\nIn 2016, 1.6 million DALYs were estimated to be caused by hookworm infections, leading to annual productivity losses of up to US$139 billion.\nCurrently, the control of hookworm and other soil-transmitted helminths is based on periodic deworming (so called preventive chemotherapy) of school-aged children and other high-risk groups by regularly administering a single dose of either albendazole (400\u202fmg) or mebendazole (500\u202fmg).\nBoth drugs are highly effective against Ascaris lumbricoides but show poor performance against Trichuris trichiura when administered as a single dose.\nMoreover, based on a recent systematic review and network meta-analysis, a single dose of albendazole shows acceptable efficacy against hookworm (CR\u202f=\u202f80%), while a single dose of mebendazole fails to clear most of these infections (CR\u202f=\u202f33%).\nThe main anthelmintic drugs available to the control programmes are, therefore, variably efficacious depending on the drug and parasite.\nAdditionally, there are worries that drug resistance will emerge, a problem commonly described in veterinary medicine.\nWith preventive chemotherapy being the predominant tool for helminthiasis control, it might not be surprising that soil-transmitted helminthiasis persists in many settings.\nAs an example, recent studies on Pemba Island reported that prevalence of hookworm continues to range at high levels from 36 to 97%, despite regular treatment of school-aged children.\nTherefore, additional strategies are required to control and eliminate soil-transmitted helminth infections, including access to improved water, sanitation and hygiene, as well as adapted treatment regimens, such as optimised dosing or combination chemotherapy which would improve drug therapy.\nA multiple dose (100\u202fmg twice per day over three consecutive days) treatment of mebendazole is recommended globally and in Tanzania for individual treatment.\nHowever, surprisingly, only a few small studies have evaluated the multiple dose regimen of mebendazole.\nMoreover, results obtained varied considerably with cure rates (CRs) ranging from 31 to 100% and, therefore, did not allow drawing final treatment recommendations.\nFinally, only a single study evaluated both treatment arms in an open label trial design, more than 15\u202fyears ago.\nNo high quality clinical trial conducted to date did a side-by-side comparison of multiple dose versus single dose treatment of mebendazole.\nThus, the present trial is, to our knowledge, the first randomised, double-blind trial comparing the efficacy and safety of a single dose (500\u202fmg) to a multiple dose (six doses of 100\u202fmg over three consecutive days twice per day) regimen of mebendazole against hookworm.\nMethods\nStudy Design and Participants\nThis randomised, double-blind clinical trial was conducted at Piki Primary School, on Pemba Island, Tanzania, from July 24 to September 15, 2017.\nPrior to the study initiation, ethical approval was obtained from the Zanzibar Medical Research and Ethical Committee (ZAMREC, reference number 0002/May/17) and from the Ethics Committee of Northern and Central Switzerland (EKNZ, reference number 2017-00320).\nThis trial is registered with ClinicalTrials.gov, number NCT03245398.\nBefore enrolment, all caregivers of children aged 6\u201312\u202fyears attending the primary school of Piki village were invited to information sessions at school during which the research staff explained the purpose and procedures of the study, as well as the benefits and potential risks of participating.\nCaregivers had the chance to clarify any questions they may had before they were asked whether they wanted their child to be included in the study or not.\nCaregivers who chose to allow the participation of their child were asked to sign a written informed consent.\nIlliterate caregivers provided a thumbprint while an impartial witness signed to verify that all information in the informed consent form was conveyed appropriately.\nConsenting children were eligible if they had provided two stool samples, were positive for hookworm eggs in the stool (\u2265\u00a0100 eggs per gram [EPG] or at least two Kato-Katz thick smears slides with more than one hookworm egg).\nAfter the initial clinical examination, children were excluded from the trial if any of the following exclusion criteria were present: had menarche (for females); presence of severe anaemia (haemoglobin <\u00a08.0\u202fg/dl was considered severe anaemia); had any known or reported history of chronic illness such as cancer, diabetes, chronic heart, liver or renal disease; received any recent anthelminthic treatment (within past 4\u202fweeks); had any known allergy to mebendazole or albendazole.\nRandomisation and Masking\nThe trial statistician (JH), who was not involved in any field work, provided a computer-generated stratified (by baseline infection intensities) block randomisation code (blocks of size ten).\nParticipants were allocated 1:1 to one of the two treatment arms: single dose (500\u202fmg) or multiple dose (100\u202fmg twice a day during three consecutive days) of mebendazole.\n500\u202fmg and 100\u202fmg mebendazole tablets were obtained from  (Zug, Switzerland).\nMatching placebos were manufactured at the  (100\u202fmg mebendazole matching placebo) or purchased at Fagron, Germany (500\u202fmg mebendazole matching placebo).\nTrial medications were prepared in advance in identical plastic envelopes labelled with the children's unique treatment identification numbers and sealed.\nThe treatment lasted 3\u202fdays and children received tablets at six different time points (mornings and evenings of each of the 3\u202fdays).\nAt the first time point, all participants received two tablets: either 500\u202fmg mebendazole plus 100\u202fmg mebendazole matching placebo, or 500\u202fmg mebendazole matching placebo and 100\u202fmg mebendazole; at the remaining five time points, children only received one tablet: either 100\u202fmg mebendazole or matching placebo, depending to which treatment arm they were allocated.\nParticipants, investigators, caregivers, outcome assessors and the trial statistician were blinded.\nAllocation was concealed: the envelopes containing the drugs were in bags of ten and, within each group of ten, envelopes were stacked on each other, preventing the administrator from seeing the next envelope.\nStudy Procedures\nThe name, age, sex and school grade of eligible children were recorded.\nChildren received containers labelled with their unique identification number (ID) and were asked to provide two stool samples, preferably on consecutive days.\nStool samples were transferred to the Public Health Laboratory Ivo de Carneri where duplicate Kato-Katz thick smear slides were prepared from each sample.\nSlides were examined by experienced laboratory technicians under a light microscope within 1\u202fh after slide preparation, to avoid clearing of hookworm eggs.\nHelminth eggs were enumerated and recorded for each parasite (hookworm, A. lumbricoides and T. trichiura) separately.\nTo ensure quality of hookworm diagnosis, 10% of the samples were divided into two stool containers; one of the containers kept its original participant ID, whereas the second container was labelled with a new ID (assigned by the co-investigator).\nDuplicate Kato-Katz were prepared from both containers and the findings compared.\nQuality control for A. lumbricoides and T. trichiura consisted of re-reading 10% of all slides.\nAny discrepancies between the original and the quality control read were discussed.\nAll children found positive for hookworm underwent a physical and clinical examination by a physician.\nHeight was measured with a standard meter (to the nearest 0.1\u202fcm), weight with an electronic balance (to the nearest 0.1\u202fkg), and temperature using an electronic ear thermometer (to the nearest 0.1\u202f\u00b0C).\nHaemoglobin levels were measured in capillary blood using the finger-prick method (HemoCue\u00ae 301).\nChildren were examined for any acute or chronic illness and questioned about their medical history.\nEligible children were enrolled (by MSP) for treatment which took place at Piki Primary school during the 3\u202fdays following the physical and clinical examination.\nChildren received treatment with clean water and a package of biscuits.\nAfter every morning treatment children were monitored for 3\u202fh and then actively questioned for adverse events by the study physician and nurses using a questionnaire.\nThe same procedure took place 24 and 48\u202fh after every morning treatment.\nAt follow-up, between 18 to 22\u202fdays after treatment, another two stool samples were collected from each treated child.\nParticipants who remained infected with hookworm, A. lumbricoides and/or T. trichiura at the end of the study were treated with albendazole (400\u202fmg).\nSimilarly, children who did not fulfil the eligibility criteria but were found infected with at least one of the parasites during screening were treated with albendazole.\nOutcomes\nThe primary outcome of this study was the CR against hookworm 20\u202fdays (\u00b1\u00a03\u202fdays) after treatment using the Kato-Katz thick smear method.\nCR was defined as percentage of hookworm-positive children being negative at follow-up.\nSecondary outcomes were (i) egg reduction rate (ERR) against hookworm, (ii) CR and ERR against A. lumbricoides and T. trichiura, and (iii) tolerability (number of adverse events) assessed 3, 24 and 48\u202fh post-treatment.\nAn additional secondary outcome, which will not be addressed in the current manuscript, was caregiver's knowledge related to the clinical trial after attending an information session.\nThis outcome was assessed using a short questionnaire.\nSample Size\nA CR of 20% was assumed for the single dose mebendazole and a CR of 40% was assumed for the multiple dose treatment regimen.\nTo detect a difference with 80% power at a two-sided 5% significance level, a minimum of 79 participants per study arm was required.\nAccounting for a loss to follow-up of 12%, we obtained a final sample size of 180 participants (90 per arm).\nStatistical Analysis\nData were double entered by two staff members into a database (Access 2003, Microsoft) and crosschecked using the Data Compare tool of EpiInfo version 3.3.2.\nAny discrepancies between both data entries were resolved by consulting the original records.\nStatistical analyses were performed using STATA 14.0 (StataCorp) and R 3.4.3 (R Development Core Team).\nOnly children who submitted samples before and after treatment were included in the available case analysis, which followed intention to treat principles.\nOdds ratios (OR) were calculated using unadjusted (primary analysis) and adjusted logistic regression (adjustment for age, sex, weight and strata).\nFor ERR calculation, the mean egg count of the four Kato-Katz thick smears was multiplied by a factor of 24 to calculate the EPG.\nERR was defined as the percentage of mean reduction at follow-up compared to baseline.\nERR was calculated using both the geometric mean (GM) and the arithmetic mean (AM).\nConfidence intervals for ERRs were calculated using the bootstrap resampling method with 5000 replicates.\nThe significance level was set at p-value\u202f\u2264\u202f0.05.\nInfection intensities (light, moderate or heavy) were determined according to WHO cut-offs.\nRole of Funding Source\nThe funder of the study had input into the study design, but no role in data collection, data analysis, data interpretation, or writing of the report.\nThe corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.\nResults\n364 consenting participants were screened for hookworm.\n354 children provided two baseline stool samples.\nOf these, 206 were found hookworm positive and were invited for the clinical and physical examination.\nTwo children were excluded because they had haemoglobin levels below 8.0\u202fg/dl and 18 were absent from school on the clinical and physical examination day.\nThe remaining 186 children were present for treatment and 93 participants were randomly allocated to each treatment arm.\nOne child from the single dose of mebendazole arm was lost to follow-up because he/she was not on Pemba Island during the follow-up period.\nWith the exception of 11 children, who provided the second baseline stool sample between six and 19\u202fdays after the first sample, both stool samples from all other children were collected within a five-day window.\nAll remaining 185 participants who provided follow-up stool samples were included in the available case analysis (Fig. 1).\nAt the second treatment time point, two children's envelopes were swapped.\nTherefore, by the end of the six-treatment time points, one child had erroneously received 100\u202fmg of mebendazole in addition to the single dose of 500\u202fmg of mebendazole plus placebo and the other child received only five doses instead of six doses of 100\u202fmg of mebendazole plus 100\u202fmg of placebo.\nThese two subjects were included in the available case analysis but excluded from the per protocol analysis (n\u202f=\u202f183) (Appendices 1 and 2).\nAppendix 3 presents the results of the analysis using the intention-to-treat population (n\u202f=\u202f186).\nThe treatment arms were balanced according to age, sex, weight, height, and hookworm baseline infection intensity (Table 1).\nDuring the physical examination, three children were found to have Tinea capitis, one child reported having asthma and one sickle cell anaemia.\nThese children were not excluded.\nAt baseline, among the 354 children who provided two stool samples, 94.3% of children were infected with at least one soil-transmitted helminth and 29.4% were co-infected with all three soil-transmitted helminths.\nThe prevalences of hookworm, A. lumbricoides and T. trichiura were 58.2, 36.7 and 92.6%, respectively.\nIn terms of intensity of infections, 4% of hookworm, 31% of T. trichiura and 45% of A. lumbricoides infections were moderate or heavy.\nThe CR of the multiple dose of mebendazole against hookworm was significantly higher (CR\u202f=\u202f97.9%) than that of the single dose (CR\u202f=\u202f13.0%, Odds ratio [OR] 303.3, 95% confidence interval [CI] 81.6 to 1999.4, p\u202f<\u202f0.001).\nSuperiority was confirmed by the adjusted logistic regression model (CR\u202f=\u202f13.0% versus CR\u202f=\u202f97.8%, OR 389.1, 95% CI 95.2 to 2885.7%, p\u202f<\u202f0.001).\nIn terms of ERR, the multiple dose (GM ERR\u202f=\u202f100.0%) was also significantly more effective than the single dose (GM ERR\u202f=\u202f68.0%) (Difference\u202f=\u202f\u2212\u00a00.32, 95% CI \u2212\u00a00.46 to \u2212\u00a00.22) (Table 2).\nERRs obtained with the arithmetic mean were 99.8% versus 52.7%.\n42.9% of children were cured against T. trichiura following six doses of mebendazole, compared to 6.8% in the single dose arm; this difference was statistically significant (OR 42.9, 95% CI 4.3 to 28.5, p\u202f<\u202f0.001 with the unadjusted model; OR 13.4, 95% CI 5.4 to 39.6, p\u202f<\u202f0.001 with the adjusted model).\nGM ERRs against T. trichiura were 71.7% (95% CI 56.7\u201378.5) in the single and 98.1% (95% CI 96.8\u201398.7) in the multiple dose arm.\nThe corresponding ERRs based on AM were 49.1% and 91.6%, respectively.\nThe single dose of mebendazole cured all children with an A. lumbricoides infection and the multiple dose cured all but one child (Table 2).\nIn both arms the ERRs against A. lumbricoides were >\u00a099.9%.\nTable 3 shows that among the 65 children with moderate T. trichiura infections at baseline, 10 were cured (all in the multiple dose arm), 43 turned from moderate into light infections (19 in the single dose arm and 24 in the multiple dose arm), and 11 children remained at moderate infection intensity (10 in the single and one in the multiple dose arm).\nAt the clinical examination, which took place right before treatment, a total of 20 children (11.3%) reported symptoms; 10 in the single dose arm and 10 in the multiple mebendazole dose arm (Appendix 4).\nThe number of adverse events and children reporting adverse events stratified by treatment arm and evaluation time point are summarised in Table 4.\nChildren in the multiple dose treatment arm reported slightly more adverse events than those in the single dose arm.\nIn total, throughout all adverse event assessment time points, 34 children (37%) in the single treatment arm reported symptoms and in the multiple arm 42 children (45%) reported symptoms after treatment (Appendix 4).\nThe most commonly reported adverse events were abdominal pain (69 reports), headache (46 reports) and diarrhoea (17 reports) during all treatment points.\nAll events were mild.\nA visual examination of the number of children reporting each type of adverse event throughout the whole treatment is available in Fig.\n2.\nDiscussion\nPreventive chemotherapy is the mainstay of helminthiases control, since it remains among the most cost-effective global public health control measures.\nAlbendazole and mebendazole, which are variably efficacious against the different soil-transmitted helminths, are the most widely used drugs.\nWe were interested in learning whether treatment efficacy could be improved by an adapted treatment regimen in a setting such as Pemba Island which, even though community members have been receiving treatment once or twice a year for 25\u202fyears, is still characterised by intense helminth transmission and persisting high prevalence rates.\nUsing a double-blind trial design we evaluated the multiple dose (3\u202fday, 6 dose course) treatment of mebendazole, which is recommended globally and in Tanzania for individual treatment, versus the single dose regimen widely used for population-based treatment.\nWe found clear evidence that the multiple, six-dose treatment schedule of mebendazole is significantly more effective at curing hookworm infections than a single dose of mebendazole.\nIn our trial, only 13% of children were cured after a 500\u202fmg single dose of mebendazole.\nOn the other hand, the multiple dose regimen of mebendazole cured almost all hookworm-infected children (CR\u202f=\u202f98%) which is in agreement with exploratory studies in the early 1970s.\nTo our knowledge, only four RCTs, conducted in Iran, Thailand, Brazil and Papua New Guinea, assessed the effect of the multiple dose mebendazole on hookworm infections.\nIn these studies, CRs ranged from 35 to 94%.\nAlthough these studies reported different baseline infection intensities, there seems to be no correlation between the intensity of infection and CRs.\nConcerns have been raised that mebendazole resistance had developed in the setting of Pemba since reduced efficacy of this drug was observed.\nIn more detail, treatment of hookworm infections resulted in significantly lower cure (7.6%) and ERR (52.1%) in 2003 than the ones reported before the beginning of periodic chemotherapy (CR\u202f=\u202f22.4%, ERR\u202f=\u202f82.4%).\nHowever, our findings clearly demonstrate that with using the recommended treatment regimen (which does not include dose intensification or dose density of chemotherapy, strategies commonly employed for example in the treatment of cancer) for mebendazole, high efficacy against hookworm can still be obtained and that speculations on drug resistance of mebendazole against hookworm should therefore be considered with caution.\nOn the other hand, it is interesting to note that the above mentioned strategies of dose intensification did not result in higher efficacy of mebendazole against soil-transmitted helminth infections.\nFor example, RCTs evaluating the effect of 500\u202fmg of mebendazole daily for 3\u202fdays (1500\u202fmg in total) found CRs ranging of 26 to 59%.\nThus, the higher dose does not seem to be the most important determinant driving the drug's effect.\nOverall, the hookworm CR for single dose mebendazole we observed on Pemba Island (13%) is in line with results from RCTs conducted in the same setting, however is considerably lower than a CR of 33% calculated from 13 trials in a recent meta-analysis.\nThis discrepancy between our results and the summary estimate by means of meta-analysis could be due to several factors such as the diagnostic method used, the parasite strain, or the study location: three of the four RCTs which reported CRs below 20% following a single dose of mebendazole against hookworm took place on Pemba Island.\nAnother influencing factor could be study quality and the sample size: the two RCTs reporting highest CRs using the single dose had small sample sizes of 35 and 45 participants in the single dose mebendazole arms.\nOur study showed that the multiple dose of mebendazole was also considerably more effective against T. trichiura infections than the single 500\u202fmg dose.\nNot only CRs were higher (CR\u202f=\u202f42.9% versus CR\u202f=\u202f6.8% respectively) but moderate infection intensities were less commonly observed in the multiple dose arm.\nOur results for the multiple dose regimen against T. trichiura are in line with previous studies conducted in Papua New Guinea and Brazil (CR\u202f=\u202f65% and CR\u202f=\u202f39%).\nHowever, similarly to what we found for the efficacy on hookworm, our results for the single dose were markedly lower than summary estimates reported by Moser and colleagues (CR\u202f=\u202f42%).\nUnlike the other two parasites, a multiple dose of mebendazole does not present an advantage against A. lumbricoides infections over a single mebendazole dose.\nOverall, both mebendazole treatments were well tolerated.\nInterestingly, available data on adverse events following multiple mebendazole doses is sparse.\nNone of the three RCTs which tested the efficacy of the multiple dose mebendazole regimen documented information on adverse events.\nHeadache, abdominal pain or diarrhoea were most commonly reported in both treatment arms, which is in line with previous studies exploring the efficacy and safety of 500\u202fmg mebendazole.\nIn the current study, we found a sudden increase in the number of adverse events 48\u202fh after the last treatment which is rather unexpected due to the short half-life of mebendazole (in the range of 2\u20139\u202fh).\nFor comparison, Speich et al. assessed adverse events at 3, 24, 27 and 48\u202fh post-treatment following a single dose of mebendazole and found the peak number of symptoms at 24\u202fh.\nIt is important to highlight that the same trend was observed in both treatment arms (hence 48\u202fh and 96\u202fh after the last active treatment), which suggests that this finding might not be triggered by the treatment but rather by differences in the reporting.\nOne limitation of this study is that the Kato-Katz technique has low sensitivity, particularly for light infections.\nAs an effort to overcome this limitation, two stool samples were collected on different days and two slides were prepared from each sample.\nHowever, this may still result in an overestimation of CRs as light infections at follow-up might have been missed and falsely considered as cured.\nAdditionally, the collection of two follow-up samples from 11 children was spaced by more than 5\u202fdays.\nAlthough there is no literature on the issue of how many days between sample collection are recommendable, this could had some impact on the outcome.\nIn conclusion, our study has shown that the multiple dose regimen of mebendazole is unarguably more effective against hookworm and concomitant T. trichiura infections.\nThe findings of our study add to recent results demonstrating that adapted treatments, including improved dosages, regimen or drug combinations considerably increase and broaden the spectrum of activity against soil-transmitted helminth infections.\nA multiple dose regimen clearly involves more resources than the administration of a single dose or drug combinations and comes with additional logistic challenges.\nHowever, in hotspot settings such as Pemba Island where the prevalences of hookworm and T. trichiura are so high despite decades of treatment, this could be a strategy to consider.\nIn the framework of preventive chemotherapy, drugs are distributed by non-medical personnel (such as teachers, volunteers or community drug distributors) in non-medical settings such as schools.\nThus, although more challenging, teachers could provide six doses instead of a single dose of mebendazole to each school-child.\nImproved treatments would trigger a considerable decrease of infections which would lead to a reduction of reservoirs that sustain reinfections in the population.\nIn parallel, efforts should continue to discover and develop novel drugs and vaccines, which n the long-term would aid in the elimination of these diseases.\nContributors\nMSP, Said MA, Shaali MA, JH, and JK planned and designed the study; MSP, Said MA, Shaali MA, and JK conducted the study; MSP, JH and JK analysed and interpreted the trial data; MSP and JK wrote the first draft and JH revised the manuscript.\nAll authors read and approved the final version of the manuscript.\nTrial profile.Fig. 1\nSpider plot indicating the percentage of observed clinical symptoms (before treatment) and adverse events in both treatment arms (throughout all seven adverse event assessment time points).Fig. 2\n\nBaseline characteristics of all randomised children stratified by treatment arm. Data are n (%), median (IQR), mean (SD). BMI\u202f=\u202fbody-mass index, EPG\u202f=\u202feggs per gram of stool.\nTable 1\n\n | Single dose (n\u202f=\u202f93) | Multiple dose (n\u202f=\u202f93)\nMean age (years) | 10.1 (1.6) | 10.1 (1.6)\nGirls | 39 (42%) | 46 (50%)\nMean weight (kg) | 26.7 (5.3) | 26.2 (5.1)\nMean height (cm) | 132.0 (10.2) | 131.8 (9.6)\nMean weight-for-age Z-score | \u2212\u00a01.4 (1.2) | \u2212\u00a01.3 (0.8)\nMean height-for-age Z-score | \u2212\u00a01.1 (1.1) | \u2212\u00a01.1 (0.7)\nMean BMI-for-age Z-score | \u2212\u00a01.1 (0.9) | \u2212\u00a01.2 (0.8)\nHookworm |  | \n\u00a0Infected children | 93 (100%) | 93 (100%)\n\u00a0Median | 222 (78-534) | 222 (96-606)\n\u00a0EPG geometric mean | 219.0 | 234.2\n\u00a0Infection intensity |  | \n\u00a0Light (1\u20131999 EPG) | 89 (96%) | 90 (97%)\n\u00a0Moderate (2000\u20133999 EPG) | 3 (3%) | 3 (3%)\n\u00a0Heavy (\u2265\u00a04000 EPG) | 1 (2%) | 0\nTrichuris trichiura |  | \n\u00a0Infected children | 88 (96%) | 91 (98%)\n\u00a0EPG geometric mean | 661.8 | 725.7\n\u00a0Infection intensity |  | \n\u00a0Light (1\u2013999 EPG) | 59 (66%) | 56 (62%)\n\u00a0Moderate (1000\u20139999 EPG) | 30 (34%) | 35 (38%)\n\u00a0Heavy (\u2265\u00a010,000 EPG) | 0 | 0\nAscaris lumbricoides |  | \n\u00a0Infected children | 47 (51%) | 51 (55%)\n\u00a0EPG geometric mean | 2691.2 | 4095.9\n\u00a0Infection intensity |  | \n\u00a0Light (1\u20134999 EPG) | 28 (60%) | 20 (39%)\n\u00a0Moderate (5000\u201349,999 EPG) | 15 (32%) | 28 (55%)\n\u00a0Heavy (\u2265\u00a050,000 EPG) | 4 (8%) | 3 (6%)\n\n\nCure rates (CRs) and egg reduction rates (ERRs) against hookworm, Ascaris lumbricoides and Trichuris trichiura after the administration of a single or multiple doses of mebendazole. CR\u202f=\u202fcure rate, CI\u202f=\u202fconfidence interval, EPG\u202f=\u202feggs per gram of stool, ERR\u202f=\u202fegg reduction rate.\nTable 2\n\n | Single dose | Multiple dose\nHookworm |  | \n\u00a0Children positive before treatment | 92 | 93\n\u00a0Children cured after treatment | 12 | 91\n\u00a0CR (95% CI) | 13.0% (6.9\u201321.7) | 97.9% (92.4\u201399.7)\n\u00a0EPG geometric mean |  | \n\u00a0Before treatment | 220.2 | 234.9\n\u00a0After treatment | 70.5 | 0.1\n\u00a0ERR (95% CI) | 68.0% (51.5\u201378.6) | 100% (99.9\u2013100)\n\u00a0EPG arithmetic mean |  | \n\u00a0Before treatment | 442.6 | 465.3\n\u00a0After treatment | 209.3 | 1\n\u00a0ERR (95% CI) | 52.7% (40.3\u201363.6) | 99.8% (99.3\u2013100)\nTrichuris trichiura |  | \n\u00a0Children positive before treatment | 88 | 91\n\u00a0Children cured after treatment | 6 | 39\n\u00a0CR (95% CI) | 6.8% (4.6\u201317.8) | 42.9% (33.8\u201354.8)\n\u00a0EPG geometric mean |  | \n\u00a0Before treatment | 655.9 | 726.8\n\u00a0After treatment | 185.5 | 14.0\n\u00a0ERR (95% CI) | 71.7% (56.7\u201378.5) | 98.1% (96.8\u201398.7)\n\u00a0EPG arithmetic mean |  | \n\u00a0Before treatment | 1017.4 | 1263.8\n\u00a0After treatment | 517.6 | 105.8\n\u00a0ERR (95% CI) | 49.1% (31.7\u201361.0) | 91.6% (88.4\u201394.6)\nAscaris lumbricoides |  | \n\u00a0Children positive before treatment | 47 | 51\n\u00a0Children cured after treatment | 47 | 50\n\u00a0CR (95% CI) | 100.0% | 98.0% (94.2\u2013100)\n\u00a0EPG geometric mean |  | \n\u00a0Before treatment | 2698.5 | 4113\n\u00a0After treatment | 0 | 0.2\n\u00a0ERR (95% CI) | 100.0% | 100%\n\u00a0EPG arithmetic mean |  | \n\u00a0Before treatment | 14,597.5 | 14,859.9\n\u00a0After treatment | 0 | 130.9\n\u00a0ERR (95% CI) | 100.0% | 99.1% (96.9\u2013100)\n\n\nNumber of children with moderate T. trichiura infections at baseline which, post-treatment, were either cured, evolved into light or heavy infections, or remained with moderate infection intensity.\nTable 3\n\n\nNumber of symptoms reported during the clinical examination and number of children reporting the symptoms during the clinical examination; number of adverse events (AEs) reported by children and number of children reporting AEs at each of the AE assessment time points by treatment arm.\nTable 4\n\nTimepoint | Number of | Single dose (n\u202f=\u202f93) | Multiple dose (n\u202f=\u202f93) | Total\nClinical examination before treatment | Symptoms | 17 | 13 | 30\nChildren | 9 | 9 | 18\n3\u202fh after 1st morning treatment | AEs | 10 | 13 | 23\nChildren | 10 | 11 | 21\n24\u202fh after 1st morning treatment | AEs | 4 | 4 | 8\nChildren | 4 | 4 | 8\n3\u202fh after 2nd morning treatment | AEs | 12 | 12 | 24\nChildren | 6 | 10 | 16\n24\u202fh after 2nd morning treatment | AEs | 11 | 13 | 24\nChildren | 9 | 9 | 18\n3\u202fh after 3rd morning treatment | AEs | 7 | 12 | 19\nChildren | 5 | 10 | 15\n24\u202fh after 3rd morning treatment | AEs | 9 | 6 | 15\nChildren | 7 | 5 | 12\n48\u202fh after 3rd morning treatment | AEs | 21 | 28 | 49\nChildren | 17 | 22 | 39\nTotal during all AE assessment time points | AEs | 91 | 101 | 192\nChildren | 34 | 42 | 76\n",
        "label": "low",
        "id": "task4_RLD_test_971"
    },
    {
        "paper_doi": "10.2166/wh.2005.0016",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: RCT\n\n\nParticipants: Number: 50 HIV+ people, all over 30 yearsInclusion criteria: confirmed HIV+ status, uses tap water 75% of the time, no children residing in the home\n\n\nInterventions: Countertop water filtration device\n\n\nOutcomes: Episodes of \"highly credible gastrointestinal illness\"Diarrhoea episodes calculated\n\n\nNotes: Location: San Francisco, USALength: 12 monthsPublication status: journa\n\n",
        "objective": "To assess the effectiveness of interventions to improve water quality for preventing diarrhoea.",
        "full_paper": "A pilot randomized, controlled trial of an in-home drinking water intervention among HIV 1 persons\nA pilot randomized, controlled trial of an in-home drinking water intervention among HIV 1 persons\nAlthough immunocompromised persons may be at increased risk for gastrointestinal illnesses, no trials investigating drinking water treatment and gastrointestinal illness in such patients have been published.\nEarlier results from San Francisco suggested an association (OR 6.76)  between tap water and cryptosporidiosis among HIV \u00fe persons.\nThe authors conducted a randomized, triple-blinded intervention trial of home water treatment in San Francisco, California, from April 2000 to May 2001.\nFifty HIV-positive patients were randomized to externally identical active (N = 24) or sham (N = 26) treatment devices.\nThe active device contained a filter and UV light; the sham provided no treatment.\nForty-five (90%) of the participants completed the study and were successfully blinded.\nIllness was measured using 'highly credible gastrointestinal illness' (HCGI), a previously published measure.\nThere were 31 episodes of HCGI during 1,797 person-days in the sham group and 16 episodes during 1,478 person-days in the active group.\nThe adjusted relative risk was 3.34 (95% CI: 0.99-11.21) times greater in those with the sham device.\nThe magnitude of the point estimate of the risk, its consistency with recently published observational data, and its relevance for drinking water choices by immunocompromised individuals support the need for larger trials.\nINTRODUCTION\nA recent case control study in San Francisco reported an elevated risk (OR 6.76,95% confidence interval: 1.37,33.5) for cryptosporidiosis among HIV-positive persons consuming tap water (Aragon et al. 2003).\nNo randomized trials have been published evaluating the benefits, if any, of supplemental in-home drinking water treatment among HIV-positive persons.\nRandomized trials evaluating in-home drinking water treatment among immunocompetent persons have been published and reached conflicting results.\nFor example, two studies by Payment and colleagues in Canada suggested a significant reduction in gastrointestinal illness arising from the use of in-home drinking water treatment (Payment et al. 1991(Payment et al. , 1997)).\nA study by Hellard et al. (2001) in Australia and a study by our group in California (Colford et al. 2002), found no significant reduction in gastrointestinal illness from the use of in-home drinking water treatment devices.\nThese studies differed with respect to several important features.\nThe Payment studies were not blinded (i.e. individuals knew the group to which they were assigned in the trial) p Portions of this manuscript were presented at the USEPA National Science Forum on 6 May 2003 in Washington, DC. and the source water was challenged (subject to industrial and human contaminants).\nThe Hellard study was a blinded trial and was conducted in a system with a reportedly pristine surface water source.\nA blinded study in California was conducted in a system with a challenged surface water source (Colford et al. 2002).\nAnother blinded trial in Davenport, Iowa enrolled 1,296 participants and found no benefit to an in-home drinking water intervention (Colford et al. 2005).\nThe present study was designed as a pilot to apply the randomized trial design to the issue of tap water consumption and gastrointestinal illness in HIV-positive persons.\nOur principal objectives were: 1) to confirm that enrolment and participation rates among this population would be high; 2) to replicate our earlier results suggesting that blinding can be achieved in drinking water trials; and 3) to develop a preliminary estimate of the relative rates of gastrointestinal illness between groups of HIV-positive persons receiving tap water with or without supplementary in-home treatment.\nSuch an estimate, reliably obtained in a randomized trial, was felt to be necessary before the design and conduct of a large-scale trial in an immunocompromised population.\nMETHODS\nThe study and the informed consent process were reviewed,\nPreliminary work: cross-sectional survey\nIn preparation for the intervention trial, we conducted and published a cross-sectional survey to analyse the prevalence of gastrointestinal illness and drinking water patterns in our potential study population (Eisenberg et al. 2002).\nBetween October 1998 and January 2000, the survey was administered to 226 patients at the same Infectious Diseases Clinic at the San Francisco Veterans' Affairs Medical Center from which trial participants would later be recruited.\nFortyseven per cent of respondents reported diarrhoea in the 7 days prior to being surveyed.\nEighty-one per cent of respondents were unaware of CDC drinking water guidelines for HIV-infected individuals, though 34% reported being very concerned about the health effects of their drinking water.\nStudy area, water supply and water distribution system\nThe trial was performed primarily within the city of San All participants received a study binder containing contact information for study staff, instructions for use of the device and the collection of a stool sample, and the current CDC safe drinking water guidelines for immunocompromised persons regardless of their treatment assignment (CDC 1999a, b).\nWe believed that, given the existence of published CDC recommendations about drinking water safety for immunocompromised persons, it would not have been ethical to conduct the study without making participants aware of these guidelines.\nRandomization and (triple) blinding\nParticipants were randomized 50:50 to receive either an active or sham water treatment device in blocks of ten.\nRandom allocation within each block was accomplished using computer-generated random numbers.\nThe manufacturer provided a list of device serial numbers and their corresponding active/sham status to facilitate device assignment.\nAll study participants, the study investigators (including clinic personnel and those performing data analysis) and the device installer were blinded throughout the trial as to device assignment.\nActive and sham water treatment devices and installation\nThe device chosen for this study was a countertop unit custom manufactured for the study by Tri H2O (San Leandro, California) and based on their commercially available 'Ultimate II' water filtration device.\nA tamper-proof seal prevented the filter casing from being opened.\nWe chose an active device that selectively removes microorganisms from the water without affecting other water quality parameters that could lead to unblinding of participants.\nOur active device used a 1micron filter followed by ultraviolet radiation to maximize the microbiological disinfection and physical removal capabilities of the treatment device without significantly affecting the taste and odour of the treated water.\nThe specification of a 1-micron absolute filter was chosen in order to enable the device to remove Cryptosporidium oocysts, a waterborne pathogen of great concern for HIV \u00fe populations.\nThe ultraviolet lamp was designed to emit wavelength at 254 nm, the optimum for disinfection, and a total minimum dose of 26,000 mwatt-sec cm 22 .\nThis dosage inactivates 99.99% of bacteria and viruses and conforms to 'Class B' standards for ultraviolet treatment devices as specified by the National Sanitation Foundation (USEPA 1996).\nThe sham device consisted of an empty filter casing, and an ultraviolet lamp secured within a glass sleeve in order to block ultraviolet light, without unblinding the device by having significant weight disparities between the active and sham devices.\nFollowing consent by all household members, the study technician came to the participant's residence to install the device.\nThe device was attached to the main faucet used for accessing drinking water in the home using a connector hose and a diverter valve that allowed for water to either be directed to the device or into the sink.\nIf a device could not be adjusted or repaired without opening the casing of the device, the study technician was instructed to replace the device to ensure that he and the participants remained blinded as to device type.\nStatistical methods: Blinding index\nOne goal of the study was to examine the feasibility of blinding of participants in such a trial among HIV-positive persons.\nFor this goal, we used the 'blinding index' (BI) of James et al. (1996) in which scores above 0.5 are viewed as evidence of effective blinding.\nAt the end of every 2 weeks, participants answered questions on which device, active or sham, they believed was installed.\nColford et al. (2002) reported the use of this same index in an earlier drinking water trial.\nWith this index (analogous to the kappa statistic) a score of 0.0 suggests all participants accurately identified device assignment, a score of 0.5 suggests random guessing by participants, and a score of 1.0 suggests all participants guessed assignment incorrectly or answered 'don't know'.\nHealth outcomes\nParticipants recorded daily occurrences of diarrhoea, nausea, vomiting, abdominal cramps and fever in their health diaries.\nDiarrhoea was defined as the occurrence of two or more loose stools in one day.\nThe principal health outcome measured in the trial was episodes of 'highly credible gastrointestinal illness' (HCGI), a measure based on that reported in several prior drinking water intervention trials (Payment et al. 1991(Payment et al. , 1997;;Hellard et al. 2001;Colford et al. 2002).\nA new episode was defined as any of the following four conditions, preceded by at least 6 symptom-free days: 1) vomiting, 2) watery diarrhoea, 3) soft diarrhoea and abdominal cramps, or 4) nausea and abdominal cramps.\nDays with missing data were not counted as 'disease-free'.\nThe requirement for 6 diseasefree days was first used by others to increase the likelihood that separate episodes truly represented distinct infections (rather than a prolonged course of one infection) (Payment et al. 1991(Payment et al. , 1997;;Hellard et al. 2001).\nHCGI data were analysed using logistic regression with the outcome being either HCGI (1) or no-HCGI (0) for every day at risk (see above).\nPoisson regression provided a poor fit to the summary counts per subject, as HCGI rates varied widely between subjects in the same treatment group.\nTherefore, logistic regression with a generalized estimating equation (GEE) -robust variance estimation approach was used on the daily data.\nWhen calculating standard errors, this approach both adjusts for residual correlation of the repeated (daily) outcome measurements within a subject and allows for different underlying rates between subjects within treatment groups (Liang K 1986).\nThe attributable risk from drinking water was calculated as (OR 2 1)/(OR)\nwhere OR is the estimated odds ratios of HCGI in the sham group compared with that in the active group (Hennekens & Buring 1987).\nIn addition to simple bivariate analyses, we also examined whether the direct effect of the device differed by baseline gastrointestinal symptoms.\nIn addition to the primary health outcome (episodes of HCGI), we calculated the total days of HCGI experienced by each participant.\nThis measure is an attempt to quantify the total burden of gastrointestinal disease experienced by the two groups.\nFor example, although a prolonged episode of HCGI could last for many days, it would only be recorded as one episode in the primary analysis.\nWith respect to the principal analysis of the causal relationship between use of the water treatment device and HCGI, the analysis of episodes of HCGI, as stated above, was the a priori defined analysis.\nParticipant medical records were reviewed to obtain CD4 count count (a measure of the current immune status of an individual), viral load and current medications.\nWater consumption\nWater consumption was self-reported using questions inserted into the health diary at 2-week intervals.\nParticipants estimated (in numbers of 240 ml (eight ounce) glasses) their daily consumption of drinking water at home (separately through the study device and through all other sources at home) and outside the home.\nParticipants were provided with water bottles and encouraged to carry water from the home device for use when outside the home.\nMean water consumption was compared by study group using the twosample t-test.\nA second participant (assigned to sham group) expired with Pneumocystis carinii pneumonia.\nRESULTS\nRecruitment\nBaseline characteristics of participants and completeness of data collection (Table 1)\nForty-four (98%) of the 45 participants were HIV-positive males, reflecting the demographic composition of our clinic.\nThe median age was 51.9 years in the active and 52.1 in the sham group.\nRandomization appeared to successfully balance the baseline characteristics of the two groups with respect to age, race, education, income, CD4 count, viral load, HIV medication usage and water consumption patterns.\nRecent symptoms of gastrointestinal illness (e.g. cramps, diarrhoea, nausea, vomiting, fever), however, were 2 -3 times more common in the participants randomized to the active group ( p = 0.028).\nThe 45 participants completed health diaries with 4,682 days of total observation time (2,087 active and 2,595 sham).\nThis represents diary completion rates of 89.3% for the active group and 97.4% for the sham group.\nEffectiveness of blinding of participants (Table 2)\nResponses from the final health diary (week 16) were evaluated using the blinding index.\nThirty-nine (87%) of the 45 participants completed the week 16 health diary.\nThe most frequent guess about treatment assignment in both the active (59%) and sham (50%) groups was\nAnalysis of gastrointestinal illnesses\nParticipants randomized to the active device experienced 16 episodes of HCGI; those in the sham group experienced 31 episodes (Table 3).\nBecause of the baseline imbalance in the frequency of gastrointestinal symptoms, we examined the data to determine if there was any interaction present between the presence of GI symptoms at baseline and treatment group assignment.\nNo evidence of an interaction was found.\nBecause the presence of gastrointestinal symptoms at baseline was strongly predictive of HCGI during the trial and was not balanced in the two treatment groups, this factor did appear to confound the relationship between device assignment and the incidence of HCGI.\nBecause confounding factors not balanced at baseline by randomization should be adjusted in any analyses of data from randomized trials (Freidman et al. 1998), we adjusted for the presence of GI symptoms at baseline in our analysis.\nThe adjusted odds of disease in the sham group were 3.34 (95% CI: 0.99-11.21) times higher in the sham group than in the group receiving treated water.\nThe attributable risk associated with such an odds ratio would be 0.70 (95% CI: 0.00-0.91).\nIn addition to this adjustment for the difference in baseline symptoms, we also stratified the participants directly (Table 4) by the presence of these symptoms.\nThese stratified results were qualitatively consistent with those found in the multivariate model (Table 3) in that the rates of HCGI were higher in the sham group.\nBecause of the small sample sizes in these strata, formal statistical testing was not undertaken within stratified groups.\nParticipants randomized to the active device reported 253 days of HCGI; those in the sham group reported 322 days of HCGI.\nAdjusted for the baseline imbalance in GI symptoms, the odds ratio for the two groups with respect to days of HCGI was 2.27 (95% CI, 0.64, 8.01).\nFurther analysis (Table 5) did not suggest that the difference in HCGI between the two groups was caused by individuals experiencing numerous (i.e. .5) episodes during the study.\nWater consumption (exposure) patterns during the trial (Table 6)\nThere was no significant difference between the two groups with respect to water consumption patterns.\nThere were p A new episode of HCGI was defined as the presence of any of the four definitions of HCGI preceded by 6 HCGI-free days.\nThe difference in total episodes of HCGI was the principal a priori health outcome measure for the study.\n1 Because individual participants could report multiple symptoms of HCGI on the same day, the total episodes of HCGI (and total days of HCGI) are less than the sums of the individual definitions.\npp Adjusted for baseline differences in the presence of GI symptoms (diarrhoea, vomiting, nausea or cramps) in the prior 7 days using logistic regression with generalized estimating equations.\nBlinding index = 0.67 (95% CI, 0.53 -0.82).\ninsufficient data with which to evaluate the presence of any dose-response trend based on amount of drinking water consumed.\nDISCUSSION\nThis study is the first randomized controlled trial of a drinking water intervention among HIV-positive persons, a group potentially at risk both for increased susceptibility to waterborne infections as well as to increased clinical severity once infected (Gerba et al. 1996).\nOur findings suggest that a randomized controlled trial of an in-home drinking water intervention in HIV-positive persons is feasible with respect to recruitment and enrolment.\nAdditionally, the increased point estimate of risk of gastrointestinal illness (of borderline statistical significance) in the sham treatment group in this small trial, combined with a recent case control investigation suggesting an elevated risk (Aragon et al. 2003) raises issues about what recommendations should be given to HIV-positive individuals about tap water treatment.\nBlinding of participants\nThere are few published approaches to the measurement of blinding and we chose to evaluate blinding in a method suggested by James et al. (1996) which uses a summary statistic analogous to the kappa statistic and has been used in other studies (Noseworthy et al. 1994;James et al. 1996;Colford et al. 2002).\nInability to properly blind the participants in drinking water trials could lessen the credibility of reported results (Noseworthy et al. 1994). It\nis interesting to note that there is a tendency for participants (in this study and others) in both the active and sham groups to more frequently believe that they are in the active group than is true (Hellard et al. 2001;Colford et al. 2002).\nWe speculate that this may arise from a desire of participants for assignment to the active arm.\nGastrointestinal illness\nThere is no doubt, in light of the reports of numerous water risks, such as cancers arising from waterborne chemicals, in which the latency period could be years.\nOur study design was similar to that of the trial reported by Australian investigators (Hellard et al. 2001).\nLike the Australian study, our investigation was blinded and conducted in a municipality believed to have excellent source water.\nThis is in contrast to the Canadian studies (Payment et al. 1991(Payment et al. , 1997) ) which were not blinded and were conducted in the setting of challenged source water.\nBoth the Canadian and Australian studies were conducted among immunocompetent participants.\nWhether or not the differences between these studies and our trial are due to differences in the source water, the treatment device, the immune status of the participants, or differences in the distribution systems cannot be answered using the existing data.\nIt is important to note that both groups in our study (but not in the earlier studies) received a form of intervention: for the ethical reasons described in the Methods section, all participants in our trial (regardless of their randomization assignment) received counselling at the start of the trial about current federal recommendations for drinking water safety for immunocompromised persons (CDC 1999a, b).\nTheoretically, this could lead to an underestimation of the effect in our study.\nOur estimate of an attributable fraction of 0.70 (CI 0.00-0.91) of cases of HCGI attributable to tap water consumption is consistent with that estimated by Aragon Second, although a number of baseline characteristics were balanced between the two groups, suggesting proper randomization, there was a difference at baseline in the number of participants with recent gastrointestinal symptoms which is itself strongly associated with HCGI.\nBecause the presence of gastrointestinal symptoms was associated with both the exposure variable (the device assignment) and the outcome variable (HCGI) it confounded the crude relationship between device assignment and HCGI and we adjusted for this baseline GI illness and included it in our final model.\nOur principal health outcome, HCGI, has been used repeatedly in prior studies (Payment et al. 1991(Payment et al. , 1997;;Hellard et al. 2001;Colford et al. 2002).\nHowever, such use does not ensure validity and validation studies of the measure itself.\nSuch validation would require expensive, close observation of each participant's actual bowel habits compared with their reports of illness.\nUnless there was a systematic difference in reporting of HCGI, the use of randomization in the trial design should minimize the introduction of bias into the results.\nOne theoretical cause for a difference between the two groups would be degradation of the water by the sham device.\nWe conducted a limited water sampling programme and found no evidence of such degradation.\nA larger water sampling programme should be a part of future trials to confirm these findings.\nWe do not believe that a firm conclusion can be drawn from this trial about the risk of HCGI from the consumption of tap water among HIV \u00fe individuals.\nFuture studies must be larger to further reduce the potential for any chance baseline imbalance in important covariates.\nAn additional limitation of our study is that the  et al. 2003), and the potential public health impact all support the need for larger trials of optimal drinking water treatment for immunocompromised persons.\napproved and monitored throughout by four Institutional Review Boards at the University of California, Berkeley, the University of California, San Francisco, the Centers for Disease Control and Prevention (CDC) and the San Francisco Veterans' Affairs Research and Development Committee.\nFrancisco, California. Forty-nine (98%) of the 50 participants were residents of San Francisco and one participant was a resident of Daly City, California. San Francisco receives its water from the largest unfiltered water supply on the West Coast, the Hetch Hetchy Water and Power Project(SFPUC 2002). Although all of the water supply is chlorinated only a small proportion of the water supply is fully filtered. Consumers receive either filtered, primarily unfiltered (approximately 82% unfiltered) or a mixed water supply depending on their location. The Hetch Hetchy watershed is a 1188.8 sq km (459 square mile) area located in Yosemite National Park at the headwaters of the Tuolumne River. Although the consistently high quality of surface source water has resulted in filtration exemption status, Cryptosporidium has been found in low levels in both the source and treated water (SFPUC 2002). A more detailed water characterization is available at www. sfwater.org. This is one of several unfiltered city water supplies in the US (others include New York, Boston and Seattle). Recruitment, enrolment and compensation of participants Participants for this trial were recruited from patients enrolled in the Infectious Diseases Clinic at the San Francisco Veterans Affairs Medical Center from April through December 2000. The first device was installed in May 2000 and the last participant completed the trial in May 2001. Our proposed sample size was 64 participants, estimated to be sufficiently large to detect successful blinding (blinding index . 0.50, see Methods) (Colford et al. 2002). Initial contact was made by a research nurse or pharmacist during scheduled or drop-in clinic visits, from patient response to a mailed flyer, or by telephone. The inclusion criteria required that each participant: has a confirmed diagnosis of HIV; routinely (75% or more of the time) uses municipal (tap) water at home with neither home filtration devices nor bottled water; confirms that all household members were aware of the HIV status of the participant and were willing to give consent to have the study device installed; and has no children residing at home. Research staff reviewed the health diary with each participant. The first section consisted of a daily log of gastrointestinal symptoms. Two weeks of responses could be entered in each health diary. The second section contained questions on water consumption, blinding and potential risk factors for gastrointestinal illness. Participants were to complete the log every day and the second section at the end of the 2-week period. Participants received US$50 upon completion of the enrolment questionnaire and US$15 for each of the eight diaries submitted during the 16-week study.\n, enrolment, randomization and adverse events We began recruitment in April 2000. As shown in Figure 1, 339 potential participants were screened and 50 were enrolled and randomized (24 active, 26 sham). The principal reasons for non-eligibility were: residence outside of the study area (45%) or use of bottled water as a primary source of drinking water (11%). Five (10%, 3 active, 2 sham) of the 50 randomized participants dropped out of the study before any blinding or health data were collected. The remaining 45 participants are the source of data for all analyses. The first device was installed in May 2000 and the last participant completed the trial in May 2001. No adverse events were attributed to trial participation. One consented participant committed suicide before device installation.\nFigure 1 | HIVWET screening and enrolment flow diagram.\net al. | Gastroenteritis and drinking water in HIV \u00fe persons Journal of Water and Health | 3.2 | 2005\ngeneralizability of our findings to other municipalities (with differing water systems) or to other participants (with differing forms of immune compromise and demographic composition, including age) is unclear. Such risk estimation must await further research in those geographic and participant communities. Recruitment for such a study in other, younger or gender-balanced HIV participant groups could differ from that which we experienced.CONCLUSIONSOur findings suggest that it is feasible to conduct randomized controlled trials among HIV \u00fe persons to investigate the risk of gastrointestinal illness from the consumption of drinking water. The presence of very large numbers of immunocompromised persons in the United States implies that even a slight elevation of risk from infection due to waterborne pathogens would carry a significant public health impact(USEPA 2000). Despite the borderline statistical significance of the findings in this small trial, the magnitude of the relative risk (OR 3.34, 95% CI: 0.99-11.21), its consistency with recently published data on cryptosporidiosis and tap water in San Francisco(Aragon    \nTable 1 | Participant baseline characteristics (n = 45)    \n | Active | Sham\n | device | device\nCharacteristic | (n = 21) | (n = 24)\nAge (years) | n (%) | n (%)\n30 -39 | 2 (9.5) | 3 (12.5)\n40 -49 | 5 (23.8) | 6 (25)\n50 -59 | 10 (47.6) | 11 (45.8)\n60 -69 | 4 (19) | 3 (12.5)\n70 \u00fe | 0 (0) | 1 (4.2)\nGender |  | \nMale | 20 (95.2) | 24 (100)\nFemale | 1 (4.8) | 0 (0)\nRace |  | \nWhite | 12 (57.1) | 14 (58.3)\nAfrican -American | 5 (23.8) | 6 (25)\nLatino | 3 (14.3) | 2 (8.3)\nOther | 1 (4.8) | 1 (4.2)\nNot available | 0 (0) | 1 (4.2)\nHighest level of education |  | \n2 -3 years high school | 1 (4.8) | 2 (8.3)\nHigh school graduate | 6 (28.6) | 7 (29.2)\n1 -3 years college | 8 (38.1) | 7 (29.2)\nCollege graduate | 3 (14.3) | 5 (20.8)\n1 -2 years post-graduate | 3 (14.3) | 2 (8.3)\nNot available | 0 (0) | 1 (4.2)\nAnnual income |  | \n, US$20,000 | 13 (61.9) | 14 (58.3)\nUS$20,000 -30,000 | 1 (4.8) | 3 (12.5)\nUS$30,000 -40,000 | 2 (9.5) | 4 (16.7)\nTable 3 | Episodes p of highly credible gastrointestinal illness (HCGI) and days of illness\n | Active device (n = 21) | Sham device (n = 24) | Total (n = 45) | Odds ratio (adjusted) p p\nTotal episodes \u00fe of HCGI defined by: | 16 | 31 | 47 | 3.34 (0.99 -11.21)\nVomiting | 2 | 9 | 11 | \nWatery diarrhoea | 11 | 17 | 28 | \nSoft diarrhoea with abdominal cramps | 0 | 0 | 0 | \nNausea with abdominal cramps | 6 | 6 | 12 | \nTotal days at risk for HCGI episodes | 1,478 | 1,797 | 3,275 | \nTotal days of HCGI defined by: \u00fe | 253 | 322 | 575 | 2.27 (0.64, 8.02)\nVomiting | 15 | 22 | 37 | \nWatery diarrhoea | 234 | 284 | 518 | \nSoft diarrhoea with abdominal cramps | 0 | 2 | 2 | \nNausea with abdominal cramps | 30 | 35 | 65 | \nTotal days of observation | 2,087 | 2,595 | 4,682 | \nTable 2 | Final (week 16) device blinding questionnaire\nAll participants who completed blinding questionnaire at 16 weeks (n = 39) | \nGuess | Active device | Sham device | Total\nActive | 10 | 11 | 21\nSham | 1 | 2 | 3\nDon't know | 6 | 9 | 15\nTotal | 17 | 22 | 39\nTable 4 | Episodes p of highly credible gastrointestinal illness (HCGI) and days of illness stratified by baseline GI symptoms\n | Active device | Sham device\n | (n = 21) | (n = 24)\nEpisodes of highly credible gastrointestinal illness (HCGI) | \nParticipants with baseline GI symptoms | \nTotal episodes of HCGI | 15 | 17\nTotal days at risk for HCGI episodes 790 | 310\nCrude rate | 0.019 | 0.055\nParticipants without baseline GI symptoms | \nTotal episodes of HCGI | 1 | 14\nTotal days at risk for HCGI episodes 688 | 1,487\nCrude rate | 0.001 | 0.009\nDays of HCGI |  | \nParticipants with baseline GI symptoms | \nTotal days of HCGI | 246 | 262\nTotal days of observation | 1,313 | 784\nCrude rate | 0.187 | 0.334\nParticipants without baseline GI symptoms | \nTotal days of HCGI | 7 | 60\nTotal days of observation | 774 | 1,811\nCrude rate | 0.009 | 0.033\np A new episode of HCGI was defined as the presence of any of the four definitions of HCGI | \npreceded by 6 HCGI-free days. |  | \nTable 5 | Distribution of highly credible gastrointestinal illness episodes and days\nNumber of participants | \nexperiencing listed number | \nin active device group | \nTable 6 | Water consumption patterns\n | Mean number of 240 ml (8 oz) glasses of water consumed per day (95% CI) | \n | Active group (n = 21) | Sham group (n = 24) | Total (n = 45)\nBottled water | 2.98 [2.44, 3.52] | 3.40 [2.85, 3.96] | 3.21 [2.83, 3.60]\nUnheated tap water at home | 2.73 [2.08, 3.39] | 3.40 [2.88, 3.92] | 3.10 [2.68, 3.51]\nUnheated tap water away from home | 1.98 [1.32, 2.63] | 1.63 [1.32, 1.96] | 1.79 [1.45, 2.13]\nJohn M. Colford et al. | Gastroenteritis and drinking water in HIV \u00fe persons Journal of Water and Health | 3.2 | 2005 Downloaded from http://iwaponline.com/jwh/article-pdf/3/2/173/396288/173.pdf by guest\nDownloaded from http://iwaponline.com/jwh/article-pdf/3/2/173/396288/173.pdf by guest\nACKNOWLEDGEMENTS\nFunding for this work was provided entirely through the University of California University-wide AIDS Research Program Award No. M98-B-1300 and Cooperative Agreement No. UR2/CCU916252-02-2 from the Centers for Disease Control and Prevention.\nThe authors acknowledge the following individuals without whose contributions this project could not have been completed: Sherline Lee,",
        "label": "low",
        "id": "task4_RLD_test_650"
    },
    {
        "paper_doi": "10.1136/bmj.h1019",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Trial design: cRCTUnit of randomization: community with >= 1 chemical shopNumber of clusters: 24 (12 per arm) to obtain 80 adults and 114 children per clusterData collection: seller kept records on test results, medications dispensed, and whether customer was referred. Slide for parasitology reading at laboratory.Length of follow-up: 17 monthsAdjustment for clustering: yes\n\n\nParticipants: Target treatment group: adults and children > 6 monthsSample size: 4208 (2719 intervention, 2029 comparison)Exclusion criteria: pregnancy, age < 6 months, signs of severe disease, prescription from health facility\n\n\nInterventions: Staff who received training: chemical sellersDuration of training: 1 day over and above standard 3-day malaria case management training for both groupsContent of training: treating Pfalciparum malaria after positive mRDT with AL, AQAS, or DP, and referring after negative mRDTSupervision: fieldworkers and supervisors provided technical support. Accuracy of records of drugs dispensed validated by random checks of forms and 'mystery clients'. Direct observation of interactions between chemical sellers and customers by checklist on weekly basis for first month and a further week midway through trialAntimalarials free to participants: no, but subsidized through the Affordable Medicines Facility malariamRDTs free to participants: yesAdditional details: chemical sellers can also sell analgesics, antibiotics (co-trimoxazole), multivitamins/minerals, haematinics, and antacids.\n\n\nOutcomes: All-cause mortality and malaria mortality (risk of bias combined), use of antimalarial when microscopy-negative, appropriate treatment (defined as antimalarial provision to microscopy-positive participants and no antimalarial provision to microscopy-negative participants), number receiving an antimalarial\n\n\nNotes: Control: chemical sellers dispensing medicines without test results (community-based treatment of suspected malaria by clinical diagnosis)Country: GhanaSetting: ruralMalaria endemicity: not statedStudy dates: August 2011-January 2013Study sponsor: the Malaria Capacity Development Consortium of the London School of Hygiene & Tropical Medicine, with funding from the Welcome Trust and the Bill and Melinda Gates Foundatio\n\n",
        "objective": "To evaluate community\u2010based management strategies for treating malaria or fever that incorporate both a definitive diagnosis with an mRDT and appropriate antimalarial treatment.",
        "full_paper": "Objective\nTo examine the impact of providing rapid diagnostic tests for malaria on fever management in private drug retail shops where most poor rural people with fever present, with the aim of reducing current massive overdiagnosis and overtreatment of malaria.\nDesign Cluster randomized trial of 24 clusters of shops.\nSetting Dangme West, a poor rural district of Ghana.\nParticipants Shops and their clients, both adults and children.\nInterventions Providing rapid diagnostic tests with realistic training.\nMain outcome measures\nThe primary outcome was the proportion of clients testing negative for malaria by a double-read research blood slide who received an artemisinin combination therapy or other antimalarial.\nSecondary outcomes were use of antibiotics and antipyretics, and safety.\nResults Of 4603 clients, 3424 (74.4%) tested negative by double-read research slides.\nThe proportion of slide-negative clients who received any antimalarial was 590/1854 (32%) in the intervention arm and 1378/1570 (88%) in the control arm (adjusted risk ratio 0.41 (95% CI 0.29 to 0.58), P<0.0001).\nTreatment was in high agreement with rapid diagnostic test result.\nOf those who were slide-positive, 690/787 (87.8%) in the intervention arm and 347/392 (88.5%) in the control arm received an artemisinin combination therapy (adjusted risk ratio 0.96 (0.84 to 1.09)).\nThere was no evidence of antibiotics being substituted for antimalarials.\nOverall, 1954/2641 (74%) clients in the intervention arm and 539/1962 (27%) in the control arm received appropriate treatment (adjusted risk ratio 2.39 (1.69 to 3.39), P<0.0001).\nNo safety concerns were identified.\nConclusions Most patients with fever in Africa present to the private sector.\nIn this trial, providing rapid diagnostic tests for malaria in the private drug retail sector significantly reduced dispensing of antimalarials to patients without malaria, did not reduce prescribing of antimalarials to true malaria cases, and appeared safe.\nRapid diagnostic tests should be considered for the informal private drug retail sector.\nRegistration Clinicaltrials.gov NCT01907672\nIntroduction\nMalaria remains one of the most common diagnoses in Africa, and fever the commonest outpatient presentation, although most children and adults with fever present solely to the private drug retail sector (shops) rather than formal health services.\nA serious problem of over-prescription of antimalarials to patients without malaria has been found across the continent.\nThere has been a major shift in emphasis on treatment of febrile illness in the formal health sector in Africa that antimalarial treatment should always be guided by parasitological testing.\nThis has been driven by several factors, including that a large proportion of patients with fever are being given antimalarials when they have other, potentially serious causes of infection which are being missed,    the waste of relatively expensive antimalarials on patients without malaria, and that unnecessary antimicrobial use helps drive drug resistance and exposes patients to unnecessary side effects.\nAs a result of these efforts, overdiagnosis of malaria in the public sector is now falling rapidly.\nHowever, most children and adults with febrile illness, including the poorest, are treated in the private drug retail sector in most countries in Africa, including Ghana, even where public sector treatment is free.\nStrengthening the use of diagnostic tests for malaria in public health facilities therefore affects only a minority of cases of over-prescription of antimalarials.\nIn rural areas in particular, most of the poorest often access antimalarials through the private drug retail sector, largely because of the transport cost and opportunity cost (things people could otherwise spend money on) of accessing healthcare.\nIn recognition of this, several schemes, including the Affordable Medicines Facility for malaria have been deployed to ensure that effective antimalarials are available through private outlets that poorer patients use.\nThese schemes have provided drugs, but generally not diagnostic support, and poor targeting of antimalarials is recognized as a major risk to these efforts.\nReducing overtreatment of malaria in the private sector is becoming increasingly important as the proportion of fevers attributable to malaria reduces so the relative importance of non-malaria causes increases.\nRapid diagnostic tests for malaria are potentially usable in peripheral private drug retail facilities.\nThey are sensitive and can be used with minimal training.\nIn the formal public sector, substituting these for clinical (non-test based) diagnosis in peripheral settings without access to laboratories generally leads to significantly better targeting of antimalarials and less over-prescription, including in Ghana, provided it is linked to training.\nFor this reason, providing free or subsidized rapid diagnostic tests has been strongly advocated by some to be provided for use in the private drug retail sector including, in the UK, the Public Accounts Committee of the House of Commons and National Audit Office, but without a strong evidence base that they can be used correctly.\nThe scale of subsidy this would require is substantial; it is estimated there are 656 million fevers a year in African children aged 0-4 years old alone, of which only a minority go to the public sector.\nIt would change the nature of healthcare provision in Africa, making private providers diagnosticians, until now the preserve of clinicians.\nThe evidence base to support this significant policy change is weak with little reliable data from Africa outside observational studies, and some less encouraging data from the very different setting of South East Asia.\nThere are several reasons why a technology that works in the formal public sector may not work for shops.\nThe most obvious is that, if they are used correctly, a major effect of rapid diagnostic tests is likely to be to reduce over-prescription of antimalarials, so the effect in the private drug retail sector could be to reduce sales of a major item, for which there may be no rational economic advantage to the seller.\nConversely, incorrectly used rapid diagnostic tests in untrained hands could lead to true cases of malaria, a potentially fatal disease, not being treated because of false negative readings.\nThere are also safety concerns around bloodborne viruses (hepatitis, HIV, Ebola); sharps disposal is difficult and sharps reuse may have some economic advantage to shopkeepers (risking clients).\nAdditionally, sharps risk to shopkeepers is non-trivial in a setting where hepatitis B prevalence is often over 10% in blood donors.\nPrivate sector providers have training, incentives, backgrounds, and population served that are different from the formal public sector, so extrapolating directly from public sector data to the private sector is unwise.\nThis study set out to examine the impact of providing rapid diagnostic tests on antimalarial and other prescribing in the most commonly used part of the private drug retail sector in rural and semi-rural Ghana, chemical shops.\nIn particular, we hypothesized that providing rapid diagnostic tests will reduce overprescribing of antimalarials to clients with no malaria.\nWe also aimed to look at safety, both in terms of deaths or adverse illness from missed malaria when rapid diagnostic tests are introduced and the use of sharps.\nThis is one of the first trials of diagnostics in the private drug retail sector in Africa, and specifically West Africa.\nMethods\nStudy site and population\nThe trial was carried out in Dangme West, a rural district with widespread poverty in Ghana with an estimated population of 142\u00a0633 in 2009 mid-year.\nThe people live in scattered small communities and are mostly subsistence farmers or fishermen.\nVehicular transport is unavailable in many parts of the district, making access to formal care difficult.\nHealth services are organized around a three tier system at district level, with the district hospital as the referral level, the health centre at subdistrict level, and the Community Health and Planning and Services (CHPS) compounds, which deliver close-to-client services mostly at home, at community level.\nThere is one district hospital, three health centres, 13 CHPS compounds, and five formal private health facilities.\nIn addition, 56 chemical shops and six pharmacies sell pharmaceutical products (fig 1 ); Licensed Chemical Sellers are the main outlet for antimalarials in Ghana..\nEarlier studies carried out in the district showed that for presumed \u201cmalaria\u201d in the household, the first action taken is, in order of the most common: home treatment, chemical shop, health centre, hospital, drug peddler, and traditional healer.\nThe study was conducted across both the high and low transmission seasons.\nChemical sellers in Ghana are regulated by the Pharmacy Council.\nTo be eligible, a person must possess a minimum of secondary school level qualification, with basic knowledge in healthcare delivery being an advantage.\nApplicants attend pre-licensing training, and the licenses are renewed annually.\nThe shops are legally allowed to retail over-the-counter medicines to members of the public in communities that the Pharmacy Council considers poor.\nThe authorized medicines include analgesics and antimalarials.\nThe only antibiotic allowed is co-trimoxazole.\nTrial design\nA cluster randomized trial design was used, with a cluster defined as a community with at least one chemical shop.\nTwenty four communities, with one to five shops per community, were eligible for inclusion in the study.\nThe trial was evaluated among clients reporting to a participating chemical shop complaining of fever or requesting an antimalarial medicine.\nClients were excluded if they were pregnant, <6 months old, had signs of severe disease, a prescription from a health facility, or would be in the district for less than 28 days.\nAssessments for eligibility were carried out by the chemical sellers, who were trained to do so.\nClients with severe disease or who were pregnant were referred to the nearest health centre or the district hospital by the chemical seller.\nAn initial baseline study was carried out involving a census, including Global Positioning System (GPS), of all chemical shops to document the location of each shop and antimalarials available.\nThe results of the baseline informed the study design.\nAs part of formative research, focus group discussions were held with community members and chemical sellers to find out how acceptable the idea of chemical sellers testing for malaria in shops was, to community members and the best way to introduce the intervention.\nThe intervention\nShops in communities randomised to the control arm were expected to dispense medicines without test results as per current practice, while those in the intervention arm carried out a test for malaria using an rapid diagnostic test before dispensing any medication.\nIn positive malaria tests, rapid diagnostic tests form a coloured line in the test area.\nTests generally take 15\u201320 minutes depending on the test kit.\nCareStart Malaria HRP2 (Pf) rapid diagnostic test kits, which fulfil all the WHO assessment criteria, were provided by the study team on a monthly basis and were free of charge to chemical sellers and clients.\nThe study team carried out regular quality control procedures by picking a random sample of test kits from the shop for testing using standard positive blood samples.\nThis was to ensure that the rapid diagnostic test kits continued to perform optimally under the storage conditions in the chemical shop.\nAll chemical sellers attended a three day training on malaria, covering topics that included the antimalarial drug policy of Ghana; causes, signs, and symptoms of malaria; signs of severe disease and indications for referral; how to take a blood sample and make a blood slide; blood safety and handling of sharps; infection prevention procedures; and standard operating procedures of the study.\nSellers in the intervention arm attended an additional day of training, covering how to carry out and interpret rapid diagnostic tests for malaria and further management of clients with a negative rapid diagnostic test result.\nThe training involved a demonstration by a laboratory technologist followed by practice sessions by chemical sellers on their colleagues.\nThey practised how to take a blood sample safely and also tested for malaria with a rapid diagnostic test.\nTest kits, sharps bins, and documentation were introduced and used during the training.\nLecture sessions were interspersed with group discussions, role plays, and individual exercises.\nAssignments were marked and discussed in class.\nA minimum of two attendants were trained per shop to ensure continuity of service.\nAfter the training, all chemical sellers were provided with reference charts for doses of artemisinin combination therapies.\nWhenever a trained shop attendant resigned and was replaced by a new one, training was carried out in the shop for the new attendant.\nSince Ghana was one of the phase one countries benefitting from the Affordable Medicines Facility for malaria, subsidized artemisinin combination therapy was available at the lowest level of the private drug retail sector from mid-2011, and chemical sellers were encouraged to use their regular source of supply of antimalarials.\nBins for disposal of sharps and a certificate identifying the seller as a trained provider of malaria testing were also provided.\nFieldworkers and supervisors were also trained separately in taking of blood samples and rapid diagnostic tests and the standard operating procedures of the study to enable them to provide technical support to the chemical sellers.\nStudy outcomes\nThe primary outcome was the proportion of clients who tested negative for malaria by a double-read research blood slide and who received an artemisinin combination therapy or other antimalarial.\nSecondary outcomes included treatment with antibiotics by test outcome, treatment of patients with a negative rapid diagnostic test result, treatment with other antimalarials or alternative treatments, the operational sensitivity of the rapid diagnostic tests used by chemical sellers against research slide results, the safety of use of rapid diagnostic tests at community level, and the impact on numbers of referrals to formal medical care.\nThe presence of malaria was assessed through a blood slide, which was prepared by health centre laboratory staff from a blood sample taken by the chemical seller as part of the consultation with the client regardless of study arm.\nSlides were double-read in research laboratories, and discrepant slides were re-read by a third and senior microscopist.\nInformation on the test results obtained, medications dispensed, and whether the client was referred were documented by the seller on a form designed specifically for the study.\nThe accuracy of the records of drugs dispensed was validated by conducting random checks of the forms and through \u201cmystery clients.\u201d\nThe \u201cmystery client\u201d complained of symptoms for himself or his child at home and submitted to a test when it was offered, observing the procedure the chemical seller followed for later documentation using a checklist.\nTo ensure that chemical sellers correctly conducted the rapid diagnostic tests and adhered to safety and infection control procedures, direct observation of interactions between chemical sellers and clients was carried out by the study team using a checklist on a weekly basis during the first month of evaluation and for a further week mid-way through the trial in both arms.\nIn both arms clients were assessed for eligibility and requested to provide written consent for participation in the study.\nBasic demographic details were collected from all consenting clients, and a blood sample was taken by the chemical seller for the research slide.\nDetailed directions to the client\u2019s home were also collected for the purposes of follow-up.\nChemical sellers had access to the subsidized green leaf logo artemisinin combination therapy from the open market via the Affordable Medicines Facility for malaria.\nIn the control arm, services were provided to clients following usual practice without test results.\nThe chemical sellers collected a blood sample in a \u201cmicrotainer\u201d for study testing but did not do rapid diagnostic test testing.\nControl chemical sellers used the same client record form as the intervention arm.\nIn the rapid diagnostic test arm consent provided by clients included consent for rapid diagnostic testing.\nIf the test result was positive, chemical sellers were trained to encourage clients to purchase any of the three recommended artemisinin combination therapies; amodiaquine-artesunate, arthemeter-lumefantrine, or dihydroartemisinine-piperaquine.\nClients with a negative test were recommended to be referred to a nearby health facility or a facility of their choice.\nClients who refused consent for a rapid diagnostic test received presumptive treatment (current practice), and reasons for the refusal of consent were documented.\nBlood slide results were provided to all shops after three days; the results of the blood slide were therefore not used by the chemical seller for initial dispensing decisions.\nPatients who were slide-positive but who had not been given an antimalarial in the chemical shop were contacted within 24 hours and requested to return to the chemical shop for an antimalarial.\nFollow-up was carried out on day 28 after the visit to the shop for all research slide-positive patients to determine the clinical course and outcome of the illness and care seeking behaviour after the visit to the chemical seller.\nRandomization and registration\nCommunities were allocated to intervention or control by a process of restricted randomisation to ensure balance in covariates expected to be important correlates of the primary outcome.\nBecause of the differing number of shops in each community, the balance was restricted to a difference of two in the number of chemical shops per arm, the same number of communities containing only one shop per arm, and a maximum relative difference of 20% in the average number of cases per day.\nRandomisation used a program written in R (version 2.14.2) by a statistician not otherwise involved in the study.\nAll chemical sellers from shops assigned to each arm were invited to a meeting with the study team, informed about their assignment, and enrolled into their study arm.\nFollowing this, community sensitization meetings and durbars (a large community meeting usually hosted by the traditional leaders of the community during which issues important to the community are discussed and consensus reached) were carried out to explain the intervention to community members.\nThose preparing and reading the blood slides were blind to the study allocation and rapid diagnostic test result.\nPermission was given by the Ghana Pharmacy Council for chemical shop involvement.\nThe trial received ethical clearance from the Ethical Review Committee of the Ghana Health Service and the ethics committee of the London School of Hygiene & Tropical Medicine.\nIt was registered on ClinicalTrials.gov NCT01907672 before July 2011, although, because of a technical misunderstanding, there was a delay in releasing the protocol until September after trial start.\n(Formal release of the protocol was after randomization but before any outcome data were collected or analyzed, and no major changes were made between the attempt to register the study and the date when registration was complete.)\nStatistical methods\nAfter sample size calculations, the aim was for 24 communities (12 per arm) to be included.\nCalculations were conducted to determine a trial size sufficient to detect a 25% relative reduction in the primary outcome from 60% in the control arm among adults and 50% among children.\nUsing methods for cluster randomised trials and assuming a coefficient of variation between clusters of 0.2, 80 adults and 114 children per community were required to detect the specified reductions with 80% power at the 5% significance level.\nData were double entered and validated using Epidata 6.1, and analyzed using STATA version 12.0 (STATA Corporation, College Station, Texas).\nThe effect of the intervention was analyzed using published methods suitable for cluster randomized trials with fewer than 20 clusters per arm.\nThe observed proportion with the primary outcome was calculated for each community (cluster).\nWithin each arm, a weighted average of the cluster proportions was taken with weights provided by the sample size for each cluster, to give the prevalence of the primary outcome in each arm.\nDue to the skewness in the distribution of the cluster proportions, a log transformation was applied and the mean of the log-proportions over clusters (log of the geometric mean) was estimated in each arm.\nThe unadjusted risk ratio was then computed from exponentiation of the log of the ratio of these geometric means.\nThe pooled variance of the log-proportions over clusters within each arm was calculated and used to obtain 95% confidence intervals and carry out formal hypothesis testing using a t\u00a0test on the logarithm of the risk ratio.\nAdjustment for covariates was made by firstly fitting a standard logistic regression model to the individual-level data, including terms for the covariates of interest, but excluding the intervention effect.\nFrom the fitted model, the expected number of events in each cluster was computed.\nThe ratio of the observed to expected number of events in each cluster was used to calculate cluster-specific ratio residuals.\nThe next stage of analysis was to compare the residuals between study arms using the above methods for estimating the risk ratios and 95% confidence intervals as well as hypothesis testing with the residuals replacing cluster-level proportions.\nAnalysis by age group and analyses of secondary outcomes were conducted using the same methods.\nResults\nFigure 1  shows the trial profile.\nThe trial was conducted from August 2011 to January 2013; last recruitment was 7 December 2012.\nOf the 4748 clients (2719 intervention, 2029 controls) enrolled in the study, 4603 (97%) were included in the analysis (fig 1) with 145 (44 negative rapid diagnostic test result, 34 positive rapid diagnostic test result, 67 controls) not assessable because of haemolysed or clotted blood sample for the research slide.\nThe study was unable to achieve the planned sample size specifically for the subgroup of children because of difficulties in recruiting this age group at the chemical sellers.\nCommunities in both arms were similar (table 1), although control arm sellers were more likely to have had formal training in medicine dispensing.\nTreatment based on research slide results\nOf 4603 clients who had research blood slide results, 1854/2641 (70%) in the intervention arm and 1570/1962 (80%) in the control arm had a negative slide reading (no malaria).\nThe proportion of clients negative for malaria by blood slide who received an antimalarial (artemisinin combination therapy or other antimalarial) was significantly lower in the intervention arm (590/1854, 32%) than the control arm (1378/1570, 88%) (P<0.0001); the adjusted risk ratio was 0.41 (95% confidence interval 0.29 to 0.58) (table 2).\nSimilar results were observed when the analysis was conducted separately in children and adults (table 2).\nIn both study arms, among slide-negative clients who received an antimalarial, a high proportion of the antimalarials were artemisinin combination therapy; 1285/1378 (93%) in the control arm and 557/590 (94%) in the intervention arm.\nRapid diagnostic tests did not lead either to a reduction or increase in true malaria cases getting an effective antimalarial.\nOf those who were slide-positive 690/787 (88%) in the intervention arm and 347/392 (89%) in the control arm received an artemisinin combination therapy (adjusted risk ratio 0.96 (0.84 to 1.09)).\nOverall, 737 out of 3424 (21.5%) of all slide-negative clients received an analgesic or non-steroidal anti-inflammatory drug (NSAID) without an antimalarial or antibiotic, while 579 of them received no medicine.\nAmong all the 1179 slide-positive clients, 23 (1.9%) received analgesics while 17 (1.4%) received no medicine.\nAmong slide-negative clients, 629/1854 (33.9%) in the intervention arm and 108/1570 (6.9%) in the control arm received an analgesic, suggesting that shopkeepers switched appropriately from selling antimalarials to selling analgesics or antipyretics to clients with non-severe illness and a negative rapid diagnostic test result (fig 2 ).\nThe secondary outcome of \u201cappropriate malaria treatment\u201d differed between clients attending shops in communities in the rapid diagnostic test arm (74%) compared with clients in the control arm (27%); adlusted risk ratio 2.39 (1.69 to 3.39).\nThis composite indicator comprised clients with a positive blood slide who received an artemisinin combination therapy and clients with a negative slide who did not receive an artemisinin combination therapy or other antimalarial.\nThere was no evidence of a difference between the arms in the former but strong evidence of an increase in the latter (adjusted risk ratio 5.28 (2.5 to 10.77)).\nNo clients with a positive slide result and only seven with a negative slide reading received antibiotics, which chemical sellers are not allowed to dispense in Ghana (table 2).\nTreatment on the basis of rapid diagnostic test results\nAll 2719 clients visiting shops in intervention communities were tested for malaria with rapid diagnostic tests.\nOf these, 1368 (50.3%) were reported negative, and 1351 (49.7%) were positive.\nTreatment was in good agreement with the rapid diagnostic test result, as recorded by chemical sellers.\nAmong those who were reported negative for malaria by rapid diagnostic test, only 38 (3%) received an artemisinin combination therapy or other antimalarial, while 656 (48%) received an analgesic and eight (0.6%) received an antibiotic.\nOverall, 570/1368 (42%) clients with a negative rapid diagnostic test received no medicine.\nAmong clients with a positive rapid diagnostic test, 1344/1351 (99.5%) received an antimalarial.\nOnly five (0.4%) received no medicine while two (0.2%) received an analgesic (fig 3 ).\nSensitivity and specificity of rapid diagnostic tests under operational conditions\nOf the 2641 clients in the rapid diagnostic test arm who had a valid research blood slide result, 29 who tested negative by rapid diagnostic test were found to be positive by research slide (table 3).\nThe operational sensitivity of rapid diagnostic tests as used by chemical sellers was therefore 96% and specificity was 70% judged against the gold standard of double-read research slides.\nThere was significant variation between chemical shops in terms of apparent specificity.\nSensitivity and specificity of the rapid diagnostic test were calculated by shop, restricted to those shops that enrolled more than 100 clients.\nSensitivity ranged from 98% to 100%, suggesting that all shops were able to identify true cases.\nMost shops had specificity from 73% to 98%, but two had a specificity of 30% and 31% and one of 52%, despite the sellers having successfully completed training, raising the possibility that they were reporting positive rapid diagnostic tests to justify sales of antimalarials.\nFollow-up, referral, and safety\nThere was no evidence that introducing rapid diagnostic tests led to major negative outcomes because true malaria cases were missed.\nOf 1179 clients with a positive slide reading, 1076 (91.3%) were successfully followed up (670 intervention arm, 406 control) on day 28 after diagnosis, with missing data for five.\nOf these, 95.3% (1021/1071) completed the treatment they were sold by the chemical seller.\nThere was no significant difference between the two study arms with regards to client completion of medication on follow-up (P=0.34).\nFive of them (0.5%) did not receive any medicines from the chemical seller, while 4.2% (45/1071) admitted not completing their medications, of whom 34/45 did not because they felt better.\nOnly four stopped taking their medications because of side effects.\nLess than 1.5% (13/1071) of all the slide-positive clients followed up admitted to being referred to another health facility after their visit to the chemical seller, but significantly more of these were from the rapid diagnostic test arm (P=0.024).\nNone of the clients who were followed up during the study period died.\nOne adult female client died after the study ended; the death was not linked to the illness presented at the chemical shop nor the treatment given.\nOf 1368 clients, 1088 (80%) who tested negative with the rapid diagnostic test were referred to the formal healthcare sector, compared with only one client in the control arm.\nThese were contacted by phone to find out whether the referral was completed.\nOf the 1088, 248 recalled being referred, 154 (62%) reported that they went to the referral facility, while 94 (38%) admitted to not going as they had been requested, and 54 said they did not go because they felt better.\nMystery client and 133 other chemical seller-client interaction observations in the intervention shops showed that chemical sellers adhered largely to safety instructions for handling sharps.\nFor all 18 safety indicators assessed, the chemical sellers performed well.\nThe range was between 87.2% and 100%.\nThe lowest indicator was the proportion of chemical sellers who discarded the lancet immediately into the sharps bin, which was 87.2% (116/133).\nWaste material from both bins was disposed of regularly in the incinerators of the nearest public sector health centre through a prior arrangement.\nDiscussion\nReducing overuse of antimalarials in the private sector is a public health priority.\nMost people with malaria or febrile symptoms, the commonest syndrome for severe illness in West Africa, are treated in the private drug retail sector, but the effect of introducing malaria diagnostics to this sector has been little studied.\nThere has been considerable debate on whether to introduce rapid diagnostic tests into the private drug retail sector, but with limited current evidence of impact or safety.\nThe incentives for drug shops (which want to sell drugs) are completely different than for clinicians, and turning shops into diagnostic centers for the most common acute condition (febrile illness) has significant policy implications.\nDespite this, pressure to introduce rapid diagnostic tests has been strong.\nFor example, the Public Accounts Committee of the UK parliament recommended in 2013: \u201cThe [UK Government] should extend its support for rapid diagnostic tests to the private drug retail sector on a national or regional scale.\nto seize the unquestionable benefits this would bring.\u201d\nThis trial showed that introducing rapid diagnostic testing for malaria in chemical shops, the major private providers to poorer patients in Ghana, to guide dispensing practice of chemical sellers had a substantial impact on the dispensing of antimalarials.\nThere was no change in prescribing artemisinin combination therapys to true malaria cases, but a significant reduction in dispensing antimalarials of all types, and artemisinin combination therapy specifically, was observed for malaria-negative cases.\nAlthough there was still some prescribing of antimalarials for test-negative clients, there was a substantial reduction in the number of potentially serious causes of infection which were being missed compared with standard care.\nA study carried out in two shops in Nigeria had similar findings.\nIn the context of a country where antibiotics cannot be sold by shops, sales were diverted to antipyretics by the chemical sellers.\nNo safety concerns were identified as a result of introducing rapid diagnostic tests after a short period of training, whether in not treating malaria cases, unexpected mortality, or sharps handling; this is an important issue in areas where hepatitis B is common and HIV prevalence is non-trivial.\nThere was no evidence of reuse of lancets or improper disposal of sharps that could pose a danger to the public.\nThere was an increase in appropriate referral of clients from the chemical shop to the formal healthcare sector.\nAttempts to increase the proportion of children with fever who get antimalarials from the private drug retail sector such as the Affordable Medicines Facility for malaria have often been successful, but evidence of how to reduce overuse of antimalarials in those without malaria in the private sector has been lacking.\nDespite our not achieving the target sample size for the subgroup of children, the large effect estimates, small P values, and moderately narrow 95% confidence intervals provide evidence that an intervention effect exists and has been estimated with reasonable precision in this setting.\nThere is now a wide body of literature on the effects of introducing rapid diagnostic tests in the formal health sector and community workers in Africa, albeit mainly from East Africa, and this has generally shown that introducing rapid diagnostic tests after a short period of training (like that provided here) is not sufficient to change behavior of healthcare workers.\nMore prolonged training and ongoing supervision have generally been needed to achieve good results, probably because malaria treatment practices are driven by ingrained behaviours with multiple drivers  in both providers and clients, but when undertaken is successful.\nOur study in Ghana demonstrates a more rapid change in behavior among shopkeepers when rapid diagnostic tests were first introduced than that found among health staff in the same setting.\nHealth workers, particularly in a health facility where testing for malaria with microscopy was known, still prescribed antimalarials for a high proportion of test-negative patients.\nClients visiting chemical shops in the intervention arm with complaints of fever were three times more likely to receive appropriate treatment for malaria fever than those visiting shops where no rapid diagnostic test testing was being done.\nThere is evidence from several settings that rapid diagnostic tests for malaria are not widely available in the private sector, but that in principle they are acceptable to clients and caregivers.\nAlthough this study lends good support to the idea that rapid diagnostic tests can be introduced to the private drug retail sector and be both effective and safe, three caveats should be borne in mind.\nThe first is the obvious one relevant to all trials; results found under trial conditions, even relatively operational trials such as this, are generally better than those found in subsequent operational practice.\nThe second is that in this trial rapid diagnostic tests were introduced free.\nThis establishes the principle that rapid diagnostic tests can be well used by shopkeepers if they are introduced without cost, such as by the Global Fund, but the financial incentives will change if the shopkeeper or client is paying for the test, and this could change behavior.\nHow much rapid diagnostic tests are sold for is likely to influence uptake.\nThe third is that chemical sellers are not supposed to prescribe most antibiotics in Ghana (although a few do), so one of the major risks of introducing rapid diagnostic tests in some other countries\u2014that they lead to a switch from overuse of antimalarials to overuse of antibiotics where they are freely sold\u2014was not tested here; sales were diverted almost exclusively to antipyretics or analgesics.\nNot using antibiotics seems to be safe in children with a negative rapid diagnostic test and without signs of severity; bacteremia rates are low in outpatient settings except in children <1 year old.\nOverprescription of antimalarials and the possible effects of rapid diagnostic tests in Ghana in the private drug retail sector is probably relevant to other settings.\nThere is the possibility that clients who are refused an antimalarial because of a negative rapid diagnostic test may go to another shop to get one, but follow-up data suggest this is rare (data not shown).\nAnthropological data from other countries suggest patients want a diagnosis they trust, not a specific drug.\nThere is the risk that shopkeepers changed their behaviour because blood samples for slides were being taken, but this was true in both trial arms and we think it unlikely that the large effect difference seen between arms is fully explained by this.\nThe major problem of overprescription of antimalarials to malaria-negative patients occurs in both the public and the private sector, but the private sector sees the majority of patients in Ghana.\nAs the incidence of malaria drops in many settings, the importance of targeting of antimalarials to those who need them will increase.\nConsiderable advances have been achieved in formal public sector settings by the use of rapid diagnostic tests.\nHowever, since most patients with fever and malaria are treated before they get to the formal sector, ensuring best targeting of antimalarials in the places such as shops where patients, and especially rural poorer patients, go, is essential.\nThis study supports the idea that introduction of rapid diagnostic tests in the private drug retail sector with a realistic training package has the potential to be safe and effective.\nWhat is already known on this topic\nOverdiagnosis of malaria is a major problem in Africa, and fever is the commonest potentially serious presentation\nMost poorer patients in Africa are treated in the private drug retail sector, yet almost all evidence for improving diagnosis come exclusively from the public sector.\nHowever, the relationship between private sector seller and buyer is completely different from that between physician and patient, leading to different behaviours on both sides.\nAsking shopkeepers to not sell a drug in case of a negative test is potentially going against their self interest, and patients are generally coming to a shop for a product (rather than to a healthcare worker for a diagnosis), meaning their expectations are also different\nWhat this study adds\nThis randomized trial provides rigorous evidence that rapid tests for malaria can substantially improve targeting of drugs safely in the private retail sector used by the poorest patients\nTreatment in the chemical shop can be in good agreement with test results in the context of use of rapid diagnostic tests by chemical sellers\nThe results may not be generalizable to other settings such as urban areas and where drug retail shops are allowed to sell antibiotics, and the provision of free rapid diagnostic tests may not be possible outside schemes such as the Global Fund\nCite this as: BMJ 2015;350:h1019\nRelated links\nthebmj.com\nResearch: Investigation and treatment of imported malaria in non-endemic countries (BMJ 2013; 346: f2900)\nResearch: Overdiagnosis and mistreatment of malaria among febrile patients at primary healthcare level in Afghanistan (BMJ 2012; 345: e4389)\nResearch: Risk factors for mortality from imported falciparum malaria in the United Kingdom over 20 years (BMJ 2012; 344: e2116)\nResearch: Protective efficacy of co-trimoxazole prophylaxis against malaria in HIV exposed children in rural Uganda (BMJ 2011;342:d1617)\nFig 1 Trial profile: rapid diagnostic test versus normal practice in Licensed Chemical shops\nFig 2 Medicines dispensed by chemical sellers with or without a rapid diagnostic test (RDT) for malaria, by the subsequent blood slide results (not available to chemical seller to guide dispensing)\nFig 3 Medicines dispensed by chemical sellers based on rapid diagnostic test (RDT) for malaria \n\n Baseline characteristics of the study population by treatment arm. Values are numbers (percentages) unless stated otherwise\n | RDT arm | Control arm\nCluster level | (n=12) | (n=11)\nMedian (range) No of shops per cluster | 2 (1\u20135) | 2 (1\u20135)\nMedian (range) No of clients per cluster | 214 (17\u2013572) | 213 (51\u2013231)\nLocation: |  | \n Urban only | 0 | 0\n Rural only | 1 (8) | 3 (27)\n Both | 11 (92) | 8 (73)\nShop level | (n=27) | (n=24)\nNo of sellers per shop: |  | \n 1-3 | 26 (96) | 23 (96)\n 4-6 | 1 (4) | 1 (4)\nRegular seller: |  | \n Owner | 12 (44) | 14 (58)\n Employee/assistant | 14 (52) | 10 (42)\n Family | 1 (4) | 0 (0)\nEducation level of respondent: |  | \n Primary | 1 (4) | 0 (0)\n Secondary | 20 (74) | 18 (75)\n Tertiary | 6 (22) | 6 (25)\nFormal training in medicine dispensing*: |  | \n No | 11 (41) | 6 (25)\n Yes | 16 (59) | 18 (75)\nRefresher training in past 6 months*: |  | \n No | 15 (56) | 6 (25)\n Yes | 12 (44) | 18 (75)\nReported average No of clients per day: |  | \n <20 | 6 (22) | 8 (33)\n 21-30 | 7 (26) | 5 (21)\n \u226530 | 14 (52) | 11 (46)\nClient level | (n=2719) | (n=2029)\nMedian (IQR) age (years) | 15 (6\u201329) | 19 (6\u201332)\nAge group: |  | \n <5 years | 534 (20) | 390 (19)\n 5-12 years | 719 (26) | 458 (23)\n \u2265 13 years | 1466 (54) | 1181 (58)\nSex: |  | \n Male | 1343 (49) | 997 (49)\n Female | 1376 (51) | 1032 (51)\nCommunity of residence: |  | \n Urban | 1198 (44) | 730 (36)\n Rural | 1521 (56) | 1299 (64)\n\nRDT=Rapid diagnostic test for malaria.\n*Formal training refers to a 1 year training in medicine dispensing. Refresher training refers to training lasting from half a day to a maximum of three days usually organized by the Pharmacy Council (the regulatory body), pharmaceutical companies, or other health agencies in specific subject areas.\n\n Crude and adjusted treatment outcomes based on research slide result, by study arm\n | Control (no RDT) arm |  | Intervention (RDT) arm |  | Crude results* |  | Adjusted results\u2020 | k statistic\nNo of clusters | Prevalence (%) among clients | No of clusters | Prevalence (%) among clients | Risk ratio (95% CI) | P value | Risk ratio (95% CI) | P value\nPrimary outcome: slide result negative, client receives antimalarial\u2021\nAll clients | 11 | 1378/1570 (88) |  | 12 | 590/1854 (32) |  | 0.34 (0.26 to 0.44) | <0.0001 |  | 0.41 (0.29 to 0.58) | <0.0001 | 0.52\nAdults | 11 | 853/1013 (84) |  | 12 | 334/1179 (28) |  | 0.33 (0.25 to 0.43) | <0.0001 |  | 0.40 (0.28 to 0.56) | <0.0001 | 0.55\nChildren | 11 | 525/557 (94) |  | 12 | 256/675 (38) |  | 0.40 (0.29 to 0.56) | <0.0001 |  | 0.52 (0.36 to 0.74) | 0.006 | 0.45\nSecondary outcomes\nAppropriate treatment\u00a7 | 11 | 539/1962 (27) |  | 12 | 1954/2641 (74) |  | 2.86 (2.24 to 3.63) | <0.0001 |  | 2.39 (1.69 to 3.39) | <0.0001 | \u2014\nACT to slide-positive client\u00b6 | 11 | 347/392 (89) |  | 12 | 690/787 (88) |  | 0.94 (0.81 to 1.09) | 0.42 |  | 0.96 (0.84 to 1.09) | 0.47 | \u2014\nNo antimalarial to slide-negative client | 11 | 192/1570 (12) |  | 12 | 1264/1854 (68) |  | 5.94 (3.32 to 10.65) | <0.0001 |  | 5.28 (2.59 to 10.77) | <0.0001 | \u2014\nAntibiotic to slide-negative client | 11 | 1/1570 (0.1) |  | 12 | 6/1854 (0.3) |  | \u2014 | \u2014 |  | \u2014 | \u2014 | \u2014\nAnalgesic or antipyretic to slide-negative client | 11 | 108/1570 (7) |  | 12 | 629/1854 (34) |  | 4.29 (2.09 to 8.84) | 0.004 |  | 3.46 (1.64 to 7.32) | 0.013 | \u2014\n\nRDT=rapid diagnostic test for malaria. ACT=artemisinin combination therapy\n*Adjusted for clustering, based on geometric mean of cluster summaries.\n\u2020Adjusted for clustering, age, sex, number of clients per cluster, number of shops per cluster, and characteristics of sellers (including whether they were the regular seller and had formal training or refresher training in the previous six months).\n\u2021Receiving an ACT or other antimalarial. Restricting the outcome to ACT only provides similar results; 1285 (82%) in the control arm; 557 (30%) in the RDT arm; adjusted risk ratio 0.37 (95% CI 0.28 to 0.51), P<0.0001.\n\u00a7Appropriate treatment defined as slide-positive client receiving an ACT or slide-negative client not receiving an antimalarial.\n\u00b6Among the 45 slide-positive clients in the control arm not receiving an ACT, 22 (49%) received other antimalarial, 8 (18%) received analgesic, 5 (11%) received no medicines, and 10 (22%) received other medicines. In the RDT arm, 70/97 (72%) slide-positive clients not receiving an ACT received other antimalarial, 15 (15%) received analgesic, and 12 (12%) received no medicines.\n\n Comparative results of the rapid diagnostic test (RDT) as used by chemical sellers and research blood slide results\nSlide results | RDT results\nPositive | Negative | Total\nPositive | 758 | 29 (false negative) | 787\nNegative | 559 (false positive) | 1295 | 1854\nInadequate sample  | 34 | 44 | 78\nTotal | 1351 | 1368 | 2719\n",
        "label": "low",
        "id": "task4_RLD_test_978"
    },
    {
        "paper_doi": "10.1186/1475-2875-5-108",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Individually RCTDate of trial: July to Oct 2001.\n\n\nParticipants: 117 malaria cases with P. falciparum >= 400 asexual stages/mL (thick film) recruited by mass blood survey and passive case detection. Symptoms not required.Age: >= 15 yearsSite: Central Java, Indonesia, an area with high CQ resistance and resurgent malaria approximately equal P. falciparum and P. vivax.Exclusion criteria: Pregnancy, breast feeding, body weight < 40 kg, G6PD deficiency, history of antimalarial or antibiotic in last 7 days, severe or complicated malaria, history or allergy or adverse reaction to trial medications, P. vivax or mixed infection.\n\n\nInterventions: CQ only (not included in this review).CQ+SP: CQ 150 mg base, 10, 10 and 5 mg/kg on days 1, 2, 3 (reported as days 0, 1, 2). SP 500 mg S 25 mg P on day 1 (reported as day 0).CQ+SP as for group 2 above plus PQ 45 mg on day 1 (reported as day 0).CQ+SP as for group 2 above plus PQ 45 mg on day 3 (reported as day 2).\n\n\nOutcomes: Parasite clearance time assessed at days 1, 3, 8, 15, 22, 29 or day of recurrent parasitaemia (reported as days 0, 2, 7, 14, 21, 28)Fever clearance time at days 2, 3, 4, 5, 8, 12, 15, 19, 22, 29Proportion of people with gametocytes (from chart) days 1 to 29Adverse events\n\n\nNotes: Some comparisons in the results reported include the CQ only group\n\n",
        "objective": "To assess the effects of single dose or short\u2010course PQ (or an alternative 8AQ) alongside treatment for people with P. falciparum malaria.",
        "full_paper": "Background\nChloroquine (CQ) or sulfadoxine-pyrimethamine (SP) monotherapy for Plasmodium falciparum often leads to therapeutic failure in Indonesia.\nCombining CQ with other drugs, like SP, may provide an affordable, available and effective option where artemisinin-combined therapies (ACT) are not licensed or are unavailable.\nMethods\nThis study compared CQ (n = 29 subjects) versus CQ + SP (with or without primaquine; n = 88) for clinical and parasitological cure of uncomplicated falciparum malaria in the Menoreh Hills region of southern Central Java, Indonesia.\nGametocyte clearance rates were measured with (n = 56 subjects) and without (n = 61) a single 45 mg dose of primaquine (PQ).\nResults\nAfter 28 days, 58% of subjects receiving CQ had cleared parasitaemia and remained aparasitaemic, compared to 94% receiving CQ combined with SP (p < 0.001).\nMsp-2 genotyping permitted reinfection-adjusted cure rates for CQ and CQ combined with SP, 70% and 99%, respectively (p = 0.0006).\nConclusion\nPrimaquine exerted no apparent affect on cure of asexual stage parasitaemia, but clearly accelerated clearance of gametocytes.\nCQ combined with SP was safe and well-tolerated with superior efficacy over CQ for P. falciparum parasitaemia in this study.\nBackground\nChloroquine (CQ) was the mainstay of antimalarial therapy from 1946 until the past decade when parasite resistance rendered it clinically ineffective in most areas of the world.\nAdult CQ therapy costs less than $0.20, is widely available almost anywhere in endemic areas, and has a generally good safety and tolerability profile.\nResource-strapped health agencies in endemic areas struggle with the decision to abandon this drug.\nAlternative drugs, such as artemisinin combination therapies (ACT), carry significant limitations linked to cost, ease of compliance and safety in vulnerable populations like infants and pregnant women.\nCombining antimalarials of differing mechanisms of action may diminish risk of onset of parasite resistance, but there are many other determinants of effectiveness and the more effective the therapeutic options available, the more likely treatment will result in a favorable clinical outcome.\nCQ combined with sulfadoxine-pyrimethamine (SP) has been proven safe and effective in clinical trials (Table 1) and has been adopted as the primary therapy for uncomplicated falciparum malaria in several countries.\nDespite the slight increase in cost of using two medications (from approximately $0.10 to $0.20), CQ combined with SP was proven cost-effective.\nIn some nations, this combination is first-line therapy (e.g. Papua New Guinea, Vanuatu, Philippines, Uganda, and Ethiopia).\nCombining CQ and SP offers compelling advantages in places like Indonesia, principally because both drugs are licensed, widely available, familiar to providers and patients, inexpensive, relatively safe, well tolerated, and easily administered.\nFurthermore, CQ + SP combined therapy was recently adopted by Papua New Guinea where the level of drug resistance is on par with the most drug resistant areas of Indonesia.\nAdoption of CQ + SP combined therapy for Plasmodium falciparum would require little more than a modification of malaria therapeutic policy and practice.\nIn contrast, most other combined therapy options would require licensing, acquisition, socialization, and carry unfamiliar and perhaps more severe risks linked to safety or compliance.\nThis is an especially salient problem for pregnant women and young children, for whom almost no safety data exists.\nClinical data for CQ combined with SP is still lacking in many countries, including Indonesia, where CQ and SP-resistant P. falciparum is prevalent.\nThis clinical trial evaluated the efficacy of CQ combined with SP in Central Java, a setting where 47% of P. falciparum infections are resistant to CQ and 22% resistant to SP.\nSince single dose primaquine is mandated on Java and Bali for transmission blocking activity (only adds an additional $0.05 to the cost of therapy), a single dose of primaquine was added to evaluate its impact on parasitological cure and on post-treatment gametocytaemia.\nMaterials and methods\nStudy location\nThe study was conducted between July and October 2001 in Purworejo District in the Menoreh Hills near the southern coast of Central Java, a region experiencing resurgent malaria at that time.\nThe social, economic and demographic characteristics of malaria in the region are described elsewhere.\nFalciparum and vivax malaria (approximately 1:1) occurs among all ages with prevalence in high-risk areas ranging from 5% to 38%.\nAmong 40 and 54 subjects treated with CQ or SP for uncomplicated falciparum malaria in this region, 47% and 22%, respectively, developed recurrent parasitaemia within 28 days.\nDuring the close study follow up, none of the treatment failures developed complicated or severe malaria.\nStudy subjects, screening and enrollment\nSubject enrollment is depicted in Figure 1.\nSubjects with uncomplicated falciparum malaria were recruited, in part, by mass blood survey and by passive case detection (PCD) at local government-run clinics.\nThis was necessary because few adults in this study area suffered from symptomatic malaria.\nAfter informed consent, all subjects were screened by medical history, physical examination, semi-quantitative glucose-6-phosphate dehydrogenase (G6PD) assay (G-6-PDH Screening Test 203-A, Sigma Diagnostics\u00ae, St. Louis, MO USA), and if female of child-bearing potential, a urine human chorionic gonadotropin (hCG) test (TestPack\u00ae +Plus\u2122 hCG Urine, Abbott, USA).\nInclusion criteria included age \u2265 15, asexual stage P. falciparum parasite density \u2265 400/\u03bcl on Giemsa thick smear examined by standard microscopy, and availability for 28-day follow-up.\nExclusion criteria included pregnancy, breast feeding, body weight < 40 kg, G6PD deficiency, history of antimalarial or antibiotic consumption during the previous 7 days, severe or complicated malaria, history of allergy or adverse reaction to study medications, and P. vivax or mixed species infection.\nAny parasitemic potential study subject not meeting these criteria was referred back to the cooperating clinic for standard therapy.\nThose patients deemed to meet enrollment criteria and willing to participate were assigned a sequential study subject code by the screening physician.\nThe study subject codes were pre-assigned to treatment arms by a random process and all treatment was pre-packaged accordingly.\nTreatment and follow-up\nSubjects were randomized to one of four treatment arms: (A) CQ 25 mg base/kg (body weight) over three days, (B) CQ 25 mg/kg over three days + SP 25/1.25 mg/kg single dose, (C) CQ 25 mg/kg over three days + SP 25/1.25 mg/kg single dose + primaquine 45 mg single dose on day 0, or (D) CQ 25 mg/kg over three days + SP 25/1.25 mg/kg single dose + primaquine 45 mg single dose on day 2.\nStudy investigators directly observed and documented administration of each dose of medication: CQ (Resochin\u00ae; chloroquine diphosphate tablets 150 mg base; PT Bayer, Indonesia) in three once daily doses of 10, 10 and 5 mg base/kg on days 0,1, and 2 respectively; SP (Fansidar\u00ae tablets; 500 mg sulfadoxine/25 mg pyrimethamine; Hoffman La Roche, Indonesia) in a single dose on day 0; Primaquine phosphate (generic tablets, 15 mg base; PT Kimia Farma, Bandung, Indonesia) in a single dose of 3 tablets on day 0 or day 2.\nSubjects were followed for 28 days after initiating therapy.\nStudy personnel visited subjects on 10 separate days (1, 2, 3, 4, 7, 11, 14, 18, 21 and 28) to assess symptoms, clinical recovery, adverse events and to obtain finger-stick blood samples for preparation of Giemsa-stained blood smears.\nSubjects complaining of illness on any day of follow-up were immediately brought to clinic and evaluated.\nBlood blot specimens (100 \u2013 200 \u03bcl) were also prepared on days 0, 2, 7, 14, 21, 28 or day of recurrent parasitaemia from fingerstick blood samples collected by heparinized, 100 \u03bcl micro-capillary tubes and expelled onto Whatman No. 1 filter paper (Whatman International, United Kingdom) for drying and storage for later analysis.\nMicroscopy\nThe microscopists in this study were certified as competent by a standardized, proctored examination.\nMicroscopists examined all ocular fields of standard Giemsa (Sigma, USA) stained thick smears at high power (1000\u00d7 oil immersion) for asexual and sexual forms of P. falciparum and other plasmodial species, and recorded sexual and asexual parasite densities as number of parasites per 200 white blood cells.\nParasite densities are reported as parasites/\u03bcL assuming an average white blood cell count of 8,000/\u03bcl.\nMerozoite Surface Protein-2 (msp-2) genotyping\nMsp-2 genotyping was performed on blood blot samples from day 0 and day of recurrent P. falciparum parasitaemia following Felger et al..\nRecrudescence was defined by the presence of the same genotype on day 0 and day of recurrent parasitaemia while a new infection (i.e. reinfection) was defined by the presence of a different genotype on day 0 and day of recurrent parasitaemia.\nDetermination of therapeutic outcomes and data analysis\nTherapeutic efficacy (parasitological cure) was defined as the proportion of subjects reaching day-28 without recrudescence.\nTherapeutic failures occurred when asexual parasitaemia increased between day 0 and day 2, failed to clear by day 4, or recurred between days 4 and 28.\nThese data were used to determine the crude therapeutic efficacy of each regimen.\nInfections initially classified as therapeutic failures but later proven to be reinfections by msp-2 genotyping were reclassified as therapeutic successes with abbreviated follow-up (until the point of censure).\nThese results were used to determine the \"adjusted\" efficacies of CQ and CQ + SP.\nEpiInfo 2000 1.1.2 (CDC, Atlanta USA) and SPSS version 9.0 (SPSS, Inc. USA) were used to calculate confidence intervals for odds ratios (OR) and the 95% confidence interval around those ratios, and for estimating the p-value using the Yates' corrected or Fisher's exact test when appropriate.\nCumulative incidence of therapeutic failure was estimated by actuarial analysis as described elsewhere.\nResults\nEnrollment\nAmong 124 parasitaemic persons identified during mass blood screening and passive case detection from outpatient clinics, 117 were enrolled and randomized to one of the 4 treatment regimens.\nSeven subjects did not meet study eligibility criteria and were returned to the referring clinic for treatment for the following reasons: taking quinine (1), pregnancy (1), G6PD deficiency (1), low body weight (2), declined participation (1) and administrative error (1).\nDemographics\nAmong the 117 subjects enrolled and randomly assigned, no statistically significant demographic differences between the four study arms were identified; age ranged from 16 to 65 years, and the male to female ratio was 1.2:1 (Table 2).\nNo statistically significant differences were noted between treatment arms with regards to fever rates (range 24\u201341%) or geometric mean asexual parasite densities (range 653-1465/\u03bcL) before enrollment.\nParasite and fever clearance times\nThere were no significant differences in parasite or fever clearance times among the 4 treatment groups.\nVirtually all (115 of 117) subjects cleared parasitaemia by day 3 regardless of the therapy assignment, and all 38 subjects with fever prior to therapy had a normal axillary temperature by day 2 (Figure 2).\nCrude therapeutic efficacy of CQ vs. CQ + SP\nAsexual parasitemia outcomes were similarly efficacious among the three groups receiving (CQ+SP, CQ+SP+PQ0, and CQ+SP+PQ2), so these groups were combined specifically for comparison with CQ monotherapy against asexual blood stages.\nAmong the 29 subjects treated with CQ alone, 1 (3%) withdrew from the study after day 18 and 3 (10%) developed intercurrent P. vivax during the second and third week of post-therapy observation.\nFourteen (48%) completed the study without recurrent parasitaemia.\nThe remaining 11 (38%) subjects developed recurrent P. falciparum during the 28-day follow-up.\nThe crude cumulative incidence of failure of CQ therapy for uncomplicated P. falciparum was 42% (Table 3); therefore, the crude therapeutic efficacy was 58%.\nAmong the 88 subjects treated with CQ + SP, six (7%) withdrew before day 28 (three declined further participation between days two and 21 and three moved away from the study site between days one and 18).\nOne subject (1%) developed intercurrent P. vivax (day 22).\nSeventy-six (86%) completed the study without recurrent parasitaemia.\nThe remaining 5 (6%) subjects developed recurrent P. falciparum during follow-up, therefore providing a crude therapeutic efficacy of 94%.\nThe crude 28-day cumulative incidence of failure of CQ + SP therapy was 6% (Table 3).\nThe relative risk of treatment failure with CQ compared to CQ + SP among those who completed the 28-day observation period was 7.13 (95% CI 2.74 \u2013 18.57, p = 0.00003).\nAdjusted efficacy of CQ vs. CQ + SP\nThree of 11 apparent therapeutic failures among subjects receiving CQ alone could not be evaluated because P. falciparum DNA could not be amplified for msp-2 genotyping, leaving eight evaluable recurrent infections.\nOne case of day 28 recurrent parasitaemia was reclassified as a cure with late reinfection based on differing day 0 and day of recurrence msp-2 genotypes.\nFrom the remaining seven recurrent infections, the parasite genotypes matched and their classification as recrudescence affirmed.\nThe adjusted cumulative incidence of failure of CQ therapy for uncomplicated P. falciparum was therefore 30% (therapeutic efficacy 70%) (Table 4).\nIn the combined CQ + SP groups, one of five recurrent parasitaemias was excluded from analysis because parasite DNA could not be amplified, leaving four evaluable recurrent infections.\nBased on discordant day 0 and day of recurrence msp-2 genotypes, three apparent therapeutic failures were reclassified as successful cures with reinfection.\nThe 28-day cumulative incidence of failure of CQ+SP therapy for uncomplicated P. falciparum was therefore only 1% (therapeutic efficacy = 99%) (Table 4).\nSubjects treated with CQ alone were 25 times more likely to suffer therapeutic failure compared to CQ combined with SP (95% CI 3.3 \u2013 196.06; p = 0.0006).\nEffect of primaquine on gametocytaemia\nFigure 3 illustrates the course of gametocyte rates among groups after treatment.\nComparing the two groups receiving primaquine, gametocyte rates declined steadily to 0% by day 11 after therapy and did not reappear during follow-up.\nGametocyte rates declined more slowly in the non-primaquine groups.\nThe difference in gametocyte clearance rates on day 11 between groups receiving primaquine (0%) and those not (33%) was significant (p = 0.025).\nA single dose of primaquine greatly improved the clearance rate of gametocytes, and administering the dose on day 2 vice day 0 accelerated clearance time (0% vs. 7% on day 7); however this difference was not statistically significant.\nConclusion\nThe combination of CQ+SP proved 99% effective among 87 subjects with uncomplicated falciparum malaria in Central Java, Indonesia.\nSimilarly excellent efficacy has been reported from other regions, including Papua New Guinea.\nHowever, in eastern Indonesia, where resistance to CQ (up to 95%) and SP (range 15\u201354%) is much more prevalent than in Central Java, the combination of CQ + SP fails in 38% to 55% (Maguire, personal communication) of subjects with uncomplicated falciparum malaria.\nPoor efficacy of CQ+SP has been reported elsewhere as well and studies in The Gambia, Uganda, Laos and Nigeria, for example, showed similarly poor efficacy or negligible difference with combining CQ and SP.\nThe apparent wide range of efficacy of CQ+SP requires information-based decisions about its use as a therapeutic option in any given area.\nDespite the marked differences between day-28 cure rates, CQ vs. CQ+SP regimens had essentially identical parasite and fever clearance times.\nThese two indices parallel clinical recovery and almost certainly exacerbate the persistent over-the-counter marketing and unsupervised use of CQ monotherapy.\nIndeed, surveys of knowledge, attitudes and practices in the region of study affirmed this practice as common.\nMoreover, CQ remains first-line therapy against vivax malaria, even though accurate diagnostic services to distinguish these species are available to relatively few people.\nFor all of these reasons, CQ remains a widely used drug against P. falciparum and other species of malaria in Indonesia.\nCQ+SP can not be suggested as a first-line treatment policy that would be supported by comprehensive national programs of patient awareness, health provider training, and widespread distribution and marketing.\nThe efficacy of the combination may not be sufficiently long-lived to support these costly adjustments and measures.\nHowever, CQ combined with SP may be a feasible therapeutic option in regions with evidence of treatment efficacy where health care providers are unable to use ACT because of problems with availability, fear of side effects or counterfeit drug.\nUse should be limited to areas like Central Java, where data suggests that CQ+SP efficacy was better than 95%.\nPrimaquine has proven gametocytocidal activity against P. falciparum.\nA single 45 mg dose reduces gametocyte clearance time to 2 to 3 days, nearly one week faster than with other asexual stage antimalarials alone.\nThe addition of primaquine to standard blood schizonticidal therapy significantly reduced the point prevalence of malaria in one study.\nVarious primaquine regimens appear equally efficacious; however, its use may be limited by the frequency of G-6-PD deficiency and sporadic supplies.\nIn our study, addition of a single 45 mg dose of primaquine on either day 0 or day 2 significantly offset the persistence of gametocytaemia.\nCurrent P. falciparum malaria treatment policy in Indonesia includes administration of 45 mg of primaquine on day 0 of treatment in hypoendemic areas.\nThis study supports its continued use in Indonesia, especially with combination regimens containing SP, which has been linked to sulfonamide-associated gametocyte proliferation.\nCombining artesunate with SP similarly reduced gametocyte rates below that observed with SP alone in one study in Indonesia; similarly, artesunate combined with amodiaquine significantly reduced gametocyte carriage in comparison to CQ+SP in Nigerian children.\nIn summary CQ+SP combination was highly effective for parasitological cure in subjects with uncomplicated P. falciparum in Central Java, Indonesia.\nThe regimen was safe and efficacious, and it is inexpensive and readily available throughout Indonesia.\nA single dose of 45 mg primaquine, either on day 0 or day 2, suppressed or quickly eliminated gametocytaemia.\nGiven wide regional differences in efficacy of CQ+SP, this regimen should be evaluated where its use is advocated as a therapeutic option for falciparum malaria.\nThe recruitment, enrollment and randomization process for this study evaluating four treatment regimens for uncomplicated P. falciparum in residents of Purworejo District, Central Java, Indonesia.\n(A) Asexual parasite clearance time in 117 subjects with falciparum malaria after treatment with one of 4 different regimens. (B) Time of fever clearance in 38 subjects with fever at the time of study enrollment. CQ = chloroquine, SP = sulfadoxine/pyrimethamine, and PQ = primaquine on either day 0 (d0) or day 2(d2) of therapy.\nGametocyte rates in 117 subjects with falciparum malaria after treatment with one of 4 different regimens. CQ = chloroquine, SP = sulfadoxine/pyrimethamine, and PQ = primaquine on either day 0 (d0) or day 2(d2) of therapy.\n\nComparison of clinical trials to evaluate the efficacy of chloroquine (CQ) + sulfadoxine/pyrimethamine (SP)\nStudy Location | Year | N | CQ Efficacy | SP Efficacy | CQ + SP Efficacy | Endpoint in Days | Comparison of CQ+SP vs. CQ, CQ+SP vs. SP | Reference\nNigeria | 2004 | 153 | n/a | n/a | 90% | 28 | n/a | [30]\n\nLaos | 2003 | 110 | n/a | n/a | 93% | 42 | n/an/a | [14]\n\nBangladesh | 2002 | 133 | n/a | n/a | 72% | 28 | n/a | [18]\n\nEastern Uganda | 2002 | 280 | n/a | 42% | 60% | 28 | n/ap = ns | [21]\n\nSouthern Uganda | 2002 | 448 | n/a | 70% | 83% | 14 | n/ap < 0.001 | [12]\n\nWestern Uganda | 2001 | 141 | 93% | 100% | 100% | 14 | p = nsp = ns | [13]\n\nSouthern Laos | 2001 | 119 | 55% | 82% | 83% | 14 | p = 0.029p = ns | [19]\n\nPapua, Indonesia | 1999 | 169 | 17% | 79% | 62% | 28 | p < 0.001p = 0.047 | [17]\n\nSouthwest Nigeria | 1999 | 111 | 93% | n/a | 96% | 28 | n/ap = ns | [22]\n\nPapua New Guinea | 1998\u20131999 | 513* | n/a | n/a | 95% | 28 | n/a | [3]\n\nThe Gambia | 1995 | 405 | n/a | 90% | 95% | 28 | n/ap = ns | [20]\n\nMumbai, India | N/a | 49 | 36% | n/a | 87% | 42 | p < 0.026n/a | [2]\n\nns = not significant\nn/a = not available\n* children were treated with amodiaquine + SP (303 of the 513), ** excluding children\n\nDemographic, clinical and parasitological characteristics of subjects treated with 4 malaria regimens in Purworejo District, Indonesia\n | Treatment group | Summary statistic\n |  | \n | CQ alone | CQ + SP | CQ + SP + PQ0 | CQ + SP + PQ2 | \nSample size | 29 | 32 | 28 | 28 | n/a\nMale: Female | 1.2 : 1 | 1.5 : 1 | 1 : 1.2 | 1.3 : 1 | p = 0.771a\nMean age (range) | 36 (17\u201355) | 40 (20\u201360) | 40 (20\u201362) | 35 (16\u201365) | p = 0.256b\nGeometric mean asexual parasite density (95% CI) | 653 (382\u20131118) | 1465 (849\u20132531) | 1148 (582\u20132263) | 1409 (720\u20132760) | p = 0.286b\nFever Rate | 24% | 41% | 29% | 36% | p = 0.529a\n\nCQ = chloroquine, SP = sulfadoxine/pyrimethamine, PQ0 = primaquine on day 0, PQ2 = primaquine on day 2, 95% CI = 95% confidence interval\nn/a \u2013 not applicable\na) Chi-square test\nb) Kruskal-Wallis test\n\nLife Table representation of cumulative incidence of crude therapeutic failure attributable to drug resistance among subjects treated with chloroquine (CQ) and CQ + sulfadoxine/pyrimethamine (SP) without distinguishing recrudescence from reinfection\nCQ | CQ+SP\nD | N | I | W | IR | CIF | D | N | I | W | IR | CIF\n\n0 | 29 | 0 | 0 | 0.00 | 0.00 | 0 | 88 | 0 | 0 | 0.00 | 0.00\n2 | 29 | 0 | 0 | 0.00 | 0.00 | 2 | 88 | 0 | 2 | 0.00 | 0.00\n4 | 29 | 0 | 0 | 0.00 | 0.00 | 4 | 86 | 1 | 1 | 0.01 | 0.01\n7 | 29 | 0 | 0 | 0.00 | 0.00 | 7 | 84 | 1 | 0 | 0.01 | 0.02\n11 | 29 | 1 | 0 | 0.03 | 0.03 | 11 | 83 | 0 | 0 | 0.00 | 0.02\n14 | 28 | 2 | 1 | 0.07 | 0.10 | 14 | 83 | 0 | 0 | 0.00 | 0.02\n18 | 25 | 2 | 1 | 0.08 | 0.18 | 18 | 83 | 0 | 2 | 0.00 | 0.02\n21 | 22 | 2 | 2 | 0.10 | 0.26 | 21 | 81 | 0 | 1 | 0.00 | 0.02\n28 | 18 | 4 | 0 | 0.22 | 0.42 | 28 | 80 | 3 | 1 | 0.04 | 0.06\n\nN = sample at risk of therapeutic failure at start of interval\ni = the number of cases of therapeutic failure during the interval\nw = the number of subjects withdrawn from the test during the interval due to exclusion from analysis, loss to follow-up after latest blood smear, intercurrent infection with another plasmodia species, or reinfection by different genotype P. falciparum or relapse/reinfection by P. vivax based on day of recurrent parasitemia CQ +DCQ concentrations\nIR = risk of therapeutic failure in the interval, calculated as i/(N-w/2)\nCIF = cumulative incidence of therapeutic failure, calculated as 1-[(1-IRn)(1-CIFn-1)], where IRn is the IR in the current interval and CIFn-1 is the cumulative incidence up to the previous interval\n\nLife Table representation of cumulative incidence of adjusted therapeutic failure attributable to drug resistance among subjects treated with chloroquine (CQ) and CQ + sulfadoxine/pyrimethamine (SP) with msp-2 genotyping to distinguish recrudescence from reinfection\nCQ | CQ+SP\nD | N | I | W | IR | CIF | D | N | I | W | IR | CIF\n\n0 | 26 | 0 | 0 | 0.00 | 0.00 | 0 | 87 | 0 | 0 | 0.00 | 0.00\n2 | 26 | 0 | 0 | 0.00 | 0.00 | 2 | 87 | 0 | 2 | 0.00 | 0.00\n4 | 26 | 0 | 0 | 0.00 | 0.00 | 4 | 85 | 1 | 1 | 0.01 | 0.01\n7 | 26 | 0 | 0 | 0.00 | 0.00 | 7 | 83 | 0 | 0 | 0.00 | 0.01\n11 | 26 | 1 | 0 | 0.04 | 0.04 | 11 | 83 | 0 | 0 | 0.00 | 0.01\n14 | 25 | 1 | 1 | 0.04 | 0.08 | 14 | 83 | 0 | 0 | 0.00 | 0.01\n18 | 23 | 1 | 1 | 0.04 | 0.12 | 18 | 83 | 0 | 2 | 0.00 | 0.01\n21 | 21 | 1 | 2 | 0.05 | 0.16 | 21 | 81 | 0 | 1 | 0.00 | 0.01\n28 | 18 | 3 | 0 | 0.17 | 0.30 | 28 | 80 | 0 | 2 | 0.00 | 0.01\n\nN = sample at risk of therapeutic failure at start of interval\ni = the number of cases of therapeutic failure during the interval\nw = the number of subjects withdrawn from the test during the interval due to exclusion from analysis, loss to follow-up after latest blood smear, intercurrent infection with another plasmodia species, or reinfection by different genotype P. falciparum or relapse/reinfection by P. vivax based on day of recurrent parasitemia CQ +DCQ concentrations\nIR = risk of therapeutic failure in the interval, calculated as i/(N-w/2)\nCIF = cumulative incidence of therapeutic failure, calculated as 1-[(1-IRn)(1-CIFn-1)], where IRn is the IR in the current interval and CIFn-1 is the cumulative incidence up to the previous interval",
        "label": "unclear",
        "id": "task4_RLD_test_929"
    },
    {
        "paper_doi": "10.1371/journal.pmed.1002433",
        "bias": "blinding of participants and personnel (performance bias) all outcomes",
        "PICO": "Methods: Non-blinded, cluster-RCT. 10 clinics in Mozambique were randomized to either intervention (CIS) or control (standard care). A pre-post intervention 2-sample design was nested within the intervention arm to assess the additional effectiveness of an enhanced version of the CIS, referred to as CIS+. Consequently, the standard care arm enrolled 1 cohort of patients, while the intervention arm enrolled 2 sequential cohorts of patients (CIS and CIS+). CIS+ participants were enrolled after CIS enrolment was completed at each facility randomized to the intervention arm.\n\n\nParticipants: 5327 participants5 clinics were selected from urban areas and 5 from rural areasInclusionAll adults testing HIV-positive in the VCT clinics within the participating health facilitiesExclusion< 18 years of age.Pregnant.Planned to move from their community of residence in the next 12 months.Had enrolled in HIV care or initiated ART in the past 6 months.Did not understand Portuguese or Xitsua.Were incapable of providing informed consent.\n\n\nInterventions: Intervention arm: 4 evidence-based interventions that simplified the clinic flow and encouraged linkage to and retention in care:Pima (Inverness Medical Innovations) CD4 assay machines in the VCT clinics to enable HIV testing counsellors to provide real-time, POC CD4 test results immediately following diagnosis.Participants with Pima CD4 cell count < 350 cells/mm3 were provided with rapid ART initiation. These individuals received an individual ART preparatory counselling session in the VCT clinic immediately following CD4 testing, on the day of diagnosis. Facility receptionists were instructed to expedite appointments for these participants when they presented to schedule their clinical consultations. Clinicians were encouraged to initiate ART at the first clinical visit.Participants received health messages and appointment reminders via SMS messaging.participants in the CIS+ cohort received the CIS interventions plus a series of non-cash FIs in the form of prepaid cellular air-time cards.Control: standard care - participants were managed as per prevailing Ministry of Health guidelines.Individuals diagnosed with HIV received post-test counselling in the VCT clinic and were referred verbally to HIV services.Participants presenting to the facility receptionist to schedule a clinical consultation for HIV care were referred to the laboratory for CD4 cell count, chemistry, and haematology testing, and provided with an appointment 2-4 weeks later to allow sufficient time for the laboratory results to be received.ART eligibility was determined at that first clinical consultation based on CD4 cell count < 350 cells/mm3 and/or WHO stage 3/4.Those found to be eligible for ART received at least 1 individual counselling session before initiating treatment.For ART-eligible participants, the time interval between enrolment in HIV care and ART initiation was estimated at 1-2 months at the time the study started.Participants initiating ART were requested to return every 2 weeks for the first month, at 2 months, at 6 months, and every 6 months thereafter.ART-ineligible participants were instructed to return at 6 months for repeat clinical evaluation and laboratory testing.\n\n\nOutcomes: Mortality and viral suppression at 12 months, time to ART initiation, linkage to care at 1 month, retention in care at 6 months, disease progression\n\n\nNotes: \n\n",
        "objective": "To assess the effects of interventions for rapid initiation of ART (defined as offering ART within seven days of HIV diagnosis) on treatment outcomes and mortality in people living with HIV. We also aimed to describe the characteristics of rapid ART interventions used in the included studies.",
        "full_paper": "Background\nConcerning gaps in the HIV care continuum compromise individual and population health.\nWe evaluated a combination intervention strategy (CIS) targeting prevalent barriers to timely linkage and sustained retention in HIV care in Mozambique.\nMethods and findings\nIn this cluster-randomized trial, 10 primary health facilities in the city of Maputo and Inhambane Province were randomly assigned to provide the CIS or the standard of care (SOC).\nThe CIS included point-of-care CD4 testing at the time of diagnosis, accelerated ART initiation, and short message service (SMS) health messages and appointment reminders.\nA pre\u2013post intervention 2-sample design was nested within the CIS arm to assess the effectiveness of CIS+, an enhanced version of the CIS that additionally included conditional non-cash financial incentives for linkage and retention.\nThe primary outcome was a combined outcome of linkage to care within 1 month and retention at 12 months after diagnosis.\nFrom April 22, 2013, to June 30, 2015, we enrolled 2,004 out of 5,327 adults \u226518 years of age diagnosed with HIV in the voluntary counseling and testing clinics of participating health facilities: 744 (37%) in the CIS group, 493 (25%) in the CIS+ group, and 767 (38%) in the SOC group.\nFifty-seven percent of the CIS group achieved the primary outcome versus 35% in the SOC group (relative risk [RR]CIS vs SOC = 1.58, 95% CI 1.05\u20132.39).\nEighty-nine percent of the CIS group linked to care on the day of diagnosis versus 16% of the SOC group (RRCIS vs SOC = 9.13, 95% CI 1.65\u201350.40).\nThere was no significant benefit of adding financial incentives to the CIS in terms of the combined outcome (55% of the CIS+ group achieved the primary outcome, RRCIS+ vs CIS = 0.96, 95% CI 0.81\u20131.16).\nKey limitations include the use of existing medical records to assess outcomes, the inability to isolate the effect of each component of the CIS, non-concurrent enrollment of the CIS+ group, and exclusion of many patients newly diagnosed with HIV.\nConclusions\nThe CIS showed promise for making much needed gains in the HIV care continuum in our study, particularly in the critical first step of timely linkage to care following diagnosis.\nTrial registration\nClinicalTrials.gov NCT01930084\nIn a cluster-randomized trial done in Mozambique, Batya Elul and colleagues study a combined intervention for linkage to and retention of people with HIV in care.\nAuthor summary\nWhy was this study done?\nIn sub-Saharan Africa, HIV testing, care, and treatment programs have been widely scaled up over the past decade, but suboptimal outcomes across the HIV care continuum\u2014particularly with regards to timely linkage to and sustained retention in care\u2014compromise their effectiveness.\nPatients experience multiple barriers to linkage to and retention in HIV care including health system barriers, structural barriers, and behavioral barriers, yet prior studies have largely evaluated individual interventions targeting a single barrier to care.\nOur study was designed specifically to examine the effectiveness of a combination intervention strategy (CIS) composed of several scalable evidence-based interventions targeting the multiple and prevalent health system, structural and behavioral barriers that patients face across the HIV continuum.\nWhat did the researchers do and find?\nWe randomly assigned 10 primary health facilities in the city of Maputo and Inhambane Province in Mozambique to provide the standard of care (SOC) or the CIS, which included point-of-care CD4 testing at the time of diagnosis, accelerated ART initiation, and short message service (SMS) health messages and appointment reminders.\nA pre\u2013post intervention 2-sample design was nested within the intervention arm to assess the effectiveness of CIS+, an enhanced version of the CIS that additionally included conditional non-cash financial incentives for linkage and retention.\nWe enrolled 2,004 adults diagnosed with HIV in the voluntary counseling and testing clinics of participating health facilities, and compared the proportion who achieved a combined outcome of linkage to HIV care within 1 month of diagnosis and retention in care at 12 months across the 3 study groups.\nWe found an increased likelihood of achieving the combined outcome in the CIS group compared to the SOC group, driven primarily by very large increases in same-day linkage, but no difference between the CIS+ and CIS groups.\nWhat do these findings mean?\nThe CIS may help improve outcomes across the HIV care continuum in high-burden settings, particularly in the critical first step of timely linkage to care following diagnosis.\nFurther research is needed to understand whether financial incentives can be optimized in this setting, given their effectiveness in enhancing other health outcomes.\nIntroduction\nAlthough the extraordinary scale-up of HIV testing, care, and treatment programs in sub-Saharan Africa over the past decade has resulted in more than 19 million persons accessing antiretroviral therapy (ART), the effectiveness of these programs has been significantly hindered by high levels of attrition across the HIV care continuum.\nObservational studies and systematic reviews have repeatedly reported disturbing gaps in care as patients move from HIV testing clinics to HIV care clinics (i.e., linkage to care) and that patient dropout among those enrolled in HIV care is far too common, both before and after ART initiation (i.e., retention in care).\nIndeed, available data suggest that less than 1/3 of individuals who are diagnosed with HIV are successfully linked to and remain engaged in HIV care 12 months later.\nBarriers to timely linkage to and sustained retention in HIV care have been well documented, and include health system barriers (e.g., multiple HIV clinic visits for counseling and clinical and laboratory assessments prior to ART initiation), structural barriers (e.g., transport costs and distances, work and childcare constraints), and behavioral barriers (e.g., forgetting appointments, lack of understanding of required care).\nPrior studies have overwhelmingly evaluated individual interventions targeting a single barrier at a single point in the HIV care continuum such as mobile phone short message service (SMS) messaging to augment linkage to care following diagnosis, or point-of-care CD4 testing to enhance retention among patients enrolled in HIV care.\nHowever, it is increasingly recognized that multi-component approaches composed of several practical, evidence-based interventions that simultaneously target the multiple and recurrent barriers that patients face as they navigate across the HIV care continuum are needed to maximize individual and population health.\nFurther, implementation science research that evaluates proposed multi-component approaches in real-world settings is needed to assess not only effectiveness, but also implementation outcomes including reach, adoption, and sustainability.\nTo this end, we designed a combination intervention strategy (CIS) composed of several scalable evidence-based interventions targeting prevalent health system, structural, and behavioral barriers across the HIV care continuum, and determined its effect on a combined outcome of linkage to and retention in HIV care among adults newly diagnosed with HIV in Mozambique, while also collecting information on its implementation and potential for broader scale-up.\nData regarding intervention feasibility and patient acceptability have been published, and thus we present here the effectiveness results.\nBecause the interventions included in the CIS are expected to be implemented at the facility level, as opposed to targeted at specific individuals, should they be scaled up, we evaluated effectiveness using a cluster design, which best mirrors this implementation approach.\nMethods\nA detailed description of the study protocol has been published.\nEthics statement\nEthical approval was provided by Mozambique\u2019s National Committee for Bioethics for Health and Columbia University\u2019s institutional review board (IRB) (protocol AAAL1354).\nInformed written consent was obtained from all participants.\nStudy design\nBetween April 22, 2013, and June 30, 2016, we conducted a 2-arm cluster-randomized study (effectiveness\u2013implementation hybrid design, Type 1) in health facilities in Maputo and Inhambane Province in Mozambique in order to assess the effectiveness of the CIS.\nAdditionally, a pre\u2013post intervention 2-sample design was nested within the intervention arm to assess the additional effectiveness of an enhanced version of the CIS, referred to as CIS+.\nConsequently, the standard of care (SOC) arm enrolled 1 cohort of patients, while the intervention arm enrolled 2 sequential cohorts of patients (CIS and CIS+).\nCIS+ participants were enrolled after CIS enrollment was completed at each facility randomized to the intervention arm.\nStudy setting\nThe city of Maputo, the nation\u2019s capital, has an area of 300 km2 and an estimated population of 1,225,868, with an HIV prevalence of 16.9% among those aged 15 to 59 years.\nThe Maputo City Health Network has a total of 37 health facilities, 32 of which offered comprehensive HIV care and treatment services at the time of study implementation.\nIn contrast, Inhambane is a rural province, with an estimated 1,475,318 people spread across 68,615 km2.\nHIV prevalence among adults aged 15 to 59 years is 14.1%.\nThe ratio of doctors to population (5.96/100,000) is one of the lowest in the country.\nOf the 135 health facilities in the province, 76 offered HIV care and treatment services when our study was initiated.\nSuboptimal health facility infrastructure, long distances to facilities, and weak referral systems in the province are all believed to compromise health service uptake.\nRandomization\nPrimary health facilities providing HIV testing, care, and treatment services and operated by the Ministry of Health with technical support from the Center for Collaboration in Health, a local PEPFAR implementing partner, were the unit of randomization.\nWe focused on primary health facilities, rather than larger provincial hospitals, to reflect the increasingly decentralized nature of HIV service delivery in Mozambique.\nTen facilities in Maputo (N = 4) and Inhambane Province (N = 6) were selected from the 66 primary health facilities receiving technical support from the Center for Collaboration in Health in those regions.\nParticipating facilities were purposely chosen because they had the highest volume of adults testing HIV positive and enrolling in HIV care in the year prior to study start and thus were expected to have sufficient participants for appropriate power.\nFacilities were matched into pairs by region (Maputo or Inhambane), level of urbanicity (urban versus rural), and average number of patients testing HIV positive in voluntary counseling and testing (VCT) in the year prior to study initiation (high versus low), resulting in 5 matched pairs.\nMatched pairs were randomized by one of the authors (MRL) using a computerized random number generator to either the CIS arm or the SOC arm using matched-pair randomization.\nSequences were concealed until interventions were assigned.\nThe study was non-blinded.\nStudy population\nParticipants were enrolled in the SOC group beginning on April 22, 2013, and in the CIS group beginning on April 25, 2013.\nThe last patient was enrolled in the SOC group on November 20, 2014, and the last patient in the CIS group was enrolled on February 11, 2015.\nEnrollment in the CIS+ group began after each clinic randomized to the intervention arm completed CIS enrollment, and ran from June 16, 2014, through June 30, 2015.\nAll participants were followed for 12 months, with the last patient completing follow-up on June 30, 2016.\nBroad inclusion criteria were used to reflect as accurately as possible the population of adults newly diagnosed with HIV in VCT clinics at the participating health facilities.\nWe focused on individuals newly diagnosed in VCT clinics, as opposed to those diagnosed in antenatal clinics and tuberculosis clinics, because the latter groups of patients typically follow a modified clinic flow.\nAll adults testing HIV positive in the VCT clinics within the participating health facilities were informed of the study by HIV testing counselors following diagnosis, and those who were interested were referred to study staff for further information, eligibility screening, and consent procedures.\nPatients were excluded if they were less than 18 years of age, were pregnant, planned to move from their community of residence in the next 12 months, had enrolled in HIV care or initiated ART in the past 6 months, did not understand Portuguese or Xitsua, or were incapable of providing informed consent.\nStudy participants agreed to be referred to HIV care and treatment services at the same facility where they were diagnosed (referred to as the \u201cdiagnosing facility\u201d); to complete a baseline, 1-month, and 12-month interview; to be traced at their homes if they could not be reached by phone for follow-up interviews; to provide contact information for a family member or friend who could provide information on their vital status if they could not be located for a follow-up interview; and, if they enrolled in HIV care and treatment services at the diagnosing facility, to have their clinical data abstracted from the facility\u2019s existing electronic medical records.\nStudy interventions\nStandard of care\nParticipants at health facilities randomized to receive the SOC were managed as per prevailing Ministry of Health guidelines.\nIndividuals diagnosed with HIV received post-test counseling in the VCT clinic and were referred verbally to HIV services, typically in the diagnosing facility.\nPatients presenting to the facility receptionist to schedule a clinical consultation for HIV care were referred to the laboratory for CD4 cell count, chemistry, and hematology testing, and provided with an appointment 2\u20134 weeks later to allow sufficient time for the laboratory results to be received.\nART eligibility was determined at that first clinical consultation based on CD4 cell count \u2264 350 cells/mm3 and/or WHO stage 3/4.\nThose found to be eligible for ART received at least 1 individual counseling session before initiating treatment.\nFor ART-eligible patients, the time interval between enrollment in HIV care and ART initiation was estimated at 1\u20132 months at the time the study started.\nParticipants initiating ART were requested to return every 2 weeks for the first month, at 2 months, at 6 months, and every 6 months thereafter.\nART-ineligible patients were instructed to return at 6 months for repeat clinical evaluation and laboratory testing.\nCombination intervention strategy\nAt facilities randomized to the intervention arm, we introduced 4 evidence-based interventions that simplified the clinic flow and encouraged linkage to and retention in care.\nThese interventions targeted several known health system, structural, and behavioral barriers across the HIV care continuum, and were adapted for the on-the-ground realities\u2014including practice norms, physical space, and available staffing\u2014at the health facilities.\nFirst, we introduced Pima (Inverness Medical Innovations) CD4 assay machines in the VCT clinics to enable HIV testing counselors to provide real-time, point-of-care CD4 test results immediately following diagnosis, and thus addressed a health system barrier by reducing the number of visits required for CD4 testing.\nWe also hypothesized that receipt of additional information on one\u2019s health at the time of diagnosis would advance patient understanding of the need for care, a documented behavioral barrier.\nAll patients regardless of CD4 count were provided with a paper-based referral to on-site HIV services that included their CD4 count, and were instructed to present for their first clinical consultation within 1 week.\nSecond, to address additional health system barriers, patients with Pima CD4 cell count \u2264 350 cells/mm3 were provided with accelerated ART initiation, with the ultimate goal of decreasing the HIV morbidity and mortality that contributes to significant attrition among ART-eligible patients.\nThese individuals received an individual ART preparatory counseling session in the VCT clinic immediately following CD4 testing, on the day of diagnosis.\nFacility receptionists were instructed to expedite appointments for these patients when they presented to schedule their clinical consultations.\nAlthough the patients were directed to the laboratory to have their blood drawn for baseline laboratory tests required by national ART guidelines, clinicians were encouraged to initiate ART at the first clinical visit rather than await the results of the laboratory tests unless the patient presented with comorbid conditions.\nPatients who initiated ART received a 2-week supply and followed the visit schedule dictated by national guidelines, similar to the SOC procedures.\nOnce baseline laboratory results were available, they were reviewed by clinic staff, and if abnormalities were noted, the participant was contacted to return to the clinic.\nThird, participants received health messages and appointment reminders via SMS messaging to address behavioral barriers associated with deferring care engagement and forgetting appointments.\nThe messages were sent from the central study office to the participant\u2019s phone or to a friend or relative\u2019s phone per participant preference, and did not refer to HIV or a specific health facility or reveal any personal information.\nThe health messages encouraged participants to care for their health, and were sent weekly for 1 month following diagnosis and then monthly (e.g., \u201cHi.\nYour health is the most important thing.\nPlease remember to come to the health center for health services.\u201d).\nAppointment reminders were sent only to participants who linked to care at the diagnosing facility, and were sent 3\u20137 days before each scheduled clinic visit (e.g., \u201cHi.\nYour health is the most important thing.\nWe expect to see you at your upcoming appointment scheduled for the day ___.\u201d).\nParticipants were not asked to confirm receipt or reply to the messages.\nFinally, patients in the CIS+ cohort received the CIS interventions plus a series of non-cash financial incentives (FIs) in the form of prepaid cellular air-time cards to offset structural barriers associated with the direct and indirect costs of coming to the health facility to receive HIV care.\nAir-time cards rather than cash were selected as the incentive based on discussion with the Ministry of Health.\nEach card was valued at approximately US$5 and was provided conditionally upon the following achievements: linkage to care within 1 month of diagnosis, retention in care 6 months after diagnosis, and retention in care 12 months after diagnosis, for a total of approximately US$15.\nParticipants who completed each achievement received the card when presenting for routine services.\nParticipants without cellular phones could opt to give them to a family member, sell them for cash, or trade them for other goods.\nBoth the point-of-care CD4 testing and accelerated ART initiation interventions were provided by health facility staff to all individuals diagnosed with HIV in the VCT clinic regardless of whether they were enrolled in the study, while the SMS messages and FIs were provided by study staff and only to study participants.\nData collection and outcomes\nSite assessments\nData on the configuration of HIV services at the 10 participating study sites were collected at the beginning and at the end of the study using a standardized site assessment form.\nThe purpose of the site assessments was to identify important similarities and differences between participating health facilities, as well as to better understand how services at the site could impact study implementation.\nBaseline interview\nParticipants completed closed-ended questionnaires administered by trained research assistants at the time of study enrollment.\nThe questionnaire took about 30 minutes to complete, and gathered information on sociodemographic characteristics, social and family support, mental health, alcohol use, HIV testing history, HIV knowledge and beliefs, and anticipated stigma and barriers to care.\nAnticipated stigma was assessed through 6 items adapted from the 12-item anticipated HIV stigma index developed by Earnshaw and Chaudoir.\nStigma scores were summed, then dichotomized into 2 groups: highest (>75th percentile) versus lower anticipated stigma.\nMental health was assessed via a 7-question evaluation based on the Kessler 10-item scale for psychological distress.\nMental health scores were summed, then dichotomized into 2 groups: highest (<75th percentile) versus lower level of distress.\nPerceived availability of social support was assessed with 4 questions adapted from a 9-item scale by Wortman and colleagues.\nSocial support scores were summed, then dichotomized into 2 groups: higher (>50th percentile) versus lower social support.\nQuestions assessing HIV-related knowledge and attitudes were based on those used by one of the authors in a previous study.\nHIV knowledge scores were summed, then dichotomized into 2 groups: higher (>50th percentile) versus lower knowledge.\nBaseline interview data were double-entered into a study database, and a computer program identified discrepant double-entered results for correction against the paper-based forms.\nPatient tracing and follow-up interviews\nOne and 12 months after enrollment, up until June 30, 2016, trained research assistants contacted participants by phone to ascertain their vital status and HIV care status, and to administer follow-up questionnaires.\nIf the participant could not be contacted by phone after 3 attempts, research assistants visited the participant\u2019s home up to 3 times.\nParticipants who were located completed closed-ended interviews that gathered updated information on key domains from the baseline questionnaire, as well as self-reported information on linkage to (1- and 12-month questionnaires) and retention in HIV care (12-month questionnaire only), reasons for linkage/non-linkage (1- and 12-month questionnaires) and retention/non-retention (12-month questionnaire only), ART status, hospitalizations, and anticipated stigma.\nIn cases where the participant could not be located, research assistants contacted a friend or family member as specified by the participant at study enrollment.\nResearch assistants did not refer to HIV or the health facility during contact tracing but rather attempted to determine whether the participant was alive or dead.\nFor those whose vital status could not be determined through contact tracing, research assistants searched existing electronic medical records at other primary health facilities supported by the Center for Collaboration in Health in the same district to assess whether patients had enrolled in HIV care at another facility, and reviewed death registers at the municipal and provincial levels to ascertain their vital status.\nSimilar data entry and reconciliation procedures to those used for the baseline interview data were used for the tracing and follow-up data.\nAbstraction of clinical data for patients linking to HIV care at the diagnosing facility\nAs part of routine clinical practice for HIV patients, clinicians documented patient information at every clinic visit on national HIV care forms, and trained data clerks entered those data into an Access-based electronic medical record.\nIn its role as a PEPFAR implementing partner supporting the study sites, the Center for Collaboration in Health assessed the completeness and accuracy of these electronic data every 4 months and initiated targeted interventions to enhance data quality if there was greater than 15% disagreement on key data elements between the electronic and paper-based systems.\nDuring the study period, research assistants reviewed the electronic medical records to identify study participants who had linked to care at their diagnosing facility.\nFor those located, we extracted the complete electronic medical record, capturing information on visit dates, vital status, transfer status, ART status, laboratory test results, and opportunistic infections.\nOutcomes\nThe primary outcome was a combined outcome of linkage to HIV care within 1 month of diagnosis plus retention in care 12 months after diagnosis measured at the individual level.\nWe used a combined outcome to reflect the fact that improvements are needed across the HIV care continuum in order to maximize individual and population health.\nLinkage to care was defined by at least 1 clinical consultation for HIV that included assessment of the patient\u2019s medical history and a physical exam.\nRetention in care was defined by a clinic visit in the 90 days prior to the end of the 12-month study follow-up period, with no documentation that the patient had transferred to another facility or had died.\nWe assessed the combined outcome from the perspective of the diagnosing health facility using data from the electronic medical records maintained by the HIV clinics.\nAll study participants were included in these analyses, including those who did not complete follow-up interviews.\nParticipants whose electronic medical records were not located were considered not to have achieved the combined outcome for this analysis.\nAs a secondary approach, we evaluated the combined outcome from the perspective of the Mozambican health program by supplementing data from the electronic medical records with patient reports of linkage to and retention in care at HIV clinics at different health facilities (obtained during follow-up interviews) and information obtained from electronic medical records at other health facilities.\nIn these analyses, participants whose self-reported linkage and retention status suggested they were linked to and/or retained at a health facility other than their diagnosing clinic were considered to have achieved the respective linkage/retention outcomes.\nParticipants who either did not complete follow-up interviews or did not self-report linkage to or retention at another clinic maintained their initial outcome designation.\nAll study participants were included in these analyses.\nSecondary outcomes included linkage to care at several predefined time points, ART eligibility assessment (defined as receipt of WHO staging and/or CD4 cell count), ART initiation, disease progression (defined as a new WHO stage 3/4 condition or hospitalization noted in the electronic medical records or self-reported during follow-up interviews), retention in care 6 and 12 months after diagnosis regardless of the timing of linkage, and death.\nStatistical analysis\nThe trial was designed and powered to measure outcomes at the individual level, with outcomes assessed within each cluster (5 clusters per arm).\nIn our initial power calculations, we anticipated that an average of 200 patients per clinic (in the CIS and SOC arms) would be eligible for enrollment based on historical data on the annual number of adults testing positive in the VCT clinics at the participating health facilities.\nWith 5 facilities per study arm, an average of 200 patients per facility, an intraclass correlation coefficient (ICC) of 0.05, and an alpha of 0.05 and assuming that 35% of participants in the SOC arm would achieve the primary outcome, we estimated that the study would have 80% power to detect as statistically significant 55% of participants in the CIS group achieving the primary outcome, and greater than 80% power to detect as statistically significant 75% of participants in the CIS+ group achieving the primary outcome.\nBecause enrollment proceeded slower than originally planned, at study midpoint we assessed the implications for power if each health facility enrolled an average of 150 participants rather than 200.\nOur calculations revealed minimal change in power with this reduction in the number of participants per health facility.\nCalculations were performed using PASS 8.0 software for 2 independent proportions in a cluster randomization study design and a 2-sided Farrington and Manning Likelihood Score Test.\nOur power estimations and statistical analyses did not take into account the pair matching prior to randomization but rather followed recommendations from Diehr et al. to break matches in statistical analyses of clustered studies when the number of pairs is between 3 and 9.\nAn intent-to-treat analysis determined the relative risk (RR) of achieving study outcomes between the CIS and SOC groups, and between the CIS+ and CIS groups.\nFor analyses of the primary outcome, we used random-intercept multilevel log-Poisson models to account for clustering within health facilities with an empirical variance adjustment for small numbers of sampling units described by Morel et al..\nWe also assessed whether the primary outcome differed after adjustment for patient-level factors by constructing propensity scores that estimated the probability of inclusion in the CIS, CIS+, and SOC groups by age, sex, region, education, income, employment status, marital status, religion, prior year history of being away from home for more than 1 month, travel time to clinic, tuberculosis status, past hospitalizations, diagnosis history, and whether another family member was known to be living with HIV.\nThe propensity score was included as a covariate in the multivariable log-Poisson models (adjusted analyses).\nIn post hoc analyses, we further estimated the likelihood of key subgroups achieving the primary outcome using interaction contrast ratios.\nThe subgroups assessed included subgroups based on baseline age, sex, region of health facility, employment status, marital status, whether the participant was away from home for more than 1 month in the year prior to study enrollment, travel time to clinic, whether a household member was known to be HIV positive, and dichotomous variables based on scales for self-reported anticipated stigma, HIV knowledge, mental health, and perceived social support as described earlier.\nFor analyses of secondary outcomes, log-Poisson models were used for dichotomous outcomes, and t tests and 2-way median tests as appropriate for continuous outcomes, adjusting for clustering but not for patient-level differences.\nResults\nHealth facility characteristics\nAs noted above, 10 primary health facilities participated in the study, 4 in Maputo and 6 in Inhambane.\nAt study start, the 5 health facilities randomized to the intervention arm reported that they had experienced disruptions of 3 or more days in VCT services in the prior 12 months, while only 1 facility randomized to the SOC arm reported experiencing a similar disruption.\nBy study end, no facilities\u2014whether in the intervention or SOC arm\u2014had experienced such disruptions.\nThroughout the study, only intervention sites conducted point-of-care CD4 testing using Pima machines in the VCT clinic.\nTwo SOC sites reported that they had Pima machines available in their laboratories but only used them to monitor CD4 counts after patients had enrolled in HIV care.\nNone of the SOC sites used SMS messaging for health messages or appointment reminders on a routine basis for all patients, but 2 sites sent SMS appointment reminders for patients participating in community ART groups.\nThough the 2013 national HIV treatment guidelines stipulate that 1 ART preparatory counseling session is required for ART-eligible patients, all the facilities participating in the study typically conducted 2 to 3 sessions prior to ART initiation, with a slight reduction in the number of sessions observed between study start and end.\nEnrollment and participant characteristics\nFig 1 shows the enrollment, exclusion, and flow of the patients by study group.\nDuring the study period, 5,327 adults \u226518 years of age were diagnosed with HIV in the VCT clinics at the 10 study facilities.\nA total of 265 of those individuals were not referred to the study staff for further information on the study because they informed the HIV testing counselor that they were not interested in the study, were already receiving HIV services, or were not willing to be referred to the diagnosing health facility.\nAmong the 5,062 who were referred to the study staff for further information, 3,058 did not meet study eligibility criteria.\nThe main reasons for exclusion were inability to provide informed consent due to distress following diagnosis (19%), inability to understand Portuguese or Xitsua (12%), and refusal to be referred to the diagnosing health facility for HIV services (10%).\nA total of 2,004 adults \u226518 years of age enrolled in the study at the 10 health facilities: 744 (37%) in the CIS group, 493 (25%) in the CIS+ group, and 767 (38%) in the SOC group.\nThe majority of participants were female (64%), and the median age of participants was 34 years of age, with no meaningful differences observed by study group (Table 1).\nMore than half of the participants (53%) were living with a partner at the time of diagnosis, and 65% of participants had a primary or lower level of education.\nMost participants (74%) were employed, and 43% had a monthly income of less than 1,500 meticais (approximately US$50).\nOne-quarter (27%) reported that another household member was living with HIV.\nWhile no serious adverse events were reported during the study period, there was 1 unanticipated event of a female participant reporting intimate partner violence.\nThe Mozambican National Committee for Bioethics for Health and the Columbia University IRB were informed of this event, and the participant asked to remain in the study but to conduct all study interviews at the facility (i.e., no follow-up phone calls).\nIntervention effect on linkage to and retention in HIV care at the diagnosing facility\nAs shown in Table 2, the CIS was associated with statistically significant improvements in the combined outcome of linkage to care within 1 month of diagnosis and retention in care 12 months following diagnosis when compared to the SOC.\nAnalyses using data from electronic medical records to examine linkage to and retention at the diagnosing health facility showed that 57% of participants in the CIS group achieved the primary outcome versus 35% of those in the SOC group (RRCIS vs SOC = 1.58, 95% CI 1.05\u20132.39).\nPost hoc calculation of the ICC for the primary outcome according to the methods of Snijders and Bosker for binary outcome data estimated an ICC of 0.066, similar to but slightly higher than the assumed ICC of 0.05 used in power and sample size estimation.\nThese results were robust to adjustment for patient-level differences (adjusted RR [aRR]CIS vs SOC = 1.55, 95% CI 1.07\u20132.25).\nAs shown in Fig 2, the greatest intervention effects were observed among young adults age 18\u201324 years (RRCIS vs SOC = 2.39, 95% CI 1.51\u20133.80, p-value for interaction between age and treatment arm = 0.07), those in Maputo (RRCIS vs SOC = 2.31, 95% CI 1.90\u20132.79, p-value for interaction between region and treatment arm < 0.0001), those who did not report that another household member was living with HIV (RRCIS vs SOC = 1.81: 95% CI 1.52\u20132.16, p-value for interaction between household member with HIV and treatment arm = 0.11), and those reporting high levels of anticipated stigma at enrollment (RRCIS vs SOC = 1.95, 95% CI 1.53\u20132.49, p-value for interaction between stigma and treatment arm = 0.10).\nEighty-nine percent of participants in the CIS group linked to the diagnosing facility on the same day as diagnosis compared to 16% (RRCIS vs SOC = 9.13, 95% CI 1.65\u201350.40) in the SOC group, 91% within 1 week compared to 46% (RRCIS vs SOC = 2.43, 95% CI 0.70\u20138.41), and 94% within 1 month compared to 63% (RRCIS vs SOC = 1.48, 95% CI 0.93\u20132.35).\nBy 12 months, nearly all CIS participants (96%) had linked to care compared to 77% (RRCIS vs SOC = 1.23, 95% CI 1.03\u20131.48) of SOC participants.\nAmong those linking to care, the median (interquartile range [IQR]) time from diagnosis to linkage was 0 days (0\u20130) in the CIS group and 3 days (1\u201326) in the SOC group (median test p < 0.001 for CIS versus SOC).\nThe effect of the intervention on retention in care, regardless of the timing of linkage, was more modest but statistically significant (6-month retention: 62% CIS versus 53% SOC, RRCIS vs SOC = 1.18, 95% CI 1.00\u20131.39; 12-month retention: 58% CIS versus 44% SOC, RRCIS vs SOC = 1.32, 95% CI 1.12\u20131.54).\nIn analyses restricted to the participants initiating ART, the median (IQR) time from diagnosis to ART initiation in the CIS and SOC groups was 32 (12\u2013135), and 63 (33\u2013230) days, respectively, while the median (IQR) time from enrollment in HIV care to ART initiation was 32 (11\u2013127), and 50 (15\u2013205) days, respectively.\nMedian time from ART eligibility to ART initiation for the CIS, CIS+, and SOC groups was 21 (9\u201340), and 25 (11\u201356) days, respectively.\nThere was no additional benefit of adding FIs to the CIS, with 55% (RRCIS+ vs CIS = 0.96, 95% CI 0.81\u20131.16; aRRCIS+ vs CIS = 0.94, 95% CI 0.76\u20131.18) of those in the CIS+ group achieving the primary outcome; 95% (RRCIS+ vs CIS = 1.00, 95% CI 0.83\u20131.13) linking to HIV care within 1 month of diagnosis, regardless of retention at 12 months; and 55% (RRCIS+ vs CIS = 0.95, 95% CI 0.79\u20131.13) being retained in care 12 months after diagnosis, regardless of the timing of linkage to care.\nIntervention effect on linkage to and retention in care at any health facility\nAnalyses supplementing data from electronic medical records from participating facilities with data from patient interviews and other health facilities in the study regions to examine linkage to and retention at any health facility showed similar effects of the intervention package.\nA total of 74% (RRCIS vs SOC = 1.47, 95% CI 1.08\u20132.01) of participants in the CIS group and 47% in the SOC group were found to have linked to HIV care at any health facility within 1 month of diagnosis and were retained in HIV care 12 months after diagnosis (Table 2).\nAdjustment for patient-level differences did not result in any change in this finding (aRRCIS vs SOC = 1.46, 95% CI 1.05\u20132.04).\nInclusion of FIs in the CIS also showed no additional benefit for linkage to and retention at any health facility, with 73% (RRCIS+ vs CIS = 0.98, 95% CI 0.85\u20131.15; aRRCIS+ vs CIS = 0.96, 95% CI 0.83\u20131.11) of those in the CIS+ group known to have linked to and been retained in HIV care at any health facility compared to the CIS group.\nIntervention effect on ART eligibility and initiation, disease progression, and death\nData from electronic medical records at study sites indicated that compared to patients in the SOC group, patients in the CIS group were more likely to ever have their ART eligibility assessed (100% versus 76.9%, RRCIS vs SOC = 1.29, 95% CI 1.08\u20131.54), be identified as ART eligible (75% versus 60%, RRCIS vs SOC = 1.24, 95% CI 1.07\u20131.43), and initiate ART (65% versus 54%, RRCIS vs SOC = 1.20, 95% CI 1.00\u20131.43) (Table 3).\nVery few participants were diagnosed with a new WHO stage 3/4 event at the diagnosing facility or self-reported a hospitalization in the 12 months after HIV diagnosis.\nThose in the CIS group had a non-significantly but modestly decreased risk compared to those in the SOC group (1% versus 3%, RRCIS vs SOC = 0.38, 95% CI 0.07\u20132.03), while similar results were observed between the CIS and CIS+ groups (1% versus 1%, RRCIS+ vs CIS = 0.65, 95% CI 0.12\u20133.64).\nNeither the CIS nor the CIS+ interventions had a significant effect on mortality within 12 months of diagnosis, with 6%, 5%, and 7% of participants in the CIS, CIS+, and SOC groups, respectively, known to have died during study follow-up (RRCIS vs SOC = 0.87, 95% CI 0.40\u20131.91; RRCIS+ vs CIS = 0.88, 95% CI 0.45\u20131.74).\nThe CIS also did not have a significant impact on mortality before (3%, RRCIS vs SOC = 0.78, 95% CI 0.46\u20131.32) or after ART initiation (3%, RRCIS vs SOC = 0.96, 95% CI 0.26\u20133.48); participants in the CIS+ group were less likely to die, though non-significantly so, before initiating ART compared to those in the CIS group (1% versus 3%, RRCIS+ vs CIS = 0.34, 95% CI 0.09\u20131.29).\nDiscussion\nWe conducted a cluster-randomized study in Mozambique to examine the effectiveness of a multi-component approach to increase linkage to and retention in HIV care\u20142 critical elements of the HIV care continuum\u2014among adults newly diagnosed with HIV.\nThe operational model of the CIS that we evaluated addresses known structural, biomedical, and behavioral barriers across the HIV care continuum and was composed of evidence-based, practical, and scalable interventions, including CD4 testing in VCT clinics with immediate turnaround of results, accelerated ART initiation for eligible individuals, and SMS health messages and appointment reminders.\nAn enhanced version of the CIS additionally included FIs.\nIn the spirit of implementation science, 2 of the interventions were implemented by existing health facility staff, rather than study staff, providing information on the real-world successes and challenges associated with the CIS that can be extrapolated to a range of settings with similar implementation contexts.\nOur study showed that participants receiving the CIS were 1.58 times more likely to link to HIV care at their diagnosing facility within 1 month of diagnosis and be retained in care at that same facility 12 months following diagnosis, representing not only a statistically significant but also a programmatically meaningful improvement.\nParticularly impressive gains were observed in timely linkage to care at the diagnosing facility: 89% of CIS participants linked to care on the day of diagnosis, representing a greater than 5-fold improvement compared to the SOC, and nearly universal linkage (96%) was achieved within 1 month of diagnosis.\nNotably, the intervention effect was greatest in subpopulations documented to have particularly poor outcomes across the HIV care continuum, including young adults and those with high stigma perceptions.\nThe intervention also had beneficial effects on other important milestones in the HIV care continuum in the 12 months following diagnosis, including the likelihood of patients having their ART eligibility assessed and initiating ART.\nWhile the intervention significantly increased retention in HIV care at both 6 and 12 months following diagnosis, retention in the CIS group remained concerningly low and far short of what is needed to end the HIV epidemic in Mozambique and other high-burden countries.\nWe found no additional gain in effectiveness from adding FIs to the CIS.\nPrior studies examining the effect of FIs in enhancing outcomes across the HIV care continuum among people living with HIV have shown inconsistent results.\nStudies from India, Uganda, and Democratic Republic of the Congo reported reductions in time to ART initiation and improvements in retention with the provision of incentives, while in the United States, randomized trials did not show any effect of FIs on linkage to care or viral load suppression.\nWhile 89% of participants in the current study reported that the type of FI provided and the amount of the FIs (i.e., mobile phone air-time vouchers worth approximately US$5 at 3 points in time) were adequate, it is possible that the FIs were not sufficiently optimized to affect behaviors.\nIndeed, as reported elsewhere, patient reactions to the FIs were surprisingly tepid, with only 21% reporting it to be the \u201cmost useful\u201d intervention for retention in care 12 months following diagnosis.\nAdditionally, fidelity to the FI component of the intervention package was imperfect, with, for example, 86% of participants eligible to receive the first incentive actually receiving it, which may have further limited the effect of this intervention.\nHowever, given the benefits of FIs in other health sectors, further research is needed to understand whether and how they may be optimized to enhance outcomes across the HIV care continuum.\nThis study has several important strengths.\nIt is among the first studies to evaluate the impact of a multi-component approach on 2 important HIV care and treatment indicators: timely linkage to care following an HIV diagnosis and sustained retention in care.\nImproving performance for these 2 elements of the HIV care continuum is critical for realizing the individual and population benefits of HIV programming in sub-Saharan Africa.\nFurther, while studies have examined the effectiveness of multi-component intervention packages that include FIs on HIV care outcomes, this study is the first to our knowledge to use a design that permits estimation of the additional benefit of including FIs as part of such a package.\nOur study also had limitations.\nFirst, in alignment with recent recommendations for implementation science studies, we used existing electronic medical records in the HIV clinics at the study sites to ascertain outcomes at the diagnosing facility, but such records may have limited data quality.\nHowever, data quality assessments were conducted regularly during the study period and ensured at least 85% concurrence between paper-based and electronic medical records on key data elements.\nSecond, aside from the FI, we cannot unpack the effect of individual intervention components.\nThird, the relevance of point-of-care CD4 count testing may change as countries adopt \u201ctreatment for all\u201d strategies, although our results suggest that providing people living with HIV with additional information on their health status immediately following diagnosis may be important in facilitating same-day linkage to care and likely same-day ART initiation.\nFourth, the CIS+ cohort was enrolled once the target sample size had been reached in the CIS cohort, thus introducing the potential for secular trends to have biased the comparison of the CIS and CIS+ packages.\nHowever, because we found no difference in the primary outcome between the CIS+ and CIS groups, secular trends would have had to have operated in the direction of reducing overall linkage and retention for this bias to result in the failure to observe an additional benefit of FIs for linkage and retention.\nWhile this is plausible, we do not have any evidence that a substantial reduction in overall linkage and retention occurred over the relatively limited time frame of the study.\nFinally, while the study was implemented in 2 contrasting settings within Mozambique, study facilities were located primarily in urban and semi-urban areas within the city of Maputo and Inhambane Province, which may limit generalizability.\nIndeed, settings with lower education and cell phone coverage than those included in our study may experience greater challenges implementing the SMS health messages and appointment reminders.\nSimilarly, while we set broad inclusion criteria, we did exclude people who did not understand Portuguese or Xitsua, were planning on leaving the community, or were not willing to receive services at the diagnosing facility, all factors that may have reduced generalizability.\nFinally, due to slower-than-expected enrollment, we enrolled fewer participants in the CIS+ group than intended, which decreased our power to detect statistically significant differences in study outcomes between the CIS+ and CIS groups.\nHowever, as the proportion achieving the combined outcome in the 2 groups was extremely similar (CIS 57% versus CIS+ 55%), it is unlikely that the inability to detect significant differences was primarily due to lack of power.\nConclusion\nMulti-component intervention strategies have been proposed to address troubling gaps in the HIV care continuum.\nTo our knowledge, this is amongst the first studies to rigorously evaluate such an approach.\nThe CIS we examined, comprising 3 evidence-based, practical, and scalable interventions, holds great promise as an approach to make much needed gains in the HIV care continuum in sub-Saharan Africa, particularly in the critical first step of timely linkage to care following diagnosis.\nFlow chart for study participation.CIS, combination intervention strategy; SOC, standard of care; VCT, voluntary counseling and testing.\nRelative risk of the CIS compared to the SOC on the primary outcome at the diagnosing health facility by patient characteristics.a Fifteen patients with missing information were excluded from this estimate. A description of the variables examined and categories used are provided in the Methods section.\n\nParticipant characteristics at study enrollment in the 3 study groups (N = 2,004).\nCharacteristic | TotalN = 2,004 | CISN = 744 | CIS+N = 493 | SOCN = 767 | p-Value\nRegion |  |  |  |  | \nMaputo | 1,077 (54%) | 396 (53%) | 275 (56%) | 406 (53%) | 0.58\nInhambane | 927 (46%) | 348 (47%) | 218 (44%) | 361 (47%) | \nSex |  |  |  |  | 0.50\nFemale | 1,292 (64%) | 490 (66%) | 319 (65%) | 483 (63%) | \nMale | 712 (36%) | 254 (34%) | 174 (35%) | 284 (37%) | \nAge (years) | 34.2 (9.6) | 34.9 (9.8) | 33.8 (9.9) | 33.8 (9.3) | 0.045\n18\u201324 | 265 (13%) | 90 (12%) | 70 (14%) | 105 (14%) | 0.12\n25\u201339 | 1,233 (62%) | 440 (59%) | 301 (61%) | 492 (64%) | \n40\u201349 | 348 (17%) | 148 (2%) | 87 (18%) | 113 (15%) | \n50+ | 158 (8%) | 66 (9%) | 35 (7%) | 57 (7%) | \nMarital status |  |  |  |  | <0.001\nMarried/partner and living together | 1,068 (53%) | 376 (51%) | 255 (52%) | 437 (57%) | \nMarried/partner, but not living together | 222 (11%) | 101 (14%) | 86 (17%) | 35 (5%) | \nSingle | 713 (36%) | 266 (36%) | 152 (31%) | 295 (38%) | \nMissing/refused | 1 (0%) | 1 (0%) | 0 (0%) | 0 (0%) | \nEducation |  |  |  |  | 0.003\nNone | 164 (8%) | 59 (8%) | 33 (7%) | 72 (9%) | \nPrimary | 1,149 (57%) | 442 (59%) | 256 (52%) | 451 (59%) | \nSecondary | 471 (24%) | 164 (22%) | 130 (26%) | 177 (23%) | \nAbove secondary | 219 (11%) | 78 (1%) | 74 (15%) | 67 (9%) | \nMissing/refused | 1 (0%) | 1 (0%) | 0 (0%) | 0 (9%) | \nEmployment |  |  |  |  | 0.46\nEmployed | 1,473 (74%) | 537 (72%) | 361 (73%) | 575 (75%) | \nUnemployed | 531 (26%) | 207 (28%) | 132 (27%) | 192 (25%) | \nMonthly income |  |  |  |  | <0.001\n\u22641,500 meticais | 871 (43%) | 342 (46%) | 165 (33%) | 364 (47%) | \n>1,500 meticais | 936 (47%) | 343 (46%) | 271 (55%) | 322 (42%) | \nMissing/refused | 197 (1%) | 59 (8%) | 57 (12%) | 81 (11%) | \nAnother household member has HIV |  |  |  |  | 0.28\nYes | 550 (27%) | 187 (25%) | 144 (29%) | 219 (29%) | \nNo | 913 (46%) | 361 (49%) | 219 (44%) | 333 (43%) | \nDon\u2019t know | 539 (27%) | 196 (26%) | 130 (26%) | 213 (28%) | \nMissing/refused | 2 (0%) | 0 (0%) | 0 (0%) | 2 (0%) | \n\nData given as N (percent).\nCIS, combination intervention strategy; SOC, standard of care.\n\nLinkage to and retention in HIV care: CIS versus SOC and CIS+ versus CIS.\nCategory | Outcome | CISN = 744 | CIS+N = 493 | SOCN = 767 | RR1 (95% CI), p-Value | aRR2 (95% CI), p-Value\nN | Percent | N | Percent | N | Percent | CIS versus SOC | CIS+ versus CIS | CIS versus SOC | CIS+ versus CIS\nPrimary outcome |  |  |  |  |  |  |  |  |  |  | \nAt diagnosing facility | Linked to care within 1 month of diagnosis and retained 12 months after diagnosis | 425 | 57% | 273 | 55% | 268 | 35% | 1.58 (1.05\u20132.39)p = 0.03 | 0.96 (0.81\u20131.16)p = 0.66 | 1.55 (1.07\u20132.25)p = 0.04 | 0.94 (0.76\u20131.18)p = 0.52\nAt any health facility | Linked to care within 1 month of diagnosis and retained 12 months after diagnosis | 547 | 74% | 360 | 73% | 363 | 47% | 1.47 (1.08\u20132.01)p = 0.02 | 0.98 (0.85\u20131.15)p = 0.91 | 1.46 (1.05\u20132.04)p = 0.03 | 0.96 (0.83\u20131.11)p = 0.52\nSecondary outcomes |  |  |  |  |  |  |  |  |  |  | \nLinkage at diagnosing facility | Same day as HIV test | 659 | 89% | 457 | 93% | 120 | 16% | 9.13 (1.65\u201350.40)p = 0.02 | 1.04 (0.92\u20131.20)p = 0.38 | N/A | \n | Within 1 week of HIV test | 678 | 91% | 461 | 94% | 349 | 46% | 2.43 (0.70\u20138.41)p = 0.14 | 1.03 (0.91\u20131.16)p = 0.59 | N/A | \n | Within 1 month of HIV test | 703 | 94% | 467 | 95% | 482 | 63% | 1.48 (0.93\u20132.35)p = 0.09 | 1.00 (0.89\u20131.13)p = 0.96 | N/A | \n | Within 12 months of HIV test | 716 | 96% | 467 | 95% | 592 | 77% | 1.23 (1.03\u20131.48)p = 0.03 | 0.98 (0.87\u20131.11)p = 0,74 | N/A | \nRetention at diagnosing facility | 6 months after diagnosis | 462 | 62% | 322 | 65% | 405 | 53% | 1.18 (1.00\u20131.39)p = 0.06 | 1.05 (0.88\u20131.26)p = 0.48 | N/A | \n | 12 months after diagnosis | 435 | 58% | 273 | 55% | 341 | 44% | 1.32 (1.12\u20131.54)p = 0.004 | 0.95 (0.79\u20131.13)p = 0.45 | N/A | \n\n1RR accounts for clustering within sites using random-intercept log-Poisson regression with empirical standard error estimates.\n2aRR adjusts for patient-level differences using propensity scores.\naRR, adjusted relative risk; CIS, combination intervention strategy; N/A, not applicable; RR, relative risk; SOC, standard of care.\n\nART determination and initiation, disease progression, and death: CIS versus SOC and CIS+ versus CIS.\n\u00a0Outcome | CIS(N = 744) | CIS+(N = 493) | SOC(N = 767) | RR1 (95% CI), p-value\nN | Percent | N | Percent | N | Percent | CIS versus SOC1 | CIS+ versus CIS1\nART eligibility assessed | 744 | 100% | 493 | 100% | 590 | 77% | 1.29 (1.08\u20131.54)p = 0.01 | 1.00 (0.89\u20131.12)p = 1.00\nIdentified as ART eligible | 557 | 75% | 372 | 75% | 464 | 60% | 1.24 (1.07\u20131.43)p = 0.01 | 1.01 (0.85\u20131.19)p = 0.91\nInitiated ART | 484 | 65% | 332 | 67% | 416 | 54% | 1.20 (1.00\u20131.43)p = 0.05 | 1.03 (0.88\u20131.22)p = 0.59\nNew WHO stage 3/4 or hospitalization | 7 | 1% | 3 | 1% | 23 | 3% | 0.38 (0.07\u20132.03)p = 0.22 | 0.65 (0.12\u20133.64)p = 0.53\nDeath within 12 months | 46 | 6% | 27 | 5% | 54 | 7% | 0.87 (0.40\u20131.91)p = 0.69 | 0.88 (0.45\u20131.74)p = 0.63\nDeath before ART initiation | 22 | 3% | 5 | 1% | 29 | 4% | 0.78 (0.46\u20131.32)p = 0.31 | 0.34 (0.09\u20131.29)p = 0.09\nDeath after ART initiation | 24 | 3% | 22 | 4% | 25 | 3% | 0.96 (0.26\u20133.48)p = 0.94 | 1.38 (0.62\u20133.07)p = 0.33\n\n1RR accounts for clustering within sites using random-intercept log-Poisson regression with empirical standard error estimates.\nART, antiretroviral therapy; CIS, combination intervention strategy; RR, relative risk; SOC, standard of care.",
        "label": "high",
        "id": "task4_RLD_test_79"
    },
    {
        "paper_doi": "10.1371/journal.pone.0119772",
        "bias": "allocation concealment (selection bias)",
        "PICO": "Methods: Cluster-RCT in Bulgan, Mongolia.\n\n\nParticipants: Sample size: 501 women randomised.Clusters: the unit of randomisation was the Soum and bag, small geographic areas in Mongolia. Each Soum has a healthcare facility where women must register their newborn. 18 geographic areas were randomised, after selection for administrative convenience and to avoid contamination.Individuals: pregnant women living in Bulgan, Mongolia.\n\n\nInterventions: Target: community.Arm 1: distribution of maternal and child health handbooks during pregnancy. The MCH handbook logged maternal health and personal information, pregnancy, delivery and postpartum health and weight, dental health, parenting classes, child developmental milestones from 0-6 years, immunisation records and height and weight charts for children.Arm 2: women received standard care.\n\n\nOutcomes: Trial primary outcome: number of antenatal visits; proportion of women attending 6 or more antenatal visits. (The national standard for ANC in Mongolia is 6 visits.)Review outcomes reported:Primary: ANC coverage of at least 4 visits, maternal mortalitySecondary: maternal outcomes: morbidity during pregnancy, mode of delivery, breastfeeding initiation, maternal depression and health (EPDS and GHQ). Infant outcomes: birthweight, Apgar score, NICU admission, neonatal mortality at discharge. Maternal healthy behaviours.Follow-up: data collection at 1 month postpartum.\n\n\nNotes: Funders: this study was funded by the National Center for Global Health and Medicine, Tokyo, Japan.Significant group differences noted for distances travelled to nearest health centre (greater in the intervention group) and for wealth index (the control group was poorer). The authors report that travel time did not function as an effect modifier; however, women from a higher socioeconomic background attended more ANC visits.Trial authors provided unpublished outcome data upon request. The trial statistician (HN) calculated ORs and 95% confidence intervals using the generalised estimating equations (GEE) method to adjust for cluster design and baseline differences, including wealth\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Objective\nTo assess the effectiveness of the Maternal and Child Health (MCH) handbook in Mongolia to increase antenatal clinic attendance, and to enhance health-seeking behaviors and other health outcomes.\nMethods\nA cluster randomized trial was conducted using the translated MCH handbook in Bulgan, Mongolia to assess its effectiveness in promoting antenatal care attendance.\nPregnant women were recruited from 18 randomly allocated districts using shuffled, sealed envelopes.\nThe handbook was implemented immediately for women at their first antenatal visit in the intervention group, and nine months later in the control group.\nThe primary outcome was the number of antenatal care visits of all women residing in the selected districts.\nCluster effects were adjusted for using generalized estimation equation.\nMasking was not possible among care providers, pregnant women and assessors.\nFindings\nNine districts were allocated to the intervention group and the remainder to the control group.\nThe intervention group (253 women) attended antenatal clinics on average 6\u20229 times, while the control group (248 women) attended 6\u20222 times.\nSocioeconomic status affected the frequency of clinic attendance: women of higher socioeconomic status visited antenatal clinics more often.\nPregnancy complications were more likely to be detected among women using the handbook.\nConclusion\nThe MCH handbook promotes continuous care and showed an increase in antenatal visits among the intervention group.\nThe intervention will help to identify maternal morbidities during pregnancy and promote health-seeking behaviors.\nTrial Registration\nUMIN Clinical Trial Registry UMIN000001748\nIntroduction\nMaternal and child health continues to present a significant public health challenge in Mongolia.\nDespite a marked improvement in the maternal and neonatal mortality ratios over the past 20 years, with 89\u00b76 per 100,000 births in 2007 and 14 per 1,000 births during 2001\u20132003, respectively, as well as a decline in the mortality of older children, the quality of antenatal care is still low and complications during pregnancy remain a significant hurdle for improving maternal health in Mongolia.\nEffective interventions to enhance maternal and child health outcomes are crucial to address these challenges and to maintain the achievement of health-related Millennium Developmental Goals (MDGs) 4 and 5.\nAn ongoing challenge for researchers and health professionals is how to deliver effective interventions to reduce maternal and neonatal mortality in resource-limited settings.\nEffective maternal health interventions should aim to encourage health-seeking behaviors among pregnant women and increase their maternal knowledge.\nAs the role of health workers is to promote healthcare-seeking behaviors and initiate preventive action, a health record book such as Japan\u2019s Maternal and Child Health (MCH) handbook could be used as an effective tool by community healthcare workers and professional hospital staff to enhance client\u2013provider communication about health, raise health awareness, and identify complications earlier in the pregnancy.\nThe purpose of introducing the handbook to Mongolia, which was proposed by the Mongolian Ministry of Health, was to increase antenatal visits and enhance client-provider communication during pregnancy to improve long-term health outcomes for mother and child.\nThe handbook was first considered by the Mongolian government as a key intervention in maternal and child health in 2007, and our study initiated the national adoption of the MCH handbook in Mongolia in 2010.\nRegarded as Japan\u2019s flagship intervention in the context of health aid, the handbook has been adopted in other countries, such as Indonesia and Bangladesh, and previous studies have evaluated its impact on perinatal health.\nHowever, a high-quality study that assesses the effectiveness of the handbook to facilitate long-term information-sharing has not previously been undertaken.\nThe World Health Organization (WHO) emphasises the importance of effective interventions that focus on delivering a continuum of care.\nThe MCH handbook facilitates continuum of care throughout pregnancy, delivery and postpartum as well as the child\u2019s infancy using the handbook\u2019s continuous record of basic educational information that encompasses antenatal care and the milestones of child development from the ages of 0\u20136 years.\nWomen use the handbook by filling out relevant sections with their personal, maternal and child health information, and bringing the handbook with them to all antenatal and postnatal appointments.\nAt the appointment, the midwife and/or doctor then check the relevant section of the handbook pertaining to the woman\u2019s stage of pregnancy or the child\u2019s stage of development, and record in it results of tests, such as protein in the urine during pregnancy, or other notes.\nThe handbook also contains information on MCH care and serves as a valuable communication and educational tool between pregnant women and healthcare professionals, through which women can raise specific health concerns and healthcare professionals can convey important health messages and guidance at point of care.][\nIn this study, we aim to measure improved health-seeking behavior by increased antenatal clinic attendance in the Mongolian province of Bulgan.\nThe effectiveness of the intervention will be investigated through a cluster randomized control trial evaluating antenatal attendance, maternal physical and mental health, neonatal health and healthy behaviors.\nImplementation of the MCH handbook\u2014a communication tool between women and healthcare professionals\u2014can only be conducted at cluster level, and therefore a cluster-randomized trial was employed.\nMethods\nStudy design\nA cluster-randomized controlled study was conducted from 1 May 2009 until 1 September 2010 among pregnant women and their infants who lived in Bulgan, Mongolia.\nThe allocation ratio was 9/9 = 1.00 and the unit of randomisation in this study was the soum\u2014a small administrative unit in Mongolia\u2014and the bag, which is a subdivision of a soum.\nParticipants/ Study population\nEligible participants included pregnant women living in the Bulgan province of Mongolia.\nHealth centres in Bulgan are located in each soum and all women must register their newborn infants at their local health centre, regardless of the infant\u2019s birthplace.\nData confidentiality was strictly maintained throughout all steps of this study.\nRandomisation and masking\nSoums and bags were selected for administrative convenience and to avoid contamination.\nBulgan province is comprised of 17 soums and 4 bags and they differ in size, health outcomes, and available healthcare facilities.\nOf the combined soums and bags (21), 18 units (16 soums and 2 bags) were selected for inclusion in this study, and randomized in equal number between intervention and control group.\nThree units were excluded because one soum was the subject of a pilot study, and two bags were included in another health promotion project.\nRandomisation was conducted using shuffled, sealed envelopes, and an envelope was selected by each soum representative.\nSince the unit of randomisation is a soum and the intervention is visible, the intervention and outcomes could not be masked.\nWritten informed consent was sought from all women for permission to use the collected data in the study.\nInterventions\nThe Mongolian edition of the handbook was translated into Mongolian from the original Japanese version.\nThe MCH handbook contained a log for recording information on maternal health and personal information, course of pregnancy, delivery and postpartum health, weight during and after pregnancy, dental health, parenting classes, child development milestones from the ages of 0\u20136 years, immunization and illnesses, and height and weight charts for children.\nThe handbook was used as the intervention at both the cluster and individual participant level.\nThe handbooks were implemented at the beginning of the study observational period, and after a delay of seven months in the control group.\nOutcomes\nThe primary outcome was the number of antenatal care visits and the proportion of women who made six antenatal care visits or more.\nIn Mongolia, the national standard of antenatal care visits is a minimum of six.\nHealthcare professionals working in each cluster recorded each antenatal visit for their soum.\nData was collected for all participants at one month postpartum.\nSecondary outcomes included clinical outcomes (mortality and morbidities) of women and their infants, as well as healthy behaviors of women and their families.\nCharacteristics and other outcomes for mothers and their infants were collected 28 days after childbirth via self-reported questionnaires and interviews conducted by trained data collectors.\nThe data collectors visited the family clinic or regional hospital as well as the household to undertake routine check-ups of mothers and infants using a questionnaire.\nAll mortality and morbidity ratios are derived from routinely collected national statistics using the ICD-10 classification system.\nStatistical analyses\nThe primary analyses followed the intention-to-treat principle and compared the proportion of women who visited health centres for antenatal check-ups and their number of visits between the intervention and control groups.\nIn the analysis of a cluster-randomized trial, correlations of the outcomes of participants in the same cluster should be adequately adjusted.\nTo do this, the generalized estimating equations (GEE) method was adopted to estimate mean difference, risk ratio and risk difference as a measure of the effect, and to calculate their 95% confidence intervals (CI).\nMultivariate GEE analyses was performed to adjust for possible effects of baseline variables.\nTo quantify household wealth status, principle component analysis was used according to the procedure outlined in the Demographic and Health Survey (DHS) guidelines.\nThe whole population sample in the Bulgan province of Mongolia was used to create the wealth index.\nThe sample size was determined to detect one mean difference in antenatal care visits between the two groups with a two-sided alpha level of 0.05 and 80% power.\nAssuming 0.01 intra-cluster correlation, it was estimated that approximately 500 women were required in total.\nAll statistical analyses were conducted with SAS version 9.2 (SAS Institute, Cary, NC, USA).\nThis clinical trial is registered at the UMIN Clinical Trial Registry (UMIN000001748).\nBoth the protocol and CONSORT checklist of the present trial are presented as S1 and S2 Documents.\nRole of the funding source\nThis study was supported by the National Center for Global Health and Medicine, Tokyo, Japan.\nThe funding source did not affect the conduct, analyses or results of the study in any way.\nResults\nBaseline characteristics of the soums\nThis study had nine clusters in the intervention group and the total intervention population was 253 women with an average of 28\u00b70 people in a cluster.\nThe intervention was implemented between May 2009 and January 2010, and data was collected between February 2010 and August 2010.\nOf the whole intervention group, only 210 participants received the intervention.\nThere were nine clusters in the control group and the total number of participants for this group was 248 women.\nThere was no difference in the size of the cluster between the two groups.\nFig. 1 shows the selection process in detail.\nBaseline characteristics of women and infants\nAll baseline characteristics were similar in both the intervention and control groups, presented in Table 1.\nIn this study, 32\u00b74% of participants from the intervention group and 31\u00b70% of the control group were experiencing their first pregnancy; 94\u00b71% of the intervention group and 95\u00b72% of the control group were married and the mean age of both groups was 27 years of age.\nFrom the intervention group, 9\u00b75% of participants were educated to elementary school level compared with 10\u00b75% of the control group.\nStatistically significant differences in travel time were observed between women\u2019s homes and the nearest health centres, (p = 0\u00b7008), and in the wealth index (p = 0\u00b7001) of the intervention group compared to the control group.\nPrimary outcome\nWomen in the intervention group attended antenatal clinics an average of 6\u00b79 times, while women from the control group attended 6\u00b72 times, as shown in Table 2.\nIn the primary GEE analysis, there is no significant difference between the two groups in the number of antenatal care visits and the proportion of women who have visited more than 6 times.\nThe travel time to antenatal clinics did not significantly affect the association between the intervention and the primary outcome.\nHowever, socioeconomic status was found to influence the frequency of clinic attendance: women of a higher socioeconomic status visited antenatal clinics more often than women from a lower socioeconomic background.\nSocioeconomic status acted as a statistically significant effect modification on outcomes by the multivariate GEE analyses.\nTherefore the analysis of primary outcomes was stratified by socioeconomic-status quintile.\nResults of the GEE analysis of primary outcomes stratified by wealth index are presented in Figs. 2, 3 and 4.\nWomen\u2019s health\nComplications in maternal health were more likely to be identified, with maternal morbidity during pregnancy at 12\u00b73% in the intervention group compared with 5\u00b77% in the control group.\nThis difference was statistically significant (p-value 0\u00b701).\nNo evidence of difference was observed in women who scored both higher than 12 points in EPDS (RR 0.99 [0.94\u20131.04], p = 0.56) and higher than 4 points in GHQ (RR 1.01 [0.99\u20131.03], p = 0.41).\nInfant health\nA higher rate of early breastfeeding initiation amongst the intervention group was a significant neonatal health outcome.\nIn the intervention group, 94\u00b71% of infants initiated breastfeeding within one hour of childbirth compared to 87\u00b75% of infants in the control group.\nThis difference, though notable, was not statistically significant.\nHealthy behaviors\nAn increase in healthy behaviors was observed amongst the intervention group.\nThe majority of women stopped drinking alcohol during pregnancy: only 7\u00b79% of women from the intervention group continued to drink alcohol when pregnant compared with 14\u00b71% in the control group.\nIn the intervention group, a statistically significant reduction in smoking was found among women\u2019s family members living in the same household, with 51\u00b70% of family members continuing to smoke during women\u2019s pregnancies compared with 60\u00b79% living with control group participants.\nDiscussion\nOur findings show that pregnant women who used the MCH handbook increased their number of antenatal visits from the national requirement of six visits to a mean of 6.9 visits, compared to a mean of 6\u00b72 visits in the control group.\nSocioeconomic background was also found to play a significant role in clinic attendance for both groups.\nAfter adjusting for confounders in the GEE model, the intervention effect was statistically significant, but only among the wealthy.\nParticipants in the wealthiest two quintiles were more likely to attend antenatal clinics more than six times.\nComplications in maternal health were more likely to be detected among pregnant women who used the handbook.\nHealthy behaviors were also adopted by partners and other family members of pregnant women in the intervention group: the majority of women did not drink alcohol (7\u00b79% in the intervention group compared with 14\u00b71% in the control group), and approximately half of family members stopped smoking at home, thereby reducing the harm of passive smoking for expectant mothers.\nTo our knowledge, this study is the first to assess the MCH handbook using a cluster-randomized design in Mongolia.\nThis is also the first cluster-randomized controlled trial to assess a Japanese health aid intervention.\nThe MCH handbook provided pregnant women with a useful educational aid that promotes healthcare-seeking behaviors, fosters continuity of care and enhances communication between pregnant women and their healthcare providers.\nThe intervention raised women\u2019s awareness of maternal and child health concerns and prompted them to seek out healthcare, as illustrated by an increase in antenatal visits.\nNot only does the handbook instigate the delivery of key health messages from healthcare providers to pregnant women during antenatal visits, but also from pregnant women to their families at home.\nThis study used a randomized cluster design and population-based data collection so its results may be more representative of the effectiveness of the MCH handbook in the community; however, several limitations are present.\nMasking was not possible among care providers, pregnant women and assessors.\nRecall bias likely exists in the analysis, because data collection was performed at one month after birth.\nThe unbalanced distribution of socioeconomic status acts as an effect modifier.\nThe per-protocol analysis showed a significant increase in women\u2019s clinic attendance, and therefore it is likely that women of a lower socioeconomic status did not receive the handbook.\nA systematic review of a similar intervention highlights the potential benefits of giving women their own health record to use during pregnancy.\nThe review included three randomized trials.\nThough none of the trials assessed the rate of antenatal care visits as an outcome, the intervention resulted in favourable outcomes such as enhancing a mother\u2019s control over her health, and satisfaction with the care provided.\nA cross-sectional study conducted by Osaki et al showed the MCH handbook increased utilisation of health services and deliveries with trained personnel.\nAlthough the rate of antenatal visits is not reported in Osaki et al\u2019s study, the findings of our study are compatible with this and other studies.\nA significant outcome of this study was an increase in the proportion of women who attended antenatal care visits among pregnant women who used the MCH handbook.\nTravel time did not function as an effect modifier; however, socioeconomic background was particularly relevant, with women from a higher socioeconomic background visiting antenatal clinics more often than those of a lower socioeconomic background.\nThe handbook also facilitated the identification of maternal morbidities during pregnancy and minimized passive smoking in the households of intervention group participants.\nIn response to the study\u2019s main findings, the MCH handbook was implemented as part of the national health policy in Mongolia in 2010 soon after the trial was finished, and the results support the policy.\nHowever, policies to reach women of lower socioeconomic status are yet to be developed and more research is required to address this issue.\nOur study showed the effectiveness of the MCH handbook to promote long-term information sharing through an increase in antenatal clinic attendance among women who used the handbook.\nThe intervention promotes better communication between women and healthcare specialists and acts as a reference point for women to raise particular concerns and questions about their own health at antenatal clinics, while at the same time giving healthcare workers the opportunity to deliver important health messages.\nThe handbook\u2019s role in enhancing long-term information sharing can make an important contribution to maintaining MDGs 4 and 5.\nFurther interventions are also necessary to specifically target pregnant women from a lower socioeconomic background in outreach efforts that aim to increase antenatal clinic attendance.\nFuture research should also focus on the effectiveness of the handbook in other provinces within Mongolia as well as other low- to middle-income countries, where the handbook can be used as an effective tool in maternal health education to further promote maternal health awareness and healthy behaviors, enable early interventions, reduce adverse birth outcomes in developing settings and sustain the achievement of MDGs 4 and 5.\nUse of the latest information technology, such as a smartphone application of the MCH handbook to facilitate use of the intervention, may also provide a valuable opportunity to enhance accessibility of the handbook, and would benefit from further research.\nFlow diagram of the study population.\nPrimary outcome: Mean difference by wealth index.The y-axis shows the mean difference with confidence intervals of the number of antenatal care visits between the intervention and control groups. The x-axis shows the wealth index quintile.\nPrimary outcome: Risk ratio by wealth index.The y-axis shows the risk ratio with confidence intervals of the number of women who made six antenatal care visits during their pregnancy in the intervention and control groups. The x-axis shows the wealth index quintile.\nPrimary outcome: Risk difference by wealth index.The y-axis shows the risk differences with confidence intervals of the number of women who made six antenatal care visits during their pregnancy in the intervention and control groups. The x-axis shows the wealth index quintile. Note: 1st quintile represents the highest wealth index, and the 5th represents the lowest.\n\nBaseline characteristics of women and infants.\n |  | Intervention | Control | p-value\nN = 253 | N = 248\nFirst pregnancy | N (%) | 82 (32.41) | 77 (31.05) | 0.743\nNumber of pregnancies | Mean (SD) | 2.49 (1.37) | 2.32 (1.24) | 0.154\nmissing | 0 | 1\nOutcome of previous pregnancies |  |  |  | \nLive birth | Mean (SD) | 1.42 (1.35) | 1.29 (1.19) | 0.229\nAbortion | 0.11 (0.41) | 0.09 (0.43) | 0.556\nMiscarriage | 0.11 (0.39) | 0.07 (0.32) | 0.233\nAdoption | 0.00 (0.00) | 0.02 (0.16) | 0.099\nPre-pregnancy weight | Mean (SD) | 61.10 (9.02) | 60.15 (8.76) | 0.237\nmissing | 2 | 2\nWeight at first antenatal care visit | Mean (SD) | 63.13 (9.20) | 61.88 (9.19) | 0.132\nmissing | 1 | 6\nTravel time from home to antenatal care clinic | Median | 40 | 40 | 0.008\n(25\u201375%) | (20\u201399) | (20\u201360)\n(min., max.) | (4, 1440) | (2, 180)\nMarital status Married | N (%) | 238 (94.1) | 236 (95.2) | 0.590\nMean maternal age (SD) | Mean (SD) | 27.3 (6.13) | 27.7 (5.67) | 0.390\nmissing | 1 | 3\nMaternal educational attainment (up to elementary level education) | N (%) | 24 (9.49) | 26 (10.48) | 0.947\nNumber of family members in the household | Mean (SD) | 4.332 (1.23) | 4.185 (1.196) | 0.177\nWealth index | Mean (SD) | 0.448 (2.194) | -0.225 (2.356) | 0.001\n\n\nPrimary outcome and outcomes for mothers, infants and healthy behaviors.\n |  | Intervention N = 253 | Control N = 248 | Effect of measure [MD: Mean difference, RR: Risk ratio, RD: Risk difference] (95%CI), p: p-value, *GEE analysis\nPrimary outcome\nAntenatal care visits | Mean (SD) | 6.615 (1.525) | 6.407 (1.765) | [MD] 0.208 (\u20130.710\u20131.125) (p = 0.66)*\nAntenatal care visits | \u2265 6 N(%) | 206 (81.7%) | 175 (70.6%) | [RR] 1.158 (0.876\u20131.532), p = 0.30*, [RD] 11.2% (-9.9%-32.3%), p = 0.30*\nWomen\u2019s outcomes\nComplications identified during pregnancy | N (%) | 31 (12.25) | 14 (5.65) | P = 0.012\nmissing | 1 | 1\nMultiple pregnancies | N (%) | 6 (2.37) | 4 (1.61) | \nGestational age | Mean (SD) | 38.95 (1.25) | 39.06 (1.18) | \nMedian (25\u201375%) | 39 (38\u201340) | 39 (39\u201340)\nmissing | 7 | 14\nCephalic fetal presentation | N (%) | 246 (97.23) | 236 (95.16) | \nSpontaneous vaginal deliveries | N (%) | 202 (79.84) | 212 (85.48) | \nEPDS: Postnatal depression Over cut-off 12 points | N(%) | 15 (5.93) | 11 (4.44) | RR 0.99 (0.94\u20131.04), p = 0.560, RD\u20140.014 (\u20130.062\u20130.034), p = 0.561\nGHQ: General Health Questionnaire Over cut-off 4 | N (%) | 3 (1.2%) | 5 (2.0%) | RR 1.01 (0.99\u20131.03), p = 0.412, RD 0.0085 (\u20130.012\u20130.029), p = 0.411\nInfant outcomes\nApgar score 5 minutes | Mean(SD) | 7.55 (0.89) | 7.34 (1.25) | MD 0.210 (\u20130.212\u20130.632), p = 0.330\nMedian (25\u201375%) | 8 (7\u20138) | 7 (7\u20138)\nmissing | 7 | 6\nBirthweight | Mean(SD) | 3388.61(449.00) | 3429.11(486.40) | MD\u201340.50 (\u2013141.53\u201360.53), p = 0.432\nmissing | 2 | 1\nFemale | N (%) | 123 (48.6) | 120 (48.39) | \nAny congenital malformation | N (%) | 6 (2.37) | 3 (1.21) | \nAdmission of newborn to the Intensive Care Unit | N (%) | 6 (2.37) | 5 (2.02) | \nWhen did breastfeeding start? | N (%) |  |  | RR 1.07 (0.97\u20131.18), p = 0.186, RD 0.062 (\u20130.028\u20130.153), p = 0.176\n1) Within one hour after birth | 238 (94.07) | 217 (87.50)\n2) Between one hour and 24 hours after birth | 10 (3.95) | 25 (10.08)\n3) After 24 hours | 3 (1.19) | 2 (0.81)\n4) Breastfeeding not initiated before discharge/ after birth | 1 (0.40) | 2 (0.81)\nNeonatal status at discharge Death | N (%) | 1 (0.40) | 2 (0.81) | RR 1.00 (0.99\u20131.02), p = 0.512, RD 0.0041 (\u20130.0082\u20130.016), p = 0.512\nHealthy behaviors\nDrinking during pregnancy | N (%) | 20 (7.91) | 35 (14.11) | RR 1.07 (0.97\u20131.18), p = 0.166, RD 0.061 (\u20130.024\u20130.15), p = 0.161\nmissing | 2 | 0\nMaternal smoking | N (%) | 5 (1.98) | 7 (2.82) | RR 1.01 (0.98\u20131.04), p = 0.572, RD 0.0086 (\u20130.021\u20130.038), p = 0.571\nmissing | 0 | 1\nSmoking among other members of the household during pregnancy | N (%) | 129 (50.98) | 151 (60.89) | RR 0.841 (0.71\u20130.99), p = 0.039, RD \u22120.097 (\u22120.194\u2013\u22120.001), p = 0.048\nmissing | 1 | 1\n",
        "label": "low",
        "id": "task4_RLD_test_487"
    },
    {
        "paper_doi": "10.1371/journal.pmed.1001630",
        "bias": "random sequence generation (selection bias)",
        "PICO": "Methods: Study design: cRCT with 2 intervention armsUnit of allocation: clusters (villages)Number of units: 25 randomized villages in each arm. A subset of 20 villages per arm was used for entomological assessment.Outcome assessment/surveillance type: see below in 'Outcomes' sectionLength of follow-up: 3 postintervention cross-sectional household surveys were undertaken in 2012. Survey A (23 February to 31 March) was after the short rainy season and 2 months after the first spray round. Survey B (25 June to 31 July) was after the long rainy season, 6 months after the first spray round, and 2 months after the second spray round. Survey C (25 October to 4 December) was 6 months after the second spray round and 10 months after the first. Baseline surveys were conducted in 2011 during the same periods as surveys A and B.Adjustment for clustering: yes\n\n\nParticipants: Number of participants: for each of the surveys, a different number of participants were used in each cohortSurvey A: 2192 children in control arm, 2348 in intervention armSurvey B: 2045 children in control arm, 2207 in intervention armSurvey C: 2101 children in control arm, 2303 in intervention armPopulation characteristics: cohort of children aged 6 months to 14 years, villages had to be sprayed with IRS in the baseline year.Withdrawal and loss to follow-up: 82.2-84.4% of intervention participants tested in each survey. 78.3-80.8% of control participants tested\n\n\nInterventions: IRSActive ingredient and dosage: bendiocarb 400 mg/m2Formulation: 80% WPFrequency of spraying: 2 rounds of spraying (December 2011 to January 2012) and (April 2012 to May 2012), timed to precede the peak in malaria cases that normally occurs at the end of each rainy season.Coverage: survey A: 92.1% (95% CI 88.4% to 94.7%) (1215); survey B: 89.5% (95% CI 84.0% to 93.2%) (1138); survey C: 89.3% (95% CI 83.6% to 93.2%) (1209)Buffer size between clusters: each village was divided into a core surveillance area consisting of >= 200 houses and approximately 1 km radius, where the surveys were conducted, and an outer buffer zone of approximately 1 km width which also received treatment but in which there was no outcome monitoring.ITNActive ingredient and dosage: permethrin 2% w/w (Olyset Net)Coverage measured as % of households with >= 1 ITN per sleeping space: survey A: 57.2 (range 53.6-60.7) (1215); survey B: 57.4 (range 54.0-60.9) (1142); survey C: 56.8 (range 51.7-61.8) (1211)Coverage measured as % of households with >= 1 ITN: survey A: 89.0 (range 87.1-90.6) (1216); survey B: 88.2 (range 85.7-90.3) (1142); survey C: 83.8 (range 79.9-87.1) (1211)Compliance measured as % of study children that reported sleeping under an ITN the night previous to the survey: survey A: 53.0 (range 47.5-58.3) (2349); survey B: 44.1 (range 39.2-49.2) (2207); survey C: 36.1 (range 31.0-41.5) (2303)ControlITN only as aboveCoverage measured as % of households with >= 1 ITN per sleeping space: survey A: 52.2 (range 47.8-56.5) (1178); survey B: 51.6 (range 47.0-56.0) (1094); survey C: 52.8 (range 47.6-58.0) (1168)Coverage measured as % of households with >= 1 ITN: survey A: 85.8 (range 83.7-87.7) (1177); survey B: 82.5 (range 78.7-85.7) (1096); survey C: 78.2 (range 74.3-81.6) (1170)Compliance measured as % of study children that reported sleeping under an ITN the night previous to the survey: survey A: 46.6 (range 41.7-51.6) (2193); survey B: 40.7 (range 34.7-47.0) (2045); survey C: 36.0 (range 29.8-42.6) (2101)Cointerventions: none reported\n\n\nOutcomes: P falciparum parasite rate in children aged 6 months to 14 years, 80 households in each cluster. Up to 3 children per household selected. Aimed for a mean of 80 children per cluster. Tested with RDT (Carestart (Pan) Malaria, DiaSys)Anaemia in children aged < 5 yearsMean Hb in children aged < 5 years. Tested with HemoCue Hb 201+ (Aktiebolaget Leo Diagnostics)EIR: 20/25 clusters per arm were monitored for 1 night each month from April 2011 to December 2012. 8 randomly selected houses in each clusterSporozoite rate\n\n\nLocation profile: Study location: north-west Tanzania, Muleba Distract, Kagera Region, the study area included 68,108 households at an altitude of 1100-1600 m above sea level. Rainfall occurred in 2 seasons: the 'short rains' in October-December (mean monthly rainfall 160 mm) and the 'long rains' in March-May (mean monthly rainfall 300 mm).Malaria endemicity: perennial with peaks after the rainy seasonEIR: baseline characteristics measured by the study reported a mean per month in the control arm of 1.1 (range 0.4-2.8) and 1.3 (range 0.4-4.4) in the intervention armPopulation proximity/density: not reportedPlasmodium spp:P falciparum\n\n\nVector profile: Primary (and secondary) vector species:An gambiae s.s. and An arabiensisVector behaviour (nature, stability, adult habitat, peak biting times, exophilic/endophilic, exophagic/endophagic, anthropophilic/zoophilic): not reportedPhenotypic resistance profile: resistance to pyrethroids in An gambiae s.s.Genotypic resistance profile: not reportedMethod of mosquito collection: CDC light traps indoors\n\n\nNotes: For inclusion in the review meta-analyses, we calculated adjusted risk ratios for prevalence from the reported adjusted odds ratios following the methodology stated in Section 12.5.4.4 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a)\n\n",
        "objective": "To summarize the effect on malaria of additionally implementing IRS, using non\u2010pyrethroid\u2010like or pyrethroid\u2010like insecticides, in communities currently using ITNs.",
        "full_paper": "Philippa West and colleagues compare Plasmodium falciparum infection prevalence in children, anemia in young children, and entomological inoculation rate between study arms.\nPlease see later in the article for the Editors' Summary\nBackground\nInsecticide-treated nets (ITNs) and indoor residual spraying (IRS) of houses provide effective malaria transmission control.\nThere is conflicting evidence about whether it is more beneficial to provide both interventions in combination.\nA cluster randomised controlled trial was conducted to investigate whether the combination provides added protection compared to ITNs alone.\nMethods and Findings\nIn northwest Tanzania, 50 clusters (village areas) were randomly allocated to ITNs only or ITNs and IRS.\nDwellings in the ITN+IRS arm were sprayed with two rounds of bendiocarb in 2012.\nPlasmodium falciparum prevalence rate (PfPR) in children 0.5\u201314 y old (primary outcome) and anaemia in children <5 y old (secondary outcome) were compared between study arms using three cross-sectional household surveys in 2012.\nEntomological inoculation rate (secondary outcome) was compared between study arms.\nIRS coverage was approximately 90%.\nITN use ranged from 36% to 50%.\nIn intention-to-treat analysis, mean PfPR was 13% in the ITN+IRS arm and 26% in the ITN only arm, odds ratio\u200a=\u200a0.43 (95% CI 0.19\u20130.97, n\u200a=\u200a13,146).\nThe strongest effect was observed in the peak transmission season, 6 mo after the first IRS.\nSubgroup analysis showed that ITN users were additionally protected if their houses were sprayed.\nMean monthly entomological inoculation rate was non-significantly lower in the ITN+IRS arm than in the ITN only arm, rate ratio\u200a=\u200a0.17 (95% CI 0.03\u20131.08).\nConclusions\nThis is the first randomised trial to our knowledge that reports significant added protection from combining IRS and ITNs compared to ITNs alone.\nThe effect is likely to be attributable to IRS providing added protection to ITN users as well as compensating for inadequate ITN use.\nPolicy makers should consider deploying IRS in combination with ITNs to control transmission if local ITN strategies on their own are insufficiently effective.\nGiven the uncertain generalisability of these findings, it would be prudent for malaria control programmes to evaluate the cost-effectiveness of deploying the combination.\nTrial registration\nwww.ClinicalTrials.gov NCT01697852\nPlease see later in the article for the Editors' Summary\nEditors' Summary\nBackground\nEvery year, more than 200 million cases of malaria occur worldwide, and more than 600,000 people, mainly children living in sub-Saharan Africa, die from this parasitic infection.\nMalaria parasites, which are transmitted to people through the bites of infected night-flying mosquitoes, cause a characteristic fever that needs to be treated promptly with antimalarial drugs to prevent anaemia (a reduction in red blood cell numbers) and organ damage.\nPrompt treatment also helps to reduce malaria transmission, but the mainstays of global malaria control efforts are the provision of insecticide-treated nets (ITNs) for people to sleep under to avoid mosquito bites, and indoor residual spraying (IRS) of houses with insecticides, which prevents mosquitoes from resting in houses.\nBoth approaches have been scaled up in the past decade.\nAbout 54% of households in Africa now own at least one ITN, and 8% of at-risk populations are protected by IRS.\nAs a result of the widespread deployment of these preventative tools and the increased availability of effective antimalarial drugs, malaria-related deaths in Africa fell by 45% between 2000 and 2012.\nWhy Was This Study Done?\nSome countries have chosen to use ITNs and IRS in combination, reasoning that this will increase the proportion of individuals who are protected by at least one intervention and may provide additional protection to people using both interventions rather than one alone.\nHowever, providing both interventions is costly, so it is important to know whether this rationale is correct.\nIn this cluster randomised controlled trial (a study that compares outcomes of groups of people randomly assigned to receive different interventions) undertaken in the Muleba District of Tanzania during 2012, the researchers investigate whether ITNs plus IRS provide more protection against malaria than ITNs alone.\nMalaria transmission occurs throughout the year in Muleba District but peaks after the October\u2013December and March\u2013May rains.\nNinety-one percent of the district's households own at least one ITN, and 58% of households own enough ITNs to cover all their sleeping places.\nAnnual rounds of IRS have been conducted in the region since 2007.\nWhat Did the Researchers Do and Find?\nThe researchers allocated 50 communities to the ITN intervention or to the ITN+IRS intervention.\nDwellings allocated to ITN+IRS were sprayed with insecticide just before each of the malaria transmission peaks in 2012.\nThe researchers used household surveys to collect information about ITN coverage in the study population, the proportion of children aged 0.5\u201314 years infected with the malaria parasite Plasmodium falciparum (the prevalence of infection), and the proportion of children under five years old with anaemia.\nIRS coverage in the ITN+IRS arm was approximately 90%, and 50% of the children in both intervention arms used ITNs at the start of the trial, declining to 36% at the end of the study.\nIn an intention-to-treat analysis (which assumed that all study participants got the planned intervention), the average prevalence of infection was 13% in the ITN+IRS arm and 26% in the ITN arm.\nA per-protocol analysis (which considered data only from participants who received their allocated intervention) indicated that the combined intervention had a statistically significant protective effect on the prevalence of infection compared to ITNs alone (an effect that is unlikely to have arisen by chance).\nFinally, the proportion of young children with anaemia was lower in the ITN+IRS arm than in the ITN arm, but this effect was not statistically significant.\nWhat Do These Findings Mean?\nThese findings provide evidence that IRS, when used in combination with ITNs, can provide better protection against malaria infection than ITNs used alone.\nThis effect is likely to be the result of IRS providing added protection to ITN users as well as compensating for inadequate ITN use.\nThe findings also suggest that the combination of interventions may reduce the prevalence of anaemia better than ITNs alone, but this result needs to be confirmed.\nAdditional trials are also needed to investigate whether ITN+IRS compared to ITN reduces clinical cases of malaria, and whether similar effects are seen in other settings.\nMoreover, the cost-effectiveness of ITN+IRS and ITN alone needs to be compared.\nFor now, though, these findings suggest that national malaria control programs should consider implementing IRS in combination with ITNs if local ITN strategies alone are insufficiently effective and cannot be improved.\nAdditional Information\nPlease access these websites via the online version of this summary at http://dx.doi.org/10.1371/journal.pmed.1001630.\nInformation is available from the World Health Organization on malaria (in several languages), including information on insecticide-treated bed nets and indoor residual spraying; the World Malaria Report 2013 provides details of the current global malaria situation\nThe US Centers for Disease Control and Prevention provides information on malaria, on insecticide-treated bed nets, and on indoor residual spraying; it also provides a selection of personal stories about malaria\nInformation is available from the Roll Back Malaria Partnership on the global control of malaria and on the Global Malaria Action Plan (in English and French); its website includes fact sheets about malaria in Africa and about nets and insecticides\nMedlinePlus provides links to additional information on malaria (in English and Spanish)\nIntroduction\nIn the past decade, insecticide-treated net (ITN) distribution has been scaled up across Africa in line with the Abuja Declaration in 2000.\nThe percentage of households that owned at least one ITN in Africa increased from 3% in 2000 to 54% in 2013.\nThe World Health Organization (WHO) policy that ITNs should be provided to everyone in malaria risk areas (universal coverage) has been adopted by 34 of the 44 malaria endemic countries in Africa.\nIndoor residual spraying (IRS) of houses, the second major vector control tool used to prevent malaria, has similarly been scaled up.\nThe proportion of at-risk populations protected by IRS increased from less than 5% in 2005 to 8% in 2012.\nAs a result of the increase in the deployment of these preventive tools and the increased availability and use of artemisinin-based combination therapies, malaria-related mortality fell by 45% between 2000 and 2012 in Africa, but there remained an estimated 165 million cases and 562,000 deaths due to malaria in 2012.\nIn an attempt to reduce the malaria burden further, a number of countries have chosen to use ITNs and IRS in combination.\nFifty-seven countries, 31 of which are in Africa, use both IRS and ITNs, in at least some areas.\nApplying ITNs and IRS in the same area can increase the proportion of individuals who are protected by at least one intervention or, more optimally, may provide additional protection for those protected by both interventions compared to those receiving one method alone.\nSince the cost of implementing both IRS and universal coverage of ITNs is much greater than the cost of implementing only one of the interventions, it is important to know what extra protection is gained by adding a second intervention, to help national malaria control programmes and international funding agencies such as the President's Malaria Initiative (PMI) and the Global Fund to Fight AIDS, Tuberculosis and Malaria make decisions that are based on evidence of likely impacts and costs.\nThis is particularly significant now, since it is estimated that global funding for malaria is less than half of what is needed to attain universal coverage of malaria vector control, i.e., access to either ITNs or IRS.\nIt is unclear from current evidence whether combined use of ITNs and IRS provides an additional benefit compared to using either intervention alone, and whether this will be similar across transmission settings.\nA recent trial in Benin found no added benefit to using IRS in combination with ITNs compared to ITNs alone.\nHowever, this trial had a relatively small sample size, and its findings may be applicable to only a particular transmission setting in west Africa.\nTo help define future malaria control policy in Africa, the PMI decided to sponsor an independent two-arm cluster randomised controlled trial (CRT) to compare the protective effectiveness of IRS in combination with high coverage of ITNs with high coverage of ITNs alone for malaria transmission control.\nTanzania has a high malaria disease burden, with a national average of 9% of children under 5 y being infected with malaria parasites.\nMalaria control activities have been scaled up nationally since 2005.\nA universal coverage campaign (UCC) primarily funded by the Global Fund to Fight AIDS, Tuberculosis and Malaria distributed long-lasting insecticidal nets (LLINs) free of charge in 2011 to top up coverage from previous distributions.\nIRS, funded by the PMI, commenced in 2007 in two districts of Kagera Region, in northwest Tanzania, and has since been extended to cover 18 districts.\nBecause IRS is costly and logistically intensive, there is an urgent need to know whether it is necessary to continue with IRS after an ITN UCC has been successfully completed.\nThe trial was carried out in 109 rural villages in Muleba District (1\u00b045\u2032S 31\u00b040\u2032E), Kagera Region.\nThe study area includes 68,108 households at an altitude ranging from 1,100 to 1,600 m above sea level.\nRainfall occurs in two seasons: the \u201cshort rains\u201d in October\u2013December (average monthly rainfall 160 mm) and the \u201clong rains\u201d in March\u2013May (average monthly rainfall 300 mm), with malaria transmission occurring throughout the year and peaking after the rainy seasons.\nAnnual rounds of IRS with the pyrethroid lambda-cyhalothrin (ICON 10CS, Syngenta) were conducted between 2007 and 2011 in Muleba District, i.e., in the entire study area.\nThe predominant malaria vectors are Anopheles gambiae s.s. and An. arabiensis .\nTests of mosquito susceptibility using standard WHO bioassays showed resistance to pyrethroids in An. gambiae s.s. in 2011.\nAs a result, IRS policy was changed to use the carbamate insecticide bendiocarb (Ficam 80% wettable powder, Bayer) by the PMI in 2012.\nMethods\nEthics and Community Sensitisation\nThe trial was approved by the ethics review committees of the Kilimanjaro Christian Medical College, the Tanzanian National Institute for Medical Research, and the London School of Hygiene and Tropical Medicine.\nWritten informed consent was obtained from all respondents.\nPrior to the baseline surveys, village and hamlet leaders were invited to sensitisation sessions conducted by district health officers.\nThe trial was registered with ClinicalTrials.gov (registration number NCT01697852) in September 2012.\nThe trial was not registered earlier because the authors were not aware of journal requirements for prospective registration.\nAll authors have affirmed that any trials they are involved in on the same or a related drug or intervention are registered.\nAn accurate summary of the trial's results has been submitted to ClinicalTrials.gov.\nStudy Design\nA CRT was conducted, comparing the Plasmodium falciparum prevalence rate (PfPR) in children 0.5\u201314 y old between communities targeted to receive both high-coverage IRS and high coverage of ITNs (ITN+IRS arm) and communities targeted for high coverage of ITNs only (standard-care control arm).\nSecondary outcomes were moderate/severe anaemia (haemoglobin <8 g/dl) in children under 5 y old and entomological inoculation rate (EIR) due to An. gambiae s.l.\nPower calculations showed that 25 clusters per study arm were required, with 80 children per cluster, to give 80% power to detect a true absolute difference in PfPR of at least 3% between study arms (relative difference 31%) with 5% significance (two-sided), based on an expected prevalence in the ITN only arm of 9% (PfPR in first baseline survey).\nThe between-cluster coefficient of variation (k) was calculated as 0.25 from the first pre-randomisation baseline survey.\nEach cluster consisted of at least one village and was divided into a core surveillance area consisting of at least 200 houses and approximately 1 km radius, where the surveys were conducted, and an outer buffer zone, 1 km in width, which also received the allocated treatment but in which no outcome monitoring was done.\nVillages were eligible for inclusion in the study if they were within daily commuting distance for survey work and had been sprayed with IRS in the baseline year.\nAll clusters received LLINs from the UCC in 2011.\nTwenty-five clusters were randomly allocated to receive IRS, in addition to ITNs, using restricted randomisation to limit potential imbalance between study arms.\nBaseline surveys provided data on seven criteria for which the study arms were balanced by constraining the randomisation (Table 1).\n200,000 random allocations were generated.\nMean values for each arm were calculated from cluster summaries for each of the seven restriction variables; 25,119 randomisations fulfilled the restriction criteria and were therefore eligible.\nThese allocations were tested for independence between any two clusters.\nThe large number of acceptable allocations, of which one was randomly selected, ensured that the restriction did not affect the validity of inference.\nThere was no evidence of dependence between any pair of clusters.\nInterventions\nHouseholds in the study area with children aged under 5 y received LLINs from a national distribution campaign in 2009.\nIn 2011, the district health authority, supported by Mennonite Economic Development Associates, completed a UCC that distributed 144,000 LLINs (Olyset, Sumitomo Chemicals) to the population of Muleba District, including all study clusters.\nThe campaign aimed to top up net coverage, so that every sleeping place had one ITN.\nAfter the UCC, 91% of households owned at least one ITN, and 58% of households owned enough ITNs to cover all their sleeping places.\nSpraying was conducted by RTI International on behalf of PMI in the ITN+IRS study arm.\nThe interior walls of each dwelling were sprayed with the carbamate insecticide bendiocarb (Ficam 80% wettable powder, Bayer) at 400 mg/m2 between December 2011 and January 2012 (round 1), and between April and May 2012 (round 2).\nSpray rounds were timed to precede the peak in malaria cases that normally occurs at the end of each rainy season, taking into account the relatively short residual duration of bendiocarb.\nBendiocarb is a carbamate insecticide recommended by WHO for IRS.\nIt is one of the few insecticides evaluated and approved by the WHO Pesticide Evaluation Scheme that has the potential to control pyrethroid-resistant mosquitoes, is odour-free, and is safe to house occupants at the recommended application rate.\nBefore obtaining WHO approval, all IRS insecticides are subject to risk assessment by WHO toxicologists.\nBendiocarb is an acetylcholinesterase inhibitor, but no serious adverse effects due to bendiocarb IRS have been reported in the recent medical literature.\nSurveys\nThree post-intervention cross-sectional household surveys were undertaken in 2012 (see Figure 1).\nSurvey A (23 February\u201331 March) was after the short rainy season and 2 mo after the first spray round.\nSurvey B (25 June\u201331 July) was after the long rainy season, 6 mo after the first spray round, and 2 mo after the second spray round.\nSurvey C (25 October\u20134 December) was 6 mo after the second spray round and 10 mo after the first.\nBaseline surveys were conducted in 2011 during the same periods as surveys A and B.\nFor each survey, 80 households were randomly selected in the core area of each cluster.\nHouseholds were eligible for the study if they had children aged 0.5\u201314 y.\nAny child aged 0.5\u201314 y was eligible to be included in the study.\nUp to three children per household were randomly selected for testing.\nAllowing for ineligible households, absence on the day of the survey, and refusals at the household and individual level, it was estimated that this would provide on average 80 children for testing per cluster.\nThe household head or another responsible adult from the household was interviewed, after seeking written informed consent.\nData on IRS coverage, bed net ownership and usage, demographics of household members, and other household characteristics were gathered using an adapted version of the standard Malaria Indicator Survey.\nSelected children were tested on the following day for malaria parasites using a rapid diagnostic test (RDT) (CareStart [Pan] Malaria, DiaSys) and had haemoglobin levels measured using HemoCue Hb 201+ (Aktiebolaget Leo Diagnostics).\nIndividuals testing positive by RDT were treated with artemether/lumefantrine (Artefan 20/120, Ajanta Pharma) following national treatment guidelines.\nEntomological surveillance was carried out in the core surveillance areas of a subset of 40 of the 50 clusters from April 2011 to December 2012.\nFor one night of each month US Centers for Disease Control and Prevention light traps for mosquito collections were set up in eight randomly selected houses in each cluster (320 houses per month).\nAnopheles mosquitoes collected were identified to species using a simplified morphological key adapted from Gillies and Coetzee.\nA sub-sample of An. gambiae s.l. individuals were tested using real-time PCR TaqMan assay to distinguish between the two sibling species An. gambiae s.s. and An. arabiensis .\nMosquitoes were also tested for P. falciparum sporozoites (P. falciparum circumsporozoite protein) using ELISA.\nStatistical Analysis\nStatistical analysis was done in Stata 12 (Statacorp) and R version 2.13.1 (R Foundation for Statistical Computing).\nThe odds of PfPR and moderate/severe anaemia for individuals were compared between study arms in intention-to-treat (ITT) analysis using logistic regression.\nMean haemoglobin was compared between the study arms using linear regression.\nA robust variance estimator was used to calculate standard errors to adjust for within-cluster correlation of responses (Stata survey commands, first-order Taylor-series linearization method).\nPfPR was considered as P. falciparum alone or mixed infections as detected by the RDT.\nThe overall odds ratio (OR) for the three surveys combined was calculated accounting for survey.\nAn adjusted Wald test was performed to test whether there was evidence for effect modification between study arm and survey round.\nA sensitivity analysis was conducted excluding one cluster from the ITN only arm that mistakenly received IRS, to assess the impact of this protocol violation on the results of ITT analysis.\nBecause of the wide variation in cluster-level estimates of PfPR at baseline, an OR for ITN+IRS versus ITN alone was calculated adjusting for baseline PfPR.\nA secondary per-protocol analysis was performed, in which individuals from the ITN+IRS arm who used an ITN and lived in a house sprayed in the most recent round of IRS were compared to individuals who used an ITN in the ITN only arm.\nThe cluster that violated the protocol was excluded from the per-protocol analysis.\nThe monthly EIR was calculated as the daily EIR found during the one night collection multiplied by the number of days in the month.\nMean EIRs were compared between study arms using negative binomial regression and adjusting for within-cluster correlation.\nResults\nAt baseline, PfPR, anaemia, ITN ownership, ITN usage, and mean EIR per month (Table 2) were similar in the two study arms.\nPfPR in children aged 6 mo to 14 y old was 9.3% (95% CI 5.9%\u201314.5%) after the short rains (survey A, February\u2013March) and 22.8% (95% CI 17.3%\u201329.4%) after the long rains (survey B, June\u2013July).\nAnaemia in children 0.5\u20134 y was 6.2% (95% CI 4.5%\u20138.5%) after the long rains.\nOf the 2,000 houses selected in each study arm for each post-intervention survey, 20% to 24% had no children between 0.5 and 14 y old (were ineligible), 13% to 18% were vacant on the day of survey, fewer than 1% refused to participate, and 55% to 61% participated in the survey (Figure 2).\nOf the children selected for RDT, 81%\u201384% were tested.\nPost-intervention IRS coverage reported by householders was 92.1% after the first spray round and 89.5% after the second (Table 3).\nIn the intervention year, the percentage of houses with sufficient ITNs for each sleeping place remained stable over successive surveys and was similar between study arms (range 52%\u201357%; Table 3).\n82.2% and 87.0% of households owned at least one ITN in the ITN only arm and the ITN+IRS arm, respectively (all surveys combined), with weak evidence that the percentage of households that owned at least one ITN was lower in the ITN only arm, and that it decreased from survey A to survey C in both arms (Table 3).\nITN usage in children was similar between study arms but declined from 50% in survey A to 36% in survey C.\nThe primary outcome PfPR was lower in the ITN+IRS arm than in the ITN only arm in all three surveys in the intervention year (Table 4).\nFor all three surveys combined, the overall OR was 0.43 (95% CI 0.19\u20130.97), with weak evidence that the intervention effect differed between surveys (interaction p\u200a=\u200a0.08).\nThe strongest effect was observed in survey B (OR 0.33, 95% CI 0.15\u20130.75), which was conducted at the peak of malaria transmission after the long rains, 6 mo after the first IRS and 2 mo after the second IRS.\nThe evidence for an effect was weaker in survey A (OR 0.51, 95% CI 0.24\u20131.09), conducted shortly after the first IRS round, and in survey C (OR 0.48, 95% CI 0.18\u20131.24), conducted several months after the main transmission season and 6 mo after last spray round.\nThe range of cluster-specific estimates for PfPR was 0% to 92% in the ITN only arm and 0% to 68% in the ITN+IRS arm.\nThe sensitivity analysis showed that excluding the cluster from the ITN only arm that had received IRS did not affect the results of the ITT analysis (Table S1).\nThe overall OR for all three surveys combined was very similar after adjusting for baseline PfPR, OR\u200a=\u200a0.41, but the precision of the estimate was increased (95% CI 0.29\u20130.59, p<0.0001).\nPrevalence of moderate to severe anaemia in children under 5 y old, a secondary outcome, was lower in the ITN+IRS arm in all post-intervention surveys, but the difference was statistically significant only in survey B (Table 5).\nMean haemoglobin was higher in children under 5 y old in the ITN+IRS arm than in the ITN only arm in all three surveys.\nThe evidence for an effect was greatest in survey B (0.49 g/dl, 95% CI 0.10\u20130.89, p\u200a=\u200a0.016), with a non-significant result in survey A (0.28 g/dl, 95% CI \u22120.02 to 0.59, p\u200a=\u200a0.065) and survey C (0.36 g/dl, 95% CI \u22120.02 to 0.73, p\u200a=\u200a0.060).\nMean EIR per month, a secondary outcome, was 0.22 in the ITN+IRS arm and 1.26 in the ITN only arm (rate ratio\u200a=\u200a0.17, 95% CI 0.03\u20131.08, p\u200a=\u200a0.059; Table 6).\nThe between-cluster coefficient of variation (k) was 0.20, 0.28, and 0.26 in the three post-intervention surveys, respectively.\nFor each survey, k was similar in the two arms.\nFor all surveys, per-protocol analysis showed statistically significant evidence for a protective effect of the combined intervention on PfPR (survey A: OR 0.39, 95% CI 0.18\u20130.81; survey B: OR 0.21, 95% CI 0.09\u20130.49; and survey C: OR 0.27, 95% CI 0.10\u20130.73; Table 7).\nDiscussion\nThis is the first randomised trial to our knowledge that provides evidence that IRS, when used in combination with ITNs, can give significant added protection against malarial infection compared to ITN use alone.\nThere was also some evidence that anaemia prevalence was lower in communities with the combination.\nExposure to infectious mosquito bites was about one-sixth in communities with the combined intervention compared to those in the ITN only arm.\nTwo rounds of IRS with bendiocarb were conducted to overcome the short residual activity of the insecticide and to ensure that there was active ingredient on the walls of sprayed homes throughout the transmission season.\nIRS coverage in the ITN+IRS arm was high at approximately 90% in both spray rounds, which would have optimised its effectiveness.\nOn the other hand, whilst 85% of households owned at least one ITN, use of ITNs was modest, declining to 36% by the end of the study.\nThe low usage of ITNs means that the addition of IRS may have simply protected those who were not using an ITN, thus compensating for low ITN usage rather than offering additional protection to net users.\nThis interpretation is contradicted by the results of a per-protocol analysis, which excluded those not using ITNs, showing strong evidence that ITN users whose houses were sprayed were additionally protected by IRS.\nThe estimated reduction in PfPR associated with the combination of interventions was greater in the per-protocol analysis than in the ITT analysis in each survey.\nPer-protocol analysis excludes non-compliers (for IRS and ITN) and therefore may have been influenced by confounders.\nIt is likely that the observed overall effect of the intervention combination was a result of both IRS protecting those not using ITNs, and IRS additionally protecting ITN users.\nA potential negative impact of the combination of interventions is that having their house sprayed may encourage some residents to stop sleeping under an ITN.\nThis was not observed in this study; ITN usage was similar between the villages with and without IRS in each post-intervention survey.\nITN usage and ownership was slightly higher at baseline in the ITN+IRS arm compared to the ITN only arm, but the 95% confidence intervals for these estimates overlapped.\nThis non-significant difference could have led to a slight overestimation of the effect size.\nPfPR was slightly lower at baseline in the ITN+IRS arm compared to the ITN only arm, but the effect size did not change after adjusting for PfPR at baseline.\nThis suggests that baseline PfPR was not confounding the relationship between study arm and PfPR (the outcome).\nIn the baseline year, malaria prevalence was higher in June\u2013July after the long rainy season than in February\u2013March after the short rains.\nIn the intervention year, the prevalence similarly increased in June\u2013July (survey B) in the ITN only arm, but prevalence in the ITN+IRS arm remained low, suggesting IRS and ITNs in combination prevented the seasonal increase in infections.\nThe added protective effect of IRS peaked in the second survey, at the height of transmission after the long rains.\nThis was probably the optimal time for the insecticide to reduce the abundance of the mosquito population (N. Protopopoff, personal communication) and thus to observe the impact of IRS on the prevalence of malarial infections.\nThe limited residual activity of bendiocarb IRS has been shown to reduce its protective effectiveness 3\u20135 mo after spraying, which probably accounts for the loss of added benefit seen in the third survey, which was 6 mo after the last spray round at the beginning of the short rains.\nImplementing IRS with long-lasting insecticide formulations might be necessary to maintain the effectiveness of the combination throughout the year.\nAlternatively, the time between IRS rounds could be reduced, but this would considerably raise the cost of the combined intervention.\nThe secondary outcomes anaemia and EIR also pointed to added protection being provided by the combination of IRS and ITNs, but the evidence for these endpoints was weaker.\nThe combination intervention was associated with higher haemoglobin levels in children under 5 y, particularly at the peak of the transmission season.\nThe study had been powered to show a difference in the primary outcome (PfPR), and therefore may have been underpowered for these secondary outcomes.\nNevertheless, the results for all outcomes are consistent.\nOne of the limitations of this study is that clinical incidence of malaria could not be recorded in addition to infection prevalence because recording of confirmed malaria cases was unreliable because of stock-outs of RDTs at health facilities.\nImplementing both IRS and universal coverage of ITNs is obviously considerably more costly than ITNs alone.\nEstimating the cost-effectiveness of the combination compared to ITNs alone was beyond the scope of this particular research.\nAlthough IRS is known to be highly cost-effective, the marginal cost per case averted through using IRS in combination with ITNs should ideally be assessed in future studies.\nThis is particularly important in light of the funding gap that has been identified for meeting the demand for universal coverage of vector control for populations in malaria endemic regions.\nPrevious studies have investigated the combined use of multiple vector control methods versus one method alone, but the results have been inconsistent.\nThe only published trial data are from a 28-cluster, four-arm CRT carried out in Benin that compared (1) targeted coverage of LLINs (pregnant women and children only), (2) universal coverage of LLINs, (3) targeted coverage of LLINs combined with bendiocarb IRS, and (4) universal coverage of LLINs combined with bendiocarb-treated wall linings.\nThe study found no difference in malaria incidence, geometric mean parasite density, or mosquito abundance between any of the study arms.\nThe lack of any evidence of an added benefit of the combined interventions over the use of LLINs alone has to be viewed against the modest sample size, and hence potentially low power of this trial, and the lack of a comparator arm with universal coverage of ITNs.\nThere are a number of differences between the Benin trial and the current study that may have contributed to the discordant results.\nIn the Benin trial, the interval between IRS rounds was 8 mo, whereas it was only 4 mo in the current study, as IRS was timed according to the seasonal peaks in cases, and taking account of its short residual duration on walls.\nThe first two cross-sectional surveys for the current trial were timed to coincide with the seasonal peaks in cases and were only 2 mo after each IRS round, whereas in Benin the cases were recorded at 6-wk intervals for 18 mo, so that the measured effect of the additional IRS may include a period when the insecticide, which is known to have a short residual duration, was no longer effective.\nIn the Benin trial, LLINs were given only to target groups in the reference arm and in the study arm with IRS, whereas in the current trial ITNs were distributed to all age groups.\nLarge CRTs have recently been conducted in the Gambia and in Sudan comparing villages with IRS and LLINs to villages with only LLINs, but the results have not yet been published.\nEvidence of an added benefit from the combination intervention compared to IRS or ITNs alone has been shown in a number of observational studies.\nFor example, children 2\u201314 y old consistently received added personal protection from using nets in addition to IRS on the island of Bioko, Equatorial Guinea (OR 0.71, 95% CI 0.59\u20130.86), and in Zambezia, Mozambique (OR 0.63, 95% CI 0.50\u20130.79).\nIn Pakistan, nets provided added protection against P. vivax and P. falciparum in refugee camps where IRS was conducted.\nHowever, other studies observed no additional benefit from the combination compared to one intervention alone.\nOne interpretation of these divergent conclusions is that if the intervention present in both study arms is compromised or poorly implemented, the second method compensates for the deficiency of the first, providing apparent added protection that would otherwise not be seen.\nOn the other hand, if the reference arm intervention is well implemented and efficacious in both study arms, there may be little or no scope for additional protection by a second intervention.\nITN usage in the present trial was moderate, and hence the IRS protected many people who were not using a net in the ITN+IRS arm, whilst non-users in the ITN only arm remained unprotected.\nAny community or \u201cmass effect\u201d of ITNs on mosquito population size would have been limited because of the low community net usage.\nTherefore, the protective effect of ITNs in this study was possibly suboptimal.\nIn Bioko, ITNs provided personal protection in the presence of IRS that was rendered only partially effective by moderate coverage (77%\u201379%) and use of an insecticide that did not outlast the long malaria season.\nProtopopoff et al. reported that in Burundi there was no additional reduction in infection prevalence in children from adding LLINs to IRS because high coverage (90%) of IRS had already reduced the sporozoite rate to a level where nets had no further impact.\nIn Sao Tome, where the IRS programme was poorly implemented, with low coverage and long intervals between spray rounds, there was an additional benefit from using ITNs and IRS compared to IRS alone.\nHowever, on the neighbouring island of Principe, where IRS coverage was high (85%) and implemented on schedule, there was no added protection from ITNs in combination with IRS compared to IRS alone.\nInsecticide resistance may be another reason why differences have been seen for the effectiveness of the combination of IRS and ITNs, resulting in either an apparent \u201cadded\u201d effect of the second effective intervention, if the first was ineffective due to insecticide resistance, or no added effect if the second intervention was ineffective due to insecticide resistance.\nIn the study area of this trial, there was evidence for high levels of resistance to pyrethroids in An. gambiae s.s.\nThe epidemiological impact of pyrethroid resistance on the effectiveness of ITNs is currently not known.\nHowever, if the effectiveness of the ITNs was compromised because of insecticide resistance, this would have enhanced our estimate of the additional benefit of non-pyrethroid IRS.\nIf pyrethroid-treated nets were to be rendered partially ineffective in the presence of resistance, there would be a compelling case for combining ITNs with non-pyrethroid IRS.\nAn experimental hut trial in an area of Tanzania where the main vector is An. arabiensis found that if ITNs were used, the addition of IRS using insecticides with high irritancy such as dichlorodiphenyltrichloroethane (DDT) or lambda-cyhalothrin did not increase mosquito mortality or repel mosquitoes from the house.\nHowever, the addition of IRS using pirimiphos-methyl, an organophosphate that has high toxicity and low irritancy, did increase mosquito mortality.\nThese findings underscore that the interaction between the two interventions is complex and that the added protective effect will be dependent on the feeding and resting behaviours of particular malaria vectors, on the type of IRS insecticide used, on the susceptibility of local vectors to each of the insecticides in the combination, and on ITN usage.\nAs a result, added protection may not be observed in all situations.\nA systematic review of all the trial results estimating the effectiveness of the combination of ITNs and IRS should be undertaken once the results of the trials in Sudan and the Gambia are available.\nNevertheless, this trial provides encouraging evidence for an additional benefit from applying IRS in combination with ITNs compared to ITNs alone.\nTo our knowledge it is the first CRT to do so.\nThe added protection from the supplementary use of IRS may in the case of bendiocarb be limited to only a few months, raising the question of whether residual insecticides of short duration are cost-effective when used in combination with ITNs.\nThis study was conducted as an effectiveness study and not an efficacy study.\nThe LLINs were distributed by a national UCC and therefore represented a real-life malaria control programme, including the challenges faced in achieving high coverage and usage of ITNs.\nIn conclusion, national malaria control programmes should consider implementing IRS in combination with ITNs if local ITN strategies alone are insufficiently effective and cannot be improved.\nA key consideration would be the additional cost of providing the combined intervention.\nGiven the inconsistent trial evidence and the unproven generalisability of the findings of all studies that have investigated this question, it would be prudent for malaria control programmes implementing the two methods simultaneously to monitor the impact and cost-effectiveness of the combination to verify whether the additional resources have the desired effect.\nStudy timetable.Surveys 1 and 2 are baseline surveys. Surveys A, B, and C are post-intervention.\nTrial profile for study households and children in the ITN only and ITN+IRS study arms.Survey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray. *No children 0.5\u201314 y old. 1Dwelling vacant for survey duration. 2Includes not found (91.0%), not visited (2.4%), and missing data (6.6%). 3Households (HH) that were included and where children attended for testing.\n\nRestriction variables for randomisation and realisation of balance between the study arms.\nVariable | Maximum Difference in Means between Study Armsa | ITN Arma | ITN+IRS Arma | Actual Difference\nPfPRb in February\u2013March 2011c | 3% | 9.9% | 9.3% | 0.5%\nPfPR in June\u2013July 2011d | 3% | 22.4% | 19.6% | 2.7%\nHousing densitye | 20 HH/km2 | 165.1 HH/km2 | 152.6 HH/km2 | 12.5 HH/km2\nMean elevation | 50 m | 1,364.8 m | 1,330.7 m | 34.1 m\nITN usaged,f | 5% | 35.0% | 30.4% | 4.6%\nAdequate LLIN ownershipe,g | 5% | 61.3% | 56.3% | 5.0%\nClusters with entomological surveillance | Count of 2 | 20 clusters | 20 clusters | 0 clusters\n\nMeans for each study arm were calculated from cluster summaries.\n\nPfPR from RDTs.\nRecorded in baseline survey 1(February\u2013March 2011).\nRecorded in baseline survey 2 (June\u2013July 2011) after the UCC.\nHousing density in surveillance area of clusters.\nNet used the night before the survey in all age groups.\nPercentage of households with at least one LLIN per two people.\nHH, household.\n\nBaseline characteristics of individuals and households by study arm, Muleba District, 2011.\nCharacteristic | ITN Only ArmPercent [95% CI] (n) | ITN+IRS ArmPercent [95% CI] (n)\nPfPR in March 2011a,b,c | 10.3 [5.2\u201319.3] (2,487) | 8.4 [4.5\u201315.3] (2,655)\nPfPR in July 2011a,b,d | 24.6 [17.0\u201334.3] (2,121) | 21.0 [13.8\u201330.5] (2,185)\nModerate/severe anaemiaa,d,e | 6.4 [3.9\u201310.2] (785) | 6.1 [4.1\u20138.9] (841)\nMean haemoglobin (g/dl)a,d, | 10.6 [10.4\u201310.9] (785) | 10.6 [10.4\u201310.9] (841)\nITN use in all age groupsa,d,f | 53.3 [48.2\u201358.3] (6,755) | 58.2 [53.8\u201362.5] (6,913)\nHouseholds with adequate ITNsd,g,h | 54.5 [49.5\u201359.5] (1,243) | 62.3 [57.3\u201367.1] (1,250)\nHouseholds with \u22651 ITNd,g | 88.9 [86.0\u201391.3] (1,248) | 92.6 [90.8\u201394.0] (1,251)\nHouseholds received IRS in 2011c,g,i | 94.4 [91.3\u201396.5] (1,598) | 95.5 [93.5\u201396.9] (1,640)\nMean An. gambiae mosquitoes per house per nightg,j | 3.1 [1.0\u20139.6] (1,055) | 2.2 [0.5\u20139.1] (1,120)\nSporozoite ratea,k | 1.1 [0.8\u20131.4] (1,359) | 2.0 [1.4\u20132.8] (1,466)\nMean EIR/monthl | 1.1 [0.4\u20132.8] | 1.3 [0.4\u20134.4]\n\nCalculated from individual-level data.\n\nPfPR from RDTs.\nRecorded in baseline survey 1 (February\u2013March 2011).\nBaseline survey 2 (June\u2013July 2011) after the UCC.\nHaemoglobin <8 g/dl.\nReported sleeping under an ITN the night previous to the survey.\nCalculated from household-level data.\nAt least one ITN per sleeping place.\nApproximately 1 mo after spraying.\nArithmetic mean.\nProportion of mosquitoes positive for P. falciparum sporozoites.\nNumber of infective bites per month.\n\nIRS coverage, ITN ownership, and ITN usage in the intervention year, Muleba District, 2012.\nSurvey | Arm | Reported IRS CoverageaPercent [95% CI] (nb) | Adequate ITN OwnershipcPercent [95% CI] (nb) | \u22651 ITN OwneddPercent [95% CI] (nb) | ITN UseePercent [95% CI] (nf)\nSurvey A | ITN only | 3.3 [1.8\u20135.9] (1,177) | 52.2 [47.8\u201356.5] (1,178) | 85.8 [83.7\u201387.7] (1,177) | 46.6 [41.7\u201351.6] (2,193)\n | ITN+IRS | 92.1 [88.4\u201394.7] (1,215) | 57.2 [53.6\u201360.7] (1,215) | 89.0 [87.1\u201390.6] (1,216) | 53.0 [47.5\u201358.3] (2,349)\nSurvey B | ITN only | 5.2 [1.3\u201318.6] (1,094) | 51.6 [47.0\u201356.0] (1,094) | 82.5 [78.7\u201385.7] (1,096) | 40.7 [34.7\u201347.0] (2,045)\n | ITN+IRS | 89.5 [84.0\u201393.2] (1,138) | 57.4 [54.0\u201360.9] (1,142) | 88.2 [85.7\u201390.3] (1,142) | 44.1 [39.2\u201349.2] (2,207)\nSurvey C | ITN only | 13.0 [6.6\u201324.1] (1,165) | 52.8 [47.6\u201358.0] (1,168) | 78.2 [74.3\u201381.6] (1,170) | 36.0 [29.8\u201342.6] (2,101)\n | ITN+IRS | 89.3 [83.6\u201393.2] (1,209) | 56.8 [51.7\u201361.8] (1,211) | 83.8 [79.9\u201387.1] (1,211) | 36.1 [31.0\u201341.5] (2,303)\n\nSurvey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\nReported spray status of household in the spray round preceding the survey.\nHouseholds.\nPercentage of households with sufficient ITNs for at least one per sleeping place.\nPercentage of households with at least one ITN.\nPercentage of study children that reported sleeping under an ITN the night previous to the survey. ITN usage in all age groups was very similar to ITN use in the study children.\nIndividuals.\n\n\nPfPR in children 0.5\u201314 y old in the ITN only and ITN+IRS arms (intention to treat) in survey A, B, and C, Muleba District, Tanzania, 2012.\nSurvey | Arm | PfPRaPercent [95% CI] (n) | OR [95% CI], p-Value\nSurvey A | ITN only | 23.6 [15.4\u201334.2] (2,191) | 1.00\n | ITN+IRS | 13.6 [8.3\u201321.4] (2,342) | 0.51 [0.24\u20131.09], p\u200a=\u200a0.082\nSurvey B | ITN only | 30.5 [20.2\u201343.4] (2,033) | 1.00\n | ITN+IRS | 12.7 [7.4\u201321.0] (2,204) | 0.33 [0.15\u20130.75], p\u200a=\u200a0.009\nSurvey C | ITN only | 24.5 [14.2\u201338.9] (2,091) | 1.00\n | ITN+IRS | 13.4 [7.3\u201323.4] (2,285) | 0.48 [0.18\u20131.24], p\u200a=\u200a0.127\nAll three surveys combined | ITN only | 26.1 [16.7\u201338.4] (6,315) | 1.00\n | ITN+IRS | 13.3 [7.9\u201321.5] (6,831) | 0.43 [0.19\u20130.97], p\u200a=\u200a0.043b\n\nSurvey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\n\nPfPR from RDTs.\nAdjusted for survey.\n\nAnaemia and mean haemoglobin in children under 5+IRS arms (intention to treat), for survey A, B, and C, Muleba District, Tanzania, 2012.\nSurvey | Arm | Anaemia Prevalencea | Mean Haemoglobin (g/dl)\n |  | Percent [95% CI] (n) | OR [95% CI], p-Value | Mean [95% CI] (n) | Difference [95% CI], p-Value\nSurvey A | ITN only | 6.0 [4.1\u20138.7] (815) | 1.00 | 10.6 [10.4\u201310.8] (815) | \n | ITN+IRS | 3.9 [2.5\u20136.2] (864) | 0.64 [0.34\u20131.19], p\u200a=\u200a0.155 | 10.9 [10.7\u201311.1] (864) | 0.28 [\u22120.02 to 0.59], p\u200a=\u200a0.065\nSurvey B | ITN only | 4.7 [2.6\u20138.6] (737) | 1.00 | 10.9 [10.6\u201311.2] (737) | \n | ITN+IRS | 2.2 [1.3\u20133.6] (784) | 0.44 [0.20\u20131.01], p\u200a=\u200a0.053 | 11.4 [11.2\u201311.6] (784) | 0.49 [0.10 to 0.89], p\u200a=\u200a0.016\nSurvey C | ITN only | 3.2 [1.8\u20135.7] (739) | 1.00 | 10.8 [10.6\u201311.1] (739) | \n | ITN+IRS | 2.6 [1.6\u20134.4] (831) | 0.81 [0.37\u20131.77], p\u200a=\u200a0.590 | 11.2 [11.0\u201311.4] (831) | 0.36 [\u22120.02 to 0.73], p\u200a=\u200a0.060\nAll three surveys combined | ITN only | 4.7 [3.2\u20136.9] (2,291) | 1.00 | 10.8 [10.5\u201311.0] (2,291) | \n | ITN+IRS | 2.9 [2.0\u20134.3] (2,479) | 0.62 [0.34\u20131.10], p\u200a=\u200a0.102b | 11.2 [11.0\u201311.3] (2,479) | 0.37 [0.07 to 0.68], p\u200a=\u200a0.017b\n\nSurvey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\nPrevalence of moderate/severe anaemia (haemoglobin <8 g/dl).\nAdjusted for survey.\n\nMean number of An. gambiae mosquitoes per household, sporozoite rate, and EIR in the ITN only and ITN+IRS arms during the post-intervention period, Muleba District, Tanzania, 2011\u20132012.\nArm | Mean or Percent [95% CI] (n)a | Effect [95% CI], p-Value\nMeanbAn. gambiae per house per night\nITN only | 1.7 [0.5\u20136.4] (1,892) | \nITN+IRS | 0.4 [0.1\u20131.4] (1,893) | Rate ratio\u200a=\u200a0.23 [0.04\u20131.44], p\u200a=\u200a0.113\nSporozoite ratec\nITN only | 2.5 [2.1\u20133.1] (3,059) | \nITN+IRS | 1.8 [0.5\u20136.2] (717) | OR\u200a=\u200a0.72 [0.21\u20132.53], p\u200a=\u200a0.600\nMean EIR/monthd\nITN only | 1.3 [0.3\u20134.6] | \nITN+IRS | 0.2 [0.1\u20130.8] | Rate ratio\u200a=\u200a0.17 [0.03\u20131.08], p\u200a=\u200a0.059\n\nData are mean [95% CI] (number of houses) for mean An. gambiae per house per night and percent [95% CI] (number of An. gambiae) for sporozoite rate.\nArithmetic mean.\nProportion of mosquitoes positive for P. falciparum sporozoites.\nNumber of infective bites per month.\n\nPer-protocol analysis of PfPR in children 0.5\u201314 y old and anaemia in children under 5 y old in surveys A, B, and C.\nSurvey | Arm | PrevalencePercent [95% CI] (n) | OR [95% CI], p-Value\nPfPRa |  |  | \nSurvey A | ITNb | 26.7 [17.5\u201338.6] (954) | 1.00\n | ITN+IRSc | 12.3 [7.8\u201318.9] (1,142) | 0.39 [0.18\u20130.81], p\u200a=\u200a0.013\nSurvey B | ITNb | 35.5 [23.2\u201350.2] (782) | 1.00\n | ITN+IRSc | 10.2 [5.7\u201317.7] (892) | 0.21 [0.09\u20130.49], p\u200a=\u200a0.001\nSurvey C | ITNb | 29.4 [16.7\u201346.4] (707) | 1.00\n | ITN+IRSc | 10.1 [5.4\u201318.2] (770) | 0.27 [0.10\u20130.73], p\u200a=\u200a0.011\nAnaemiad |  |  | \nSurvey A | ITNb | 5.9 [3.5\u20139.7] (390) | 1.00\n | ITN+IRSc | 3.8 [1.8\u20137.5] (453) | 0.62 [0.25\u20131.55], p\u200a=\u200a0.301\nSurvey B | ITNb | 5.4 [2.2\u201312.5] (295) | 1.00\n | ITN+IRSc | 1.9 [0.8\u20134.1] (374) | 0.33 [0.10\u20131.12], p\u200a=\u200a0.076\nSurvey C | ITNb | 4.0 [2.2\u20137.0] (303) | 1.00\n | ITN+IRSc | 2.3 [1.0\u20135.0] (305) | 0.57 [0.21\u20131.55], p\u200a=\u200a0.264\n\nMuleba, Tanzania, 2012; analysis restricted to ITN users in both study arms. Survey A\u200a=\u200a2 mo after first intervention spray. Survey B\u200a=\u200a6 mo after first intervention spray and 2 mo after second intervention spray. Survey C\u200a=\u200a10 mo after first intervention spray and 6 mo after second intervention spray.\n\nPfPR from RDTs.\nITN used by the individual the night preceding the survey in the ITN only arm.\nITN used by the individual the night preceding the survey, and household with IRS in the ITN+IRS arm. One cluster that was allocated to be in the ITN only arm but received IRS in the second spray round was excluded from this analysis.\nPrevalence of moderate/severe anaemia (haemoglobin <8 g/dl).",
        "label": "low",
        "id": "task4_RLD_test_274"
    },
    {
        "paper_doi": "10.1186/1472-6963-11-92",
        "bias": "blinding of outcome assessment (detection bias) all outcomes",
        "PICO": "Methods: A parallel arm cluster-RCT conducted in Anhui province, Eastern China, between Aug 2000 to Jul 2002.\n\n\nParticipants: Sample size: 20 clusters (1264 individuals).Clusters: townships were selected and paired according to: place of birth (hospital, family planning centre or other); per capital income; average number of prenatal care visits; and location. Population, proportion of farmers, infant death rate, number of midwives and number of hospital beds were also taken into account. Townships were required to have an existing health facility and the staff necessary to implement the trial.Individuals: women who had given birth in the past year were eligible for the interview.\n\n\nInterventions: Target: health system (health worker education and equipment provision) and community (IEC).Arm 1 (10 clusters, 673 women): the intervention had 3 health system components: training of community midwives, a public awareness campaign with posters and leaflets about prenatal care, and provision of equipment to health centres.Arm 2 (10 clusters, 591 women): usual health system.\n\n\nOutcomes: Trial primary outcomes: prenatal care utilisation and perinatal outcomes.Primary: ANC coverage (at least 4 visits).Secondary: ANC initiation in first trimester, health facility deliveries, stillbirths, perinatal mortality, neonatal mortality.\n\nFollow-up: data were collected from health centre records monthly. Observation in intervention hospitals monthly. Training of midwives involved initial sessions over 2 days and meetings every 3 months. Poster and leaflets in the community throughout trial. Interviews with pregnant women conducted after delivery (mothers of dead infants were not approached for interviews).\n\n\nNotes: Funders: Academy of Finland, Finnish Ministry of Education (DPPH-program), European Commission INCO Programme \"Structural hinders to and promoters of good maternal care in rural China - C HIMACA (015396).Results of a hospital-based survey are not included in this trial report.We have excluded the perinatal mortality data reported for this trial due to multiple risk of bias concerns, including unclear denominators\n\n",
        "objective": "To assess the effects of health system and community interventions for improving coverage of antenatal care and other perinatal health outcomes.",
        "full_paper": "Background\nA community-based randomized control prenatal care trial was performed in a rural county of China during 2000-2003.\nThe purpose of this paper is to describe the trial implementation and the impact of the trial on the utilization of prenatal care and perinatal outcomes.\nMaterials and methods\nIn the study county, 10 townships (from a total of 55) were each paired with a control (20 study townships in total), with the criteria for pairing being the township's socioeconomic development, perinatal health, and maternal care utilization and provision.\nOne of each township pair was randomly allocated to the intervention or control groups.\nThe trial interventions were: 1) training township hospital midwives and instructing them in how to provide systematic maternal care, 2) informing women in the community of the importance of prenatal care, 3) if needed, providing basic medical instruments to the hospitals.\nA variety of data sources were used to describe the trial implementation (observations, group discussions, field notes, survey to women).\nThe data on pregnancy and perinatal outcomes were from the original hand-written work-records in the village family planning centers of the study townships.\nResults\nImplementation of the intervention was deficient.\nThe factors hindering the trial implementation included poor coordination between midwives and family planning officers, broader policy changes implemented by the provincial government during the trial, the decentralization of county governance, and the lack of government funding for maternal care.\nThere was only little difference in the use of maternal care, in women's opinions related to maternal care or content of prenatal care, and no difference in the perinatal outcomes between the intervention and control townships.\nConclusions\nA community based randomized controlled trial could not be fully carried out in rural China as planned due to the changing political landscape, the complexity of the socio-economic situation and a lengthy planning stage.\nThe study could not answer if perinatal outcomes could be improved by increased use of prenatal care.\nTrial registration\nNCT 01054235\nBackground\nPrenatal care has become routine care for pregnant women in the developed world, and increasingly in developing countries too.\nPrenatal care is assumed to improve maternal and neonatal health notably, though this has not been demonstrated conclusively in studies.\nThe wide variation in the design and implementation of prenatal programs has made it difficult to determine the effectiveness of prenatal care.\nAnother major challenge in evaluating prenatal care is adequately controlling confounders and selection bias, with some studies trying to achieve this with the use of quasi-experimental evaluation designs.\nOnly a few studies have adequately controlled for socioeconomic status, which is a composite variable of occupation, education and income.\nThe Chinese government has promoted nationwide prenatal care since 1987.\nThe Ministry of Health issued guidelines entitled \"Perinatal Care Management Approaches for Urban Women\" and \"Systematic Maternal Care Management Approaches for Rural Women\" in 1987 and 1989, respectively.\nSince then, both the provision and utilization of prenatal care have increased, although discrepancies continue both between urban and rural areas and within them.\nEspecially in rural areas, prenatal care varies from place to place in terms of the timing, frequency and the content of prenatal care.\nBy the late 1990s a few rural areas still had not organized systematic prenatal care.\nThis situation offered the possibility to conduct a randomized controlled trial to evaluate the effectiveness of prenatal care.\nWe developed and implemented a community-based randomized prenatal care trial in a rural county in eastern China in 2000-2003.\nThe purposes of the trial were to 1) define the optimal content of prenatal care for the context of rural China; 2) to evaluate the effectiveness of such prenatal care on infant and maternal outcomes, using a community-based, well-designed controlled trial in a rural county;3) to describe the process of conducting a controlled study using community resources.\nThe purpose of this paper is to describe the trial implementation and the impact of this community-based trial on the utilization of prenatal care and perinatal outcomes.\nContext of the trial\nCarried forward by a wave of economic reforms that began in the late 1970s, the health sector in rural China has experienced major changes in recent decades.\nUp until the 1980s, the Cooperative Medical Scheme (CMS) had been a collective-based medical care scheme that ensured basic curative and preventive care for most of the rural population.\nMost of these services were provided free of charge or at a very low cost that was paid by collectives.\nAfter the rural collective economy shifted to a market economy in the 1980s, the so-called \"household responsibility system\" that underpinned the CMS collapsed.\nThat had a major impact on the use and payment of health care, including prenatal care, because farmers had to pay for their health care out of their own pocket.\nWith the economic reforms the responsibility for financing and managing health centers shifted from the county to township authorities.\nWith decentralization, the development of local hospital care plans and the appointment of hospital directors became the responsibility of the township authority.\nHowever, the County Health Bureau still transmitted national guidelines and provided technical support when requested.\nChinese family planning practices and policy have clearly impacted reproductive health care.\nIn many rural areas, local family planning authorities require married women of childbearing age to have a pregnancy test every 2-3 months.\nOnce a woman is found to be pregnant, the family planning authorities determine whether the pregnancy is according to family planning regulations, or whether the woman is persuaded to abort the baby.\nIn most rural areas, if a couple has a girl, they are allowed to have another child after 4 years, otherwise the one-child family planning policy applies.\nA pregnancy test gives an early opportunity to provide maternal care education, but it may also lead to women avoiding health care.\nSome township family planning authorities have required pregnant women to give birth in a designated health facility.\nThe facility was usually the township hospital or township family planning center.\nThe study was made in a rural county in Anhui province in Eastern China.\nThe area of the county is almost 3000 square kilometers, consisting of mostly flatland.\nIn early 2000, the population was some 900 000, mostly farmers.\nThe GDP (gross national product) ranking of the county is typical of less developed rural areas in China.\nThe healthcare system in the study county consisted of county hospitals and decentralized township- and village-level healthcare services.\nSpecialized obstetric services were available only at the county level.\nIn township hospitals, midwives cared for normal births, although typically there were no systematic prenatal care services.\nThe village-level private clinics were not officially involved in the pregnancy or birth care.\nAll three levels of health care in the county functioned on a fee-for-service basis, and most farmers paid for all health services out-of-pocket.\nThe Chinese household registration system (hukou-system) stipulates where public services-including health care and education-can be obtained.\nHukou operates differently for agricultural and non-agricultural people, and is also used to regulate rural people's migration to cities.\nIn brief, free or subsidized services are available only in the community one has his or her hukou; births are registered to the hukou community, regardless of where the parents live.\nEven though there have been several modifications to the system to allow people to have services in the area they live, in the area of our study, the hukou-system resulted in many registered births to women who were no longer living in the area.\nCounty health officials estimated that 30-40% of young women had left their hometowns to work elsewhere.\nHowever, many come back at the time of delivery and stay for a few months thereafter.\nLater, children may be left in the care of the grandparents or they move with their parents.\nMaterials and methods\nThe trial was conducted from August 2000 to March 2003; pregnant woman were enrolled in the study up until July 2002.\nThe trial was approved by the ethics committee of the National Research and Development Centre Welfare and Health (STAKES), Helsinki, Finland in January 11, 1999.\nFor power calculations, perinatal mortality rate (PNMR) was chosen for the indicator.\nAssuming a 30% reduction in PNMR (estimated 91 per 1000 birth, \u03b1 = 0.05, and \u03b2 = 0.1), 1638 births in both groups were needed.\nBased on the estimated birth rate and population size, we needed 20 townships.\nOf the 55 townships in the county, 20 townships were selected that had existing health facilities for the implementation of the trial.\nThe primary criteria for pairing the townships in control and intervention groups were the place of birth (township hospital, township family planning services center and others), fiscal income per capita, average number of prenatal care visits, and the location of the township.\nSecondary criteria applied were township population, proportion of farmers among the population, assigned birth place (township hospital or township family planning services centre), infant death rate, number of midwives, and number of hospital beds.\nThese data were collected from local government documents and by interviewing local county health officials.\nOne township was assigned to the intervention and one to control group in each matched pair by the toss of a coin.\nThe randomization was done by the investigators.\nAfter the randomization, results were compared once more to ensure the paired townships met the criteria for matching.\nIntervention\nBefore the trial, we conducted a pilot study in one experimental and one control township to test the feasibility of the data collection methods.\nIn the pilot data were collected from women at different stages of their pregnancy rather than following one group through pregnancy.\nThe intervention trial was not piloted due to difficulties with the local heath authorities in negotiating for a pilot and time constraints.\nThe intervention consisted of 1) training township midwives, 2) informing women and men in the community of the importance of prenatal care, 3) providing intervention township hospitals with basic medical instruments used in prenatal care (i.e. blood pressure monitors, weighing scales for mothers and newborns, stethoscopes).\nTraining of township midwives\nIn the intervention group (10 townships), the training of midwives was done in four sessions by a researcher (ZW) and two experienced obstetricians from the county's Maternal and Child Health Care Institute.\nThe first training session lasted two days and later sessions a single day.\nThe midwives were lectured on the content and procedures of prenatal care, as given in the regulation issued by the China Ministry of Health (MOH) in 1989.\nThe core instruction material used by the trainers was the \"Maternal Care Management Approaches for Rural Women\", which was issued by the Ministry of Health (MOH) in 1989.\nIt stipulates that the first prenatal care visit should be made within 13 weeks of gestation, and there should be at least five visits during pregnancy and three after delivery.\nIt specifies that prenatal care should include health education, routine check-ups, laboratory tests, measurements (including height and weight), referral of high risk pregnancy to a high level hospital, and it gives instructions for safe delivery.\nDuring the intervention, the midwives were asked by the researchers to attend meetings organized in the county's maternal and child care institute roughly every three months (eight times) to report and discuss the situation in the prenatal care provision and utilization, including the state of progress and any obstacles encountered.\nThe researchers asked the township midwives to give a standard prenatal care card to each pregnant woman at the first prenatal care visit.\nThe card was for recording the progress of the pregnancy, the use of maternal care services and the pregnancy outcome.\nThe woman was asked to carry the card with her until three postnatal home visits were completed.\nIf a woman discontinued her visits during the course of the pregnancy, the midwives were asked to make home visits.\nIn addition, the midwives from both the intervention and control groups were asked to attend four one-day training sessions together on keeping medical records and doing women's interviews in hospitals.\nInforming women and men in the community\nThe County Health Bureau officials contacted the township's family planning officers and asked them to supply a leaflet on maternal care, entitled \"letter to the mother-to-be\", to pregnant woman during regular pregnancy testing in the villages.\nThe leaflet was written by the researchers and signed by the County Health Bureau, the Family Planning Committee and the Maternal and Child Health Care Institute.\nThe leaflet advised the pregnant woman to seek prenatal care and to have their delivery in the township hospital.\nIt said that the woman can receive further information about maternity and infant care by consulting midwives in the township hospital.\nThe township midwives facilitated and monitored the distribution of the leaflet.\nSecondly, the field research assistant hung posters on the walls of the township hospital, the family planning center, and the village administration offices.\nThe posters were obtained from the provincial Maternal and Child Health Care Institute and consisted of a set of educational posters (12 in total) on daily life during pregnancy (relating to food, nutrition, sleep, work, activities, avoiding poisons etc.) and when and where to have prenatal care and delivery, with emphasis on the importance of a hospital delivery.\nTrial monitoring\nTo monitor the trial regularly and to collect data for ongoing evaluations, the researchers hired a medical graduate from the county to work as a field research assistant.\nShe visited every intervention township hospital once a month, monitored the progress of the intervention implementation and gave feedback to the researchers.\nThe researchers made monthly visits to the county to discuss with the County Health Bureau officials and field research assistant any obstacles hindering the trial implementation and to find timely solutions.\nData collection\nSeveral kinds of data covering the processes and outcomes were collected.\nQualitative data included researchers' observations, interviews, group discussions, and field visits by the field assistant and researchers (ZW, YW, KV).\nQuantitative data included a community-based survey of mothers, a survey conducted by midwives at township hospitals of women giving birth, and routine pregnancy and birth records in township family planning centers.\nObservations, interviews and discussions\nThe field research assistant visited the intervention township hospital and their midwives at least once a month.\nShe collected progress data, identified problems and barriers to the intervention, and reported back to the Deputy Director of the County Health Bureau's and also to a researcher (ZW).\nA researcher (ZW) visited the study county at least every other month, had meetings with the Director and Deputy Director of the Health Bureau and with the field research assistant to discuss progress and to improve the implementation of the intervention.\nThe trial progress was also discussed at routine visits made by the Directors of the County Health Bureau to the intervention townships.\nThe minutes of the meetings and discussions were kept by a researcher (ZW).\nThe field research assistant and a researcher (ZW) documented the eight workshops attended by midwives, the township hospital manager and county health officials.\nThe Director and Deputy-Director of the County Health Bureau were (unsystematically) interviewed several times by a researcher (ZW) in regard to the obstacles to and promoters of systematic maternal care.\nTwo researchers (ZW, YW) also interviewed other health administrators, including the Deputy-Director of the county's Mother and Child Health Institute, four directors of the township hospitals (two from the intervention group, two from the control group) and five township midwives (three from the intervention group, two from the control group), again addressing the obstacles to and promoters of maternal care.\nWe also attempted to interview family planning officials, though without success.\nEach interview lasted between 30 minutes and an hour.\nNotes were taken during the interview.\nHospital based survey\nMidwives in the intervention and control townships and in the county level hospitals collected data on women's care, pregnancy and birth, and on the child born, either by asking the mother or extracting from her own medical notes, using a structured data collection form.\nThe information also included the women's home town, thus allowing data to be linked to the study groups.\nThe questionnaire included 49 questions covering parents' background, the mother's reproductive history, infant outcomes, prenatal care and infant health.\nThese data were not used in this paper.\nCommunity-based mothers' survey\nA survey of mothers was carried out between October 2001 and December 2003.14\nThis survey covered mothers who had been pregnant during the intervention period and gave birth between April 1, 2001 and March 31, 2003 in the 10 intervention and 10 control townships.\nEach township in the study county was administratively divided into 6-15 villages.\nOf the 20 townships, 10 percent were randomly selected to be surveyed in October-December in 2001, with 20 and 30 percent of townships selected in 2002 and 2003, respectively.\nAll mothers who had given birth in the 12 months prior to the survey in these villages were approached for an interview.\nDuring that time, 1306 mothers were eligible for the study according to the records kept by the local family planning center.\nMothers with dead infants were not approached for interviews.\nInterviewers were recruited from among local health workers (not midwives or family planning workers).\nThey were trained to conduct interviews by a researcher (ZW).\nThe structured questionnaire used in the survey included 60 questions covering infant outcomes, infant health and women's knowledge, attitudes, and practices relevant to prenatal, delivery, and postnatal care.\nThe interview was conducted at the mothers' homes.\nIf the mother was not at home at the time of the survey, the father or some other family member responded on her behalf.\nAltogether, survey data was collected for 1264 of the 1306 mothers, with 90% of survey respondents being the mother.\nData on 42 mothers (3%) were missing: 2 (0.15%) refused, 27 (2.1%) were out of the village; and 13 (1.0%) cases of missing data were for other reasons.\nThere were no differences in the demographic characteristics of the respondents in the intervention and control groups (Table 1).\nMost (86%) mothers were farmers, and 96% were aged 20-34.\nA clear majority (63%) were first time mothers, 22% had two children and 14% had three or more children.\nThe intervention group had a larger proportion of higher parity than in the control group.\nFamily planning (FP) routine pregnancy and birth records\nThe field research assistant abstracted data on pregnancies and their outcomes from the original hand-written work records of the village family planning centers in the 20 study townships.\nThe centers record data on all women who had a positive pregnancy test result.\nPregnancy outcome (abortion, including miscarriage, or birth) is also recorded.\nDue to the hukou-system (regulated residence system) women living and giving birth elsewhere than in the home-community are still marked in the home-community's village records.\nThe records contained data on women's background (birth date, township of residence, and number of living children), pregnancy-related variables (time of last menstrual period, abortion), and infant variables (birth date, gender, plurality, birth order, live birth, stillbirth and early neonatal death).\nFor non-resident women living in the townships, pregnancy data were not usually recorded (information given by local record keepers).\nEach village family planning center brought their month's records to the administrative family planning office for the purpose of statistics compilation.\nThe outcomes of the pregnancies beginning between August 1, 1999 and July 31, 2002 in the 20 study townships were extracted from the family planning records.\nData analysis\nAll interview data and observational notes were reviewed and then analyzed using the framework method.\nOnly the main conclusions from the qualitative data analyses are presented in this paper.\nQuantitative data were entered into the computer using EPI-INFO (version 6.04) software and analyzed using SPSS 8.0 for Windows.\nSince randomization was by community, we adjusted for cluster randomization in the statistical analyses.\nDifferences between the experimental and control groups were tested for with adjusted chi-square tests.\nThe results describing the trial implementation were studied separately for the first and second trial year to see whether the intervention intensified or contamination increased over time.\nThe health outcomes did not differ between the two years, and so combined results are given.\nResults\nResults on trial implementation\nLocal government involvement\nThis section is based on the researcher's (ZW) and the field research assistant's observations, field notes and surveys of the government officials.\nThe County Health Bureau officials helped to organize the training of township midwives and also arranged eight meetings between township midwives and the researchers.\nThe County Health Bureau officials tried to co-ordinate the relationship between the township hospital (midwives) and the township Family Planning Centers by asking the FP officers at the centers to help supply the information leaflet to mothers-to-be during regular pregnancy testing.\nThis, however, did not work very well because the township Family Planning Center was under the leadership of the County Family Planning Committee rather than the County Health Bureau.\nAbout eight months into the trial (March 2001), the Department of Maternal and Child Health Care at the Provincial Health Bureau decided to carry out a systematic program of maternal care in the whole Anhui province, and assigned this task to the County Health Bureaus.\nTo avoid contamination between the intervention and control groups, we (ZW) tried to persuade the County Health bureau to begin this overlapping initiative in the control townships later than in the other townships.\nAlthough the County Health Bureau officials understood the researcher's concern, the maternal care program started only a few months later than in the non-study townships.\nThe managers and midwives of all the study township hospitals were assembled by the County Health Bureau in May 2001 and informed of the provincial government's decision to carry out a province-wide systematic reform of maternal care in May 2001.\nTwo townships in the intervention group and four in the control group had a prepayment maternal care scheme implemented during our trial, introduced already in the 1990s: every pregnant woman paid 200 RMB (about 24 dollars) in advance to the township hospital, which covered prenatal care, hospital delivery and postnatal care.\nThe scheme promoted the use of maternity care also in the four control townships mentioned above, and thus diluted the trial effects.\nTraining of midwives and their data collection\nAnalysis of the field notes (ZW) and the hospital-based survey revealed that the training sessions for the midwives in the intervention and control townships were completed as planned.\nAbout 95% of available midwives were present in each training session.\nThe 5% of midwives who did not attend were absent because of illness or were not able to leave their work station; they received training on a later day.\nInforming women in the community\nAccording to the midwives' daily work records, 74% of women who delivered in the hospital in the intervention townships and 71% of the women who delivered in control townships had a prenatal care card.\nThe distribution of the information leaflet to the pregnant women was not particularly successful (Table 2).\nAccording to the discussions between researchers and midwives, in some townships, the FP officers did not welcome the involvement of hospital midwives, thinking that such involvement was beyond the scope of their work and would increase the workload of the FP officers.\nFor midwives, it was also difficult to set their hospital work aside to go with FP officers to visit women in the villages.\nAnother important reason for the failure of this measure was that the midwives lacked motivation since they did not receive subsidies for their extra work.\nHanging posters (12) relating to prenatal care on the walls of the intervention township hospitals and family planning service centers was carried out as planned.\nIn addition, the field research assistant gave the posters to township midwives to hang on the walls in the village administration office or in the village doctor's office.\nBased on the researcher assistant's observations, about 70% of villages had the posters in place.\nAccording to the community-based survey, about 40% of women in the intervention group saw the study leaflet and poster, more so in the second year of the trial.\nSome women in the control group had also seen the leaflet and posters, but less often (Table 2).\nThe proportion of women who received information related to prenatal care from township midwives was higher in the intervention group (70%) than that in control group (49%) in the first year of the trial, though not in the second year (Table 2).\nThe proportion of women who received information related to prenatal care from FP officers did not differ between the two groups.\nOutcome results\nMaternal care utilization\nThe community-based survey revealed that differences between the intervention and control groups in the timing of the first prenatal care visit and the frequency of prenatal care visits were small (Table 3).\nIn the intervention group 93% of women gave birth in public hospitals, which was a little higher than in the control group (88%), but the difference was not statistically significant.\nThere was no statistically significant difference in the cesarean section rate between the intervention and control groups (Table 3).\nMost women (more than 97%) were satisfied with maternal care provided by township midwives, with no difference between the two groups.\nThere was no difference between the intervention and control groups in the variables measuring women's' knowledge or attitudes towards pregnancy care.\n91% of women in the intervention group and 93% in the control group thought that prenatal visits are important and 97% in both groups thought that hospital delivery is important.\nContent of prenatal care\nThe community-based survey revealed that only about half of the women in either the intervention or control group had their blood pressure measured at each visit.\nMore women in the control group did not have their blood pressure measured, but the difference between the two groups was not statistically significant.\nAbout 43% of the women in both groups had blood tests.\nThere were more women in the control group who had more than one ultrasound scan, but the difference between the two groups was not statistically significant (Table 4).\nWomen would have liked more health promotion information during pregnancy (Table 4).\nBased on women's reports, the midwives in the intervention group had given somewhat more information on health promotion (Table 4).\nHowever, in cases where the information was given, the women considered it equally good in the two groups.\nImpact of the trial on the pregnancy and perinatal outcomes\nThere were 1830 recorded pregnancies in the intervention and 1718 in the control group before the intervention period (from August 1, 1999 to July 31, 2000), and 2580 pregnancy records in the intervention and 2550 in the control group during the trial (Table5).\nIn addition, there were 669 women in the intervention and 790 women in the control group who had given birth between January 1, 2001 and December 31, 2002, but with no record of a pregnancy test; these were very likely women who lived elsewhere, but whose deliveries were recorded because of the hukou-system.\nThese women were not included in the analysis\nFamily planning records showed that the abortion rate, sex ratio at birth and early neonatal death rate (ENDR) both in the intervention and control groups were somewhat lower after the intervention than before it.\nThere were no statistically significant differences between the intervention and control groups in the abortion rate, sex ratio at birth, still birth rate, early neonatal death rate or perinatal mortality (Table 5).\nWhen we studied the early neonatal mortality (ENDR) by gender, we found that girls' neonatal death rate was much higher than that of boys both in the intervention and control groups.\nComparing the time before and after the trial in the intervention group, boys' ENDR dropped from 2.9% to 2.0%, while girls' ENDR increased from 5.2% to 5.7%.\nIn the control group, boys' ENDR increased from 1.3% to 1.8%, while girls' ENDR dropped from 6.1% to 3.6%\nDiscussion\nThis is the first community-based randomized controlled prenatal care trial carried out in rural China.\nThe trial was not fully implemented as planned.\nAlthough the training of midwives did occur as planned and some local government support was obtained, the distribution of information on the importance of prenatal care was not particularly successful, partly due to poor cooperation between the township hospital midwives and the family planning officers.\nIn respect of the similar results for health indicators and prenatal care obtained in the intervention and control groups, the likely cause was contamination in the controls and weak implementation of the intervention.\nThere was no impact on still-birth and neonatal death rates or perinatal mortality.\nA strength of this study was that we conducted both a process and outcome evaluation concurrently.\nThe main difficulties centered on the rapidly evolving situation in maternity care and also contamination in the control groups.\nIn the late 1980s, the Ministry of Health issued a regulation requiring local authorities to carry out systematic prenatal care.\nThe regulation was not mandatory and no funding was provided for it.\nThus, the provision and use of maternal care varied from place to place.\nIt was up to county or township authorities to decide whether to carry out a systematic prenatal care program.\nBased on the information of the mid-1990s (at the time when the trial was being planned) we assumed the use of prenatal care to be very low in the study county and township hospitals were under the leadership of county health bureau (including financing and managing township hospitals).\nHowever, as shown by our survey, by early 2000 most women in the control townships had already been making several prenatal visits and many were starting care early.\nAnd the responsibility for financing and managing township hospitals had shifted from the county to township authorities in late 1990s.\nThis is an example of the very rapidly changing health care in China, reflecting the rapid changes in Chinese society as a whole.\nThe situation should have been mapped more thoroughly and at a time closer to the point of intervention.\nThe use of prenatal care in the control townships may also be a result of contamination.\nIn the middle of the trial the provincial authorities required the County Health Bureau to carry out county-wide systematic maternal care.\nThat may have resulted in information materials on prenatal care being distributed in the control townships, leading to increased use of prenatal care.\nThis may explain why 56% of mothers in the control group had received advice or information on maternal care from township midwives and 26% from family planning officers.\nIt may also explain why there was a bigger difference between the experimental and control groups in the proportion of women who had received the material in the first year than in the second year.\nAnother source of contamination may have been contacts between the townships.\nSince midwives from different township hospitals meet regularly, the midwives from the intervention and control townships may have influenced each other.\nFurthermore, in the countryside, people often visit relatives and friends in other townships.\nSome control women may have seen the health education posters while visiting and they may have discussed prenatal care and delivery with relatives or friends they had in intervention townships.\nBasic criteria for selecting study townships were that they must have an existing health facility and available hospital staff to support the trial implementation.\nTherefore, there was little difficulty for township hospitals to carry out prenatal care and delivery services.\nOn the other hand, the criteria may have worked to select the best townships, which of itself could have yielded an increased proportion of women using prenatal care, irrespective of our intervention.\nEven though the county officials were supportive, they did not have full control over the townships.\nHealth care administration was partly decentralized from the county to the township level, with township hospitals partially funded by township authorities.\nThe County Health Bureau supervised the townships and their professionals, but they could not appoint staff or regulate their work assignments.\nIn the study county, the family planning (FP) sector has more public resources than the health sector.\nThe township's FP officers were under the leadership of the County Family Planning Committee and township authorities.\nSo the County Health Bureau officials were not able to coordinate the work of the FP officers with the township midwives.\nThis could explain why one of our intervention procedures, the dissemination of leaflets on maternal care to future mothers during pregnancy testing, was not successfully implemented.\nWe were not given permission to discuss the matter with family planning officials at the county or higher level so as to remedy the situation.\nIn addition, the FP officers were not paid for supplying the leaflets and it was not part of their ordinary work.\nSo they lacked enthusiasm to cooperate with the health sector.\nFurthermore, there were only one or two midwives engaged in the trial work in each township hospital, and they were often too busy to leave the hospital, which decreased their participation in the information dissemination.\nUnfortunately, we did not collect data on how many leaflets were actually supplied to pregnant women.\nTo profit from fees-for-services, township hospitals have to be of high quality and acceptable to the patients.\nThe midwife training was consistent with the interests of hospitals to remain competitive and it was not unexpected that the training sessions were carried out smoothly.\nSome of our indicators turned out to be unsuitable for measuring the impacts of the trial intervention.\nOur study has shown that the early neonatal death rate is much higher among girls than boys, both before and after the trial.\nIf the impact of the family planning policy is larger on perinatal mortality than maternal care, then it is hard for any health care intervention to have an effect on perinatal health outcomes.\nTypically, in large-scale cluster randomization trials that evaluate public health interventions at community level, the expected effects on morbidity and mortality tend to be modest, although still of public health relevance.\nA second unsuitable indicator was the hospital delivery rate.\nIt was already high, and in such a situation it is hard to distinguish a difference between study groups.\nWe could not find any other studies that were sufficiently similar to allow a comparison with our trial.\nIf such was the case, it would have been useful to compare the factors affecting success or failure in introducing a community-based intervention aimed at increasing the utilization of prenatal care.\nConclusions\nOur study showed that a community based randomized controlled trial could not be fully carried out in rural China as planned due to the changing political landscape, the complexity of the socio-economic situation and a lengthy planning stage, The trial was adversely affected by policy changes at a higher level of government, the decentralization of the decision-making in health care, and the poor cooperation between family planning and health sectors at township level.\nThis kind of cooperation is crucial to the success any intervention in maternity care in rural China, given that the family planning sector is better financed than the health sector in townships.\nThis trial could be well planned but was hard to implement.\nThus, it was not able to answer if prenatal care is able to improve perinatal outcomes.\n\nMothers' background by group (1)\n | Intervention | Control | Total | P (2)\n |  |  | % (n) | \n(n) | (n = 673)% | (n = 591)% | (n = 1264)% | \nAge |  |  |  | \n< 20 | 0.4 | 0.7 | 0.6 | 0.295\n20- | 73.4 | 76.1 | 74.7 | \n30- | 20.8 | 15.7 | 18.4 | \n> = 35 | 2.4 | 4.6 | 3.4 | \nMissing | 3.0 | 2.9 | 2.9 | \nOccupation |  |  |  | \n\u2003Farmer | 80.1 | 86.8 | 83.2 | 0.056\n\u2003Other | 15.2 | 11.8 | 13.6 | \n\u2003Missing | 4.8 | 1.4 | 3.2 | \nParity |  |  |  | \n\u20031 | 60.5 | 66.7 | 63.4 | 0.031\n\u20032 | 24.2 | 20.0 | 22.2 | \n\u20033+ | 15.3 | 13.4 | 14.4 | \n\u2003Missing | 0.0 | 0.0 | 0.0 | \n\n(1) Data from community-based survey\n(2) P value refers to the difference between the intervention and control group\n\nProportions of women receiving information of prenatal care, by group and intervention year (%) (1)\n | Intervention | Control | \n(n, first year) | (n = 343) | (n = 350) | \n(n, second year) | (n = 330) | (n = 241) | P (2)\n(n, both years) | (n = 673) | (n = 591) | \nReceived the study leaflet |  |  | \n\u2003First year | 37.0 | 20.0 | 0.005\n\u2003Second year | 48.2 | 44.4 | 0.748\n\u2003Both years | 42.5 | 29.9 | 0.090\nSeen the study poster |  |  | \n\u2003First year | 27.7 | 12.6 | 0.016\n\u2003Second year | 50.6 | 34.4 | 0.142\n\u2003Both years | 38.9 | 21.5 | 0.014\nReceived the advice on maternal care from township midwives |  |  | \n\u2003First year | 70.0 | 49.1 | 0.031\n\u2003Second year | 63.9 | 64.7 | 0.942\n\u2003Both years | 67.0 | 55.5 | 0.196\nReceived the advices on maternal care from FP worker |  |  | \n\u2003First year | 16.0 | 17.4 | 0.797\n\u2003Second year | 33.3 | 37.8 | 0.698\n\u2003Both years | 24.5 | 25.7 | 0.858\n\n(1) Data from the community base survey\n(2) P value refers to the difference between the intervention and control group\n\nTiming and frequency of prenatal visit and delivery care by group (%) (1)\n | Intervention (n = 673) | Control (n = 591) | P(2)\n1st prenatal visit < = 13 weeks of gestation | 40.0 | 44.5 | 0.524\nNo prenatal visit | 3.4 | 3.9 | 0.758\nPrenatal visit > = 5 times | 55.9 | 42.8 | 0.246\nHospital delivery | 92.9 | 87.5 | 0.231\nC-section | 6.5 | 8.5 | 0.313\n\n(1) Data from the community base survey\n(2) P value refers to the difference between the intervention and control group\n\nContent of prenatal care by group (%)\n |  | Intervention (n = 673) | Control (n = 591) | P value\nBlood pressure taken |  |  |  | \n | At every visit | 49.6 | 51.3 | 0.997\n | sometimes | 38.6 | 31.1 | \n | Not at all | 8.2 | 12.5 | \nBlood test |  |  |  | \n | once | 34.2 | 28.9 | 0.738\n | More than once | 7.0 | 12.5 | \n | no | 55.0 | 53.3 | \nUltrasound test |  |  |  | \n | 1 time | 42.8 | 30.5 | 0.803\n | 2 time | 13.2 | 19.0 | \n | 3 and plus | 3.3 | 5.3 | \n | No | 37.0 | 40.3 | \nInformation on prenatal care |  |  |  | \n | Too simple | 8.2 | 11.5 | 0.065\n | Too difficult | 2.7 | 1.9 | \n | Just OK | 52.8 | 47.9 | \n | Not sure | 20.2 | 14.9 | \n | Did not get information | 11.9 | 18.1 | \nInformation on nutrition |  |  |  | \n | Needed more | 47.1 | 41.0 | 0.143\n | Needed less | 0.2 | 1.0 | \n | Just ok | 27.3 | 27.1 | \n | Not sure | 4.9 | 2.9 | \n | Did not get information | 16.3 | 23.2 | \nInformation on baby care |  |  |  | \n | Needed more | 50.1 | 45.4 | 0.041\n | Needed less | 0.2 | 1.2 | \n | Just ok | 24.2 | 17.6 | \n | Not sure | 3.3 | 2.7 | \n | Did not get information | 18.1 | 28.1 | \nInformation on disease prevention |  |  |  | \n | Needed more | 49.0 | 47.0 | 0.228\n | Needed less | 0.2 | 0.3 | \n | Just ok | 30.3 | 28.4 | \n | Not sure | 4.6 | 3.1 | \n | Did not get information | 11.7 | 15.7 | \n\n1) Data from the community base survey. Women with no information (varying from 3.7% to 5.8%) are not shown.\n2) P value refers to the difference between the intervention and control group\n\nPregnancy and perinatal outcomes by group (1)\n | Intervention | Control | P (2)\nNumber of pregnancies |  |  | \n\u2003before intervention | 1830 | 1718 | \n\u2003after intervention | 2580 | 2550 | \nAbortion rate ( %) |  |  | \n\u2003before intervention | 19.1 | 19.2 | 0.847\n\u2003after intervention | 17.1 | 16.8 | 0.912\nnumber of live births |  |  | \n\u2003before intervention | 1423 | 1362 | \n\u2003after intervention | 2094 | 2062 | \nSex ratio at birth |  |  | \n\u2003before intervention | 153.0 | 163.5 | 0.454\n\u2003after intervention | 144.5 | 152.8 | 0.608\nStillbirth rate ( \u2030 ) |  |  | \n\u2003before intervention | 15.5 | 8.8 | 0.153\n\u2003after intervention | 7.6 | 14.1 | 0.120\nNeonatal mortality ( \u2030 ) |  |  | \n\u2003before intervention | 37.9 | 30.8 | 0.694\n\u2003after intervention | 37.7 | 26.2 | 0.368\nPerinatal mortality, \u2030 |  |  | \n\u2003before intervention | 52.6 | 39.3 | 0.343\n\u2003after intervention | 42.5 | 39.3 | 0.429\n\n1) Data from township family planning records\n2) P values are the results of cluster adjusted ordered logistic regression",
        "label": "unclear",
        "id": "task4_RLD_test_513"
    }
]