electricsheepafrica/contraceptive-fp-commodity

Contraceptive & Family Planning Commodity Dataset

Abstract

This dataset provides 30,000 simulated facility-level observations (10,000 per scenario) of contraceptive and family planning commodity availability across three tiers of FP service delivery in sub-Saharan Africa. Each record represents one FP commodity assessed at one facility during one quarterly period. The dataset captures 35+ variables including method type, availability, stockout metrics, method choice breadth, LARC access barriers, client impact, and supply chain performance. Three scenarios: urban FP clinic (91% availability), district hospital (65%), rural health centre (32%).

This dataset is entirely simulated. It must not be used for clinical or programme decisions.

1. Introduction

1.1 Contraceptive Supply in Sub-Saharan Africa

Unmet need for family planning in SSA remains among the highest globally, with an estimated 60 million women of reproductive age wanting to avoid pregnancy but not using modern contraception. Supply-side barriers — particularly commodity stockouts and limited method choice — are major contributors to unmet need.

1.2 Method Mix and LARC Access

The CHAI/RHSC Family Planning Market Report (2024) documented a 15% increase in LARC market volumes from 2022 to 2023, driven by implant scale-up. However, LARC access remains constrained at peripheral facilities where trained providers and insertion equipment are unavailable (PMC, 2014; PMC4168608). Rural health centres frequently offer only short-acting methods (injectables, pills, condoms), limiting informed method choice.

1.3 Commodity Stockouts

Frequent commodity stockouts hinder public-sector provision of family planning services. The UNFPA Supplies Partnership supports 54 countries with FP commodities, but stockouts persist due to quantification errors, distribution delays, and funding gaps (UNFPA, 2023). DMPA injectable — the most widely used method in SSA — faces periodic stockouts despite high demand.

1.4 Rationale

This dataset integrates FP commodity availability, method choice breadth, LARC access barriers, and client impact indicators for supply chain modelling and reproductive health equity research.

2. Methodology

2.1 Parameterization

Parameter	Urban FP Clinic	District Hospital	Rural HC	Source
Short-acting availability	88%	70%	45%	UNFPA Supplies 2023
LARC availability	72%	45%	10%	PMC4168608
Stocked out in 6m	23%	47%	73%	Multi-country data
Method choice offered	85%	55%	25%	FP programme data
Order fill rate	82%	58%	30%	UNFPA data

2.2 FP Commodity Selection

12 commodities: DMPA-IM, DMPA-SC (Sayana Press), COC pill, POP pill, emergency contraceptive, male condom, female condom, levonorgestrel implant, etonogestrel implant, copper IUD, hormonal IUD, CycleBeads/SDM.

2.3 Scenario Design

Scenario A — Urban FP Clinic: Trained provider, implant and IUD insertion capability, electronic LMIS, UNFPA-supported, wide method choice.

Scenario B — District Hospital: Trained provider, implant insertion but no IUD, paper LMIS, UNFPA-supported.

Scenario C — Rural Health Centre: No trained FP provider, no LARC insertion capability, no LMIS, short-acting methods only.

3. Schema

Column	Type	Description
facility_level	categorical	Facility tier
commodity_name	categorical	12 FP commodities
method_type	categorical	injectable / oral / barrier / implant / IUD / natural
duration_category	categorical	short_acting / LARC / emergency / natural
available_on_survey_day	binary	Commodity available
stocked_out_in_last_6m	binary	Any stockout in 6 months
stockout_days_last_6m	int	Total stockout days
stockout_cause	categorical	11 root cause categories
method_choice_offered	float	Proportion of methods available
new_acceptors	int	New FP acceptors this period
continuing_users	int	Continuing users served
clients_turned_away	int	Clients unable to receive method
method_switched_due_to_stockout	binary	Forced method switch
unmet_need_contribution	binary	Stockout contributed to unmet need
months_of_stock	float	Current stock in months
order_fill_rate_pct	float	Order fulfilment rate
can_insert_implant	binary	Facility can provide implants
can_insert_iud	binary	Facility can provide IUDs

4. Validation

5. Usage

from datasets import load_dataset

dataset = load_dataset(
    "electricsheepafrica/contraceptive-fp-commodity",
    "district_hospital_fp"
)
df = dataset["train"].to_pandas()

# LARC vs short-acting availability
print(df.groupby('duration_category')['available_on_survey_day'].mean())

6. Limitations

• Simulated: Not from real FP programme data.
• No demand-side factors: Cultural, religious, and partner-related barriers not modelled.
• No discontinuation: Method switching due to side effects not captured.

7. References

1. CHAI/RHSC (2024). Family Planning Market Report. LARC volumes up 15%.
2. UNFPA (2023). Supplies Partnership Performance Report. 54 countries.
3. PMC (2014). Scaling up delivery of contraceptive implants in SSA. PMC4168608.
4. UNFPA Kenya. Contraceptive implants delivered to MOH.

Citation

@dataset{esa_fp_commodity_2025,
  title   = {Contraceptive and Family Planning Commodity Dataset},
  author  = {{Electric Sheep Africa}},
  year    = {2025},
  publisher = {Hugging Face},
  url     = {https://huggingface.co/datasets/electricsheepafrica/contraceptive-fp-commodity},
  note    = {Simulated dataset. Not for clinical or programme decisions.}
}

License

CC-BY-4.0