File size: 10,564 Bytes
4d8a13f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
#!/usr/bin/env python3
"""
Literature-Informed Oral Health & Dental Disease Dataset
=========================================================

Generates realistic synthetic records of oral health patients in
sub-Saharan Africa, including dental caries, periodontal disease,
noma, oral cancer, treatment access, and outcomes.

References (web-searched):
-----------
[1]  WHO Africa 2024. Africa has largest global increase in
     oral diseases. 6 major conditions: caries, periodontal,
     oral cancer, oral HIV, noma, cleft lip/palate.
[2]  WHO 2022. Global Oral Health Status Report. Untreated
     caries is most prevalent condition globally.
[3]  PubMed 2021. DMFT in East Africa: 2.57 at age 12,
     4.04 at age 15. High caries burden.
[4]  BMC Public Health 2021. Dental caries in adults SSA.
     Limited access to dental care.
[5]  WHO Africa. Noma (cancrum oris) persists in extreme
     poverty. CFR 70-90% untreated. Disfiguring.
[6]  PubMed 2015. Oral health South Africa: DMFT trends,
     national surveys, fluoride.
[7]  Dentist ratio in SSA: <1 per 100,000 population in
     many countries. WHO target 1:7500.
"""

import numpy as np
import pandas as pd
import argparse
import os

SCENARIOS = {
    'dental_clinic': {
        'description': 'Urban dental clinic with dentist, basic '
                       'restorative/extraction capability, X-ray '
                       '(e.g., university dental clinics Nairobi, '
                       'Lagos, Addis Ababa)',
        'dentist_available': True,
        'restorative_available': True,
        'xray_available': True,
        'fluoride_programme': True,
        'treatment_mod': 1.0,
    },
    'district_hospital': {
        'description': 'District hospital with dental officer, '
                       'extraction only, no restorative '
                       '(e.g., district hospitals Tanzania, Malawi)',
        'dentist_available': True,
        'restorative_available': False,
        'xray_available': False,
        'fluoride_programme': False,
        'treatment_mod': 0.6,
    },
    'rural_health_centre': {
        'description': 'Rural health centre, no dental professional, '
                       'basic pain relief, referral '
                       '(e.g., rural CHCs DRC, Niger, Chad)',
        'dentist_available': False,
        'restorative_available': False,
        'xray_available': False,
        'fluoride_programme': False,
        'treatment_mod': 0.2,
    },
}


def generate_dataset(n=10000, seed=42, scenario='district_hospital'):
    rng = np.random.default_rng(seed)
    sc = SCENARIOS[scenario]

    records = []

    for idx in range(n):
        rec = {'id': idx + 1}

        # ── 1. Demographics ──
        rec['age_years'] = max(2, min(80, int(rng.normal(28, 18))))
        rec['sex'] = rng.choice(['M', 'F'], p=[0.48, 0.52])
        rec['child'] = 1 if rec['age_years'] < 18 else 0
        rec['education'] = rng.choice(
            ['none', 'primary', 'secondary', 'tertiary'],
            p=[0.20, 0.35, 0.35, 0.10])
        rec['urban'] = 1 if rng.random() < 0.40 else 0

        # ── 2. Risk factors ──
        rec['tobacco_use'] = 0
        if rec['age_years'] >= 15:
            rec['tobacco_use'] = 1 if rng.random() < 0.12 else 0
        rec['sugary_diet'] = 1 if rng.random() < 0.55 else 0
        rec['fluoride_toothpaste'] = 1 if rng.random() < (0.50 if rec['urban'] else 0.20) else 0
        rec['brushing_frequency'] = rng.choice(
            ['never', 'occasional', 'once_daily', 'twice_daily'],
            p=[0.15, 0.25, 0.40, 0.20])
        rec['hiv_positive'] = 1 if rng.random() < 0.06 else 0
        rec['diabetes'] = 0
        if rec['age_years'] >= 30:
            rec['diabetes'] = 1 if rng.random() < 0.08 else 0
        rec['malnutrition'] = 0
        if rec['child']:
            rec['malnutrition'] = 1 if rng.random() < 0.15 else 0

        # ── 3. Dental caries [2][3] ──
        caries_prob = 0.40
        if rec['sugary_diet']:
            caries_prob += 0.15
        if rec['brushing_frequency'] in ('never', 'occasional'):
            caries_prob += 0.10
        if not rec['fluoride_toothpaste']:
            caries_prob += 0.05
        rec['dental_caries'] = 1 if rng.random() < min(caries_prob, 0.80) else 0

        rec['dmft_score'] = 0
        if rec['dental_caries']:
            if rec['child']:
                rec['dmft_score'] = max(0, min(20, int(rng.exponential(3))))
            else:
                rec['dmft_score'] = max(0, min(32, int(rng.exponential(5))))

        rec['untreated_caries'] = 0
        if rec['dental_caries']:
            rec['untreated_caries'] = 1 if rng.random() < 0.80 else 0

        # ── 4. Periodontal disease ──
        rec['periodontal_disease'] = 0
        if rec['age_years'] >= 15:
            perio_prob = 0.20
            if rec['tobacco_use']:
                perio_prob *= 1.5
            if rec['diabetes']:
                perio_prob *= 1.5
            if rec['hiv_positive']:
                perio_prob *= 1.3
            rec['periodontal_disease'] = 1 if rng.random() < min(perio_prob, 0.60) else 0

        rec['periodontal_severity'] = 'none'
        if rec['periodontal_disease']:
            rec['periodontal_severity'] = rng.choice(
                ['mild', 'moderate', 'severe'],
                p=[0.30, 0.45, 0.25])

        rec['tooth_loss'] = 0
        if rec['periodontal_severity'] == 'severe' or rec['dmft_score'] > 8:
            rec['tooth_loss'] = max(0, min(20, int(rng.exponential(3))))

        # ── 5. Other conditions [1][5] ──
        rec['oral_cancer'] = 0
        if rec['age_years'] >= 40:
            oc_prob = 0.005
            if rec['tobacco_use']:
                oc_prob *= 3
            rec['oral_cancer'] = 1 if rng.random() < oc_prob else 0

        rec['noma'] = 0
        if rec['child'] and rec['malnutrition']:
            rec['noma'] = 1 if rng.random() < 0.005 else 0

        rec['oral_hiv_manifestation'] = 0
        if rec['hiv_positive']:
            rec['oral_hiv_manifestation'] = 1 if rng.random() < 0.30 else 0

        rec['cleft_lip_palate'] = 0
        if rec['age_years'] < 10:
            rec['cleft_lip_palate'] = 1 if rng.random() < 0.002 else 0

        rec['dental_trauma'] = 0
        if rec['age_years'] < 18:
            rec['dental_trauma'] = 1 if rng.random() < 0.05 else 0

        rec['dental_abscess'] = 0
        if rec['untreated_caries']:
            rec['dental_abscess'] = 1 if rng.random() < 0.10 else 0

        rec['dental_pain'] = 0
        if rec['dental_caries'] or rec['periodontal_disease'] or rec['dental_abscess']:
            rec['dental_pain'] = 1 if rng.random() < 0.60 else 0

        # ── 6. Treatment access [4][7] ──
        rec['sought_dental_care'] = 0
        if rec['dental_pain'] or rec['dental_abscess']:
            rec['sought_dental_care'] = 1 if rng.random() < (0.50 * sc['treatment_mod'] + 0.10) else 0

        rec['treatment_received'] = 'none'
        if rec['sought_dental_care']:
            if sc['restorative_available']:
                rec['treatment_received'] = rng.choice(
                    ['extraction', 'filling', 'scaling', 'antibiotics', 'pain_relief'],
                    p=[0.35, 0.25, 0.10, 0.15, 0.15])
            elif sc['dentist_available']:
                rec['treatment_received'] = rng.choice(
                    ['extraction', 'antibiotics', 'pain_relief'],
                    p=[0.50, 0.25, 0.25])
            else:
                rec['treatment_received'] = rng.choice(
                    ['pain_relief', 'traditional_remedy', 'referral'],
                    p=[0.40, 0.35, 0.25])

        rec['barrier_to_care'] = 'none'
        if not rec['sought_dental_care'] and rec['dental_pain']:
            rec['barrier_to_care'] = rng.choice(
                ['cost', 'distance', 'no_dentist', 'fear',
                 'not_severe_enough', 'traditional_preference'],
                p=[0.25, 0.20, 0.20, 0.15, 0.10, 0.10])

        rec['fluoride_varnish'] = 0
        if rec['child'] and sc['fluoride_programme']:
            rec['fluoride_varnish'] = 1 if rng.random() < 0.20 else 0

        rec['oral_health_education'] = 0
        if rec['sought_dental_care'] and sc['dentist_available']:
            rec['oral_health_education'] = 1 if rng.random() < 0.30 else 0

        # ── 7. Outcome ──
        rec['pain_resolved'] = 0
        if rec['treatment_received'] not in ('none', 'referral'):
            rec['pain_resolved'] = 1 if rng.random() < 0.70 else 0

        rec['complication'] = 0
        if rec['dental_abscess'] and rec['treatment_received'] == 'none':
            rec['complication'] = 1 if rng.random() < 0.15 else 0

        rec['noma_disfigurement'] = 0
        if rec['noma']:
            rec['noma_disfigurement'] = 1 if rng.random() < 0.80 else 0

        records.append(rec)

    df = pd.DataFrame(records)

    print(f"\n{'='*65}")
    print(f"Oral Health — {scenario} (n={n}, seed={seed})")
    print(f"{'='*65}")
    print(f"\n  Dental caries: {df['dental_caries'].mean()*100:.1f}%")
    print(f"  Untreated: {df['untreated_caries'].mean()*100:.1f}%")
    print(f"  Periodontal: {df['periodontal_disease'].mean()*100:.1f}%")
    print(f"  Dental pain: {df['dental_pain'].mean()*100:.1f}%")
    print(f"  Sought care: {df['sought_dental_care'].mean()*100:.1f}%")
    print(f"  Mean DMFT: {df[df['dental_caries']==1]['dmft_score'].mean():.1f}")

    return df


if __name__ == '__main__':
    parser = argparse.ArgumentParser(
        description='Generate oral health dataset')
    parser.add_argument('--scenario', type=str, default='district_hospital',
                        choices=list(SCENARIOS.keys()))
    parser.add_argument('--n', type=int, default=10000)
    parser.add_argument('--seed', type=int, default=42)
    parser.add_argument('--output', type=str, default=None)
    parser.add_argument('--all-scenarios', action='store_true')
    args = parser.parse_args()

    os.makedirs('data', exist_ok=True)

    if args.all_scenarios:
        for sc_name in SCENARIOS:
            df = generate_dataset(n=args.n, seed=args.seed, scenario=sc_name)
            out = os.path.join('data', f'oral_{sc_name}.csv')
            df.to_csv(out, index=False)
            print(f"  -> Saved to {out}\n")
    else:
        df = generate_dataset(n=args.n, seed=args.seed, scenario=args.scenario)
        out = args.output or os.path.join('data', f'oral_{args.scenario}.csv')
        df.to_csv(out, index=False)
        print(f"  -> Saved to {out}")