Table 1. Characteristics of the derivation and external validation cohorts.
| Ventanilla derivation cohort (n=2017) | Callao external-validation cohort (n=1910) | p value | |
|---|---|---|---|
| Contacts | |||
| Age at recruitment (years) | 30 (22–43) | 38 (25–52) | <0.0001* |
| High-risk age group (15–19 or >50 years) | 602 (30%) | 761(40%) | <0.0001 |
| Men/women | 814 (40%)/1203 (60%) | 914 (48%)/996 (52%) | <0.0001 |
| Body-mass index (kg/m2)† | 25.2 (4.2) | 25.6 (3.9) | 0.005‡ |
| History of tuberculosis | 222 (11%) | 197 (10%) | 0.5 |
| Index cases | |||
| Age at recruitment (years) | 26 (20–36) | 28 (21–42) | <0.0001* |
| Men/women | 1210 (60%)/807 (40%) | 1240 (65%)/670 (35%) | 0.001 |
| Sputum smear status | <0.0001 | ||
| Negative | 53 (3%) | 126 (7%) | |
| 1+ | 726 (36%) | 1104 (58%) | |
| 2+ | 635 (31%) | 402 (21%) | |
| 3+ | 603 (30%) | 278 (15%) | |
| Drug sensitivity | 0.004 | ||
| Sensitive | 1620 (80%) | 1498 (78%) | |
| Resistant to isoniazid | 171 (9%) | 223 (12%) | |
| Multidrug resistant§ | 226 (11%) | 189 (10%) | |
| Sustained exposure to index case¶ | 1071 (53%) | 1188 (62%) | <0.0001 |
| Migrant from coastal, mountainous, or jungle area of Peru | 1054 (52%) | 414 (22%) | <0.0001 |
| Households | |||
| Exposure to indoor air pollution|| | 698 (35%) | 20 (1%) | <0.0001 |
| Any household member with a history of tuberculosis | 742 (37%) | 801 (42%) | 0.001 |
| Fewer windows per room** | 797 (40%) | ||
| Wall material | <0.0001 | ||
| Adobe | 223 (12%) | 75 (4%) | |
| Wood | 842 (42%) | 257 (13%) | |
| Cement or brick | 942 (47%) | 1578 (83%) | |
| Floor material | <0.0001 | ||
| Dirt | 641 (32%) | 105 (6%) | |
| Cement | 1212 (60%) | 1301 (68%) | |
| Tiles or laminated surface | 164 (8%) | 504 (26%) | |
| Access to piped water inside the house | 1049 (52%) | 1852 (97%) | <0.0001 |
| Access to a toilet inside the house | 972 (48%) | 1851(97%) | <0.0001 |
| Electric lighting | 1911 (95%) | 1868 (98%) | <0.0001 |
| Asset ownership | <0.0001 | ||
| Television | 1859 (92%) | 1865 (98%) | |
| Stove | 1979 (98%) | 1843 (96%) | 0.002 |
| Fridge | 988 (49%) | 1604(84%) | <0.0001 |
| Head of household did not complete secondary education | 1155 (57%) | 730 (38%) | <0.0001 |
Data are median (IQR), number (%), or mean (SD). All p values stated represent χ2 tests unless otherwise stated. See appendix pp 1–6 for a full description of these variables and the rationale for their definition.
Mann-Whitney U test.
Adjusted with WHO BMI-for-age charts (for ages 15, 16, 17, and 18 years, multiplied by 1·12, 1·09, 1·05, and 1·02, respectively).
Two sample t test.
Defined by initial prescription of a multidrug-resistant tuberculosis regimen or by microbiological evidence of resistance to rifampicin and isoniazid.
≥5 h/day to index case in the 2 weeks before index case diagnosis (derivation cohort) or ≥60 h while index case had cough (validation cohort). ||Living in a household that cooked predominantly with kerosene (or occasionally solid fuels: wood, coal, animal dung, or crop wastes).
Defined as <0.67 windows per room in the derivation cohort; data not available for validation cohort and, therefore, to calculate a risk score for contacts we gave all participants an average value (appendix pp 1–3).