Background. Active case finding (ACF) among contacts of patients with tuberculosis is a global health priority, but the effects of ACF compared to passive case finding (PCF) are poorly characterised. We assessed the contribution of ACF versus PCF to tuberculosis detection among contacts and compared sex and disease characteristics between contacts diagnosed through these strategies.

Methods. In shantytowns in Callao, Peru, we identified index patients with tuberculosis and followed their contacts aged ≥15 years for tuberculosis for a median 10 years. All were offered free programmatic ACF involving sputum smear microscopy and clinical evaluation. Additionally, all contacts were offered intensified ACF with sputum smear and culture testing monthly for six months and then once every four years. PCF at local health facilities was ongoing throughout follow up.

Findings. 8.7% (232/2,666) of contacts were diagnosed with tuberculosis: a two-year cumulative risk of 4.6% and overall incidence of 0.98/100 person-years. 23% (53/232) were diagnosed through ACF and 77% (179/232) through PCF. During the first six months, 45% (23/51) were diagnosed through ACF and 55% (28/51) through PCF. Contacts diagnosed through ACF versus PCF were more frequently female (68% [36/53] versus 47% [85/179], p=0.009); had a shorter symptom duration (25% [9/36] had symptoms for less than 15 days versus 8% [10/127], p=0.03); and were more likely to be sputum smear negative (62% [33/53] versus 35% [62/179], p<0.001).

Interpretation. Although ACF made an important contribution to tuberculosis detection among contacts, PCF detected the majority of the tuberculosis burden. Compared to PCF, ACF was equitable, diagnosed tuberculosis earlier and usually before it became sputum smear positive, and revealed a high burden of undetected tuberculosis among women.