HIV prevalence among persons suspected of tuberculosis: policy implications for India.

BACKGROUND: HIV testing of persons referred for tuberculosis diagnosis (TB suspects) is recommended by World Health Organization but is not a policy in India, where HIV prevalence among TB suspects has never been reported. The current Indian policy of offering HIV testing only to TB cases may limit opportunities for early HIV diagnosis and treatment.
METHODS: All adult TB suspects examined for diagnostic sputum microscopy in Mandya district (2 million population), in December 2010, were offered voluntary HIV counseling and testing. Participants were assessed for subsequent TB diagnosis.
RESULTS: Of 1668 eligible TB suspects, HIV status was ascertained for 1539 (92%). Among these, 108 (7%) were HIV positive. Of the 108, 43 (40%) were newly diagnosed as HIV (ie, not previously known to have HIV infection). To detect a new case of HIV infection, the number needed to screen among TB patients was 13, as compared to an number needed to screen of 37 among "TB suspects not diagnosed as TB". Applied annually in 2010, HIV testing of TB suspects in 2010 could have identified approximately 534 newly diagnosed HIV cases, a 51% increase in district HIV case finding.
CONCLUSIONS: Routine HIV testing of TB suspects was feasible and yielded a large number of HIV cases in absolute terms and would increase district HIV case finding by 51%. The number of patients needed to be HIV tested to find a previously undetected HIV case among TB suspects was greater than for TB cases but was potentially acceptable. Given heterogeneity of HIV epidemic in India, broader surveillance is required before national policy decision.