Antiretroviral program associated with reduction in untreated prevalent tuberculosis in a South African township.

RATIONALE: In 2005, we reported high prevalence of untreated pulmonary tuberculosis (TB) in a South African community. Prevalent untreated TB is the main source of transmission. In settings with large burdens of human immunodeficiency virus (HIV) and TB, highly active antiretroviral therapy (HAART) may contribute to TB control. OBJECTIVES: To assess the community-level impact of HAART on TB prevalence, we repeated a community-based TB prevalence cross-sectional survey in 2008 following HAART roll-out. METHODS: A random 10% adult population sample was identified from the community. Participants provided two sputum specimens for acid-fast bacilli microscopy and TB culture. Oral transudate specimen was collected for anonymous HIV testing, linked to TB diagnosis. An interviewer-administered, structured questionnaire identified TB and HIV history and risk factors. MEASUREMENTS AND MAIN RESULTS: In the 2008 survey, 1,250 adults participated (90% response rate); 306 (25%) tested HIV positive, of which 60 (20%) were receiving HAART. A total of 20 TB cases were identified (12 receiving TB treatment), representing a significant decline in prevalence from 3.2 to 1.6% (P = 0.02) between the surveys. TB prevalence in participants not infected with HIV was unchanged (P = 0.90). The decline occurred among participants not infected with HIV, decreasing from 9.2 to 3.6% in 2005 to 2008, respectively (P = 0.003). In participants infected with HIV, prevalence of treated TB declined from 4 to 2.3% (P = 0.06), and untreated TB prevalence from 5.2 to 1.3% (P = 0.02). The proportion of untreated TB in patients receiving HAART decreased significantly, from 22 to 0% (P < 0.001). CONCLUSIONS: Prevalence of undiagnosed TB declined significantly over a period of increasing HAART availability. The decline was predominantly in individuals infected with HIV receiving HAART.