Timing of antiretroviral therapy and effects on tuberculosis treatment outcomes in HIV-co-infected patients in Malawi

Setting: Queen Elizabeth Central Hospital, Blantyre, Malawi. Objectives: To determine 1) the proportion of human immunodeficiency virus (HIV) infected tuberculosis (TB) patients started on antiretroviral therapy (ART), 2) the timing of ART and 3) the effect of the timing on TB treatment outcomes. Design: A retrospective record review of HIV-infected TB patients registered from January to December 2009. Results: A total of 3376 TB patients were registered, of whom 2665 (79%) were HIV-tested and 2042 (77%) were HIV-infected. A total of 1190 HIV-infected TB patients who were not on ART at the time of starting TB treatment were studied. Of 688 (58%) who started ART, 61% started therapy within 2 months of anti-tuberculosis treatment and 39% started later (>= 2 months). Treatment success for patients with TB who started ART within 2 months was higher than for those starting ART later (RR 1.6, 95% CI 1.4-1.8), and death rates were lower (RR 0.25, 95% CI 0.19-0.35). Conclusion: Under routine programme conditions in Malawi, a higher proportion of HIV-infected TB patients who started ART did so within 2 months of starting TB treatment, and early ART intervention was associated with better treatment outcomes. This confirms recommendations that co-infected TB patients should start ART early.