In this prospective observational cohort study from Taiwan the outcome of HIV infected TB patients and HIV infected non-TB patients treated with highly active antiretroviral therapy (HAART) was compared.

A total of 46 TB and 230 non-TB antiretroviral naive patients were included between June 1994 and October 2002. The median duration of antituberculous therapy was 9 months (range 1–24). Viral clearance (20 of 46 v 107 of 230, RR 0.93 (95% CI 0.65 to 1.34); p = 0.71) at week 4 of HAART was similar in the two groups, as was the virological failure during HAART (RR 1.49 (95% CI 0.92 to 2.41); p = 0.14). The CD4 cell count increased in both groups (71 v 64 × 10⁶ cells/l, p = 0.70). The risk for HIV progression to new opportunistic illnesses (adjusted RR 1.16 (95% CI 0.764 to 1.77)) and the adjusted hazard ratio for death of TB patients compared with non-TB patients was also similar in the two groups (1.18 (95% CI 0.65 to 2.32) before HAART era and 0.89 (95% CI 0.57 to 1.69) in HAART era).

This study addresses the important issue of HIV and TB co-infection and shows that TB does not alter the virological and immunological responses and the clinical outcome when these patients are treated with HAART and an appropriate antituberculous therapy. The authors do point out, however, that the rate of multidrug resistant TB was low and few of these patients were intravenous drug users or homeless, distinguishing them from some other study populations. They had access to high quality follow up and treatment. Ideally, we would like to have a situation where we could provide care of an equally high standard to all patients with HIV and TB co-infection to ensure that these results are widely applicable.