OBJECTIVE: To assess the prognosis of human immunodeficiency virus (HIV) associated tuberculosis in a high tuberculosis prevalence setting. METHODS: Survival and subsequent AIDS-defining illnesses (ADI) of patients with tuberculosis were compared with patients remaining tuberculosis-free in a prospective cohort study in two university-affiliated adult HIV clinics in Cape Town. RESULTS: Tuberculosis without prior or simultaneous ADI was diagnosed in 141 patients. CD4+ T-lymphocyte count was >200 cells/microl in 67% of the incident tuberculosis cases. Survival in tuberculosis patients was comparable to that of patients with oral hairy leukoplakia and/or oral candidiasis (median = 23.6 vs. 27.8 months respectively; P = 0.59, adjusted hazard ratio [AHR] = 0.87; 95%CI 0.63-1.58), and better than in patients with AIDS (median = 11.5 months; P < 0.001, AHR = 2.37; 95% CI 1.93-4.66). Subsequent ADI were less frequent in tuberculosis than in AIDS patients (AHR = 0.36; 95%CI 0.23-0.58). Survival of patients with pulmonary or extra-pulmonary tuberculosis was similar (P = 0.32). CONCLUSION: Tuberculosis in HIV-infected patients from areas endemic with tuberculosis occurs across a wide spectrum of immune suppression and has a considerably better prognosis than other ADI. Inclusion of tuberculosis in the clinical case definition of AIDS in such areas should be reconsidered.
OBJECTIVE: To assess the prognosis of human immunodeficiency virus (HIV) associated tuberculosis in a high tuberculosis prevalence setting. METHODS: Survival and subsequent AIDS-defining illnesses (ADI) of patients with tuberculosis were compared with patients remaining tuberculosis-free in a prospective cohort study in two university-affiliated adult HIV clinics in Cape Town. RESULTS:Tuberculosis without prior or simultaneous ADI was diagnosed in 141 patients. CD4+ T-lymphocyte count was >200 cells/microl in 67% of the incident tuberculosis cases. Survival in tuberculosispatients was comparable to that of patients with oral hairy leukoplakia and/or oral candidiasis (median = 23.6 vs. 27.8 months respectively; P = 0.59, adjusted hazard ratio [AHR] = 0.87; 95%CI 0.63-1.58), and better than in patients with AIDS (median = 11.5 months; P < 0.001, AHR = 2.37; 95% CI 1.93-4.66). Subsequent ADI were less frequent in tuberculosis than in AIDSpatients (AHR = 0.36; 95%CI 0.23-0.58). Survival of patients with pulmonary or extra-pulmonary tuberculosis was similar (P = 0.32). CONCLUSION:Tuberculosis in HIV-infectedpatients from areas endemic with tuberculosis occurs across a wide spectrum of immune suppression and has a considerably better prognosis than other ADI. Inclusion of tuberculosis in the clinical case definition of AIDS in such areas should be reconsidered.