Twice-weekly, directly observed treatment for HIV-infected and uninfected tuberculosis patients: cohort study in rural South Africa

Objective: To determine the effectiveness of twice-weekly directly observed therapy (DOT) for tuberculosis (TB) in HIV-infected and uninfected patient s, irrespective of their previous treatment history. Also to determine the predictive value of 2-3 month smears on treatment outcome. Methods: Four hundred and sixteen new and 113 previously treated adults wit h culture positive pulmonary TB (58% HIV infected, 9% combined drug resista nce) in Hlabisa, South Africa. Daily isoniazid (H), rifampicin (R), pyrazin amide (Z) and ethambutol (E) given in hospital (median 17 days), followed b y HRZE twice a week to 2 months and HR twice a week to 6 months in the comm unity. Results: Outcomes at 6 months among the 416 new patients were: transferred out 2%; interrupted treatment 17%; completed treatment 3%; failure 2%; and cured 71%. Outcomes were similar among HIV-infected and uninfected patients except for death (6 versus 2%; P = 0.03). Cure was frequent among adherent HIV-infected (97%; 95% CI 94-99%) and uninfected (96%; 95% CI 92-99%) new patients. Outcomes were similar among previously treated and new patients, except for death (11 versus 4%; P = 0.01), and cure among adherent previous ly treated patients 97% (95% CI 92-99%) was high. Smear results at 2 months did not predict the final outcome. Conclusion: A twice-weekly rifampicin-containing drug regimen given under D OT cures most adherent patients irrespective of HIV status and previous tre atment history. The 2 month smear may be safely omitted. Relapse rates need to be determined, and an improved system of keeping treatment interrupters on therapy is needed. Simplified TB treatment may aid implementation of th e DOTS strategy in settings with high TB caseloads secondary to the HIV epi demic. (C) 1999 Lippincott Williams & Wilkins.