Directly observed therapy for tuberculosis given twice weekly in the workplace in urban South Africa.

Effective models of delivery of directly observed therapy (DOT) for tuberculosis in resource-poor settings are needed. Intermittent chemotherapy may be an important component of DOT delivered in the community as it means fewer visits to supervisors. There is no published evidence on the efficacy of twice weekly therapy given from the start of treatment without an intensive daily phase. We analysed data from 3 large cohort studies in a migrant, urban workforce in South Africa between 1975 and 1983. All patients received 4 drugs (isoniazid, rifampicin, pyrazinamide and streptomycin) twice weekly under direct observation by a nurse in the workplace. Of 444 patients, 378 (85.1%) completed treatment. Cure could be assessed in 362, and 348 (96.1%, 95% confidence interval 93.7-97.8%) were bacteriologically cured. Sputum status was assessed at 2-3 months in 343 patients and 309 (90.1%) were sputum negative. Eighteen patients relapsed (5.7%; 2.9/100 patient-years of observation). DOT can be effectively delivered to a migrant, urban workforce, and 4-drug therapy given twice weekly under direct observation is efficacious.