Directly observed therapy for tuberculosis: the Harlem Hospital experience, 1993.

OBJECTIVES: A directly observed therapy program was established at Harlem Hospital, New York, NY, in 1993 to promote high tuberculosis treatment completion rates. METHODS: The Harlem program used an on-site surrogate family model. Treatment completion rate, visit adherence rate, human immuno-deficiency virus seroprevalence, and time to sputum culture conversion were assessed. RESULTS: Out of 145 enrolled patients with suspected and confirmed tuberculosis, 95 (92 confirmed and 3 suspected) continued treatment. The visit adherence rate was 91.1 +/- 7.9%, with one patient (1%) lost to follow-up. CONCLUSION: High rates of treatment completion and visit adherence were achieved because of unique program characteristics. Thus, directly observed therapy is advocated as a means of ensuring treatment completion.