Setting: A large urban tuberculosis (TB) control program.
Objectives: To identify factors associated with directly observed therapy (DOT) participation and to quantify how early use of DOT affected treatment duration.
Design: A retrospective study of 731 Asian-born patients with drug-susceptible Mycobacterium tuberculosis isolates who were verified in New York City between 1993 and 1997 and completed treatment.
Results: Overall, 297 (41%) of 731 patients in the study participated in DOT for some or all of their TB treatment. DOT participation was significantly associated with TB disease in a pulmonary site (adjusted odds ratio [aOR] 2.85, 95% CI 1.86-4.35), more recent year of diagnosis (aOR 1.70, 95% CI 1.50-1.94) and male sex (aOR 1.86, 95% CI 1.30-2.66). Patients who received > or = 70% of their TB treatment at a health department chest clinic were also significantly more likely to participate in DOT (aOR 3.83, 95% CI 2.55-5.74). Among 297 DOT patients, those who completed treatment by 9 months received a greater amount of treatment by DOT during the first 4 months of treatment than those who took longer to complete treatment.
Conclusion: Earlier DOT participation can lead to overall shorter treatment duration. Health care providers should encourage TB patients to participate in DOT as early as possible in their TB treatment.