Objective: To identify risk factors associated with default, failure and death among tuberculosis patients treated in a newly implemented DOTS programme in South India.
Design: Analysis of all patients registered from May 1999 through April 2000. A community survey for active tuberculosis was underway in the area; patients identified in the community survey were also treated in this programme.
Results: In all, 676 patients were registered during the period of the study. Among new smear-positive patients (n = 295), 74% were cured, 17% defaulted, 5% died and 4% failed treatment. In multivariate analysis (n = 676), higher default rates were associated with irregular treatment (adjusted odds ratio [AOR] 4.3; 95%CI 2.5-7.4), being male (AOR 3.4; 95%CI 1.5-8.2), history of previous treatment (AOR 2.8; 95%CI 1.6-4.9), alcoholism (AOR 2.2; 95%CI 1.3-3.6), and diagnosis by community survey (AOR 2.1; 95%CI 1.2-3.6). Patients with multidrug-resistant tuberculosis (MDR-TB) were more likely to fail treatment (33% vs. 3%; P < 0.001). More than half of the patients receiving Category II treatment who remained sputum-positive after 3 or 4 months of treatment had MDR-TB, and a large proportion of these patients failed treatment. Higher death rates were independently associated with weight <35 kg (AOR 3.8; 95%CI 1.9-7.8) and history of previous treatment (AOR 3.3; 95%CI 1.5-7.0).
Conclusions: During this first year of DOTS implementation with sub-optimal performance, high rates of default and death were responsible for low cure rates. Male patients and those with alcoholism were at increased risk of default, as were patients identified by community survey. To prevent default, directly observed treatment should be made more convenient for patients. To reduce mortality, the possible role of nutritional interventions should be explored among underweight patients.