Objectives: To determine treatment completion among patients with tuberculosis (TB), and to analyse factors associated with treatment default and fatality.
Methods: A prospective cohort study of patients who began treatment between 1 June 1999 and 31 May 2000 in areas where members of the SEPAR Tuberculosis and Respiratory Infections Group work. Factors associated with treatment default and fatality were studied using logistic regression, calculating odds ratios (OR) and their 95% confidence intervals (95%CI).
Results: The study involved 142 physicians from 76 different hospitals who provided information on 1515 cases. Eighty-two per cent of the patients completed treatment correctly, 14% defaulted, 5% died, 0.5% failed, and 8.7% interrupted treatment due to transfer or other reasons. The variables associated with default were intravenous drug use (IVDU) (OR 6.00, 95%CI 2.59-13.89) and immigration (OR 8.57, 95%CI 3.78-19.45); sex, age, homelessness, incarceration, directly observed treatment (DOT) or hospitalisation were not associated with default. Variables found to be predictive of fatality were alcoholism (OR 6.38, 95%CI 2.09-19.48), human immunodeficiency virus (HIV) infection (OR 7.08, 95%CI 2.08-29.15) and age >64 years (OR 10, 95%CI 2.9-34.07), whereas sex, IVDU, homelessness, DOT and hospitalisation were not.
Conclusions: In industrialised countries, IVDU patients and immigrants should be targeted for DOT, while to reduce fatality rates stricter monitoring is required for patients who are alcoholic, HIV-infected, or aged >64 years.