Cohort analysis of directly observed treatment outcomes for tuberculosis patients in urban Pakistan.

BACKGROUND This quasi-experimental cohort study aimed to evaluate World Health Organization (WHO) defined tuberculosis (TB) treatment outcomes for patients under directly observed treatment at a health facility (clinic DOT) or at home (family DOT) in urban Pakistan. METHODS We enrolled 582 sputum smear-positive TB patients being treated by either clinic DOT (n = 295) or family DOT (n = 287) in 11 treatment centres. Patients and/or family members were interviewed for baseline measurements. WHO-defined treatment outcomes were evaluated at the end of treatment. Proportions of 'cured' patients were computed. A log-binomial model was used to evaluate the associations of various factors with 'cured' status. RESULTS The proportion of 'cured' patients was respectively 66% and 34% in the clinic DOT and family DOT groups (risk difference 0.32; 95%CI 0.24-0.39). Patients on clinic DOT were more likely to achieve cure (adjusted relative risk [RR(adj)] 1.85; 95%CI 1.43-2.39) than those on family DOT, as were patients satisfied with their health care worker's attitude (RR(adj) 5.73; 95%CI 2.54-12.96). CONCLUSION Clinic DOT nearly doubled the proportion of cured patients compared to family DOT. Efforts to improve care-provider attitudes to enhance patient satisfaction, and effective implementation of the WHO's public-private mix approach, may enhance TB control in this and similar settings.