ABSTRACT
BACKGROUND: Tuberculosis, declared a global emergency by the World Health Organization, continues to be an important public health problem in Mexico, included in the first twenty causes of death. OBJECTIVE: To know the impact of drug resistance of Mycobacterium tuberculosis on treatment outcome, need of re-treatment and mortality in a cohort of patients with pulmonary tuberculosis receiving directly observed therapy, short course (DOTS). METHODS: We conducted a population-based study in a suburban region in Southern Mexico. People who had been coughing for more than two weeks underwent sputum acid-fast bacilli smear. Patients with a positive smear were recruited and underwent clinical exam, chest X-ray, HIV testing, and sputum cultures. Identification, drug susceptibility testing and restriction fragment length polymorphism analysis (RFLP) were performed in all isolates. Patients were followed every 12 months for new episodes of tuberculosis and vital status. Patients were referred for clinical care to the local program of tuberculosis. Deaths were corroborated with death certificates. Informed consent was obtained from participants. RESULTS: Between March 1995 and February 1999, tuberculosis was diagnosed in 371 patients who were followed for an average of 32 months. M. tuberculosis was cultured from 316 patients; resistance to any drug occurred in 25.0% of isolates (primary 18.8%, acquired 49.2%); only to isoniazid in 6.8% (primary 7.3%, acquired 4.8%); to isoniazid and rifampin in 6.2% (primary 1.6%, acquired 23.8%). Patients with drug resistance had a higher probability of treatment failure (OR = 16.9, CI 95% 4.5-63.0) and patients with MDR strains had a higher probability of need of re-treatment (RR = 24.4, CI 95% 8.8-67.6), and of death (RR = 4.0, CI 95% 1.5-10.7). Additional variables were found to be associated with subsequent episodes of disease and mortality: Cocaine use, chronic disease, type of radiological lesions, HIV co-infection, non-compliance and treatment delay, as well as RFLP clustering. CONCLUSIONS: In this study, we observed that drug resistance showed a severe impact on the outcome and survival; drug-resistance was the most significant factor for these negative outcomes; DOTS may not be sufficient in areas where drug resistance is considerable, and patient follow-up for longer periods of time, as compared to evaluation at the end of treatment, provides additional information which is useful for prevention and control programs.