Objective: To determine the impact of drug resistance (DR) on the clinical outcome and transmission of tuberculosis under programmatic conditions.
Methods: Prospective cohort and molecular epidemiologic study in the Orizaba Health Jurisdiction of Mexico. Between March 1995 and July 1999, chronic coughers with positive acid-fast bacilli (AFB) detected in sputum smear underwent clinical and mycobacteriologic evaluation (species identification, drug susceptibility testing and IS6110-based genotyping). Treatment was provided in accordance with official norms.
Results: Mycobacterium tuberculosis was isolated from 326/387 AFB-positive cases. The rate of DR was 24.2% and that of multidrug resistance (MDR, defined as resistance to both isoniazid and rifampin at least) was 7.7%; 78% were cured, 8% abandoned treatment, 6% failed treatment, and 5% died. An additional 13.5% received retreatment and 8.9% died during a median 28.6 months of follow up. Factors associated with DR by multivariate analysis were chronicity of tuberculosis (OR 4.8, 95%CI 2.7-8.4, P < 0.001), age >40 years (OR 1.9, 95%CI 1.1-3.2, P = 0.02) and indigenous origin (OR 0.3, 95%CI 0.13-0.75, P = 0.01). Cox-adjusted relative risks showed that MDR (RR 2.5, 95%CI 1.02-6.16, P = 0.04), HIV infection (RR 31.3, 95%CI 11.6-84.8, P < 0.001), and chronicity of tuberculosis (RR 2.1, 95%CI 1.0-4.4, P = 0.06) were associated with mortality, controlling for age. Predictors of retreatment were DR (not including MDR) (RR 2.2 95%CI 0.89-5.31, P < 0.087), MDR (RR 12.6, 95%CI 5.46-28.88, P < 0.001), and living in a household with an earthen floor (RR 2.8, 95%CI 1.27-6.13, P = 0.011). Being infected with MDR-TB was the only factor associated with a decreased likelihood of being in an RFLP cluster (OR 0.31, 95%CI 0.12-0.81, P = 0.02).
Conclusions: Although MDR-TB may have a decreased propensity to spread and cause disease, it has a profoundly negative impact on tuberculosis control.