Setting: Tuberculosis (TB) drug resistance survey in six hospitals in Rio de Janeiro, Brazil.
Objective: To estimate resistance to at least one drug (DR) and multidrug resistance (MDR) and identify associated factors.
Design: One-year cross-sectional survey. Hospitals were included as a convenience sample.
Results: Of 595 patients investigated, 156 (26.2%) had previously undergone anti-tuberculosis treatment, 433 (72.8%) were not previously treated and information on the remaining 6 was not available. Overall, DR and MDR rates were high, at respectively 102 (17.1%, 95%CI 14.3-20.5) and 44 (7.4%, 95%CI 5.5-9.9) cases. Among individuals not previously treated, 17 had MDR (3.9%, 95%CI 2.4-6.3) and diagnosis in a TB reference hospital was independently associated with MDR (prevalence ratio [PR] 3.3, 95%CI 1.2-8.7) after multivariate analysis. Among previously treated individuals, 27 had MDR (17.3%, 95%CI 11.7-24.2). MDR-TB was independently associated with diagnosis in a TB reference hospital (PR 3.6, 95%CI 1.5-8.7), male sex (PR 2.3, 95%CI 1.2-4.4) and dyspnoea (PR 0.3, 95%CI 0.1-0.7).
Conclusion: We found high levels of DR- and MDR-TB. Our study design did not permit us to determine the contribution of community versus nosocomial transmission. Further studies are needed to establish this. Nevertheless, hospitals should be recognised as a potential source of transmission of resistant TB strains and urgent measures to avoid nosocomial TB transmission should be taken.