RESUMO
SETTING: Tomsk, Siberia, Russian Federation. OBJECTIVE: To evaluate the relationship between TB susceptibility patterns and risk factors among a civilian cohort of new cases in Tomsk city in 1999. DESIGN: Population-based study. The association between MDR-TB or PROMDR-TB, defined as resistance to isoniazid and rifampicin (MDR) or to isoniazid, ethambutol, and streptomycin (rifampicin mono-sensitive), and hypothesized risk factors was determined. Univariable analysis with and without stratification for history of incarceration and stepwise logistic regression modeling were used. RESULTS: Overall, 49.6% of participants were infected with a Mycobacterium tuberculosis strain resistant to at least one prescribed anti-tuberculosis medication. PROMDR-TB and MDR-TB were prevalent in 17.2% and 13.1% of participants, respectively. Logistic regression modeling indicated that good residence (OR 3.1, 95%CI 1.4-6.9), treatment default (OR 4.4, 95%CI 2.1-9.3) and psychological disorder (OR 3.3, 95%CI 1.0-10.9) were associated with PROMDR-TB. Both good residence (OR 2.6, 95%CI 1.1-6.0) and treatment default (OR 5.3, 95%CI 2.4-11.6) were associated with MDR-TB. History of incarceration was not found to be significant. CONCLUSION: Our findings support the hypothesis that drug-resistant disease among the Tomsk city population is not directly linked to history of incarceration, nor is it an extension of drug resistance in prisons. Rather, drug resistance in the civil sector reflects problems specific to the sector itself.