Prevalence and risk factors for positive tuberculin skin tests among active drug users at a syringe exchange program

To evaluate the prevalence and predictors of tuberculin skin test (TST) reactions > or =10 mm among active injection drug users (IDUs) at a syringe exchange program in New York City.

From August 1995 to January 1996, participants were offered TB screening, an interview, and received $15.00 upon returning for skin test interpretation.

610/650 (94%) consented to screening. Of the 566 (93%) who returned for skin test readings, skin test data were available for 564 (99.8%); 14% (95% CI 11.6-17.4) had TSTs > or =10 mm. When the > or =5 mm threshold for interpretation of TST among HIV-infected persons was used, the prevalence of TST positivity increased by only 1%. In univariate analysis, the prevalence of TST > or =10 mm increased with age and with increasing years of IDU (both P = 0.001). Because of a strong correlation between age and duration of IDU, two logistic regression models were examined. In the model with age alone, a history of self-reported TST positivity (OR 8.88; 95% CI 4.9-16.09; P = 0.0001) and increasing age (OR per 10 years increase in age, 1.69; 95% CI 1.24-2.29; P = 0.0008) were independent predictors of TSTs > or =10 mm. In the model with duration of IDU, a history of TST positivity (OR 8.82; 95% CI 4.74-16.41; P = 0.0001) and duration of IDU (OR per 10 years of IDU, 1.46; 95% CI 1.10-1.94; P = 0.0081) were independent predictors of TST > or =10 mm.

Use of the reduced cutoff point for TST positivity from 10 mm to 5 mm did not significantly affect the prevalence of positive TSTs in this cohort of active drug users. Increased prevalence of TB infection with age suggests a high annual incidence of TB infection in this population, and the increased risk of TB infection with increasing duration of IDU suggests that the duration spent in IDU environments may increase infection risk.