Setting: A residential program in Barcelona for drug addicts (therapeutic community) admitted between November 1988 and March 1992, and followed until September 1994.
Objective: To study the incidence of tuberculosis as related to the presence of tuberculosis infection and/or human immunodeficiency virus (HIV) infection, and to evaluate the protective effect of chemoprophylaxis with isoniazid.
Design: Prospective cohort study. Incidence rates were compared using the Chi-square test for cohort studies. The effectiveness of chemoprophylaxis was evaluated by the Kaplan-Meier method at the univariate level, and by logistic regression models and proportional risks analysis at the multivariate level.
Results: During the study of 361 individuals without previous known tuberculosis or history of anti-tuberculosis chemoprophylaxis, 25 developed tuberculosis, an overall incidence rate of 1.79/100 person-years. For HIV-positive persons, the incidence rate was 3.25/100 person-years, compared with 0.30/100 in those who were HIV-negative (P < 0.05). The highest incidence rates occurred among HIV-positive persons who did not receive chemoprophylaxis and who were either anergic (HIV-positive, purified protein derivative [PPD]-negative, Multitest-negative) or who were infected with Mycobacterium tuberculosis (PPD+), 10.0/100 person-years and 4.64/100 person-years, respectively. Of the 53 persons who received chemoprophylaxis, three developed tuberculosis, an incidence rate of 1.4/100 person-years. In comparison, in the group of 51 patients who were designated to receive chemoprophylaxis but where none was actually taken, 17 developed tuberculosis, an incidence rate of 5.7/100 person-years (P = 0.03).
Conclusion: HIV-infected intravenous drug users, particularly those who are anergic or who are PPD positive, are at increased risk of developing tuberculosis. Anti-tuberculosis chemoprophylaxis proved effective in this population.