ACTIVE TUBERCULOSIS IN HIV-INFECTED INJECTING DRUG-USERS FROM A LOW-RATE TUBERCULOSIS AREA

This article describes the features of active tuberculosis in HIV-infe cted injecting drug users (IDUs) from a low-rate tuberculosis area. Th e cohort was followed in a hospital-based HIV/AIDS registry, and data were extracted from the registry, patient charts, and the Tuberculosis Control Program of the Connecticut Health Department. The setting was an acute care inner-city hospital-based health care system, with a hi gh incidence of AIDS, serving a small-to-medium urban area in Connecti cut. The patients were 905 HIV-infected IDUs whose time of HIV diagnos is (TOHD) was between 1984 and 1992. The outcome measures were demogra phics, clinical characteristics, and morbidity rates of active tubercu losis. Of the 27 IDUs who developed active tuberculosis, none were whi te, ail but one were male, and only one was known to have had a positi ve purified protein derivative (PPD) reaction prior to TOHD: 59% of ca ses developed in patients known to be HIV infected, 11% occurred in es tablished AIDS patients, and 67% qualified as extrapulmonary tuberculo sis (that is, AIDS defining by pre-1993 definitions). In 22% of cases, both Mycobacterium tuberculosis and M. avium-intracellulare were isol ated. Mycobacterium tuberculosis was most commonly isolated from a res piratory specimen (67%). The annual incidence rate has been less than or equal to 1.0% since 1988. The cumulative incidence rate was highest for patients with a positive PPD reaction or a history of tuberculosi s (1.4 cases/100 patient years; 52 patients; mean follow-up 4.0 years) . The demographics and clinical characteristics of active tuberculosis in our HIV-infected IDUs are similar to those described elsewhere in the United States; the morbidity rates are low and stable. The implica tions of our findings on tuberculosis control in HIV-infected IDUs may be applicable to health care systems with low tuberculosis rates.