Long-term effectiveness of diagnosing and treating latent tuberculosis infection in a cohort of HIV-infected and at-risk injection drug users

Introduction: Between 1990 and 1998, tuberculin skin tests (TST) and isoniazid preventive therapy (IPT) were provided to injection drug users participating in the AIDS Linked to the Intravenous Experiences cohort. Methods: A registry match was conducted with the AIDS Linked to the Intravenous Experiences cohort database and the Maryland State Department of Health and Mental Hygiene tuberculosis registry. Results: Of 2010 participants, 1753 (74%) had a TST placed and read; 536 (31%) were positive. TST positivity was 16% in HIV positives; 39% in HIV negatives (P < 0.01). Overall, 299 (56%) TST reactors started IPT; 165 (55%) completed 6 months. Three tuberculosis (TB) cases were diagnosed among HIV negatives (incidence rate = 0.16/1000 person-years); 19 among HIV positives (1.94/1000 person-years; incidence rate ratio = 12.3 (3.61-64.70). Within the entire cohort, TB rates were 0.81 per 1000 person-years for those not receiving IPT, 0.48 per 1000 person-years for those receiving any IPT, 0.29 per 1000 person-years for those completing at least 30 days, and 0 per 1000 person-years for completers. Ten cases of TB occurred in HIV-infected individuals with negative TSTs. Discussion: IPT was associated with protection against TB, but uptake was modest. Although it is likely that TB incidence would have increased, especially in HIV-positive subjects, if the IPT program had not occurred, more significant declines in TB incidence in this population will require improved methods for ensuring uptake and completion of IPT and preventing disease in TST-negative individuals.