We enrolled 427 consecutive patients with tuberculosis diagnosed in Cité Soleil, Haiti in a trial of short-course intermittent therapy. All patients received supervised therapy with isoniazid, rifampin, pyrazinamide, and ethambutol thrice weekly for 8 wk, followed by isoniazid and rifampin thrice weekly for 18 wk. At entry, the 177 human immunodeficiency virus (HIV)-infected patients (42%) were found significantly more likely to have extrapulmonary tuberculosis and negative tuberculin skin tests (p < 0.05). Treatment was well tolerated by both groups of patients, and adherence to the treatment regimen was over 90%. Among patients with pulmonary or intrathoracic tuberculosis, 9% of HIV-seropositive and 1% of HIV-seronegative patients died during therapy (p < 0.001), whereas 81% and 87%, respectively, of those in the two groups were cured. Relapses occurred in 5.4% of HIV-seropositive and 2.8% of HIV-seronegative patients who completed treatment (p = 0.36). Survival after tuberculosis was poorer in HIV-seropositive patients, whose probability of dying was 33% at 18 mo after diagnosis as compared with 3% for HIV-seronegative patients (p < 0.001). HIV-seropositive patients who died had significantly lower median CD4 lymphocyte counts than did HIV-seropositive patients who survived (p < 0.001). Treatment of tuberculosis with short-course, thrice-weekly, supervised therapy in the setting of a developing country is highly efficacious in both HIV-seropositive and -seronegative patients.

Infection with the human immunodeficiency virus type 1 (HIV-1) results in decreased cell-mediated immunity, which includes decreased delayed hypersensitivity to skin test antigens. HIV-1 seropositivity and skin test reactivity to purified protein derivative (PPD) were determined among 2042 healthy Haitian adults with normal chest radiographs. Among HIV-1-seropositive individuals, 52.3% (146/279) had PPD reactions greater than or equal to 10 mm compared with 67.2% (1184/1763) of the seronegative adults (P less than .001). However, the percentage of HIV-1-seropositive individuals with PPD reactions greater than or equal to 5 mm was similar to the percentage of seronegative adults with PPD reactions greater than or equal to 10 mm (180/279 [64.5%] vs. 1184/1763 [67.2%]). Assuming that the rate of prior infection with Mycobacterium tuberculosis was similar for HIV-1-seronegative and -seropositive populations, these data provide support for the recent recommendations to use induration of greater than or equal to 5 mm as evidence of past infection with M. tuberculosis in HIV-1 seropositive adults.