To determine the prevalence of Mycobacterium tuberculosis infection and the incidence of tuberculosis in HIV-infected and uninfected urban Rwandan women, 460 HIV-positive and 998 HIV-negative childbearing women were recruited from pediatric and prenatal care clinics and were enrolled in a prospective study in 1988 and followed for 2 yr. Tuberculin testing was administered 12 to 18 months after enrollment. Fifty-three percent of HIV-negative women had positive tuberculin tests (induration > or = 10 mm), with higher rates among older women and among women who had received BCG vaccine. Only 21% of HIV-positive women had positive tuberculin tests, with no relationship to BCG vaccine. Follow-up was available for 93% of subjects. Tuberculosis was diagnosed in 20 HIV-positive women and in two HIV-negative women. Features associated with an increased risk of tuberculosis among HIV-positive women included: age > or = 30, body mass index in the lowest quartile, low income, erythrocyte sedimentation rate > 75, positive tuberculin test, and chronic cough, chronic fever, and weight loss. Among Rwandan women who are infected with HIV, approximately half of those who are infected with M. tuberculosis do not have positive tuberculin tests. The rate ratio for development of tuberculosis among HIV-positive women was 22 (95% CI, 5 to 92). New algorithms are needed to improve the early detection of tuberculosis among HIV-positive patients in Africa.
PIP: This study determines the prevalence of Mycobacterium tuberculosis (TB) infection and the incidence among HIV infected and uninfected women in urban Rwanda. The sample population includes 460 HIV-positive women and 998 HIV-negative women who were recruited from pediatric and prenatal care clinics at the Centre Hospitalier de Kigali. The sample is considered representative of childbearing women from the capital city. Initial interviews were conducted in 1988 and followed-up in 1990. HIV-1 diagnosis was determined on the basis of enzyme immunoassay and western blot tests or indirect immunofluorescence that showed reactivity to both a core protein and an envelope protein. A positive tuberculin test was defined as induration of 10 mm or more. Routine visits were made every 6 months. Comparisons were made between women who were HIV positive at their first HIV test (250 women), women who were negative at their first test but seroconverted between the first HIV test and the TB test 3 years later (80 women), and women who were HIV negative at the time of TB testing (687). 55% of HIV-negative women had positive tests with induration of more than 10 mm. 25% of HIV-positive women and 66% of HIV-negative women had TB tests with over 5 mm induration. 31% of HIV-positive women and 70% of HIV-negative women had induration of over 2 mm. 77% of women had TB vaccine scars. Prevalence of a positive test was significantly higher in the HIV-negative vaccinated group than in the nonvaccinated group. The proportion with low white cell counts, low lymphocyte counts, and high sedimentation rates was higher among HIV-positive women than HIV-negative women. During the 2-year follow-up period, 20 of the 401 HIV-positive women and 2 of the 917 HIV-negative women were diagnosed with TB. The risk ratio was 22.9. The incidence of TB was 3 times higher among women who had been infected with HIV at least 18 months than among women who had been infected less than 18 months. Low income and low body mass were associated with an increased risk of TB. 9 out of 17 HIV-infected women with TB had negative TB tests.