Predictors of mortality among HIV-infected women in Kigali, Rwanda.

ABSTRACT

PIP: In 1988, researchers recruited 18-35 year old women from pediatric and prenatal care clinics at the Centre Hospitalier de Kigali in Rwanda to observe HIV disease progression. They compared probability of survival of the 460 HIV-positive women with that of the 998 HIV-negative women. They used simple clinical and laboratory variables as predictors of mortality from AIDS. The researchers did not use the WHO clinical case definition of AIDS as the outcome measure since 40 and 30 women from each group, respectively, met the criteria for AIDS at entry. Only 66% (25) of the HIV=infected women who died met the criteria for AIDS during the study. After 2 years, mortality among HIV-infected women stood at 7% (39) which was more than 20 times higher than that among women not HIV infected (0.3%; p .001). Mortality was 21% for those who met the WHO criteria for AIDS. The wasting syndrome was the cause of the death in 51% of HIV-infected death cases. The baseline predictors of mortality in HIV-infected women in descending order of prevalence of predictor included an at most body mass index of 21 kg.sq. (48%; relative hazard [RH] 2.3), low income (46%; RH=2.6), mm/hour erythrocyte sedimentation rate (39%; rh = 4.9), chronic diarrhea (10%; RH = 2.6), a history of herpes zoster (9%; RH 5.3), and oral candidiasis (1%; RH 7.3). The erythrocyte sedimentation rate was a better predictor than lymphocyte counts (p .001) and p .11, respectively). Of the 40 HIV-infected women who met the criteria for AIDS, the health of 32 women improved so the physicians no longer considered them to have AIDS. Thus health workers should treat symptomatic HIV-positive cases. AIDS was responsible for 90% of all deaths among reproductive age women living in Kigali. Health workers in Africa can use the simpler erythrocyte sedimentation rate instead of the more costly CD4 counts as a predictor of progression to AIDS.