Maternal mortality associated with tuberculosis-HIV-1 co-infection in Durban, South Africa

Objectives: To document the impact of tuberculosis and HIV-1 on maternal mo rtality. Design: Prospective study, 1997 and 1998; retrospective analysis, 1996. Participants: Known maternal deaths, defined as the death of a mother withi n a year post-delivery, were studied in Durban, KwaZulu Natal. The HIV-1 st atus, presence of tuberculosis, maternal clinical features and perinatal ou tcomes were documented. The overall as well as HIV-1 and tuberculosis-speci fic maternal mortality rates for the hospital were calculated. The attribut able fraction of deaths as a result of HIV-1 was calculated in the overall group and in those with tuberculosis co-infection. Results: A total of 50 518 deliveries and 101 maternal deaths were recorded . Of the deaths, 29.7% (30/101) were HIV-1 infected. The overall mortality rate was 200/ 100 000; for HIV-1-infected women this was 323.3/100 000, HIV -1-negative mothers, 148.6/100 000 live births. The attributable fraction o f overall deaths as a result of HIV-1 was 15.9% Fourteen of the 15 mothers with tuberculosis were HIV-1 co-infected. The mortality rate for tuberculos is and HIV-1 co-infection was 121.7/1000; for tuberculosis without HIV-1 co -infection, 38.5/1000. Fifty-four per cent of maternal deaths caused by tub erculosis were attributable to HIV-1 infection. Thirty-five per cent of mat ernal deaths were associated with stillbirths; perinatal outcomes were no d ifferent between groups of mothers with tuberculosis, HIV-1 or neither infe ction. Conclusion: Tuberculosis and HIV-1 are emerging as significant contributors to maternal mortality in KwaZulu Natal. Any attempt to improve maternal he alth must also include careful screening and investigation for tuberculosis in high-risk pregnant women. (C) 2001 Lippincott Williams & Wilkins.