The association between maternal HIV infection and perinatal outcome: a systematic review of the literature and meta-analysis.

To investigate the association between maternal HIV infection and perinatal outcome by a systematic review of the literature and meta-analysis. Appropriate publications were identified using electronic and hand searching of relevant journals from 1983 to 1996. Studies were included in the review if they were prospective cohorts with pregnant women identified as being HIV-infected with a control group of pregnant women who were not infected with HIV. Methodological quality was assessed for each study. Data were extracted for pre-determined outcome measures. Sensitivity analyses were performed to explore the association between HIV infection and an adverse perinatal outcome for the following study characteristics: clinical setting (developed or developing countries), methodological quality (high or poor) and whether studies controlled for potential confounding. Thirty-one studies were eligible to be included in the review. The summary odds ratio of the risk of pre-defined adverse perinatal outcomes related to maternal HIV infection were as follows: spontaneous abortion 4.05 (95% CI 2.75-5.96); stillbirth 3.91 (95% CI 2.65-5.77); fetal abnormality 1.08 (95% CI 0.7-1.66); perinatal mortality 1.79 (95% CI 1.14-2.81); neonatal mortality 1.10 (95% CI 0.63-1.93); infant mortality 3.69 (95% CI 3.03-4.49); intrauterine growth retardation 1.7 (95% CI 1.43-2.02); low birthweight 2.09 (95% CI 1.86-2.35) and pre-term delivery 1 83 (95% CI 1.63-2.06). Sensitivity analyses showed that the association between infant mortality and maternal HIV infection was stronger in studies conducted in developing countries when compared with developed countries [odds ratios (OR) 3.72 (95% CI 3.05-4.54) and 8.6 (95% CI 0.53-141.05), respectively]; studies of higher methodological quality compared with those of poorer quality [odds ratios 14.57 (95% CI 6.93-30.65) and 3.37 (95% CI 2.74-4.14), respectively] and studies which had used restriction or matching to control for potential confounding factors compared with those that had not [OR 11.60 (95% CI 5.71-23.58) and 3.35 (95% CI 2.73-4.12), respectively]. The findings of this review have implications for women infected with HIV who are planning a pregnancy or who find themselves pregnant. There appears to be an association, although not strong, between maternal HIV infection and an adverse perinatal outcome. This relationship may be due to bias including uncontrolled or residual confounding. There does, however, appear to be a real and large increase in the risk of infant death in developing countries associated with maternal HIV infection, especially so when there has been an attempt to control for confounding.