Duration of ruptured membranes and vertical transmission of HIV-1: a meta-analysis from 15 prospective cohort studies.

OBJECTIVE: To test the a priori hypothesis that longer duration of ruptured membranes is associated with increased risk of vertical transmission of HIV. DESIGN: The relationship between duration of ruptured membranes and vertical transmission of HIV was evaluated in an individual patient data meta-analysis. METHODS: Eligible studies were prospective cohort studies including at least 100 mother-child pairs, from regions where HIV-infected women are counselled not to breastfeed. Analyses were restricted to vaginal deliveries and non-elective Cesarean sections; elective Cesarean section deliveries (those performed before onset of labour and before rupture of membranes) were excluded. RESULTS: The primary analysis included 4721 deliveries with duration of ruptured membranes < or = 24 h. After adjusting for other factors known to be associated with vertical transmission using logistic regression analysis to assess the strength of the relationship, the risk of vertical HIV transmission increased approximately 2% with an increase of 1 h in the duration of ruptured membranes [adjusted odds ratio, 1.02; 95% confidence interval, 1.01-1.04; for each 1 h increment]. There were no significant interactions of duration of ruptured membranes with study cohort or with any of the covariates, except maternal AIDS. Among women diagnosed with AIDS, the estimated probability of transmission increased from 8% to 31% with duration of ruptured membranes of 2 h and 24 h respectively (P < 0.01). CONCLUSIONS: These results support the importance of duration of ruptured membranes as a risk factor for vertical transmission of HIV and suggest that a diagnosis of AIDS in the mother at the time of delivery may potentiate the effect of duration of ruptured membranes