ObjectiveTo compare rates of cearean birth, endometritis, chorioamnionitis, and serious neonatal infections among pregnancies complicated by premature rupture of membranes (PROM) at term and managed by immediate oxytocin induction, by conservative management (or delayed oxytocin induction), or by vaginal (or endocervical) prostaglandin E2gel, suppositories, or tablets.Data SourcesThe English-language literature in MED-LINE and other databases was searched through April 1996 using the terms “fetal membranes,” “premature rupture,” and “term.”Methods of Study SelectionWe included randomized trials comparing two or more management schemes for PROM at term.Tabulation, Integration, and ResultsTwenty-three studies with a total of 7493 subjects met the inclusion criteria and were included for analysis. Data regaridng chorioamnionitis, endometritis, neonatal infections, and cesarean delivery were extracted. Meta-analyses were performed for the three interventions for these outcomes of interest using the Dersimonian and Laird and Mantel-Haenszel techniques to estimate the pooled odds ratios (ORs). No statistically significant differences in cesarean deliveries or neonatal infections were noted among management schemes. Vaginal prostaglandins resulted in more chorioamnionitis than immediate oxytocin (OR 1.55, 95% confidence interval [CI] 1.09, 2.21), but less chorioamnionitis than immediate oxytocin induction resulted in fewer cases of chorioamnionitis (OR 0.67, 95% CI 0.52 0.85) and endometritis (OR 0.71, 95% CI 0.51, 0.99) than conservative management, although these results achieved significance only with the Mantel-Haenszel technique.ConclusionConservative management may result in more maternal infections than immediate induction with oxytocin or prostaglandins.