Management of PROM at term - premature rupture of fetal membranes - POEMS:Patient-Oriented Evidence that Matters

Journal of Family Practice, Sept, 1997 by Mindy Smith

Clinical question Does immediate induction of labor for premature rupture of the fetal membranes (PROM) at term improve maternal and neonatal outcomes compared with conservative management?

Background Rupture of the fetal membranes before the onset of labor occurs in about 10% of pregnancies at term; 90% of these women will be in spontaneous labor by 24 hours after rupture.[1] Numerous small trials and several randomized clinical trials offer conflicting recommendations for the optical management of PROM. Early induction appears to be associated with an increased rate of cesarean section, while expectant management may result in increased maternal and neonatal infectious morbidity.

Population studied Twenty-three randomized clinical trials (RCTs) examining the management of women with PROM at term (36 or more weeks' gestation) by immediate oxytocin induction conservative management, or intervention with prostaglandin [E.sub.2] were identified. Authors performed a MEDLINE search cross-checked reference lists, and identified studies from the Controlled Trials Register of the Cochrane Collaboration.

Study design and validity For this meta-analysis, trial results were combined appropriately using the DerSimonian and Laird (random-effects) technique to estimate pooled odds ratios (OR). Where findings between studies were similar (homogeneous), they were confirmed using the Mantel-Haenszel (fixed-effects) pooled OR Three comparisons were made: immediate oxytocin vs conservative management; immediate prostaglandin use vs conservative management; and immediate prostaglandin use vs immediate oxytocin. Subgroup analyses based on the absence of a digital examination before randomization or an unfavorable cervix (Bishop score 4 or less) were also conducted. Using the above search scheme, it is unlikely that any RCTS were overlooked. The explicit criteria used for serious neonatal infection lend confidence to the validity of this study. However, the use of clinical criteria for the maternal infectious outcomes is poor, and reporting was not uniform or blinded to treatment group. There was also no apparent control for the duration of ruptured membranes prior to hospital arrival, the number of vaginal examinations, or the use of intrapartum antibiotics. In addition, the definition of immediate and late induction varied considerably across studies. Finally, the inclusion of the one large RCT may have overwhelmed the results of the pooled analysis for chorioamnionitis.

Outcomes measured The four outcomes examined were cesarean birth, chorioamnionitis (clinical diagnosis), endometritis (requiring clinical management), and serious neonatal infection (defined as culture-proven neonatal septicemia, meningitis, or pneumonia).

Results A total number of 7493 subjects were pooled across the trials. There were no significant differences in cesarean section rates (10.5% and 9.1% for oxytocin induction and conservative treatment, respectively). While there was no significant difference in the rate of clinical chorioamniotis between the immediate oxytocin and conservative management strategies, a significant reduction in chorioamnionitis was found for the immediate use of vaginal prostaglandin compared with conservative management (5.6% vs 8.1%, OR = 0.68, 95% CI 0.51 to 0.91). An increase in chorioamnionitis between the immediate use of vaginal prostaglandin compared with oxytocin (6% vs 4%, OR = 1.55, 95% CI 1.09 to 2.21) was also noted. There was a small, but significant reduction in postpartum endometritis between the oxytocin and conservative management groups (3.2% vs 4.4%, OR = 0.71, 95% CI 0.51 to 0.99) but not for the other two comparisons. There were no differences between management strategies for serious neonatal infections. Subgroup analyses failed to show differences between management strategies with the exception of more cases of endometritis in the immediate induction group compared with conservative management for women undergoing no digital examinations prior to randomization (OR = 2.8, 95% CI 1.02 to 7.7).

Recommendation for clinical practice The decision to proceed with immediate induction rather than conservative management (with induction of labor at some time beyond 24 hours) in otherwise healthy women with PROM is best made with the woman or couple after discussing the risks and benefits of these approaches. Benefits of early induction appear to include a shorter time to delivery and possibly a lower risk of maternal infection. If the rate of chorioamnionitis is reduced from 8% to 6%, 50 women would have to be treated to prevent one case of chorioamnionitis. This infection, however, results in neonatal infection in only 10% to 20% of neonates.[2] The rate of chorioamnionitis reported here is also higher than in most studies reported in the literature (range 0.5% to 4%).[2] The risks Of early induction include a possible increase in cesarean deliveries and exposure to oxytocin. The role of prostaglandin in reducing the risk of cesarean delivery, particularly among women with unfavorable cervices, is not clear but appears safe.