Maternal morbidity after preterm premature rupture of membranes at

Abstract

After preterm premature rupture of membranes at <24 weeks' gestation, pregnant women may choose continuation (expectant management) or termination of pregnancy, via either dilation and evacuation or labor induction. Neonatal outcomes after expectant management are well described. In contrast, limited research addresses maternal outcomes associated with expectant management compared to termination of pregnancy. This study aimed to compare maternal morbidity after preterm premature rupture of membranes at <24 weeks' gestation in women who choose either expectant management or termination of pregnancy. This retrospective cohort study included women with preterm premature rupture of membranes between 14 0/7 and 23 6/7 weeks' gestation with singleton or twin pregnancies at 3 institutions from 2011 to 2018. We excluded pregnancies complicated by fetal anomalies, rupture of membranes immediately after obstetrical procedures (chorionic villus sampling, amniocentesis, cerclage placement, fetal reduction), spontaneous delivery <24 hours after membrane rupture, and contraindications to expectant management. Our primary outcome was the difference in composite maternal morbidity between women choosing expectant management and women choosing termination of pregnancy. We defined composite maternal morbidity as at least 1 of the following: chorioamnionitis, endometritis, sepsis, unplanned operative procedure after delivery (dilation and curettage, laparoscopy, or laparotomy), injury requiring repair, unplanned hysterectomy, unplanned hysterotomy (excluding cesarean delivery), uterine rupture, hemorrhage of >1000 mL, transfusion, admission to the maternal intensive care unit, acute renal insufficiency, venous thromboembolism, pulmonary embolism, and readmission to the hospital within 6 weeks. We compared the demographic and antenatal characteristics of women choosing expectant management with that of women choosing termination of pregnancy and used logistic regression to quantify the association between initial management decision and composite maternal morbidity. We identified 350 women with pregnancies complicated by preterm premature rupture of membranes at <24 weeks' gestation, and 208 women were eligible for the study. Of the 208 women, 108 (51.9%) chose expectant management as initial management, and 100 (48.1%) chose termination of pregnancy as initial management. Among women selecting termination of pregnancy, 67.0% underwent labor induction, and 33.0% underwent dilation and evacuation. Compared to women who chose termination of pregnancy, women who chose expectant management had 4.1 times the odds of developing chorioamnionitis (38.0% vs 13.0%; 95% confidence interval, 2.03-8.26) and 2.44 times the odds of postpartum hemorrhage (23.1% vs 11.0%; 95% confidence interval, 1.13-5.26). Admissions to the intensive care unit and unplanned hysterectomy only occurred after expectant management (2.8% vs 0.0% and 0.9% vs 0.0%). Of women who chose expectant management, 36.2% delivered via cesarean delivery with 56.4% non-low transverse uterine incisions. Composite maternal morbidity rates were 60.2% in the expectant management group and 33.0% in the termination of pregnancy group. After adjusting for gestational age at rupture, site, race and ethnicity, gestational age at entry to prenatal care, preterm premature rupture of membranes in a previous pregnancy, twin pregnancy, smoking, cerclage, and cervical examination at the time of presentation, expectant management was associated with 3.47 times the odds of composite maternal morbidity (95% confidence interval, 1.52-7.93), corresponding to an adjusted relative risk of 1.91 (95% confidence interval, 1.35-2.73). Among women who chose expectant management, 15.7% avoided morbidity and had a neonate who survived to discharge. Expectant management for preterm premature rupture of membranes at <24 weeks' gestation was associated with a significantly increased risk of maternal morbidity when compared to termination of pregnancy.