Analysis of Obstetric Outcomes by Hospital Location, Volume, and Teaching Status Associated With Non–Medically Indicated Induction of Labor at 39 Weeks

ABSTRACT Previous research has shown that non–medically indicated induction of labor contributes to some favorable outcomes compared with expectant management in low-risk pregnancies. Cesarean birth rates have risen in recent years and represent a significant impact on long-term health of both mother and child. Research has connected these 2 phenomena, showing that non–medically indicated induction may reduce the overall likelihood of cesarean delivery, but it is not yet known how these outcomes are affected by hospital facility variables. This study was designed to examine outcome differences in women undergoing non–medically indicated induction at 39 weeks' gestation based on hospital characteristics such as obstetric volume, location, and teaching status. This study was designed as a retrospective cohort study, analyzing births between January 1, 2007, and December 31, 2011. Inclusion criteria were singleton, nonanomalous births at 39 weeks 0 days' gestation to 41 weeks 6 days' gestation among nulliparous mothers. Births with missing data on induction or hospital type were excluded, as well as births to pregnant individuals with comorbid conditions, placenta previa, breech presentation, stillbirths, and elective or planned cesarean delivery. The primary outcome was cesarean birth, and secondary outcomes included perinatal outcomes of severe maternal morbidity, chorioamnionitis, postpartum hemorrhage, operative vaginal birth, and obstetric anal sphincter injury, as well as neonatal outcomes of neonatal intensive care unit admission for more than 24 hours, respiratory distress syndrome, and shoulder dystocia. Analysis included 455,044 births, with 24,272 (5.3%) experiencing non–medically indicated induction of labor. Significant differences were found between non–medically indicated induction and expectant management when the sample was stratified by urban versus rural settings. Cesarean birth was significantly less likely among those who underwent non–medically indicated induction, with lower odds for rural (adjusted odds ratio [aOR], 0.68; 99% confidence interval [CI], 0.53–0.86) versus urban hospitals (aOR, 0.78; 99% CI, 0.74–0.81). For other outcomes assessed based on location, urban hospitals showed significantly lower odds of severe maternal morbidity (aOR, 0.78; 95% CI, 0.61–0.98), chorioamnionitis (aOR, 0.26; 99% CI, 0.22–0.30), postpartum hemorrhage (aOR, 0.73; 99% CI, 0.65–0.83), operative vaginal birth (aOR, 0.85; 99% CI, 0.79–0.90), and obstetric anal sphincter injury (aOR, 0.90; 99% CI, 0.82–0.98). Odds were also decreased for the neonatal outcomes of neonatal intensive care unit admission for more than 24 hours (aOR, 0.71; 99% CI, 0.66–0.77) and respiratory distress syndrome (aOR, 0.64; 99% CI, 0.57–0.71). In an analysis stratified by obstetric volume, odds of cesarean delivery were lower for those with non–medically indicated induction in both medium- and high-volume hospitals (aORs, 0.86 [99% CI, 0.78–0.94] and 0.73 [99% CI, 0.69–0.77], respectively). Stratified by hospital teaching status, significant differences were found in women who experienced non–medically indicated induction in the areas of race and ethnicity, age, BMI, and insurance type. Odds of cesarean delivery were again lower with induction, both in community and academic hospitals (aORs, 0.78 [99% CI, 0.74–0.81] and 0.67 [99% CI, 0.56–0.80], respectively). No significant differences between hospital types were observed. These findings support previous research showing that cesarean birth rates are lowered by non–medically indicated induction of labor at 39 weeks' gestation and show that previous findings hold across hospitals of different sizes, locations, and teaching status. With the increase in non–medically indicated induction, it is necessary for clinicians to optimize outcomes for mothers and infants using any evidence-based methods available. Further research could focus on variations between clinicians or hospital guidelines or policies. In addition, methods of induction have yet to be analyzed and tend to change over time; these could be opportunities for reanalysis to ensure the most benefit to patients in labor and delivery.