The dramatic rise in cesarean delivery rates worldwide in recent decades, without evidence of a concomitant decrease in cerebral palsy rates, has raised concerns about its potential negative consequences for maternal and infant health. In 2014, the American College of Obstetricians and Gynecologists and the Society for Maternal-Fetal Medicine jointly published an Obstetric Care Consensus for safe prevention of the primary cesarean delivery.

We sought to assess whether modification of our protocol to implement these recommendations helped to decrease our primary cesarean delivery rate safely.

This is a before-and-after retrospective cohort study at a university referral hospital. In March 2014, the threshold for defining active labor changed from 4 to >6 cm and arrest of first-stage labor from lack of cervical change despite regular contractions after 3 hours of oxytocin administration with amniotomy and epidural anesthesia to no change after 4 hours of adequate or 6 hours of inadequate contractions in women with an epidural. The definition of second-stage arrest of labor changed simultaneously from lack of progress for 3 hours with adequate contractions in women with epidural anesthesia to no progress for ≥4 hours in nulliparas or 3 hours in multiparas with an epidural. We compared maternal and neonatal outcomes over two 1 year periods: from March 2013 to February 2014 (before, preguideline) and from June 2014 to May 2015 (after, postguideline). We included all women with singleton pregnancies at ≥37 weeks’ gestation, in vertex presentation, in spontaneous or induced labor, and with epidural anesthesia. We excluded women with an elective or previous cesarean delivery and those with obstetric or fetal complications.

This study included 3283 and 3068 women in the before and after periods, respectively. The groups had similar general and obstetric characteristics. The global cesarean delivery rate decreased significantly from 9.4% in the preguideline to 6.9% in the postguideline period (odds ratio, 0.71; 95% confidence interval, 0.59–0.85; P < .01). The cesarean delivery rate for arrest of first-stage labor fell by half, from 1.8% to 0.9% (odds ratio, 0.51; 95% confidence interval, 0.31–0.81; P < .01) but was significant only among nulliparous women. The cesarean delivery rate for second-stage arrest of labor decreased but not significantly between periods (1.3% vs 1.0%; odds ratio, 0.73; 95% confidence interval, 0.44–1.22; P = .2), and the cesarean delivery rate for failure of induction remained similar (3.7% vs 3.5%; odds ratio, 1.06; 95% confidence interval, 0.06–13.24; P = .88). The median duration of labor before cesarean delivery also became significantly longer among nulliparous women during the later period. Maternal and neonatal outcomes did not differ between the 2 periods, except that the rate of 1 minute Apgar score <7 fell significantly in the later period (8.4% vs 6.9%; odds ratio, 0.80; 95% confidence interval, 0.66–0.97; P = .02).

The modification of our protocol by implementing the new consensus recommendations was associated with a reduction of the rate of primary cesarean delivery performed for arrest of labor with no apparent increase in immediate adverse neonatal outcomes in nulliparous women at term with singleton pregnancies in vertex presentation and with epidural anesthesia. Further studies are needed to assess the long-term maternal and neonatal safety of these policies.