We appreciate Dr Leveno’s interest in our recent article.

We conducted the first randomized controlled trial of nulliparous women with singleton gestations at 36 weeks to 41 weeks who reached the American Congress of Obstetricians and Gynecologists definition of prolonged second stage of labor: 3 hours with epidural anesthesia. Women were assigned to extended labor (1 additional hour) or to usual labor (expedited delivery via cesarean delivery [CD] or operative vaginal delivery). The incidence of CD was 19.5% in the extended labor group and 43.2% in the usual labor group (relative risk, 0.45; 95% confidence interval, 0.22–0.93). Extending the length of labor in nulliparous women with singleton gestations, epidural anesthesia, and prolonged second stage decreased the incidence of CD by 55%, compared with usual guidelines. Maternal or neonatal morbidity were not different between the groups; however, our study was underpowered to detect small, but potentially clinical important, differences.

First, our decision to analyze the 1 CD for nonreassuring fetal heart tracing that occurred in a patient randomized to extended labor was because of the intention-to-treat principle. We considered it a crossover because the patient could not be given at least 1 additional hour to push. CD for nonreassuring fetal heart tracing would be an interesting secondary outcome, but our sample size was underpowered.

Second, all of the crossovers from the usual labor (ie, 3 hours) to extended labor (≥4 hours) group resulted in vaginal deliveries. Therefore, the crossover strengthens our study’s results because most of the spontaneous vaginal deliveries occurring in the women in the usual labor group actually occurred in women who crossed over, ie, in the women who instead of having an operative vaginal delivery or CD soon after 3 hours of second stage preferred and were allowed to labor longer.

Third, we cannot hypothesize about the interpretation of the results as if they were different than they are. Of course, if some women had different outcomes, the results would have been different. The unbalanced number of crossovers possibly reflects a trend occurring on labor and delivery in the United States in what some patients and providers seem to want: a bit more time than 3 hours to achieve a vaginal delivery.

Lastly, we agree that safety is a major concern, for both the mother and the fetus. While we conclude that maternal or neonatal morbidity were not statistically different between groups, we eagerly encourage more randomized controlled studies to corroborate our results.