The effect of labor epidural analgesia on labor, delivery, and neonatal outcomes: a propensity score-matched analysis in a single Japanese institute

Abstract

Purpose: Lumbar epidural analgesia (LEA) is the most widely used method in reducing labor pain. Previous RCTs have shown that LEA does not increase cesarean section rates; however, the results are inconsistent and may vary depending on the different backgrounds. Therefore, we aimed to study whether LEA would affect the course of labor in our institute.

Methods: Delivery records from October 2013 to April 2016 were collected. Deliveries at gestational age < 36 weeks and multiple pregnancies were excluded. All cases were divided into the non-epidural labor (NEL) group or the epidural labor (EL) group. We performed a propensity score matching analysis to balance intergroup differences. Our primary outcome was a mode of delivery (spontaneous, assisted vaginal, cesarean). Secondary outcomes were lengths of labor and outcomes of the neonates.

Results: During the study period, 2632 cases met the inclusion criteria. All analyses were performed after propensity score matching (218 pairs). The percentage of assisted vaginal delivery increased by the use of LEA (11.5% in NEL group vs 25.7% in EL group; p < 0.001), but the rate of cesarean section was similar (12.8% vs 17.0%; p = 0.23). The durations of the first and second stages of labor were prolonged by the use of LEA in both primipara and multipara women. Outcomes of the neonates were similar in both groups.

Conclusion: Use of LEA did not increase the rate of cesarean section when analyzed by propensity score-matched analysis in our institute.

Keywords: Assisted vaginal delivery; Cesarean section; Epidural labor; Labor; Propensity score-matched analysis.

Fig. 1

All deliveries in our institute were retrospectively investigated as shown in the flow chart. During the study period, there were 4206 deliveries. Final cases that met the criteria were 2632 cases. Within these cases, epidural analgesia was performed in 226 cases

Fig. 2

Each panel shows the Kaplan–Meier Curve for the first stage of labor for primipara (a), second stage for primipara (b), first stage for multipara (c), and second stage for multipara (d). The difference in the average time between the EL and NEL group was statistically significant between two groups compared with log-rank test (p < 0.01)