Effect of intrapartum epidural analgesia on rate of emergency delivery for presumed fetal compromise: nationwide registry-based cohort study

Objectives: To determine the rate of emergency delivery for presumed fetal compromise after epidural analgesia (EDA) compared with that after alternative analgesia or no analgesia, and to assess whether this rate is increased in pregnancies with reduced placental reserve. Methods: This was a nationwide registry-based cohort study of 629 951 singleton pregnancies delivered at 36 + 0 to 42 + 0 weeks of gestation that were recorded in the Dutch national birth registry between 2014 and 2018, including 120 426 cases that received EDA, 86 957 that received alternative analgesia and 422 568 that received no analgesia during labor. Pregnancies with congenital anomaly, chromosomal abnormality, fetal demise, planned Cesarean delivery, non-cephalic presentation at delivery and use of multiple forms of analgesia were excluded. The primary outcome was emergency delivery for presumed fetal compromise. Secondary outcomes included delivery characteristics and neonatal outcome. Negative binomial regression analysis was stratified by parity and results are presented according to birth-weight centile, after adjusting for confounding. Results: Among women who received EDA, 13.2% underwent emergency delivery for presumed fetal compromise, compared with 4.1% of women who had no analgesia (relative risk (RR), 3.23 (95% CI, 3.16–3.31)) and 7.0% of women who received alternative analgesia (RR, 1.72 (95% CI, 1.67–1.77)). Independent of birth weight, the RR of presumed fetal compromise after EDA vs no analgesia was higher in parous women (adjusted RR (aRR), 2.15 (95% CI, 2.04–2.27)) compared with nulliparous women (RR, 1.88 (95% CI, 1.84–1.94)). Stratified for parity, the effect of EDA was modified significantly by birth-weight centile (interaction P-value, < 0.001 for nulliparous and 0.004 for parous women). The emergency delivery rate following EDA was highest in those with a birth weight < 5^th centile (25.2% of nulliparous and 16.6% of parous women), falling with each increasing birth-weight centile category up to the 91^st–95^th centile (11.8% of nulliparous and 7.2% of parous women). Conclusions: Intrapartum EDA is associated with a higher risk of emergency delivery for presumed fetal compromise compared with no analgesia and alternative analgesia, after adjusting for relevant confounding. The highest rate of emergency delivery for presumed fetal compromise was observed at the lowest birth-weight centiles. RRs of emergency delivery for presumed fetal compromise after EDA were modestly but consistently modified by birth-weight centile, supporting the hypothesis that the adverse effects of EDA are exacerbated by reduced placental function. While EDA provides effective pain relief during labor, alternative strategies for pain management may be preferable in pregnancies with a high background risk of fetal compromise.