Determinants of labor epidural analgesia uptake and associations with maternal-neonatal outcomes: a stratified cohort study with risk prediction modeling

Comprehensive Summary

Epidural analgesia (EA) is recognized as one of the most effective measures to control labor pain, a significant physiological and psychological stressor women experience during childbirth. In this article, Guo et al. explore factors that impact utilization of EA and its association with both labor progression and post-partum hemorrhage (PPH). Using data collected from 1035 women at a single hospital over the span of three years, the authors found multiple independent predictors of EA uptake. These include younger age, first-time motherhood, and lower pre-delivery fibrinogen levels. The authors then built a logistic regression prediction model with acceptable accuracy (AUC = 0.70) that performed on par with more complex machine learning approaches and provided superior interpretability. EA was shown to prolong both first and second stages of labor amongst vaginal deliveries, however, it did not increase the rate of postpartum hemorrhage or worsen any neonatal outcomes such as Apgar scores or NICU admission. Importantly, the authors also show that EA is not an independent risk factor for PPH as compared to elevated D-dimer and fibrinogen levels which are associated with this negative outcome. The authors conclude that EA remains a safe and viable option for most women undergoing labor. Monitoring pre-delivery hematologic status is a more meaningful tool for predicting PPH than avoiding EA.

Outcomes and Implications

This study provides evidence that can guide discussions with patients regarding EA usage, its associated risks, and maternal-neonatal outcomes. Beyond guiding clinicians and patients through individualized decisions, the performance of the developed predictive model shows the utility of this technology in estimating EA demand in a real world setting. Advancements of these tools can help better allocate resources and training to best meet the needs of a dynamic patient population.