Predicting Vaginal Birth After Previous Cesarean: Using Machine-learning Models and a Population-based Cohort in Sweden

Abstract

(Acta Obstet Gynecol Scand. 2021;100:513–520. doi: 10.1111/aogs.14020. Epub October 31, 2020) There are risks and benefits to choosing a trial of labor after cesarean (TOLAC) or a repeat cesarean delivery (CD). When TOLAC is successful (international success rate is 40% to 80%) and the patient has a vaginal birth (VD) after cesarean (VBAC), a benefit is decreased epidemic CD rate and decreased risk for repeat CD maternal morbidity. However risks of TOLAC include unplanned repeat CD, uterine rupture, and greater likelihood of maternal and neonatal adverse outcomes. Due to these tradeoffs, it is important to make accurate predictions regarding potential TOLAC success, thus leading to improved clinical counseling, and outcomes for women. This study aims to evaluate individualized predictions using machine-learning models for TOLAC for women with no previous VD and only one previous CD. It also aims to compare machine-learning methods with established regression prediction models.