Background: Global attempts are made to reduce cesarean deliveries rates. Special focus has been directed towards parturients undergoing a trial of labor after cesarean deliveries in order to reduce the burden of repeated cesarean deliveries.

Big data methods enable investigation of large-scale sets of data and can combine input parameters better than traditional statistical analysis. We aimed to use these methods to aid in decision making prior to a trial of vaginal labor after cesarean delivery.

Objective: This study aims to evaluate the feasibility of using machine-learning methods to predict a successful vaginal birth after cesarean delivery.

Study design: The electronic medical records of labors in a tertiary referral center were explored. Using gradient boosting, a model incorporating multiple maternal and neonatal variables was created to predict the risk for emergency cesarean deliveries in those undergoing trial of vaginal labor after cesarean deliveries. Elective repeated cesarean deliveries, multifetal gestations and preterm deliveries were excluded.

Results: We identified 7,486 cases of women who attempted a vaginal labor after one prior cesarean delivery. A machine learning based model to predict when vaginal delivery would be successful was developed. The area under the curve of the model was 0.78.

Additionally, a risk stratification tool was built to allocate parturients into low and high-risk groups for failed trial of labor after cesarean delivery.

Conclusion: An individualized prediction model of the success of vaginal birth after previous cesarean delivery was developed and showed good performance. Such a model can be used to reduce the rate of unnecessary cesarean deliveries.