An integrated model with classification criteria to predict vaginal delivery success after cesarean section

Abstract

Background: Cesarean delivery (CD) is the most frequently performed surgical procedure worldwide. Trial of labor after cesarean (TOLAC) is associated with an increase in perinatal complications related to uterine rupture. However, in general, vaginal birth after cesarean (VBAC) is considered safe and women have less morbidity than those who undergo an elective repeat CD.

Objective: To develop an integrated model with the best performance criteria for predicting vaginal delivery success after CD.

Study design: Retrospective observational study including 2367 women who underwent a TOLAC. A predictive model using classification and regression tree modeling was constructed to predict vaginal delivery using maternal demographic, medical history, and labor predictors.

Results: Vaginal delivery was best predicted by spontaneous onset of labor, estimated fetal weight <3775 g, maternal body mass index <25, previous CD as an elective or for fetal distress reasons, and interdelivery interval <2290 days. The algorithm showed a sensitivity of 75%, a specificity of 53%, and the area under the curve was 0.69.

Conclusions: The classification and regression tree algorithm can be used to develop a predictive model for the success of TOLAC.

Keywords:

• Classification and regression trees
• vaginal delivery after cesarean section
• trial of labor
• vaginal delivery
• cesarean delivery
• risk assessment

Disclosure statement

No potential conflict of interest was reported by the authors.