Establishment and validation of a machine learning based prediction model for termination of pregnancy via cesarean section

Purpose This study aimed to investigate the risk factors related to the termination of pregnancy via cesarean section and establish a prediction model for cesarean section based on the characteristics of pregnant women.Patients and methods: The clinical characteristics of 2552 singleton pregnant women who delivered a live baby between January 2020 and December 2021 were retrospectively reviewed. These women were divided into vaginal delivery group (n = 1850) and cesarean section group (n = 802). Lasso regression analysis was employed to screen the independent risk factors of cesarean section. Multivariate logistic regression analysis was used to establish the prediction model, followed by delineation of nomogram, receiver operating characteristic curve (ROC), calibration curve, Decision Curve Analysis (DCA) and confusion matrix.Results There were 1850 women in the vaginal delivery group and 702 women in the cesarean section group. There were significant differences in the age and height of women, maternal weight at birth, pregestational weight, weight gain during pregnancy, gravida, weeks of pregnancy, use of assisted reproductive technology, abnormal blood glucose, hypertension disorders of pregnancy(HDP), scarred uterus, premature rupture of membrane༈PROM༉, placenta previa, floating head, abnormal fetal presentation, cord entanglement and labor analgesia between two groups (P Conclusion The logistic regression prediction model of cesarean section has favorable discrimination, accuracy and consistency and can be employed as a reference for clinicians to improve the outcomes of pregnant women and neonates.