The following is a summary of “Prediction of cesarean delivery in class III obese nulliparous women: An externally validated model using machine learning,” published in the September 2023 issue of Gynecology Obstetrics and Human Reproduction by Lodi et al.
Class III obese women are more likely to undergo cesarean section during labor, and cesarean section contributes to increased maternal and neonatal morbidity in this population. This project’s objective was to devise a method for quantifying the risk of a cesarean section before labor. This is a retrospective cohort study involving 410 primiparous class III obese pregnant women who attempted vaginal delivery in two French university institutions. The researchers created two predictive algorithms (a logistic regression model and a random forest model) and evaluated and contrasted their performance levels.
According to the logistic regression model, only initial weight and labor induction were predictive of an unplanned cesarean section. The probability forest was able to predict the likelihood of a cesarean section using only initial weight and labor induction as pre-labor characteristics. Calculated for a cut-off point of 49.5% risk, its performance was higher: area under the curve 0.70 (0.62, 0.78), accuracy 0.66 (0.58, 0.73), specificity 0.87 (0.77, 0.93), and sensitivity 0.44 (0.32, 0.54).
This innovative and effective method for predicting the risk of an unplanned cesarean section in this population may influence the decision between a trial of labor and a planned cesarean section. Additional research is required specifically a prospective clinical trial. Funding: “Plan Investissements d’Avenir” and “Agence Nationale de la Recherche” from the French government