Explanatory Variables and Nomogram of a Clinical Prediction Model to Estimate the Risk of Caesarean Section After Term Induction

The aims of this study were to identify the explanatory variables associated with failure of induction of labour (IOL) and to designate nomograms that predict probability. This retrospective study included 1328 singleton term pregnant women (37-42 weeks). The penalised maximum likelihood estimation (PMLE) method was used instead of traditional logistic regression. Of the 25,678 deliveries that occurred during the study period, 1328 (5.1%) women underwent term delivery. Of those, 1125 (84.7%) had successful vaginal deliveries and 203 (15.3%) had failed vaginal deliveries following use of a dinoprostone slow-release vaginal insert. Explanatory variables were discovered that were associated with delivery failure in term pregnancy undergoing induction of labour with an unfavourable cervix, and a nomogram that predicted probability was developed.IMPACT STATEMENT The caesarean rate has continued to climb worldwide over the past decade. Most caesarean sections are performed because of suspected foetal distress or failure to progress. In absolute numbers, most caesarean deliveries are performed in women with a term pregnancy with a foetus in cephalic presentation. Despite these numbers, predicting the mode of delivery by which these women will deliver remains a challenge. Five explanatory variables were strongly associated with failure of dinoprostone delivery of term pregnancies: nulliparity, induction time, premature rupture of membranes, Bishop score and foetal gender The developed nomograms enable fast and easy implementation in clinical practice. After external validation and proof of generalisability, the present model could be used in obstetric clinical management.