Predictors of shoulder dystocia at the time of operative vaginal delivery: a prospective cohort study

Abstract Our aim was to identify factors associated with shoulder dystocia following an attempted operative vaginal delivery (aOVD) in a prospective cohort study and to evaluate whether these factors can be used to accurately predict shoulder dystocia by building a score of shoulder dystocia risk. This was a planned secondary analysis of a prospective cohort study of deliveries with aOVD at term from 2008–2013. Cases were defined as women with shoulder dystocia following an aOVD defined as a delivery that requires additional obstetric maneuvers following failure of gentle downward traction on the fetal head to effect delivery of the shoulders. Multivariate logistic regression analyses were performed to determine risk factors for shoulder dystocia. Shoulder dystocia occurred in 57 (2.7%) of the 2118 women included. In the whole cohort, women with shoulder dystocia more often had a history of shoulder dystocia (3.5% vs. 0.2%, p = 0.01), and there was a significant interaction between aOVD and gestational age and the duration of the second stage of labor: women with shoulder dystocia more often had a gestational age > 40 weeks and a second stage of labor longer than 3 h specifically for midpelvic aOVD. In multivariable analysis, a history of shoulder dystocia was the only factor independently associated with shoulder dystocia following aOVD (aOR 27.00, 95% CI 4.10–178.00). The AUC for the receiver operating characteristic curve generated using a multivariate model with term interaction with head station was 0.70 (95% CI 0.62–0.77). The model failed to accurately predict shoulder dystocia.