651 Clinical efficacy of a machine learning model for prediction of shoulder dystocia

A newly introduced shoulder dystocia (SD) machine learning (ML) model (Tsur et al, 2019) uses maternal demographics, obstetric history, and sonographic evaluation within 5 weeks from delivery. Our primary aim was to externally validate the calculator for prediction of SD or birth injury in patients who attempted a vaginal Secondary objective assess clinical utility optimizing number cesarean delivery (CD) suspected large gestational age. We conducted multicenter retrospective study all non-anomalous singleton pregnancies with estimated fetal weight (EFW) ? 35 between January 2013 June 2019. All people labor underwent due macrosomia EFW ?4000g 4 were included. matched 3500g 3999g also included analysis. The ROC area under curve (AUC) used as metric performance compared (ACOG recommendations). total 490 met inclusion criteria, whom 381 (77.8%) 109 (22.2%) had CD macrosomia. 19 cases our cohort. ML achieved an AUC 0.77, significantly better than alone (0.61, p-value = 0.028). Furthermore, we evaluated various risk scores above 0.5 (Table). score 0.92 could identify prevent 53% neonatal while reducing rate by 6%. In this validation cohort, using found be more accurate injury. Based on these findings, are planning prospective investigating efficacy current recommendations SD, associated morbidity indication