Development and Internal Validation of a Multivariable Prediction Model for Complications After External Cephalic Version in Singleton Term Pregnancies

Objectives: This study aimed to describe the incidence and clinical spectrum of external cephalic versión (ECV)-related complications, to evaluate pre-procedural predictors of their occurrence, and to develop and internally validate a prediction model for ECV-related complications. Methods: This single-center cohort study included ECV procedures performed in singleton pregnancies with non-cephalic presentation at term between January 2014 and December 2025. All procedures followed a standardized protocol, including tocolysis and either deep sedation or spinal anesthesia. ECV-related complications were defined as maternal or fetal adverse events occurring during the procedure or within 24 h after ECV. A prespecified multivariable logistic regression model was developed using clinically relevant pre-procedural variables. Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC), and internal validation was performed by bootstrap resampling. Results: A total of 1050 ECV procedures were included, of which 733 (69.7%) were successful. Complication status was available for 1050 procedures, among which 113 ECV-related complications were recorded (10.8%). Complications were more frequent after failed than successful ECV (17.6% vs. 7.8%). The most frequent events were non-reassuring fetal heart rate patterns, major vaginal bleeding, spotting, and uterine contractions. Emergent cesarean delivery within 24 h occurred in 72 complication cases (63.7%). No neonatal deaths occurred among cases with ECV-related complications. In the final multivariable model, lower maternal body mass index (adjusted OR 0.92 per kg/m2), lower deepest vertical pocket (adjusted OR 0.69 per 10 mm), and breech rather than transverse lie (adjusted OR 0.24) were independently associated with ECV-related complications. Adding deepest vertical pocket significantly improved model discrimination from an AUC of 0.645 to 0.714. Bootstrap internal validation demonstrated limited optimism, with an optimism-corrected AUC of 0.699. Conclusions: ECV performed under a standardized protocol achieved a high success rate with an acceptable safety profile, although approximately one in ten procedures was associated with an ECV-related complication when broadly defined. Lower maternal BMI, reduced amniotic fluid volume, and breech presentation were the main pre-procedural predictors of complications. These findings may facilitate individualized counselling and procedural planning, although external validation is required before clinical implementation.