# PRE-ECV Score: Validation of a Pre-Specified Tool for Predicting the Success of External Cephalic Version at Term—Proof of Concept

## TL;DR

This study validates a tool called PRE-ECV Score to predict the success of a procedure to turn a baby to head-down position before birth.

## Contribution

The study introduces and validates a pre-specified clinical prediction tool for ECV success using expert consensus and scientific evidence.

## Key findings

- The PRE-ECV Score showed moderate discrimination with an AUC of 0.76.
- ECV success rates increased with higher score strata, from 34.6% to 100%.
- Calibration was moderate, with a corrected intercept of −3.02 and slope of 0.70.

## Abstract

Objectives: We aimed to validate the predictive performance of the PRE-ECV Score, a pre-specified clinical tool, using logistic recalibration and bootstrap optimism correction to estimate the probability of successful external cephalic version (ECV) at term. Methods: The PRE-ECV Score was defined a priori based on a literature review and expert consensus, incorporating eight variables supported by level A or B scientific evidence (total score range 0–13). Validation of this pre-specified score was performed in a retrospective, single-center cohort of 100 consecutive ECV procedures between November 2023 and October 2025. Study conduct and reporting adhered to the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) statement. Results: The revised PRE-ECV Score demonstrated moderate discrimination (AUC = 0.76; 95% CI, 0.66–0.85). Calibration was moderate, with a bootstrap-corrected intercept of −3.02 (95% CI, −5.42 to −1.51) and a calibration slope of 0.70 (95% CI, 0.44–1.18). Observed ECV success rates increased across score strata, from 34.6% (0–4 points) to 79.4% (5–8 points) and 100% (≥9 points). Conclusions: The PRE-ECV Score demonstrated moderate discrimination and satisfactory calibration in this single-center validation. Given the small sample and model optimism, these results should be interpreted strictly as proof-of-concept.