Data and MATLAB(R2025b) Code for Preoperative Prediction of Intraoperative Transfusion in Pediatric Craniosynostosis Surgery

This repository contains the original dataset, MATLAB analysis code, and selected output files used for the exploratory development and internal validation of preoperative prediction models for intraoperative transfusion in pediatric craniosynostosis surgery. The analysis was performed using ridge-penalized logistic regression with prespecified baseline and full models, where fused suture extent was evaluated as an additional predictor. Internal validation was conducted using bootstrap-based optimism correction, and model tuning was performed using inner 5-fold cross-validation. The repository includes the following files: the original analytic dataset (data.xlsx), the main MATLAB analysis script (fivefoldCI.m), and summary output files generated after model execution, including model performance summaries, paired comparison summaries, model specifications, and patient-level predicted probabilities. These materials are provided to support transparency, reproducibility, and secondary methodological review.