An Ensemble Boosting Model for Predicting Transfer to the Pediatric Intensive Care Unit.

•1–3% of pediatric patients in the general ward will be transferred to the PICU.•Early warning scores rely on expert defined rules and spot check observations.•We present a data-driven early deterioration prediction algorithm.•Our model is trained using an ensemble of gradient boosted learning algorithms•Inter-facility evaluation shows improved AUC compared to a rule-based PEWS system.