A new predictive model for cardiopulmonary resuscitation (CPR) outcomes in surgical patients offers crucial insights for patient selection and risk assessment.

• Among 6,405 patients, 57.9% died within 30 days following CPR.
• The extreme gradient boosting model showed 80% accuracy for predicting 30-day mortality.
• Key predictors included ASA status, case urgency, and frailty, allowing for better-informed surgical decisions.

This model helps tailor interventions and supports goal-concordant care in surgical settings.