A new machine learning-based index enhances social risk evaluation in surgical patients, crucial for better outcomes and health equity.

• Models using machine learning outperformed conventional indices, showing a median 8.15-fold increase in predictive power across 14 surgical outcomes.
• The study analyzed over 3.2 million patients from 688 hospitals, providing tailored social risk insights for different surgical results.

Surgeons should consider these advanced risk assessments to refine patient selection and enhance surgical outcomes.

• Machine learning variables illuminate specific social factors impacting postoperative results.