Automated machine learning model identifies patients at risk for surgical complications

The model outperformed the established calculator of surgical risk and requires less input from clinicians.