A new model predicts perioperative mortality in low-resource settings.

A novel perioperative mortality risk (POMR) model was developed and validated for humanitarian surgical care, utilizing electronic health record data from 49,277 patients. Eight critical, accessible variables were identified: age over 65 years, male sex, Glasgow Coma Scale under 13, high glucose levels, low hemoglobin, elevated heart rate, fever, and high systolic blood pressure. The model showcased an impressive area under the curve, revealing high sensitivity and specificity across different validation sets, making it practical for low-resource environments.