Comparison of Artificial Intelligence and Logistic Regression for Postoperative Mortality Prediction

A systematic review and meta-analysis was conducted comparing artificial intelligence and machine learning models with logistic regression in predicting 30-day and 90-day postoperative mortality in adult patients undergoing abdominal or visceral surgery. The study protocol was registered in the PROSPERO database on May 5, 2026.

Postoperative mortality constitutes a significant burden on global health, with an estimated 4.2 million deaths occurring each year within the first 30 days following surgery. While traditional surgical risk prediction models rely heavily on logistic regression, artificial intelligence and machine learning approaches offer flexible methods for working with high-dimensional data. This review was designed to examine studies directly comparing artificial intelligence-based models with logistic regression within the same surgical population.