Validation of Artificial Intelligence-Based POTTER Calculator in Emergency General Surgery Patients Undergoing Laparotomy: Prospective, Bi-Institutional Study.

Abstract

The Predictive Optimal Trees in Emergency Surgery Risk (POTTER) calculator, a widely used interpretable artificial intelligence risk calculator, has been validated in population-based studies and shown to predict outcomes in patients who underwent emergency general surgery better than surgeons. We sought to prospectively validate POTTER. Patients undergoing an emergency exploratory laparotomy for nontrauma indications at 2 academic medical centers between June 2020 and March 2022 were included. POTTER preoperative risk calculations and postoperative outcomes were systematically recorded. POTTER's performance in predicting 30-day postoperative mortality, septic shock, respiratory failure, bleeding, and pneumonia was assessed using the c-statistic methodology. A total of 361 patients were included. The median age was 63 years (interquartile range 51 to 72 years), 45.4% were women, and the overall mortality and morbidity were 24.1% and 51.4%, respectively. POTTER predicted mortality accurately with a c-statistic of 0.90. POTTER also accurately predicted the occurrence of individual postoperative complications, with c-statistics ranging between 0.80 and 0.89. This is the first prospective validation of the artificial intelligence-enabled POTTER calculator. The superior accuracy, user-friendliness, and interpretability of POTTER make it a useful bedside tool for preoperative patient and family counseling.