Using the Geriatric Emergency Perioperative Risk Index Derived From Artificial Intelligence Algorithms to Predict...

Area of Science:

• Geriatric Medicine
• Surgical Oncology
• Artificial Intelligence

Background:

• Elderly patients undergoing emergency general surgery (EGS) face significant risks.
• Accurate prediction of surgical outcomes is crucial for this demographic.
• Existing risk assessment tools may not fully capture the complexities of EGS in older adults.

Purpose of the Study:

• To develop an artificial intelligence (AI) model for forecasting EGS outcomes in elderly patients.
• To create an innovative visual scoring system, the Geriatric Emergency Perioperative Risk Index (GEPR), based on the AI model.
• To enhance risk stratification and clinical decision-making for geriatric EGS.

Main Methods:

• Retrospective analysis of a geriatric EGS patient database.
• Development of an AI model using a specialized four-step algorithm.
• Derivation of the GEPR scoring system based on AI model feature importance.

Main Results:

• The RandomForestClassifier algorithm demonstrated superior performance among AI models.
• The GEPR score ranges from 0-26, with a C-statistic of 0.872 for predicting in-hospital mortality.
• In-hospital mortality risk increases significantly with higher GEPR scores, reaching 100% at a score of 15.

Conclusions:

• An AI-derived GEPR model effectively predicts postoperative in-hospital mortality in geriatric EGS patients.
• The GEPR model integrates patient-related and technical parameters for robust risk assessment.
• Further validation through clinical studies and prospective multicenter trials is recommended to confirm the model's stability and precision.