Value of Some Risk Scores in Predicting Cardiovascular Events After Gastrointestinal Surgery

What this Trial Is About

Researchers are evaluating an artificial intelligence (AI)-based multimodal model designed to predict major cardiovascular events within 30 days after gastrointestinal surgery in adults. This observational study compares the AI model's predictive performance with commonly used traditional risk scores, such as the Revised Cardiac Risk Index (RCRI), ACS NSQIP MICA, and ACS NSQIP Surgical Risk Calculator. The study addresses important questions about whether the AI model can better identify risks in surgical patients by analyzing complex clinical data. The study includes adult patients undergoing gastrointestinal surgery at Bach Mai Hospital. It uses a mixed retrospective and prospective design, collecting data from patients treated in 2025 and those treated prospectively in 2026. The study does not change routine clinical care but reviews medical records and collects clinical information including demographics, medical history, surgical details, laboratory results, electrocardiographic findings, and biomarkers when available. Participants' data will be analyzed to predict major cardiovascular events such as cardiovascular death, nonfatal heart attacks, cardiac arrest with recovery, new stroke, and arrhythmias needing treatment within 30 days after surgery. Researchers will assess the AI model's discrimination, calibration, and reclassification improvements compared to traditional risk scores during the preoperative period through 30 days post-surgery. The study maintains confidentiality by coding data and focuses on improving risk stratification without altering patient management.

Comparing Traditional Risk Scores and an AI-Based Multimodal Model for Predicting Cardiovascular Events After Gastrointestinal Surgery