A New Predictive Model for In-Hospital Major Adverse Cardiac and Cerebrovascular Events in Chinese Patients After Major Noncardiac Surgery

Abstract

Prediction tools focused on cardiovascular and cerebrovascular events after noncardiac surgery are lacking, particularly for Chinese patients. We developed and validated what we believe is a new predictive tool for postoperative major cardiovascular and cerebrovascular events (MACCEs) in Chinese patients in this study. Overall, 401 variables derived from 598 patients who received noncardiac surgery at our center were retrospectively analyzed to develop and validate the new predictive model for MACCEs during hospitalization. The 7 strongest predictors for MACCEs in the development cohort were chronic heart failure, age, atrial fibrillation, general anesthesia, history of coronary heart disease, high-risk procedures, and lymphocyte count. The area under the receiver operating characteristic curve was 0.698 (95% confidence interval 0.616 to 0.780) for the new predictive tool with the validation cohort. Receiver operating characteristic curve analysis showed the new predictive tool had better performance than the Revised Cardiac Risk Index and the American College of Surgeons National Surgical Quality Improvement Program scores. This new predictive tool is effective for the prediction of postoperative MACCEs in patients who undergo noncardiac surgery.