PREDICTION OF CARDIOVASCULAR RISK IN ABDOMINAL SURGERY: RESULTS OF AN OBSERVATIONAL MULTICENTER STUDY

Abstract

HighlightsThe developed prediction model is a simple and accurate method for assessing the postoperative risk of cardiovascular complications in a large population of patients undergoing non-cardiac surgery. Its widespread use will optimize perioperative management tactics and improve surgical results. Aim. To develop a model for predicting cardiovascular risk in patients undergoing abdominal surgery and compare its accuracy with foreign analogs validated on a domestic cohort of patients.Methods. The multicenter prospective study included 8 241 patients over 18 years of age undergoing elective abdominal surgery. The following postoperative complications were assessed: acute myocardial infarction, stroke, cardiac arrest, cardiogenic pulmonary edema, pulmonary embolism, and 30-day mortality. First of all, we compared baseline characteristics of patients with and without complications, and factors, associated with surgery and anesthesia. Next, we performed a logistic regression analysis to assess the contribution of factors to the development of postoperative cardiovascular complications. Following that, we developed a model for predicting postoperative cardiac risk based on the data of multivariate logistic regression analysis. Finally, we compared the obtained model with other prediction models found in the literature.Results. Out of 8241 patients, 53 patients (0.64%) presented with cardiovascular complications (62 cases): cardiac arrest in 39 patients (0.47%), cardiogenic pulmonary edema in 4 patients (0.049%), stroke in 3 patients (0.036%), pulmonary embolism in 9 patients (0.11%), and acute myocardial infarction in 7 patients (0.084%). 36 patients (0.43%) had lethal outcome. Retrospectively, the obtained model assigned 2251 patients to the high-risk group for developing cardiovascular complications, the incidence of cardiovascular complications in the group was 2.1%. The low-risk group consisted of 5 990 patients; the incidence of cardiovascular complications in the group was 0.13%.Conclusion. Eight independent variables associated with postoperative cardiovascular complications were identified: high and moderate surgical trauma, smoking, statin use, Stange test less than 40 seconds, American Society of Anesthesiologists functional class 3, intraoperative need for vasopressors and transfusions. The cardiovascular risk prediction model has good predictive power (AUROC = 0.880).