A Long-Term Survival Score Improves Preoperative Prediction of Survival Following Major Vascular Surgery

Abstract

In a previous study it has been shown that a long-term survival score (LTSS), composed of Lee's Revised Cardiac Risk Index (RCRI) criteria supplemented by age, preoperative electrocardiography (EKG) features, and all types of diabetes to the RCRI criteria, predicts long-term (3-15 years) survival after major vascular surgery. The present study aimed to investigate the performance of LTSS in predicting earlier survival (3 months-3 years) as compared with the RCRI. Data from 921 consecutive patients undergoing major vascular surgery (624 patients at Hadassah Medical Center [HMC] and 296 patients in Beth Israel Deaconess Medical Center [BIDMC]) were collected retrospectively. The LTSS was seven points that included the five RCRI factors as well as age >65 years and ST-segment depression on preoperative EKG. Logistic regression and receiver operating characteristic curve (ROC) curve analyses were used to compare the 3 months-3 years mortality between the RCRI and LTSS. The Beth Israel Deaconess Medical Center patients were sicker than the Hadassah Medical Center patients, with higher RCRI (1.2 ± 1.0 vs. 0.81 ± 0.83, p < 0.001) and LTSS (2.6 ± 1.4 vs. 1.7 ± 1.2, p < 0.001) and higher 3-years mortality (36.3% vs. 20.7%, p = 0.005). The LTSS predicted mortality better than RCRI as measured by the area under the ROC curves at all time points between 6 months (0.66 ± 0.03 vs. 0.57 ± 0.04, p = 0.02) and 3 years (0.70 ± 0.02 vs. 0.61 ± 0.02, p < 0.0001) in both institutions, but not 3-months mortality. The LTSS also provided better discrimination between each adjacent two-risk score than the RCRI. Age >65 years, ST-segment depression on preoperative 12-lead EKG, and all types of diabetes added to the RCRI significantly improved the preoperative prediction of mortality after 6 months following major vascular surgery.