Revised cardiac score index is a predictor of long-term outcomes after carotid endarterectomy

Abstract

Summary: Background: Cardiac complications represent the main cause of mortality after non-cardiac surgery and the Revised Cardiac Risk Index (RCRI) was created to estimate the perioperative risk of these events. It considers history of ischaemic heart disease, congestive heart failure, diabetes requiring preoperative insulin, stroke or transient ischaemic attack and renal impairment. We aim to describe the accuracy of the RCRI for predicting perioperative major adverse cardiovascular events (MACE) - a composite of heart failure, ischemic events and all-cause death. Also, the authors aimed to review the score for better prediction of cardiovascular outcomes. Patients and methods: From January 2012 to January 2020, patients who underwent Carotid endarterectomy (CEA) with regional anaesthesia (RA) were selected. RCRI was calculated for each case. Estimated and reported cardiovascular complications were compared using multivariate logistic regression and cox proportional hazards. An alternative and optimized carotid-RCRI (CtRCRI) was obtained. Overall predictive accuracy was assessed and compared by measuring model discrimination. Adjustments for overfitting and evaluation of the new model were performed by bootstrap. Results: 186 patients were selected, of which 80% were male with a mean age of 70.0±9.05 years old. The median follow-up was 50 months, interquartile range 21–69 months. None of the scores were able to predict MACE in the perioperative period. Both were associated with 30-day Clavien-Dindo ≥2 (p=0.022 and p=0.041, respectively). Regarding long-term prognosis, both were able to predict MACE (RCRI: hazard ratio (HR) 3.54 (95% confidence interval [CI] 1.04–11.48) vs. CtRCRI: HR 2.08 (95%CI 1.08–3.98) and all-cause mortality (RCRI: HR 3.33, 95%CI 0.99–11.11 vs. CtRCRI: HR 1.57, 95%CI 1.14–7.04). Conclusions: RCRI and CtRCRI did not predict MACE in the perioperative period but are good predictors of 30-day complications (Clavien-Dindo ≥2). Both RCRI and CtRCRI have good prognostic value as predictors of long-term cardiovascular events.