Association between Medicare high-risk criteria and outcomes after carotid revascularization procedures

Background: The U.S. Centers for Medicare and Medicaid Services (CMS) has defined a set of high-risk criteria to help define patients who are appropriate for carotid artery stenting (CAS), but these criteria have never been validated. We aimed to validate the CMS high-risk criteria in a nationally representative cohort of patients undergoing CAS and carotid endarterectomy (CEA). Methods: All patients undergoing CAS (with embolic protection) or CEA in the Vascular Quality Initiative (VQI) database (2013-2016) were included. Patients were stratified as being at normal risk (Nr) or high risk (Hr) for undergoing CEA on the basis of CMS criteria. Thirty-day and 2-year stroke outcomes were compared for CAS vs CEA in both the Nr and Hr groups using 1:1 coarsened exact matching and multivariable Cox proportional hazards modeling. Results: A total of 51,942 patients (CAS, 7030; CEA, 44,912) underwent carotid revascularization during the study period. Thirty-day (Nr, 1.7% vs 1.0%; Hr, 2.5% vs 1.4%) and 2-year (Nr, 1.9% vs 1.0%; Hr, 2.4% vs 1.3%) stroke occurred more frequently after CAS vs CEA on crude analysis (P <.001). After matching 2920 pairs of patients on 18 preoperative variables, the risk of 30-day and 2-year stroke remained higher after CAS in the Hr group (30-day risk: hazard ratio [HR], 1.90; 95% confidence interval [CI], 1.26-2.85; 2-year risk: HR, 1.65; 95% CI, 1.05-2.60) but was similar for CAS vs CEA in the Nr group (30-day risk: HR, 0.97; 95% CI, 0.48-1.95; 2-year risk: HR, 1.49; 95% CI, 0.76-2.90). Conclusions: These data suggest that the utility of CAS in Nr patients may be underappreciated, whereas the potential benefit of CAS in Hr patients may be overestimated. Re-evaluation of the criteria for identifying patients at high risk for CEA and the national guidelines on the indications for CAS is strongly indicated.