¤ ¤Purpose: To examine the 4-year outcomes from Carotid Revascularization using Endarterectomy or Stenting Systems (CaRESS) in light of the current reimbursement guidelines for carotid artery stenting (CAS) from the Centers for Medicare and Medicaid Services (CMS).Methods: CaRESS was designed as a prospective, nonrandomized comparative cohort study of a broad-risk population of symptomatic and asymptomatic patients with carotid stenosis. In all, 397 patients (247 men; mean age 71 years, range 43-89) were enrolled and underwent carotid endarterectomy (CEA; n5254) or protected CAS (n5143). More than 90% of patients had .75% stenosis; two thirds were asymptomatic. The primary endpoints included (1) all-cause mortality, (2) any stroke, and (3) myocardial infarction (MI), as well as the composite endpoints of (4) death and any nonfatal stroke and (5) death, nonfatal stroke, and MI. The secondary endpoints were restenosis, repeat angiography, and carotid revascularization. All patients were classified with respect to surgical risk, symptom status, and stenosis grade based on criteria published by the CMS. In addition, separate analyses were performed comparing genders and octogenarians to those ,80 years old.Results: No significant differences in the primary outcome measures were found between the CEA and CAS groups in the 4-year analysis. The incidences of any stroke at 4 years were 9.6% for CEA and 8.6% for CAS (p50.444); when combined with death, the composite death/nonfatal stroke rates were 26.5% for CEA versus 21.8% for CAS (p50.361). The composite endpoint of death, nonfatal stroke, and MI at 4 years was 27.0% in CEA versus 21.7% in CAS (p50.273) patients. The secondary endpoints of restenosis (p50.014) and repeat angiography (p50.052) were higher in the CAS arm. There were no differences in any of the subgroups stratified according the CMS guidelines or in the gender comparison. Four-year incidences of death/nonfatal stroke and death/nonfatal stroke/MI were higher in the CEA arm among patients ,80 years of age (p50.049 and p50.030, respectively). There were no significant differences between these incidences in the octogenarian subgroup.