The value of carotid intervention is predicated on long-term survival for patients to derive a stroke prevention benefit. Randomized trials report no significant difference in survival after carotid endarterectomy (CEA) vs carotid artery stenting (CAS), whereas observational studies of “real-world” outcomes note that CEA is associated with a survival advantage. Our objective was to examine long-term mortality after CEA vs CAS using a propensity-matched cohort.
We studied all patients who underwent CEA or CAS within the Vascular Quality Initiative from 2003 to 2013 (CEA, n = 29,235; CAS, n = 4415). Long-term mortality information was obtained by linking patients in the registry to their respective Medicare claims file. We assessed the long-term rate of mortality for CEA and CAS using Kaplan-Meier estimation. We assessed the crude, adjusted, and propensity-matched (total matched pairs, n = 4261) hazard ratio (HR) of mortality for CEA vs CAS using Cox regression.
The unadjusted Kaplan-Meier estimated 5-year mortality was 14.0% for CEA and 18.3% for CAS. The crude HR of all-cause mortality for CEA vs CAS was 0.75 (95% confidence interval [CI], 0.70-0.81), indicating that patients who underwent CEA were 25% less likely to die before those who underwent CAS. This survival advantage persisted after adjustment for age, sex, and comorbidities (adjusted HR, 0.75; 95% CI, 0.69-0.82). This effect was confirmed on a propensity-matched analysis, with an HR of 0.76 (95% CI, 0.69-0.85). Finally, these findings were robust to subanalyses that stratified patients by presenting symptoms and were more pronounced in symptomatic patients (adjusted HR, 0.69; 95% CI, 0.61-0.79) than in asymptomatic patients (adjusted HR, 0.80; 95% CI, 0.71-0.90).
During the last 15 years, patients who underwent CEA in the Vascular Quality Initiative have a long-term survival advantage over those who underwent CAS in real-world practice. Despite no difference in long-term survival in randomized trials, our observational study demonstrated a survival benefit for CEA that did not diminish with risk adjustment.