Key result
TCAR shows no difference in stroke or death risk compared to CEA in symptomatic stenosis.
Why the study?
Quality benchmarks for symptomatic carotid stenosis rely on NASCET evidence, which antedates significant advances in optimal medical therapy and surgical interventions.
What are the contemporary trends and outcomes of symptomatic carotid revascularization procedures (CEA and TCAR) in real-world practice?
Observational (n=53,891)
Yes
What are the contemporary trends and outcomes of symptomatic carotid revascularization procedures (CEA and TCAR) in real-world practice?
p-value: p=.23
Contemporary real-world outcomes for symptomatic carotid revascularization show a composite stroke/death risk of 3.2%, well below the historical 6.5% benchmark, suggesting a need to update quality standards.
Improved stroke/death rates below 6.5% benchmark may warrant updated thresholds; Level 5 data leaves open whether to revise guidelines.
Objective Contemporary quality benchmarks for symptomatic carotid stenosis remain predicated on foundational level I evidence from the North American Symptomatic Carotid Endarterectomy Trial. However, this important study antedates significant advances in both optimal medical therapy and surgical interventions since its initial publication. Therefore, this study aimed to document trends in symptomatic carotid revascularization procedures and associated outcomes. Methods We identified all patients with symptomatic carotid stenosis undergoing either carotid endarterectomy (CEA) or transcarotid artery revascularization (TCAR) at centers with continuous participation in the Vascular Quality Initiative registry from January 2016 to July 2023. We stratified patients by age (<70, 70-80, and >80 years old, and assessed trends in procedure use and in-hospital stroke and/or 30-day death rates, by year. We used linear P trend tests to measure changes in procedure use and stroke/death risk over time. Results Among 53,891 patients treated at 211 centers, 17% underwent TCAR (n = 9218) and 83% underwent CEA (n = 44,673). The total number of procedures increased modestly from 6447 in 2016 to 6853 in 2022 ( P trend < .01). TCAR use increased substantially during the study period from 0.5% to 31.0%, with a corresponding decrease in CEA from 95.5% to 69.0% ( P trend < .01). The overall in-hospital stroke/30-day death risk was 3.2%, decreasing from 3.4% in 2016 to 2.5% in 2023 ( P trend < .01), with no significant difference between TCAR and CEA ( P = .23). This trend was largely driven by in-hospital stroke, decreasing from 2.7% to 2.0% in that same period ( P trend = .01); perioperative death remained stable at approximately 1.0% ( P trend = .13). Stroke/death risk was highest among patients >80 years old (3.1%), compared with 2.7% in those aged 70 to 80 and 2.0% in patients <70 years old (P < . 01). Both stroke (<70 years, 1.6% vs 70-80 years, 1.9% vs >80 years, 2.3%) and perioperative death (<70 years, 0.6% vs 70-80 years, 0.8% vs >80 years, 1.2%) risks independently increased with advancing age (all P < .01). Conclusions Although the overall volume of symptomatic carotid interventions has remained stable over time, it appears that the composite stroke/death risk has improved across all age strata and remains well below the societal benchmark of 6.5%. Based on these findings, it seems appropriate to update current quality benchmarks for symptomatic carotid procedures to better reflect concordant improvements in care delivery.
No takes yet. Share an insight, caveat, or question.
John et al. (2026) conducted an observational in Symptomatic carotid stenosis (n=53,891). Transcarotid artery revascularization (TCAR) vs. Carotid endarterectomy (CEA) was evaluated on In-hospital stroke and/or 30-day death (p=.23). In patients with symptomatic carotid stenosis, the in-hospital stroke or 30-day death risk decreased from 3.4% in 2016 to 2.5% in 2023, with no significant difference between TCAR and CEA (P=0.23).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: