Regional variation in patient selection, practice patterns, and outcomes based on techniques for carotid artery revascularization in the Vascular Quality Initiative.

OBJECTIVE: Significant regional variation is known with multiple surgical procedures. This study describes regional variation in carotid revascularization within the Vascular Quality Initiative (VQI).

METHODS: Data from the VQI carotid endarterectomy (CEA) and carotid artery stenting (CAS) databases from 2016 to 2021 were used. Nineteen geographic VQI regions were divided into three tertiles based on the average annual volume of carotid procedures performed per region (low-volume: 956 cases [range, 144-1382]; medium-volume: 1533 cases [range, 1432-1589]; and high-volume: 1845 cases [range, 1642-2059]). Patients' characteristics, indications for carotid revascularization, practice patterns, and outcomes (perioperative and 1-year stroke/death) of different revascularization techniques were compared between these regional groups. Regression models that adjust for known risk factors and allow for random effects at the center level were used.

RESULTS: CEA was the most common revascularization procedure (>60%) across all regional groups. Significant regional variation was observed in the practice of CEA such as variability in the use of shunting, drain placement, stump pressure and electroencephalogram monitoring, intraoperative protamine, and patch angioplasty. For transfemoral CAS, high-volume regions had a higher proportion of asymptomatic patients with <80% stenosis (30.5% vs 27.8%) in addition to higher use of local/regional anesthesia (80.4% vs 76.2%), protamine (16.1% vs 11.8%), and completion angiography (81.6% vs 77.6%) during transfemoral carotid artery stenting (TF-CAS) compared with low-volume regions. For transcarotid artery revascularization (TCAR), high-volume regions were less likely to intervene on asymptomatic patients with <80% stenosis (32.2% vs 35.8%) than low-volume regions. They also had a higher proportion of urgent/emergent procedures (13.6% vs 10.4%) and were more likely to use general anesthesia (92.0% vs 82.1%), completion angiography (67.3% vs 63.0%), and poststent ballooning (48.4% vs 36.8%). For each carotid revascularization technique, no significant differences were noted in perioperative and 1-year outcomes between low-, medium-, and high-volume regions. Finally, there were no significant differences in outcomes between TCAR and CEA across the different regional groups. In all regional groups, TCAR was associated with a 40% reduction in perioperative and 1-year stroke/death compared with TF-CAS.

CONCLUSIONS: Despite significant variation in clinical practices for the management of carotid disease, no regional variation exists in the overall outcomes of carotid interventions. TCAR and CEA continue to show superior outcomes to TF-CAS across all VQI regional groups.