Clinical Experience amongst Surgeons in the Asymptomatic Carotid Surgery Trial-1

Abstract

INTRODUCTION: Hospital volume may influence the outcomes of carotid revascularization, but in trials the effect of the clinical experience of individual surgeons on procedural outcome is less certain. We assessed perioperative event rates amongst centers with different trial entry volumes and also the effects of individual operator experience in the first Asymptomatic Carotid Surgery Trial-1 (ACST-1). METHODS: In 126 centers participating in ACST-1, surgeons were classified according to their in-trial experience (group A: 50 cases; group B: 51-100 cases; group C: >100 cases), center enrolment volume (group I: <30 patients; group II: 30-75 patients; group III: >75 patients) and center annual hospital volume (group 1: <40 carotid endarterectomies (CEAs); group 2: 40-75 CEAs; group 3: >75 cases). Differences in perioperative event rates were compared using logistic regression analysis. RESULTS: In centers with the most clinical experience compared with those with least experience (groups C vs. A), the number of strokes or deaths was 8 of 275 (2.9%) versus 24 of 810 (3.0%) with OR 0.99 (95% CI 0.44-2.25, p = 0.986). Numbers of strokes or death in high enrolment centers compared with those in low enrolment centers (groups III vs. I) was 20 of 680 (2.9%) versus 21 of 580 (3.6%) with OR 0.81 (95% CI 0.43-1.51, p = 0.921). In centers with a high annual volume compared with those of low annual volume (groups 3 vs. 1), numbers of strokes and death were non-significantly lower, 26 of 823 (3.2%) versus 19 of 422 (4.5%) with OR 0.68 (95% CI 0.37-1.26, p = 0.386). Cumulative stroke risk at 5 and 10 years were similar among different levels of reported clinical experience, enrolment volume and annual hospital volume. CONCLUSION: Although our data did not demonstrate an association between perioperative complications and operators' experience, enrolment volume or annual hospital volume, rates of stroke or death were numerically lower in both high enrolment and high annual volume centers. This lack of association could be explained by an overall low procedural risk in ACST-1.