Surgeon specialty significantly affects outcome of asymptomatic patients after carotid endarterectomy

Abstract Background This study evaluates the impact of surgical specialty, specifically vascular surgery (VS) versus non-VS (NVS; namely, cardiac surgery, thoracic surgery, general surgery, or neurosurgery) on perioperative carotid endarterectomy (CEA) outcomes stratified by symptom status on presentation. Methods The National Surgical Quality Improvement Program Vascular Procedure Targeted database was queried for elective asymptomatic or symptomatic CEA (excluding concomitant CEA and cardiac surgery) from 2011 to 2016. Data were stratified by VS versus NVS and symptom presentation. Primary end points were 30-day stroke and stroke/death; secondary end points included perioperative complications. Multivariable logistic regression determined predictors of all assessed primary outcomes and propensity-weight analysis was used to confirm results. Results Overall, 21,060 CEA (12,671 [59%] asymptomatic) were identified with 19,687 (93%) done by VS. In the asymptomatic CEA cohort, VS had lower unadjusted stroke (1.3% vs 2.4%; P = .021) and stroke/death (1.7% vs 3.2%; P = .006) rates. In addition, VS had fewer deaths (0.6% vs 1.3%; P = .033) and pulmonary complications (1.6% vs 2.7%; P = .036). After risk adjustment, the NVS asymptomatic cohort predicted stroke (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.1-3.1; P = .032), driven by neurosurgery (OR, 3.1; 95% CI, 1.3-7.2; P = .008). This NVS cohort also predicted stroke/death (OR, 1.8; 95% CI, 1.1-2.9; P = .013), driven by neurosurgery (OR, 2.5; 95% CI, 1.1-5.7; P = .035). After propensity weighting, these differences persisted (stroke: OR, 1.9; 95% CI, 1.1-3.3; P = .030; stroke/death: OR, 1.9; 95% CI, 1.2-3.0; P = .011). Among symptomatic CEA, there was no difference between VS and NVS in unadjusted primary end points of stroke (3.1% vs 4.2%; P = .106) or stroke/death (3.8% vs 4.6%; P = .275). However, in this cohort, VS had fewer major complications (12.7% vs 15.5%; P = .029). Conclusions This study identifies the VS specialty as having significantly better outcomes after CEA in patients presenting with asymptomatic disease than NVS specialty, as evidenced by lower rates of stroke and stroke death, which persisted after risk adjustment and propensity weighting. This difference in stroke and stroke/death was not apparent in the symptomatic cohort; however, NVS did have increased unadjusted rates of major complications. Although this finding may reflect multiple factors, including higher operative volume, training, or technical approach, these differences in 30-day CEA outcomes may be crucial for the proper interpretation of ongoing national outcome trials such as CREST2.