Risk of emergent carotid endarterectomy varies by type of presenting symptoms

Background: The timing of carotid revascularization in symptomatic patients is a matter of ongoing debate. Current evidence indicates that carotid endarterectomy (CEA) within 2 weeks of symptoms is superior to delayed treatment. However, there is little evidence on the outcomes of emergent CEA (eCEA). The purpose of this study was to compare outcomes of emergency eCEA vs nonemergent CEA (non-eCEA), stratified by type of presenting symptoms. Methods: We analyzed the Vascular Targeted-National Surgical Quality Improvement Program dataset from 2011 to 2016. Symptomatic patients were divided into two groups: eCEA and non-eCEA. Univariable and multivariable methods were used to compare patient characteristics and to evaluate stroke, death, myocardial infarction (MI), stroke/death, and stroke/death/MI within 30 days of surgery adjusting for all potential confounders. A further subgroup analysis was done to compare the outcomes of eCEA vs non-eCEA stratified by the type of presenting symptoms (amaurosis, transient ischemic attack [TIA], and stroke). Results: A total of 9271 patients were identified, of which 10.7% were eCEA vs 89.3% non-eCEA. Comparing eCEA vs non-eCEA, the two groups were similar in age (70.8 vs 70.5), female gender (36.3% vs 36.9%), diabetes (26.2% vs 28.9%), and smoking status (31.9% vs 28.7%; all P > .05). Patients undergoing eCEA were less likely to be hypertensive (76.2% vs 80.2%; P = .025), but more likely to belong to non-white race (51.5% vs 20.5%; P < .001). The eCEA patients were less likely to be on preprocedural medication vs non-eCEA (antiplatelets, 76.8% vs 89.2%; statins, 74.2% vs 79.9%; beta-blockers, 44.6% vs 50.4%; all P < .05). The 30-day outcomes comparing eCEA vs non-eCEA were: stroke, 6.2% vs 3.1%; death, 2% vs 1%; and stroke/death, 6.9% vs 3.7% (all P < .05). After risk adjustment, perioperative stroke (odds ratio [OR], 2.04; 95% confidence interval [CI], 1.36-3.0), stroke/death (OR, 1.66; 95% CI, 1.13-2.45), and stroke/death/MI (OR, 1.58; 95% CI, 1.18-2.23) were higher after eCEA (all P < .01). When stratified by the type of presenting symptom, eCEA vs non-eCEA stroke outcomes were similar in patients who presented with stroke or amaurosis fugax. However, in the subset of patients presenting with TIA, eCEA had much worse outcomes compared with non-eCEA (stroke, 8.3% vs 2.5%; stroke/death, 8.3% vs 3.2%) and had significantly higher odds of stroke (OR, 3.12; 95% CI, 1.71-5.68) and stroke/death (OR, 2.24; 95% CI, 1.25-4.03) in the adjusted analysis (all P < .05). Conclusions: In patients presenting with stroke, eCEA does not seem to add significant risk compared with non-eCEA. However, patients presenting with TIA might be better served with non-emergent surgery as their risk of stroke is tripled when CEA is performed emergently.