Examination of the interaction between method of anesthesia and shunting with carotid endarterectomy

Background: Although the choice of anesthesia during carotid endarterectomy (CEA) does not seem to increase the risk of perioperative stroke, it might affect the outcomes of shunting during CEA. This study aims to evaluate whether the choice of anesthesia modifies the association between shunting and in-hospital stroke/death after CEA. Methods: We retrospective reviewed all CEA cases performed between 2003 and 2017 in the Vascular Quality Initiative. Patients were divided into three groups: (1) no shunting during CEA (n = 29,227 [48.4%]), (2) routine shunting (n = 28,673 [47.5%]), and (3) selective shunting based on an intraoperative indication (n = 2499 [4.1%]). Multivariable logistic regression analysis was used to study the interaction between anesthesia (local anesthesia [LA]/regional anesthesia [RA] vs general anesthesia [GA]) and intraoperative shunting (no shunting vs routine and selective shunting) during CEA in predicting the risk of in-hospital stroke/death after CEA. Results: The final cohort included 60,399 patients. The majority of CEA cases (90.2%) were performed under GA. Of the study cohort, 29,227 (48.4%) underwent CEA without shunting, 28,673 patients (47.5%) had routine shunting, and the remaining (n = 2499 [4.1%]) were selectively shunted. The interaction between intraoperative shunting and anesthesia in predicting in-hospital stroke/death was statistically significant (P < .05). When CEA is performed under LA/GA, routine shunting was associated with 3.5 times the adjusted odds of in-hospital stroke/death after CEA (odds ratio [OR], 3.5; 95% confidence interval [CI], 1.8-6.8; P < .001) compared with no shunting, whereas selective shunting was associated with 7.1 the odds (OR, 7.1; 95% CI, 3.5-14.7; P < .001). In contrast, under GA, there was no significant association between routine shunting and in-hospital stroke/death (OR, 1.2; 95% CI, 1.0-1.5; P = .12), whereas selective shunting was associated with 1.7 times the odds (OR, 1.7; 95% CI, 1.2-2.4; P < .01) compared with not performing shunting during CEA. Conclusions: The use of LA/RA is associated with increased odds of stroke/death compared with GA when intraoperative shunting is performed. The effect of anesthesia is more pronounced in patients who develop clamp-related ischemia and undergo selective shunting. More controlled studies are needed to explain these findings and validate them.