Preoperative anemia is associated with mortality after carotid endarterectomy in symptomatic patients

Preoperative anemia and blood transfusions are associated with worse outcomes after surgery. However, the impact of preoperative anemia and transfusions on outcomes after carotid endarterectomy (CEA) is unknown.

CEA patients from 2011 to 2015 in the American College of Surgeons National Surgical Quality Improvement Program Targeted Vascular module were compared by the presence of preoperative anemia (hematocrit <36%) after stratification by symptom status. Multivariable analysis accounted for differences in baseline characteristics. We included an interaction term in our multivariable model to assess whether the effect of anemia differed significantly between patients who received a perioperative transfusion and those who did not, with 30-day mortality as our primary outcome.

Of 16,068 patients, 6734 (42%) were symptomatic, of whom 1500 (22%) had anemia. Of the 9334 asymptomatic patients, 1935 (21%) had anemia. Both symptomatic and asymptomatic anemic patients were more likely to be transfused perioperatively compared with nonanemic patients, with 7.0% vs 0.4%, and 5.8% vs 0.7% (both P < .001). Among symptomatic patients, those with anemia compared with those without had a higher rate of 30-day mortality (2.5% vs 0.7%; P < .001). After adjustment, anemic symptomatic patients had a higher 30-day mortality risk (odds ratio [OR], 3.1; 95% confidence interval [CI], 1.9-5.0; P < .001) compared with nonanemic symptomatic patients. In addition, in symptomatic patients, we found a significant interaction between anemia and perioperative transfusion on the outcome of 30-day mortality (P = .004), with a higher risk in perioperatively transfused symptomatic patients with anemia (OR, 7.8; 95% CI, 3.4-18.0; P < .001) than in symptomatic patients with anemia who did not receive a perioperative transfusion (OR, 2.3; 95% CI, 1.4-3.9; P = .002). In asymptomatic patients, anemic and nonanemic patients had comparable 30-day mortality rates (0.9% vs 0.6%; P = .2). After adjustment, anemia was not associated with 30-day mortality in asymptomatic patients (OR, 1.0; 95% CI, 0.5-2.0; P = .9), nor did we identify an interaction between anemia and perioperative transfusion in asymptomatic patients (P = .1). Patients who received a preoperative transfusion had a higher 30-day mortality rate than anemic patients not receiving preoperative transfusion in both symptomatic (n = 31, 9.7% vs 2.5%; P = .04) and asymptomatic patients (n = 21, 9.5% vs 0.9%; P = .02).

Preoperative anemia is a risk factor for 30-day mortality after CEA in symptomatic patients but not in asymptomatic patients. These results should be factored into the selection of symptomatic patients for CEA and dissuade treatment of asymptomatic patients scheduled for CEA who need a preoperative transfusion.