The use of blood and blood products in aortic surgery is associated with adverse outcomes.

Abstract

To report long-term outcomes after deep hypothermic circulatory arrest (DHCA) with or without perioperative blood or blood products. All patients who underwent proximal aortic surgery with DHCA from 2011 to 2018 were propensity matched according to baseline characteristics. Primary outcomes included short- and long-term mortality. Stratified Cox regression analysis was performed for significant associations with survival. A total of 824 patients underwent aortic replacement requiring circulatory arrest. After matching, there were 224 patients in each arm (transfusion and no transfusion). All baseline characteristics were well matched, with a standardized mean difference (SMD) <0.1. Preoperative hematocrit (41.0 vs 40.6; SMD = 0.05) and ejection fraction (57.5% vs 57.0%; SMD = 0.08) were similar between the no transfusion and blood product transfusion cohorts. Rate of aortic dissection (42.9% vs 45.1%; SMD = 0.05), hemiarch replacement (70.1% vs 70.1%; SMD = 0.00), and total arch replacement (21.9% vs 23.2%; SMD = 0.03) were not statistically different. Cardiopulmonary bypass and cross-clamp time were higher in the blood product transfusion cohort (P < .001). Operative mortality (9.4% vs 2.7%; P = .003), stroke (7.6% vs 1.3%; P = .001), reoperation rate, pneumonia, prolonged ventilation, and dialysis requirements were significantly higher in the transfusion cohort (P < .001). In stratified Cox regression, transfusion was an independent predictor of mortality (hazard ratio, 2.62 [confidence interval, 1.47-4.67]; P = .001). One- and 5-year survival were significantly reduced for the transfusion cohort (P < .001). In patients who underwent aortic surgery with DHCA, perioperative transfusions were associated with poor outcomes despite matching for preoperative baseline characteristics.