Endocarditis is not an Independent Predictor of Blood Transfusion in Aortic Valve Replacement Patients with Severe Aortic Regurgitation

Objective This study sought to evaluate if the presence of endocarditis was independently associated with increased perioperative blood transfusion in patients undergoing aortic valve replacements (AVR) with aortic regurgitation. Design This was a retrospective study. Setting Large Canadian tertiary care hospital. Participants Six hundred sixty-two consecutive patients with aortic regurgitation score of 3 or higher undergoing AVR from 1995 to 2012. Interventions No interventions were performed in this retrospective study. Measurements and Main Results After REB approval, data were obtained from a center-specific database. Univariate analysis was performed to identify variables that may be associated with transfusion of any allogeneic blood product perioperatively. A multivariate logistic regression was generated to identify independent predictors of perioperative transfusion. Unadjusted transfusion rates in patients with no endocarditis and with endocarditis were 32% and 70% (p<0.001), respectively. Independent predictors of any transfusion were moderate-to-severe preoperative anemia, preoperative renal failure, non-isolated AVR, age>70, urgent/emergent surgery, BMI<25, and female sex. Endocarditis was not an independent predictor of transfusion (OR = 0.748; 95% CI = 0.35-1.601). Conclusions In patients undergoing AVR, unadjusted perioperative transfusion rates were higher when endocarditis was present. However, after adjustment, aortic valve endocarditis was not independently associated with blood transfusion. The authors' observation could be explained by the higher prevalence of many independent predictors of transfusion, such as comorbidities or more complex surgery, within the endocarditis group. Thus, AV endocarditis, in the absence of other risk factors, was not associated with increased perioperative transfusion risk.