Cost Analysis of Transfusion Therapy in Coronary Artery Surgery

Background In patients undergoing coronary artery bypass grafting (CABG), wide variability in transfusion rate (7.8% to 92.8%) raises the question of the amount of unnecessary transfusions. The aim of the study was (1) to identify CABG patients at low risk of bleeding to whom transfusion treatment should be avoided and (2) to calculate the amount of possible cost savings that would be achieved by avoiding transfusion in low bleeding risk patients. Methods This retrospective observational study enrolled patients undergoing isolated elective CABG from January 2010 to January 2018. Patients were divided with respect to the presence of excessive bleeding and transfusion costs were compared between the two groups. Predictors for postoperative excessive bleeding were defined and multivariable logistic regression analysis and risk modeling were performed. The use of a model to predict patients at low risk of bleeding allowed for the estimation of transfusion cost savings assuming the patients who were found to be at low risk of bleeding should not be transfused. Results A total of 1,426 patients were enrolled in the analysis. Of those, 28.3% had excessive postoperative bleeding. The multivariate logistic regression analysis model was developed to identify/predict patients without excessive bleeding (receiver operating characteristic curve analysis, area under the curve 72.3%, p <0.001). When applied to the existing database, the use of the developed model identifying patients at low risk of bleeding may result in a 39.1% reduction of transfusions. Specifically, cost savings would be 48.2% for packed red blood cells, 38.9% for fresh frozen plasma, 10.9% for platelets concentrate, and 17.9% for fibrinogen concentrate. Conclusion The clinical and economic burdens associated with unnecessary transfusions are significant. Avoiding transfusion in CABG patients found to be at low risk of bleeding may result in significant reduction of transfusion rate and transfusion-associated costs.