The influence of cardiopulmonary bypass residual volume processing technique on blood management in cardiac surgical patients

<i>Background:</i> Post-cardiopulmonary bypass (CPB) blood processing is an important component of blood management during cardiac surgery. <i>Purpose:</i> The purpose of this study is to evaluate several methods of processing post-CPB residual blood.<i>Research Design:</i> Using a multi-institutional national database (SpecialtyCare Operative Procedural rEgistry [SCOPE]), 77,591 cardiac surgical operations performed in adults (>18 years) between January 2017 and September 2022 were reviewed.<i>Study Sample:</i> Blood processing methods included: Cell washing (CW, <i>n</i> = 63,592), Ultrafiltration (UF, <i>n</i> = 6286), Whole blood (WB, <i>n</i> = 3749), Hemobag (HB, <i>n</i> = 2480), and No processing (NO, <i>n</i> = 1484). The primary outcome was intraoperative post-CPB allogenic red blood cell (RBC) transfusion.<i>Data Analysis:</i> Group differences in RBC transfusion were assessed using a Bayesian mixed-effects logistic regression model controlling for multiple operative variables.<i>Results:</i> Across blood processing groups, patients had similar ages, body mass index and surgical procedures performed as well as preoperative hematocrit and nadir operative hematocrit. Median hematocrit change from last-in-operating room to first-in-ICU were highest in UF and HB groups (3.0 [IQR = 2.0-4.8] and 2.5 [IQR = 0.4-5.0]), respectively. The model-predicted probability of intraoperative post-CPB RBC transfusion was lowest in the HB group (0.79% [95% CrI = 0.37%-1.26%]), and highest in NO group (2.12% [95% CrI = 1.47%-2.82%]). Relative to CW, the odds of RBC transfusion for HB cases were reduced by half (OR = 0.5 [95% CrI = 0.28-0.89], statistical reliability = 99.1%), while odds for NO were 1.41 greater (OR = 1.41 [95% CrI = 1.03-1.93], statistical reliability = 98.2%).<i>Conclusions:</i> Post-CPB blood processing affects the likelihood for both receiving an intraoperative post-CPB RBC transfusion and for hematocrit change, with HB use resulting in the lowest predicted risk for transfusion, and NO the highest.