Transfusion practice varies widely in cardiac surgery: results from a national registry

Objectives Evidence is accumulating of adverse outcomes associated with transfusion of blood components. If there are differences in perioperative transfusion rates in cardiac surgery, and what hospital factors may contribute, requires further investigation. Methods Analysis of 42,743 adult patients who underwent 43,482 procedures from 2005 to 2011 at 25 Australian hospitals, according to the Australian and New Zealand Society of Cardiac and Thoracic Surgeons Cardiac Surgery Database. Multiple logistic regression examined associations of patient and hospital characteristics with transfusion of ≥1 red blood cell (RBC) unit; platelet (PLT), fresh frozen plasma (FFP), and cryoprecipitate (CRYO) doses; and ≥5 RBC units, from surgery until hospital discharge. Results Procedures included 24,222 (55%) isolated coronary artery bypass grafts, 7299 (17%) isolated valve, 4714 (11%) coronary artery bypass graft and valve, and 7247 (17%) other procedures. After adjustment for various patient and procedure characteristics, transfusion rates varied across hospitals for ≥1 RBC unit from 22% to 67%, ≥5 RBC units from 5% to 25%, ≥1 PLT dose from 11% to 39%, ≥1 FFP dose from 11% to 48% and ≥1 CRYO dose from 1% to 20%. Hospital characteristics, including state or territory, private versus public, and teaching versus nonteaching, were not associated with variation in transfusion rates. Conclusions Variation in transfusion of all components and large volume RBC was identified, even after adjustment for patient and procedural factors known to influence transfusion, and this was not explained by hospital characteristics.