Limitations
The main limitations of the current study are the retrospective nature of the study, with most cases performed by a single surgeon, and the population sample not being random, although it did include all isolated primary CABG patients during the 2 study periods. In addition, the protocol was implemented as a unit, so we could not analyze which techniques were more successful than others. For example, several of the protocol measures, such as cessation of blood thinners and use of aminocaproic acid and cell saver, were part of our practice to varying degrees during the first study period because their efficacy in reducing transfusions had been previously proven. The purpose of the protocol was to reduce transfusion rates, with the understanding that some variables could be more impactful than others. The study did not test any of these variables independently. However, a major change during the second period was tolerance of anemia both intraoperatively and postoperatively. Commonly, during the first period (group A), the hematocrit “trigger point” for transfusion was 26% compared with 21% for the second period (group B).
Some may argue that this study was underpowered to demonstrate improved outcomes from a restrictive transfusion protocol. First, the low risk of blood transfusion-related complications would require a much larger sample to demonstrate any meaningful difference. Second, outcomes typically measured with coronary revascularization would not necessarily capture some of the potential long-term transfusion-related complications. Third, the purpose of this study was to demonstrate the feasibility and safety of implementing a transfusion reduction practice in one of the most commonly performed cardiac operations, not to prove or disprove the risks of blood transfusions already well documented in the literature. Finally, there is a fundamental flaw with such an argument. When subjecting a patient to a treatment or intervention, the burden of proof should be on the intervening rather than the non-intervening group. To our knowledge, there are no studies demonstrating improved outcomes with transfusions. Should patients undergo an operation or receive chemotherapy with no proven benefit? Blood products are a scarce resource,31,32 and exhausting them without a proven justification in an era of evidence-based medicine imposes a moral question. Should we transfuse a stable patient post-CABG merely for a predetermined hematocrit trigger point, while it could be life saving for another patient?
Two key issues are at hand. First is the realization and acceptance of the team caring for the patient that a blood transfusion does not lead to better outcomes and may actually be detrimental. Second, and perhaps more challenging, is taking ownership and interest in reducing blood use. This would require buy-in from all team members. At our facility, the primary caretaker is the surgeon, who must approve or deny all transfusion requests. This could be more challenging in other settings with multiple services (such as intensivists, residents, etc.), mandating strict adherence to a protocol agreed on by all.