Further Study Needed on the Effects of Blood Transfusion on Lower-Extremity Revascularization Outcomes


Submitted on Tue, 07/06/2010 - 10:35

Nicholas J. Morrissey, MD

In the current issue of Vascular Disease Management, the effects of blood transfusion on outcomes of lower-extremity revascularization are discussed in the paper by Eleftherios Xenos and colleagues. The paper uses the NSQIP database as a source for cases of lower-extremity revascularization for analysis. The authors use a preoperative propensity analysis to stratify patients for their risk of needing a transfusion. They then go on to assess the effect of blood transfusion on mortality and morbidity of lower-extremity bypass with vein and prosthetic grafts. Mortality and other major morbidity were increased in patients receiving intraoperative transfusion and the effect seemed more drastic with increasing amounts of transfusion. One of the interesting findings was that a return to the operating room was more common in patients receiving transfusion. In addition, pulmonary complications were much higher in patients receiving transfusions intraoperatively. The authors suggest that pulmonary complications such as transfusion-related lung injury may be directly related to the toxicity of blood transfusion. In addition, the higher rate of septic complications could be related to the immunosuppressive effects of banked blood. The findings in this paper certainly raise the possibility that blood transfusion alone could result in adverse outcomes in patients undergoing lower-extremity revascularization and suggests the need for more rigorous study. The challenges of proving these hypotheses with the NSQIP database and the limitations of this study are clear. While the authors attempted to stratify the patients preoperatively based on risk for transfusion, they are unable to provide clear data on intraoperative blood loss. Clearly, the adverse outcomes associated with higher transfusion rates could reflect higher rates of intraoperative blood loss. While the preoperative propensity scoring is clever and useful, it does not reflect true intraoperative blood loss, and therefore we must consider that some of the morbidity and mortality seen in these cases is due to intraoperative issues leading to high blood loss. Such complications as renal insufficiency, sepsis, mortality and return to the operating room can certainly be the result of excessive blood loss. Perhaps information on the total duration of an operation and the total intraoperative fluid requirements or estimated blood loss itself could provide more insight. While there are shortcomings based on the database, these results raise interesting concepts. It is important to consider that the potential immunosuppressive effects of transfused blood can adversely impact patients undergoing all types of surgery, and not just oncology patients. Importantly, it is time to once again make the discussion of the need for transfusion a critical one. Clearly, there are many different thresholds for initiating transfusion of banked blood, and this may result in excess use of this therapy. It is necessary to initiate a discussion between vascular interventionists, whose patients often have cardiac issues, and the anesthesia team, which must initiate transfusion in a responsible way based on good evidence. The paper presented provides interesting data which suggest the need for larger prospective studies that assess the effect of blood transfusion on outcomes in vascular patients. Prospective analysis with control for rates of blood loss will provide the clearest answers to these questions.

From the Division of Vascular Surgery, Columbia University College of Physicians and Surgeons, Weill Medical College of Cornell University, New York, New York. The author reports no conflicts of interest regarding the content herein. Address for correspondence: Nicholas J. Morrissey, MD, Assistant Professor of Surgery, Division of Vascular Surgery, Columbia University College of Physicians and Surgeons, Weill Medical College of Cornell University, New York, NY 10032. E-mail: njm2106@columbia.edu