Regarding "The association of perioperative transfusion with 30-day morbidity and mortality in patients undergoing major vascular surgery"With great interest, we read the recent article by Obi and colleagues 1 assessing relationships of perioperative transfusion with short-term outcomes after major vascular surgery in a retrospective study. They showed that perioperative transfusion is independently associated with increased risks of postoperative 30-day morbidity and mortality. Strengths of this study include the large sample of patients from a multicenter statewide quality improvement registry and use of appropriate statistical analyses to determine relationships of perioperative transfusion with postoperative outcomes. In our view, however, two issues of this study should be clarified and discussed.First, health status, comorbidities, and operative risk are important determinants for postoperative morbidity and mortality. 2 In this study, patients receiving a transfusion had multiple baseline differences compared with those not receiving a transfusion; transfused patients were more likely to have preoperative anemia, cardiovascular diseases, diabetes mellitus, cerebrovascular accident, and renal failure requiring hemodialysis. Moreover, transfused patients were more likely to undergo open abdominal aortic aneurysm repair, open revascularization bypass, and urgent surgery. It suggests that transfused patients may have worse health status and heavier surgery and comorbidity burdens than those not receiving a transfusion. In our opinion, no matter how refined the adjustment is for differences in health status, surgery, and comorbidity burdens, it is never possible to ensure a complete adjustment for differences between patients with and without a transfusion. Furthermore, this study showed that independent predictors of perioperative transfusion were mainly the comorbidities and known factors significantly affecting postoperative morbidity and mortality, such as preoperative anemia, congestive heart failure, prior myocardial infarction, abdominal aortic aneurysm repair, open revascularization bypass, and urgent surgery. 3 This can further make it difficult to differentiate the independent role of perioperative transfusion on postoperative outcomes. Even sophisticated multivariable analysis is also probably inadequate to differentiate whether perioperative transfusion is a true determinant of postoperative outcomes or simply a synthetic manifestation of worse health status and heavier surgery burdens that can markedly increase postoperative morbidity and mortality. Although perioperative transfusion per se in this study is independently associated with increased postoperative morbidity and mortality, we do not wish to imply that restrictive perioperative transfusion can improve the short-term outcomes after major vascular surgery. As clinicians, we argue that perioperative transfusion should be tailored to requirements of the individual patient.Second, the authors did not provide the intraoperative blood loss and ...