Abstract
An increased risk of postoperative infection (including pneumonia) attributable to the receipt of allogeneic blood transfusion has been investigated as a possible manifestation of transfusion-related immunomodulation (TRIM) in 16 randomized controlled trials (RCTs) and approximately 40 observational studies. This review categorizes RCTs and observational studies with regard to the inference that they permit about possible mediators of TRIM-allogeneic white cells (WBCs), WBC-derived soluble mediators, and/or allogeneic plasma-and examines whether the totality of the clinical evidence supports an association between allogeneic blood transfusion and postoperative infection. When all available studies are considered together in meta-analyses, three types of studies show no increased risk of postoperative infection in association with allogeneic blood transfusion: a) RCTs comparing recipients of buffy-coat-reduced and prestorage-filtered, WBC-reduced allogeneic red cells; b) RCTs comparing recipients of allogeneic and autologous blood; and c) observational studies comparing patients transfused before and after implementation of WBC reduction. RCTs comparing recipients of nonbuffy-coat-reduced and WBC-reduced red blood cells may point to a TRIM effect, but they cannot yet be subjected to formal meta-analysis. No overwhelming clinical evidence has been presented to establish the existence of a TRIM effect that relates allogeneic blood transfusion to postoperative infection.