N this month's Journal, Fergusson and colleagues describe a meta-analysis of clinical trials that evaluated the efficacy of leukoreduction of red blood cell (RBC) transfusion. 1 The authors found that when all patients in the randomized trials were included in the meta-analysis, statistically significant benefits were not demonstrated. However, if only patients that actually received a transfusion were included in the analysis, a 40% reduction in postoperative infections was demonstrated. This report raises a number of questions. These include: 1) Do the results presented in this report add convincing evidence that leads to the conclusion that leukoreduction improves clinical outcomes? 2) Is it methodologically proper to include only patients who received transfusion in the analysis? We begin this commentary with some background information on leukoreduction and its rationale.Universal leukoreduction is performed in Canada and many European countries. It remains voluntary in the United States. Leukoreduction involves the removal of white blood cells by use of filtration and may be performed either at the time of the preparation of RBC (prestorage) or concurrent with administration of (poststorage) RBC. There is a concomitant loss of red cells from 4 to 19% associated with filtration. 2,3 During blood storage, white blood cells produce cytokines, which are involved with intracellular communication and influence development and maintenance of immunity and the inflammatory processes. Because these cytokines may interfere with immune function, 4-7 in theory, prestorage should be superior to poststorage leukoreduction, although this has not been confirmed in clinical trials.Allogeneic transfusion has been thought to alter immune function for many years. Allogeneic blood transfusion has been demonstrated to promote tumour growth in experimental animals. 8,9 Renal allograft survival was better in patients receiving allogeneic blood transfusion prior to transplant 3,10 which is consistent with a reduction in immune function. The risk of recurrent cancer has not been demonstrated in randomized control trials. 11 However, it remains very controversial whether allogeneic blood transfusion increases the risk of bacterial infection.Ten randomized clinical trials have evaluated the effects of leukoreduction. 1 Scanning the results of clinical trials displayed in the table and figures in the article by Fergusson et al. reveals significant heterogeneity in the design and results of the studies. The majority of studies were performed in patients undergoing colorectal or cardiac surgery and used prestorage leukodepletion. Buffy-coated depleted blood was the standard RBC preparation used in many studies, which may make it more difficult to demonstrate a difference, since about 75% of leukocytes are removed by this process. Most of the individual studies did not demonstrate an improvement in outcome. In two of the three studies that showed differences, the observed effects were much larger than seems biologically plausible. 12,13 O...