Background: Between January 1995 and November 1998, at Yale-New Haven Hospital, 25 percent of RBCs transfused were processed through prestorage or bedside leukoreduction filters, chosen on a per patient basis (selective leukoreduction [SLR]). Between January 1995 and July 1999, 30 percent of platelet concentrates (PCs) were infused through bedside leukoreduction filters. In an attempt to decrease febrile nonhemolytic transfusion reactions (FNHTR), a change was made from SLR to universal prestorage leukoreduction (UPL) for RBCs between November 1998 and December 1999 and for random donor PCs between July 1999 and January 2000. FNHTR and allergic transfusion reactions (ATR) reported from January 1995 through December 2002 were reviewed.
Study design and methods: For retrospective observational analysis, blood bank data were available on the number of RBCs and PCs transfused, percent products leukoreduced, and rate of FNHTR and ATR from 1995 through December 2002. After dividing this time period into three phases (SLR, transition, and UPL), these data were evaluated using odds ratio (ORs) and Student's t tests.
Results: A total of 145,369 RBCs and 137,982 PCs (29,487 PC pools) transfused between January 1995 and December 2002 were evaluated. For RBCs, the relative FNHTR rate decreased 47.1 percent, from 0.34 percent (SLR) to 0.18 percent (UPL) (p < 0.0001). ATR rates for RBCs showed 0.09 percent for both SLR and UPL groups (p > 0.05, NS). For PCs, the FNHTR relative rate decreased 93.1 percent, from 2.18 percent for SLR to 0.15 percent for UPL (p < 0.0001). Rates for ATR were 0.49 percent (SLR) and 0.35 percent (UPL) (p > 0.05, NS).
Conclusions: A significant decrease in the frequency of posttransfusion FNHTR, but not ATR, for RBCs and PCs followed introduction of 100-percent UPL. The data support the hypothesis that the practice of UPL of RBCs and PCs decreases the frequency of FNHTR and thus improves patient care over the practice of selective leukoreduction.