Background: Additional randomized controlled trials (RCTs) comparing recipients of non-white-blood-cell-(WBC)-reduced and WBC-reduced allogeneic red blood cells (RBCs) have been reported since the undertaking of previous meta-analyses of the association of allogeneic blood transfusion (ABT) with postoperative infection and/or mortality. Because no further RCTs are underway, a final meta-analysis of all available RCTs was conducted.
Methods: RCTs reporting on the association of ABT with postoperative infection and/or short-term (up to 3-month post-transfusion), all-cause mortality were retrieved. Twelve RCTs reporting on infection and 11 RCTs reporting on mortality were eligible for meta-analysis. Summary odds ratios (ORs) of infection or mortality in recipients of WBC-containing ABT vs. WBC-reduced ABT were calculated across the studies.
Results: An association of ABT with postoperative infection was demonstrated across RCTs transfusing RBCs WBC-reduced after storage [summary OR = 2.25; 95% confidence interval (CI), 1.12-4.25] but not before storage (summary OR = 1.06; 95% CI, 0.91-1.24). An association of ABT with mortality was demonstrated across RCTs conducted in cardiac surgery (summary OR = 1.72, 95% CI, 1.05-2.81) and across RCTs transfusing buffy-coat-reduced RBCs vs. RBCs WBC-reduced before storage (summary OR = 1.60; 95% CI, 1.14-2.24), but not across RCTs transfusing non-buffy-coat-reduced RBCs vs. RBCs WBC-reduced before storage (summary OR = 1.01; 95% CI, 0.73-1.40).
Conclusions: An association between ABT and postoperative infection or short-term mortality is not detected across all clinical settings and transfused RBC products. An association between ABT and mortality is detected in cardiac surgery, but the other associations found in subgroup analyses contradict current theories about mechanism(s) of the ABT effect.