In a 5-week prospective observational study in 47 Australian & New Zealand ICUs, Pettila et al (the ANZICS group) followed 757 ICU patients receiving red blood cell transfusions for anemia of any cause. Those in the quartile receiving the youngest RBCs (mean age 7.7 days) had an absolute 8% lower mortality than the other 3 quartiles (13% vs 21%, CI 2.2-14.0), who were transfused 23-day old RBCs, on average. This was before adjusting for any covariates. After adjusting for what they could, the odds ratio for death in the "freshest RBC" quartile was ~0.5 (this was presented as an OR of 2.0 in the 3 "old RBC" quartiles, which sounds scarier). Critical Care 2011;15:R116 FULL FREE TEXT
A 2009 review in Transfusion could not establish an association between age of transfused RBCs and survival, from 24 studies that were too heterogeneous to combine into a meta-analysis.