Transfusion of fresher versus older red blood cells for all conditions

Background Red blood cell transfusion is a common treatment for anaemia in many clinical conditions. One current concern is uncertainty as to the clinical consequences (notably efficacy and safety) of transfusing red blood cell units that have been stored for different durations of time before a transfusion. If evidence from randomised controlled trials were to indicate that clinical outcomes are affected by storage age, the implications for inventory management and clinical practice would be significant. Objectives To assess the effects of using fresher versus older red blood cells in people requiring a red blood cell transfusion. Search methods We ran the search on 29th September 2014. We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (OvidSP), Embase (OvidSP), CINAHL (EBSCO), PubMed (for e-publications), three other databases and trial registers. Selection criteria We included randomised controlled trials comparing fresher red blood cell transfusion versus active transfusion of older red blood cells, and comparing fresher red blood cell transfusion versus current standard practice. All definitions of 'fresher' and ' older'/'standard practice' red blood cells were included. Data collection and analysis Two review authors independently assessed trial quality and extracted from the trial report data on adverse red blood cell transfusion reactions, when reported. Main results We included 16 trials (1864 participants) in the review. Eight trials (279 participants) compared transfusion of fresher red blood cells versus transfusion of older stored red blood cells ('fresher' vs 'older'). Eight trials (1585 participants) compared the transfusion of fresher red blood cells versus current standard practice ('fresher' vs 'standard practice'). Five trials enrolled neonates, one trial enrolled children and 12 trials enrolled adults. Overall sample sizes were small: only two trials randomly assigned more than 100 participants. We performed no meta-analyses for a variety of reasons: no uniform definition of 'fresher' or 'older' red blood cell storage; overlap in the distribution of the age of red blood cells; and heterogeneity in measurements and reporting of outcomes of interest to this review. We tabulated and reported results by individual trial. Overall risk of bias was low or unclear, with four incidences of high risk of bias: in allocation concealment (three trials) and in incomplete outcome data (one trial). No trial measured all of the outcomes of interest in this review. Four trials comparing 'fresher' with 'older' red blood cells reported the primary outcome: mortality within seven days (one study; 74 participants) and at 30 days (three trials; 62 participants). Six trials comparing 'fresher' with 'standard practice' red blood cells reported the primary outcome: mortality within seven days (three studies; 159 participants) and at 30 days (three trials; 1018 participants). All 10 trials reported no clear differences in mortality at either time point between intervention arms. Three trials comparing 'fresher' with 'standard practice' red blood cells reported red blood cell transfusion-associated adverse events. No adverse reactions were reported in two trials, and one incidence of cytomegalovirus (CMV) infection was described in the 'standard practice' arm in one trial. Overall the trials reported no clear difference between either of the intervention comparisons in long-term mortality (three trials; 478 participants); clinically accepted measures of multiple organ dysfunction (two trials: 399 participants); incidence of in-hospital infection (two trials; 429 participants); duration of mechanical ventilation (three trials: 95 participants); and number of participants requiring respiratory organ support (five trials; 528 participants) or renal support (one trial; 57 participants). The outcome 'physiological markers of oxygen consumption or alterations in microcirculation' was reported by 11 studies, but the measures used were highly varied, and no formal statistical analysis was undertaken. Authors' conclusions Several factors precluded firm conclusions about the clinical outcomes of transfusing red blood cell units that have been stored for different periods of time before transfusion, including differences in clinical population and setting, diversity in the interventions used, methodological limitations and differences in how outcomes were measured and reported. No clear differences in the primary outcome -death-were noted between 'fresher' and 'older' or 'standard practice' red blood cells in trials that reported this outcome. Findings of a large number of ongoing trials will be incorporated into this review when they are published. Updates of this review will explore the degree of overlap in trials between 'fresher', 'older' and 'standard practice' storage ages of red blood cells and will consider whether the size of any observed effects is dependent on recipient factors such as clinical background, patient age or clinical presentation.