Will clinical studies elucidate the connection between the length of storage of transfused red blood cells and clinical outcomes? An analysis based on the simulation of randomized controlled trials

Authors


Arturo Pereira, Service of Hemotherapy and Hemostasis, Hospital Clínic, Villarroel 170, 08036 Barcelona, Spain; e-mail: apereira@clinic.ub.es.

Abstract

BACKGROUND: The temporal pattern of the biologic mechanism linking red blood cell (RBC) storage duration with clinical outcomes is yet unknown. This study investigates how such a temporal pattern can affect the power of randomized controlled trials (RCT) to detect a relevant clinical outcome mediated by the transfusion of stored RBCs.

STUDY DESIGN AND METHODS: This study was a computer simulation of four RCTs, each using a specific categorization of the RBC storage time. The trial's endpoint was evaluated assuming five hypothetical temporal patterns for the biologic mechanism linking RBC storage duration with clinical outcomes.

RESULTS: Power of RCTs to unveil a significant association between RBC storage duration and clinical outcomes was critically dependent on a complex interaction among three factors: 1) the way the RBC storage time is categorized in the trial design, 2) the temporal pattern assumed for the RBC storage lesion, and 3) the age distribution of RBCs in the inventory from which they are picked up for transfusion. For most combinations of these factors, the power of RCTs to detect a significant treatment effect was below 80%. All the four simulated RCTs had a very low power to disclose a harmful clinical effect confined to last week of the maximum 42-day shelf life of stored RBCs.

CONCLUSIONS: Ongoing RCTs may lack enough power to settle the issue of whether or not the transfusion of stored blood has a negative clinical impact. A precautionary reduction of the maximum storage time to 35 days is advisable.

Ancillary