A dynamic mathematical model of red blood cell clinical demand to assess the impact of prolonged blood shortages and transfusion restriction policies. Issue 10 (3rd January 2014)

• Record Type: Journal Article
• Authors: McQuilten, Zoe K.; Mercer, Geoff; Phillips, Louise; Luangwilai, Thiansiri; Brown, Richard; Ozolins, Ieva; Cheng, Allen C.; Wood, Erica M.
• Is Part Of: Transfusion. Volume 54:Issue 10:Part 2(2014)
• Page Start: 2705
• Page End: 2715
• Publication Date: 2014-01-03
• Subjects: Hematology -- Periodicals; Blood -- Transfusion -- Periodicals; Blood Group Antigens -- Periodicals; Blood Preservation -- Periodicals; Blood Transfusion -- Periodicals; 615
• DOI: 10.1111/trf.12525
• Languages: English
• ISSNs: 0041-1132

Abstract

Abstract : Background: Estimating change in clinical demand for red blood cells (RBCs) from a disaster, as well as triaging introduced in response, is essential to plan effectively for a major blood shortage. We aimed to develop a RBC demand model to assess the impact of restriction policies on RBC use and patient outcomes. Study Design and Methods: A compartmental dynamic model was developed in which patients require RBCs acutely (within 1 hr), urgently (24 hr), semiurgently (1‐7 days), or nonurgently; outcomes included death or remaining at or transitioning to more or less urgent categories. A mathematical model was developed with transitions governed by differential equations and calibrated to a baseline scenario of adequate blood supply (using population‐based hospital data sets, registries, and RBC issues). Distribution into urgency categories was based on a prospective study of 5132 randomly selected RBC units. Scenarios when the blood supply is limited compared to baseline were investigated. Transition rates between urgency categories under these scenarios were established by clinician survey. Results: In the baseline 21‐day scenario, patients requiring the most RBCs were other surgery (2162, 22%), medical anemia (1916, 12%), malignant hematology (1092, 16%), and gastrointestinal hemorrhage (1115, 8%). A policy of withholding RBCs for all nonurgent indications results in an estimated reduction of only 1007 (11.2%) RBC units and, if extended to semiurgent, a reduction of 2567 (28.5%) RBC units. Conclusions: Based on this model, restrictions that withhold transfusion from nonurgent patients have minimal impact on RBC demand and may not be sufficient to address changed demand and/or decreased supply during a prolonged disaster.