Implementation of an algorithm for optimized multisite distribution of red blood cell concentrates: an efficacy analysis: Improving multisite RBC distribution through algorithms

Background - Managing red blood cell (RBC) inventory across multi-site transfusion networks is challenging, often resulting in expiry wastage or costly external compensation. We report the development and evaluation of a spreadsheet-based decision-support tool implemented in the Pavia Transfusion Medicine and Hematology Department (DMTE), Lombardy, Italy. Materials and methods - A Microsoft Excel workbook with visual basic for applications (VBA) macros implemented tiered inventory thresholds (safety stock, critical level, and maximum stock) derived from 2022-2024 transfusion data. Automated algorithms redistributed RBC units according to shortages and surpluses, with dedicated rules for Rh-negative units with C/E/Kell antigens and AB-group units. Expiry-wastage rates before (January-June 2025) and after implementation (July-December 2025) were compared using Fisher’s exact test, multivariable logistic regression, Mantel-Haenszel analysis, and difference-in-differences (DiD) analysis against 2024. Results - Over 100,000 RBC unit movements were analyzed. Expiry wastage decreased from 1.30% to 1.00% (RR 0.77, 95% CI 0.65-0.90; p=0.002), corresponding to a relative reduction of 23.3%. Multivariable analysis confirmed the protective effect (OR 0.72, 95% CI 0.61-0.86; p<0.001). However, benefits were confined to Rh(D)-positive units; wastage among Rh(D)-negative groups increased from 1.59% to 2.27% (RR 1.43, 95% CI 1.03-1.99; p=0.034). DiD analysis showed a net reduction of 0.50 percentage points. Improvements were greater in lower-complexity facilities, whereas high-complexity centers showed nonsignificant increases, consistent with their inherently rapid RBC turnover and their role as recipients of last-resort redistribution. Discussion - This low-cost digital tool reduced overall RBC expiry wastage while maintaining departmental self-sufficiency in a multi-site Italian transfusion network. However, its effectiveness was not uniform across blood groups. Rh(D)-negative inventory requires dedicated safety-stock calibration and mandatory recirculation of near-expiry units. Algorithm-driven inventory management may improve transfusion network efficiency, provided that rare phenotypes are governed by tailored rules. Future developments should incorporate predictive analytics and integration with transfusion management systems.