Algorithm‐based selection of automated red blood cell exchange procedure goals reduces blood utilization in chronically transfused adults with sickle cell disease

Background: Automated red cell exchange (RCE) is a common treatment for patients with sickle cell disease (SCD). Two key parameters are used to determine the volume of blood for RCE to reduce sickle hemoglobin (eg, HbS): fraction of cells remaining (FCR) and target hematocrit. We evaluated how the calculated FCR-using the manufacturer's algorithm-impacted blood utilization and incidence of acute care encounters. Study design and methods: Retrospective chart review was conducted of 15 adults with SCD who underwent chronic RCE from July 1, 2015 to August 31, 2019. Blood utilization and acute care encounters were compared across three time periods: (a) when a fixed FCR of 30% was used (12 months); (b) transition period during which physicians made ad hoc changes to the FCR (25 months); (c) algorithm phase when a procedural FCR between 30% and 50% was selected using an algorithm generated by the manufacturer's built-in software to target a HbS fraction of 8% post-procedure (12 months). Wilcoxon signed rank test was used to determine statistical significance. Results: Median blood utilization per procedure decreased from 2398 mL (interquartile range [IQR]: 2271-2759 mL) during the fixed FCR phase to 1887 mL (IQR: 1495-2241 mL) during the algorithm phase (P < 0.001). Similarly, median number of units transfused decreased from 10 (9-11) to 7 (5-9) during the respective phases (P < 0.001). Visits to the emergency department were 1 (0-4) in the fixed FCR phase and 0 (0-3) in the algorithm phase. Conclusion: Algorithm-based selection of a procedural FCR significantly reduced blood utilization (similar to 21%) without appearing to increase acute care encounters.