The University of Texas MD Anderson Cancer Center Houston, Texas, United States
Background/Case Studies:
Background: Daily blood inventory management became a priority during the COVID-19 pandemic as demand for blood far exceeded the availability due to decreased collections. The latest Blood Inventory dashboard or all blood types is updated twice a day. Intent: For key stakeholders to be aware of the restrictions placed for the release of blood products based on current Inventory.
Study
Design/Methods: Study Designs and Methods: A look forward Data driven IT model created for the Transfusion Service (TS) contained a link to the available blood products dashboard in EPIC and included: Total number of inpatients/EC current patients with their latest Hgb levels/blood types; location of O pos patients, patients with transfusion orders, those currently transfused by each ABO Rh blood type and total number of patients who were currently not transfused and with no active orders. The hemoglobin thresholds were > 8, <= 8, <= 7.5, <= 7, <= to 6.5 gm/dL. The model also calculates the number of patients impacted if the inventory decreased to below minimal normal levels. In addition, it assumed that a certain number of O pos red cells would be provided by various suppliers and predicts the demand for O pos units based on the current census. The inventory for platelets was categorized by ABO type with days to expiration and available number of platelet units for use. Graphs for expected supply and average demand for the week as well as days to expiration, next 3 days projected demand and doses on hand (including a scenario of how much the inventory could be extended by partial doses) by expiration date for Platelets were also illustrated. The model also predicts minimal surgical inventory needed utilizing current total number of cases with the expected number of O Positive cases and other blood types aggregated together as well as risk of MTP/Emergent orders. For outpatients, the total number of appointments, expected total outpatient transfusions and expected outpatient O Pos transfusions are predicted by the model.
Results/Findings: Results/Findings: The number of inpatient/EC patients segregated by their ABO Rh blood types and patients being currently being transfused allows the Transfusion Service to judiciously dispense available pRBC and platelet units in inventory and keep track of the number of Rh Neg patients. Conclusions: Conclusion The look forward model is unique as it includes a report of the daily patient census by each ABO blood group, by Hgb levels, the expected demand by weekday based on historical data, current census, and the number of units needed from all suppliers to keep the inventory at established levels Also importantly, the predictive model informs the Transfusion Service of the total number of Rh negative patients with Hgb < 8.0 gm/dL, < 7.5, < 7.0, < 6.5 gm/dL, includes predictions for surgery and outpatients, and O pRBC Risk Level report if the anticipated blood supply is not received.
Importance of research: The Data Driven Look Forward IT model allows the Transfusion Service to accurately predicts the number of patients who may need transfusion of blood and blood products based on daily census by ABO Rh blood group allowing for restrictions to be placed dependent on established Inventory levels with notification of key stakeholders .
