ECU Health Medical Center Greenville, North Carolina, United States
Background/Case Studies: Transfusion data visualization in the form of a dashboard can provide insight into the appropriateness of transfusions and support PBM initiatives. Our hospital transfusion service in collaboration with our data analytics group developed a real-time, interactive web based dashboard to assess transfusion guideline compliance, provide feedback to clinicians, and minimize unnecessary transfusions.
Study
Design/Methods: Our health system transfusion service supports nine hospitals. The largest is an academic medical center with 974 beds and is a Level I Trauma center. This hospital transfuses over 13000 RBCs, 3000 platelets, and 3500 plasma units annually. The patient population supported include trauma, hematology/ oncology, cardiovascular, pediatrics, and maternity.
Data is extracted from EPIC Clarity (Epic Systems Corp, Verona, WI), and from blood bank software (WellSky, Overland Park, KS). The dashboards are created using Microsoft Power BI. RBC, platelets, and plasma transfusion data are updated at midnight every day. The dashboard has drill-down capability for each department, provider, and patient.
The key designated performance indicators include: greater than 95% with pre-transfusion laboratory orders; and less than 5% of orders with pre-transfusion value above or below “target”. The “target” for RBC is order with hemoglobin (Hb) value above 7g/dL, platelet with count above 30000, and plasma with INR below 1.6. The rate is calculated by tracking all patients with pre-transfusion orders and transfused. The drill-down capability includes patient and provider specific information, primary diagnosis, and procedure codes for all departments.
Results/Findings: Figure A shows RBC transfusion data from January to December 2022. Thirteen thousand four hundred twenty two (13422) RBC units transfused to 4011 patients; 3.35 units/patient; orders with pre-transfusion Hb value available in 88.5% (below the target of 95%); order for pre-transfusion Hb greater than 7 g/dL is 36.4% (above the target of 5%); the top five ICD10 Codes; average and median pre-transfusion Hb by department. Further drill-down of each department shows the various sections within the department, provider, and patient specific information (not shown). Conclusions: We developed a robust interactive web-based transfusion dashboard to improve patient transfusion outcomes, and reduce blood product wastage.
Importance of research: In house development of customizable transfusion dashboards provide real-time assessment of transfusion guideline compliance. The dashboard is unique with the overview page providing blood product specific transfusion data display and reporting with actionable information. Transfusion dashboards can be provided by external vendors. However, these products are expensive, and require patient information sharing with third party.
