Bridging the gap between protocol and practice: A quality improvement initiative to optimize platelet use in a community healthcare setting

Abstract Introduction: Platelet transfusion (PLT) is a critical intervention used to prevent or manage bleeding in individuals with platelet function disorders or thrombocytopenia. Electronic medical record (EMR) systems like Epic provide common clinical indications for platelet transfusions. Additionally, established guidelines offer evidence-based criteria for transfusion thresholds and appropriate indications. However, there is concern regarding discrepancy in EMR indications in relation to established guidelines resulting in overuse of platelet products which could impose a financial burden on the healthcare system as well as increase transfusion-related reactions. This study evaluates the concordance of PLT transfusion practices with Epic indications and established transfusion guidelines, aiming to identify areas of alignment and deviation in clinical application to optimize PLT product use in a community healthcare setting. Methods: A retrospective population study was conducted at Landmark Medical Center in Rhode Island. Patients who received platelet transfusions between 2021 and 2023 were identified through blood bank records. Relevant data were extracted from the Epic EMR system, including the number of platelet units ordered and administered from two different PLT products (PLT pheresis leukoreduced (LR) and platelet apheresis acid citrate dextrose-reduced leukocytes (PLT APH ACD-A Resleu), pre-transfusion platelet counts, and documented clinical indications. A proportional analysis was performed to assess the alignment of platelet transfusion practices with established transfusion guidelines. Results: A total of 188 patients who received platelet transfusions were analyzed and subgrouped by product type: PLT Pheresis Leukoreduced (LR) (n=42) and PLT APH ACD-A Resleu (n=146). In the PLT Pheresis LR group (n=42), 50% were aged 61–75 years, with near-equal gender distribution (47.6% female, 52.4% male). Most patients (90.3%, n=38) met institutional (Epic) criteria, but only 66.6% (n=28) met established guideline criteria. Of those meeting Epic criteria, 23.7% (n=9) transfusions did not align with guideline recommendations. Among non-guideline-concordant transfusions, 12.0% had pre-transfusion platelet counts >100K, suggesting potential overuse in cases of mild thrombocytopenia. In the PLT APH ACD-A Resleu group (n=146), 84.8% (n=124) met Epic criteria; however, 26.6% (n=33) of these did not meet guideline standards. Notably, among all non-guideline-concordant transfusions (n=46), 12.3% occurred at platelet counts between 10–20K, indicating potential overuse due to institutional patterns or provider preference. Conclusion: While a high percentage of platelet transfusions conformed to Epic EMR indications, a substantial portion did not adhere to clinical guidelines. This discrepancy highlights a potential gap between EMR-based decision support and evidence-based practice. These findings highlight the need to refine decision support algorithms within EMR systems to better align with current guidelines and to provide targeted education or oversight to reduce unnecessary transfusions and associated wastage.