A quality improvement initiative to reduce low-value red blood cell transfusions in hospitalized oncology patients.

Abstract

116 Background: Red blood cell (RBC) transfusions are costly, a limited resource, and pose serious medical risk. While most disciplines now transfuse sparingly, hospitalized oncology patients at our institution are often transfused liberally. We conducted a QI initiative to reduce practice variation by adopting a standardized restrictive transfusion strategy. Methods: We performed a retrospective analysis of the total number of pre-transfusion hemoglobin-adjusted RBC transfusions from 02/2015 to 08/2015. Patients with severe gastrointestinal bleeding and acute coronary syndrome were excluded. Extensive feedback from providers was solicited and a comprehensive review of the evidence was performed to understand the process driving existing practices. This information and a root cause analysis were presented to key members of Oncology leadership to gain consensus on a RBC transfusion algorithm. The final algorithm recommended transfusing to a Hg goal of 8 g/dL if the platelet count was ≤ 50K; otherwise the Hg goal was 7 g/dL. We defined low-value RBC transfusions as a transfusion occurring outside the parameters of this algorithm. A multifaceted educational campaign comprising in-person meetings and web-based postings was launched to target ward nurses, advanced practitioners, residents, and attending physicians. RBC transfusions were then tracked to monitor adherence to the algorithm. Results: Analysis of the current process using QI methods found that RBC transfusion practices are most influenced by uncertainty of best practice and variation in attending preference. After the implementation of a RBC transfusion algorithm, the overall number of low-value RBC transfusions decreased from 29.6% to 19.7% (p < 0.001). This reduction was most marked in patients with platelet counts > 50K where there was a reduction from 78.7% to 55.6% (p < 0.001). A 9.9% overall reduction corresponds to roughly 520 fewer RBC transfusions per year, which suggests an annual estimated cost savings of $520,000. Conclusions: By using a standardized QI framework, we were able to successfully decrease variation in practice that results in low-value RBC transfusions in hospitalized oncology patients.