BLOOD MANAGEMENT A data-driven approach to patient blood management Claudia S. Cohn, Julie Welbig, Robert Bowman, Susan Kammann, Katherine Frey, and Nicole Zantek BACKGROUND: Patient blood management (PBM) has become a topic of intense interest; however, implement- ing a robust PBM system in a large academic hospital can be a challenge. In a joint effort between transfusion medicine and information technology, we have devel- oped three overlapping databases that allow for a com- prehensive, semiautomated approach to monitoring up-to-date red blood cell (RBC) usage in our hospital. Data derived from this work have allowed us to target our PBM efforts. STUDY DESIGN AND METHODS: Information on transfusions is collected using three databases: daily report, discharge database, and denominator database. The daily report collects data on all transfusions in the past 24 hours. The discharge database integrates transfusion data and diagnostic billing codes. The denominator database allows for rate calculations by tracking all patients with a hemoglobin test ordered. A set of algorithms is applied to automatically audit RBC transfusions. The transfusions that do not fit the algo- rithms’ rules are manually reviewed. Data from audits are compiled into reports and distributed to medical directors. Data are also used to target education efforts. RESULTS: Since our PBM program began, the per- centage of appropriate RBC orders increased from an initial 70%-80% to 90%-95%, and the overall RBC transfusions/1000 patient-days has decreased by 67% in targeted areas of the hospital. Our PBM program has shaved approximately 3% from our hospital’s blood budget. CONCLUSION: Our semiautomated auditing system allows us to quickly and comprehensively analyze and track blood usage throughout our hospital. Using this technology, we have seen improvements in our hospi- tal’s PBM. P atient blood management (PBM) has become a topic of intense interest for many reasons. Blood is a precious and limited resource that carries inherent risks when transfused. Although infec- tious disease screening has made blood extremely safe, it has contributed to higher costs, which are straining hos- pital budgets. The importance of this topic was recently highlighted by the AABB when it produced the “Best Practices for Patient Blood Management” guidance document. 1 Implementing a robust PBM system in a major uni- versity hospital can be a challenge. Many physicians were trained during a time when good evidence-based guide- lines for red blood cell (RBC) transfusions were lacking. They were taught to follow the maxim “if you are going to transfuse, why not give two.” 2,3 Now, however, there are multiple investigations of RBC 4-11 and platelet (PLT) use, 12-16 which have led to guidelines, published by the AABB and other groups, to help guide practice. 17-20 None- theless, some clinicians can be slow to change. To help foster a good transfusion culture in our hos- pital, we have developed a three-pronged semiautomated approach to monitor RBC usage. In a joint effort between transfusion medicine and information technology, we have developed three overlapping databases that allow us ABBREVIATIONS: EMR = electronic medical record; ICD-9 = International Classification of Diseases; LIS = laboratory information system; PBM = patient blood management. From the Department of Laboratory Medicine and Pathology, University of Minnesota Medical School, and Fairview Health Services, University of Minnesota Medical Center, Minneapolis, Minnesota; and Fairview Health Services, Southdale Hospital, Edina, Minnesota. Address reprint requests to: Claudia S. Cohn, MD, PhD, Laboratory Medicine & Pathology, University of Minnesota, MMC 609, 420 Delaware Street SE, Minneapolis, MN 55455; e-mail: cscohn@umn.edu. Received for publication January 9, 2013; revision received April 10, 2013, and accepted April 17, 2013. doi: 10.1111/trf.12276 TRANSFUSION 2014;54:316-322. 316 TRANSFUSION Volume 54, February 2014