PS1589 RETROSPECTIVE ANALYSIS OF HEMOGLOBIN LEVEL IN 500 TRANSFUSION EVENTS AS A STUDY MODEL FOR THE EVALUATION OF TRANSFUSION APPROPRIATENESS

Background: The hemoglobin (Hb) value adopted in the real life setting for the clinical decision of packed red blood cell (RBC) transfusion (transfusion trigger) is sometimes difficult to accurately retrieve and could be different to what is indicated in the blood request. The identification of the transfusion trigger is critical for the analysis of the distinct physicians’ attitudes across the different units within the same hospital and for the dissection of transfusion appropriateness. Aims: The aims of the present analysis is to compare the transfusion trigger across the multiple medical and surgical units at our Center and to describe the Hb values before and after the transfusion events. Further, considerations about the feasibility and the potential applications of this analytical approach are discussed. Methods: Data extraction procedure focused on the first semester 2018, dealing with transfusion episodes of 1–2 packed RBCs administered to patients hospitalised at ASST GOM Niguarda, Milan, Italy. Each single episode was registered, regardless of the number of the total packed RBCs transfused, as well as the Unit where the patient was admitted and the Hb values at admission, before and after the transfusion, together with the respective date and time. Comparisons analyses were made using the classical statistical tests. The Hb value before the transfusion is considered as the trigger for transfusion decision by the physician(s) in charge of the patients. Results: On a total of n = 501 episodes, n = 497 were evaluable for completeness of data. The corresponding number of packed RBCs represents the 2% of all the packed RBCs transfused at our Hospital in one year (n = 22170 in 2018). Median Hb was 10.70 (range: 6.30–16.70), 8.60 (range:3.00–15.10) and 10.30 (range: 5.00–16.30) at admission, before and after transfusion, respectively. Analyzing separately by unit, we observed that the transfusion trigger was higher in Cardiac Surgery, Vascular Surgery and Intensive Care Units, with median values of 9.50, 8.75 and 8.70 respectively. Conversely, transfusion trigger was comparable among the other surgical units (including Orthopedics) and the medical ones: Hb median values were 8.20, 8.20 and 8.30 respectively. Cardiology showed a slightly higher trigger: 8.65 when compared with the other non‐surgical units. Summary/Conclusion: The present analysis, accounting for « only » the 2% of the total transfusion burden at our Hospital in 2018, allows for a satisfactory definition of the transfusion trigger across the multiple medical and surgical units as well as the confirmation of the expected post‐transfusion Hb values. Our analysis model supports the feasibility of such an approach as a potential audit instrument and intervention policies in the area.