442 Liberal Transfusion Strategy As a Potential Risk Factor for Mortality in Acute Nonvariceal Upper Gastrointestinal Bleeding (Nvugib)

Acute NVUGIB is a frequent indication for red cell blood (RBC) transfusion, especially in patients with massive bleeding, where comorbid conditions can be decompensated by an insufficient availability of oxygen to tissues. Nonetheless, recent data show that a liberal approach to transfusion may be a risk factor for death in patients with upper GI bleeding. Aim of the study was to assess the role of liberal transfusion policy on the risk for mortality in patients admitted for acute NVUGIB. 30-day mortality was the primary outcome measures. The Rockall score was used to assess clinical conditions prior to transfusion. Only transfused patients without recurrent bleeding were included. A propensity risk score, using the Rockall score and the pre-transfusion hemoglobin value, was constructed to stratify the patients in 2 homogeneous death risk groups (low and high); discriminant analysis was performed in order to identify the number of RBC transfusion better separating deceased patients from survivors. A total of 2.257 patients were analyzed (mean age 68.0±16.4 years, 66.2% males). Overall, 851 pts (36.7%) were transfused and 771 had no recurrent bleeding. In the 1406 patients not transfused the mean Hb value was 10.02 [SD ± 2.5] g/dL vs. 7.5 [SD ± 1.8] g/dL of those transfused (p<0.0000). The Rockall scores were 4.1 [2] vs. 5.2 [2], respectively (p<0.0000). Of the transfused patients 50% received 1 or 2 RBC units and 50 (6.5%) died. 24/596 patients (4.4%) died in the propensity low risk and 24/175 (13.7%) in high risk group (p<0.000). The mean number of RBC units was 2.9 [1.7] vs. 3.7[1.9] respectively (p<0.000). A number of transfusion > 3 had an OR for death of 3.3 (95% CI 1.67-6.31, p<0.0001). Baseline risk profile and number of transfusion were independent risk factors for mortality at logistic regression analysis (table). RBC transfusion may be a risk factor for mortality in patients hospitalized for NVUGIB. A cut-off of >3 units is the number of RBC transfusion better discriminating the risk of death, which increases by 27% per each transfusion after the first one. Tabled 1 OR for mortality 95% CI p value Number of RBC transfusions 1.27 1.13 to 1.44 0.000 High risk patient at baseline 2.05 1.25 to 3.36 0.004 Open table in a new tab