ABSTRACT
Background: The aim of this study was to investigate the impact of blood transfusion (BT) on mortality and rebleeding in patients with gastrointestinal bleeding (GIB) and whether BT at a threshold of ≤7 g/dL may improve these outcomes. Methods: A prospective study was conducted in patients admitted with GIB between 2013 and 2021. Antithrombotic (AT) use and clinical outcomes were compared between transfused and non-transfused patients, and between those transfused at a threshold of ≤7 vs. >7 g/dL. Multivariate analysis was performed to identify predictors of mortality and rebleeding. Results: A total of 667 patients, including 383 transfused, were followed up for a median of 56 months. Predictors of end-of-follow-up mortality included: age-adjusted Charlson Comorbidity Index, stigmata of recent hemorrhage (SRH), and being on anticoagulants only upon presentation (P=0.026). SRH was a predictor of end-of-follow-up rebleeding, while having been on only antiplatelet therapy (AP) upon presentation was protective (P<0.001). BT was not associated with mortality or rebleeding at 1 month or end of follow up. Among transfused patients, being discharged only on AP protected against mortality (P=0.044). BT at >7 g/dL did not affect the risk of short or long-term rebleeding or mortality compared to BT at ≤7 g/dL. Conclusions: Short- and long-term mortality and rebleeding in GIB were not affected by BT, nor by a transfusion threshold of ≤7 vs. >7 g/dL, but were affected by the use of AT. Further studies that account for AT use are needed to determine the best transfusion strategy in GIB.