Gastrointestinal bleeding in the oldest old: predictors of emergency outcomes and 30-day mortality

Aims: “Oldest old” (patients aged 85 years and older) represent a particularly vulnerable group, yet data identifying predictors of adverse outcomes in this population remain limited. This study aimed to identify clinical, physiological, and laboratory predictors of adverse emergency department (ED) outcomes and 30-day mortality in patients aged ≥85 years presenting with gastrointestinal (GI) bleeding, and to evaluate the impact of antithrombotic therapy on these outcomes.Methods: This retrospective observational study included patients aged ≥85 years who presented to a tertiary ED with GI bleeding between May 2020 and May 2023. Demographic characteristics, comorbidities, antithrombotic therapy, vital signs, laboratory parameters, and clinical outcomes were extracted from electronic medical records. Primary outcomes were ED disposition (discharge, ward admission, intensive care unit admission, or ED death). Secondary outcomes included 30-day mortality, erythrocyte transfusion within 24 hours, and emergency gastroscopy. Univariable and multivariable logistic regression analyses were performed to identify predictors of severe emergency outcomes and 30-day mortality. Results: A total of 88 patients were included. Shock index, serum lactate, neutrophil-to-lymphocyte ratio (NLR), and erythrocyte transfusion requirement were independently associated with adverse outcomes. Antithrombotic therapy was common (61.4%) and demonstrated a stepwise increase in adverse outcomes from no therapy to monotherapy and dual therapy; however, antithrombotic agent class was not independently associated with mortality after adjustment for physiological severity.Conclusion: In oldest olds with GI bleeding, markers of acute physiological instability are stronger predictors of adverse outcomes than antithrombotic therapy alone. Larger multicenter studies are needed to validate these findings and better define anticoagulant safety in this high-risk population.