Patient-Important Upper Gastrointestinal Bleeding

• In ICU
• Fri, 26 Sep 2025

Critically ill patients are prone to stress-related erosions that can lead to upper gastrointestinal (GI) bleeding. While previous studies defined clinically important bleeding by investigator criteria, a new outcome measure, patient-important upper GI bleeding, was developed with input from ICU survivors and families. This definition excludes haemodynamic or lab thresholds and instead focuses on overt bleeding leading to blood transfusion, vasopressor use, endoscopy, CT angiography, surgery, death, disability, or prolonged hospitalisation.

In the REVISE (Re-Evaluating the Inhibition of Stress Erosions) trial, which randomised 4,821 invasively ventilated ICU patients to pantoprazole or placebo, pantoprazole significantly reduced both clinically important and patient-important GI bleeding (numbers needed to treat: 40 and 37, respectively), without affecting mortality or infection-related outcomes.

Given evolving ICU care and the competing risk of death, a new study identified modern risk factors for patient-important upper GI bleeding in invasively ventilated critically ill adults, to inform contemporary stress ulcer prophylaxis strategies. This analysis of an international trial database examined baseline and recent (past 3 days) risk factors for patient-important upper gastrointestinal bleeding, while adjusting for illness severity and the competing risk of death.

Patient-important upper gastrointestinal bleeding occurred in 2.7% (131/4,821) of ICU patients. Higher risk was linked to greater illness severity (APACHE II score), vasopressor use, severe thrombocytopaenia, and platelet inhibitor therapy. Lower risk was associated with pantoprazole and higher volumes of enteral nutrition. Pantoprazole reduced bleeding risk regardless of enteral nutrition, and the protective effect of enteral feeding was consistent with or without pantoprazole. Corticosteroids were not linked to increased risk. Sensitivity analyses also implicated dialysis-dependent renal failure and coagulopathy.

Both pantoprazole and higher volumes of enteral nutrition were associated with reduced bleeding risk, though causality cannot be inferred, and pantoprazole remained protective regardless of nutrition. Previous evidence similarly links renal dysfunction and illness severity to bleeding, but differences in study populations, definitions, and analytic methods complicate comparisons.

Although gastrointestinal bleeding carries attributable mortality, it is usually recognised and treated early. Pantoprazole reduces bleeding but does not lower overall mortality, consistent with other trials and meta-analyses.

Overall, these findings refine the understanding of bleeding risk in critically ill patients and highlight the value of integrating patient and family perspectives into research.

Source: AJRCCM
Image Credit: iStock 

References:

Dean AM, Lauzier F, Adhikari NKJ et al. (2025) Risk Factors for Patient-Important Upper Gastrointestinal Bleeding. AJRCCM.