Acute upper gastrointestinal bleeding in over 16s: management

Risk assessment tools

New evidence was identified on tools for assessing risk of poor outcomes after acute upper gastrointestinal bleeding (mainly AIMS65, Blatchford score and Rockall score). Evidence indicated that no tool appears to be sufficient to be used alone; which is consistent with the current recommendation to use both the Blatchford score and the Rockall score after endoscopy.

Resuscitation and initial management

Evidence on resuscitation and initial management was generally consistent with current recommendations. Two randomised controlled trials suggested benefits of tranexamic acid after acute upper gastrointestinal bleeding. Effects were not consistent across outcomes, and no effect on mortality was seen. We are awaiting results of the ongoing Haemorrhage alleviation with tranexamic acid – intestinal system (HALT‑IT) study. This National Institute for Health Research (NIHR)-funded trial aims to study the effects of tranexamic acid in 12,000 people with acute gastrointestinal bleeding. When results from this study are published, the impact on the guideline will be assessed.

Oral versus intravenous proton pump inhibitors (PPIs)

We found several studies comparing oral and intravenous administration of PPIs for non-variceal bleeds. Overall, little difference between the methods was found, although results were inconsistent between studies. Therefore, we decided that the case was not strong enough for an update to cover routes of administration for PPIs at this time.

Variceal bleeding

A large number of new studies covering various interventions, comparators and outcomes for variceal bleeding were identified. Results were inconsistent across interventions and outcomes. However, overall, band ligation and transjugular intrahepatic portosystemic shunts (TIPS) appear to be effective for oesophageal varices. Similarly, cyanoacrylate and TIPS appear to be effective for gastric varices. These findings support current recommendations for treating variceal bleeding. There was no strong indicator of a need to update to consider other interventions for variceal bleeding.

Stress ulcer prophylaxis

Evidence indicated that stress ulcer prophylaxis appears to reduce gastrointestinal bleeding, and there was no consistent evidence of increased infections, a potential adverse event noted to be of concern. These findings support the current recommendations to offer acid suppression therapy as stress ulcer prophylaxis in people admitted to critical care.

We are awaiting results of the ongoing Stress ulcer prophylaxis in the intensive care unit (SUP‑ICU) study. This aims to study the effects of stress ulcer prophylaxis in more than 3,000 people with acute gastrointestinal bleeding. When results from this study are published, we will assess the impact on the guideline.

For further details and a summary of all evidence identified in surveillance, see appendix A.