2022 Comparative Audit of Acute Upper Gastrointestinal Bleeding

Key objectives

• To collect data from all NHS acute admitting hospitals in the UK regarding numbers, demographics, management and outcomes of patients presenting with AUGIB.
• To assess changes to patient population related to aetiology and clinical presentations with AUGIB compared to 2007.
• To audit resource availability, both within normal working hours and OOH (including at weekends), regarding access and use of emergency endoscopy, IR and surgery.
• To audit the following against UK / NICE standards / AUGIB care bundle recommendations and 2007 audit results and identify variation in practice:
  • time taken from presentation to any specialist intervention (endoscopy/IR/surgery);
  • use of endoscopic therapies for patients with AUGIB;
  • use of specific drug therapies (e.g. PPIs, terlipressin, antibiotics and TXA);
  • transfusion practices for these patients including thresholds for red cell transfusion and use of FFP, platelets and other products;
  • quantify the use of antiplatelet and anticoagulant medications in patients presenting with AUGIB, and audit the management of these patients.
• To measure the use and impact of risk scoring systems for patients presenting with AUGIB and compare the utility of commonly used risk scores i.e., Glasgow-Blatchford score (GBS), Rockall score, and other risk scores including the recently developed ABC score. [20–24]
• To make recommendations based on findings on OOH care, blood use, endoscopy, optimal use and timing of IR and surgery and any other factors that are highlighted as having a clinically significant impact on patient outcomes.
• To explore the use of ML to develop tools for risk assessment.
• To review the involvement of GI trainees in the endoscopic management of AUGIB.

What did we audit?

• 225 NHS Trusts and Board in the UK that accept acute, adult admissions were invited to participate. Hospitals that focus on children or non-related specialities such as maternity hospitals or neurological units were not asked to participate. Independent hospitals were not invited to participate since GI bleeds are managed in the NHS. Each NHS hospital site in the UK admitting acute medical and surgical admissions was eligible for enrolment for this audit.
  
  

Who took part?

• 147 sites contributed data
• 5142 patients were analysed
  
  

What did we find?

• Full report (PDF 2.49MB)

Main findings 

• In-hospital mortality, rebleeding rates, and the need for surgery have declined compared to 2007, dropping from 10% to 8.8%, 13.3% to 9.7% and 1.9% to 0.7 %, respectively, despite patients being older, having more comorbidities (including chronic liver disease), and increased use of anticoagulants.
• Routine implementation of care bundles remains limited, with usage reported in only 43% of hospitals. Additionally, many patients (40%) lacked pre-endoscopy risk stratification, leading to missed opportunities for improved planning and early intervention.
• Adherence to recommendations for managing variceal bleeding remains inconsistent with less than half of eligible patients receiving essential treatments such as antibiotics (44%) and terlipressin (49%).
• The use of red cell transfusions often deviated from national guidelines with 57% of patients eligible for a restrictive approach receiving transfusions outside the guidelines. This was particularly common among stable patients without clinically significant bleeding, thus exposing them to unnecessary risks including higher risk of rebleeding and mortality.
• In-patient endoscopy has increased compared to 2007, rising from 74% to 83%, along with greater use of therapeutic endoscopic interventions (23% in 2007 to 27.1% in 2022) and interventional radiology (1.2% in 2007 to 2.6% in 2022).
• While the cause of bleeding was identified for most patients on endoscopy, about a third (34%) had no abnormalities detected and may not have required urgent endoscopy or endoscopic therapy. This underscores the need for better risk stratification to optimise resource allocation.
• Access to out-of-hours endoscopy remains available in 92% of hospitals, similar to 2007, but is not yet universal. Notable improvements in on-call staffing and trained nursing support reflect a more structured approach to emergency care.
• On-site interventional radiology availability has increased from 23% in 2007 to 65% in 2022, but only 44% of hospitals offered 24/7 IR service, leaving critical gaps in care.
• Trainees face limited opportunities to gain hands-on experience in managing AUGIB, emphasising the importance of supervised training and formal haemostasis courses.

Our recommendations:

Clinical care

• Ensure consistent implementation of validated risk scores and the British Society of Gastroenterology (BSG) AUGIB consensus care bundle at presentation, particularly in emergency departments (ED) and acute medical units (AMU).
• Adhere to national guidelines for restrictive thresholds for red cell transfusions (Haemoglobin (Hb) <70 g/L for stable patients, except in acute coronary syndrome (ACS)). Use single-unit red blood cell (RBC) transfusions for stable patients and reassess the patient’s clinical status and Hb before transfusing further units.
• Increase adherence to guideline-recommended management plans for patients with variceal and non-variceal bleeding.
• Focus on strategies to reduce unnecessary endoscopies, especially for low-risk patients, to optimise resource utilisation.

Organisational care

• Ensure protected daily emergency endoscopy slots and formal 24/7 on-call endoscopy rotas.
• Address gaps in access to interventional radiology, including formal networks for transfer and repatriation. Aim for universal availability of minimally invasive haemorrhage control techniques. Establish clear pathways for timely access to interventional radiology (IR) and transfer for centres lacking on-site 24/7 IR or surgical services.
• Conduct annual local audits on AUGIB management, focusing on transfusion practices, care bundle compliance, and training gaps.

Training

• Improve trainee access to AUGIB cases and therapeutic endoscopy through increased supervision and structured involvement on semi-elective inpatient lists and in on-call rotas during the final years of training.
• Promote attendance at the Joint Advisory Group on Gastrointestinal Endoscopy (JAG) Haemostasis Course for trainees managing AUGIB.
• Ensure future iterations of training curricula include endoscopic haemostasis as a core competency.

Additional recommendations

• Encourage regional collaboration between hospitals to standardise AUGIB care delivery, particularly for complex cases requiring IR or surgery.

Please cite this document as:

Nigam and others on behalf of the National Comparative Audit of Blood Transfusion. 'UK 2022 Comparative Audit of Acute Upper Gastrointestinal Bleeding (AUGIB) and the use of Blood' (2025). https://doi.org/10.71745/ybck-p873