Acute management

GI Bleeding


  • In the UK 2,500 patients die each year from internal bleeding; this accounts for 4% of acute hospital admissions, but has a mortality of 13-15%, mainly because most patients are often elderly with significant co-morbidities.
  • Currently less than 40% of trusts have a comprehensive 24/7 service and care is often fragmented across clinical teams.
  • The recent CROMES report provides a framework for commissioners to support trusts in improving quality of care and patient outcomes.

Patient View

  • Admission to hospital with upper gastrointestinal bleeding is involuntary, unexpected, frightening and sometimes rather degrading with soiling by vomited blood and maleana.

Current Practice

  • Admission policies and out of hours endoscopy services are highly variable and not organised in a planned and focussed way: the recent CROMES report showed that 45% of trusts admitting patients with GI bleeding do not have a comprehensive out of hours service.

Recommended Practice and Opportunities for Integrated Working

  • Approximately 20-25% of patients are at low risk (with a Blatchford score of 0: REF) and do not require admission.
  • For those with significant bleeding, early endoscopy plays a central role in management.
  • Commissioners should require Trusts, either alone or in collaboration to provide effective management through a defined pathway for a 24 hours a day, 7 days a week as set out in the recently published CROMES report, Scope for Improvement (see references)
  • Critical components are that each unit must demonstrate competent assessment and risk scoring and resuscitation with patients whose Blatchford score shows it to be appropriate being discharged without urgent endoscopy and those where risk is higher gaining access to treatment as soon as appropriate after resuscitation.
  • There is substantial room to improve safety by implementation of the CROMES recommendations for a cooperative comprehensive GI Bleeding service
  • The principal recommendations are that all patients should have access to endoscopy, interventional radiology and surgery. The CROMES document outlines 3 models for an upper gastrointestinal bleeding service in which hospitals either provide a full 24 hour bleeding service autonomously or by use of a network. This model requires a degree of planning and co-ordination between individual services, particularly
    • the ambulance and A&E emergency services
    • the admissions unit
    • a gastroenterology team
    • specialist staff (gastroenterology or surgery) in a dedicated bleed ward area
    • HDU or ITU where appropriate for resuscitation
    • organisation of diagnostic and interventional endoscopy and radiology
    • Involvement of emergency care surgery.

Opportunities for Savings

  • CROMES can be implemented at little or no cost and better practice with early discharge of low risk patients should be cost saving
  • More low risk patients should be discharged without in-patient endoscopy from the acute medicine area. For some of the higher risk patients who are admitted, time to endoscopy should reduce and never be more than 24 hours.

Quality Indicators (Outcome Metrics)

Outcome measures

  • Deaths from PU bleeding
  • Deaths from variceal bleeding
  • Proportion of patients with Blatchford score of 0 that are inappropriately kept in hospital for investigation.
  • Practice by practice assessment of the proportion of patients on NSAIDs who receive PPI prophylaxes.

Process measures likely to impact on outcome

Adherence to CROMES standards, in particular

  • Nominated individual in charge of the service
  • All patients with suspected UGIB should undergo risk assessment on presentation.
  • All high risk patients with UGIB should be endoscoped within 24 hours, preferably on a planned list in the first instance.
  • For patients who require more urgent intervention either for endoscopy, interventional radiology or surgery formal 24/7 arrangements must be available.
  • Each stage of the patient pathway should be carried out in an area with appropriate facilities, equipment and support including staff experienced in UGIB management.
  • All hospitals must collect a minimum data set in order to measure service provision against auditable outcomes (case-mix adjusted as appropriate).

Social Policy and Public Understanding

  • There are indications that too liberal blood transfusion may, somewhat counter intuitively, be associated with less favourable outcomes. A randomised HTA trial is planned to investigate this.
  • Commissioners should take steps to ensure that all patients who receive NSAIDs have PPI protection, as recommended by NICE.