Introduction Audit is a mandatory, and important, part of postgraduate training but many audits do not capture sufficient data to drive service change. The audit cycle can be broken when doctors in training rotate away from institutions. An audit network within a regional rotation can address both these issues. Parallel, multi-centre data collection allows benchmarking between institutions. Trainee-led networks are common in other disciplines. We have formed the Gastroenterology Audit and Research Network (GARNet) for East Midlands Trainees. Our aim was to demonstrate the feasibility of a multi-centre audit delivered through the GARNet.

Method We audited national guidance NICE CG141, updated August 2016, stating that patients with suspected acute upper gastrointestinal bleeding (AUGIB) should not be offered acid suppression drugs prior to endoscopy. Setting: a prospective audit in 9 hospitals in the East Midlands, UK Patients: aged ≥16, admitted in November 2016 with AUGIB who underwent endoscopy. Current in-patients who developed AUGIB were excluded. Standard: 0% should receive acid suppression. Secondary standard: 100% of patients should be offered endoscopy within 24 hours of admission. Data on other aspects of CG141 were also collected. Each hospital registered with its audit department. Anonymised data was pooled for analysis.

Results 166 patients were identified from a catchment population of 4.5 million, an incidence of 45 per 1 00 000. Complete data were available on 154, of whom 78% presented via the emergency department and 22% were referred from primary care. Mean age was 66y ±18.1. 59% were male. 51% presented between 0700 and 1900 on a week day; 29% presented on a week night and 20% at the weekend.

35/154 were already on regular treatment with proton pump inhibitors. 80/154 (52%) received acid suppression before their endoscopy, including 48 not previously on treatment. In 37 this was given by injection or infusion. Median time to endoscopy was 24 hour (IQR 16.6, 44.4). 49.7% of endoscopies occurred within 24 hours of admission. Figure 1 shows individual hospital data.

A Blatchford score was recorded at first assessment in 43%. A post-endoscopy Rockall score was recorded in 1%, by one endoscopist. 38/154 (25%) were treated at endoscopy, of whom 14 had banding and 24 received other forms of endotherapy.

• Download figure
• Open in new tab
• Download powerpoint

Abstract PTU-045 Figure 1

Conclusion A multi-centre trainee-led audit is feasible. Data gathered in a network give a better picture of local and regional areas for improvement than separate individual audits. Variation in performance against NICE standards will be reviewed within institutions before re-audit and re-pooling. GARNet East Midlands welcomes involvement and collaboration with other regions in the future. Funding: Midland Gastroenterological Society

Disclosure of Interest None Declared