Introduction NICE recommends that patients with non-variceal acute upper gastrointestinal bleeding (AUGIB) and stigmata of recent haemorrhage (SRH) are offered endoscopic treatments (combination or mechanical method).¹

We audited the appropriateness and outcomes of endoscopic therapy in patients with non-variceal bleeding in a University Hospital with a dedicated consultant led out of hours endoscopy service.

Methods AUGIB patients who underwent an emergency endoscopy (OGD) were identified from endoscopy, theatre and electronic records between 01.01.15 and 30.04.15. Our electronic clinical database was used to ascertain clinical details and outcomes.

Results A total of 116 patients received an emergency OGD. Rebleeding occurred in 18 (15.5%) and 30 day mortality was 11%.

24 patients had non-variceal bleeding with SRH. 22 (92%) of these patients received endotherapy. Haemostasis was achieved after endotherapy in 20 (91%) patients.Two patients required CT angiography/embolization of the gastroduodenal artery. Two received no therapy.

19/24 (79%) patients received at least dual endoscopic therapy with 9/24 (37.5%) patients >2 modalities. 5/24 (9%) patients did not receive minimal recommended therapy; 3 patients (13%) received single (non-mechanical) modality. 2 patients received no treatment.

In patients with endoscopic therapy, haemostasis was achieved in 13/22 (59%) patients at first OGD, 5/22 (27%) at second OGD and 2 (9%) at third OGD.

In the two patients who failed endoscopic therapy, haemostosis was achieved at CTA.

7/22 patients (32%) did not achieve haemostasis at index endoscopy and required repeat endoscopy. Of these patients 5 (71%) had suboptimal endoscopic therapy at index endoscopy. In the 15 patients who achieved haemostasis 3 (20%) had suboptimal therapy at index endoscopy. (p = 0.05, Fisher’s exact test)

Conclusion In this audit he failure to deliver effective endoscopic therapy at the index endoscopy led to ongoing bleeding in patients with non-variceal bleeding and SRH. In this audit we found 37% of patients needed triple endoscopic therapy to achieve haemostasis. The impact of such intensive endotherapy requires further study.

Reference 1 NICE: https://www.nice.org.uk/guidance/gs38

Disclosure of Interest K. Waddell: None Declared, G. White: None Declared, A. Stanley Conflict with: Cook Medical, Boston Scientific, A. Morris Conflict with: Cook Medical, Boston Scientific.