PWE-147 The Use of Angiography with A View to Embolisation In Non Variceal Gastrointestinal Haemorrhage Uncontrolled with Endoscopic Therapy, A 10 Year Experience at A Teaching Hospital

Abstract

Introduction Gastrointestinal (GI) bleeding is a common medical emergency. The incidence is 1.33/1000 population, approximately 85,000 cases/year. The NCEPOD review of GI haemorrhage emphasised the role of embolisation in patients with GI bleeding who have failed endoscopic therapy. Transcatheter arterial embolisation (TAE) has been shown to be as effective as emergency surgery in the treatment of refractory acute non variceal upper GI bleeding (ANVGIB) with reduced 30 day mortality rates. The study aim was to review the utilisation and outcomes of patients undergoing emergency angiography (EA) with a view to embolisation in ANVGIB not amendable to, or controlled by, endoscopic therapy. Methods A retrospective analysis of patients undergoing EA for ANVGIB between January 2005 and February 2016 was performed. Patients were identified from endoscopy and radiology databases. Endoscopy reports and discharge letters were reviewed electronically. Successful embolisation was deemed when extravasation had stopped and/or no bleeding post embolisation. Results Over the study period, 22412 patients were diagnosed with GI bleeding. 57.9% (12967/22412) had an endoscopy. EA was performed in 107/12967 (0.83%) of patients undergoing endoscopy due to uncontrolled GI bleeding. Mean age was 69 years for males and 74 years for females (range 30–98); 74/107 were male. EA and embolization was performed in hours in 71% of cases and out of hours in 29%. Pre-angiography endoscopy report was available in 86/107 patients. The commonest endoscopic diagnosis was duodenal ulcer which was present in 25.6% (22/86) of the cases. A normal upper GI endoscopy was noted in 20.9% (18/86) of patients. Lower GI bleeding was the cause in 17.4% (15/86) of patients. Other causes included; unidentified bleeding point, dieulafoy lesion, gastric ulcer. Of those patients undergoing angiography 54/106 (50.9%) had TAE. TAE was deemed successful in 90.7% (49/54) of cases. Surgical intervention was necessary in 2/5 of the unsuccessful cases. The 30 day mortality for the patients undergoing angiography with or without intervention was 18/107 (16.8%). 57% (61/107) were alive at 1 year. Conclusion The results demonstrate that angiography and embolisation is a valuable tool in the management of GI bleeding not controlled by endoscopic therapy with a high success rate avoiding the need for surgery in the majority of cases. Our figures are comparable to published studies. The results support the need for 24/7 access to interventional radiology which is part of the recommendations from the NCEPOD review of the care received by patients who had a severe gastrointestinal haemorrhage. Disclosure of Interest None Declared