Introduction Despite advances in endoscopic therapy for non-variceal upper gastrointestinal bleeding (NV-UGIB), achieving definitive haemostasis remains a challenge.1 Radiological intervention with embolisation is an alternative to surgery where endoscopic therapy has failed, yet there is little outcome data. We describe our experience of outcome following endoscopic therapy where both radiological and surgical interventions are readily available.
Methods A retrospective observational study of all patients undergoing therapeutic endoscopy as primary treatment for NV-UGIB at the John Radcliffe Hospital, Oxford, was performed. All 180 patients eligible over a 2-year period (January 2009 to December 2010) were included. The main outcome measures were failure of primary endoscopy, defined as continuing bleeding or rebleeding requiring further intervention or causing death, and definitive haemostasis rate after all intervention (repeat endoscopy, radiological embolisation or surgery).
Results 180 patients underwent therapeutic endoscopy; median age 75 years, 114 male (63.3%). 128 (71.1%) had peptic ulcer disease. Haemostasis was achieved at endoscopy in 165 (91.7%). In four patients endoscopic therapy was not attempted due to inaccessibility of the lesion. There was failure of primary therapeutic endoscopy in 40 (22.2%), with continuing bleeding in 13 and rebleeding in 27. A second intervention was undertaken in 37; embolisation in 21, repeat endoscopy in 14 and surgery in 2. 13 required three or more interventions Definitive haemostasis was achieved in 18/25 (72%) of patients undergoing embolisation and 8/8 (100%) of patients undergoing surgery. All cause mortality was 20% in the embolisation group, with one patient dying from ischaemic complications. There were no deaths in the surgical group. Overall, definitive haemostasis was achieved in 174 patients (96.7%) with all cause 30-day mortality 10% and bleeding-related mortality 3.3%. Failure of primary endoscopy was associated with an increased risk in all cause mortality (RR 2.80, CI 1.18 to 6.62, p=0.02).
Conclusion The failure rate of therapeutic endoscopy for NV-UGIB was comparable with the published literature. The combination of endoscopic, radiological and surgical therapy achieved definitive haemostasis in a high proportion (96.7%). When endoscopic therapy failed, interventional radiology was an effective salvage modality in the majority of cases, avoiding the need for surgery. Failure of primary endoscopic therapy was associated with all cause mortality.
Competing interests None declared.
Reference 1. Hearnshaw SA, Logan RF, Lowe D, et al. Acute upper gastrointestinal bleeding in the UK: patient characteristics, diagnoses and outcomes in the 2007 UK audit. Gut 2011;60:1327–35.