ADWE-02 Diagnostic yield and early discharge in gastrointestinal bleeding: comparative trial between capsule endoscopy and gastroscopy

Abstract

Introduction Capsule endoscopy may have a role in triaging patients with suspected acute upper gastrointestinal (GI) bleeding who need further invasive investigation. Method The Mirocam Navi (Intromedic Ltd., Korea) is a single camera capsule endoscope steerable by a handheld magnet. Magnetically assisted capsule endoscopy (MACE) was performed in stable patients with suspected acute upper GI bleeding and a decision made as to whether same day discharge was appropriate, prior to gastroscopy (OGD) performed by a different, blinded endoscopist followed by standard care by the admitting team during which small bowel capsule endoscopy was completed. Diagnostic yield and comfort were compared. Potential impact on hospital stay was examined and MACE mucosal visualisation quality was assessed. Results A total of 90 upper GI lesions were detected in 30 patients (80% male, median age 57 years (IQR=26)): 21 (23%) concomitantly identified by MACE and OGD, 49 (54%) by MACE alone and 20 (22%) by OGD alone (CI 0.05–0.58; p=0.0007). MACE and OGD concomitantly detected 50% of major pathologies (gastric and duodenal ulcers and oesophageal and gastric varices (n=2 each)). MACE alone identified 31% (D1 ulcers (n=4) and varices (n=1)) whereas OGD alone identified three D1 ulcers (19%; p=0.7). Ulcers (all classified as Forest class III at OGD) missed by MACE were due to rapid D1 transit and insufficient imaging. Review of small bowel images identified additional causes for GI bleeding distal to D2 in five cases: small bowel lymphoma (n=1), angioectasia (n=2), fresh bleeding in the proximal and distal small bowel with no culprit lesion seen (n=1 for each). MACE correctly suggested safe discharge for 21/21 patients. The median hospital stay of uncomplicated admissions where MACE suggested discharge was 53 (range 14–103) hours. Pain, discomfort and distress scores were lower with MACE than OGD (p 2 =133 (p Conclusions MACE has better diagnostic yield than OGD in suspected upper GI bleeding (with both modalities missing lesions), is better tolerated and used ‘at the front door’ might allow immediate discharge for two thirds of patients saving a median hospital stay for each of over two days. In addition, it suggests that about 10% of suspected upper GI bleeding may have a small bowel source.