PWE-103 Capsule endoscopy in suspected GI bleeding-is it worth repeating a negative study?

Introduction The most common indications for capsule endoscopy (CE) are iron deficiency anaemia (IDA) and occult gastrointestinal bleeding (OGIB). This study aimed to assess the diagnostic yield (DY) of repeat CE in patients for whom there is ongoing clinical concern, despite an initially negative CE. Methods Patients who underwent ≥2 CEs for IDA/OGIB at our centre, from 2005 to 2017, were identified from a prospectively-designed database. Data were extracted on indications and CE findings. The capsule examination was considered to have DY if the findings accounted for the patient’s presentation. Results 85 patients underwent repeat CE during the study period, median age 65.8 years (range 11.5–89.8; 42F/55M). The median interval between procedures was 463 days (range 1–3066). 14 patients underwent repeat CE due to a retained or incomplete initial capsule and were excluded from analysis. In the remaining 71 patients, initial CE findings were: normal (22), vascular lesions/bleeding (26), small bowel inflammation (5), others (8; including polyps, portal hypertensive enteropathy (PHE), celiac disease, small bowel (SB) varices and SB lymphoma), nonspecific findings of unclear significance (6) and non-SB findings (5). In patients with a normal initial CE, repeat CE identified a cause for IDA/OGIB in 9/22 (40.9%). 12/22 CEs (54.5%) were normal and 1 was incomplete. Of the 19 patients with vascular lesions seen initially, the initial lesion was confirmed in 13/19 (68.4%) CEs. The diagnosis was revised in 2 patients: 1 was found to have PHE and 1 likely NSAID enteropathy. 3/19 (15.8%) patients had normal repeat CEs and 1 was retained. 7 patients had active bleeding on initial CE but no lesion seen. Repeat CE in this group had DY 5/7 (71.4%): angioectasias (3), polyp (1), SB inflammation (1). 2 repeat CEs were normal. In the 6 patients with nonspecific initial findings, repeat CE identified specific findings in 2/6 (33.3%) patients (1 NSAID enteropathy, 1 jejunal ulcer). 10 patients underwent a 3rd CE. In 7/10 patients with concordant initial CEs, the DY of repeat CE was 0/7. Where the 2 initial CEs disagreed, DY was 2/3. Conclusion 1. In patients with a negative or inconclusive initial CE for IDA or OGIB, repeating the procedure has an overall DY of 25% (7/28). The DY is highest when fresh blood was seen in the initial procedure (71.4%) even if no lesions were found initially. Patients with initially normal studies had lower DY (22.7%). 3rd CE is only warranted by a change in presentation or discordance in the previous results, especially when one examination has identified active bleeding.