Introduction Chronic liver disease (CLD) is commonly associated with anaemia. Whilst varices represent the commonest cause of gastrointestinal bleeding in patients with CLD, in patients where iron deficiency anaemia (IDA) persist, capsule endoscopy (CE) may have a useful role to investigate the small bowel (SB). We conducted a study to evaluate the utility of CE in patients with CLD and obscure gastrointestinal bleeding (OGB) and their subsequent management consequences.

Methods We retrospectively analysed our data set and isolated patients with OGB and CLD. Data collected included demographics, clinical indication (overt bleeding (OB) or IDA) the presence of co-morbidity, diagnostic yield (DY) and subsequent follow up.

Results Of the 1324 patients investigated for OGB using CE, 3%(n = 41) had CLD. The mean age was 61 years (range = 26–88) with 59% males. The indications for CE was IDA in 66%(n = 27) and OB in the remaining 34%. All patients in this cohort had other significant co-morbidity in addition to CLD. Five patients were on non-steroidal anti-inflammatories whilst 2 patients were transfusion dependent. The DY (as defined by lesions responsible for OGB) identified on CE was 49%(n = 20).The commonest finding was SB ulcer and erosions 27%(n = 11) and SB angioectasia (AE) 24%(n = 10). Other findings included SB varices (2), blood without definite source (5), a tumour (metastatic renal tumour) and other miscellaneous lesions (4). In 13 patients (32%), lesions found were within the upper GI tract, which had been underestimated at the index gastroscopy. These included gastric antral vascular ectasia (3), varices (oesophageal and duodenal)(2), blood without definite source (5) and others lesions (erosions, ulcers, portal hypertension and polyps )(9). In 2 patients, colonic lesions were identified (erosions and AE). There was no significant difference in the DY between those with IDA and OB (p = 0.59) and between the sexes (p = 0.41). In our cohort, management was altered in 90%(n = 18) of those with a DY, in the form of further procedures (25%, n = 5) which included repeat OGD (2), colonoscopy (2), double balloon enteroscopy (1) and the patients with renal metastasis avoided surgery. 25%(n = 5) of patients within this cohort also received argon photocoagulation therapy. On logistic regression, factors that were associated with a subsequent change in management included previous transfusions (p = 0.04) and SB AE (p = 0.03).

Conclusion CE is a useful tool for investigation of OGB in patients with chronic liver disease and persistent anaemia. Ulcers and AE were the commonest pathology seen in the SB in patients with CLD, in keeping with the published literature. CE is also useful to pick up pathology in the upper GI tract which may have been underestimated.

Disclosure of Interest None Declared.