Non-small-bowel lesions identification by capsule endoscopy: A single centre retrospective study

•Small bowel capsule endoscopy is currently the first-line approach for the investigation of obscure gastro-intestinal bleeding. It also identifies non-small-bowel lesions missed at conventional endoscopy.•This study confirms the high diagnostic yield of capsule endoscopy, in patients with both occult and overt obscure gastro-intestinal bleeding.•The high number of non-small-bowel lesions detected suggests that a second look with conventional endoscopy might be recommended in selected cases, resulting in quicker diagnosis and considerable economic savings. Capsule endoscopy has been considered the first-line approach for the investigation of obscure gastro-intestinal bleeding since its approval in 2001. Our study aims to evaluate the diagnostic yield of capsule endoscopy in the investigation of this condition. We also analyse the incidence of non-small-bowel lesions missed after conventional endoscopy and later detected by capsule endoscopy in patients with suspected obscure bleeding. A total of 290 patients with negative conventional endoscopy referred to our centre to undergo a capsule endoscopy examination for the investigation of obscure gastro-intestinal bleeding. We considered as non-small-bowel lesions those outside the tract between the second duodenal portion and the ileocecal valve. We also looked for actively bleeding lesions at the time of the exam. Intestinal preparation was good, adequate or poor in 74.1%, 8.4%, and 17.5% of the tests, respectively. Caecum was reached in 92.4%. Capsule retention occurred in 0.7%. Mean small bowel transit time was 5hours and 13minutes. Diagnostic yield was 73.8%. An actively bleeding lesion was noticed in 39.3% of positive tests. Capsule endoscopy revealed clinically significant non-small-bowel lesions missed at gastroscopy or colonoscopy in 30.3% of patients, 43.2% of which were bleeding. Capsule endoscopy has high diagnostic yield and safety in the investigation of obscure gastro-intestinal bleedings. Given the high percentage of non-small-bowel lesions detected, it may be appropriate to consider an endoscopic second look before performing a capsule endoscopy study.