Inter-observer agreement for detection of small bowel Crohn's disease with capsule endoscopy

Compared to other modalities, capsule endoscopy (CE) has a high diagnostic yield for diagnosing small bowel Crohn's disease (CD). The aim of this study was to determine the inter-observer agreement for detection of small bowel CD with predefined diagnostic criteria.

Thirty patients with suspected or known CD were included in the study. Observers were blind to patient histories, the results of ileo-colonoscopies, and small bowel examinations. More than three ulcerations (aphthous lesions or ulcers) or the presence of stenosis caused by inflammation or fibrosis was diagnostic of small bowel CD. Three observers with experience in gastrointestinal endoscopy and CE participated in the study.

The presence or absence of small bowel CD was determined with complete agreement in 23 patients, nine patients with and 14 without small bowel CD. The inter-observer agreement was substantial for the diagnosis (kappa = 0.68) and moderate for the localization of CD (kappa = 0.44). Aphthous lesions were detected with only fair agreement (kappa = 0.38). The time intervals to passage of the pylorus and ileo-caecal valve were detected with excellent intra-class correlation.

CE is performed with substantial inter-observer agreement for detection of small bowel CD. In the majority of patients, the presence or absence of small bowel CD is unequivocal. However, in patients with few or minor lesions, the diagnosis is observer dependent.