P168 Interobserver agreement in bowel ultrasonography for diagnostic assessment in patients with Crohn’s disease
E. Calabrese*1, T. Kucharzik2, C. Maaser3, G. Maconi4, D. Strobel5, S. Wilson6, F. Zorzi7, K. Novak8, D. Bruining9, M. Iacucci8, M. Watanabe10, E. Lolli7, C. Chiaramonte7, S. Hanauer11, R. Panaccione8, F. Pallone7, S. Ghosh8
1University of Rome Tor Vergata, Gastroenterology Unit, Department of Systems Medicine, Rome, Italy, 2Department of Internal Medicine and Gastroenterology Staedtisches Klinikum, Luneburg, Germany, 3Ambulanzzentrum Gastroenterologie am Klinikum, Luneburg, Germany, 4Department of Clinical Sciences, L. Sacco University Hospital, Milan, Italy, 5University Hospital Erlangen, Erlangen, Germany, 6Division of Gastroenterology, Department of Medicine, University of Calgary,, Calgary, Canada, 7Gastroenterology Unit, Department of Systems Medicine, University of Rome “Tor Vergata”, Rome, Italy, 8Division of Gastroenterology, Department of Medicine, University of Calgary, Calgary, Canada, 9Division of Gastroenterology and Hepatology, Mayo Clinic College of Medicine, Rochester, Minnesota, United States, 10Department of Gastroenterology and Hepatology, Tokyo Medical and Dental University, Tokyo, Japan, 11Northwestern University Feinberg School of Medicine, Digestive Disease Centre, Chicago, Illinois, United States
Bowel ultrasonography (US) constitutes an attractive imaging modality in Crohn’s disease (CD) because it can be repeated frequently to assess and monitor lesions over time. Bowel US does not involve radiation and its low cost provides an alternative to other techniques, especially for children and young patients. Currently, however, a lack of standardisation of the bowel US parameters commonly used in CD and the still reduced number of skilled ultrasonographers in this technique is limiting the use of this imaging modality around the world.
The aim of this study was to assess the interobserver agreement amongst different sonographers, experts in bowel US, from different centres, in the evaluation of predefined bowel US parameters commonly used in the detection and follow-up of patients with CD to develop an unequivocal imaging interpretation.
In the study, 15 patients with established CD underwent bowel US performed blindly by 6 sonographers from different IBD centres around the world, expert in bowel US (>1000 exams/year). Before the evaluation, the operators and clinical and radiological IBD experts dedicated a meeting to formally define the US parameters (site lesion, bowel wall thickness [3–7 mm vs > 7 mm]), bowel wall pattern (preserved vs disrupted), vascularisation (absent/mild vs moderate/severe), lymph nodes, mesenteric adipose tissue alteration, stenosis, fistulas, phlegmon, and abscess) that have been included in the study. Interobserver agreements were tested with Fleiss’ kappa statistics.
Fleiss’ k values were moderate for bowel wall thickness (k = 0.54, p = 0.01), site lesion (k = 0.46, p = 0.01), bowel wall pattern (k = 0.43, p = 0.001), vascularisation (k = 0.59, p = 0.001), and mesenteric adipose tissue alteration (k = 0.43, p = 0.001). Agreement was good and excellent for prestenotic dilation and penetrating complications (k = 0.73–1, p = 0.001). Fair and poor agreements were observed for stenosis, lymph nodes, and lesion extent (k < 0.3, p = 0.001).
In this pilot study, most of the US parameters used in CD patients showed a satisfactory reproducibility. The development of a common US imaging interpretation amongst bowel sonographers will allow the better comparability of US results amongst various centres, improve the quality of multicentre US studies and bowel US training with a further diffusion of this technique.