Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P231 Development of a small bowel capsule endoscopic index of severity in patients with Crohn’s disease

A. Bourreille*1, A. Attar2, V. Maunoury3, M. Simon4, A. Aubourg5, M. Fumery6, J.-M. Reimund7, C. Reenaers8, I. Nion-Larmurier2, B. Bonaz9, J. Filippi10, O. Dewit11, X. Dray2, J. Moreau12, R. Altwegg13, X. Roblin14, C. Stefanescu2, J.-y. Mary15

1CHU Nantes, Institut des Maladies de l’Appareil Digestif, Nantes, France, 2CHU Paris, Gastroenterologie, Paris, France, 3CHU Lille, Gastroenterologie, Lille, France, 4CHU, Gastroenterologie, Paris, France, 5CHU Tours, Gastroenterologie, Tours, France, 6CHU Amiens, Gastroentérologie, Amiens, France, 7CHU Caen, Gastroenterologie, Caen, France, 8CHU Liège, Gastroenterologie, Liège, Belgium, 9CHU Grenoble, Gastroenterologie, Grenoble, France, 10CHU Nice, Gastroenterologie, Nice, France, 11CHU Bruxelles, Gastroenterologie, Bruxelles, Belgium, 12CHU toulouse, Gastroenterologie, Toulouse, France, 13CHU Montpellier, Gastroenterologie, Montpellier, France, 14CHU Saint Etienne, Gastroenterologie, Saint Etienne, France, 15Inserm, Biostatistique, Paris, France


The capsule endoscopy is the most powerful device to detect small bowel (SB) lesions in patients with Crohn’s disease (CD), but disease severity remains difficult to evaluate. The aim of our study was to develop a severity index using the Pilcam® SB capsule


Patients with CD were prospectively enrolled between June 2007 and May 2013. Inclusions were stratified according to SB disease severity at endoscopy or radiology and to clinical activity to ensure a wide range of severity. The SB was divided into tertiles. When possible, the terminal ileum (TI) was defined as the last 15 cm above the caecal valve. Within each segment, each of the endoscopists noted patchy or diffuse erythema, oedema, villous denudation, nodularity, lymphangiectasia, pseudo-polyp, presence of blood, numbers according to predefined frequencies of aphtous lesions, ulcerations taking into account size and depth, characteristics of fistulas, extension according to predefined percentages in length of oedema and of ulcerations and in circumference of the largest ulceration, estimated surface and length of lesions and of ulcerations on a 10cm linear analogue scale. The physician global assessment of severity (PGAs) was assessed by the endoscopist in the same way for each segment and globally for the whole SB. The severity index was constructed through multiple linear mixed modelling as a linear combination of the various lesions as quantified, of the lesion and ulceration lengths and surfaces, highly correlated (correlation r) to the segmental PGAs, testing if the combination could be assumed constant across small bowel segments through interaction terms and using investigator as a random factor. Global severity index was estimated through the same type of model as a linear combination of the segmental PGAs highly correlated to the global PGAs, taking into account the observation of the terminal ileum or not


In total, 118 patients were enrolled in 19 GETAID Centres. Nine films were unusable because of a prolonged stay in the stomach (8) or missing data (1). The colon was identified in 80 patients (73%). A linear combination of the extension of oedema, the numbers of large ulcerations, deep and superficial ulcerations, the characteristics of stenosis, the presence of erythema, blood and nodularity, was highly correlated to the segmental PGAs (r > 0.92). The SB severity was predicted by adding the severity of the first tertile and twice that of the third tertile, or by adding the severities of the first plus the third and the TI (r > 0.95)


This prospective study has allowed us to construct the first SB severity index related to the severity as judged by the endoscopists. Simplifications are ongoing to obtain a tool that is easy to use in clinical practice