Search in the Abstract Database

Abstracts Search 2019

P110 Inter-rater validity of a new scoring index for Crohn’s disease (Crohn's disease activity in capsule endoscopy)

T. Omori*1, K. Yasuhiro1, S. Murasugi1, H. Kambayashi1, T. Hara1, A. Ito1, M. Yonezawa1, S. Nakamura1, K. Tokushige1

1Tokyo Women's Medical University, Institute of gastroenterology, Tokyo, Japan

Background

The Lewis Score (LS) and the Capsule Endoscopy Crohn's Disease Activity Index (CECDAI) are scoring indices for small bowel capsule endoscopy (SBCE) in patients with Crohn’s disease (CD) and small bowel lesions. We proposed the new capsule endoscopic scoring index (CDACE) correlated with existing scores. CDACE is evaluated by dividing the small intestine into four sections, determining the sum (range: 0–16) of the degree of inflammation at each section (range: 0–4; inflammatory score: (A) the number of sections with inflammation (range: 0–4; zone score: (B) and the degree of stenosis (range: 0–3; stenosis score: (C) and using the equation CDACE = A × 100 + B × 10 + C (range: 0–1643).

Methods

In purpose, we evaluate the validity and rate of concordance of CDACE. An expert with experience calculated CDACE scores of 184 SBCE sessions performed on 102 patients with CD. Twenty patients forming a representative score range were included in the analysis (CDAI 168 ± 115, LS 566 ± 1191; range: 0–3961), CECDAI 6.6 ± 4.4 (range: 0–13). After anonymizing and randomising these cases, two gastroenterologists (reader A, B) independently interpreted the image records, determined the LS, CECDAI, and CDACE score, and we determined the concordance between the three gastroenterologists including the original expert. We evaluated the concordance using the intraclass correlation coefficient (ICC) (2.1). We also compared the correlation between the scores.

Results

The average CDACE scores for the expert were 594 ± 395 (range: 0–1243), for reader A, 760 ± 351 (ranging: 110–1342), and for reader B, 546 ± 357 (range 0–1340). The ICC (2.1) equalled 0.618, indicating a somewhat strong concordance. In addition, CDACE exhibited a correlation with the existing scores, and was strongly correlated with CECDAI (expert: CDACE:LS (r = 0.662, p < 0.0001), CDACE:CECDAI (r = 0.911, p < 0.0001), LS:CECDAI (r = 0.784, p < 0.0001).

Conclusion

The CDACE can be used to determine the range of inflammation of the small bowel (second digit of the score) and the presence or absence of stenosis (first digit of the score). Moreover, to some degree, it is possible to infer inflammation morphology (third and fourth digits of the score) from these results. CDACE scores were found to have a somewhat strong concordance among readers, as well as a correlation with existing scores.