Endoscopy 2015; 47(04): 330-335
DOI: 10.1055/s-0034-1390894
Original article
© Georg Thieme Verlag KG Stuttgart · New York

Validation of the Lewis score for the evaluation of small-bowel Crohn’s disease activity

José Cotter
1   Gastroenterology Department, Centro Hospitalar do Alto Ave, Guimarães, Portugal
2   Life and Health Sciences Research Institute, School of Health Sciences, University of Minho, Braga, Portugal
3   Life and Health Sciences Research Institute/3B’s, PT Government Associate Laboratory, Braga/Guimarães, Portugal
,
Francisca Dias de Castro
1   Gastroenterology Department, Centro Hospitalar do Alto Ave, Guimarães, Portugal
,
Joana Magalhães
1   Gastroenterology Department, Centro Hospitalar do Alto Ave, Guimarães, Portugal
,
Maria João Moreira
1   Gastroenterology Department, Centro Hospitalar do Alto Ave, Guimarães, Portugal
,
Bruno Rosa
1   Gastroenterology Department, Centro Hospitalar do Alto Ave, Guimarães, Portugal
› Author Affiliations
Further Information

Publication History

submitted 15 June 2014

accepted after revision: 22 September 2014

Publication Date:
20 November 2014 (online)

Preview

Background and study aims: The Lewis score was developed to measure mucosal inflammatory activity as detected by small-bowel capsule endoscopy (SBCE). The aim of the current study was to validate the Lewis score by assessing interobserver correlation and level of agreement in a clinical setting.

Patients and methods: This was a retrospective, single-center, double-blind study including patients with isolated small-bowel Crohn’s disease who underwent SBCE. The Lewis score was calculated using a software application, based on the characteristics of villous edema, ulcers, and stenoses. The Lewis score was independently calculated by one of three investigators and by a central reader (gold standard). Interobserver agreement was assessed using intraclass correlation (ICC) coefficient and Bland – Altman plots.

Results: A total of 70 patients were consecutively included (mean age 33.9 ± 11.7 years). The mean Lewis score was 1265 and 1320 for investigators and the central reader, respectively. There was a high correlation, both for scores obtained for each tertile (first tertile r = 0.659 – 0.950, second tertile r = 0.756 – 0.906, third tertile r = 0.750 – 0.939), and for the global score (r = 0.745 – 0.928) (P < 0.0001). Interobserver agreement was almost perfect between the investigators and the central reader (first tertile ICC = 0.788 – 0.971, second tertile ICC = 0.824 – 0.943, third tertile ICC = 0.857 – 0.968, global score ICC = 0.852 – 0.960; P < 0.0001). The inflammatory activity was classified as normal (score < 135) in 2.9 % vs. 2.9 %, mild (score ≥ 135 – < 790) in 51.4 % vs. 55.7 %, and moderate to severe (score ≥ 790) in 45.8 % vs. 41.4 % of patients, respectively (P < 0.001).

Conclusion: A strong interobserver agreement was demonstrated for the determination of the Lewis score in a practical clinical setting, validating this score for the reporting of small-bowel inflammatory activity. The Lewis score might be used for diagnosing, staging, follow-up, and therapeutic assessment of patients with isolated small-bowel Crohn’s disease.