1051 Assessment of Wireless Capsule Endoscopy in Patients with Ulcerative Colitis

Abstract

Background: Several non-invasive indices are available to measure disease activity in ulcerative colitis (UC), but sparse evidence exists to guide their use. We aimed to systematically compare the psychometric properties of all existing indices and to determine cutoff scores that correspond to remission and response. Methods: In this prospective longitudinal study, 86 adults with UC were recruited (52% males, mean age 37.6 ± 13.7 years, median disease duration 6.6 (2.1-13.1) years). Items from the following indices were scored: Mayo score, Rachmilewitz, Lichtiger, Seo, Pediatric UC Activity Index (PUCAI), Partial Powell-Tuck, Endoscopic-Clinical Correlation Index (ECCI), Beattie, and Walmsley (SCCAI). Physician global assessment, colonoscopic score, blood tests and the Mayo score were used to assess construct and discriminative validity. Follow-up evaluation of 61 patients was used to assess test-retest reliability and responsiveness. Results: The PUCAI and Walmsley indices ranked the best for each of the four psychometric properties evaluated. For construct validity, the average correlation for the five included constructs were r=0.79 and r=0.80 for the PUCAI andWalmsley respectively (both p<0.001); discriminative validity AUC=0.95 (95% CI, 0.900.99) and AUC=0.93 (95% CI, 0.87-0.98); test-retest reliability ICC=0.89 (95%CI, 0.760.95) and ICC=0.94 (95% CI, 0.86-0.97) and responsiveness AUC=0.84 (95% CI, 0.720.96) and AUC=0.86 (95%CI, 0.74-0.99) respectively. The partial Mayo score ranked well on three of the four properties evaluated: construct validity r=0.79 (p<0.001); discriminative validity AUC=0.95 (95% CI, 0.90-0.99); and responsiveness AUC=0.84 (95%CI, 0.72-0.96). However, this index demonstrated inferior test-retest reliability ICC=0.82 (95% CI, 0.600.92). The Rachmilewitz index showed good validity (construct validity r=0.81 (p<0.001); discriminative validity AUC 0.92 (95% CI, 0.87-0.98) and reliability ICC=0.89 (95% CI, 0.76-0.95) but inferior responsiveness AUC=0.74 (95%CI, 0.58-0.90). Cutoff scores that define clinical-endoscopic remission and response were determined using ROC curve analysis for all instruments and were broadly similar to those cutoffs that had been empirically derived and used with each of the instruments. Conclusion: The Walmsley, PUCAI and partial Mayo are valid, reliable and responsive non-invasive indices to assess disease activity in adult UC. Given their robust psychometric properties, these instruments may permit less frequent endoscopic assessment in clinical practice and in research.