Diagnosing pouchitis: comparative validation of two scoring systems in routine follow-up

Pouchitis represents a serious threat to patients with ulcerative colitis after restorative proctocolectomy with ileal pouch-anal anastomosis. The frequency of pouchitis is high, and it implies the risk of pouch failure and the risk of malignant mucosal transformation in the pouch. Early detection and precise classification of the inflammatory process are required for adequate therapy, which might be facilitated using a scoring system. The aim of the present study was to validate two existing scoring systems in routine outpatient practice.

The Heidelberg Pouchitis Activity Score and the Pouchitis Disease Activity Index developed at the Mayo Clinic were simultaneously prospectively applied in a consecutive series of 103 outpatient consultations of 41 patients at our hospital and comparatively validated against the diagnosis of " pouchitis" or "no pouchitis" concurrently made by a physician and a surgeon.

The median score of examinations in which the clinicians' diagnosis was consistent with pouchitis were significantly higher than those of examinations inconsistent with pouchitis in both scoring systems (Heidelberg Pouchitis Activity Score, 17 (interquartile range, 14-21) and 8 (interquartile range, 5-10), respectively, P < 0.001; Pouchitis Disease Activity Index, 7 (interquartile range, 5-8) and 2.5 (interquartile range, 1-4), respectively, P < 0.001). The sensitivity and specificity in the two total scores were 84 and 79.5 percent, respectively (Heidelberg Pouchitis Activity Score), and 60 and 96.2 percent, respectively (Pouchitis Disease Activity Index); in the field clinical manifestations 44 and 73.1 percent, respectively (Heidelberg Pouchitis Activity Score), and 20 and 87.2 percent, respectively (Pouchitis Disease Activity Index); in the field endoscopic manifestations 88 and 83.3 percent, respectively (Heidelberg Pouchitis Activity Score), and 60 and 89.7 percent, respectively (Pouchitis Disease Activity Index); and in the field histologic manifestations 72 and 76.9 percent, respectively (Heidelberg Pouchitis Activity Score), and 44 and 96.2 percent, respectively (Pouchitis Disease Activity Index). Lowering the cutoff point for diagnosis of pouchitis in the Pouchitis Disease Activity Index by 2 points (pouchitis: score >or= 5) would result in an 88 percent sensitivity and a 67 percent specificity.

Specificity and sensitivity of the Heidelberg Pouchitis Activity Score were satisfactory. The cutoff point for diagnosing pouchitis in the Pouchitis Disease Activity Index would have to be lowered to reach an acceptable sensitivity and specificity. The very poor validity of the field clinical manifestations in diagnosing pouchitis emphasizes the need for endoscopic and histologic examination for detection of pouchitis. The issue of whether the diagnosis of pouchitis should be based on endoscopic and histologic features alone, instead of additionally taking clinical features into account, should be addressed in future studies.