The primary objective SP600125 of the present study was to assess the reliability of UCEIS scoring and perform an initial validation in an independent cohort of videos and investigators after appropriate training. Secondary objectives included an assessment of the impact of endoscopists’ knowledge

of clinical details on the evaluation of endoscopic disease severity. For consistency in the text, the word “index” refers to an instrument for assessing activity, “descriptor” refers to an item within that index with severity allocated on a Likert scale, and “level” refers to the severity graded for an item. “Score” is the overall measure provided by an index. Initial development of the UCEIS has been reported.6 In brief, a library of 670 video sigmoidoscopies from patients with

Mayo Clinic scores of 0 to 11, supplemented by 10 videos from 5 people without UC and 5 hospitalized patients with acute, severe UC, was used. Phase 1 mapped inconsistency in overall endoscopic assessment of 16 of 24 video sigmoidoscopies by specialists (the clinical authors) and defined word for word by common agreement 10 endoscopic descriptors that evaluated components of the visual image. check details Phase 2 was conducted in a separate cohort of 30 investigators from 13 countries. The investigators rated descriptors in 25 of 60 randomly assigned videos and assessed overall endoscopic severity on a VAS from 0 to 100. An index (the UCEIS) consisting of the sum of 3 descriptors, each with 3 or 4 levels of severity, was then constructed that could be tested for reliability Bcl-w and validation (Table 1). Interobserver and intraobserver variations in these descriptors were also quantified. Phase 3 of the study is reported here. Investigators were recruited to reflect a range of geographic and institutional characteristics (see Acknowledgments) from gastroenterologists known to have endoscopic training in trials of inflammatory bowel disease or known to the authors to have an interest in endoscopy and inflammatory bowel disease. Each investigator was then

further trained to ensure consistency in understanding and use of the descriptors for assessing endoscopic severity. Training involved assessing video clips of each descriptor at each level, each with an agreed definition of severity. During training, investigators scored 4 standardized videos from phase 2 that included characteristics of the 3 descriptors. To qualify, investigators had to identify correctly the level of the descriptor “erosions and ulcers” on each video and the descriptors “vascular pattern” and “bleeding” within one level of the correct response on each video. Investigators failing to qualify at first assessment were permitted one retest that consisted of correctly scoring 2 of 3 different examples (from different videos) of the descriptor(s) that they had previously incorrectly scored.