Article Text

Download PDFPDF
PTH-119 A Comparison of Colonoscopists Performance in Recording and Labelling Images Taken During Endoscopy Using The Recorded Image Quality Index (RIQI) Tool
  1. D Samuel,
  2. JB Berrill,
  3. J Hurley,
  4. B Lee,
  5. E Hawkes,
  6. S Yosief,
  7. N Hawkes
  1. Gastroenterology, Cwm Taf University Health Board, Mid Glamorgan, UK


Introduction Recording images on the endoscopic reporting tool and report writing are key skills for a colonoscopist. Together they represent the data set on which other clinicians have to base decisions. The performance level of colonoscopists in this area of endoscopy is not commonly measured. We therefore aimed to examine the variation in image recording amongst a cohort of colonoscopists performing regular colonoscopy using the validated Recorded Image Quality Index (RIQI) tool.

Methods We searched the HICCS Endoscopic Reporting System for all colonoscopists performing regular colonoscopy (n = 11). All procedures performed between July and December 2015 were identified (screening cases were excluded). All images and the endoscopy report for the first 10 cases with pathological findings for each colonoscopist were obtained, ordered into folders and data anonymised. The 110 image sets were scored by 3 independent assessors using the validated RIQI score. This measures 4 domains representation, labelling, caecal landmarks and image quality high scores indicate high utility of the images as a base for decision-making. Kruskal-Wallis non-parametric test was used to compared differences in rank sums between groups.

Results 110 data sets were reviewed by 3 assessors generating 330 RIQI scores. These observations were pooled for each colonoscopist yielding a median RIQI score from 30 observations. Inter-observer rating scores for the 3 assessors were in the moderate to good range. Median values for colonoscopists ranged from 2 (scores less than 4 indicate a < 2% utility value as a base for decision-making) to 10 (scores 9–10 have a 96.5% utility rate). Upper quartile performers on RIQI score were significantly better than the lower quartile of colonoscopists (Kruskal-Wallis, H = 1132 (1, n = 180, p < 0.00001). 4/11 (37%) colonoscopists met acceptable standards (median 9–10), 5/11 (45%) were rated as needing improvement (median 6–8) and 2/11 (18%) demonstrated poor performance (median ≤5).

Conclusion The RIQI tool demonstrates widely varying performance in capture of quality images during endoscopy. High quality image capture and reporting is an important aspect of endoscopic practice. RIQI provides a performance indicator that can be used as an audit and training tool to improve performance in this area of practice.

Disclosure of Interest None Declared

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.