PTH-013 Development of the upper GI recorded image quality index (UGI-RIQI) score and quality assurance tool

Background Endoscopic images saved on the Electronic Reporting System are the only visible representation of completeness of examination and pathological findings. Together with the endoscopy report these become the only reference for other clinicians not present at the original endoscopy on which to base further decisions. Standards for image recording form a component part of the Quality Standards in Upper Gastrointestinal (UGI) endoscopy.1 Aims and methods We aimed to develop a systematic scoring system for quality of images recorded at UGI endoscopy and validate this UGI Recorded Image Quality Index (UGI-RIQI) scoring system. We searched the HICSS Endoscopic Reporting System for endoscopists performing regular UGI endoscopy (n=14) between January and June 2018. All images and the endoscopy report for the first 10 cases with pathological findings for each endoscopist were obtained, ordered into folders and the data anonymised. An UGI-RIQI scoresheet was devised, based on the validated lower GI RIQI tool,2 3 assessing 4 domains: Representation, Image Labelling, Extent of examination and Image Quality, and the clinical utility (CU) of the image set - rating its ability to inform further decision-making. The UGI-RIQI total score range was 0 to 12. 140 image sets were scored by 3 independent assessors. Cohen’s kappa values for intra observer variation were calculated for individual domains and total RIQI scores. These results informed recommended RIQI standard levels of performance. The correlation of these levels and CU scores was tested with Spearman’s test. Results 140 data sets were reviewed by 3 assessors generating 420 domain scores. Inter-rater agreement (IRA) for assessors for the total RIQI score were in the moderate to good range (0.6, 0.46 and 0.47). Performance levels were defined in terms of total RIQI score: poor 0–6, below standard 7–8 and meets standards 9–12. The correlation between the derived RIQI levels and clinical utility scores were high (0.71, 0.64 and 0.71). Conclusions The UGI-RIQI tool provides a method for assessing the quality of image capture across ten procedures with scores in 4 domains. The UGI-RIQI score correlates well with clinical utility of the images, with acceptable inter-rater reliability. It shows potential both as an audit and training tool to improve performance in this area of endoscopic practice. References Beg S, et al. Gut 2017;0:1–14. doi:10.1136/gutjnl-2017–314109. Samuel D, et al. Gut 2016; 65:Suppl 1A224-A225doi:10.1136/gutjnl-2016-312388.420. Samuel D, et al. Gut 2016;65:Suppl 1 A278doi:10.1136/gutjnl-2016-312388.522.