Roel Bogie

Development and validation of an educational web-based system for endoscopic classification of laterally spreading tumors

international study on IOA of endoscopic classification of LSTs in the colon and rectum. To simulate the real-life practice, we assembled a systematic case-collection of LSTs. For didactic purpose, only high-quality images were used. We developed a case-based system for LST classification with e-learning and validated it among endoscopy fellows at two different centers. This educational program had significantly improved the performance of endoscopy trainees and can now be used worldwide. It should be taken into account that the durability of the learning effect has not been assessed in the present study. The effect of not using chromoendoscopy on the IOA of the endoscopic Kudo classification remains unknown, although we expect the IOA to be lower. Chromoendoscopy is not commonly used in all centers worldwide, but is advised by experts for lesion characterization. 28 In the present study, only images of chromoendoscopy with contrast dye were used. The effect of using digital chromoendoscopy instead of dye chromoendoscopy on the IOA of the endoscopic Kudo classification remains to be clarified. Because the availability of ESD differs among centers, both experts’ and trainees’ views on most suitable resection technique could have been biased. In this study, we compared the IOA of the fellows for LST classification with that of experts, who were considered to be the reference. Because of the lack of a gold standard, diagnostic accuracy could not be calculated. The IOA of the endoscopic Kudo LST classification was examined instead. Intraobserver agreement (the consistency in scoring of individual raters in a single case) could not be assessed because of single measurements. The use of short endoscopy movies showing the LST instead of still images could further improve the classification. Of note, IOA was lower in the cases scored as ‘of insufficient quality’ by at least one rater. Future studies should use high-definition endoscopy movies that mimic real life better. A more objective assessment method to identify pseudo-depressions, dominant nodules, and subtle granulesmay help to improve the IOA of the endoscopic Kudo LST classification. Identification of a dominant nodule may become more straightforward when using 5mm as cut off for a dominant nodule, as suggested by the study of Shigita et al. 12 The risk of SMI is leading in the treatment strategies 1 and is higher for LST-G with dominant nodules of at least 5mm. Small granules may be difficult to distinguish with standard colonoscopy while dye-based chromoendoscopy provides more detail. Notably, recognition of pseudo-depression remains a challenge for both experts and (trained) endoscopy fellows. Conclusion In summary, our study confirms that the Kudo endoscopic classification of LSTs is a practical tool. Substantial IOA was found for classification of LSTs among international experts. A specific designed e-learning significantly improved the IOA among fellows. Training of endoscopists in applying the endoscopic Kudo classification is a meaningful first step in improving endoscopic treatment of LSTs.

2

33