VALIDATION OF AN ONLINE LEARNING TOOL FOR OPTICAL DIAGNOSIS OF DIMINUTIVE COLORECTAL POLYPS

Aims 1) To assess the efficacy of an online self-learning tool for OD of diminutive colorectal polyps; 2) To describe in detail the learning process and possible variables related to learning efficacy.

Methods Online web page-based tool divided in five sections: 1) pre-tutorial test; 2) OD tutorial (NICE and WASP classifications); 3) First post-tutorial test; 4) Second post-tutorial test (different images); 5) Test with videos. The OD learning tool is prepared to gather information of each student individually as they pass through every phase. For each phase, number of right answers, percentage of high confidence diagnosis and number of attempts to pass the test were collected. Each phase was passed when there were at least 90% corrects answers. A multicenter study was designed to assess the performance of the OD tool on endoscopists of different experiences and background.

Results 36 endoscopists from 7 different Spanish centers entered the study [23 (63,9%) endoscopists; 5 (13,9%) gastroenterologists; 8 (22,2%) fellows].

Mean difference on correct answers between pre- and post-tutorial tests was 5 (0-10), and mean difference on the percentage of high-confidence diagnosis was 19,7 (-8 - 56).

Tab. 1 Overall results of the learning process through the online tool

	Pre-tutorial test	Post-tutorial test	1st pass of the 2nd post-tutorial test	1st pass of the video test
Correct answers, mean (range)	19.0 (13.0-25.0)	24.0 (23.0-25.0)	25.8 (20.0-30.0)	24.6 (18.0-29.0)
High-confidence diagnosis, %(range)	67.5 (32.0-92.0)	87.3 (64.0-100)	78.7 (53.0-100)	75.2 (37.0-97.0)
Students passed at 1st attempt, n (%)	-	24 (66.7)	18 (50.0)	6 (16.7)
Students needing 3 attempts to pass		2 (5.6)	1 (2.8)	7 (19.4)

Previous endoscopic experience had no influence on the efficacy of learning.

Conclusions 1) Our web-based online tool is efficient for OD learning of diminutive colorectal polyps; 2) A drop in the diagnostic efficacy occurs when moving from evaluating still pictures to videos. Therefore any learning tool has to include a video phase; 3) Efficacy results are homogeneous not depending on previous endoscopic experience.