James Ansell, Joanna Hurley, James Horwood, Chantelle Rizan, Konstantinos Arnaoutakis, Stuart Goddard, Neil Warren, Jared Torkington
Welsh Institute for Minimal Access Therapy, University Hospital Llandough, University Hospital of Wales
Objectives: The aim of this study is to establish if the WIMAT colonoscopy suitcase has construct and concurrent validity for colonic polypectomy training. This is a novel ex-vivo porcine animal simulator which can be used to recreate a range of polypectomy procedures.
Methods and procedures: A prospective cross sectional study design was used to assess 40 participants. This included 10 “novices” (junior doctors with no prior experience of endoscopy), 10 “intermediates” (Surgical / Gastroenterology registrars), 10 “advanced” (Joint Advisory Group on GI Endoscopy (JAG) accredited independent colonoscopists) and 10 “experts” (JAG Bowel screening colonoscopists). All subjects completed a standardised, video recorded polypectomy task. This involved snare polypectomy and Roth net retrieval of 2 pedunculated polyps with 2 degrees of difficulty; polyp A (simple) polyp B (complex). Two JAG accredited colonoscopists, independent to the trial and blinded to group allocation scored the videos according to 8 previously validated Direct Observation of Polypectomy Skills (DOPyS) parameters. Real-life DOPyS scores from the intermediate group were compared with their simulator DOPyS results. Median overall group scores with Inter Quartile Ranges (IQR) were compared using a Wilcoxon matched-pair signed-rank test for non-parametric data. The Kappa statistic was used to assess Interrater reliability.
Results: Median completion times (seconds) for polyp A were 472s (IQR 317-600) for novices, 324s (IQR 251-424) for intermediates, 263s (IQR 197-299) for advanced and 355s (IQR 182-457) for experts. Polyp B completion times were 600s (IQR 536-600) for novices, 599s (IQR 460-600) for intermediates, 364s (IQR 243-455) for advanced and 448s (IQR 382-528) for experts. Median overall DOPyS scores (out of 4) for polyp A were 1.00 (IQR 0.88-1.63) for Novices, 2.00 (IQR 1.88-2.63) for Intermediates, 3.00 (IQR 2.88-3.5) for Advanced and 3.25 (IQR 3.00-4.00) for Experts (N vs I p=0.02, N vs Ad p=0.005, N vs E p=0.005, I vs Ad p=0.04, I vs E p= 0.006 and Ad vs E p=0.196). Median overall DOPyS scores for polyp B were 0.50 (IQR 0.38-1.00) for Novices, 1.25 (IQR 0.50-2.50) for Intermediates, 2.75 (IQR 2.00-3.50) and 3.0 (IQR 2.5-4.00) for Experts (N vs I p=0.033, N vs Ad p=0.007, N vs E p=0.005, I vs Ad p=0.041, I vs E p= 0.08 and Ad vs E p=0.150). Interrater reliability was favourable (k= 0.5). There was no difference between median, overall real-life DOPyS scores and those obtained on the simulator (polyp A p=0.69 and polyp B p=1.00).
Conclusion: This simulator demonstrates good construct and concurrent validity for use in colonic polypectomy training. The model can be reliably adjusted to provide varying degrees of difficulty according to the experience of the user.
Session: Poster Presentation
Program Number: P144
