Endoscopic Mucosal Resection of Non-pedunculated Colorectal Polyps ≥20mm: Outcome in a Self-taught Skills Environment.

Background and aims: Endoscopic mucosal resection (EMR) of non-pedunculated colorectal polyps ≥20mm is technically demanding and should preferentially be performed by specialist endoscopists in referral centres. Little is known about the outcome in institutions establishing this competency. Here, we report the learning curve on 100 consecutive large non-pedunculated polyps resected by a single endoscopist with self-taught acquisition of skills.

Methods: We analysed data on 100 non-supervised EMR procedures performed at our academic endoscopy centre (2016-2021), representing a single endoscopist's learning curve beginning with the first polyp ≥20 mm.

Results: The median polyp size was 30 mm (20-70mm), and 61% of all polyps were ≥30 mm. Predominant polyp morphology was 0-Is (34%) or 0-IIa (47%), and most polyps developed in the ascending colon (36%). In total, 20% of polyps showed high-grade intraepithelial neoplasia, and 8% included pT1 carcinoma. Adenoma recurrence rate after piecemeal resection was 21%. All but one recurrent adenoma were treated endoscopically. Deep mural injury, intra-procedural bleeding and post-procedural bleeding were detected and managed endoscopically in 3%, 21%, and 4% of procedures, respectively. Overall, surgery could be avoided in 91% of all and 98% of non- malignant polyps. Results for the first 50 polyps did not differ from results for the following polyps.

Conclusions: Structured training is advisable to acquire advanced EMR skills. Our data show that autonomous acquisition of skills after finishing a training course represents an acceptable alternative with good results in the setting of an open error culture. Continuous review of outcome parameters and complication rate is mandatory during the learning process.