The learning experience for endoscopic submucosal dissection in a non-academic western hospital: a single operator's untutored, prevalence-based approach.

BACKGROUND: Studies of learning experience in endoscopic submucosal dissection (ESD) commonly originate from the East. Little is known about the performance of ESD in low-volume western centers. Furthermore, it is unclear whether ESD can be self-taught without a tutored approach. METHODS: We performed a retrospective analysis of consecutive ESDs, performed in an untutored prevalence-based fashion by a single operator at a private Greek hospital from 2016-2020. Out of 60 lesions, standard ESD was applied for 54 and enucleation for 6; 41 were mucosal and 19 submucosal; 3 esophageal, 24 gastric, one duodenal, 12 colonic, and 20 rectal. RESULTS: Pathology revealed carcinoma (n=14), neuroendocrine tumor (n=7), precancerous lesion (n=27), or other submucosal tumors (n=12). The rates of en bloc and R0 resection were 98% and 91%, respectively. The median resection speed was <3 cm2/h for the first 20 cases, but improved progressively to ≥9 cm2/h after 40 cases. Two patients underwent laparoscopic surgery for colonic perforation, and one received a blood transfusion because of delayed bleeding (serious adverse event rate: 5%). No deaths occurred. The median hospital stay was 1.3 days. Variables associated with improvement in ESD speed during the second period of the study were the application of countertraction and the experience acquired through other endosurgical techniques. CONCLUSIONS: ESD was safe and effective in a low-volume center, with an acceptable adverse events rate. At least 40 mixed cases were needed to achieve a high resection speed. Additive experience gained through other endosurgical procedures probably contributed to the improvement in performance.