# Does development of submucosal dissection models influence quality of training? Comparison of existing models

## TL;DR

This study compares different training models for endoscopic submucosal dissection and finds that ex-vivo models are most effective for skill development.

## Contribution

The study provides a direct comparison of ESD training models and identifies the most effective ones for skill progression.

## Key findings

- Ex-vivo models are effective for learning ESD skills.
- The bovine colon model is the most discriminating for skill assessment.
- Synthetic models are better suited for beginners.

## Abstract

Use of endoscopic submucosal dissection (ESD) is growing, but access to it remains limited. The aim of this study was to compare the performance of various existing models and the progress made by students on them.

Four training models (bovine colon (ex-vivo), ex vivo porcine model, live porcine model, and an artificial model (Endogel)) were evaluated during a 1-week training course. Each participant was evaluated at the beginning (D1) and at the end of the training (D5). Learners performed a standardized ESD of 2 cm on the four models in a randomized order. Experts evaluated the ability of participants to perform ESD using the objective structured assessment of technical skill score (OSATS).

Sixteen students were involved, the average age was 35.6 years (+/- 4.6) and they practiced endoscopy for 10 years (+/-5.3). The OSATS significantly increased in each model during the week, with mean scores increasing from 8.6 to 23.3, from 10.7 to 12.9, from 8.8 to 21.3 and from 8.2 to 12.5 for the bovine colon, ex vivo porcine model, live porcine model, and Endogel models, respectively.

Ex-vivo models are good models for learning ESD skills. The bovine colon model seems to be the most discriminating. Synthetic models should be reserved for novices.