Structured Endoscopic Tip Control Training – An Ex-Vivo Model of Snare Tip Soft Coagulation Improves Participant Tip Control (HAM-2 Study)

Aims There are no objective performance metrics for basic endoscopic skills to guide training and career trajectory. The fidelity of endoscopic tip manipulation is a key skill akin to hand steadiness in surgery. Two recent studies using ex-vivo and in-vivo models of snare tip soft coagulation (STSC) applied to a polypectomy defect margin reliably stratified endoscopists by interventional profile, number, and complexity of polypectomies performed.

Methods Tip-control of 13 endoscopists with varied expertise was assessed using a STSC model containing 3D-printed figures constructed from 1-4 intersecting circles (radius=17.5mm) (shape 1=1 circle, shape 4=4 circles) printed onto the centre of a single piece of cooked ham. The study consisted of two assessments (= all 4 shapes, maximum 5 minutes/shape), 1 initial, two sessions of training on the model, and a second final after the training. Each assessments' video was rated three times by three different raters from a pool of 7 raters, for number of correct/incorrect hits and subjective tip-control score using a web application. Resulst are presented as estimated marginal means and correcponding confidence intervals from gneralized linear mixed models.

Results We observed that pre-training, experienced endoscopists achieved more correct hits per minute 42.8 (26–70.4) vs. 16.9 (14 – 20.4) in less experienced groups, leading to a higher effective speed (40.7 [23.2–71.4] vs. 19.7 [15.3–25.3] mm/min). Subjective impressions of tip control were also significantly higher in experienced endoscopists (97.3 [90.7–99.3] vs. 59.7 [50.2–68.5], p<0.001), as was accuracy (99.2 [97.4–99.8] vs. 90.6 [86.6–93.5], p<0.001). After two training sessions endoscopists were able to apply 6.1 (3.7-8.5) more correct hits per second and were 5.5 (3.1-7.9) mm/min faster, compared with pre-training. Accuracy did not increase after training (difference: 1.3 (-0.2-2.8), p=0.083). Baseline lower performance metrics (speed OR: -0.84 [-1.45 – -0.24], p=0.01 and accuracy OR: -1.64 [-2.62 – -0.67], p=0.002) were independent predictors of improvement in tip-control after training.

Conclusions An ex-vivo model to assess tip control was found to stratify endoscopists by expertise. Furthermore, performing two structured training sessions using the model led to a significant increase in the tip control of fellows, and trainees but not experts. These results have implications for standardising the assessment of endoscopic competency, benchmarking progress during training, and allowing expedited acquisition of complex endoscopic skills in an ex vivo setting.

Publikationsverlauf

Artikel online veröffentlicht:
27. März 2025