Ben Selvan, MD, Amy Cha, MD, Jon B Morris, MD, Joe Palmerie, MD, Krisoffel Dumon, MD, Noel Williams, MD, Gary Korus, MD. Penn Medicine Clinical simulation center, Hospital of University of Pennsylvania; Philadelphia
Background: The Surgical Council on Resident Education has identified endoscopy as a “common” or “essential” procedure that requires competency by the end of training. While this competency should be attainable primarily by case volume, limited patient resources mandate efficient training protocols.
AIM. Development of an efficient introductory curriculum for resident training in basic endoscopy skills.
Methods
PGY1 and PGY2 surgical residents participated in the study during their skills sessions on endoscopy at Penn Medicine Clinical Simulation Center. The interventions used were EndoBubble exercise on GI Mentor™ and low fidelity colon model as described in ACS/APDS skills curriculum. The GI Mentor™ exercise has simulated colon with 20 balloons on the mucosa that have to be popped with the simulated endoscopy probe without touching the mucosa [error]. The residents were given 1-4 attempts to reach the proficiency time of 90 seconds. In the low fidelity model, we embedded 17 letters from A to Q over the simulated mucosa and the residents have to find them using a colonoscopy. The proficiency level set was to identify 12. The residents were divided into two groups. Group I did the GI Mentor™ module first followed by the colon model and Group II did the colon module first followed by GI Mentor™. Errors in tasks I were defined as the learner touching the colonic wall/missing the balloons which were recorded by the system automatically. Time required to complete the tasks and the error committed were analyzed with student t test.
Results
20 residents completed the study. Mean time taken by GI to complete the tasks on GI mentor was 156s; 163.1s and 156.5s were duration for the first and second attempt which was not statistically significant [P 0.28]. Group II had mean time of 140.8s; [166.1 and 135.5]. The average errors made in the first did not improve on subsequent attempts. In colon model, the average letters identified was 12 and required time to complete the tasks was 337 seconds on first attempt by GII and GI was 376s. There was no improvement in the outcome between G I first attempt and GII subsequent attempt on GI mentor [p 0.18] though low fidelity model has shortened the GI mentor tasks by 23s.
| Tasks | GI Mentor | Low fidelity colon model |
| GI –Bubble pop first | 156 [163.1,156.5] | 337.33 |
| G II- finding the alphabets first | 140.8[146.1, 135.5] | 376 |
| Proficiency level | > 90 s | >12 letters to be identified |
Conclusion; It is feasible to implement endoscopy curriculum for surgical residents. The tool could be either low fidelity or virtual trainer such as GI mentor. These tasks could form the basis of endoscopy skills evaluation to certify a resident/physician. Our data suggests that a low fidelity model may be superior though it needs to be studied on a larger number. When a low fidelity model has the added benefit of developing familiarity with the same instruments to be used in patient care there may be improved transference of skills.
Session: Poster
Program Number: P189
View Poster
