Training on an ex vivo animal model improves endoscopic skills: a randomized, single-blind study

Abstract

Background: Animal models are used for training of different endoscopic procedures. Whether this really improves endoscopic skills remains controversial.

Objective: To assess the effectiveness of training by using an ex vivo animal gastric model on the performance of two therapeutic procedures-hemostasis and treatment of perforation.

Design: A randomized, single-blind study.

Setting: An experimental endoscopy center in a university hospital.

Participants: Thirty-one gastroenterology fellows with comparable endoscopic experience.

Methods: Participants were randomized into two groups: with (T, n = 16) and without (S, n = 15) training. All fellows continued with standard endoscopic practice. Baseline skills were assessed at enrollment. All physicians in group T underwent 2 full days of a hands-on course over a 3-month period, in addition to their standard endoscopic practice. Both groups then underwent a blinded, final evaluation. Endoscopic skills were scored from 1 (best) to 5 (poorest) by two expert, blinded tutors. Outcomes of clinical hemostatic procedures also were analyzed.

Main outcome measurements: Successful hemostasis and successful perforation closure.

Results: Thirty physicians completed the study. Hemostasis results (n = 15): The number of physicians who carried out a successful hemostasis procedure increased significantly in the group with training (27% vs 73%; P = .009) but did not change in the group without training (20% vs 20%). The mean scores of injection and clipping technique improved significantly only after training. The number of clips used decreased significantly only in the group with training; the time of clipping did not change significantly in either group. Perforation results (n = 15): The number of physicians with a successful and complete perforation closure increased nearly significantly in the group with training (40% vs 73%, P = .06) as opposed to the group without training (27% vs 47%; P = .27). The procedure time decreased significantly in the group with training only. In clinical practice, fellows in group T had a significantly higher success rate with respect to hemostatic procedures (83.2%, range 67-100 vs 63.6%, range 25-100; P = .0447). The majority of participants (93%) agreed that such courses should be compulsory in gastroenterological credentials.

Limitations: A retrospective analysis of clinical outcomes. Clinical outcome data were based on self-reporting of the participants.

Conclusion: Hands-on training by using an animal ex vivo model improves endoscopic skills in both hemostasis and perforation closure. In clinical practice, the training improves the outcome of hemostatic procedures.