Although there is growing evidence that practice on bench model simulators can improve the acquisition of technical skill in surgery, the degree to which these models have to approximate real-world conditions (model fidelity) to optimize learning is unclear. Previous research suggests that low-fidelity models may be adequate for novice learners. The purpose of this study was to assess the effect of model fidelity and surgical expertise on the acquisition of vascular anastomosis skill.Methods:
Twenty-seven surgical residents participated in this institutional review board-approved study. Junior residents (postgraduate year 1 and 2) and senior residents (postgraduate year 4 or higher) were randomized into two groups: low-fidelity (n = 13) and high-fidelity (n = 14) model training. Both groups were given a 3-hour hands-on training session: the low-fidelity group used plastic models, and the high-fidelity group used human cadaver arms (brachial arteries) to practice graft-to-arterial anastomosis. One week later, all subjects participated in an animal laboratory in which they performed a single vascular anastomosis on a live, anesthetized pig (femoral artery). A blinded vascular surgeon scored candidate performance in the animal laboratory by using previously validated end points, including a checklist and final product analysis score.Results:
Acquisition of skill was significantly affected by model fidelity and level of training as measured by both the checklist (P= .03) and final product analysis (P= .01; Kruskal-Wallis). Specifically, junior residents practicing on high-fidelity models scored better on the checklist (P= .05) and final product analysis (P= .04). Senior residents practicing on high-fidelity models scored better on final product analysis (P< .05).Conclusions:
Training in the laboratory does improve skill when assessed in a realistic setting. Both expertise groups showed better skill transfer from the bench model to live animals when practicing on high-fidelity models. For vascular anastomosis, it is important to provide appropriate model fidelity for trainees of different abilities to optimize the effectiveness of bench model training.