Utility of Modern Arthroscopic Simulator Training Models: A Meta-analysis and Updated Systematic Review

作者：Frank Rachel M^*; Wang Kevin C; Davey Annabelle; Cotter Eric J; Cole Brian J; Romeo Anthony A; Bush Joseph Charles A; Bach Bernard R Jr; Verma Nikhil N

来源：Arthroscopy-The Journal of Arthroscopic and Related Surgery, 2018, 34(5): 1650-1677.

DOI：10.1016/j.arthro.2017.10.048

Purpose: To determine the utility of modern arthroscopic simulators in transferring skills learned on the model to the operating room. Methods: A meta-analysis and systematic review of all English-language studies relevant to validated arthroscopic simulation models using PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines from 1999 to 2016 was performed. Data collected included the specific simulator model, the joint used, participant demographic characteristics, participant level of training, training session information, type and number of tasks, pre-and post-training assessments, and overall outcomes of simulator performance. Three independent reviewers analyzed all studies. Results: Fifty-seven studies with 1,698 participants met the study criteria and were included. Of the studies, 25 (44%) incorporated an arthroscopic training program into the study methods whereas 32 (56%) did not. In 46 studies (81%), the studies' respective simulator models were used to assess arthroscopic performance, whereas 9 studies (16%) used Sawbones models, 8 (14%) used cadaveric models, and 4 (7%) evaluated subject performance on a live patient in the operating room. In 21 studies (37%), simulator performance was compared with experience level, with 20 of these (95%) showing that clinical experience correlated with simulator performance. In 25 studies (44%), task performance was evaluated before and after simulator training, with 24 of these (96%) showing improvement after training. All 4 studies that included live-patient arthroscopy reported improved operating room performance after simulator training compared with the performance of subjects not participating in a training program. Conclusions: This review suggests that (1) training on arthroscopic simulators improves performance on arthroscopic simulators and (2) performance on simulators for basic diagnostic arthroscopy correlates with experience level. Limited data suggest that simulator training can improve basic diagnostic arthroscopy skills in vivo.