Simulation Training Improves Surgical Proficiency and Safety During Diagnostic Shoulder Arthroscopy Performed by Residents

摘要

Although virtual reality simulators have established construct validity, no studies have proven transfer of skills from a simulator to improved in vivo surgical skill. The current authors hypothesized that simulation training would improve residents' basic arthroscopic performance and safety. Twenty-two orthopedic surgery trainees were randomized into simulation or standard practice groups. At baseline testing, all of the participants performed simulator-based testing and a supervised, in vivo diagnostic shoulder arthroscopy with video recording. The simulation group subsequently received 1 hour of total instruction during a 3-month period, and the standard practice group received no simulator training. After intervention, both groups were reevaluated with simulator testing and a second recorded diagnostic shoulder arthroscopy. Two blinded, independent experts evaluated arthroscopic performance using the anatomic checklist, Arthroscopic Surgery Skill Evaluation Tool (ASSET) score, and total elapsed time. All outcome measures were compared within and between groups. After intervention, mean time required by the simulation group to complete the simulator task (30.64 seconds) was 8 +/- 1.2 seconds faster than the time required by the control group (38.64 seconds; P=. 001). Probe distance (51.65 mm) was improved by 41.2 +/- 6.08 mm compared with the control (92.83 mm; P=. 001). When comparing ASSET safety scores, the simulation group was competent (3.29) and significantly better than the control group (3.00; P=. 005) during final arthroscopic testing. This study establishes transfer validity for an arthroscopic shoulder simulator model. Simulator training for residents in training can decrease surgical times, improve basic surgical skills, and confer greater patient safety during shoulder arthroscopy.