Comparing the efficacy of immersive virtual reality and traditional physical arthroscopy simulators in arthroscopic training: a randomized controlled trial

Virtual reality (VR)–based simulation is increasingly integrated into surgical education; however, its role relative to traditional physical arthroscopy simulators for early arthroscopic skill acquisition remains incompletely defined. This randomized controlled trial compared training metrics and test performance following immersive virtual reality (IVR) training versus a non-anatomic physical arthroscopy simulator in arthroscopy-naïve medical trainees. In this single-center randomized controlled trial (NCT05830786), novice medical students were allocated to one of four groups: three IVR training sessions (IVR3), a single IVR training session (IVR1), one physical arthroscopy simulator session, or no training. IVR training performance metrics were recorded across sessions. All participants subsequently performed a standardized diagnostic knee arthroscopy on an anatomical model and were assessed by a blinded evaluator using a Global Rating Scale (GRS). Task completion time and participant-reported confidence and experience were also evaluated. Twenty-nine participants were randomized and 28 completed the post-training assessment and were included in the complete-case analysis. The IVR3 group demonstrated significant within-group improvement across repeated training sessions in precision, scope manipulation, and object-tracking metrics ( p < 0.001). At testing, no statistically significant differences were detected between IVR and physical simulator groups across all GRS domains. Participants trained on the physical simulator demonstrated higher camera and instrument dexterity scores compared with untrained controls. Task completion time differed significantly between groups, with the IVR3 cohort completing the assessment faster than IVR1 and untrained participants. Self-reported confidence did not differ between groups. Qualitative feedback highlighted perceived advantages of physical simulators in realism and of IVR in accessibility and repeatability. In arthroscopy-naive trainees, performance at diagnostic knee arthroscopy assessment did not differ significantly between immersive virtual reality and non-anatomic physical arthroscopy simulator training, while repeated IVR exposure was associated with improved efficiency. As the trial was small and was not designed as an equivalence study, these findings indicate the absence of a detectable difference rather than evidence of equivalence. These findings are consistent with a potential role for IVR as a scalable adjunct for early arthroscopic skill development, although limitations related to task comparability, anatomic fidelity, and the absence of validated arthroscopy-specific assessment tools should be considered when interpreting comparative effectiveness.