EP117 Virtual or Augmented Reality in Hip Arthroscopy Training Demonstrates Adequate Face, Content, and Construct Validity: A Systematic Review with Expert Commentary

To summarize literature outlining the efficacy and validity of virtual reality (VR) and augmented reality (AR) simulation in hip arthroscopy training.

Three databases were searched on February 9th, 2025 for studies investigating the validity of VR/AR simulation for hip arthroscopy training. Data on participant details, simulation specifics, and validity (e.g. face, construct, and content) were included. Descriptive statistics were used to report results.

Nine studies (one level I, one level II, seven level IV) comprising 218 participants were included, 37 (16.9%) being considered as “experts”. The majority of studies (7/9; 77.8%) were analyses of one session, with two studies comparing progress during sessions over time. Five studies investigated face validity, where the majority of participants (>70%) in each study found that the simulators were realistic (>7/10 or >6/7 via Likert scale) in all elements apart from tactile feedback. Eight studies evaluated construct validity. Four studies compared the safety performance of experts and non-experts, three (75%) finding statistically lower amounts of iatrogenic tissue damage in the expert group (p<=0.002). Three of four (75%) studies found that experts had lower completion times in either a diagnostic or surgical module (p<=0.03) than non-experts. One study reported lower iatrogenic damage via soft-tissue or bone collisions (p<0.0001) and completion times (p<0.001) in the seventh session of VR simulation, relative to the initial first three sessions in junior learners.

VR/AR simulation training for hip arthroscopy demonstrates adequate face, content, and construct validity. Generally, VR/AR simulations were rated realistic by users apart from tactile feel. Participants, particularly junior learners improved in minimizing iatrogenic tissue damage and decreasing completion times over multiple sessions. Experts generally outperformed non-experts in the majority of performance metrics, suggesting high construct validity.