Validation of a high-fidelity simulator for diagnostic knee arthroscopy training

To validate a high-fidelity arthroscopic simulator by assessing its construct, face and content validity for diagnostic knee arthroscopy training.

Eighty-one participants (novices = 41, intermediates = 20 and experts = 20) performed a knee arthroscopy navigation task requiring identification of nine intra-articular landmarks, repeated for up to five trials with a 10-min limit per trial. Task-completion times and attrition were recorded. Performance was analysed using Kruskal-Wallis and Mann-Whitney tests with effect sizes. Post-task questionnaires evaluated face validity, content validity, educational value and user-friendliness.

Performance times differed significantly between experience groups (p < 0.001-0.043). Experts completed Trial 1 in roughly one third the time required by novices (90 s vs. 292 s; r = 0.78). All groups improved with repetition, with steep early gains in novices and the most consistent improvement among intermediates. Between-group differences narrowed in later trials, although interpretation is limited by attrition and survivorship bias, with only six novices completing Trial 5. Face validity ratings were high (mean > 7/10), and content validity was strongly endorsed (92%-97% supporting curricular integration). Educational value was rated positively (92.3-97.4%), whereas user-friendliness received lower agreement (35.9%).

This high-fidelity simulator demonstrated strong construct, face and content validity, reliably discriminating between experience levels and supporting early-phase skills acquisition. Strong realism and educational value ratings support its potential role in structured arthroscopic training, although usability findings highlight opportunities for ergonomic and instructional refinement. Future research should assess transfer validity and operative performance impact.

Level III, prospective observational validation study.