Validity and reliability of a virtual reality game in evaluating the projected frontal plane knee angle when landing from a drop vertical jump

Objectives: To determine the validity and reliability of the peak frontal plane knee angle evaluated by a virtual reality (VR) netball game when landing from a drop vertical jump. Study Design: Laboratory. Methods: Forty participants performed 3 drop vertical jumps evaluated by 3-dimensional motion analysis and 3 drop vertical jumps evaluated by the VR game. Limits of agreement for the peak projected frontal plane knee angle and peak knee abduction were determined. Participants were given a consensus category of "above threshold" or "below threshold" based on a prespecified threshold angle of 9° during landing. Classification agreement was determined using kappa coefficient, and accuracy was determined using specificity and sensitivity. Ten participants returned 1 week later to determine intrarater reliability, standard error of the measure, and typical error. Results: The mean difference in detected frontal plane knee angle was 3.39° (95% confidence interval [CI], 1.03° to 5.74°). Limits of agreement were -10.27° (95% CI, -14.36° to -6.19°) to 17.05° (95% CI, 12.97° to 21.14°). Substantial agreement, specificity, and sensitivity were observed for the threshold classification (κ = .66; 95% CI, .42 to .88; specificity = 0.96; 95% CI, 0.78 to 1.0; and sensitivity = 0.75; 95% CI, 0.43 to 0.95). The game exhibited acceptable reliability over time (intraclass correlation coefficient, ICC_3,1 = .844), and error was approximately 2°. Conclusion: The VR game reliably evaluated a projected frontal plane knee angle. Although the knee angle detected by the VR game is strongly related to peak knee abduction, the accuracy of detecting the exact angle was limited. A threshold approach may be a more accurate approach for gaming technology to evaluate frontal plane knee angles when landing from a jump.