Functional gait assessment : Reliability and validity in people with peripheral vestibular disorders

Background/aim: Reliable and valid measures evaluating imbalance during walking are important for targeting rehabilitation efforts. This study aimed to investigate the reliability and validity of the Functional Gait Assessment (FGA) in people with peripheral vestibular disorder. Methods: This study used a cross-sectional design and tested the reliability and validity of the FGA in 43 participants with peripheral vestibular disorder. The participants' performance on the FGA was filmed. To test intra-rater reliability, three raters reassessed 39 of these participants using the video uptakes. Interrater reliability was assessed by five raters analysing the video uptakes. Concurrent validity was investigated using timed measures of dynamic balance and the Activities-specific Balance Confidence scale. Results: The intra-rater reliability for the FGA score was high, with a intraclass correlation coefficient (ICC) (model 2,1)=0.94 (95% CI=0.85; 0.97). The intra-rater percentage of agreement for all raters for separate items was moderate to high, and weighted kappa was good to very good, ranging from 0.61 for change in speed to 0.95 for stair climbing. The inter-rater reliability was ICC (model 2,1)=0.73 (95% CI=0.49; 0.86). Agreement for two random raters was 45-92%. Weighted kappa was fair to very good, ranging from 0.27 for change in speed to 0.87 for walking with a narrow base of support. The internal consistency was high (α=0.88). Gait with eyes closed had the lowest item-to-total (0.33) and item-to-item (<0.30) correlation. Concurrent validity was moderate to strong (rho=0.50-0.76) with the highest correlation coefficient for the Timed Up and Go test. Conclusions: The FGA is a reliable and valid measure in people with peripheral vestibular disorder. For improving internal consistency, removal of the item 'gait with eyes closed' may be considered.