Reliability and validity of measures obtained from stroke patients using the balance master

Abstract

Objective: To determine the test-retest reliability and validity of data obtained using the Balance Master (BM), a computerized balance assessment and training tool. Design:Data were collected on three occasions, 1 week apart. Intraclass correlation coefficients (ICCs) were calculated if significant ( p < .05) between-subject variance was found using a univariate analysis of variance for repeated measures. Concurrent validity of the BM data was determined using the Berg Balance Scale and gait velocity as criterion standards. Participants: Twenty ambulatory hemiparetic subjects who had no history of lower extremity orthopedic problems, no neurological deficits apart from stroke, and had not trained using the BM. Main Outcome Measures: BM data relating to static and dynamic balance performance, gait velocity, and the total score from the Berg Balance Scale. Results: ICCs indicated that only the BM test requiring subjects to shift their center of gravity to randomly highlighted targets (positioned in a circle representing 75% of the individual's limits of stability) was reliable, both in terms of movement path (ICC = .84 and movement time (ICC = .88). Concurrent validity of the BM data was established for the dynamic measures of balance only, which correlated with both the Berg Balance Scale and gait velocity outcomes ( r ≥ .48, p < .05). Conclusions: These findings suggest that in stroke patients the test-retest reliability of data obtained using the BM is greatest for complex tests of balance and that dynamic rather than static balance measures are valid indicators of functional balance performance.