Reliability, validity, and responsiveness of the Balance Evaluation Systems Test (BESTest) in individuals with multiple sclerosis.

BACKGROUND: The Balance Evaluation Systems Test (BESTest) measures various aspects of postural control, but little data exist in persons with multiple sclerosis (MS). The purpose of this study was to determine the psychometrics of the BESTest in MS.

DESIGN: Observational study.

METHODS: 21 ambulatory subjects with MS participated. In the first session, demographic data was collected; each subject completed a questionnaire of self-perceived disability level and the BESTest. The BESTest was re-administered 1 week later.

RESULTS: Test-retest reliability (ICC 3,1) for the total BESTest was 0.94, ranging 0.66 to 0.93 for the subsections. Internal consistency (Chronbach's alpha) for the total BESTest was 0.97; subsections scores ranged 0.79 to 0.96. Minimal detectable change (MDC) scores ranged from 2.25 to 4.58 for subsections with 9.47 points for total BESTest. Weak to moderate correlations were found between individual subsection scores (0.12 to 0.78), and BESTest total and subsection scores to fall (-0.08 to -0.62) frequency and self-perceived disability level (-0.24 to -0.64). Strongest correlations were found between BESTest total and individual subsection scores. No floor effects were found; five BESTest subsections had ceiling effects.

CONCLUSIONS: The BESTest is reliable and valid in individuals with MS. Total BESTest scores demonstrated higher reliability and a lack of a ceiling effect as compared to subsection scores, suggesting that clinicians use the BESTest in its entirety. The correlations among subsection scores indicate that each assesses a unique aspect of balance, supporting its construct validity. The MDC scores will assist clinicians in assessing patient change.