Is the Berg Balance Scale an effective tool for the measurement of early postural control impairments in patients with Parkinson's disease? Evidence from Rasch analysis.

BACKGROUND: It is unclear whether the BBS is an effective tool for the measurement of early postural control impairments in patients with Parkinson's Disease (PD). AIM: To evaluate BBS' content validity, internal construct validity, reliability and targeting in patients with PD within the Rasch analysis framework. DESIGN: Observational, cross-sectional study. SETTING: Outpatient Rehabilitation Unit. POPULATION: A sample of 285 outpatients with PD. METHODS: The content validity of the BBS was assessed using standard linking techniques. The BBS was administered by trained physiotherapists. The data collected then underwent Rasch analysis. RESULTS: Content validity analysis showed a lack of items assessing postural responses to tripping and slips and stability during walking. On Rasch analysis, the BBS failed the requirements of monotonicity, local independence, unidimensionality and invariance. After rescoring 7 items, grouping of locally dependent items into testlets, and deletion of the static sitting balance item because mistargeted and underdiscriminating, the Rasch-modified BBS for PD (BBS-PD) showed adequate internal construct validity (χ2 24=39.693; p=0.023), including absence of differential item functioning (DIF) across gender and age, and was, as a whole, sufficiently precise for individual person measurement (PSI=0.894). However, the scale was not well targeted to the sample in view of the prevalence of higher scores. CONCLUSION: This study demonstrated the internal construct validity and reliability of the BBS-PD as a measurement tool for patients with PD within the Rasch analysis framework. However, the lack of items critical to the assessment of postural control impairments typical of PD, affected negatively the targeting, so that a significant percentage of patients was located in the higher ability range of the measurement continuum, where precision of measurement is reduced. CLINICAL REHABILITATION IMPACT: These findings suggest that the BBS, even if modified, may not be an effective tool for the measurement of early postural control in patients with PD. Language: en