Purpose: The Berg Balance Scale (BBS) is one of the most commonly applied clinical measures of balance for individuals with stroke. This study used item analysis methods to investigate various psychometric aspects of the BBS in participants with subacute and chronic stroke.
Methods: Secondary analysis of data from 69 participants was included. The Rasch measurement model was used to generate item fit statistics for each item of the BBS and assess scale precision across the continuum of balance.
Results: Two of the 14 items of the BBS (standing on one foot and tandem standing) demonstrated misfit. The hierarchical scale demonstrated decreased precision at the higher end of the balance continuum.
Discussion: The original instructions for the BBS indicate that the individual tested can choose the limb utilized in the single-limb stance and tandem stance items. Administration of these test items in this way to individuals with stroke may contribute to item misfit. Standardization of administration of the BBS to routinely use the involved lower extremity may support higher item difficulty, contribute to improved scale precision, and decrease the ceiling effect of the scale previously reported.
Conclusion: The ability of participants to self-select the assessed lower extremity is most likely contributing to item misfit for 2 of the 14 BBS items. Standardized administration of these items may improve the ability of the BBS to more accurately measure initial balance impairment, as well as changes in balance due to spontaneous recovery and rehabilitation interventions.