Measurement Properties of the Functional Balance Ability Measure Across Subacute Neurologic Diagnoses

Abstract

Objective

To evaluate the measurement properties of the Functional Balance Ability Measure (FBAM) in people with subacute neurologic conditions using Rasch analysis.

Design

Retrospective, cross-sectional analysis.

Setting

Inpatient rehabilitation facility.

Participants

Data from persons (N=841, 57.5±18.6y, 39% women) with a neurologic diagnosis receiving inpatient rehabilitation who had Berg Balance Scale and Functional Gait Assessment item-level scores within three days of admission.

Interventions

Not applicable.

Main Outcome Measures

Principal component analysis of residuals evaluated measurement unidimensionality. Rasch analysis with the Andrich Rating Scale Model was used to quantify rating scale characteristics, item difficulty hierarchy, item and person fit, and person separation of the FBAM. Item difficulty hierarchy stability across 4 diagnostic groups (cerebral vascular accident, traumatic brain injury, spinal cord injury, and other nontraumatic neurologic disorders) was evaluated with differential item functioning (DIF).

Results

The principal component analysis of residuals suggested 2 dimensions, however, disattenuated correlations indicated dimensions were not measuring different constructs and therefore supported unidimensionality. The FBAM had an appropriate rating scale structure, item difficulty hierarchy, and high person reliability (0.91). Only 5.2% of persons misfit response patterns. Two items misfit but were retained because removal caused additional item misfit. The FBAM separated persons into 6 groups of balance ability. Marginal DIF was found for the items “sitting unsupported,” “gait with vertical head turns,” and ”gait with horizontal head turns” in the other nontraumatic neurologic disorders group, and “look behind” and “pick up an object” in the spinal cord injury group. Twenty-one percent of the sample received the lowest measure value, indicating a floor effect.

Conclusions

The FBAM is a unidimensional and reliable measure of balance for persons with subacute neurologic injuries. DIF analyses confirmed a stable hierarchy of balance ability across mobility levels supporting the FBAM’s construct validity. Future work should address the observed floor effect.