Item-Level Psychometrics for the Functional Gait Assessment in Persons With Stroke.

Area of Science:

• Rehabilitation Science
• Clinical Measurement
• Psychometrics

Background:

• Gait and balance impairments are significant challenges for individuals with chronic stroke.
• Accurate and reliable assessment tools are crucial for guiding rehabilitation interventions.
• The Functional Gait Assessment (FGA) is widely used, but its item-level psychometrics require detailed evaluation.

Purpose of the Study:

• To determine the item-level psychometric properties of the Functional Gait Assessment (FGA) in ambulatory individuals with chronic stroke.
• To develop a keyform (score sheet) for the FGA to aid clinical application.

Main Methods:

• Retrospective analysis of item-level data from 101 ambulatory individuals with chronic stroke.
• Rasch analysis, including principal component analysis of residuals, was employed to evaluate unidimensionality, item fit, and rating scale structure.
• The Rasch Rating Scale Model (RSM) was used to assess item difficulty hierarchy and person separation.

Main Results:

• Rasch analysis confirmed the FGA's unidimensionality and lack of item local dependence.
• The FGA demonstrated a monotonic category rating scale and an item difficulty hierarchy comparable to older adults.
• High person reliability (0.90) was observed, with no floor or ceiling effects, and the FGA effectively separated individuals into four strata.

Conclusions:

• Rasch analysis validates the FGA as a robust measure of walking balance ability in ambulatory chronic stroke survivors.
• The developed FGA keyform enables clinicians to obtain instantaneous interval measurements, enhancing clinical utility.