Development of a Short Form Assessment Combining the Fugl-Meyer Assessment–Upper Extremity and the Wolf Motor Function Test for Evaluating Stroke Recovery

Objective: To use Rasch methodologies to combine the items of the Fugl-Meyer Assessment–Upper Extremity (FMA-UE, motor skill) and the Wolf Motor Function Test (WMFT, motor function) onto a single measurement metric and create an FMA-UE+WMFT short form. Design: Secondary analysis of preintervention data from 2 upper extremity stroke rehabilitation trials. Confirmatory factor analysis and Rasch rating scale analysis were first applied to examine the properties of the pooled item bank and then item response theory methodologies were used to develop the short form. Confirmatory factor analysis and Rasch analysis were then applied to the short form to examine the dimensionality and measurement properties. Setting: Outpatient academic medical research center. Participants: Data from 167 participants who completed the FMA-UE and WMFT (rating scale score) were pooled (N=167). Participants were eligible if they had a stroke ≥3 months prior and had upper extremity (UE) hemiparesis and excluded if they had severe UE hemiparesis, severe UE spasticity, or UE pain. Intervention: Not applicable. Main Outcome Measures: The dimensionality and measurement properties of the pooled 30-item FMA-UE and the 15-item WMFT and short form were examined. Results: Five items from the pool of 45 items were misfit and were removed. The 40-item pool demonstrated adequate measurement properties. A 15-item short form was then developed and met rating diagnostic scale criteria. All items on the 15-item short form met the Rasch fit criteria, and the assessment met criteria for reliability (Cronbach alpha=.94), separation (person separation = 3.7), and strata (number of strata = 5). Conclusions: Items from the FMA-UE and WMFT can be pooled to create a psychometrically sound 15-item short form.