Development of a Computerized Adaptive Testing System of the Functional Assessment of Stroke

Abstract

Objective: To develop a computerized adaptive testing system of the Functional Assessment of Stroke (CAT-FAS) to assess upper- and lower-extremity (UE/LE) motor function, postural control, and basic activities of daily living with optimal efficiency and without sacrificing psychometric properties in patients with stroke.

Design: Simulation study.

Setting: One rehabilitation unit in a medical center.

Participants: Patients with subacute stroke (N=301; mean age, 67.3±10.9; intracranial infarction, 74.5%).

Interventions: Not applicable.

Main outcome measures: The UE and LE subscales of the Fugl-Meyer Assessment, Postural Assessment Scale for Stroke Patients, and Barthel Index.

Results: The CAT-FAS adopting the optimal stopping rule (limited reliability increase of <.010) had good Rasch reliability across the 4 domains (.88-.93) and needed few items for the whole administration (8.5 items on average). The concurrent validity (CAT-FAS vs original tests, Pearson r=.91-.95) and responsiveness (standardized response mean, .65-.76) of the CAT-FAS were good in patients with stroke.

Conclusions: We developed the CAT-FAS, and our results support that the CAT-FAS has sufficient efficiency, reliability, concurrent validity, and responsiveness in patients with stroke. The CAT-FAS can be used to simultaneously assess patients' functions of UE, LE, postural control, and basic activities of daily living using, on average, no more than 10 items; this efficiency is useful in reducing the assessment burdens for both clinicians and patients.

Keywords: Psychometrics; Rehabilitation; Reproducibility of results; Stroke; Validation studies as topic.