A Valid and Reliable Virtual Reality Tool for Assessment of Arm Nonuse Post-Stroke

Introduction: Approximately 40-80% of individuals post-stroke fail to use their paretic arm despite sufficient capacity, a phenomenon known as arm nonuse. This behavior arises from an interaction between cognitive and motor factors, necessitating evaluation approaches that capture the impact of complex task demands on arm nonuse. However, existing measures require specialized equipment, extensive setup, or prolonged monitoring, limiting their clinical feasibility. To address this gap, this study evaluated the reliability, validity, and usability of a novel virtual reality-based arm nonuse (VR-NonUse) tool designed to measure arm nonuse during naturalistic reaching under varying cognitive and motor demands. Methods: Fifty-one individuals with stroke and 23 controls completed the VR-NonUse assessment across three conditions varying in cognitive and motor demand. Usability and tolerability were evaluated. Concurrent validity was examined by assessing associations between paretic arm nonuse on the VR-NonUse tool and established nonuse assessments. Discriminant validity was evaluated by comparing these associations with a motor impairment measure. Construct validity was examined by determining the effects of cognitive and motor demand on arm use. Results: The VR-NonUse tool was usable and well-tolerated. Paretic arm nonuse on the VR-NonUse tool was associated with established nonuse measures, and more strongly related to nonuse than to motor impairment, supporting concurrent and discriminant validity. Higher cognitive demand reduced paretic arm use, supporting construct validity. A brief cognitively demanding condition yielded a robust index of nonuse. Discussion: The VR-NonUse tool is a reliable, valid, and practical assessment of arm nonuse in chronic stroke.