Robot-assisted and conventional therapies produce distinct rehabilitative trends in stroke survivors
Valero-Cuevas, Francisco J.
Winstein, Carolee J.
- Journal Article
Rights / licenseCreative Commons Attribution 4.0 International
Background Comparing the efficacy of alternative therapeutic strategies for the rehabilitation of motor function in chronically impaired individuals is often inconclusive. For example, a recent randomized clinical trial (RCT) compared robot-assisted vs. conventional therapy in 77 patients who had had chronic motor impairment after a cerebrovascular accident. While patients assigned to robotic therapy had greater improvements in the primary outcome measure (change in score on the upper extremity section of the Fugl-Meyer assessment), the absolute difference between therapies was small, which left the clinical relevance in question. Methods Here we revisit that study to test whether the multidimensional rehabilitative response of these patients can better distinguish between treatment outcomes. We used principal components analysis to find the correlation of changes across seven outcome measures between the start and end of 8 weeks of therapy. Permutation tests verified the robustness of the principal components found. Results Each therapy in fact produces different rehabilitative trends of recovery across the clinical, functional, and quality of life domains. A rehabilitative trend is a principal component that quantifies the correlations among changes in outcomes with each therapy. Conclusions These findings challenge the traditional emphasis of RCTs on using a single primary outcome measure to compare rehabilitative responses that are naturally multidimensional. This alternative approach to, and interpretation of, the results of RCTs may will lead to more effective therapies targeted for the multidimensional mechanisms of recovery Show more
Journal / seriesJournal of NeuroEngineering and Rehabilitation
Pages / Article No.
SubjectPrincipal Component Analysis; Conventional Therapy; Stroke Survivor; Nonnegative Matrix Factorization; Dominant Trend
Organisational unit03654 - Riener, Robert
MoreShow all metadata