Telerehabilitation for Motor Recovery in Adults with Neurological Disorders: A Systematic Review and Meta-Analysis Across Motor Domains, Delivery Models, and Follow-Up Effects

Abstract

Background/Objectives: Telerehabilitation may extend neurorehabilitation beyond conventional settings, but comparative effects across neurological diagnoses and motor domains remain uncertain. This systematic review and meta-analysis synthesized motor outcomes while distinguishing International Classification of Functioning, Disability and Health activity domains, delivery models, and comparator questions. Methods: Five databases were searched from inception to 8 June 2026 using two database-specific Boolean search strategies. Comparative studies of adults with neurological disorders receiving active remote, digital, virtual, wearable, or hybrid rehabilitation were eligible. Random-effects models used restricted maximum likelihood estimation; Hartung–Knapp confidence intervals were primary when at least three studies were available, and DerSimonian–Laird models were sensitivity analyses. Results: Eighty-seven studies included stroke, Parkinson’s disease, multiple sclerosis, spinal cord injury, acquired brain injury, or ataxia. For walking and mobility activities versus usual care, no additional therapy, or waitlist, the point estimate favoured telerehabilitation but was imprecise and compatible with no difference (standardized mean difference 1.06, 95% confidence interval −0.23 to 2.35; three studies; 121 participants; I2 = 57.6%). Global motor-functional and activities-of-daily-living outcomes showed similar uncertainty (0.83, −0.35 to 2.00; three studies; 304 participants; I2 = 73.6%). Both findings had very-low-certainty evidence. No pooled comparison included enough studies to assess publication bias. Two-study datasets for balance-related activity/postural control and upper-limb impairment were highly heterogeneous and interpreted descriptively. Conclusions: Comparative evidence for telerehabilitation across neurological diagnoses and motor domains remains uncertain. Programme characteristics are plausible implementation considerations, not established efficacy modifiers. Small evidence sets, wide intervals, clinical heterogeneity, risk of bias, and inconsistent adverse-event ascertainment preclude conclusions about clinical importance or established safety.

The authors' abstract, as published at the source. Brain Sciences, 2026 · DOI ↗