Comparative Effectiveness Of Virtual Reality, Mirror Therapy, And Virtual Reality–Based Mirror Therapy For Upper-Extremity Motor Recovery After Stroke: A Systematic Review And Network Meta-Analysis

Abstract

Background. Upper-extremity motor impairment is a primary determinant of persistent disability following stroke. Conventional mirror therapy (MT), standalone virtual reality (VR), and digital or VR-based mirror therapy (VRMT/DMT) represent distinct visual and sensorimotor neurorehabilitation adjuncts. Their comparative effectiveness within a unified evidence framework remains uncertain.

Objective. To systematically evaluate and compare the effectiveness of MT, standalone VR, and VRMT/DMT against conventional rehabilitation (CT) and against one another for upper-extremity motor recovery, measured by the Fugl-Meyer Assessment–Upper Extremity (FMA-UE).

Methods. MEDLINE/PubMed, Scopus, and Web of Science were searched up to the last date supported by the review records (not independently re-verified from deposited database export files) without language restrictions. Randomized controlled trials (RCTs) evaluating adults with stroke who received MT, standalone VR, or VRMT/DMT against CT or active comparators were screened using a four-node framework developed for the present review. Pairwise mean differences (MDs) in total FMA-UE scores were pooled using Paule–Mandel random-effects models with Hartung–Knapp–Sidik–Jonkman (HKSJ) 95% confidence intervals. A frequentist random-effects network meta-analysis (NMA) evaluated indirect comparisons, loop inconsistency, and treatment rankings using P-scores. Methodological quality was appraised using Cochrane RoB 2, and certainty of evidence was graded using CINeMA.

Results. Sixteen unique RCTs contributing 20 direct pairwise contrasts were included. In pairwise analyses, standalone VR demonstrated a statistically significant improvement in total FMA-UE scores compared with CT (k = 8; MD = 4.17 points, 95% CI 2.08 to 6.26; I² = 0.0%; τ² = 0.00; p = 0.002). Conventional MT showed a positive but statistically non-significant difference versus CT (k = 6; MD = 5.17 points, 95% CI −0.64 to 10.98; I² = 35.9%; τ² = 10.85; p = 0.071). Direct evidence for VRMT/DMT versus CT (k = 3; MD = 3.08 points, 95% CI −14.12 to 20.27; I² = 39.6%; τ² = 17.21) and VRMT/DMT versus MT (k = 3; MD = 4.09 points, 95% CI −11.93 to 20.11; I² = 50.7%; τ² = 21.56) was imprecise. In the NMA, VR versus CT was the only active comparison with a 95% confidence interval excluding the null (MD = 4.17 points, 95% CI 2.08 to 6.26). The authors’ CINeMA assessment rated this comparison as high certainty; that rating was not independently re-executed from a deposited CINeMA project file. Treatment rankings yielded P-scores of 72.4% for VRMT/DMT, 66.1% for MT, 58.6% for VR, and 2.9% for CT.

Conclusions. Standalone VR demonstrated the most consistent evidence of benefit for post-stroke upper-extremity motor recovery in this network. Conventional MT showed promising directional effects that warrant continued evaluation. Digital or VR-based mirror therapy remains an emerging modality, and current evidence is insufficiently precise to confirm superiority over physical mirror therapy or conventional rehabilitation. Certainty judgements are those of the review authors using CINeMA and should be interpreted in light of the incomplete public deposition of analysis files.