Comparative Effects of Neurostimulation- and Technology-Augmented Robot-Assisted Training on Upper-Limb Motor Recovery After Stroke: A Systematic Review and Network Meta-Analysis
In brief
Magnetic brain stimulation plus robot training scored 9.6 points above usual care
In a review of 103 trials, robot-assisted training combined with repetitive magnetic brain stimulation was associated with a 9.6-point higher arm-function score than usual care; low-frequency electrical stimulation plus training was associated with a 7.5-point gain. The magnetic-stimulation combination ranked highest, but varied results, limited direct comparisons and low-to-moderate evidence certainty leave the best approach unsettled.
- Journal
- Archives of physical medicine and rehabilitation (Q1)
- Published
- 3 November 2026
- Study design
- Systematic review / meta-analysis of RCTs
- Evidence level
- Level 1, High (CEBM 1a)
- Authors
- Lin Wang, Zhitao Zhu, Yue Liu, Junjie Qu, Ning Li, Haitao Liu, et al.
- PMID
- 42977063
- DOI
- 10.1016/j.apmr.2026.10.014
Why clinicians should know about it
- Picked for Rehabilitation (paper of the day, 4 November 2026): Stroke upper-limb recovery interventions; systematic review and network meta-analysis
- Picked for Neurology (clinical) (paper of the day, 4 November 2026): Systematic review compares rehabilitation interventions for stroke recovery
Abstract
OBJECTIVE: To compare robot-assisted training (RAT) alone with neurostimulation- or technology-augmented RAT for poststroke upper-limb recovery. DATA SOURCES: PubMed, Embase, Web of Science, Scopus, Cochrane Library, CNKI, and Wanfang were searched through August 1, 2026. STUDY SELECTION: Randomized trials evaluated RAT alone or RAT combined with virtual reality, a brain-computer interface, low-frequency electrical stimulation (LFES), transcranial direct current stimulation, or repetitive transcranial magnetic stimulation (rTMS). DATA EXTRACTION: Two reviewers independently selected studies and extracted data; risk of bias and certainty of evidence were assessed with Cochrane and GRADE methods. DATA SYNTHESIS: Of 103 trials (5,251 participants), 84 (4,277 participants) entered restricted Fugl-Meyer Assessment for Upper Extremity (FMA-UE) Model E. Compared with usual care (UC), mean differences (MDs) were 9.56 points (95% CI, 7.18-11.93) for rTMS plus RAT, 7.52 (95% CI, 4.48-10.57) for LFES plus RAT, and 3.83 (95% CI, 2.69-4.97) for RAT alone. Heterogeneity was substantial (τ=2.79); global inconsistency was not detected (χ²(9)=3.64; P=.934). rTMS plus RAT ranked first (SUCRA, 96.2%), but its probability of being best decreased from 77.8% to 56.0% after predictive uncertainty was incorporated. Signals for stroke stage (P=.030) and total duration (P=.040) did not meet the multiplicity-adjusted threshold (P<.01). CONCLUSIONS: rTMS plus RAT yielded the largest FMA-UE estimate, followed by LFES plus RAT. Sparse head-to-head evidence, heterogeneity, and very low-to-moderate certainty preclude a definitive hierarchy and warrant comparative trials.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.