The efficacy of robot-assisted therapy on upper limb motor function in children with cerebral palsy: a systematic review and meta-analysis

Cerebral palsy (CP) is the leading cause of neurodevelopmental motor disability in children, often resulting in impaired upper limb function. Although traditional interventions offer benefits, they are limited by therapist dependence and poor adherence. Robot-assisted therapy (RAT) has emerged as a promising alternative, providing repetitive and high-intensity training. However, its clinical efficacy, particularly for upper limb rehabilitation in CP, remains inconclusive. A systematic search was performed in eight electronic databases up to June 2025. Seven randomized controlled trials (RCTs) involving 244 children with CP met the inclusion criteria. The risk of bias was assessed using the Cochrane Risk of Bias 2.0 tool. Data were synthesized using both fixed-effects and random-effects models, with effect sizes expressed as standardized mean differences (SMD) and 95% confidence intervals (CI). RAT significantly improved fine motor function (SMD = 1.02, 95% CI: 0.56–1.48, P < 0.0001), consistent with a clinically meaningful effect. A positive trend was observed for ADL capacity (SMD = 0.65, 95% CI: 0.07–1.22, P = 0.03), although this finding was sensitive to the removal of individual studies. Due to extreme heterogeneity (I² = 96%), gross motor function results were not pooled across device types. Subgroup analyses suggested that exoskeleton-based RAT may offer benefits for proximal motor recovery; however, this finding is based on a single trial and should be considered hypothesis-generating. No significant improvements were found in muscle strength (SMD = 0.34, P = 0.10) or real-world participation (SMD = 0.22, P = 0.56). RAT produces a clinically meaningful improvement in fine motor function, with preliminary evidence of benefit for ADL capacity. Its efficacy on gross motor function appears device-dependent but requires confirmatory evidence, and effects on muscle strength and participation remain limited. This study was prospectively registered in the PROSPERO database (Registration No. CRD42025645232).