Abstract
Several studies have reported the effects of robotic arms on improving upper limb function in patients with stroke. However, previous studies have reported inconsistent findings that may lead to incorrect applications of robotic arm use. Six databases were searched for relevant randomized controlled trials. Meta-analyses were performed for upper limb performance measures, including subgroup analysis of pooled upper limb rehabilitation data such as stroke stage and intervention delivery dose. Furthermore, the Cochrane risk-of-bias tool for randomized trials version 2 (RoB 2) and sensitivity analysis were used to assess methodology and determine publication bias. The final analysis included 18 studies. Robotic arms improved upper limb and hand function in patients with stroke. Subgroup analysis revealed that robotic arm interventions lasting 30-60 min per session significantly improved upper limb function. However, no significant improvement was observed in shoulder and elbow or wrist and hand movements. This review may help develop applicable rehabilitation robots and collaboration between clinicians.