Robot-Assisted Training as Self-Training for Upper-Limb Hemiplegia in Chronic Stroke: A Randomized Controlled Trial

Area of Science:

• Rehabilitation Medicine
• Neuroscience
• Robotics in Healthcare

Background:

• Stroke survivors often experience chronic upper-extremity hemiplegia, impacting daily function.
• Conventional self-training is a common rehabilitation approach, but its efficacy can be limited.
• Exploring novel therapeutic strategies like robotic self-training is crucial for improving patient outcomes.

Purpose of the Study:

• To compare the effectiveness of robotic self-training against conventional self-training for upper-extremity function in chronic stroke.
• To evaluate robotic self-training combined with conventional therapy and constraint-induced movement therapy.

Main Methods:

• Multi-center, prospective, randomized, parallel-group study involving 129 patients with chronic upper-limb hemiplegia.
• Three groups: conventional self-training (control), robotic self-training (RT) + conventional therapy, and RT + constraint-induced movement therapy.
• Interventions consisted of 1-hour sessions, 3 times per week for 10 weeks, with blinded outcome assessments.

Main Results:

• No significant difference in the primary outcome (Fugl-Meyer Assessment for upper-extremity) between robotic self-training and conventional self-training groups.
• The robotic self-training group showed a significant improvement in the shoulder/elbow/forearm subscore within the per-protocol set (P=0.037).
• One severe adverse event was reported, unrelated to the robotic device, indicating good safety.

Conclusions:

• Robotic self-training alone did not demonstrate superior upper-limb functional improvement compared to conventional self-training in this cohort.
• Robotic self-training may offer benefits when integrated with conventional therapy, particularly in patients adhering strictly to the protocol.
• Further research is warranted to identify specific patient populations or therapeutic combinations where robotic interventions yield optimal results.