Background: Cerebral palsy has pathology in immature brain problem; ischemic brain, hypoxic brain. The cause of pathology can be prenatal, perinatal and postnatal.  lectromyographic (EMG) biofeedback is muscles training by using electrical stimulation modality to train specific weakness muscles or pathologic side. Feedback response to the patients by visual or evidence supporting sound can make the patients to train themselves specifically. However, there is only few evidences supporting efficacy of EMG biofeedback to train muscles in cerebral palsy.

Objective: To study the effect of EMG biofeedback on upper extremity to improve hand function in children with cerebral palsy.

Materials and Methods: Forty children with cerebral palsy who had impairment of upper extremity and hand function. They were randomly assigned into two groups. The biofeedback group consisted of 20 patients, each received EMG biofeedback training of 3 muscles for 30 minutes plus three-task training for 30 minutes. The conventional group consisted of 20 patients, each received three-task training for 60 minutes. Upper extremity and hand function were evaluated before starting training, at 4 weeks and 8 weeks.

Results: The biofeedback group displayed statistically significant improvement regarding 3 subtest of Jebsen hand function test (JHFT) (p=0.004, 0.017, 0.004), respectively. Comparing with before starting training, mean decreasing of time spending at 4^th week were 15.03±4.01, 232.42±74.52 and 14.24±3.80, and at 8th week were 13.32±2.70, 251.85±80.25 and 10.34±3.28. There were significant improvement in almost all aspect of range of motion (ROM) of elbow and wrist joints (p<0.05) and modified tardieuscale (MTS) of elbow flexors and wrist flexors (p<0.005). Conventional group displayed statistically significant improvement regarding 1 subtest of JHFT (p=0.006). Comparing with before starting training, mean decreasing of time spending at 4th week were 174.90±49.20. Biofeedback group showed statistically significant progress over conventional group in 1 subtest of JHFT (p=0.002, 0.005), MTS of elbow flexors (p<0.001, 0.007) and ROM of elbow extension (p=0.018).

Conclusion: The effect of EMG biofeedback on upper extremity and hand function in children with cerebral palsy especially in large muscle trained by EMG biofeedback is superior to conventional therapy.