Pilot studies of robotics suggest efficacy, but randomized clinical trials reveal little: Why?

Abstract

Robotic assistive devices to aid retraining an upper extremity or enable walking have led to gains in pilot studies of patients with spinal cord injury, as well as stroke, multiple sclerosis, and cerebral palsy. When tested scientifically in a randomized clinical trial with blinded assessments and an active control intervention, the results usually reveal equivalent efficacy at best. Confounders in pilot studies play a role in these disappointing results. The conceptual basis for robotic training-induced motor learning may also require examination, if the future of robotics is to produce better interventions for highly impaired patients. © 2011 Thomas Land Publishers, Inc.