Living Therapists Outclass Robots in Post-Stroke Rehab

By Todd Neale
CHICAGO, 09 may 2008-- A human touch appears to be better than robotic manipulation in helping stroke patients with hemiparesis improve their ability to walk, a small randomized study found.Those assisted by the warm hand of a therapist during locomotion training had greater improvements in walking speed and time spent standing on the impaired leg than those who were helped by an inanimate motorized apparatus, found T. George Hornby, Ph.D., of the University of Illinois at Chicago, and colleagues.Immediately following 12 half-hour training sessions on a treadmill, patients who were assisted by a therapist almost doubled the gains in normal walking speed made by those given robotic assistance (0.13 ± 0.11 versus 0.07 ± 0.07 m/s, P=0.04), they reported online in Stroke: Journal of the American Heart Association.
Therapist-assisted training was also associated with greater improvements in the amount of time spent standing on the impaired limb while walking at maximum speed (2.5 ± 3.7% versus 0.1 ± 0.6%, P<0.01) immediately following training.
The device consists of a harness that is attached at the patient's trunk and pelvis, with attachments down to the legs, and computer-controlled motors to move them in a way that approximates a normal human gait.
"Given the cost and continued development of rehabilitation devices, it is imperative to identify patients who may benefit from robotic-assisted training," the researchers said.
They said that stroke patients with chronic hemiparesis who can walk independently -- even slowly -- may be the right candidates for therapist-assisted training and that those who are nonambulatory may derive more benefit from robotic assistance.
Robotic devices were developed to aid locomotion training because it can be physically demanding on a therapist to provide assistance in walking, the researchers said, but the effectiveness of these devices has not been well established.
There are also some drawbacks to their use, including a possible decrease in aerobic stimulus and the reduction of the muscular activity needed to maintain a stable walking posture, they said.
To compare the benefits of each type of training, Dr. Hornby and colleagues recruited 48 stroke patients with hemiparesis lasting more than six months and with moderate to severe gait dysfunction for a randomized, controlled study.
Those who had an initial walking velocity of £0.5 m/s were considered to have severe impairments and those with a velocity of >0.5 and £0.8 m/s were classified as having moderate impairments.
All participants (mean age 57) were given 12 half-hour training sessions on a treadmill, in which they were partially supported by a harness. Half received assistance from a therapist as needed and half had constant guidance from a robotic orthosis called Lokomat.
Training speed was slightly faster in the therapist-assisted group (P=0.01) and training duration was slightly longer in the robotic-assisted group (P=0.01).
The greater improvements in normal and maximum walking speed and time spent standing on the impaired limb while walking at maximum speed in the therapist-assisted group persisted to six months after the training sessions, but the differences failed to reach statistical significance.
Overall, patients with moderate impairments at baseline had significantly greater gains in normal (P=0.03) and maximum (P=0.04) walking velocity than those with severe deficits.
Patients who had severe walking impairments and who were given therapist-assisted training were the only participants who had improvements in scores on an assessment of their quality of life relating to fewer physical limitations.
One of the possible reasons greater improvements were found in the therapist-assisted group, the researchers said, may be that the patients were expending more energy.
"Reduced metabolic activity during robotic-assisted [locomotor training] may limit adaptations which occur during treadmill exercise in individuals post-stroke, which are thought to be a primary factor contributing to improved ambulation," they said.
Also, they said, robotic-assisted training may reduce the volitional drive of patients and impair motor memory consolidation.
Finally, patients are free to make more errors in walking in therapy-assisted training than in robotic-assisted training, allowing them to learn from the mistakes.
The researchers acknowledged that the study was limited by the small size and the inability to blind the investigators or patients to the therapy received.
The study was funded by grants from the National Institute on Disability and Rehabilitation Research. The authors made no disclosures.
Primary source: Stroke: Journal of the American Heart AssociationSource reference:Hornby T, et al "Enhanced gait-related improvements after therapist- versus robotic-assisted locomotor training in subjects with chronic stroke: A randomized controlled study" Stroke 2008; DOI: 10.1161/STROKEAHA.107.504779.