On "a four-week, task-specific neuroprosthesis program ..." Dunning K, et al. Phys Ther. 2008;88:397-405.

We regard this report (1) as very valuable as it describes the use of a task-specific neuroprosthesis program in a patient characterized as severely impaired without any active residual finger and hand movements. Those patients without residual finger and hand movements usually are not included in studies dealing with rehabilitation after stroke because of their negative prognosis. Fortunately, Dunning et al present a therapeutic approach that was successful for a patient with such impairment.

There is one point we would like to raise to better understand the severity of the patient's impairment. Although it was stated that the patient was incapable of any active wrist or finger movement, before the intervention he was able to transfer 12 blocks in the Box and Block Test (2) from one side of a wall to the other side in a time frame of 60 seconds. On the Action Research Arm Test, he scored 5 points for grasping. We believe that a minimum of active finger movement is essential to grasp, transfer, and release. Therefore, the definition of "no residual movement" needs some clarification.

Preintervention testing was performed 2 weeks before intervention. But directly after baseline testing, the patient took the neuroprosthesis home for 1 week, probably starting the intervention. It, therefore, is unclear at which time after preintervention the task-specific training began.

The apparatus stimulated finger and wrist extension or finger and wrist flexion. The natural pattern during grasping is wrist extension with finger flexion and wrist flexion with finger extension. Is there any adaptive skillful movement available after this?

Please describe the physical therapy with more detail. If the patient was not able to move his fingers actively, how was feedback about the correct planning of a movement possible? Feedback and reward are main aspects of motor learning. (3-6) How did the therapist help the patient perform the complex tasks? The stimulator gave an interrupted-pulse with contraction and relaxation intervals at 7 seconds "on" and 7 seconds "off." How could the patient train the listed tasks in that rhythm? Did the apparatus have any feedback system, such as electromyography?

The clarification of the above questions could help other clinicians follow the intervention programs and eventually assist other patients with severe impairments after stroke.

D Broetz is Physical Therapist, Institute of Medical Psychology and Behavioral Neurobiology, MFG Center, University of Tuebingen, Germany.

This letter was posted as a Rapid Response on March 6, 2008, at www.ptjournal.org.

References

(1) Dunning K, Berberich A, Albers B, et al. A four-week, task-specific neuroprosthesis program for a person with no active wrist or finger movement because of chronic stroke. Phys Ther. 2008;88:397-405.

(2) Mathiowetz V, Volland G, Kashman N, Weber K. Adult norms for the Box and Block Test of manual dexterity. Am J Occup Ther. 1985;39:386-391.

(3) Cirstea CM, Ptito A, Levin MF. Feedback and cognition in arm motor skill reacquisition after stroke. Stroke. 2006;37:1237-1242.

(4) Cirstea MC, Levin ME Improvement of arm movement patterns and endpoint control depends on type of feedback during practice in stroke survivors. Neurorehabil Neural Repair. 2007;21:398-411.

(5) Bray S, Shimojo S, O'Doherty JP. Direct instrumental conditioning of neural activity using functional magnetic resonance imaging-derived reward feedback. J Neurosci. 2007;27:7498-7507.

(6) Marco-Pallares J, Muller SV, Munte TF. Learning by doing: an fMRI study of feedback-related brain activations. Neuroreport. 2007;18:1423-1426.

[DOI: 10.2522/ptj.2008.88.8.970.1]

Author Response

Thank you to Doris Broetz and colleagues for their thoughtful comments regarding this case report. As our ultimate goal is to improve function in persons with stroke, we welcome the opportunity to respond. It is our goal that, with this clarification, therapists will more clearly understand the type of patient who may benefit from this therapy and how to administer the therapy. We will provide our answers in the order of Broetz and colleagues' questions.

Broetz and colleagues are correct, in that having no active movement of the fingers would make it challenging to transfer blocks during the Box and Block Test or to grasp during the Action Research Arm Test (ARAT). Prior to starting the intervention, the patient demonstrated trace active movement in wrist extension and flexion. He demonstrated no active finger movement (flexion or extension). During pretesting, he scored zero in all grasp portions of the FuglMeyer Test. Upon reviewing the video portions of the ARAT, we observed that the patient started in forearm pronation, which resulted in a relaxed open hand. From this position, he was able to achieve a "grasp" by changing his shoulder and elbow position, resulting in wrist and finger flexion. Although we don't have video of the Box and Block testing, we assume that he used this same strategy for that test.

Two weeks before the task-specific training intervention started, the patient was pretested. One week before the task-specific training intervention started, the patient returned to the lab for orientation to the neuroprosthesis unit. He then took the unit home for 1 week to perform cyclic stimulation in order to ramp up his tolerance to the stimulation. During this week, he gradually increased the cyclic stimulation from 10 to 30 minutes daily; he did not perform task-specific activities during this week. The goal of this "ramp up" week was to acclimate him to the electrical stimulation in order avoid muscle fatigue during the intervention. This is common protocol for these neuroprosthesis units.

During a normal grasp, the wrist extensors provide stabilization for fingers to extend and flex. During the transport phases, the wrist can be in extension or flexion, depending on the task. During the release, the hand typically goes into a neutral position or even flexion; however, many tasks are functional with the wrist in extension. The neuroprosthesis helped the patient reach and grasp by stabilizing his wrist in a functional stationary extended position. The unit then stimulated his finger flexors and extensors to grasp and release. To facilitate normal motor learning, it would be ideal to not stimulate wrist flexion while finger flexors are being stimulated (as Broetz and colleagues are suggesting). Although the neuroprosthesis stimulates wrist flexion and extension during finger flexion and extension, respectively, this is minimized by electrode placement and stimulation intensity. Optimal contraction is achieved at the initial neuroprosthesis fitting session by selecting electrode placements that optimize finger contraction and minimize wrist contraction. The goal is to stimulate the primary finger muscles (eg, extensor digitorum and flexor digitorum superficialis muscles) and to minimize stimulation to the primary wrist muscles. Broetz and colleagues ask if there is "any adaptive skillful movement available after this." We believe that this refers to whether this type of activation facilitates motor learning. As we encourage patients to actively participate with the stimulation and do so in the context of feedback, motor learning can occur; however, activation may not resemble that of a "normal" hand due to the nervous system damage and subsequent control issues. This will depend on the patient. For example, someone with high wrist flexor spasticity may need much more extensor stimulation, making it difficult to isolate the finger extensors from the primary wrist extensors.

It is true that functional tasks are not performed in 7-second cycles as we used in this intervention program. These 7-second cycles, however, do allow for muscles to rest between contractions. Based on motor learning principles, we broke down the activity into components. First, the patient learned the stimulation timing (7 seconds "on" and 7 seconds "off"). Then he learned to do grasp and release during the stimulation timing. Education and feedback were provided during this learning period. The therapist observed whether the patient was doing the sequence correctly or fighting against the stimulation timing. The patient quickly learned to work with the stimulation timing. Once he learned to grasp and release simple objects using the stimulation timing, he progressed to components of functional activities. For example, with ironing, he initially worked on grasping the handle. Once that was achieved, he worked on moving the iron within the stimulation timing sequence. Once the sequencing of the subtask components were learned using the neuroprosthesis, he progressed to working on these subtask components without the neuroprosthesis. Finally, he worked on the complete task without stimulation (eg, the task of ironing). The therapist provided physical assistance as needed for safety. The therapist also provided verbal cues and encouragement but did not provide "hands-on" assistance to achieve the task (eg, hand-overhand physical assistance). If the patient was unable to perform the activity without the neuroprosthesis, we returned to practice with the device.

The apparatus does not have any electromyographic (EMG) feedback system, nor did we use any other EMG feedback during the therapy. The device does have a feedback light that turns on when the device is stimulating. Our patient didn't need this, however, because he had intact sensation so that he could feel the stimulation, and he quickly understood the sequence pattern. In our experience, this feedback light has helped patients who have limited sensation. (1)

K Dunning, PT, PhD, is Assistant Professor, Department of Rehabilitation Sciences, College of Allied Health Sciences, University of Cincinnati Academic Medical Center, Ohio.

This letter was posted as a Rapid Response on June 24, 2008, at www.ptjournal.org.

Reference

(1) Hill Hermann V, Strasser A, Albers B, et al. Task specific, patient-driven neuroprosthesis training in chronic stroke: results of a three-week clinical study. Am J Occup Ther. In press.

[DOI: 10.2522/ptj.2008.88.8.970.2]