Archives of Physical Medicine and Rehabilitation journal homepage: www.archives-pmr.org Archives of Physical Medicine and Rehabilitation 2015;96:2243-8
BRIEF REPORT
Combined Clinic-Home Approach for Upper Limb Robotic Therapy After Stroke: A Pilot Study Grace J. Kim, MS, OTR/L,a Lisa Rivera, MS, OTR/L,a Joel Stein, MDa,b,c From the aDepartment of Rehabilitation Medicine, NewYork-Presbyterian/Weill Cornell Medical Center, New York, NY; bDivision of Rehabilitation Medicine, Weill Cornell Medical College, New York, NY; and cDepartment of Rehabilitation and Regenerative Medicine, Columbia University College of Physicians and Surgeons, New York, NY.
Abstract Objective: To investigate the feasibility of a combined clinic-home intervention using a robotic elbow brace and, secondarily, to collect preliminary data on the efficacy of this clinic-home intervention. Design: Nonrandomized pre-/postinterventional study. Setting: Outpatient clinic and participants’ homes. Participants: Individuals at least 6 months after stroke (NZ11; 5 women and 6 men; mean age, 51.7y; mean time since stroke, 7.6y; mean FuglMeyer Assessment of the Upper Extremity [FMA-UE] score, 22 of 66) were enrolled from the community. Interventions: Participants received training in an outpatient clinic from an experienced occupational therapist to gain independence with use of the device (3e9 sessions) followed by a 6-week home program using the device at home. Main Outcome Measures: Five instruments were administered before and after the study intervention: Modified Ashworth Scale, Box and Blocks test, FMA-UE, Arm Motor Ability Test, and Motor Activity Log-Amount of Use and Motor Activity Log-How Well subscales (MAL-AOU, MAL-HW). Results: Nine participants completed the study. Participants used the device on average 42.9min/d, 5.3d/wk. The FMA-UE (tZ3.32; PZ.01), MAL-AOU (tZ4.40; PZ.002), and MAL-HW (tZ4.02; PZ.004) scores showed statistically significant improvement from baseline to discharge; the MAL-AOU (tZ2.61; PZ.035) and MAL-HW (tZ2.47; PZ.043) scores were also significantly improved from baseline to 3-month follow-up. Conclusions: This combined clinic-home intervention was feasible and effective. Participants demonstrated improvements in arm impairment and self-reported use of the arm from baseline to discharge; they continued to report significant improvement in actual use of the arm at 3-month follow-up. Archives of Physical Medicine and Rehabilitation 2015;96:2243-8 ª 2015 by the American Congress of Rehabilitation Medicine
The use of upper limb robotic devices has proven to be safe and comparable to dose-matched conventional therapy in improving motor impairment, activity, and participation.1-4 Most previous studies of upper limb robotic therapy have been conducted in research laboratories or clinical settings by using large endeffector or exoskeletal robots such as the MIT-Manus, ARMin,
The devices used during the study were provided by Myomo, Inc. No other funding was provided, and Myomo did not have any involvement in the design, conduct, or reporting of this study. Disclosures: J. Stein has served on the scientific advisory board of Myomo, Inc; received nonfinancial support from Tyromotion, Inc; and grant funding from Nexstim, Inc. The other authors have nothing to disclose.
MIME, and T-WREX5 that are not mobile or practical to use in a patient’s home. Moreover, these devices are complex to operate and are not designed to be used by a patient without the supervision of a trained operator. As health care delivery shifts away from traditional hospital settings and increasingly into the community,6 there is a growing need and interest in the feasibility and efficacy of home-based interventions. In our study, we examined the home use of the Myomo mPower 1000,a an electromyography-controlled wearable robotic elbow brace that provides assistance with elbow flexion and/or extension. Individuals with decreased motor control, coordination, and/or strength after neurological injury use the biofeedback provided by the device to practice activation and inhibition of the biceps and
0003-9993/15/$36 - see front matter ª 2015 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2015.06.019
2244
G.J. Kim et al
triceps muscles. Users must initiate muscle contraction or inhibition to trigger the device and modulate the degree of force assistance provided by the device. Previous studies7,8 in the clinic using the Myomo have demonstrated upper extremity improvements at impairment and participation levels. Because the device is wearable and fairly lightweight (1kg), it can potentially be worn while engaging in various tasks in a functional context at home. It has been cleared by the Food and Drug Administration for home use without a treating therapist’s direct supervision and for use by people after stroke to target muscle reeducation and to maintain or increase active range of motion. The aim of this pilot study was to test the feasibility of a combined clinic-home intervention consisting of device training of a robotic elbow brace followed by a 6-week home exercise program and, secondarily, to collect preliminary data on the efficacy of this clinic-home intervention measured by activity and impairment level clinical outcomes of the upper extremity.
The Actigraph GT3Xþ accelerometerb was attached to the device to collect objective motion data to substantiate participant reports of use.
Outcome measures Five outcome measures were used to assess the motor status of the affected arm: Modified Ashworth Scale (MAS),9 Box and Blocks test,10 FMA-UE,11 Arm Motor Ability Test,12 and Motor Activity Log-Amount of Use subscale (MAL-AOU) and Motor Activity Log-How Well subscale (MAL-HW).13 Assessments were performed at 4 time points: before training (baseline 1), after training in the use of the device, but before initiating the home program (baseline 2), at the conclusion of the home-based exercise program (discharge), and at 12-week follow-up after the completion of home training.
Results
Methods
Device use at home Participants Participation of human subjects was approved by the local institutional review board before the start of the study. A total of 44 potential subjects were screened, and 11 qualified for the study. Nine subjects completed the entire protocol, whereas 2 subjects withdrew early, citing technical difficulties (management of wires and straps with one hand, accurate placement of the brace on the arm), lack of sufficient time, and insufficient caregiver assistance available to comply with daily use of the device. Six men and 5 women participated in the study, with a mean age of 51.78.5 years and a mean time since stroke of 7.69.7 years. The mean baseline Fugl-Meyer Assessment of the Upper Extremity (FMA-UE) score for the group was 2215.5 of 66, indicating a severely impaired group (see appendix 1 for inclusion and exclusion criteria).
Procedures Individuals who met all inclusion and exclusion criteria were included in the study and consented to participate in the institutional review boardeapproved protocol. Participants were trained by an experienced occupational therapist to don/doff the device, practice electromyography electrode placement, and practice biofeedback of the elbow flexors and extensors in the various modes of the device. The training sessions were of 30 to 45 minutes long, typically 2 to 3 times a week. Participants were deemed independent once they or a caregiver were able to correctly demonstrate all the technical items listed on the user competency checklist (appendix 2). Subsequently, participants started a 6-week program practicing upper limb activities at home (appendix 3). Participants had weekly reassessments in the clinic to troubleshoot any technical issues, monitor adverse events, upgrade their home program, collect weekly logs, and download activity monitor data. Participants recorded time of daily use of the device on a weekly calendar.
List of abbreviations: FMA-UE MAL-AOU MAL-HW MAS
Fugl-Meyer Assessment of the Upper Extremity Motor Activity Log-Amount of Use subscale Motor Activity Log-How Well subscale Modified Ashworth Scale
Participants recorded time of daily use of the device in a calendar log. There was variability in time spent using the device at home, ranging from 33 to 89 minutes, with a group average of 42.9min/d. Accelerometer data captured by the Actigraph GT3Xþ accelerometer confirmed participant reports. ActiLife v.6.11.2 software,b which accompanies the accelerometer, was used to calculate daily wear time validation for each participant during the 6-week home program.
Data analysis Descriptive statistics were used to calculate demographic characteristics and average wear time for the group (see table 1 for participant summary). Paired t tests were calculated to determine mean change scores pre- and posttreatment from baseline 1 to discharge, from baseline 1 to follow-up, from baseline 2 to discharge, and from baseline 2 to follow-up. Baseline 1 and baseline 2 were also compared because a therapeutic effect was possible for the device training period. For baseline 1 to discharge, improvement in the FMA-UE (tZ3.32; PZ.01), MAL-AOU; (tZ4.40; PZ.002), and MAL-HW (tZ4.02; PZ.004) scores were statistically significant; mean change scores for the FMA-UE was 3.2 points; for MAS-AOU, .30 points; and for MAL-HW, .27 points (table 2). For baseline 1 to follow-up, improvement in the MAL-AOU (tZ2.61; PZ.035) and MAL-HW (tZ2.47; PZ.043) scores were statistically significant; mean change scores for the MAS-AOU was .14 points; and for MAL-HW, .18 points (table 3). Only the MAL-HW score (tZ2.47; PZ.033) was statistically significant from baseline 2 to discharge, with a mean change score of .19 points. There were no significant findings for baseline 1 to baseline 2 and for baseline 2 to follow-up.
Discussion The study results support our hypothesis that a combined clinichome robotic program integrating the affected arm into functional activities is feasible and a potentially effective therapeutic approach. On the basis of accelerometer wear time data and participant report, participants used the device and practiced arm www.archives-pmr.org
Clinic-home robotics study Table 1
2245
Participant summary
Participant No.
Age (y)
1 2
47 68
3 4 5 6 7 8 9 10 11 Group mean
61 57 48 47 60 49 43 40 49 51.73
Chronicity (y) 1.43 3.49 3.04 3.64 3.04 3.98 9.16 8.59 35.43 7.47 3.44 7.6
FMA-UE Score at Baseline
No. of Training Sessions
14 9
6 9
8 12 26 16 21 30 57 41 9 22.09
9 9 6 9 9 4 3 7 5 6.9
Adverse Events* Technical difficultiesy Skin breakdown; technical difficulties Skin redness; technical difficulties Technical difficulties None Technical difficulties Technical difficulties None Skin irritation Skin irritation Technical difficulties 4 of 11 had skin irritation or breakdown from electrodes; 7 of 11 had technical difficulties
Average Time of Daily Use of the Device (min)
Completed Study?
89.04 48.86
Yes Yes
34.4 43.21 41.9 NA 32.14 53.14 52.14 33.26 NA 42.9
Yes Yes Yes No Yes Yes Yes Yes No 9 of 11
* Adverse events included skin redness and irritation as well as minor breakdown. All cases were related to the adhesives from the electromyography electrodes. We used a combination of strategies to successfully address the adverse events: break from device use (1d to 1wk), reeducation of electromyography electrode placement, and skin protectants. No participants withdrew because of skin issues, and symptoms were resolved. y Technical difficulties included problems with battery charging; needing additional device training during home exercise program; difficulty with electrode placement, wires, straps with one hand; and accurate placement of the brace on the affected arm.
training on their own, averaging 42.9min/d. However, this included time required for setup, troubleshooting, and donning/doffing the device, which varied widely across participants. Therefore, the true dosage of daily training was most likely less than that indicated by our data. Participants demonstrated statistically significant improvements in both arm impairment and self-reported use of the arm from baseline to discharge; they continued to report significant improvement in actual use of the arm at 3-month follow-up. Interestingly, we did not find significant therapeutic effects of the device training period or the 6-week home program alone; however, an analysis of a combined intervention yielded significant improvement in arm impairment and function. These findings may simply represent a dose-response curve that was achieved only with the entire training and therapeutic program, or they might indicate that combining a comprehensive training period with a home program may have a synergistic effect. Most participants completed the study, with 2 early withdrawals.
Table 2 Outcome Measure FMA-UE AMAT MAL-AOU MAL-HW MAS-B MAS-T BB
Study limitations There were several limitations of this study. Because of the small sample size and the absence of a control group, the results should be interpreted with caution. Our results were statistically significant, but did not achieve minimally clinical important difference levels for the FMA-UE or MAL. For subsequent follow-up studies, we suggest a larger sample size with sufficient power to yield definitive statistical results. Because of the general nature of the data captured by commercially available accelerometers, a true dosage of arm training at home was difficult to determine. Considerable commitment and effort is required from both the stroke survivor and the clinician to make this type of program successful, limiting this approach to highly motivated patienttherapist teams (see appendix 4 for recommendations for future studies). The device used in this study was limited to the elbow joint, and no direct training of the distal wrist or hand was addressed.
Paired t test results for baseline 1 to discharge Mean Change Score (points) 3.22 0.04 0.30 0.27 0.06 0.33 0.22
SD 2.91 0.21 0.20 0.20 0.68 0.75 2.28
SEM .97 .07 .07 .07 .23 .25 .76
t 3.33 0.57 4.41 4.02 0.58 1.34 0.29
Significance (P
BRIEF REPORT
Combined Clinic-Home Approach for Upper Limb Robotic Therapy After Stroke: A Pilot Study Grace J. Kim, MS, OTR/L,a Lisa Rivera, MS, OTR/L,a Joel Stein, MDa,b,c From the aDepartment of Rehabilitation Medicine, NewYork-Presbyterian/Weill Cornell Medical Center, New York, NY; bDivision of Rehabilitation Medicine, Weill Cornell Medical College, New York, NY; and cDepartment of Rehabilitation and Regenerative Medicine, Columbia University College of Physicians and Surgeons, New York, NY.
Abstract Objective: To investigate the feasibility of a combined clinic-home intervention using a robotic elbow brace and, secondarily, to collect preliminary data on the efficacy of this clinic-home intervention. Design: Nonrandomized pre-/postinterventional study. Setting: Outpatient clinic and participants’ homes. Participants: Individuals at least 6 months after stroke (NZ11; 5 women and 6 men; mean age, 51.7y; mean time since stroke, 7.6y; mean FuglMeyer Assessment of the Upper Extremity [FMA-UE] score, 22 of 66) were enrolled from the community. Interventions: Participants received training in an outpatient clinic from an experienced occupational therapist to gain independence with use of the device (3e9 sessions) followed by a 6-week home program using the device at home. Main Outcome Measures: Five instruments were administered before and after the study intervention: Modified Ashworth Scale, Box and Blocks test, FMA-UE, Arm Motor Ability Test, and Motor Activity Log-Amount of Use and Motor Activity Log-How Well subscales (MAL-AOU, MAL-HW). Results: Nine participants completed the study. Participants used the device on average 42.9min/d, 5.3d/wk. The FMA-UE (tZ3.32; PZ.01), MAL-AOU (tZ4.40; PZ.002), and MAL-HW (tZ4.02; PZ.004) scores showed statistically significant improvement from baseline to discharge; the MAL-AOU (tZ2.61; PZ.035) and MAL-HW (tZ2.47; PZ.043) scores were also significantly improved from baseline to 3-month follow-up. Conclusions: This combined clinic-home intervention was feasible and effective. Participants demonstrated improvements in arm impairment and self-reported use of the arm from baseline to discharge; they continued to report significant improvement in actual use of the arm at 3-month follow-up. Archives of Physical Medicine and Rehabilitation 2015;96:2243-8 ª 2015 by the American Congress of Rehabilitation Medicine
The use of upper limb robotic devices has proven to be safe and comparable to dose-matched conventional therapy in improving motor impairment, activity, and participation.1-4 Most previous studies of upper limb robotic therapy have been conducted in research laboratories or clinical settings by using large endeffector or exoskeletal robots such as the MIT-Manus, ARMin,
The devices used during the study were provided by Myomo, Inc. No other funding was provided, and Myomo did not have any involvement in the design, conduct, or reporting of this study. Disclosures: J. Stein has served on the scientific advisory board of Myomo, Inc; received nonfinancial support from Tyromotion, Inc; and grant funding from Nexstim, Inc. The other authors have nothing to disclose.
MIME, and T-WREX5 that are not mobile or practical to use in a patient’s home. Moreover, these devices are complex to operate and are not designed to be used by a patient without the supervision of a trained operator. As health care delivery shifts away from traditional hospital settings and increasingly into the community,6 there is a growing need and interest in the feasibility and efficacy of home-based interventions. In our study, we examined the home use of the Myomo mPower 1000,a an electromyography-controlled wearable robotic elbow brace that provides assistance with elbow flexion and/or extension. Individuals with decreased motor control, coordination, and/or strength after neurological injury use the biofeedback provided by the device to practice activation and inhibition of the biceps and
0003-9993/15/$36 - see front matter ª 2015 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2015.06.019
2244
G.J. Kim et al
triceps muscles. Users must initiate muscle contraction or inhibition to trigger the device and modulate the degree of force assistance provided by the device. Previous studies7,8 in the clinic using the Myomo have demonstrated upper extremity improvements at impairment and participation levels. Because the device is wearable and fairly lightweight (1kg), it can potentially be worn while engaging in various tasks in a functional context at home. It has been cleared by the Food and Drug Administration for home use without a treating therapist’s direct supervision and for use by people after stroke to target muscle reeducation and to maintain or increase active range of motion. The aim of this pilot study was to test the feasibility of a combined clinic-home intervention consisting of device training of a robotic elbow brace followed by a 6-week home exercise program and, secondarily, to collect preliminary data on the efficacy of this clinic-home intervention measured by activity and impairment level clinical outcomes of the upper extremity.
The Actigraph GT3Xþ accelerometerb was attached to the device to collect objective motion data to substantiate participant reports of use.
Outcome measures Five outcome measures were used to assess the motor status of the affected arm: Modified Ashworth Scale (MAS),9 Box and Blocks test,10 FMA-UE,11 Arm Motor Ability Test,12 and Motor Activity Log-Amount of Use subscale (MAL-AOU) and Motor Activity Log-How Well subscale (MAL-HW).13 Assessments were performed at 4 time points: before training (baseline 1), after training in the use of the device, but before initiating the home program (baseline 2), at the conclusion of the home-based exercise program (discharge), and at 12-week follow-up after the completion of home training.
Results
Methods
Device use at home Participants Participation of human subjects was approved by the local institutional review board before the start of the study. A total of 44 potential subjects were screened, and 11 qualified for the study. Nine subjects completed the entire protocol, whereas 2 subjects withdrew early, citing technical difficulties (management of wires and straps with one hand, accurate placement of the brace on the arm), lack of sufficient time, and insufficient caregiver assistance available to comply with daily use of the device. Six men and 5 women participated in the study, with a mean age of 51.78.5 years and a mean time since stroke of 7.69.7 years. The mean baseline Fugl-Meyer Assessment of the Upper Extremity (FMA-UE) score for the group was 2215.5 of 66, indicating a severely impaired group (see appendix 1 for inclusion and exclusion criteria).
Procedures Individuals who met all inclusion and exclusion criteria were included in the study and consented to participate in the institutional review boardeapproved protocol. Participants were trained by an experienced occupational therapist to don/doff the device, practice electromyography electrode placement, and practice biofeedback of the elbow flexors and extensors in the various modes of the device. The training sessions were of 30 to 45 minutes long, typically 2 to 3 times a week. Participants were deemed independent once they or a caregiver were able to correctly demonstrate all the technical items listed on the user competency checklist (appendix 2). Subsequently, participants started a 6-week program practicing upper limb activities at home (appendix 3). Participants had weekly reassessments in the clinic to troubleshoot any technical issues, monitor adverse events, upgrade their home program, collect weekly logs, and download activity monitor data. Participants recorded time of daily use of the device on a weekly calendar.
List of abbreviations: FMA-UE MAL-AOU MAL-HW MAS
Fugl-Meyer Assessment of the Upper Extremity Motor Activity Log-Amount of Use subscale Motor Activity Log-How Well subscale Modified Ashworth Scale
Participants recorded time of daily use of the device in a calendar log. There was variability in time spent using the device at home, ranging from 33 to 89 minutes, with a group average of 42.9min/d. Accelerometer data captured by the Actigraph GT3Xþ accelerometer confirmed participant reports. ActiLife v.6.11.2 software,b which accompanies the accelerometer, was used to calculate daily wear time validation for each participant during the 6-week home program.
Data analysis Descriptive statistics were used to calculate demographic characteristics and average wear time for the group (see table 1 for participant summary). Paired t tests were calculated to determine mean change scores pre- and posttreatment from baseline 1 to discharge, from baseline 1 to follow-up, from baseline 2 to discharge, and from baseline 2 to follow-up. Baseline 1 and baseline 2 were also compared because a therapeutic effect was possible for the device training period. For baseline 1 to discharge, improvement in the FMA-UE (tZ3.32; PZ.01), MAL-AOU; (tZ4.40; PZ.002), and MAL-HW (tZ4.02; PZ.004) scores were statistically significant; mean change scores for the FMA-UE was 3.2 points; for MAS-AOU, .30 points; and for MAL-HW, .27 points (table 2). For baseline 1 to follow-up, improvement in the MAL-AOU (tZ2.61; PZ.035) and MAL-HW (tZ2.47; PZ.043) scores were statistically significant; mean change scores for the MAS-AOU was .14 points; and for MAL-HW, .18 points (table 3). Only the MAL-HW score (tZ2.47; PZ.033) was statistically significant from baseline 2 to discharge, with a mean change score of .19 points. There were no significant findings for baseline 1 to baseline 2 and for baseline 2 to follow-up.
Discussion The study results support our hypothesis that a combined clinichome robotic program integrating the affected arm into functional activities is feasible and a potentially effective therapeutic approach. On the basis of accelerometer wear time data and participant report, participants used the device and practiced arm www.archives-pmr.org
Clinic-home robotics study Table 1
2245
Participant summary
Participant No.
Age (y)
1 2
47 68
3 4 5 6 7 8 9 10 11 Group mean
61 57 48 47 60 49 43 40 49 51.73
Chronicity (y) 1.43 3.49 3.04 3.64 3.04 3.98 9.16 8.59 35.43 7.47 3.44 7.6
FMA-UE Score at Baseline
No. of Training Sessions
14 9
6 9
8 12 26 16 21 30 57 41 9 22.09
9 9 6 9 9 4 3 7 5 6.9
Adverse Events* Technical difficultiesy Skin breakdown; technical difficulties Skin redness; technical difficulties Technical difficulties None Technical difficulties Technical difficulties None Skin irritation Skin irritation Technical difficulties 4 of 11 had skin irritation or breakdown from electrodes; 7 of 11 had technical difficulties
Average Time of Daily Use of the Device (min)
Completed Study?
89.04 48.86
Yes Yes
34.4 43.21 41.9 NA 32.14 53.14 52.14 33.26 NA 42.9
Yes Yes Yes No Yes Yes Yes Yes No 9 of 11
* Adverse events included skin redness and irritation as well as minor breakdown. All cases were related to the adhesives from the electromyography electrodes. We used a combination of strategies to successfully address the adverse events: break from device use (1d to 1wk), reeducation of electromyography electrode placement, and skin protectants. No participants withdrew because of skin issues, and symptoms were resolved. y Technical difficulties included problems with battery charging; needing additional device training during home exercise program; difficulty with electrode placement, wires, straps with one hand; and accurate placement of the brace on the affected arm.
training on their own, averaging 42.9min/d. However, this included time required for setup, troubleshooting, and donning/doffing the device, which varied widely across participants. Therefore, the true dosage of daily training was most likely less than that indicated by our data. Participants demonstrated statistically significant improvements in both arm impairment and self-reported use of the arm from baseline to discharge; they continued to report significant improvement in actual use of the arm at 3-month follow-up. Interestingly, we did not find significant therapeutic effects of the device training period or the 6-week home program alone; however, an analysis of a combined intervention yielded significant improvement in arm impairment and function. These findings may simply represent a dose-response curve that was achieved only with the entire training and therapeutic program, or they might indicate that combining a comprehensive training period with a home program may have a synergistic effect. Most participants completed the study, with 2 early withdrawals.
Table 2 Outcome Measure FMA-UE AMAT MAL-AOU MAL-HW MAS-B MAS-T BB
Study limitations There were several limitations of this study. Because of the small sample size and the absence of a control group, the results should be interpreted with caution. Our results were statistically significant, but did not achieve minimally clinical important difference levels for the FMA-UE or MAL. For subsequent follow-up studies, we suggest a larger sample size with sufficient power to yield definitive statistical results. Because of the general nature of the data captured by commercially available accelerometers, a true dosage of arm training at home was difficult to determine. Considerable commitment and effort is required from both the stroke survivor and the clinician to make this type of program successful, limiting this approach to highly motivated patienttherapist teams (see appendix 4 for recommendations for future studies). The device used in this study was limited to the elbow joint, and no direct training of the distal wrist or hand was addressed.
Paired t test results for baseline 1 to discharge Mean Change Score (points) 3.22 0.04 0.30 0.27 0.06 0.33 0.22
SD 2.91 0.21 0.20 0.20 0.68 0.75 2.28
SEM .97 .07 .07 .07 .23 .25 .76
t 3.33 0.57 4.41 4.02 0.58 1.34 0.29
Significance (P
