Yu-wei Hsieh, Ching-yi Wu, Keh-chung Lin, Grace Yao, Kuen-yuh Wu and Ya-ju Changof Initial Motor Status

Response Relationship of Robot-Assisted Stroke Motor Rehabilitation: The Impact−Dose

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 2012 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/STROKEAHA.112.658807

2012;43:2729-2734; originally published online August 14, 2012;Stroke. 

http://stroke.ahajournals.org/content/43/10/2729World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://stroke.ahajournals.org//subscriptions/

is online at: Stroke Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer process is available in the

Request Permissions in the middle column of the Web page under Services. Further information about thisOnce the online version of the published article for which permission is being requested is located, click

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Strokein Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from

Dose–Response Relationship of Robot-Assisted StrokeMotor Rehabilitation

The Impact of Initial Motor Status

Yu-wei Hsieh, PhD; Ching-yi Wu, ScD; Keh-chung Lin, ScD; Grace Yao, PhD;Kuen-yuh Wu, PhD; Ya-ju Chang, PhD

Background and Purpose—The increasing availability of robot-assisted therapy (RT), which provides quantifiable,reproducible, interactive, and intensive practice, holds promise for stroke rehabilitation, but data on its dose–responserelation are scanty. This study used 2 different intensities of RT to examine the treatment effects of RT and the effecton outcomes of the severity of initial motor deficits.

Methods—Fifty-four patients with stroke were randomized to a 4-week intervention of higher-intensity RT, lower-intensityRT, or control treatment. The primary outcome, the Fugl-Meyer Assessment, was administered at baseline, midterm, andposttreatment. Secondary outcomes included the Medical Research Council scale, the Motor Activity Log, and thephysical domains of the Stroke Impact Scale.

Results—The higher-intensity RT group showed significantly greater improvements on the Fugl-Meyer Assessment thanthe lower-intensity RT and control treatment groups at midterm (P�0.003 and P�0.02) and at posttreatment (P�0.04and P�0.02). Within-group gains on the secondary outcomes were significant, but the differences among the 3 groupsdid not reach significance. Recovery rates of the higher-intensity RT group were higher than those of the lower-intensityRT group, particularly on the Fugl-Meyer Assessment. Scatterplots with curve fitting showed that patients withmoderate motor deficits gained more improvements than those with severe or mild deficits after the higher-intensity RT.

Conclusions—This study demonstrated the higher treatment intensity provided by RT was associated with better motoroutcome for patients with stroke, which may shape further stroke rehabilitation.

Clinical Trial Registration—URL: http://clinicaltrials.gov. Unique identifier: NCT00917605.(Stroke. 2012;43:2729-2734.)

Key Words: randomized controlled trial � robot-aided neurorehabilitation � stroke � treatment dosage

Stroke remains a common cause of acquired adult disabil-ity worldwide.1,2 Motor deficits of the upper limb are

often a devastating disability for stroke survivors3 and thus,the search for effective and efficient rehabilitation to promotemotor recovery becomes urgent. Robot-assisted therapy (RT)is an innovative approach to stroke rehabilitation that usesintensive, repetitive, interactive, and individualized practiceas an optimal strategy to enhance motor learning.4,5

The optimal dosage for specific rehabilitation regimens toinduce improvement is unclear from current evidence.6 Be-cause the research suggests that intensive therapy has apositive influence on stroke recovery,7,8 RT may offer a goodway to close the gap between limited rehabilitation resourcesand a greater amount of therapy.9 RT also provides preciseand quantifiable control of therapy, allowing better research

into treatment dosage.10 This critical factor—how the inten-sity of therapy influences the effects of RT—should beaddressed to inform the dose–response relation and to seekproper treatment intensity for patients.

Another emerging concern is which intervention is mostbeneficial for which type of patient under specific circum-stances.11 Identifying the factors affecting successful out-comes and the patients most likely to respond to the therapywould be informative for clinical guidelines. Although theinitial motor status of patients with stroke is viewed as animportant factor that influences recovery,12,13 whether thisfactor affects the outcomes of RT remains unknown. Wetherefore investigated the effects of RT on clinical outcomesin patients with chronic stroke by using higher-intensity andlower-intensity RT relative to a duration-matched control

Received March 27, 2012; final revision received July 15, 2012; accepted July 18, 2012.From the School of Occupational Therapy (K.C.L.), College of Medicine, the Department of Psychology (G.Y.), and the Institute of Occupational

Medicine and Industrial Hygiene (K.Y.W.), National Taiwan University, Taipei, Taiwan; the Departments of Occupational Therapy and Graduate Instituteof Behavioral Science (Y.W.H., C.Y.W.) and Physical Therapy (Y.J.C.), Chang Gung University, Taoyuan, Taiwan; and the Division of OccupationalTherapy, Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital, Taipei, Taiwan (K.C.L.).

Correspondence to Keh-chung Lin, ScD, School of Occupational Therapy, College of Medicine, National Taiwan University, 17, F4, Xu Zhou Road,Taipei, Taiwan 100. E-mail kehchunglin@ntu.edu.tw

© 2012 American Heart Association, Inc.

Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.112.658807

2729 by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from

treatment (CT) and also examined whether the initial severityof motor deficits and the treatment intensities of RT interactto influence the primary outcome.

MethodsParticipantsCriteria for study participants were (1) more than 6 months’ onsetfrom a unilateral stroke; (2) baseline upper extremity score on theFugl-Meyer Assessment (FMA) of 26 to 56; (3) no excessivespasticity in forearm and wrist joints (modified Ashworth scale �3);(4) able to follow study instructions and perform study tasks(Mini-Mental Status Examination �22); (5) no upper limb fracturewithin 3 months or painful arthritis of the joints; and (6) no severe

neuropsychologic impairments (eg, global aphasia or severe atten-tion deficits). The Institutional Review Boards of the participatinghospitals approved the study, and all participants providedinformed consent.

Study ProceduresThis was a randomized-block controlled trial with pretest, midterm,and posttest evaluations. Eligible participants were stratified accord-ing to side of lesion and level of motor deficits and individuallyrandomized to receive one of the 3 interventions. A random numbertable was used to generate randomization assignments, and aresearch assistant allocated the patients to an intervention groupaccordingly. All clinical measures were administered to the patientsat baseline and immediately after the intervention by the sameblinded rater. The primary outcome measure was also administered2 weeks after treatment began (midterm).

InterventionsAll participants received a duration-matched intervention for 90 to105 minutes/day, 5 days/week for 4 weeks. The patients in the 2 RTgroups practiced with the Bi-Manu-Track (Reha-Stim Co, Berlin,Germany; Figure 1), which allows 2 movement patterns: forearmpronation–supination and wrist flexion– extension.14 Each move-ment pattern is enabled by 3 computer-controlled modes: passive–passive (mode 1), active–passive (mode 2), and active–activemode (mode 3). The parameters of movement and resistance canbe adjusted individually. The robot was equipped with a computergame to provide instant visual movement feedback and toincrease participation.

Within one training session, the patients in the higher-intensity RTgroup practiced 600 to 800 repetitions of modes 1 and 2 for 15 to 20minutes and 150 to 200 repetitions of mode 3 for 3 to 5 minutes forbilateral forearm and wrist movements. One repetition indicated onemovement direction. Patients in the higher-intensity RT received

Figure 1. Bi-Manu-Track.

326 Patients with stroke were

assessed for eligibility

54 Underwent randomization

272 Were excluded

232 Did not meet inclusion

criteria

40 Refused to participate

18 Were assigned to

lower-intensity

robot-assisted therapy

18 Were assigned to

control treatment

18 Were assigned to

higher-intensity

robot-assisted therapy

18 Completed study

and 18 were analyzed

18 Completed study

and 18 were analyzed

17 Completed study and

18 were analyzed

1 Drop out (due to hydronephrosis)

Figure 2. Flow chart of participant enrollment.

2730 Stroke October 2012

by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from

twice the number of the repetitions per unit of time than patients inthe lower-intensity RT group.15

Before the RT training, 5 minutes of mobilization warm-up wereprovided. After the training, the patients received 15 to 20 minutes offunctional activities practice to help them transfer the acquired motorability to their performance of daily activities.

The CT group received an intensive therapist-administered controltherapy matched in duration with the RT groups. Occupationaltherapy techniques used in the treatment protocols included neuro-developmental treatment, muscle strengthening, fine-motor training,and functional task training.

Outcome MeasuresThe primary outcome was a change in the FMA. The 33 upperextremity items of the FMA, with scores ranging from 0 to 66, wereused to assess motor impairment.16 The reliability, validity, andresponsiveness of the FMA in patients with stroke have been wellestablished.16–18

Secondary outcomes included the following assessments: (1) theMedical Research Council scale, a reliable measurement rangingfrom 0 (no contraction) to 5 (normal power), examines muscle powerof the affected arm19; (2) the Motor Activity Log consists of a30-question interview in which patients rate the amount of use andquality of movements at the time of using their affected arm toaccomplish daily activities20; and (3) the Stroke Impact Scale (SIS)3.0 has 4 physical domains—strength, activities of daily living/instrumental activities of daily living, mobility, and hand function—that use patient report to evaluate function and quality of life.21

Two common complications after stroke, pain and fatigue,22 weremeasured to investigate if intensive rehabilitation causes adverseeffects. The therapist asked the patient to rate the severity of his orher pain and fatigue during the intervention on a scale of 0 (no painand no fatigue) to 10 (unbearable pain and exhaustion).

Statistical AnalysisAn intention-to-treat analysis was applied. Two-way repeated-measures analysis of covariance was used to evaluate efficacy of theprimary outcome among the 3 groups followed by a post hoc analysisusing the Bonferroni test for a significant effect. Analysis ofcovariance was used to evaluate treatment efficacy for the secondaryoutcomes with baseline scores as the covariates. The t test was usedto examine recovery rates of each week on each outcome (ie,improved score divided by the number of weeks) between the 2 RTgroups. We also examined whether the initial severity of motorimpairments affected the primary outcome (ie, FMA). Scatterplotswith quadratic curves were used to illustrate the relationship betweenthe baseline scores and the change scores.

ResultsThe study enrolled 54 patients. One patient in the CT groupdropped out due to a medical problem unrelated to the studytreatment (Figure 2). The 3 groups did not differ significantlyin baseline characteristics (P�0.40–0.93; Table 1). As de-termined by the obtained effect size of the primary outcome,post hoc power was calculated to be 0.80.

Primary OutcomeOn the FMA total score, there was a significant group�timeinteraction effect (F3.4,83.8�3.95, P�0.01). All 3 groupsshowed significant within-group gains on the FMA totalscore from baseline to midterm and from baseline to post-treatment (all P�0.05; Table 2). Analysis of covariancerevealed significant differences among the 3 groups at mid-term (F2,50�6.97, P�0.002) and at posttreatment(F2,50�4.80, P�0.01). Post hoc analyses showed that thehigher-intensity RT group had significantly greater improve-ments on the FMA total score than the lower-intensity RT and

CT groups at midterm (P�0.003 and P�0.02) and at post-treatment (P�0.04 and P�0.02; respectively; Table 2). Nosignificant differences were found between the lower-intensity RT and CT groups at midterm and at posttreatment.A similar effect was also found on the FMA distal score(Table 2).

Table 1. Baseline Characteristics of the Study Participants byGroup (N�54)

Characteristics*

Higher-Intensity RT

(n�18)

Lower-Intensity RT

(n�18)

ControlTreatment

(n�18) P Value

Age, y 56.51 (10.03) 52.21 (12.20) 54.83 (9.84) 0.49

Time afterstroke, mo

28.67 (13.67) 23.28 (15.37) 22.44 (15.34) 0.40

Sex, no.

Male 11 13 12 0.78

Female 7 5 6

Stroke subtype, no.

Ischemic 12 11 9 0.73

Hemorrhagic 6 7 8

Subarachnoid 0 0 1

Side of stroke, no.

Right 9 9 8 0.93

Left 9 9 10

MMSE score 28.50 (1.98) 28.00 (2.50) 28.28 (2.08) 0.79

FMA score 42.78 (8.86) 43.11 (9.18) 44.61 (11.06) 0.84

RT indicates robot-assisted therapy; MMSE, Mini-Mental State Examination;FMA, Fugl-Meyer Assessment.

*Continuous data are presented as the mean (SD); categoric data arepresented as indicated.

Table 2. Descriptive Statistics and Group Comparisons on thePrimary Outcome

Outcome

Higher-Intensity RTMean (SD)

Lower-Intensity RTMean (SD)

ControlTreatmentMean (SD)

FMA total score

Baseline 42.78 (8.86) 43.11 (9.18) 44.61 (11.06)

Midterm 46.06 (8.50)*†‡ 44.50 (9.69)* 46.33 (10.50)*

Posttreatment 48.00 (8.22)*†‡ 46.33 (10.27)* 47.56 (10.50)*

FMA distal score

Baseline 12.56 (6.17) 11.44 (6.81) 13.39 (7.65)

Midterm 14.00 (5.99)*‡ 12.22 (7.14)* 14.06 (7.67)*

Posttreatment 15.17 (5.93)*‡ 13.06 (7.53)* 14.72 (7.51)*

FMA proximal score

Baseline 30.22 (4.01) 31.67 (3.96) 31.22 (4.60)

Midterm 32.06 (3.76)* 32.28 (4.17) 32.28 (4.07)*

Posttreatment 32.83 (3.62)* 33.28 (3.72)* 32.83 (4.25)*

RT indicates robot-assisted therapy; FMA, Fugl-Meyer Assessment.With-group comparison: *P�0.05 when compared with baseline scores.Between-group comparison: †P�0.05 when the higher-intensity RT group

score was greater than the lower-intensity RT group.‡P�0.05, when the higher-intensity RT group score was greater than the

control treatment group.

Hsieh et al Dose–Response of Robot-Assisted Stroke Rehabilitation 2731

by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from

Secondary Outcomes and Adverse ResponsesThe 3 groups made significant within-group improvementsover time (all P�0.05) on the Medical Research Council andMotor Activity Log; however, the improvements were notsignificantly different among the 3 groups on the MedicalResearch Council (F2,50�1.41, P�0.25), the Motor ActivityLog quality of movements (F2,50�2.38, P�0.10), or theMotor Activity Log amount of use (F2,50�1.61, P�0.21).The higher-intensity RT group reported significant within-group improvements on the SIS–strength (P�0.002) and SIS– activities of daily living/instrumental activities of dailyliving (P�0.02) assessments. The lower-intensity RT grouphad significant within-group improvements on the SIS–strength assessment (P�0.02). The CT group, however, didnot report significant improvements on the 4 SIS physicaldomains (P�0.07–0.29). The between-group comparison didnot show a significant difference among the 3 groups forgains on the 4 SIS physical domains (F2,50�0.40–1.38,P�0.26–0.67). The 3 groups showed mild ratings for fatigueand pain (mean score of �3 of 10 possible).

Recovery RatesRecovery rates of the higher-intensity RT group on the FMAtotal and distal scores were significantly higher than those ofthe lower-intensity RT group at midterm and posttreatment(all P�0.05; Table 3). On the secondary outcomes, differ-ences in recovery rates between the 2 RT groups were notsignificant (P� 0.10–0.86; Table 3); however, the recoveryrate values for the higher-intensity RT group were generallyhigher than those for the lower-intensity RT group.

Interaction Between Initial Motor Status andTreatment IntensityFigure 3 shows the relationships between patients’ baselinescores and change scores on the FMA. The data fit a quadraticfunction, and the critical point of the fitted curve in thehigher-intensity RT group at posttreatment was calculated tobe 38.41 by differentiating the equation of the curve (Figure3); that is, patients with an FMA baseline score of approxi-mately 40 showed the most gains after the higher-intensityRT. In the lower-intensity RT group, patients with fewermotor deficits gained more benefits on the FMA; however,the general gains of this group were lower than the gains inthe higher-intensity RT group. To summarize, the patientswith motor deficits in the middle range (FMA score ofapproximately 40) had more improvement in motor abilityafter the higher-intensity RT than those with severe or mildmotor deficits.

DiscussionThe 3 treatment groups in this trial had significant within-group gains on the FMA, Medical Research Council, andMotor Activity Log, indicating that the patients benefitedfrom the intervention in motor ability, muscle power, andself-perceived performance in daily activities. The 2 RTgroups also reported significant improvements in the strengthdomain of the SIS over time. Between-group comparisonsshowed that the patients who received the higher-intensity RThad significant improvements in primary motor ability at

midterm and posttreatment compared with those who re-ceived the lower-intensity RT or CT; however, the 2 RTprotocols and the CT demonstrated comparable effects onimproving the secondary outcomes.

When these results were compared with those in the trialby Hesse et al,14 which used the same robotic device, theeffects on the primary outcome (ie, FMA) were different. Thetreatment intensity of RT used in their study was similar tothe lower-intensity RT used in this study. The RT group in theHesse et al study, however, had significant benefits comparedwith the control group, whereas we did not find a significantdifference between the lower-intensity RT and the CT. Thediffering results may be attributable to differences in treat-ment protocols of the control groups (electric stimulationversus conventional occupational therapy), stroke phases(subacute versus chronic), and baseline motor deficits ofpatients (severe versus mild to moderate). Moreover, thehigher-intensity RT (twice as many repetitions as in thelower-intensity RT) of this study led to significant gains inmotor ability compared with the other 2 groups. Our findingssuggest that for patients with chronic stroke with mild tomoderate motor deficits, the treatment intensity of RT usingthe Bi-Manu-Track can be higher than the original protocol14

to reach superior effects on motor recovery.

Table 3. Recovery Rates of Each Week at Midterm andPosttreatment on the Outcomes

Outcome

Higher-Intensity RTMean (SD)

Lower-Intensity RTMean (SD)

PValue

FMA total score

Midterm 1.64 (1.12) 0.69 (0.52) 0.003

Posttreatment 1.31 (0.65) 0.81 (0.50) 0.01

FMA distal score

Midterm 0.72 (0.43) 0.39 (0.40) 0.02

Posttreatment 0.65 (0.39) 0.40 (0.33) 0.05

FMA proximal score

Midterm 0.92 (0.94) 0.31 (0.52) 0.02

Posttreatment 0.65 (0.46) 0.40 (0.46) 0.11

MRC

Posttreatment 0.07 (0.09) 0.03 (0.05) 0.14

MAL-QOM

Posttreatment 0.11 (0.12) 0.05 (0.09) 0.10

MAL-AOU

Posttreatment 0.10 (0.09) 0.05 (0.09) 0.11

SIS–strength

Posttreatment 2.00 (2.26) 2.17 (3.40) 0.86

SIS–ADL/IADL

Posttreatment 1.17 (1.93) 0.43 (2.45) 0.32

SIS–mobility

Posttreatment 0.62 (2.25) 0.26 (2.29) 0.63

SIS–hand function

Posttreatment 2.67 (5.95) 1.07 (5.29) 0.40

RT indicates robot-assisted therapy; FMA, Fugl-Meyer Assessment; MRC,Medical Research Council scale; MAL-QOM, Motor Activity Log quality ofmovement; AOU, amount of use; SIS, Stroke Impact Scale; ADL/IADL, activitiesof daily living/instrumental ADL.

2732 Stroke October 2012

by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from

The study results for the functional-based or disabilityoutcomes are concordant with previous RT studies23,24 thatdid not find significant differences between RT and controlgroups. Within-group analyses, however, showed that the 2RT groups had significant gains in self-perceived amount ofuse and quality of movement of the affected hand in dailyactivities and in the strength domain of the SIS evaluation.These gains may be attributed to the additional 15 to 20minutes of functional training per session in our RT proto-cols. The supplementary functional practice may be necessaryto enhance functional performance but may not be sufficientfor significant between-group differences compared with theCT. Most robotic devices are designed for the practice ofrelatively simple movements, and patients who train on themhave some difficulty in transferring the gained motor abilityto the performance of functional activities.25 Further ap-proaches may consider the combination of RT and task-oriented training (eg, constraint-induced therapy) in strokerehabilitation to promote motor recovery and functionalimprovement.

The recovery rate data suggest a trend for better motoroutcome with the higher-intensity RT at midterm and post-treatment, indicating a dose–response relationship such thatproviding more intense RT may accelerate motor improve-ment for chronic stroke. Current findings support that highlyintensive therapy tends to accelerate recovery.26 The midtermevaluation can provide more information about the dynamicresponse of patients during the treatment course; however,whether the intensity effect occurs mainly in the first half orthroughout the entire period of the treatment course needsfurther research. For patients who have already reached afunctional plateau at midterm, incorporating or receiving

other interventions after midterm (eg, RT in sequentialcombination with constraint-induced therapy) might promotefurther improvement.

The initial severity levels of the patients may influence thetreatment effects on the primary outcome. The FMA iscommonly used to identify the severity levels of patients withstroke.27–29 The relation of FMA baseline scores and changescores in this study was dependent on the intensity of RTtreatment. As seen in the graph (Figure 3), patients withmoderate motor deficits (ie, FMA score of approximately 40)showed more motor improvement after the higher-intensityRT than those with severe or mild deficits. However, therewas a trend that the patients with fewer motor deficits in thelower-intensity RT group gained more benefits on the FMA.Further studies may examine the usefulness of the FMAcutoff score to more accurately stratify patients in RT trials.

Some limitations of the study warrant consideration. Onlyone intermediate assessment was conducted during the studyperiod. To more clearly define dose–response curves of RT,further investigation of the efficacy of RT on multipleintermediate data points is warranted. This study did notevaluate sensory function. Patients with sensory deficits whoreceive interactive RT with sensory inputs and feedback maybenefit from RT.30 Further trials should assess the role ofsensory function as a treatment outcome and potential mod-erator of the outcomes. Another study limitation is that theparticipants and the intervention providers were not blindedto the treatment group, which might have led to some bias.

In conclusion, our study results support that treatmentoutcomes, especially in motor improvements, were betterafter the higher-intensity RT than after the other 2 interven-tions. Higher-intensity RT may be feasible in patients with

Figure 3. The relationships of initial values and change scores on the Fugl-Meyer Assessment (FMA) for the 2 robot-assisted therapygroups. Scatterplots show the distribution of pretreatment and change scores on the FMA. Quadratic function was applied for curvefitting. The dotted lines show the critical point (38.41) of the fitted curve in the higher-intensity robot-assisted therapy (RT) group atposttreatment.

Hsieh et al Dose–Response of Robot-Assisted Stroke Rehabilitation 2733

by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from

chronic stroke to accelerate the rate of motor recovery.Patients with moderate motor deficits tended to have moremotor improvements after the higher-intensity RT than thosewith severe or mild motor deficits. The patient’s level ofmotor impairment should be considered when planning forrobot-assisted stroke rehabilitation.

Sources of FundingThis project was supported in part by the National Health ResearchInstitutes (NHRI-EX101-9920PI and NHRI-EX101-10010PI), theNational Science Council (NSC-100-2314-B-002-008-MY3 andNSC 99-2314-B-182-014-MY3), and the Healthy Ageing ResearchCenter at Chang Gung University (EMRPD1A0891) in Taiwan.

DisclosuresNone.

References1. Lawrence ES, Coshall C, Dundas R, Stewart J, Rudd AG, Howard R, et

al. Estimates of the prevalence of acute stroke impairments and disabilityin a multiethnic population. Stroke. 2001;32:1279–1284.

2. Muntner P, Garrett E, Klag MJ, Coresh J. Trends in stroke prevalencebetween 1973 and 1991 in the us population 25 to 74 years of age. Stroke.2002;33:1209–1213.

3. Cramer SC. Changes in motor system function and recovery after stroke.Restor Neurol Neurosci. 2004;22:231–238.

4. Mirelman A, Bonato P, Deutsch JE. Effects of training with a robot-virtual reality system compared with a robot alone on the gait of indi-viduals after stroke. Stroke. 2009;40:169–174.

5. Brewer BR, McDowell SK, Worthen-Chaudhari LC. Poststroke upperextremity rehabilitation: a review of robotic systems and clinical results.Top Stroke Rehabil. 2007;14:22–44.

6. Wallace AC, Talelli P, Dileone M, Oliver R, Ward N, Cloud G, et al.Standardizing the intensity of upper limb treatment in rehabilitationmedicine. Clin Rehabil. 2010;24:471–478.

7. Cooke EV, Mares K, Clark A, Tallis RC, Pomeroy VM. The effects ofincreased dose of exercise-based therapies to enhance motor recoveryafter stroke: a systematic review and meta-analysis. BMC Med.2010;8:60.

8. Kwakkel G, van Peppen R, Wagenaar RC, Wood Dauphinee S, RichardsC, Ashburn A, et al. Effects of augmented exercise therapy time afterstroke: a meta-analysis. Stroke. 2004;35:2529–2539.

9. Teasell RW, Kalra L. What’s new in stroke rehabilitation: back to basics.Stroke. 2005;36:215–217.

10. Reinkensmeyer DJ, Emken JL, Cramer SC. Robotics, motor learning, andneurologic recovery. Annu Rev Biomed Eng. 2004;6:497–525.

11. Dobkin BH. Progressive staging of pilot studies to improve phase III trialsfor motor interventions. Neurorehabil Neural Repair. 2009;23:197–206.

12. Fritz SL, Light KE, Patterson TS, Behrman AL, Davis SB. Active fingerextension predicts outcomes after constraint-induced movement therapyfor individuals with hemiparesis after stroke. Stroke. 2005;36:1172–1177.

13. Lin KC, Huang YH, Hsieh YW, Wu CY. Potential predictors of motorand functional outcomes after distributed constraint-induced therapy forpatients with stroke. Neurorehabil Neural Repair. 2009;23:336–342.

14. Hesse S, Werner C, Pohl M, Rueckriem S, Mehrholz J, Lingnau ML.Computerized arm training improves the motor control of the severelyaffected arm after stroke: a single-blinded randomized trial in two centers.Stroke. 2005;36:1960–1966.

15. Hsieh YW, Wu CY, Liao WW, Lin KC, Wu KY, Lee CY. Effects oftreatment intensity in upper limb robot-assisted therapy for chronicstroke: a pilot randomized controlled trial. Neurorehabil Neural Repair.2011;25:503–511.

16. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. 1. A method for evaluation of physical per-formance. Scand J Rehabil Med. 1975;7:13–31.

17. Platz T, Pinkowski C, van Wijck F, Kim IH, di Bella P, Johnson G.Reliability and validity of arm function assessment with standardizedguidelines for the Fugl-Meyer Test, Action Research Arm Test and Boxand Block Test: a multicentre study. Clin Rehabil. 2005;19:404–411.

18. Hsieh YW, Wu CY, Lin KC, Chang YF, Chen CL, Liu JS. Respon-siveness and validity of three outcome measures of motor function afterstroke rehabilitation. Stroke. 2009;40:1386–1391.

19. Gregson JM, Leathley MJ, Moore AP, Smith TL, Sharma AK, WatkinsCL. Reliability of measurements of muscle tone and muscle power instroke patients. Age Ageing. 2000;29:223–228.

20. van der Lee JH, Beckerman H, Knol DL, de Vet HC, Bouter LM.Clinimetric properties of the Motor Activity Log for the assessment ofarm use in hemiparetic patients. Stroke. 2004;35:1410–1414.

21. Duncan PW, Bode RK, Min Lai S, Perera S. Rasch analysis of a newstroke-specific outcome scale: the Stroke Impact Scale. Arch Phys MedRehabil. 2003;84:950–963.

22. Yorkston KM, Johnson K, Boesflug E, Skala J, Amtmann D. Communi-cating about the experience of pain and fatigue in disability. Qual LifeRes. 2010;19:243–251.

23. Housman SJ, Scott KM, Reinkensmeyer DJ. A randomized controlledtrial of gravity-supported, computer-enhanced arm exercise for indi-viduals with severe hemiparesis. Neurorehabil Neural Repair. 2009;23:505–514.

24. Volpe BT, Lynch D, Rykman-Berland A, Ferraro M, Galgano M, HoganN, et al. Intensive sensorimotor arm training mediated by therapist orrobot improves hemiparesis in patients with chronic stroke. NeurorehabilNeural Repair. 2008;22:305–310.

25. Rosati G. The place of robotics in post-stroke rehabilitation. Expert RevMed Devices. 2010;7:753–758.

26. Kwakkel G. Impact of intensity of practice after stroke: issues for con-sideration. Disabil Rehabil. 2006;28:823–830.

27. Daly JJ, Hogan N, Perepezko EM, Krebs HI, Rogers JM, Goyal KS, et al.Response to upper-limb robotics and functional neuromuscular stimu-lation following stroke. J Rehabil Res Dev. 2005;42:723–736.

28. Luft AR, McCombe-Waller S, Whitall J, Forrester LW, Macko R, SorkinJD, et al. Repetitive bilateral arm training and motor cortex activation inchronic stroke: a randomized controlled trial. JAMA. 2004;292:1853–1861.

29. Thielman G, Kaminski T, Gentile AM. Rehabilitation of reaching afterstroke: comparing 2 training protocols utilizing trunk restraint. NeurorehabilNeural Repair. 2008;22:697–705.

30. Sivan M, O’Connor RJ, Makower S, Levesley M, Bhakta B. Systematicreview of outcome measures used in the evaluation of robot-assistedupper limb exercise in stroke. J Rehabil Med. 2011;43:181–189.

2734 Stroke October 2012

by guest on May 20, 2014http://stroke.ahajournals.org/Downloaded from