Predictors of Motor,Daily Function,and Quality-of-Life Improvements After Upper-Extremity Robot-Assisted Rehabilitation in Stroke

Abstract

OBJECTIVE. A subgroup of patients benefiting most from robot-assisted therapy (RT) has not yet been described. We examined the predictors of improved outcomes after RT.

METHOD. Sixty-six patients with stroke receiving RT were analyzed. The outcome measures were the Fugl-Meyer Assessment (FMA), Wolf Motor Function Test (WMFT), Motor Activity Log (MAL), and Stroke Impact Scale (SIS). The potential predictors were age, side of lesion, time since onset, Modified Ashworth Scale (MAS) scores, accelerometer data, Box and Block Test (BBT) scores, and kinematic parameters.

RESULTS. BBT scores were predictive of FMA (29%) and MAL (9%−15%) improvements. Reduced shoulder flexion synergy, as measured by less shoulder abduction during forward reach, and MAS–distal were predictive of WMFT–function improvements. MAS–distal was predictive of SIS–physical improvements. Demographic variables did not predict outcomes.

CONCLUSION. Manual dexterity was a valuable predictor of motor impairment and daily function after RT. Outcomes at different levels may have different predictors.

Keywords

forecasting motor skills outcome assessment (health care)robotics stroke

Using robotic devices to achieve highly intensive, task-specific, reproducible, and interactive practice brings a relatively new scientific concept to clinical practice (Harwin, Patton, & Edgerton, 2006; Waldner, Tomelleri, & Hesse, 2009). Robot-assisted therapy (RT) shows great promise for patients with stroke and has had a significant impact on neurorehabilitation during the past 15 yr (Hesse et al., 2005; Hsieh et al., 2011; Lo et al., 2010; Volpe et al., 1999). The results of meta-analyses support the treatment efficacy of RT in the arm on motor and functional outcomes in stroke patients (Kwakkel, Kollen, & Krebs, 2008; Mehrholz, Platz, Kugler, & Pohl, 2008).

After a decade of research on stroke rehabilitation using RT, a subgroup of patients who may benefit most from this type of therapy still cannot be clearly defined (Dobkin, 2009). We know many factors interact with and affect the outcomes of stroke rehabilitation (Franceschini et al., 2012). However, existing evidence on this topic is limited. Given the growing accessibility of RT, the identification of predictors to tailor treatment plans and more accurately stratify patients toward a better outcome of stroke rehabilitation is also of value to clinicians and researchers (Brewer, McDowell, & Worthen-Chaudhari, 2007).

We conducted this study to identify the predictors of outcomes in stroke patients after RT to help target the optimized individuals to receive the intervention. The candidate predictors were age; side of lesion; time since stroke onset; muscle tone in distal hand; hand usage in daily life; manual dexterity; and joint movements of shoulder, elbow, and trunk. These predictors were selected on the basis of previous studies of specific stroke rehabilitation interventions other than RT. For instance, studies have found that time since stroke, side of stroke lesion, finger extension, and Fugl-Meyer Assessment (FMA; Fugl-Meyer, Jääskö, Leyman, Olsson, & Steglind, 1975) arm and distal part scores were predictive of improved outcomes after constraint-induced therapy or bilateral arm training (Fritz et al., 2006; Fritz, Light, Patterson, Behrman, & Davis, 2005; Kwakkel, Kollen, van der Grond, & Prevo, 2003; Lin, Huang, Hsieh, & Wu, 2009; Park, Wolf, Blanton, Winstein, & Nichols-Larsen, 2008; Waller & Whitall, 2005).

One unique aspect of this study is that we included kinematic measures quantifying movement quality in the paretic arm (Roby-Brami et al., 2003) at baseline assessments as potential predictors. We used these kinematic data to quantify flexion synergy (i.e., abduction and external rotation of the shoulder, elbow flexion, and forearm supination while elevating the paretic arm), which is one common upper-extremity (UE) synergy after stroke (Brunnstrom, 1970). This abnormal synergy makes extending the elbow of the paretic hand difficult (Messier, Bourbonnais, Desrosiers, & Roy, 2006). Also included as potential predictors were trunk flexion, as an indicator of compensatory movement because of UE synergy, and accelerometry data. Accelerometry is an objective measure to monitor actual use of affected extremities and has been used to validate the Motor Activity Log (MAL; Uswatte et al., 2006).

Our research question was, What are the predictors of outcomes after RT in stroke rehabilitation? We aimed to develop predictive models of outcome improvement from baseline to posttreatment. Predictors that significantly contributed to the changes in motor impairment (measured by the FMA), motor function (measured by the Wolf Motor Function Test [WMFT; Wolf et al., 2001]), daily function (measured by the MAL), and self-perceived quality of life (measured by the Stroke Impact Scale [SIS; Lai, Studenski, Duncan, & Perera, 2002]) were determined.

Method

Research Design

This study was a secondary analysis of data from cohorts receiving RT therapy in previous studies (Hsieh et al., 2012; Yang, Lin, Chen, Wu, & Chen, 2012) and in an ongoing project. Eligible participants received RT interventions with the Bi-Manu-Track device (Reha-Stim Co., Berlin, Germany) for 90–105 min/day, 5 days/wk for 4 wk. Assessments were administered at baseline and after treatment. A correlational design and linear regression analysis was used to determine the demographic and baseline scores that significantly predicted improved outcomes. The institutional review boards of the participating hospitals approved the trials, and all participants provided informed consent.

Participants

The study originally enrolled 66 participants from three studies of RT treatment effects after stroke. The inclusion criteria for all studies were (1) ≥3 mo onset from a unilateral stroke; (2) FMA UE score >18 (Hesse et al., 2005); (3) no excessive spasticity in forearm and wrist joints (Modified Ashworth Scale [MAS; Bohannon & Smith, 1987)] score <3); (4) able to follow study instructions and perform study tasks (Mini-Mental State Examination score ≥20; Castro-Costa, Fuzikawa, Uchoa, Firmo, & Lima-Costa, 2008); (5) no UE fracture ≤3 mo and no painful arthritis or contracture of the joints; and (6) no severe neuropsychologic impairments (e.g., global aphasia or severe attention deficits).

Outcome Measures

We selected outcome measures to follow the International Classification of Functioning, Disability and Health (World Health Organization, 2001) framework to facilitate interpretability of our results (Lemmens, Timmermans, Janssen-Potten, Smeets, & Seelen, 2012). Motor impairments were measured by the FMA; motor function, by the WMFT; daily function, by the MAL; and quality of life, by the SIS–physical subscale. The Upper-Extremity (UE) subscale of the FMA (Fugl-Meyer et al., 1975) was used to assess motor impairment outcomes. The FMA–UE consists of 33 items measuring the movement and reflexes of the shoulder, elbow, forearm, wrist, and hand, as well as coordination and speed, on a 3-point ordinal scale (2: can perform fully). Higher FMA–UE scores indicate less motor impairment. The change scores of the FMA–UE from baseline to posttreatment were calculated as the dependent variables in the linear regression analyses. The FMA–UE has good reliability (r > .95) and validity (ρ = .49–.73) in stroke patients (Hsieh et al., 2009; Platz et al., 2005).

UE function was assessed with the WMFT, which consists of two scales: performance time and functional ability. The WMFT has good reliability (r = .86−.99; Morris, Uswatte, Crago, Cook, & Taub, 2001) and criterion validity (ρ = .45–.77; Hsieh et al., 2009) and has been widely used as an outcome measure in stroke motor rehabilitation trials (Whitall et al., 2011). This study used the functional ability scale (WMFT–function) only, because the collected data did not show statistically significant changes in the time subscale before and after treatment.

The MAL, a functional and patient-reported outcome measure, captures the amount of use (AOU) and quality of movement (QOM) while the patients use their affected arm to accomplish 30 daily activities. The score ranges from 0 to 5; higher scores indicate better performance. The MAL in stroke patients has been reliable (r ≥ .91) and valid (Cronbach’s α > .87, ρ = .63; Lang, Edwards, Birkenmeier, & Dromerick, 2008; van der Lee, Beckerman, Knol, de Vet, & Bouter, 2004).

A self-reported quality-of-life measure, the SIS was used to assess the difficulty level in performing activities during the preceding week. A 5-point Likert scale was used in each scale, with higher scores indicating better function. We used the composite score of the SIS–physical domain, which was computed using the strength, activity of daily living, mobility, and hand function subscores. Previous studies have established the reliability (r = .96) and validity (ρ = .61; Carod-Artal, Coral, Trizotto, & Moreira, 2008) of the SIS.

Potential Predictors

This study selected three demographic, three motor functional, and four kinematic assessments as initial candidate predictors, which were recorded or assessed at baseline. The three demographic predictors were age, side of lesion, and time since stroke onset, which were found to be relevant and important in previous outcome prediction research (Kwakkel et al., 2003). The three motor functional predictors were UE distal muscle tone measured by the MAS–distal, hand usage in daily life measured by accelerometers, and manual dexterity measured by the Box and Block Test (BBT; Mathiowetz, Volland, Kashman, & Weber, 1985). The four kinematic variables selected as predictors were degrees of shoulder flexion, shoulder abduction, elbow extension, and trunk flexion during forward reach.

Motor and Functional Predictors.

The MAS is a reliable scale for assessing muscle tone in stroke patients; scores range from 0 (no increase in muscle tone) to 5 (rigid). The Spearman’s ρ reflecting the interrater reliability of MAS has been found to range from .56 to .90 (Sloan, Sinclair, Thompson, Taylor, & Pentland, 1992). MicroMini-Motionlogger accelerometers (Ambulatory Monitoring, Ardsley, NY) were used to monitor arm activity in the participants’ natural environments. Patients wore the accelerometer nearly 24 hr daily (removed when bathing) for 3 consecutive days before and after treatment. Compliance, monitored by the examiners and the analytic software, was reported as good. Accelerometers have demonstrated sufficient reliability (r = .85) and validity (r = .52−.98; Gebruers, Vanroy, Truijen, Engelborghs, & De Deyn, 2010). Values of affected hand activity level and the ratio of affected to unaffected hand activity level, measured by accelerometer, were used for this study.

The BBT is a measure of manual dexterity with satisfactory reliability (r = .94−.99) and validity (ρ = .35−.80; Lin, Chuang, Wu, Hsieh, & Chang, 2010; Mathiowetz et al., 1985). Participants use the affected hand to move as many blocks as possible from one compartment to another other within 1 min. The BBT score is the number of blocks transferred by the affected hand.

Kinematic Predictors.

Participants were instructed to reach forward with their affected hand to press a desk bell as fast as possible while sitting on a height-adjustable, straight-back chair with seat height at 100% of lower leg length. The table height was adjusted so that the initial position of the hand was on the edge of the table with the elbow flexed at 90°. A 7-camera VICON MX motion capture system (Oxford Metrics, Oxford, England), at a sampling frequency of 120 Hz, was used with a personal computer to record kinematic data. Kinematic data were processed with an analysis program coded by LabVIEW language (National Instruments, Austin, TX). Kinematic variables included degree of movement of shoulder flexion and abduction, elbow extension, and trunk flexion. Degree of movement was defined as the difference of angle change from movement onset to offset. Kinematic analysis has demonstrated good accuracy (Carse, Meadows, Bowers, & Rowe, 2013) and test–retest reliability for kinematic variables (Wu, Yang, Chen, Lin, & Wu, 2013).

Data Collection

Pretest and posttest evaluations were conducted by examiners blinded to the participant group. Before the examiners conducted the assessments, the principal investigators (Lin and Wu) of this study trained them and reviewed their performance in administering the tests.

Data Analysis

The normality of each variable was statistically verified by the value of skewness (±1) and visually verified using histograms. All the variables in this study met the assumptions of normality. Regression analyses were performed to identify significant predictors. The variance inflation factor was used to examine the presence of multicollinearity (redundancy) among the predictors. The initially chosen candidate predictors were entered into the predictive models according to the strength of the bivariate correlations (Pearson’s r; criteria of p < .20; Wee & Hopman, 2005). Multiple linear regression analyses with a forward stepwise procedure were used to predict the change scores on the FMA, WMFT–function, MAL–AOU, MAL–QOM, and the SIS–physical by entering the pretest scores of these outcomes as covariates. Statistical analyses were performed using IBM SPSS Statistics Version 16.0 (SPSS, Inc., Chicago).

Results

The study analyzed data from all 66 participants (45 men), with a mean age of 53 yr. Mean stroke onset was 23 mo, the stroke was ischemic in 39 patients, and the lesion was in the right hemisphere in 34 patients (Table 1).

Table 1.

Baseline Characteristics of the Study Participants (N = 66)

Characteristic	N (%) or M (SD)
Age, yr	53.19 (11.35)
Time since onset, mo	22.56 (15.51)
Gender
Male	45 (68.2)
Female	21 (31.8)
Stroke type
Hemorrhagic	24 (36.4)
Ischemic	39 (59.1)
Others	3 (4.5)
Side of stroke
Right	34 (51.5)
Left	32 (48.5)
MMSE score	27.75 (2.48)
FMA–UE score	41.98 (9.97)
BBT score	13.59 (11.85)
MAS–distal score	0.70 (0.34)

Note. BBT = Box and Block Test; FMA–UE = Fugl-Meyer Assessment—Upper Extremity; M = mean; MAS = Modified Ashworth Scale; MMSE = Mini-Mental State Examination; SD = standard deviation.

Table 2 summarizes the relationships among the remaining candidate predictors and the change scores of the outcomes. All outcomes showed statistically significant improvements from pretest to posttest. The four predictors of time since stroke, accelerometer-affected hand activity level, BBT score, and elbow extension were entered into the model with FMA–UE as the dependent variable. The four predictors of MAS–distal score, BBT, shoulder flexion, and shoulder abduction were entered into the model with WMFT–function as the dependent variable. The three predictors of time since stroke, MAS distal score, and BBT were entered into the models with MAL–AOU and MAL–QOM as the dependent variables. The four predictors of side of lesion, MAS distal score, shoulder flexion, and shoulder abduction were entered into the model for prediction of SIS–physical.

Table 2.

Correlations Between the Candidate Predictors and the Outcomes (Pearson’s r)

	Candidate Predictora
Outcomes	Side of Lesion	Time Since Onset	MAS–distal	Accelerometer Activity Level	BBT	Shoulder Flexion	Shoulder Abduction	Elbow Extension
FMA–UE		−0.25		0.17	0.35			−0.18
WMFT–function			−0.29		0.20	−0.24	−0.36
MAL–AOU		−0.25	−0.26		0.34
MAL–QOM		−0.25	−0.18		0.39
SIS–physical	0.21		−0.25			0.19	0.19

Note. AOU = amount of use; BBT = Box and Block Test; FMA–UE = Fugl-Meyer Assessment—Upper Extremity; MAL = Motor Activity Log; MAS = Modified Ashworth Scale; SIS = Stroke Impact Scale; QOM = quality of movement; WMFT = Wolf Motor Function Test.

Only correlations with p < .20 are shown.

Table 3 summarizes the results of multiple linear regression analyses. The BBT score was the significant predictor. The model accounted for 29% of the variance in change scores of the FMA. A higher baseline BBT score was associated with greater FMA improvement. For WMFT–function, shoulder abduction during forward reach and MAS distal score were the two significant predictors. The model accounted for 16% of variance in change in WMFT–function. Lower baseline scores of shoulder abduction and MAS distal score at baseline predicted greater WMFT–function improvement.

Table 3.

Multiple Linear Regression Analysis for Prediction of Changes in Outcomes

Model	Predictor	Ba	β^b	P
FMA–UE: F(2, 56) = 12.73, p < .01, Adjusted R ² = .288	Constant	8.01		<.01
	Covariate^c	−0.15	−.51	.02
	BBT	0.18	.78	<.01
WMFT–function: F(3, 61) = 5.12, p < .05, Adjusted R ² = .162	Constant	6.28		<.01
	Covariate^c	−0.02	−.08	.51
	Shoulder abduction	−0.12	−.35	<.01
	MAS–distal	−2.42	−.28	.03
MAL–AOU: F(2, 63) = 4.32, p < .05, Adjusted R ² = .093	Constant	4.18		.07
	Covariate^c	0.07	.10	.45
	BBT score	0.29	.29	.04
MAL–QOM: F(2, 63) = 6.68, p < .05, Adjusted R ² = .149	Constant	3.31		.12
	Covariate^c	0.12	.18	.18
	BBT score	0.28	.30	.03
SIS–physical: F(2, 62) = 9.72, p < .01, Adjusted R ² = .214	Constant	35.58		<.01
	Covariate^c	−1.80	−.44	<.01
	MAS–distal	−8.11	−.36	<.01

B indicates the nonstandardized coefficient. ^bβ indicates the standardized coefficient. ^cPretest scores as covariates.

MAL–AOU and MAL–QOM revealed one common significant predictor, BBT score. Time since stroke, although correlated at .25 (see Table 2), was not included in the final regression solution, having been thrown out by the program because of its correlation with BBT scores. The models accounted for 9% and 15% of the variance in MAL–AOU and MAL–QOM change scores, respectively.

The MAS–distal score accounted for 21% of the variance in SIS–physical, with a lower MAS–distal score predicting greater improvement in SIS–physical score. No strong multicollinearity was found among the predictors for these models (the largest variance inflation factor = 1.98).

Discussion

This study is among the first to investigate the possible clinical characteristics of stroke patients that can predict the treatment effects after UE RT. Our results indicate that manual dexterity is a valuable predictor. Patients with higher BBT baseline scores made more improvements on the FMA, the MAL–AOU, and the MAL–QOM after the RT. The BBT, however, predicted only 9%–29% of the variance associated with improvement. In addition, patients who had less shoulder abduction while performing reaching tasks at baseline were likely to make greater improvement on the WMFT. Lower ratings on the MAS–distal at baseline predicted greater positive changes on the WMFT. Lower MAS–distal scores also significantly predicted better SIS–physical scores.

Manual Dexterity

Our findings suggest that manual dexterity, as a requirement for some activities (Geyh et al., 2004), seems to be critical in regaining motor function after RT. This finding is similar to previous results. Multiple studies have found baseline performance of finger extension or FMA to be significantly associated with improvement in motor function after stroke (Fritz et al., 2005; Kwakkel & Kollen, 2007; Kwakkel et al., 2003; Lin et al., 2009). The BBT, a 1-min test for assessing manual dexterity, is easy to administer and can significantly predict a person’s change on motor and functional outcomes. Consequently, the BBT is likely to be a suitable and convenient screening criterion for further RT trials.

Patients with higher manual dexterity had greater gains in AOU and QOM while using their affected hand in executing daily activities. This result is concordant with previous studies showing that manual dexterity greatly affected stroke patients’ ability to independently perform activities of daily living (Rand & Eng, 2010; van Heuvelen, Kempen, Brouwer, & de Greef, 2000; Yancosek & Howell, 2009), was associated with the hand AOU (Rand & Eng, 2010) in older adults, and predicted the living status (totally independent, living at home with social assistance, or nursing home) in older women.

Compensatory Movement (Shoulder Abduction) and Muscle Tone (MAS–distal)

Our results reveal that less shoulder abduction during forward reach (an indication of lessened flexor synergy) and less distal muscle tone at baseline significantly predicted motor function outcome, as measured by WMFT–function. This finding suggests that patients with less shoulder–elbow synergy at baseline might benefit more than those with greater synergy. Some studies have shown that an increase in generation of elbow flexion torque was accompanied by increased levels of shoulder abduction in people with hemiparesis (Beer, Given, & Dewald, 1999; Dewald, Pope, Given, Buchanan, & Rymer, 1995). This poor synergy between the elbow and shoulder makes extending the elbow of the paretic hand difficult (Messier et al., 2006).

We also found that less initial muscle tone predicted greater improvements in the WMFT and self-perceived physical function after RT. Less initial muscle tone in the distal UE might have been associated with less interference in performing tasks involving hand manipulation in the WMFT, such as better finger extension. A prospective cohort study by Formisano et al. (2005) indicated that patients with spasticity initially showed no further improvement after 2 mo of rehabilitation, whereas flaccid patients continued to show motor recovery. Formisano et al. (2005) proposed that recovery mechanisms emerge later or proceed more slowly in flaccid patients. Therefore, flaccid patients may have more potential than spastic patients and, consequently, gain more benefits from RT.

Understanding the factors that are not important predictors of outcomes after treatment is as crucial as knowing the significant predictors (Fritz et al., 2005). The regression analyses included demographic variables, hand usage measured by accelerometers, and some kinematic parameters (i.e., shoulder flexion, elbow extension, and trunk flexion), but these data were not significant to the prediction of improvement after RT. The use of the accelerometers has challenges; for example, patients might have had unusually active or inactive routines for the sampled 3 days while wearing the accelerometers. Our results also suggested that kinematic measures of range of motion were not significant predictors for any outcomes we measured except WMFT–function. No demographic variables significantly predicted outcomes. These results are inconclusive, however. The significant predictors we found accounted for limited variation in outcomes. Other potential predictors, such as a composite score indicating synergistic versus nonsynergistic movement or sensation, were not addressed in this study. Further inquiry may include more potential predictors in a larger scale study to validate and improve the current results.

Conclusions

We found that manual dexterity seems valuable in predicting improvements in motor impairment and daily function. In addition, factors related to more normal movement, such as reduced shoulder abduction during forward reach and less distal muscle tone, are related to improvements at motor function and quality-of-life levels. We suggest further studies to include additional potential factors, such as sensation, endurance, depression, and motivation, with a larger sample to study the potential contributors of these factors in mediating treatment success after RT or other rehabilitation interventions. Our findings may also be applied to other treatments sharing similar characteristics, such as high intensity, consistency, and assistance with movement in the face of weakness of RT.

Implications for Occupational Therapy Practice

The findings of this study have the following implications for occupational therapy practice:

Patients with stroke who initially demonstrate good manual dexterity, as measured by the BBT, may have a better chance than those who do not to improve UE motor impairment and daily function; therefore, they should receive intensive therapy as is offered by RT, for example.

Factors related to typical movement, such as reduced shoulder abduction during forward reach and less distal muscle tone, are related to improvements at motor function and quality-of-life levels.

None of the demographic variables used in this study (i.e., age, side of lesion, and time since stroke onset) predicted improved motor function or daily function of UE and therefore should not be used to limit participation in intensive rehabilitation.

Only 29% of the variance associated with improved movement control was accounted for by the chosen predictor variables in this study; therefore, 71% of the variance has not been accounted for and may be found in mental attitude or other variables not included in this study. Patients should not be excluded from intensive therapy only on the basis of dexterity testing. Further study illuminating a greater understanding of characteristics associated with recovery from stroke is needed.

Footnotes

Acknowledgments

This project was supported in part by National Health Research Institutes Grants NHRI-EX101-9920PI and NHRI-EX101-10010PI; National Science Council Grants NSC-100-2314-B-002-008-MY3, NSC 99-2314-B-182-014-MY3, and NSC-102-2314-B-182-001; and Chang Gung Memorial Hospital Research Grants CMRPD1C0401 in Taiwan. Pai-Chuan Huang and Yu-Wei Hsieh contributed equally to this work.

References

Beer

R. F.

Given

J. D.

, & Dewald

J. P. A.

(1999). Task-dependent weakness at the elbow in patients with hemiparesis. Archives of Physical Medicine and Rehabilitation, 80, 766–772. http://dx.doi.org/10.1016/S0003-9993(99)90225-3

Bohannon

R. W.

, & Smith

M. B.

(1987). Interrater reliability of a Modified Ashworth Scale of muscle spasticity. Physical Therapy, 67, 206–207.

Brewer

B. R.

McDowell

S. K.

, & Worthen-Chaudhari

L. C.

(2007). Poststroke upper extremity rehabilitation: A review of robotic systems and clinical results. Topics in Stroke Rehabilitation, 14, 22–44. http://dx.doi.org/10.1310/tsr1406-22

Brunnstrom

(1970). Movement therapy in hemiplegia. New York: Harper & Row.

Carod-Artal

F. J.

Coral

L. F.

Trizotto

D. S.

, & Moreira

C. M.

(2008). The Stroke Impact Scale 3.0: Evaluation of acceptability, reliability, and validity of the Brazilian version. Stroke, 39, 2477–2484. http://dx.doi.org/10.1161/STROKEAHA.107.513671

Carse

Meadows

Bowers

, & Rowe

(2013). Affordable clinical gait analysis: An assessment of the marker tracking accuracy of a new low-cost optical 3D motion analysis system. Physiotherapy, 99, 347–351. http://dx.doi.org/10.1016/j.physio.2013.03.001

Castro-Costa

Fuzikawa

Uchoa

Firmo

J. O.

, & Lima-Costa

M. F.

(2008). Norms for the Mini-Mental State Examination: Adjustment of the cut-off point in population-based studies (evidences from the Bambuí health aging study). Arquivos de Neuro-Psiquiatria, 66, 524–528. http://dx.doi.org/10.1590/S0004-282X2008000400016

Dewald

J. P.

Pope

P. S.

Given

J. D.

Buchanan

T. S.

, & Rymer

W. Z.

(1995). Abnormal muscle coactivation patterns during isometric torque generation at the elbow and shoulder in hemiparetic subjects. Brain, 118, 495–510. http://dx.doi.org/10.1093/brain/118.2.495

Dobkin

B. H.

(2009). Progressive staging of pilot studies to improve Phase III trials for motor interventions. Neurorehabilitation and Neural Repair, 23, 197–206. http://dx.doi.org/10.1177/1545968309331863

10.

Formisano

Pantano

Buzzi

M. G.

Vinicola

Penta

Barbanti

, & Lenzi

G. L.

(2005). Late motor recovery is influenced by muscle tone changes after stroke. Archives of Physical Medicine and Rehabilitation, 86, 308–311. http://dx.doi.org/10.1016/j.apmr.2004.08.001

11.

Franceschini

Ceravolo

M. G.

Agosti

Cavallini

Bonassi

Dall’Armi

, … Sale

(2012). Clinical relevance of action observation in upper-limb stroke rehabilitation: A possible role in recovery of functional dexterity. A randomized clinical trial. Neurorehabilitation and Neural Repair, 26, 456–462. http://dx.doi.org/10.1177/1545968311427406

12.

Fritz

S. L.

Light

K. E.

Clifford

S. N.

Patterson

T. S.

Behrman

A. L.

, & Davis

S. B.

(2006). Descriptive characteristics as potential predictors of outcomes following constraint-induced movement therapy for people after stroke. Physical Therapy, 86, 825–832.

13.

Fritz

S. L.

Light

K. E.

Patterson

T. S.

Behrman

A. L.

, & Davis

S. B.

(2005). Active finger extension predicts outcomes after constraint-induced movement therapy for individuals with hemiparesis after stroke. Stroke, 36, 1172–1177. http://dx.doi.org/10.1161/01.STR.0000165922.96430.d0

14.

Fugl-Meyer

A. R.

Jääskö

Leyman

Olsson

, & Steglind

(1975). The post-stroke hemiplegic patient. 1. A method for evaluation of physical performance. Scandinavian Journal of Rehabilitation Medicine, 7, 13–31.

15.

Gebruers

Vanroy

Truijen

Engelborghs

, & De Deyn

P. P.

(2010). Monitoring of physical activity after stroke: A systematic review of accelerometry-based measures. Archives of Physical Medicine and Rehabilitation, 91, 288–297. http://dx.doi.org/10.1016/j.apmr.2009.10.025

16.

Geyh

Cieza

Schouten

Dickson

Frommelt

Omar

, … Stucki

(2004). ICF Core Sets for stroke. Journal of Rehabilitation Medicine, 44 (Suppl.), 135–141. http://dx.doi.org/10.1080/16501960410016776

17.

Harwin

W. S.

Patton

J. L.

, & Edgerton

V. R.

(2006). Challenges and opportunities for robot-mediated neurorehabilitation. Proceedings of the IEEE, 94, 1717–1726. http://dx.doi.org/10.1109/JPROC.2006.880671

18.

Hesse

Werner

Pohl

Rueckriem

Mehrholz

, & Lingnau

M. L.

(2005). Computerized arm training improves the motor control of the severely affected arm after stroke: A single-blinded randomized trial in two centers. Stroke, 36, 1960–1966. http://dx.doi.org/10.1161/01.STR.0000177865.37334.ce

19.

Hsieh

Y. W.

C. Y.

Liao

W. W.

Lin

K. C.

K. Y.

, & Lee

C. Y.

(2011). Effects of treatment intensity in upper limb robot-assisted therapy for chronic stroke: A pilot randomized controlled trial. Neurorehabilitation and Neural Repair, 25, 503–511. http://dx.doi.org/10.1177/1545968310394871

20.

Hsieh

Y. W.

C. Y.

Lin

K. C.

Chang

Y. F.

Chen

C. L.

, & Liu

J. S.

(2009). Responsiveness and validity of three outcome measures of motor function after stroke rehabilitation. Stroke, 40, 1386–1391. http://dx.doi.org/10.1161/STROKEAHA.108.530584

21.

Hsieh

Y. W.

C. Y.

Lin

K. C.

Yao

K. Y.

, & Chang

Y. J.

(2012). Dose-response relationship of robot-assisted stroke motor rehabilitation: The impact of initial motor status. Stroke, 43, 2729–2734. http://dx.doi.org/10.1161/STROKEAHA.112.658807

22.

Kwakkel

, & Kollen

B. J.

(2007). Predicting improvement in the upper paretic limb after stroke: A longitudinal prospective study. Restorative Neurology and Neuroscience, 25, 453–460.

23.

Kwakkel

Kollen

B. J.

, & Krebs

H. I.

(2008). Effects of robot-assisted therapy on upper limb recovery after stroke: A systematic review. Neurorehabilitation and Neural Repair, 22, 111–121. http://dx.doi.org/10.1177/1545968307305457

24.

Kwakkel

Kollen

B. J.

van der Grond

J. V.

, & Prevo

A. J. H.

(2003). Probability of regaining dexterity in the flaccid upper limb: Impact of severity of paresis and time since onset in acute stroke. Stroke, 34, 2181–2186. http://dx.doi.org/10.1161/01.STR.0000087172.16305.CD

25.

Lai

S. M.

Studenski

Duncan

P. W.

, & Perera

(2002). Persisting consequences of stroke measured by the Stroke Impact Scale. Stroke, 33, 1840–1844. http://dx.doi.org/10.1161/01.STR.0000019289.15440.F2

26.

Lang

C. E.

Edwards

D. F.

Birkenmeier

R. L.

, & Dromerick

A. W.

(2008). Estimating minimal clinically important differences of upper-extremity measures early after stroke. Archives of Physical Medicine and Rehabilitation, 89, 1693–1700. http://dx.doi.org/10.1016/j.apmr.2008.02.022

27.

Lemmens

R. J.

Timmermans

A. A.

Janssen-Potten

Y. J.

Smeets

R. J.

, & Seelen

H. A.

(2012). Valid and reliable instruments for arm-hand assessment at ICF activity level in persons with hemiplegia: A systematic review. BMC Neurology, 12, 2. http://dx.doi.org/10.1186/1471-2377-12-21

28.

Lin

K. C.

Chuang

L. L.

C. Y.

Hsieh

Y. W.

, & Chang

W. Y.

(2010). Responsiveness and validity of three dexterous function measures in stroke rehabilitation. Journal of Rehabilitation Research and Development, 47, 563–571. http://dx.doi.org/10.1682/JRRD.2009.09.0155

29.

Lin

K. C.

Huang

Y. H.

Hsieh

Y. W.

, & Wu

C. Y.

(2009). Potential predictors of motor and functional outcomes after distributed constraint-induced therapy for patients with stroke. Neurorehabilitation and Neural Repair, 23, 336–342. http://dx.doi.org/10.1177/1545968308321773

30.

A. C.

Guarino

P. D.

Richards

L. G.

Haselkorn

J. K.

Wittenberg

G. F.

Federman

D. G.

, … Peduzzi

(2010). Robot-assisted therapy for long-term upper-limb impairment after stroke. New England Journal of Medicine, 362, 1772–1783. http://dx.doi.org/10.1056/NEJMoa0911341

31.

Mathiowetz

Volland

Kashman

, & Weber

(1985). Adult norms for the Box and Block Test of manual dexterity. American Journal of Occupational Therapy, 39, 386–391. http://dx.doi.org/10.5014/ajot.39.6.386

32.

Mehrholz

Platz

Kugler

, & Pohl

(2008). Electromechanical and robot-assisted arm training for improving arm function and activities of daily living after stroke. Cochrane Database of Systematic Reviews, Issue 4, CD006876. http://dx.doi.org/10.1002/14651858.CD006876.pub2

33.

Messier

Bourbonnais

Desrosiers

, & Roy

(2006). Kinematic analysis of upper limbs and trunk movement during bilateral movement after stroke. Archives of Physical Medicine and Rehabilitation, 87, 1463–1470. http://dx.doi.org/10.1016/j.apmr.2006.07.273

34.

Morris

D. M.

Uswatte

Crago

J. E.

Cook

E. W.

III , & Taub

(2001). The reliability of the Wolf Motor Function Test for assessing upper extremity function after stroke. Archives of Physical Medicine and Rehabilitation, 82, 750–755. http://dx.doi.org/10.1053/apmr.2001.23183

35.

Park

S. W.

Wolf

S. L.

Blanton

Winstein

, & Nichols-Larsen

D. S.

(2008). The EXCITE Trial: Predicting a clinically meaningful Motor Activity Log outcome. Neurorehabilitation and Neural Repair, 22, 486–493. http://dx.doi.org/10.1177/1545968308316906

36.

Platz

Pinkowski

van Wijck

Kim

I. H.

di Bella

, & Johnson

(2005). Reliability and validity of arm function assessment with standardized guidelines for the Fugl-Meyer Test, Action Research Arm Test, and Box and Block Test: A multicentre study. Clinical Rehabilitation, 19, 404–411. http://dx.doi.org/10.1191/0269215505cr832oa

37.

Rand

, & Eng

J. J.

(2010). Arm–hand use in healthy older adults. American Journal of Occupational Therapy, 64, 877–885. http://dx.doi.org/10.5014/ajot.2010.09043

38.

Roby-Brami

Feydy

Combeaud

Biryukova

E. V.

Bussel

, & Levin

M. F.

(2003). Motor compensation and recovery for reaching in stroke patients. Acta Neurologica Scandinavica, 107, 369–381. http://dx.doi.org/10.1034/j.1600-0404.2003.00021.x

39.

Sloan

R. L.

Sinclair

Thompson

Taylor

, & Pentland

(1992). Inter-rater reliability of the Modified Ashworth Scale for spasticity in hemiplegic patients. International Journal of Rehabilitation Research, 15, 158–161. http://dx.doi.org/10.1097/00004356-199206000-00009

40.

Uswatte

Giuliani

Winstein

Zeringue

Hobbs

, & Wolf

S. L.

(2006). Validity of accelerometry for monitoring real-world arm activity in patients with subacute stroke: Evidence from the Extremity Constraint-Induced Therapy Evaluation trial. Archives of Physical Medicine and Rehabilitation, 87, 1340–1345. http://dx.doi.org/10.1016/j.apmr.2006.06.006

41.

van der Lee

J. H.

Beckerman

Knol

D. L.

de Vet

H. C. W.

, & Bouter

L. M.

(2004). Clinimetric properties of the Motor Activity Log for the assessment of arm use in hemiparetic patients. Stroke, 35, 1410–1414. http://dx.doi.org/10.1161/01.STR.0000126900.24964.7e

42.

van Heuvelen

M. J.

Kempen

G. I.

Brouwer

W. H.

, & de Greef

M. H.

(2000). Physical fitness related to disability in older persons. Gerontology, 46, 333–341. http://dx.doi.org/10.1159/000022187

43.

Volpe

B. T.

Krebs

H. I.

Hogan

Edelsteinn

Diels

C. M.

, & Aisen

M. L.

(1999). Robot training enhanced motor outcome in patients with stroke maintained over 3 years. Neurology, 53, 1874–1876. http://dx.doi.org/10.1212/WNL.53.8.1874

44.

Waldner

Tomelleri

, & Hesse

(2009). Transfer of scientific concepts to clinical practice: Recent robot-assisted training studies. Functional Neurology, 24, 173–177.

45.

Waller

S. M.

, & Whitall

(2005). Hand dominance and side of stroke affect rehabilitation in chronic stroke. Clinical Rehabilitation, 19, 544–551. http://dx.doi.org/10.1191/0269215505cr829oa

46.

Wee

J. Y. M.

, & Hopman

W. M.

(2005). Stroke impairment predictors of discharge function, length of stay, and discharge destination in stroke rehabilitation. American Journal of Physical Medicine and Rehabilitation, 84, 604–612. http://dx.doi.org/10.1097/01.phm.0000171005.08744.ab

47.

Whitall

Waller

S. M.

Sorkin

J. D.

Forrester

L. W.

Macko

R. F.

Hanley

D. F.

, … Luft

(2011). Bilateral and unilateral arm training improve motor function through differing neuroplastic mechanisms: A single-blinded randomized controlled trial. Neurorehabilitation and Neural Repair, 25, 118–129. http://dx.doi.org/10.1177/1545968310380685

48.

Wolf

S. L.

Catlin

P. A.

Ellis

Archer

A. L.

Morgan

, & Piacentino

(2001). Assessing Wolf Motor Function Test as outcome measure for research in patients after stroke. Stroke, 32, 1635–1639.

49.

World Health Organization. (2001). International classification of functioning, disability and health. Geneva: Author.

50.

C. Y.

Yang

C. L.

Chen

M. D.

Lin

K. C.

, & Wu

L. L.

(2013). Unilateral versus bilateral robot-assisted rehabilitation on arm-trunk control and functions post stroke: A randomized controlled trial. Journal of NeuroEngineering and Rehabilitation, 10, 35. http://dx.doi.org/10.1186/1743-0003-10-35

51.

Yancosek

K. E.

, & Howell

(2009). A narrative review of dexterity assessments. Journal of Hand Therapy, 22, 258–270. http://dx.doi.org/10.1016/j.jht.2008.11.004

52.

Yang

C. L.

Lin

K. C.

Chen

H. C.

C. Y.

, & Chen

C. L.

(2012). Pilot comparative study of unilateral and bilateral robot-assisted training on upper-extremity performance in patients with stroke. American Journal of Occupational Therapy, 66, 198–206. http://dx.doi.org/10.5014/ajot.2012.003103