Effectiveness of Virtual Reality Using Wii Gaming

Technology in Stroke Rehabilitation

A Pilot Randomized Clinical Trial and Proof of Principle
Gustavo Saposnik, MD, MSc, FAHA; Robert Teasell, MD, FRCPC;
Muhammad Mamdani, PharmD, MPH, MA; Judith Hall, MSc; William McIlroy, PhD;
Donna Cheung, OT; Kevin E. Thorpe, MMath; Leonardo G. Cohen, MD; Mark Bayley, MD, FRCPC;
for the Stroke Outcome Research Canada (SORCan) Working Group

Background and Purpose—Hemiparesis resulting in functional limitation of an upper extremity is common among stroke
survivors. Although existing evidence suggests that increasing intensity of stroke rehabilitation therapy results in better
motor recovery, limited evidence is available on the efficacy of virtual reality for stroke rehabilitation.
Methods—In this pilot, randomized, single-blinded clinical trial with 2 parallel groups involving stroke patients within 2
months, we compared the feasibility, safety, and efficacy of virtual reality using the Nintendo Wii gaming system
(VRWii) versus recreational therapy (playing cards, bingo, or “Jenga”) among those receiving standard rehabilitation to
evaluate arm motor improvement. The primary feasibility outcome was the total time receiving the intervention. The
primary safety outcome was the proportion of patients experiencing intervention-related adverse events during the study
period. Efficacy, a secondary outcome measure, was evaluated with the Wolf Motor Function Test, Box and Block Test,
and Stroke Impact Scale at 4 weeks after intervention.
Results—Overall, 22 of 110 (20%) of screened patients were randomized. The mean age (range) was 61.3 (41 to 83) years.
Two participants dropped out after a training session. The interventions were successfully delivered in 9 of 10
participants in the VRWii and 8 of 10 in the recreational therapy arm. The mean total session time was 388 minutes in
the recreational therapy group compared with 364 minutes in the VRWii group (P⫽0.75). There were no serious adverse
events in any group. Relative to the recreational therapy group, participants in the VRWii arm had a significant
Downloaded from http://ahajournals.org by on April 1, 2024

improvement in mean motor function of 7 seconds (Wolf Motor Function Test, 7.4 seconds; 95% CI, ⫺14.5, ⫺0.2) after
adjustment for age, baseline functional status (Wolf Motor Function Test), and stroke severity.
Conclusions—VRWii gaming technology represents a safe, feasible, and potentially effective alternative to facilitate
rehabilitation therapy and promote motor recovery after stroke. (Stroke. 2010;41:1477-1484.)
Key Words: stroke 䡲 virtual reality 䡲 rehabilitation 䡲 Wolf Motor Function Test 䡲 outcome research 䡲 safety
䡲 feasibility 䡲 randomized clinical trial 䡲 Wii gaming system

S troke is a devastating disease for patients and their

families and a leading cause of adult disability. Up to
85% of stroke patients experience hemiparesis immediately
evidence suggests that repetitive, task-oriented training of the
paretic extremity is beneficial.3,4 However, implementation of
these techniques is tedious, resource-intensive, and costly,
after stroke, and between 55% and 75% of survivors continue often requiring transportation of patients to specialized facil-
to experience motor deficits associated with diminished ities.5,6 Moreover, there is a gap in knowledge of how these
quality of life.1 Traditional rehabilitative therapies can help interventions influence patients within the first 3 months after
regain motor function and ameliorate disability.2 Recent stroke (subacute period). Recovery of function, both sponta-

Received March 19, 2010; accepted April 15, 2010.

From the Division of Neurology, Departments of Medicine and Health Policy Management and Evaluation and Institute for Clinical Evaluative
Sciences (G.S.), Applied Health Research Centre (M.M., J.H.), St. Michael’s Hospital and Health Policy Management and Evaluation and Institute for
Clinical Evaluative Sciences, South East Toronto Stroke Network (D.C.), St. Michael’s Hospital, and Neuro-Rehabilitation Program (M.B.), Toronto
Rehabilitation Institute, and Keenan Research Centre (K.E.T.), Li Ka Shing Knowledge Institute of St. Michael’s Hospital, Dalla Lana, School of Public
Health, University of Toronto, Ontario, Canada; Stroke Rehabilitation Program (R.T.), Parkwood Hospital, St. Joseph’s Health Care London, University
of Western Ontario, London, Canada; Applied Health Sciences (W.M.), University of Waterloo, Waterloo, Canada; and Human Cortical Physiology and
Stroke Neurorehabilitation Section (L.G.C.), National Institute of Neurological Disorders, National Institutes of Health, Bethesda, Md.
The EVREST Outcome and Statistical Analysis Team is composed of G.S., M.M., L.G.C., and K.E.T. The statistical analysis was performed by K.E.T.
L.G.C. and M.B. contributed equally to this manuscript and project to qualify as the senior investigators.
Correspondence to Dr Gustavo Saposnik, Assistant Professor, Director, Stroke Research Unit, Stroke Outcome Research Canada (SORCan;
www.sorcan.ca) Working Group, Department of Medicine, St Michael’s Hospital, University of Toronto, 55 Queen St E, Toronto, Ontario, M5C 1R6.
E-mail [email protected]
© 2010 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.110.584979

1477
 1478 Stroke July 2010

neous and secondary to intense rehabilitative treatments, is summary of inclusion and exclusion criteria, see Supplemental Table
sustained by plasticity and rewiring in the injured brain in I, available online at http://stroke.ahajournals.org.)
adults.2,7,8 Neurons in the adult human brain increase their
firing rates when a subject observes movements performed by Baseline Measures
Baseline characteristics were collected, including demographics (age
other persons. Activation of this mirror–neuron system,
and gender), handedness, comorbid conditions, stroke characteristics
including areas of the frontal, parietal, and temporal lobes, including location, type, and baseline disability based on the modi-
can induce cortical reorganization and possibly contribute to fied Rankin scale, and Barthel index for activities of daily living.
functional recovery because critical nodes in the system are Stroke severity was assessed using the Canadian Neurological Scale,
also active when subjects actually perform movements.9 –11 a simple, reliable, and validated scale (in which lower scores indicate
greater stroke severity) for estimating the neurological status, espe-
Virtual reality (VR) gaming systems are novel and poten-
cially when the National Institutes of Health Stroke Scale is not
tially useful technologies that allow users to interact in 3 available.16,17 Baseline motor function was assessed using the Wolf
dimensions with a computer-generated scenario (a virtual Motor Function Test18 and the Box and Block Test.19 Baseline
world), engaging the mirror–neuron system. The gaming quality of life was assessed using the Stroke Impact Scale.20 The
industry has developed a variety of VR systems for home use, blinded assessor was trained in the use of these scales.
All baseline, post-treatment, and 4-week follow-up assessments
making this technology both affordable and accessible with
were performed by a trained outcome assessor, blinded to patient
potential application in community settings (ie, patients’ randomization, who was not involved in administration of study
homes). In particular, these technologies allow for interactive interventions.
observation of avatar movements captured on the screen and
combine features of increasing rehabilitation intensity re- Study Interventions
quired for induction of neuroplasticity.9,11,12
Description of Wii Gaming Technology
However, there has been limited research involving the Nintendo introduced a new style of VR (2006) by using a wireless
incorporation of VR gaming systems into neurorehabilitation controller that interacts with the player through a motion detection
programs, particularly in the subacute period after stroke. system and avatar (computer user’s representation of himself or
There is an identified need for rigorous randomized con- herself or alter ego) technology. The controllers use embedded
trolled trials to establish the safety, feasibility, and efficacy of acceleration sensors responsive to changes in direction, speed, and
acceleration that enable participants to interact with the games while
VR systems as therapeutic options in stroke rehabilita- performing wrist, arm, and hand movements. A 2-point infrared light
tion.13,14 The objectives of this study were to examine the sensor, mounted on top of a television, captures and reproduces on
feasibility and safety of the VR Nintendo Wii gaming system the screen the movement from the controller as performed by
(VRWii) compared with recreational therapy (RT) in facili- participants. Because Wii is computer assisted, big sweeping move-
Downloaded from http://ahajournals.org by on April 1, 2024

ments in the games are not necessary. The feedback provided by the
tating motor function of the upper extremity required for
TV screen as well as the opportunity to observe their own move-
activities of daily living among patients with subacute stroke ments in real time, generates positive reinforcement, thus facilitating
receiving standard rehabilitation. training and task improvement. (Additional details are described
online at http://www.nintendo.com/wii/what.)
Methods As described, several distinctive features favored the selection
VRWii over other VR systems, including novel and widely available
Study Design 3D technology using gaming simulations, affordability, clinical
The Effectiveness of Virtual Reality Exercises in Stroke Rehabilita- applicability using simple graphics with real-time feedback with the
tion (EVREST) is a pilot, randomized, single-blind, parallel group possibility to reduce speed, making it usable for patients with
trial to systematically compare the feasibility and safety of VRWii to cognitive impairments after stroke, and provision of direct multimo-
RT in patients with a first stroke within 6 months before enrollment dal sensory feedback (vision, touch, and auditory) with the avatar,
to determine whether VRWii enhances motor recovery after stroke. thus allowing adjustments while performing and self-observing the
execution of diverse tasks.
Participants The software used in EVREST was the publicly available sports
Participants 18 to 85 years of age having a first-time ischemic or (ie, Wii Sports) and Cooking Mamma packages, accounting for 30
hemorrhagic stroke were eligible for the study. Although the proto- minutes each in the VRWii group.
col allowed the inclusion of patients up to 6 months after stroke,
early recruitment was favored. This time window was chosen to Description of RT
maximize the opportunity for enhancing motor recovery. All partic- RT sessions included leisure activities such as playing cards,
ipants had a clinically defined acute stroke confirmed by neuroim- stamping a seal while playing bingo, or playing Jenga. Adherence to
aging (CT or MRI) and neurological assessment and met a level of standard rehabilitation and to the study tasks were monitored with a
function of the upper extremity derived from the Chedoke–McMas- timer. RT was used as a control group to allow a fair comparison
ter scale15 ⬎3 either in the arm or hand (ie, shrug their shoulders, between the time spent in rehabilitation activities between groups
touch chin with the affected arm) at time of enrollment. and a lack of evidence that Wii gaming system is standard rehabil-
Potential participants were excluded if they were unable to follow itation therapy. Additional details of the protocol have been pub-
instructions, had a prestroke modified Rankin score of ⱖ2, were lished previously.21
medically unstable or had uncontrolled hypertension according to the
treating physician, were experiencing a severe illness with a life Study Procedures
expectancy ⬍3 months, experienced unstable angina, or had recent
myocardial infarction (within 3 months), had a history of seizures or Randomization
epilepsy (except for febrile seizures of childhood), were participating Participants admitted to the Toronto Rehabilitation Institute were
in another clinical trial involving an investigational drug or physical randomly allocated in a 1:1 ratio to the 2 study groups. The
therapy, or had any condition that might put the patient at risk (ie, randomization schedule was computer generated using a basic
known shoulder subluxation or fracture) at study entry. (For a random number generator.
 Saposnik et al Virtual Reality Using Wii Technology in Stroke Rehab 1479

Allocation (no exertion at all) to 20 (maximal exertion). Excessive fatigue is

All participants received standard rehabilitation therapy for stroke, considered any score ⬎13. For a healthy person, this represents
which accounts for an average of 1 hour of physiotherapy and slowly walking at his or her own pace and finding an exercise
another hour of occupational therapy per day on tolerance. somewhat difficult but not severe enough to discontinue the task.
Patients were instructed to rate their perception of exertion reflecting
Sessions how heavy and strenuous the exercise felt to them, combining all
Patients received an intensive program consisting of 8 interventional sensations and feelings of physical stress, effort, and fatigue.
sessions (VRWii or RT) of 60 minutes each over a 14-day period.
These 8 sessions were scheduled in a flexible manner as long as all Secondary Outcome Measures
8 sessions were completed within the 2-week period with sessions Because EVREST was designed as a feasibility study, efficacy was
separated by ⱖ5 hours. The arm movements involved in the use of a secondary outcome assessed at 4 weeks after intervention. Efficacy
the Wii included shoulder flexion and extension (bowling and was measured as an improvement in motor function determined by
tennis), shoulder rotation (tennis), elbow extension and flexion the total time elapsed to complete a shorter version of the Wolf
(Cooking Mama), wrist supination and pronation (tennis and Cook- Motor Function Test (WMFT).18 This test contains 15 timed and 2
ing Mama), and different degrees of wrist flexion and extension as strength tasks (lifting the weighted limb and grip strength), ordered
well as thumb flexion involved in all activities. The recreational from simple to complex, administered sequentially to each upper
activities engaged similar movements. Patients were instructed to extremity and controlling for patient positioning and distance the
remain in a sitting position and primarily use their more affected extremity segment must traverse. Tasks specifically measured in the
arm/hand in these activities. Participants randomized to one arm study included forearm to box (side), hand to box (front), lift a can,
were not exposed to the other intervention. lift a paper clip, flip cards, and fold a towel. The 2 strength tasks
Video games may be associated with a risk of photosensitive- included forward flexion of the shoulder in a seated position to the
induced seizures (⬇1:4000) and repetitive motion injuries.22,23 To top of a box placed on the table using weights ⱕ20 pounds strapped
reduce the likelihood of seizures, the lights were kept on during the to the forearm, as well as dynamometer grip strength for 3 seconds
VRWii gaming sessions, and patients were sitting ⱖ6 feet away from with the elbow bent to 90°.18 Other efficacy end points included a
the television screen. The study coordinator remained in the room 4-block improvement on the Box and Block Test,19,25 a performance-
during the sessions and monitored the patient for symptoms sugges- based measure assessment of gross manual dexterity. The respondent
tive of seizures or shoulder, arm, or hand pain. If the patient felt was instructed to move as many blocks as possible, one at a time,
unwell at any time, the coordinator was instructed to stop the session. from one compartment to the other for a period of 60 seconds.
Quality of life was measured by the Stroke Impact Scale.20 This is a
Blinding of Caregivers and Outcomes simple, validated, stroke-specific, health status measure. Version 2.0
Measures Assessment was composted of 64 items in 8 domains (strength, hand function,
The study coordinator and patients participating in this study were activities of daily living/instrumental activities of daily living,
not blinded to the intervention group. To limit knowledge of the Wii mobility, communication, emotion, memory and thinking, and par-
gaming technology, and to ensure other caregivers and support staff ticipation). Hand function and a composite score of the physical
domains (strength, hand function, mobility, and activities of daily
Downloaded from http://ahajournals.org by on April 1, 2024

were not aware of subject allocation, all study interventions were

conducted by dedicated trial staff out of sight of ward staff. Trial living/instrumental activities of daily living) were considered a priori
staff and subjects were instructed not to divulge the intervention as the more clinically relevant Stroke Impact Scale outcomes.
allocation to caregivers or other ward staff. Interventions were not
recorded in the medical record. All postintervention and 4-week Study Organization and Data Management
follow-up assessments were performed by a trained outcome asses- The study was completed at Toronto Rehabilitation Institute, which
sor who was not involved in administration of study interventions receives referrals for inpatient stroke rehabilitation from 4 acute care
and was blinded to patient randomization. facilities in Toronto.
Administrative activities, data management, research coordina-
Contamination tion, and statistical analyses were conducted at the Applied Health
At the time of final follow-up measurements, the blinded assessor Research Centre at St. Michael’s Hospital, University of Toronto.
was asked to select (forced choice) to which group they thought the Operational procedures, guidelines for the implementation of both
patient had been allocated. This process allowed us to assess the arms of the study, and informed consent forms were approved by the
effectiveness of blinding the assessor to group allocation. ethics review boards at St. Michael’s Hospital and the Toronto
Rehabilitation Institute. Written informed consent was obtained at
Follow-Up Visit the Toronto Rehabilitation Institute before enrollment.
Follow-up data were collected 4 weeks (⫾3 days) after the final
study intervention session.
Analysis
Given the pilot nature of this study, primarily descriptive statistics
Outcome Measures were calculated. The average total therapy time, the average therapy
Primary Outcome Measures time per session, and differences 4 weeks after intervention from
The coprimary end points of the present study relate to feasibility baseline in clinical outcomes were computed for each group, along
and safety. Time tolerance and adaptation to playing Wii have not with 95% CIs. The Welch t test was used to rule out large differences
been formally tested in stroke patients. Therefore, the primary in therapy times. For relevant clinical outcomes, descriptive statistics
feasibility outcome was defined as the total time receiving the (mean, SD, median, and quartiles) were computed for each assess-
VRWii intervention. ment. Adjusted treatment analysis was planned a priori assuming
The primary safety outcome was defined as the proportion of differences in baseline characteristics as expected in a small sample
patients experiencing intervention-related adverse events or any study. Adjustment treatment effects were obtained from the param-
serious adverse event during the study period. A serious adverse eter estimate pertaining to treatment group from multiple linear
event was defined as any untoward medical occurrence, whether or regression, for which age, Canadian Neurological Scale, and the
not considered to be causally related to the study intervention, that baseline score for the particular outcome being analyzed were
resulted in death, life-threatening illness or injury, required inpatient adjusted. An examination of the residual plots did not suggest
hospitalization or prolonged inpatient hospitalization, or resulted in concerns regarding the normality of the errors. A 2-sided P value of
persistent disability or incapacity. The Borg perceived exertion scale ⬍0.05 was considered statistically significant. All statistical analysis
was used to determine fatigability or level of effort required to was performed in the R language for statistical computing (R
perform tasks at the end of the sessions.24 The scale ranges from 6 Foundation for Statistical Computing; Vienna, Austria).
 1480 Stroke July 2010

110 Individuals Screened

78 Not Eligible
- Age criteria: 10
- Too mild or severe deficit: 30
- Medical reasons: 38

10 Refused Consent

22 Randomized

11 Assigned to Receive Wii 11 Assigned to Receive

Virtual Reality Therapy Recreational Therapy

1 Patient who discontinued 1 Patient who discontinued

after training session after training session
Figure. Flow of patients through the trial.

1 Patient who discontinued 2 Patient who discontinued

prior to completion of 8 prior to completion of 8
sessions sessions

9 Patients Who Completed All 8 Patients Who Completed All

8 Sessions 8 Sessions

1 Patient Who
Discontinued Prior to 4-
Week Follow-Up

9 Completed 4-Week Follow- 7 Completed 4-Week Follow-

Up Up
Downloaded from http://ahajournals.org by on April 1, 2024

Results period (mean time of physiotherapy and occupational therapy

From November 2008 to October 2009, 110 potential candi- 100 minutes on weekdays and 81 minutes on weekends).
dates were screened to participate in EVREST (Figure). A In assessing the success of blinding, we found no evidence
total of 88 patients were excluded. Ten patients refused of failure of concealment at the end of the study. The blinded
consent, 10 were ⬎85 years of age, 30 had a very mild or assessor was correct 9 times out of 16 decisions, or 56% of
severe deficit, and 38 had an underlying medical condition the time, which is similar to the 50% expected by chance.
(angina, recent myocardial infarction, language barrier, or
global aphasia) that affected their ability to participate in the Primary End Points: Safety and Feasibility
study. A total of 22 (20%) participants were enrolled in None of the participants in either group experienced a serious
EVREST. Although 11 patients were randomized to each
adverse event during the study period. Table 2 summarizes
group, one patient in each arm dropped out after the first
the primary outcomes. There was no difference in the
training session, leaving 20 patients randomized (10 to each
proportion of patients reporting any symptom during the
intervention arm). Eight participants randomized to RT and 9
interventions (difference VRWii-RT 0.2 (95% CI, ⫺0.229,
randomized to VRWii completed all sessions (Figure).
0.629]). Two participants in the RT and 3 participants in the
The mean age was 61 years (range 41 to 83 years), and the
mean time from stroke onset to enrollment was 25 days VRWii reported exertion fatigue (defined as a Borg ⬎13
(range 10 to 56 days). Baseline characteristics are presented during therapy). The difference in the proportion of patients
in Table 1. Relative to patients randomized to the RT groups, ever reporting fatigue was not significant.
patients randomized to VRWii were younger (mean age 55 Ten participants in each group started the intervention after
years versus 67 years), had more severe strokes (median the training session. The interventions were successfully
Canadian Neurological Scale 9.0 versus 10.0), and had lower delivered in 9 of 9 participants in the VRWii and 8 of 10 in
arm function (median Chedoke–McMaster scale 4.0 versus the RT arm (Figure).
4.5) at baseline. There were no significant imbalances be- The primary end point for time of delivered interventions
tween groups with respect to gender, comorbidities, affected showed no difference between groups. The mean time for the
side, handedness, stroke type. delivered intervention was 388⫾171 minutes in the RT and
Both groups received similar durations of standard reha- 364⫾148 minutes in the VRWii group (P⫽0.75; Table 2).
bilitation (physiotherapy and occupational therapy) during The total delivery time comprised 80% (RT) and 76%
the study period: ⬇20 hours during the 2-week intervention (VRWii) of the scheduled time.
 Saposnik et al Virtual Reality Using Wii Technology in Stroke Rehab 1481

Table 1. Study Participant Characteristics within the VRWii and RT groups. Participants in the VRWii
Characteristics RT (n⫽11) VRWii (n⫽11)
group had a significant improvement in WMFT (represented
as a shorter time in completing the tasks) and grip strength
Mean age (range), y 67.3 (46 – 83) 55.3 (41–72)
from baseline, whereas both groups had an improvement in
Sex, male 7 (64) 7 (64) Box and Block Test (Table 3).
Handedness, right 10 (91) 10 (91) Multivariable analyses using linear regression 4 weeks
Comorbidities after intervention were performed using baseline measures as
Hypertension 5 (45) 9 (82) covariates to evaluate the efficacy of VRWii versus RT. After
Dyslipidemia 2 (18) 1 (9) adjustment for age, baseline functional status (WMFT), and
Diabetes mellitus 4 (36) 2 (18) stroke severity, participants in the VRWii arm performed, on
average, significantly better on the WMFT than the RT group
Atrial fibrillation 0 (0) 1 (9)
(⫺7.4 seconds; 95% CI, ⫺14.5, ⫺0.2). Similarly, participants
Stroke type, hemorrhagic 2 (18) 2 (18)
randomized to VRWii achieved a nonsignificant improve-
Affected side, right 6 (55) 6 (55)
ment in grip strengths when compared with RT after adjust-
Baseline functional status at ing for age, baseline grip strength, and stroke severity (1.9 kg;
time of randomization
95% CI, ⫺2.5, 6.2). No significant differences were observed
Chedocke–McMaster, 4.5 关4.0–5.0兴 4.0 关3.5–4.0兴 in the adjusted analysis for the Stroke Impact Scale (hand or
median 关IQR兴
composite) and Box and Block Test.
Stroke severity, median 10.0 关8.1–10.4兴 9.0 关7.8–9.5兴
CNS 关IQR兴
Discussion
Barthel index, median 关IQR兴 65 关60–88兴 65 关55–72.5兴 The field of poststroke rehabilitation is evolving. The current
Modified Rankin scale paradigm of stroke rehabilitation strategies to improve motor
1 1 (9) 0 (0) function is focused on high-intensity, repetitive, and task-
2 3 (27) 2 (18) specific practice.2,26 Long-term potentiation, implicated in the
3 5 (45) 5 (45) acquisition of new and retrieval of learned motor patterns,
4 2 (18) 4 (36) develops from repeated stimulation or tasks. VR gaming is a
technology that allows a user to interact with a computer-
Mean 关SD兴 days from onset to 22.7 关8.6兴 26.7 关16.4兴
randomization simulated environment and receive near real-time feedback
on performance. As reported in recent studies,2,14 the extent to
Downloaded from http://ahajournals.org by on April 1, 2024

Numbers between brackets represent percentages unless otherwise specified.

IQR indicates interquartile range; CNS, Canadian Neurological Scale.
which VR systems can facilitate conventional therapy cur-
rently in use remains to be determined.
Secondary End-Points: Clinically The EVREST trial represents the first randomized con-
Relevant Outcomes trolled study to systematically test an interactive 3D VR
Table 3 summarizes the unadjusted changes in clinical technology using the Wii gaming system as neurorehabilita-
outcomes at baseline through 4 weeks after the interventions tion therapy among patients who experienced a first stroke

Table 2. Effect of VRWii on Primary End Points Between Groups

Outcomes RT (n⫽10) VRWii (n⫽10) Difference VRWii-RT (95% CI)
Primary end points
Primary feasibility outcomes
No. (%) of patients completing all 8 8 (80) 9 (90) 0.1 (⫺0.21, 0.41)*
sessions
Mean (SD) total session time, in minutes 388 (171) 364 (148) ⫺23.3 (⫺173.7, 127.1)†
Mean (SD) session time, in minutes 56.2 (6.8) 46.5 (16.0) ⫺9.7 (⫺21.7, 2.2)‡
Primary safety outcomes
Serious adverse events 0 (0) 0 (0) NA
Dizziness or nausea 1 (10) 0 (0) NA
Mean Borg exertion scale at the end of the 9.38 (4.7) 9.0 (2.8) ⫺0.4 (⫺4.1, 3.4)
interventions (SD)
Borg exertion scale (fatigability) ⬎13 in any 2 (20) 3 (33) 0.13 (⫺0.26, 0.53)§
session
No. of patients reporting any symptom 4 (40) 6 (60) 0.2 (⫺0.23, 0.63)¶
during any session
NA indicates not applicable.
*Difference in the proportion completing all 8 sessions between groups; †difference (95% CI) in the total session time of the
delivered intervention between groups; ‡difference (95% CI) in the average total session length between groups; §difference (95%
CI) in the proportion of patients ever reporting fatigue (Borg exertion scale ⬎13) during the intervention; ¶difference (95% CI) in the
proportion of patients reporting any symptom during the intervention.
 1482 Stroke July 2010

Table 3. Effect of Interventions on Secondary Outcomes at Baseline Through 4-Week Follow-Up and Change From Baseline Within
Wii and RT Groups
Virtual Reality using Wii (VRWii) Recreational Therapy (RT)

Secondary Four Weeks Four Weeks

Efficacy Baseline Postintervention After Intervention Change From Baseline Postintervention After Intervention Change From
Outcomes (n⫽9) (n⫽9) (n⫽9) Baseline (95% CI)* (n⫽10) (n⫽9) (n⫽7) Baseline (95% CI)*
WMFT, mean 29.5 19.8 19.0 ⫺10.5 (⫺19.3, ⫺1.8) 39.7 27.4 25.7 ⫺14 (⫺32.1, 4.1)
time in seconds
BBT, mean No. 33.4 40.1 42.0 8.6 (0.9, 16.2) 24.4 34.1 36.4 12 (3.8, 20.2)
of blocks
SIS hand 70.6 71.1 85.8 15.3 (⫺1.2, 31.8) 63.4 70.0 74.6 11.2 (⫺21.3, 43.8)
function, %
SIS composite 74.4 74.2 82.7 8.3 (⫺3.3, 19.9) 67.0 74.3 80.3 13.3 (0.9, 25.8)
function, %
SIS perception of 62.1 76.7 76.1 14 (⫺2.7, 12.6) 61.4 66.4 77.1 15.7 (⫺9.1, 40.5)
recovery, %
Grip strength, 22.4 24.6 28.8 6.4 (3.8, 9.0) 18.7 21.5 21.5 2.8 (⫺1.6, 7.1)
mean in kg
SIS indicates Stroke Impact Scale; BBT, Box and Block Test.
SIS composite includes the following physical domains: strength, hand function, mobility, and activities of daily living.
*Difference and 95% CI in secondary efficacy outcomes between baseline and 4-week postintervention measures in each group; †one patient was unable to
complete some baseline tasks for the WMFT.

within 2 months before enrollment. The effect of the VRWii Most of the reported studies focused on lower extremities,
gaming technology was compared with RT among patients especially in gait training, including the use of a treadmill.13
receiving usual/standard rehabilitation. We found that VRWii Only 3 studies addressed upper limb rehabilitation, and none
use was feasible and safe when performed under the prespeci- of them were randomized trials. None of the studies reported
fied criteria. For the efficacy outcomes (secondary clinical any significant adverse effects. Moreover, several studies
end points), this feasibility study showed a significant im- compared an intervention plus conventional physical therapy
Downloaded from http://ahajournals.org by on April 1, 2024

provement in motor function (WMFT) in the unadjusted versus conventional physical therapy alone, which, by neces-
(Table 3) and adjusted analysis for the VRWii group. When sity allowed for more rehabilitation time in the experimental
reporting outcomes, it is worth knowing whether an observed group.33 This creates a bias in favor of the new intervention
difference indicates a clinically significant effect.27 The because the intensity and frequency of rehabilitation per se is
observed 7-second benefit for the VRWii group in the known to directly and beneficially affect functional out-
adjusted analysis appears clinically meaningful given the 1.5- comes. A more recent systematic review on VR gaming and
to 2-second change reported as a the minimal clinically arm motor function found only 2 randomized clinical trials,
important difference in previous studies.28 with just one including patients in the acute phase of stroke.14
The stroke rehabilitation research landscape has been No significant differences were observed between virtual-
evolving.2 The results of the EXCITE (Extremity Constraint environment training and conventional therapy groups in
Induced Therapy Evaluation) trial provided new insight in motor strength and functional scores. Finally, the authors of
stroke rehabilitation by showing that constraint-induced mo- this systematic review highlighted the potential value and
tor therapy can produce a clinically relevant improvement in safety of VR gaming as a tool for stroke rehabilitation but
arm function for patients within 6 months of a stroke.4 A concluded that VR gaming in stroke rehabilitation is an
study using functional MRI showed that repetitive bimanual unproven treatment, and much more evidence is needed from
stimulation produced bilateral activation of the motor corti- well-designed randomized trials.13,14 Considering the paucity
ces, one marker of brain plasticity.8 Interestingly, the com- of well-designed randomized studies, the scarce funding for
parison of Bobath, proprioceptive neuromuscular facilitation, stroke rehabilitation research, and the limitations of conven-
or motor relearning conventional techniques has shown that no tional rehabilitation, VRWii technology is accessible for all
one approach improves functional outcomes over another.29 –32 segments of the population at a relatively low cost, without
According to a recent systematic review, constraint-induced requiring special resources, assistance, or transportation to a
motor therapy represents, thus far, the most promising interven- specific facility.
tion to date for improving upper limb function.2 The limitations of our study deserve comment. First,
Limited evidence exists regarding the use of VR gaming EVREST was a pilot study with a small sample size, limiting
systems in stroke rehabilitation. Two recent systematic re- any definitive conclusions about the efficacy of VRWii.
views have summarized the available studies on VR in stroke EVREST was designed as a feasibility study and therefore
rehabilitation.13 Among 11 identified studies, there were only not powered to detect a difference between groups. However,
3 randomized clinical trials with diverse outcomes measures these results will allow for an informed estimate of the sample
(memory retraining, walking, gait, and postural stability). size required for an adequately powered large-scale random-
 Saposnik et al Virtual Reality Using Wii Technology in Stroke Rehab 1483

ized clinical trial. Second, because this study was single- the trial. We especially thank Jacqueline Willems, manager of the
blinded, it was possibly subject to bias in that patients using South East Toronto Stroke Network, and Dr Neville Bayer (de-
ceased), former director of the Stroke Program at St. Michael’s
the “new” technology may have been more motivated by the
Hospital, for their invaluable support. We also appreciate the
use of this treatment and also may have inadvertently dis- support, comments, and suggestions from Drs Vladimir Hachinski
closed their treatment allocation to the examiner. Neverthe- and Sandra Black during conception of the study.
less, this is a common issue in the design and implementation The EVREST Research Study Group comprised G.S. (principal
of randomized clinical trials in stroke rehabilitation, and not investigator), M.B., M.M., J.H., D.C., W.M., K.E.T., Jacqueline
Willems, Alexis Dishaw (Toronto, Canada), R.T., and L.G.C.
unique to EVREST. However, we had made an effort to The EVREST team at Toronto Rehabilitation Institute was com-
determine whether concealed allocation was preserved by posed of M.B. (site principal investigator), Laura Langer, Jennifer
asking the blinded assessor to guess the allocation group for Shaw, Hannah Cheung, and W.M.
each participant. No difference was found. Third, because The EVREST Steering Committee included G.S. (principal inves-
tigator of EVREST), M.B. (principal investigator at Toronto Reha-
safety and feasibility were the primary outcome measures in bilitation), J.H. (research manager), and M.M. (M.M.) at the Applied
EVREST, we have no information on the potential effects of Health Research Centre, who made all decisions concerning the
bimanual training in this population. Fourth, the VRWii implementation and conduction of the study.
group was significantly younger than the RT group. Older
subjects typically have slower reaction times, with an appar- Sources of Funding
ent poorer performance in the RT group. However, the results This study was supported by a grant from the Ministry of Health and
Long Term Care (MoHLTC) through the Ontario Stroke System
remained consistent after adjusting for age and other baseline
(OSS), administered by Heart and Stroke Foundation of Ontario
differences. Finally, the short duration of the intervention (8 (HSFO). We are most grateful for the initial funding provided by
sessions within 2 weeks) may underestimate the effect of South East Toronto Stroke Network, which helped with the early
VRWii gaming technology. In addition, the Wii system only organization of the study design and coordination.
provides feedback on the movement itself such that persons
might adopt a variety of movement strategies to successfully Disclosures
play the game. Some of these strategies (eg, significant The authors declare no financial conflicts of interest. Specifically, we
have not received support from Nintendo or other software devel-
shoulder/trunk motions) are not necessarily ideal adaptive opment companies.
strategies to reinforce.34 As a result, patient supervision may The sponsors were not involved in the design, execution, analysis,
be an important component of subsequent trials, as might be interpretation, or reporting of the results.
the addition of additional sensors to better represent feedback We declare that we have participated in the conception, design,
analysis, interpretation of the results, drafting the manuscript, and
of limb/arm position.
Downloaded from http://ahajournals.org by on April 1, 2024

made a critical revision of the manuscript.

Despite these limitations, EVREST is the first randomized K.T. (biostatistician), G.S., and J.H. all had full access to the data.
clinical trial showing that VRWii is a feasible, safe, and G.S. was responsible for obtaining funds and is supported by the
potentially effective intervention to enhance motor function clinician-scientist award from Heart and Stroke Foundation Ontario.
recovery in patients with a recent stroke and represents a
proof-of-concept trial.14 EVREST used a novel, simple, References
1. Nichols-Larsen DS, Clark PC, Zeringue A, Greenspan A, Blanton S.
wireless, and widely available 3D VR technology, allowing Factors influencing stroke survivors’ quality of life during subacute
the implementation of proven concepts in stroke rehabilita- recovery. Stroke. 2005;36:1480 –1484.
tion to improve motor function. Repetitive intense training 2. Langhorne P, Coupar F, Pollock A. Motor recovery after stroke: a sys-
tematic review. Lancet Neurol. 2009;8:741–754.
and the observation, practice, and representation on the 3. Barreca S, Wolf SL, Fasoli S, Bohannon R. Treatment interventions for
screen of task-specific activities can facilitate brain plasticity the paretic upper limb of stroke survivors: a critical review. Neurorehabil
mechanisms that engage the mirror neuron system or long- Neural Repair. 2003;17:220 –226.
term potentiation effects. 4. Wolf SL, Winstein CJ, Miller JP, Taub E, Uswatte G, Morris D, Giuliani
C, Light KE, Nichols-Larsen D. Effect of constraint-induced movement
In summary, the EVREST study constitutes the initial step therapy on upper extremity function 3 to 9 months after stroke: the
in the understanding of potential benefits of a novel interac- EXCITE randomized clinical trial. J Am Med Assoc. 2006;296:
tive approach in neurorehabilitation after stroke that can be 2095–2104.
5. Jutai JW, Teasell RW. The necessity and limitations of evidence-based
easily implemented in routine clinical practice. Further, VR practice in stroke rehabilitation. Top Stroke Rehabil. 2003;10:71–78.
technology could be used as adjuvant therapy to other proven 6. Teasell R, Meyer MJ, McClure A, Pan C, Murie-Fernandez M, Foley N,
successful interventions (ie, constraint-induced therapy). As Salter K. Stroke rehabilitation: an international perspective. Top Stroke
such, it provides hope for enhancing motor function and Rehabil. 2009;16:44 –56.
7. Kwakkel G, Wagenaar RC, Koelman TW, Lankhorst GJ, Koetsier JC.
improving quality of life in stroke survivors. Effects of intensity of rehabilitation after stroke. A research synthesis.
Stroke. 1997;28:1550 –1556.
Acknowledgments 8. Luft AR, McCombe-Waller S, Whitall J, Forrester LW, Macko R, Sorkin
JD, Schulz JB, Goldberg AP, Hanley DF. Repetitive bilateral arm training
The authors thank the support provided by the Ontario Stroke System
and motor cortex activation in chronic stroke: a randomized controlled
(OSS), the Ministry of Health and Long Term Care of Ontario
trial. J Am Med Assoc. 2004;292:1853–1861.
(MoHLTC), and Heart and Stroke Foundation Ontario (HSFO). The 9. Buccino G, Solodkin A, Small SL. Functions of the mirror neuron system:
authors also thank Florencia E. Saposnik (9), source of inspiration implications for neurorehabilitation. Cogn Behav Neurol. 2006;19:55– 63.
for the design of EVREST. The authors thank the co-op students at 10. Celnik P, Stefan K, Hummel F, Duque J, Classen J, Cohen LG. Encoding
Li Ka Shing Knowledge Institute for their assistance with database a motor memory in the older adult by action observation. Neuroimage.
development and data entry. We also thank the Li Ka Shing 2006;29:677– 684.
Knowledge Institute at St Michael’s Hospital and Toronto Rehabil- 11. Rizzolatti G, Fabbri-Destro M. The mirror system and its role in social
itation Institute for the efficient organization and implementation of cognition. Curr Opin Neurobiol. 2008;18:179 –184.
 1484 Stroke July 2010

12. Celnik P, Webster B, Glasser DM, Cohen LG. Effects of action obser- 22. Fisher RS, Harding G, Erba G, Barkley GL, Wilkins A. Photic- and
vation on physical training after stroke. Stroke. 2008;39:1814 –1820. pattern-induced seizures: a review for the Epilepsy Foundation of
13. Crosbie JH, Lennon S, Basford JR, McDonough SM. Virtual reality in America Working Group. Epilepsia. 2005;46:1426 –1441.
stroke rehabilitation: still more virtual than real. Disabil Rehabil. 2007; 23. Bonis J. Acute Wiiitis. N Engl J Med. 2007;356:2431–2432.
29:1139 –1146; discussion 1147–1152. 24. Borg G. Perceived exertion as an indicator of somatic stress. Scand J
14. Lucca LF. Virtual reality and motor rehabilitation of the upper limb after Rehabil Med. 1970;2:92–98.
stroke: a generation of progress? J Rehabil Med. 2009;41:1003–1100. 25. Mathiowetz V, Volland G, Kashman N, Weber K. Adult norms for the
15. Gowland C, Stratford P, Ward M, Moreland J, Torresin W, Van Hullenaar Box and Block Test of manual dexterity. Am J Occup Ther. 1985;39:
S, Sanford J, Barreca S, Vanspall B, Plews N. Measuring physical 386 –391.
impairment and disability with the chedoke-mcmaster stroke assessment. 26. Dobkin BH. Training and exercise to drive poststroke recovery. Nat Clin
Stroke. 1993;24:58 – 63. Pract Neurol. 2008;4:76 – 85.
16. Cote R, Battista RN, Wolfson C, Boucher J, Adam J, Hachinski V. The 27. Barak S, Duncan PW. Issues in selecting outcome measures to assess
Canadian Neurological Scale: validation and reliability assessment. functional recovery after stroke. NeuroRx. 2006;3:505–524.
28. Lin KC, Hsieh YW, Wu CY, Chen CL, Jang Y, Liu JS. Minimal detectable
Neurology. 1989;39:638 – 643.
change and clinically important difference of the Wolf Motor Function Test
17. Bushnell CD, Johnston DC, Goldstein LB. Retrospective assessment of
in stroke patients. Neurorehabil Neural Repair. 2009;23:429–434.
initial stroke severity: comparison of the NIH stroke scale and the
29. Langhammer B, Stanghelle JK. Bobath or motor relearning programme?
Canadian Neurological Scale. Stroke. 2001;32:656 – 660.
A comparison of two different approaches of physiotherapy in stroke
18. Wolf SL, Lecraw DE, Barton LA, Jann BB. Forced use of hemiplegic
rehabilitation: a randomized controlled study. Clin Rehabil. 2000;14:
upper extremities to reverse the effect of learned nonuse among chronic 361–369.
stroke and head-injured patients. Exp Neurol. 1989;104:125–132. 30. Logigian MK, Samuels MA, Falconer J, Zagar R. Clinical exercise trial
19. Carey JR, Kimberley TJ, Lewis SM, Auerbach EJ, Dorsey L, Rundquist for stroke patients. Arch Phys Med Rehabil. 1983;64:364 –367.
P, Ugurbil K. Analysis of fmri and finger tracking training in subjects 31. Lord JP, Hall K. Neuromuscular reeducation versus traditional programs
with chronic stroke. Brain. 2002;125:773–788. for stroke rehabilitation. Arch Phys Med Rehabil. 1986;67:88 –91.
20. Duncan PW, Wallace D, Lai SM, Johnson D, Embretson S, Laster LJ. 32. van Vliet PM, Lincoln NB, Foxall A. Comparison of bobath based and
The stroke impact scale version 2.0. Evaluation of reliability, validity, and movement science based treatment for stroke: a randomised controlled
sensitivity to change. Stroke. 1999;30:2131–2140. trial. J Neurol Neurosurg Psychiatry. 2005;76:503–508.
21. Saposnik G, Mamdani M, Bayley M, Thorpe KE, Hall J, Cohen LG, 33. Kim JH, Jang SH, Kim CS, Jung JH, You JH. Use of virtual reality to
Teasell R. Effectiveness of Virtual Reality Exercises in Stroke Rehabil- enhance balance and ambulation in chronic stroke: a double-blind, ran-
itation (EVREST): rationale, design, protocol and baseline data of a pilot domized controlled study. Am J Phys Med Rehabil. 2009;88:693–701.
randomized clinical trial assessing the Wii gaming system. Int J Stroke. 34. Cirstea MC, Levin MF. Compensatory strategies for reaching in stroke.
2010;5:47–51. Brain. 2000;123:940 –953.
Downloaded from http://ahajournals.org by on April 1, 2024