Virtual Reality – Upper Extremity

Evidence Reviewed as of before: 09-08-2016
Author(s)*: Tatiana Ogourtsova, MSc OT
Editor(s): Annabel McDermott, OT; Annie Rochette PhD, OT
Expert Reviewer: Mindy Levin PhD, PT (currently under revision)
Patient/Family Information Table of contents

Introduction

Virtual Reality (VR) is an environment that is simulated by a computer. It provides an interactive multi-sensory stimulation in real-time. VR provides users with the opportunity to engage in activities within an environment that appears and feels similar to real world objects and events. Users can interact with a virtual environment through the use of standard input devices such as a keyboard and mouse, or through multimodal devices such as a wired glove. VR is becoming increasingly popular as it can be easily modified according to the needs of individuals, it is perceived as being fun and motivating for patients, and it allows researchers to incorporate elements such as feedback that have been shown to maximize motor learning. On the negative side, there is concern that the use of VR in the clinic is not possible due to the cost of the required equipment. While certainly true when this technology was created, the cost of virtual reality hardware and software has decreased and is now reasonably affordable for clinical use.

Note: In this module we did not differentiate between immersive and non-immersive VR. This categorization is determined mainly by the degree of ‘virtual presence’ the subject experienced during training, and this information was not made readily available in most of the studies reviewed.

Note: This review focuses on any type of therapy involving a virtual environment. For a specific review of commercial game systems used for physical rehabilitation (e.g. Sony Playstation EyeToy, Nintendo Wii), please see the Video Games module.

Additional support from undergraduate students, School of Physical and Occupational Therapy, McGill University: Kareim Aziz, Sara Jafri, James Moore, Sebastien Mubayed, Roshnie Shah, Samrah Sher, and Peter Yousef.

Patient/Family Information

Authors*: Amy Henderson, PhD Student, Neuroscience; Dr. Nicol Korner-Bitensky PhD OT, Mindy Levin, PhD PT; Geoffroy Hubert BSc. Lic. K. ; Elissa Sitcoff BSc. B.A.

Expert: Francine Malouin PhD, PT

Additional support from undergraduate students, School of Physical and Occupational Therapy, McGill University: Kareim Aziz, Sara Jafri, James Moore, Sebastien Mubayed, Roshnie Shah, Samrah Sher, and Peter Yousef

What is virtual reality?

Virtual Reality is an environment that is simulated by a computer. Most virtual reality environments are primarily visual experiences, displayed either on a computer screen or through special stereoscopic displays (see picture 1), and may also include auditory stimulation through speakers or headphones. Users can interact with the virtual environment through the use of devices such as a keyboard, a mouse, or a wired glove (see picture 2).

Are there different kinds of virtual reality?

Generally, there are two types of virtual reality: full immersion, and non-immersion.

Full immersive VR is when the environment is viewed through a device such as a head-mounted display to create the illusion that one is inside the environment.

Non-immersive, or partially immersive VR, is when the user views the scene on a computer screen and it appears as if he was watching TV.

Why use virtual reality after a stroke?

Loss of leg function, movement, and strength are common after a stroke, and can result in the impairment of walking and standing.

Virtual reality is becoming an increasingly popular intervention used to improve the use of one’s leg after a stroke. It can be easily modified according to the needs of the individual, is perceived as being fun and motivating for patients, and allows researchers to include elements such as feedback that have been shown to maximize learning.

Does it work for stroke?

Researchers have studied how virtual reality can help stroke patients:

  • Remapping of the brain: virtual reality has been shown useful in retraining of the brain in persons who have had a stroke.
  • Walking: virtual reality was shown to be more useful than regular rehabilitation in improving walking speed, length of step, stamina, and strength in people who have had a stroke.
  • Stepping over obstacles:evidence has shown that virtual reality does not lead to any more improvement in stepping over obstacles than regular rehabilitation therapy.
  • Stair-climbing: There are no well-designed research studies that look at the effect of virtual reality on stair-climbing ability.
  • Community living skills: There is some evidence that shows that virtual reality is more useful than regular rehabilitation in helping people who have had a stroke develop the community living skill of “cross the street” or walking. However, there is conflicting evidence as to whether virtual reality provides any further benefit compared to regular rehabilitation in developing the community living skill “taking the train” in people who have had a stroke.
  • Perceived walking performance: There is evidence from one high quality study that virtual reality does not lead to any more improvement in how well patients view their ability to walk compared to regular rehabilitation therapy.

Side effects/risks?

Use of devices such as a head-mounted display can cause nausea and vertigo.

No real risks have been reported because of the absence of external manipulation. All activities are self-paced and under individual control and perception of movement.

Who provides the treatment?

VR treatments are usually provided by a Physical Therapist or Occupational Therapist. Presently most rehabilitation centers and private clinics are not equipped with this technology other than for research purposes. But, given the promising early evidence for the value of using VR, this treatment is likely to be integrated as part of post-stroke therapy in the future.

How many treatments?

Information on the amount and intensity of VR training needed is still not available. High quality studies need to be conducted before advice can be given regarding specific programs and content of treatment sessions

How much does it cost?

There is concern that the use of VR in the clinic is not possible due to the cost of the required equipment. While certainly true when this technology was created, the cost of virtual reality hardware and software has decreased and should soon be reasonably affordable for clinical use.

Is virtual reality for me?

There is clear evidence that there are benefits to using virtual reality in comparison to regular therapy or no therapy. These benefits include walking strength, how fast you can walk, length of step, stamina, the community living skill “crossing the street”, and remapping of the brain. However, in terms of obstacle clearance, VR was not shown to be more effective than conventional therapy. More studies are needed to determine if VR is an effective intervention for stair-climbing and the community living skill “taking the train”. So, overall, VR is an effective treatment you may want to consider after a stroke. If you are interested in learning more about VR, speak to your rehabilitation provider about the possibility of using this treatment.

Clinician Information

Note: When reviewing the findings, it is important to note that they are always made according to randomized clinical trial (RCT) criteria – specifically as compared to a control group. To clarify, if a treatment is “effective” it implies that it is more effective than the control treatment to which it was compared. Non-randomized studies are no longer included when there is sufficient research to indicate strong evidence (level 1a) for an outcome.

Note: All virtual reality training in this module was focused on improving the upper extremity (UE).

This review presents 28 studies (ten high quality RCTs, 11 fair quality RCTs, one poor quality RCT and six non RCTs studies) have investigated the effect of virtual reality for the upper-extremity on rehabilitation in patients with stroke.

Results Table

View results table

Outcomes

Acute phase

Functional independence/activities of daily living (ADLs)
Not effective
1a

Two high quality RCTs (Lee et al., 2014, Yin et al., 2014) and two fair quality RCTs (Piron et al., 2003; da Silva Cameirao et al., 2011) investigated the effect of upper extremity VR training on functional independence and activities of daily living (ADLs) in patients with acute stroke.

The first high quality RCT (Lee et al., 2014) randomized patients to receive VR training, cathodal transcranial direct current stimulation (tDCS), or a combination of VR training and cathodal tDCS. Functional independence/ADLs were measured by the Korean-Modified Barthel Index at post-treatment (3 weeks). There were no significant differences between any groups.

The second high quality RCT (Yin et al., 2014) randomized patients to receive either VR training combined with conventional rehabilitation or conventional rehabilitation alone. Functional independence/ADLs were measured by the Functional Independence Measure (FIM) at post-treatment (2 weeks) and at follow-up (1 month). No significant between-group differences were found at either time point.

The first fair quality RCT (Piron et al., 2003) randomized patients to receive either VR training or conventional rehabilitation. Functional independence/ADLs were measured by the FIM at post-treatment (5-7 weeks). No significant between-group differences were found.

The second fair quality RCT (da Silva Cameirao et al., 2011) randomized patients to receive VR training, intense occupational therapy or non-specific interactive games. Functional independence/ADLs were measured by the Barthel Index at post-treatment (12 weeks) and at follow-up (3 months). No significant between-group differences were found at either times point.

Conclusion: There is strong evidence (Level 1a) from two high quality RCTs and two fair quality RCTs that upper extremity VR training is not more effective than comparison interventions (cathodal transcranial direct current stimulation – tDCS, tDCS during VR training, conventional rehabilitation, intense occupational therapy or non-specific interactive games) for improving functional independence/ADLs in patients with acute stroke.

Manual dexterity
Not effective
1B

One high quality RCT (Lee et al., 2014) investigated the effects of upper extremity VR training on manual dexterity in patients with acute stroke. This high quality RCT randomized patients to receive VR training, cathodal transcranial direct current stimulation (tDCS), or a combination of VR training and cathodal tDCS. Manual dexterity was measured by the Box and Block Test (BBT) at post-treatment (3 weeks). No significant between-group differences were found.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that VR training for the upper extremity is not more effective than comparison interventions (cathodal transcranial direct stimulation – tDCS or tDCS with VR training) for improving manual dexterity in patients with acute stroke.

Motor activity
Not effective
1b

One high quality RCTs (Yin et al., 2014) and one fair quality RCT (da Silva Cameirao et al., 2011) investigated the effects of upper extremity VR training on upper extremity motor activity in patients with acute stroke.

The high quality RCT (Yin et al., 2014) randomized patients to receive either VR training combined with conventional rehabilitation or conventional rehabilitation alone. Upper extremity motor activity was measured by the Motor Activity Log – Amount of Use (MAL-AOU) and – Quality of Movement (MAL-QOM) scales at post-treatment (2 weeks) and at follow-up (1 month). No significant between-group differences were found at either time point.

The fair quality RCT (da Silva Cameirao et al., 2011) randomized patients to receive VR training, intense occupational therapy or non-specific interactive games. Upper extremity motor activity was measured by the Chedoke Arm and Hand Activity Inventory (CAHAI) at post-treatment (12 weeks) and at follow-up (3 months). Significant between-group differences were found at post-treatment in favor of VR training vs. intense occupational therapy, and in favor of VR training vs. non-specific interactive games. These differences did not remain significant at follow-up.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that VR training for the upper extremity is not more effective than comparison interventions (e.g. conventional rehabilitation) for improving upper extremity motor activity in patients with acute stroke. However, one fair quality RCT found that VR training for upper extremity was more effective, in short-term, than comparison interventions (e.g. intense occupational therapy or non-specific interactive games) for improving upper extremity motor activity in patients with acute stroke.
Note: The duration of the intervention (2 weeks vs. 12 weeks) could account for differences in outcomes found in both studies.

Motor function
Not effective
1A

Two high quality RCTs (Lee et al., 2014, Yin et al., 2014) and two fair quality RCTs (Piron et al., 2003, da Silva Cameirao et al., 2011) investigated the effect of upper extremity VR training on upper extremity motor function in patients with acute stroke.

The first high quality RCT (Lee et al., 2014) randomized patients to receive VR training, cathodal transcranial direct current stimulation (tDCS), or a combination of VR training and cathodal tDCS. Upper extremity motor function was measured by the Fugl-Meyer Assessment–upper extremity scale (FMA-UE) and the Manual Function Test (MFT) at post-treatment (3 weeks). There was a significant between-group difference on both measures in favour of VR training with tDCS vs. VR training alone or tDCS alone, and in favour of tDCS alone vs. VR training alone.

The second high quality RCT (Yin et al., 2014) randomized patients to receive either VR training combined with conventional rehabilitation or conventional rehabilitation alone. Upper extremity motor function was measured by the FMA-UE and the Action Research Arm Test (ARAT) at post-treatment (2 weeks) and at follow-up (1 month). No significant between-group differences on both measures were found at either time point.

The first fair quality RCT (Piron et al., 2003) randomized patients to receive either VR training or conventional rehabilitation. Upper extremity motor function was measured by the FMA-UE at post-treatment (5-7 weeks). No significant between-group differences were found.

The second fair quality RCT (da Silva Cameirao et al., 2011) randomized patients to receive VR training, intense occupational therapy or non-specific interactive games. Upper extremity motor function was measured by the FMA-UE at post-treatment (12 weeks) and at follow-up (3 months). Significant between-group differences were found at post-treatment in favor of VR vs. both control interventions. However, these differences did not remain significant at follow-up (3 months).

Conclusion: There is strong evidence (Level 1a) from two high quality RCTs and one fair quality RCT that VR is not more effective than comparison interventions (cathodal transcranial direct stimulation – tDCS, VR+tDCS or conventional rehabilitation) for improving upper extremity motor function in acute stroke. In fact, one high quality RCT found that VR alone was less effective than tDCS alone or VR+tDCS.
Note: However, one fair quality RCT reported that VR was more effective than comparison interventions (intensive OT, interactive games) – this study provided intervention over a longer duration than the other studies (12 weeks compared to 2 weeks, 3 weeks and 5-7 weeks).

Spasticity
Not effective
1B

One high quality RCT (Lee et al., 2014) investigated the effects of upper extremity VR training on upper extremity spasticity in patients with acute stroke. This high quality RCT randomized patients to receive VR training, cathodal transcranial direct current stimulation (tDCS), or a combination of cathodal tDCS and VR training. Spasticity was measured by the Modified Ashworth Scale (MAS) at post-treatment (3 weeks). No significant between-group differences were found.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that VR training for the upper extremity is not more effective than comparison interventions (cathodal transcranial direct stimulation – tDCS, tDCS with VR training) for improving upper extremity spasticity in patients with acute stroke.

Strength
Not effective
1B

One high quality RCT (Lee et al., 2014) and one fair quality RCT (da Silva Cameirao et al., 2011) investigated the effects of upper extremity VR training on upper extremity strength in patients with acute stroke.

The high quality RCT (Lee et al., 2014) randomized patients to receive VR training, cathodal transcranial direct current stimulation (tDCS), or a combination of cathodal tDCS and VR training. Strength was measured by the Manual Muscle Test (MMT) at post-treatment (3 weeks). No significant between-group differences were found.

The fair quality RCT (da Silva Cameirao et al., 2011) randomized patients to receive VR training, intense occupational therapy, or non-specific interactive games. Strength was measured by the Medical Research Council Grade (MRCG) & Motricity Index –upper extremity subscale (MI-UE) at post-treatment (12 weeks) and at follow-up (3 months). No significant between-group differences were found on both measures at either time point.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT and one fair quality RCT that VR training for the upper extremity is not more effective than comparison interventions (cathodal transcranial direct stimulation – tDCS, tDCS with VR training, intense occupational therapy, non-specific interactive games) for improving upper extremity strength in patients with acute stroke.

Subacute phase

Functional independence/activities of daily living (ADLs)
Not effective
2b

One quasi-experimental study (Piron et al., 2007) investigated the effects of upper extremity VR training on functional independence and activities of daily living (ADLs) in patients with subacute. This quasi-experimental study assigned patients to receive either VR training or conventional rehabilitation. Functional independence/ADLs were measured by the Functional Independence Measure (FIM) at post-treatment (5-7 weeks). No significant between-group differences were found.

Conclusion: There is limited evidence (Level 2b) from one quasi-experimental study that upper extremity VR training is not more effective than comparison interventions (e.g. conventional rehabilitation) for improving functional independence/ADLs in patients with subacute stroke.

Grip strength
Insufficient evidence
5

One single case pre-post design study (Broeren et al., 2004) investigated the effects of upper extremity VR training on grip strength in one patient with subacute stroke. This single case pre-post design study assigned one patient to receive VR training. Grip strength was measured by hand held dynamometer at baseline, at post-treatment (4 weeks) and at follow-up (5 months). A small improvement in grip strength was found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

Conclusion: There is insufficient evidence (Level 5) regarding the effect of VR on grip strength in patients with subacute stroke. However, one pre-post design study found an improvement in grip strength following VR training.

Manual dexterity
Insufficient evidence
5

One single case pre-post study (Broeren et al., 2004) investigated the effects of upper extremity VR training on manual dexterity in one patient with subacute stroke. This single case pre-post study assigned a patient to receive VR training. Manual dexterity was measured by the Purdue Peg Board Test (PPBT) at baseline, at post-treatment (4 weeks) and at follow-up (5 months). An improvement in manual dexterity was reported at both post-treatment time points.
Note: No analysis for statistical significance was reported.

Conclusion: There is insufficient evidence (Level 5) regarding the effect of VR on manual dexterity in patients with subacute stroke. However, it should be noted that one pre-post study found an improvement in manual dexterity following VR training.

Motor function
Not effective
2B

One quasi-experimental study (Piron et al., 2007) and one single case pre-post design study (Broeren et al., 2004) investigated the effects of upper extremity VR training on upper extremity motor function in patients with subacute stroke.

The quasi-experimental study (Piron et al., 2007) assigned patients to receive either VR training or conventional rehabilitation. Upper extremity motor function was measured by the Fugl-Meyer Assessment – Upper Extremity scale (FMA-UE) at post-treatment (5-7 weeks). No significant between-group differences were found.

The single case pre-post study (Broeren et al., 2004) assigned one patient to receive VR training. Upper extremity motor function was measured by the PHANToM haptic device at baseline, at post-treatment (4 weeks) and follow-up (5 months). An improvement in upper extremity movement was found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

Conclusion: There is limited evidence (Level 2b) from one quasi-experimental study that upper extremity VR training is not more effective than comparison interventions (e.g. conventional rehabilitation) for improving upper extremity motor function in patients with subacute stroke. However, it should be noted that one single case pre-post study found an improvement in upper extremity movement following VR training.

Chronic phase

Depression
Not effective
1b

One high quality RCT (Shin et al., 2015) investigated the effect of VR training for the upper extremity on depression in patients with chronic stroke. This high quality RCT randomized patients to receive either VR training with conventional occupational therapy (OT) or conventional OT alone. Depression was measured by the Korean Hamilton Depression Rating Scale at post-treatment (4 weeks). No significant between-group differences were found.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is not more effective than comparison interventions (conventional occupational therapy) for improving depression in patients with chronic stroke.

Executive function
Insufficient evidence
5

One pre-post study (Rand et al., 2009) investigated the effects of VR training on executive function in patients with chronic stroke. This pre-post study assigned patients to receive VR training. Executive function was measured by the Multiple Errands Test (Hospital Version) and the Virtual Multiple Errands Test at baseline and at post-treatment (3 weeks). An improvement in executive function was found.
Note: No analysis for statistical significance was reported.

Conclusion: There is insufficient scientific evidence (Level 5) that VR improves executive function in patients with chronic stroke. However, one pre-post study found an improvement in executive function following VR training.

Functional independence/activities of daily living (ADLs)
Not effective
1b

One high quality RCT (Piron et al., 2010) and three pre-post study (Rand et al., 2009, Burdea et al., 2010, Burdea et al., 2011) investigated the effect of upper extremity VR training on functional independence/Activities of Daily Living (ADLs) in patients with chronic stroke.

The high quality RCT (Piron et al., 2010) randomized patients to receive VR training or conventional rehabilitation. Functional independence/ADLs were measured by the Functional Independence Measure (FIM) at post-treatment (4 weeks). No significant between-group differences were found.

The first pre-post study (Rand et al., 2009) assigned patients to receive VR training. Functional Independence/ADLs were measured by the Activities of Daily Living questionnaire at baseline and at post-treatment (3 weeks). An improvement was found at post-treatment.
Note: No analysis for statistical significance was reported.

The second pre-post study (Burdea et al., 2010) assigned patients to receive VR training. Functional independence/ADLs were measured by the Upper Extremity Functional Index (UEFI) at baseline, at post-treatment (4 weeks) and at follow-up (3 months). Improved functional independence/ADLs were found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

The third pre-post design study (Burdea et al., 2011) assigned patients to receive VR training. Functional independence/ADLs were measured by the UEFI at baseline, at post-treatment (6 weeks) and at follow-up (3 months). Improved functional independence/ADLs were found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is not more effective than comparison interventions (e.g. conventional rehabilitation) for improving functional independence/ADLs in patients with chronic stroke.
Note: However, three non-randomized studies reported improvements in functional independence / ADLs following VR training in patients with chronic stroke (but analysis for statistical significance was not reported).

Grip strength
Not effective
1B

One high quality RCT (Thielbar et al., 2014), two fair quality RCTs (Housman et al., 2009, Friedman et al., 2014), and three non-randomized studies (Holden et al., 2002, Burdea et al., 2010, Burdea et al., 2011) investigated the effect of upper extremity VR training on grip strength in patients with chronic stroke.

The high quality RCT (Thielbar et al., 2014) randomized patients to receive VR training or intensity-matched occupational therapy. Grip strength was measured by the Jamar dynamometer at post-treatment (6 weeks) and at follow-up (10 weeks). No significant between-group differences were found at either time point.

The first fair quality RCT (Housman et al., 2009) randomized patients to receive VR training or conventional rehabilitation. Grip strength was measured by hand-held dynamometer at post-treatment (8-9 weeks) and at follow-up (6 months). No significant between-group differences were found at either time point.

The second fair quality RCT (Friedman et al., 2014) randomized patients to receive VR training, isometric movement training using the IsoTrainer, or conventional table top exercises. Grip strength was measured by Jamar dynamometer at post-treatment (2 weeks) and at follow-up (1 month). No significant differences were found between any groups at either time point.

The first pre-post study (Holden et al., 2002) assigned patients to receive VR training. Grip strength was measured by Jamar dynamometer at baseline and at post-treatment (20-30 sessions). No improvement was found.
Note: No statistical analysis for significance was reported in this study.

The second pre-post study (Burdea et al., 2010) assigned patients to receive VR training. Grip strength was measured by Jamar dynamometer at baseline, at post-treatment (4 weeks) and at follow-up (3 months). No improvement was found at either post-treatment time point.
Note: No statistical analysis for significance was reported in this study.

The third pre-post study (Burdea et al., 2011) assigned patients to receive VR training. Grip strength was measured by Jamar dynamometer at baseline, at post-treatment (6 weeks) and at follow-up (3 months). Improved grip strength was found at either post-treatment time point.
Note: No statistical analysis for significance was reported in this study.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT and two fair quality RCTs that upper extremity VR training is not more effective than comparison interventions (e.g. intensity-matched occupational therapy, conventional rehabilitation, isometric movement training using the IsoTrainer or conventional table top exercises) for improving grip strength in patients with chronic stroke. Two out of three non-randomized studies also reported no improvement in grip strength following VR training (but analysis for statistical significance was not reported).

Instrumental activities of daily living
Insufficient evidence
5

One pre-post study (Rand et al., 2009,) investigated the effect of VR on Instrumental Activities of Daily Living (IADLs) in patients with chronic stroke. This pre-post study assigned patients to receive VR training. IADLs were measured by the Instrumental Activities of Daily Living questionnaire at baseline and at post-treatment (3 weeks). Improved IADLs were found.
Note: Statistical significance was not reported in this study.

Conclusion: There is insufficient scientific evidence (Level 5) regarding the effect of VR on Instrumental ADLs in patients with chronic stroke. However, one pre-post study found an improvement in IADLs following VR training (but analysis for statistical significance was not reported).

Intrinsic motivation
Effective
1B

One high quality RCT (Subramanian et al., 2013) and one poor quality RCT (Sucar et al., 2009) investigated the effect of upper extremity VR training on intrinsic motivation in patients with chronic stroke.

The high quality RCT (Subramanian et al., 2013) randomized patients to receive either VR training vs. dose-matched upper extremity training within a physical environment. Intrinsic motivation was measured by the Intrinsic Motivation Task Evaluation Questionnaire – anxiety subscale) at post-treatment (4 weeks) and at follow-up (3 months). A significant between-group difference was found on both measures time points in favor of VR training vs. upper extremity training within a physical environment. However, subjects from the physical environment group reported feeling more comfortable practicing movements than those from the VR group.

The poor quality RCT (Sucar et al., 2009) randomized patients to receive VR training or conventional occupational therapy. Intrinsic motivation was measured by the Intrinsic Motivation Scale at post-treatment (5 weeks). No significant between-group difference was found.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is more effective than comparison interventions (e.g. dose-matched training in a physical environment) for improving intrinsic motivation in patients with chronic stroke. However, one poor quality RCT reported no significant between-group difference in intrinsic motivation between VR training and conventional occupational therapy.
Note: The differences in findings between the two studies could possibly result from the dissimilarity of the nature of comparison interventions. The high quality RCT used dose-matched pointing exercises within the physical environment, whereas the poor quality RCT employed conventional occupational therapy.

Kinematics
Not effective
1a

Two high quality RCTs (Piron et al., 2010, Subramanian et al., 2013) investigated the effect of VR on upper extremity kinematics in patients with chronic stroke.

The first high quality RCT (Piron et al., 2010) randomized patients to receive either VR training or conventional rehabilitation. Kinematic outcomes were measured by a 3D motion analysis at post-treatment (4 weeks). No significant between-group differences were found for upper extremity kinematics (duration, linear velocity, submovements).

The second high quality RCT (Subramanian et al., 2013) randomized patients to receive either VR training or upper extremity training within a physical environment. Kinematic outcomes were measured by a 3D motion analysis at post-treatment (4 weeks) and at follow-up (3 months). A significant between-group difference was found for only one kinematic outcome (shoulder horizontal abduction) at post-treatment in favor of the VR training vs. upper extremity training within a physical environment. No significant between-group differences were found in other kinematic outcomes (elbow extension, shoulder flexion) at either time point.

Conclusion: There is strong evidence (Level 1a) from two high quality RCTs that VR training is not more effective than comparison interventions (e.g. conventional rehabilitation, dose-matched upper extremity training within a physical environment) for improving kinematics in patients with chronic stroke.
Note: However, one high RCT did report that VR training was more effective, in short-term, than training in a physical environment on one kinematic outcome.

Manual ability
Not effective
1B

One high quality RCT (Piron et al., 2009) investigated the effect of upper extremity VR training on manual ability in patients with chronic stroke. This high quality RCT randomized patients to receive telerehabilitation VR training or physical therapy. Manual ability was measured by the ABILHAND at post-treatment (4 weeks) and at follow-up (2 months). No significant between-group differences were found at either time points.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is not more effective than comparison interventions (e.g. physical therapy) for improving manual ability in patients with chronic stroke.
Note: This study used telerehabilitation VR training.

Manual dexterity
Not effective
1b

One high quality RCT (Thielbar et al., 2014), three fair quality RCTs (Jang et al., 2005, Sung In et al., 2012, Friedman et al., 2014), and one pre-post study (Burdea et al., 2010) investigated the effect of upper extremity VR training on manual dexterity in patients with chronic stroke.

The high quality RCT (Thielbar et al., 2014) randomized patients to receive either VR training or intensity-matched occupational therapy. Manual dexterity was measured by the Jebsen-Taylor Hand Function Test (JTHFT) and the Finger Individuation Index (FII) at post-treatment (6 weeks) and at follow-up (10 weeks). A significant within-group difference in manual dexterity at post-treatment (FII) and at follow-up (JTHFT) was found in the VR training group. The subsequent non inferiority testing was performed only for the JTHFT, indicating that the intervention was not significantly inferior to the control treatment.

The first fair quality RCT (Jang et al., 2005) randomized patients to receive either VR training or conventional rehabilitation. Manual dexterity was measured by the Box and Block Test (BBT) at post-treatment (4 weeks). Significant between-group differences were found favoring the VR training vs. conventional rehabilitation.

The second fair quality RCT (Sung In et al., 2012) randomized patients to receive either VR Reflection Therapy or sham program. Manual dexterity was measured by the BBT and the JTHFT at post-treatment (4 weeks). No significant between-group differences on both measures were found.

The third fair quality RCT (Friedman et al., 2014) randomized patients to receive either VR training or isometric movement training using IsoTrainer (control 1) or conventional table top exercises (control 2). Manual dexterity was measured by the BBT and the 9-Hole Peg Test (9HPT) at post-treatment (2 weeks) and at follow-up (1 month). Significant between-group difference on both measures was found at post-treatment, favoring the VR training group vs. conventional table top exercises (control 2). The changes in manual dexterity (BBT) in the VR training group persisted at follow-up (1 month), but no statistical between-group analysis for significance was reported for this time point.

The pre-post study (Burdea et al., 2010) assigned patients to receive VR training. Manual dexterity was measured by the JTHFT at baseline, at post-treatment (4 weeks) and at follow-up (3 months). No improvement was found at either post-treatment time point.
Note: No statistical analysis for significance was reported in this study.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT and one fair quality RCTthat VR training is not more effective than comparison intervention (e.g. intensity-matched occupational therapy, sham program) in improving manual dexterity in patients with chronic stroke. One non-randomized study also reported no improvement in manual dexterity following VR training (but analysis for statistical significance was not reported).
Note: However, two fair quality RCTs found that VR training was more effective in improving manual dexterity than convention rehabilitation.

Motor activity
Not effective
1B

One high quality RCT (Subramanian et al., 2013) and one fair quality RCT (Housman et al., 2009) investigated the effect of upper extremity VR training on motor activity in patients with chronic stroke.

The high quality RCT (Subramanian et al., 2013 ) randomized patients to receive either VR training or training within a physical environment. Motor activity was measured by the Motor Activity Log – Amount of Use (MAL-AOU) scale at post-treatment (4 weeks) and at follow-up (3 months). No significant between-group difference was found at either time point.

The fair quality RCT (Housman et al., 2009) randomized patients to receive either VR training or conventional rehabilitation. Motor activity was measured by the MAL-AOU and MAL – Quality of Movement (MAL-QOM) scales at post-treatment (8-9 weeks) and at follow-up (6 months). No significant between-group difference was found on both measures at either time point.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT and one fair quality RCT that upper extremity VR training is not more effective than comparison interventions (e.g. training in a physical environment, conventional rehabilitation) for improving motor activity in patients with chronic stroke.

Motor function
Conflicting
4

Six high quality RCTs (Piron et al., 2009, Piron et al., 2010, Crosbie et al., 2012, Subramanian et al., 2013, Thielbar et al., 2014, Shin et al., 2015), four fair quality RCTs (Jang et al., 2005, Housman et al., 2009, Sung In et al., 2012, Friedman et al., 2014), one poor quality RCT (Sucar et al., 2009) and 2 pre-post studies (Holden et al., 2002, Burdea et al., 2011) investigated the effect of upper extremity VR training on upper extremity motor function in patients with chronic stroke.

The first high quality RCT (Piron et al., 2009) randomized patients to receive either telerehabilitation VR training or conventional rehabilitation. Motor function was measured by the Fugl-Meyer Assessment – Upper Extremity scale (FMA-UE) at post-treatment (4 weeks) and at follow-up (2 months). A significant between-group difference for motor function was found at post-treatment, in favor of telerehabilitation VR vs. conventional physical therapy. This significant between-group difference was not maintained at follow-up.

The second high quality RCT (Piron et al., 2010) randomized patients to receive either VR training or conventional rehabilitation. Motor function was measured by the FMA-UE at post-treatment (4 weeks). A significant between-group difference for motor function was found at post-treatment, in favor of VR training vs. conventional rehabilitation.

The third high quality RCT (Crosbie et al., 2012) randomized patients to receive either VR training or conventional physical therapy. Motor function was measured by the Action Research Arm Test (ARAT) at post-treatment (3 weeks) and at follow-up (6 weeks). No significant between-group differences for motor function were found at either time points.

The fourth high quality RCT (Subramanian et al., 2013) randomized patients to receive either VR training or training within a physical environment. Motor function was measured by the FMA-UE, Wolf Motor Function Test (WMFT), and Reaching Performance Scale for Stroke at post-treatment (4 weeks) and at follow-up (3 months). No significant between-group differences on any of the measures were found at either time point.

The fifth high quality RCT (Thielbar et al., 2014) randomized patients to receive VR training or intensity-matched occupational therapy. Motor function was measured by the FMA-UE and ARAT at post-treatment (6 weeks) and at follow-up (10 weeks). Significant within-group differences in motor function (FMA-UE, but not ARAT) were found for the VR training group only. Subsequent non inferiority testing found that VR training was significantly superior in improving motor function (ARAT only) from pre-treatment to follow-up.

The sixth high quality RCT (Shin et al., 2015) randomized patients to receive either VR training with conventional occupational therapy (OT) or conventional OT alone. Motor function was measured by the FMA-UE at post-treatment (4 weeks). No significant between-group difference was found.

The fist fair quality RCT (Jang et al., 2005) randomized patients to receive either VR training or no therapy. Motor function was measured by the FMA-UE and the Manual Function Test (MFT) at post-treatment (4 weeks). A significant between-group difference for motor function was found in favor of VR training vs. no therapy.

The second fair quality RCT (Housman et al., 2009) randomized patients to receive VR training or conventional rehabilitation. Motor function was measured by the FMA-UE and the Rancho Function Test for the Hemiplegic/Paretic Extremity (RFTHPE) at post-treatment (8-9 weeks) at follow-up (6 months). Although no significant between-group differences were found on both measures at post-treatment, a significant between-group difference was found at follow-up (FMA-UE only), in favor of VR training vs. conventional rehabilitation.

The third fair quality RCT (Sung In et al., 2012) randomized patients to receive VR Reflection Therapy or a sham program. Motor function was measured by the FMA-UE and MFT at post-treatment (4 weeks). A significant between-group difference was found in favour of VR Reflection Therapy vs. the sham program.

The fourth fair quality RCT (Friedman et al., 2014) randomized patients to receive VR training, isometric movement training using the IsoTrainer, or conventional table top exercises. Motor function was measured by the FMA-UE, WMFT, and ARAT at post-treatment (2 weeks) and at follow-up (1 month). No significant between-group differences on any of the measures were reported at either time point.

The poor quality RCT (Sucar et al., 2009) randomized patients to receive either VR training or conventional occupational therapy. Motor function was measured by the FMA-UE at post-treatment (5 weeks). No significant between-group differences were found.

The first pre-post study (Holden et al., 2002) assigned patients to receive VR training. Motor function was measured by the FMA-UE and the WMFT at baseline and at post-treatment (20-30 sessions). An improvement in motor function was found on both measures.
Note: No analysis for statistical significance was reported.

The second pre-post study (Burdea et al., 2011) assigned patients to receive VR training. Motor function was measured by the FMA-UE at baseline, at post-treatment (6 weeks) and at follow-up (3 months). An improvement in motor function (FMA-UE) was found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

Conclusion: There is conflicting evidence (Level 4) regarding the effectiveness of upper extremity VR training compared to other interventions for improving motor function among patients with chronic stroke. While two high quality RCTs and two fair quality RCTs reported that VR training was more effective than comparison interventions (conventional physical therapy, conventional rehabilitation, no training and sham program) for improving motor function in patients with chronic stroke, four high quality RCTs, two fair quality RCTs and one poor quality RCTreported that VR training was not more effective than comparison interventions (conventional physical therapy, training within a physical environment, intensity matched occupational therapy, conventional occupational therapy, conventional rehabilitation, isometric movement training using the IsoTrainer, conventional table top exercises) for improving motor function of patients.
Note: The two pre-post studies are not considered in the conclusion.

Pinch strength
Not effective
1B

One high quality RCT (Thielbar et al., 2014), one fair quality RCT (Friedman et al., 2014) and one pre-post study (Burdea et al., 2010) investigated the effects of VR on pinch strength in patients with chronic stroke.

The high quality RCT (Thielbar et al., 2014) randomized patients to receive either VR training or intensity-matched occupational therapy. Pinch strength was measured by a lateral and 3-point pinch meter at post-treatment (6 weeks) and at follow-up (10 weeks). No significant between-group difference was found at either time point.

The fair quality RCT (Friedman et al., 2014) randomized patients to receive VR training, isometric movement training using the IsoTrainer, or conventional table top exercises. Pinch strength was measured by the pinch gauge at post-treatment (2 weeks) and at follow-up (1 month). No significant between-group differences were found at either time point.

One pre-post study (Burdea et al., 2010) assigned patients to receive VR training. Pinch strength was measured by a pinch gauge at baseline, at post-treatment (6 weeks) and at follow-up (3 months). An improvement was found at post-treatment, but this improvement was not maintained at follow-up.
Note: No analysis for statistical significance was reported.

Conclusion: There is moderate evidence (Level 1b) one high quality RCT and one fair quality RCTthat upper extremity VR training is not more effective than comparison interventions (e.g. intensity-matched occupational therapy, isometric movement training or conventional table top exercises) for improving pinch strength in patients with chronic stroke.
Note: One pre-post study found improved pinch strength immediately after VR training (but analysis for statistical significance was not reported).

Quality of life
Not effective
1B

One high quality RCT (Shin et al., 2015) investigated the effect of upper extremity VR training on quality of life in patients with chronic stroke. This high quality RCT randomized patients to receive either VR training with conventional occupational therapy (OT) or conventional OT alone. Quality of life was measured by the Korean Short Form Health Survey SF-36 at post-treatment (4 weeks). No significant between-group differences were found for most measures of quality of life (Korean Short Form Health Survey, SF-36 – Physical functioning, Pain, General health, Social functioning, Mental health, Vitality, Role limitations due to emotional problems) at post-treatment. There was a significant between-group difference in one subtest (SF-36 – Role limitations due to physical problems), in favour of VR training compared to conventional rehabilitation.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is not more effective than comparison interventions (e.g occupational therapy) in improving quality of life in patients with chronic stroke.

Range of motion (hand/finger)
Insufficient evidence
5

Two pre-post studies (Burdea et al., 2010, Burdea et al., 2011) investigated the effect of upper extremity VR training on upper extremity range of motion (ROM) in patients with chronic stroke.

The first pre-post study (Burdea et al., 2010) assigned patients to receive VR training. ROM was measured by a mechanical goniometer at baseline, at post-treatment (4 weeks) and at follow-up (3 months). An improvement was found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

A second pre-post study (Burdea et al., 2011) assigned patients to receive VR training. ROM was measured by a mechanical goniometer at baseline, at post-treatment (6 weeks) and at follow-up (12 weeks). An improvement was found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

Conclusion: There is insufficient scientific evidence (Level 5) regarding the effect of upper extremity VR on hand/finger range of motion in patients with chronic stroke. However, two pre-post studies reported improved finger active ROM following VR training.

Range of motion (shoulder/reaching)
Not effective
2a

One fair quality RCT (Housman et al., 2009) and 2 pre-post studies (Burdea et al., 2010, Burdea et al., 2011) investigated the effect of upper extremity VR training on shoulder/elbow ROM in patients with chronic stroke.

The fair quality RCT (Housman et al., 2009) randomized patients to receive either VR training or conventional therapy. ROM was measured by reach distance between the wrist and a target at shoulder or elbow height at post-treatment (8-9 weeks) and at follow-up (6 months). No significant between-group differences were found at both time points.

The first pre-post study (Burdea et al., 2010) assigned patients to receive VR training. ROM was measured by a goniometer at baseline, at post-treatment (4 weeks) and at follow-up (3 months). An improvement in active shoulder ROM was found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

The second pre-post study (Burdea et al., 2011) assigned patients to receive VR training. ROM was measured by a goniometer at baseline, at post-treatment (6 weeks) and at follow-up (3 months). An improvement in active shoulder ROM (goniometer) was found at both post-treatment time points.
Note: No analysis for statistical significance was reported.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that upper extremity VR training is not more effective than comparison interventions (e.g. conventional rehabilitation) for improving shoulder/elbow range of motion in patients with chronic stroke. However, two pre-post studies reported improved shoulder/elbow ROM following VR training.

Spasticity
Not effective
1B

One high quality RCT (Piron et al., 2009) and one fair quality RCT (Sung In et al., 2012) investigated the effect of upper extremity VR training on spasticity in patients with chronic stroke.

The high quality RCT (Piron et al., 2009) randomized patients to receive telerehabilitation VR training or conventional physical therapy. Spasticity was measured by the Modified Ashworth Scale (MAS) at post-treatment (4 weeks) and at follow-up (2 months). No significant between-group differences were found at either time point.

The fair quality RCT (Sung In et al., 2012) randomized patients to receive either VR Reflection Therapy or sham program. Spasticity was measured by the MAS at post-treatment (4 weeks). No significant difference was found.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT and one fair quality RCT that upper extremity VR training is not more effective than comparison interventions (e.g. conventional physical therapy or sham therapy) for improving spasticity in patients with chronic stroke.

Strength
Not effective
1B

One high quality RCT (Crosbie et al., 2012), one poor quality RCT (Sucar et al., 2009), and one pre-post study (Holden et al., 2002) investigated the effect of upper extremity VR training on strength in patients with chronic stroke.

The high quality RCT (Crosbie et al., 2012) randomized patients to receive VR training or conventional physical therapy. Strength was measured by the Motricity Index (MI) at post-treatment (3 weeks) and follow-up (6 weeks). No significant between-group differences were found at either time point.

The poor quality RCT (Sucar et al., 2009) randomized patients to receive either VR training or conventional occupational therapy. Strength was measured by the MI at post-treatment (4 weeks). No significant between-group difference was found.

The pre-post study (Holden et al., 2002) investigated the effect of upper extremity VR training on strength in patients with chronic stroke. Strength was measured by cuff weight placed on the forearm at baseline and at post-treatment (20-30 sessions). There was a notable improvement in the weight that was lifted at post-treatment.
Note: No analysis for statistical significance was reported.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT and one poor quality RCT that upper extremity VR training is not more effective than comparison interventions (e.g. physical therapy or occupational therapy) for improving strength in patients with chronic stroke. However, one pre-post study found improvement in strength following VR training (but analysis for statistical significance was not reported).

Phase of stroke recovery not specific to one period

Cognitive function
Not effective
2a

One fair quality RCT (Kim et al., 2011) investigated the effect of upper extremity VR training on cognitive function in patients with stroke. This fair quality RCT randomized patients with acute/subacute stroke to receive either VR training with computer-assisted cognitive rehabilitation or computer-assisted cognitive rehabilitation alone. Cognitive function was measured by the Korean version of the Mini-Mental Status Examination and computerized neuropsychological tests (visual continuous performance test, auditory continuous performance test, word color test, forward digit span test, backward digit span test, forward visual span test, backward visual span test, visual learning test, verbal learning test, Trail making test – A) at post-treatment (4 weeks). There were no significant between-group differences on any test item at post-treatment. However, in the VR group, the changes scores from pre- to post-treatment in the visual continuous performance test and the backward visual span test were significantly higher than those in the control group.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that upper extremity VR training is not more effective than comparison interventions (computer-assisted cognitive rehabilitation) for improving certain aspects of cognitive function in patients with stroke.

Executive function
Not effective
2A

One fair quality RCT (Kim et al., 2011) investigated the effect of upper extremity VR training on executive function in patients with stroke. This fair quality RCT randomized patients with acute/subacute stroke to receive either VR training with computer-assisted cognitive rehabilitation or computer-assisted cognitive rehabilitation alone. Executive function was measured by the Tower of London Test at post-treatment (4 weeks). No significant between-group differences were found.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT upper extremity VR training is not more effective than comparison interventions (computer-assisted cognitive rehabilitation) for improving executive function in patients with stroke.

Functional Independence/activities of daily living (ADLs)
Not effective
1b

One high quality RCT (Shin et al., 2014) and four fair quality RCTs (Kiper et al., 2011, Kim et al., 2011, Kwon et al., 2012, Turolla et al., 2013) investigated the effect of upper extremity VR training on functional independence and activities of daily living (ADLs) in patients with stroke.

The high quality RCT (Shin et al., 2014) randomized patients with acute/subacute stroke to receive either VR training with conventional occupational therapy (OT) or conventional OT alone. Functional independence/ADLs were measured by the Modified Barthel Index (MBI) at post-treatment (2 weeks). No significant between-group difference was found.

The first fair quality RCT (Kiper et al., 2011) randomized patients with subacute to chronic stroke to receive either VR training with traditional neuromotor rehabilitation or traditional neuromotor rehabilitation alone. Functional independence/ADLs were measured by the Functional Independence Measure (FIM) at post-treatment (4 weeks). A significant between-group difference was found in favour of VR training with traditional neuromotor rehabilitation vs. traditional neuromotor rehabilitation alone.

The second fair quality RCT (Kim et al., 2011) randomized patients with acute/subacute stroke to receive either VR training with computer-assisted cognitive rehabilitation or computer-assisted cognitive rehabilitation alone. Functional independence/ADLs were measured by the Korean-Modified Barthel Index (K-MBI) at post-treatment (4 weeks). No significant between-group differences were found.

The third fair quality RCT (Kwon et al., 2012) randomized patients with acute/subacute stroke to receive either VR training with conventional therapy or conventional therapy alone. Functional independence/ADLs were measured by the K-MBI at post-treatment (4 weeks). No significant between-group differences were found.

The fourth fair quality RCT (Turolla et al., 2013) randomized patients with subacute to chronic stroke to receive either VR training with conventional rehabilitation or conventional rehabilitation alone. Functional independence/ADLs were measured by the FIM at post-treatment (4 weeks). A significant between-group difference was found in favour of VR training with conventional rehabilitation vs. conventional rehabilitation alone.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT and two fair quality RCTs that upper extremity VR training is not more effective than comparison interventions (conventional OT, computer-assisted cognitive rehabilitation or conventional therapy) for improving functional independence/ADLs in patients with stroke. However, two fair quality RCTs found that VR training was more effective than comparison interventions (traditional neuromotor rehabilitation and conventional rehabilitation) for improving functional independence/ADLs in patients with stroke.
Note: The two fair quality RCTs that found between-group differences both used the FIM in patients with subacute / chronic stroke; whereas the three studies that found no significant between-group differences all used the K-mBI and patients with acute/subacute stroke. The present distinction in assessment and in stage of stroke recovery could contribute to the difference in findings across studies.

Manual dexterity
Not effective
1B

One high quality RCT (Shin et al., 2016) investigated the effect of upper extremity VR training on manual dexterity in patients with stroke. This high quality RCT randomized patients with acute to chronic stroke to receive either VR training or conventional rehabilitation. Manual dexterity was measured with the Purdue Pegboard Test (PPT) at baseline, at post-treatment (4 weeks) and at follow-up (1 month). No significant changes in scores were found at either post-treatment time points for both groups.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is not more effective than comparison interventions (conventional rehabilitation) for improving manual dexterity in patients with stroke.

Motor function
Conflicting
4

Two high quality RCTs (Shin et al., 2014, Shin et al., 2016) and four fair quality RCTs (Kiper et al., 2011, Jo et al., 2012, Kwon et al., 2012, Turolla et al., 2013) investigated the effect of upper extremity VR training on motor function in patients with stroke.

The first high quality RCT (Shin et al., 2014) randomized patients with acute/subacute stroke to receive VR training with conventional occupational therapy (OT) or OT alone. Motor function was measured by the Fugl-Meyer Assessment – Upper Extremity (FMA-UE) at post-treatment (2 weeks). No significant between-group differences were found.

The second high quality RCT (Shin et al., 2016) randomized patients with acute to chronic stroke to receive either VR training or conventional rehabilitation. Motor function was measured using the FMA-UE) and the Jebsen-Taylor Hand Function Test (JTHFT) at baseline, at post-treatment (4 weeks) and at follow-up (1 month). Significant changes in scores were found in motor function (FMA-UE total, proximal and distal scores, JTHFT total and gross scores) at both time points in the VR training group, but not in the conventional rehabilitation group.

The first fair quality RCT (Kiper et al., 2011) randomized patients with subacute to chronic stroke to receive either VR training with traditional neuromotor rehabilitation or traditional neuromotor rehabilitation alone. Motor function was measured by the FMA-UE at post-treatment (4 weeks). A significant between-group difference was found favoring VR training with traditional neuromotor rehabilitation vs. traditional neuromotor rehabilitation alone.

The second fair quality RCT (Jo et al., 2012) randomized patients with stroke (time since stroke not specified) to receive VR training with conventional rehabilitation or conventional rehabilitation alone. Motor function was measured by the Wolf Motor Function Test (WMFT) at baseline and at post-treatment (4 weeks). Although both groups showed significant improvement in motor function (WMFT total score, arm and hand subtest scores), no between-group analyses were reported.

The third fair quality RCT (Kwon et al., 2012) randomized patients with acute/subacute stroke to receive VR training with conventional therapy or conventional therapy alone. Motor function was measured by the FMA-UE and the Manual Function Test at post-treatment (4 weeks). No significant between-group differences were found on both measures.

The fourth fair quality RCT (Turolla et al., 2013) randomized patients with subacute to chronic stroke to receive either VR training with conventional therapy or conventional therapy alone. Motor function was measured by the FMA-UE at post-treatment (4 weeks). A significant between-group difference was found in favor of VR training with conventional therapy vs. conventional therapy alone group.

Conclusion: There is conflicting evidence (Level 4) from two high quality RCTs and four fair quality RCTs regarding the effects of upper extremity VR training on motor function in patients with stroke. While a first high quality RCT and two fair quality RCTs found that VR training was not more effective than comparison interventions (occupational therapy, conventional rehabilitation), a second high quality RCTand two fair quality RCTs found that VR training was more effective than comparison interventions (conventional rehabilitation and traditional neuromotor rehabilitation) for improving motor function among populations of patients with acute to chronic stroke.
Note: Differences in stage of stroke of participants, duration of training and type of VR training may contribute to the lack of agreement between studies.

Range of motion
Not effective
1B

One high quality RCT (Shin et al., 2014) investigated the effect of upper extremity VR training on passive range of motion in patients with stroke. This high quality RCT randomized patients with acute/subacute stroke to receive either VR training with occupational therapy or occupational therapy alone. Passive range of motion was measured (measurement tool not specified) at post-treatment (2 weeks). No significant between-group differences were found.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is not more effective than comparison interventions (occupational therapy) for improving passive range of motion in patients with stroke.

Spasticity
Effective
2A

One fair quality RCT (Kiper et al., 2011) investigated the effect of upper extremity VR training on spasticity in patients with stroke. This fair quality RCT randomized patients with subacute to chronic stroke to receive VR training with traditional neuromotor rehabilitation or traditional neuromotor rehabilitation alone. Spasticity was measured by the Modified Ashworth Scale (MAS) at post-treatment (4 weeks). Significant between-group differences were found among patients with ischemic type stroke, (in favor of VR training with traditional neuromotor rehabilitation) but not among patients with hemorrhagic type stroke.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that upper extremity VR training is more effective than control interventions (e.g. traditional neuromotor rehabilitation) for improving spasticity in patients with ischemic stroke.

Strength
Not effective
2A

One fair quality RCT (Kim et al., 2011) investigated the effect of upper extremity VR training on strength in patients with stroke. This quality RCT randomized patients with acute/subacute stroke to receive VR training with computer-assisted cognitive rehabilitation or computer-assisted cognitive rehabilitation alone. Strength was measured by the Motricity Index at post-treatment (4 weeks). No significant between-group differences were found.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that upper extremity VR training is not more effective than comparison interventions (computer-assisted cognitive rehabilitation) for improving strength in patients with stroke.

Stroke outcomes
Effective
1B

One high quality RCT (Shin et al., 2016) investigated the effect of upper extremity VR training on stroke outcomes in patients with stroke. This high quality RCT randomized patients with acute to chronic stroke to receive either VR training or conventional rehabilitation. Stroke outcomes were measured with the Stroke Impact Scale (SIS) at baseline and at post-treatment (4 weeks). Significant changes in scores was found for some measures of stroke outcomes (SIS composite and overall scores, social participation and mobility subscores) in VR training group but not in conventional rehabilitation. No significant changes in scores were found at post-treatment for other stroke outcomes (SIS memory and thinking, communication, emotion, strength and hand subscores).

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that upper extremity VR training is more effective than comparison interventions (conventional rehabilitation) for improving certain aspects of stroke outcomes in patients with stroke.

Visual perception
Effective
2A

One fair quality RCT (o et al., 2012J) investigated the effect of upper extremity VR training on visual perception in patients with stroke. This fair quality RCT randomized patients with stroke (time since stroke not specified) to receive VR training with conventional rehabilitation or conventional rehabilitation alone. Visual perception was measured by the Motor Free Visual Perceptual Test (MVPT) at post-treatment (4 weeks). There was a significant between-group difference on some measures of visual perception (MVPT total, time, visual discrimination and form constancy subtests), in favour of VR training vs. conventional rehabilitation. There were no significant differences on other measures of visual perception (MVPT visual memory, visual closure, spatial relations).

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that upper extremity VR training is more effective than comparison interventions (conventional rehabilitation) for improving some measures of visual perception in patients with stroke.

References

Broeren, J., Rydmark, M., Sunnerhagen, K. (2004). Virtual reality and haptics as a training device for movement rehabilitation after stroke:A single-case study. Archives of Physical Medicine and Rehabilitaton , 85, 1247-1250.
http://www.ncbi.nlm.nih.gov/pubmed/15295748

Burdea, G., Cioi, D., Martin, J., Rabin, B., Kale, A., & DiSanto, P. (2011). Motor retraining in virtual reality: a feasibility study for upper-extremity rehabilitation in individuals with chronic stroke. Journal of Physical Therapy Education , 25, 20-29.
https://www.questia.com/library/journal/1P3-2413003451/motor-retraining-in-virtual-reality-a-feasibility

Burdea, G.C., Cioi, D., Martin, J., Fensterheim, D., & Holenski, M. (2010). The rutgers arm II rehabilitation system—a feasibility study. IEEE Transactions On Neural Systems And Rehabilitation Engineering , 18(5), 505-514.
http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=5482032&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel5%2F7333%2F5596181%2F05482032.pdf%3Farnumber%3D5482032

Crosbie, J.H., Lennon, S., McGoldrick, M.C., McNeil, M.D.J., & McDonough, S.M. (2012). Virtual reality in the rehabilitation of the arm after hemiplegic stroke: a randomized controlled pilot study. Clinical Rehabilitation , 26(9), 798-806.
http://www.ncbi.nlm.nih.gov/pubmed/22275463

da Silva Cameirão, M., Bermúdez, I.B., Duarte, E., Verschure P.F. (2011). Virtual reality based rehabilitation speeds up functional recovery of the upper extremities after stroke: a randomized controlled pilot study in the acute phase of stroke using the rehabilitation gaming system. Restor Neurol Neurosci ., 29(5), 287-98.
http://www.ncbi.nlm.nih.gov/pubmed/21697589

Friedman, N., Chan, V., Reinkensmeyer, A.N., Beroukhim, A., Zambrano, G.J., Bachman, M., Reinkensmeyer, D.J. (2014). Retraining and assessing hand movement after stroke using the MusicGlove: comparison with conventional hand therapy and isometric grip training. Journal of Neuroengineering & Rehabilitation , 11:76.
http://www.ncbi.nlm.nih.gov/pubmed/24885076

Holden, M., & Dyar, T. (2002). Virtual environment training: A new tool for neurorehabilitation. Neurology Report , 26(2), 62-71.
http://web.mit.edu/bcs/bizzilab/publications/holden2002a.pdf

Housman, S.J., Scott, K.M., & Reinkensmeyer, D.J. (2009). A randomized controlled trial of gravity-supported, computer-enhanced arm exercise for individuals with severe hemiparesis. Neurorehabilitation and Neural Repair , 23(5), 505-14.
http://www.ncbi.nlm.nih.gov/pubmed/19237734

Jang, S.H., You, S.H., Hallett, M., Cho, Y.W., Park, C.M., Cho, S.H., Lee, H.Y., Kim TH. (2005). Cortical reorganization and associated functional motor recovery after virtual reality in patients with chronic stroke: An experimenter-blind preliminary study. Archive of Physical Medicine Rehabilitation , 86, 2218-2223.
http://www.ncbi.nlm.nih.gov/pubmed/16271575

Jo, K., Yu, J., & Jung, J. (2012). Effects of virtual reality based rehabilitation on upper extremity function and visual perception in stroke patients: a randomized control trial. Journal of Physical Therapy Science , 24, 1205–8.
https://www.jstage.jst.go.jp/article/jpts/24/11/24_1205/_pdf

Kiper, P., Piron, L., Turolla, A., Stozek, J., & Tonin, P. (2011). The effectiveness of reinforced feedback in virtual environment in the first 12 months after stroke. Neurologia i Neurochirurgia Polska , 45(5), 436–44.
http://www.ncbi.nlm.nih.gov/pubmed/22127938

Kim, B.R., Chun, M.H., Kim, L.S., & Park, J.Y. (2011). Effect of virtual reality on cognition in stroke patients. Annals of Rehabilitation Medicine , 35, 450-9.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3309247/

Kwon, J., Park, M., Yoon, I., & Park, S. (2012). Effects of virtual reality on upper extremity function and activities of daily living performance in acute stroke: a double-blind randomized clinical trial. Neurorehabilitation, 31(4), 379–85.
http://www.ncbi.nlm.nih.gov/pubmed/23232161

Lee, S.J. & Chun, M.H. (2014). Combination transcranial direct current stimulation and virtual reality therapy for upper extremity training in patients with subacute stroke. Ar chives of Physical Medicine and Rehabilitation , 95 (3), 431-438.
http://www.ncbi.nlm.nih.gov/pubmed/24239790

Piron, L., Turolla, A., Agostini, M., Zucconi, C., Cortese, F., Zampolini, M., Zannini, M., Dam, M., Ventura, L., Battauz, M., & Tonin, P. (2009). Exercises for paretic upper limb after stroke: a combined virtual-reality and telemedicine approach. Journal of Rehabilitation Medicine , 41, 1016–20.
http://www.ncbi.nlm.nih.gov/pubmed/19841835

Piron, L., Tonin, P., Atzori, A.M., Zucconi, C., Massaro, C., Trivello, E., & Dam, M. (2003). The augmented-feedback rehabilitation technique facilitates the arm motor recovery in patients after a recent stroke. Stud Health Technol Inform , 94, 265-7.
http://www.ncbi.nlm.nih.gov/pubmed/15455905

Piron L, Tombolini P,Turolla A, Zuccon Ci,Agostini M, Dam M, Santarello G, Piccione F & Tonin P. (2007). Reinforced Feedback in Virtual Environment Facilitates the Arm Motor Recovery in Patients after a Recent Stroke. International Workshop of Virtual Rehabilitation (IEEE), 121–3.
http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4362151&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D4362151

Piron, L., Turolla, A., Agostini, M., Zucconi, C.S., Ventura, L., Tonin, P., & Dam, M. (2010). Motor learning principles for rehabilitation: A pilot randomized controlled study in poststroke patients. Neurorehabilitation Neural Repair , 24, 501-508.
http://www.ncbi.nlm.nih.gov/pubmed/20581337

Rand, D., Weiss, P. L., & Katz, N. (2009). Training multitasking in a virtual supermarket: A novel intervention after stroke. American Journal of Occupational Therapy , 63, 535–542.
http://www.ncbi.nlm.nih.gov/pubmed/19785252

Shin, J.H., Ryu, H., & Jang, S.H. (2014). A task-specific interactive gamebased virtual reality rehabilitation system for patients with stroke: a usability test and two clinical experiments. Journal of NeuroEngineering and Rehabilitation , 11, 32.
http://www.ncbi.nlm.nih.gov/pubmed/24597650

Shin, J.H., Park, S.B., & Jang, S.H. (2015). Effects of game-based virtual reality on health-related quality of life in chronic stroke patients: A randomized, controlled study. Computers in Biology and Medicine , 63, 92-98.
http://www.ncbi.nlm.nih.gov/pubmed/26046499

Shin, J.H., Kim M.Y., Lee, J.Y., Jeon, Y.J., Kim, S., Lee, S., Seo, B., & Choi, Y. (2016). Effects of virtual reality-based rehabilitation on distal upper extremity function and health-related quality of life: a single-blinded, randomized controlled trial. Journal of NeuroEngineering and Rehabilitation , 13, 17.
http://www.ncbi.nlm.nih.gov/pubmed/26911438

Subramanian, S.K. Lourenco, C.B., Chilingaryan, G., Sveistrup, H., & Levin, M.F. (2013). Arm motor recovery using a virtual reality intervention in chronic stroke: Randomized control trial. Neurorehabilitation and Neural Repair , 27 (1), 13-23.
http://www.ncbi.nlm.nih.gov/pubmed/22785001

Sucar, L.E., Leder, R., Hernandez, J., Sanchez, I., Azcarate, G. (2009). Clinical evaluation of a low-cost alternative for stroke rehabilitation. IEEE 11th International Conference on Rehabilitation Robotics , 863–6.
http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=5209526&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel5%2F5188775%2F5209456%2F05209526.pdf%3Farnumber%3D5209526

Sung In, T., Sim Jung, K., Lee, S.W., & Ho Song, S. (2012). Virtual Reality Reflection Therapy Improves Motor Recovery and Motor Function in the Upper Extremities of People with Chronic Stroke. Journal of Physical Therapy Science , 24 (4), 339-43.
https://www.jstage.jst.go.jp/article/jpts/24/4/24_339/_article

Thielbar, K.O., Lord, T.J., Fischer, H.C., Lazzaro, E.C., Barth, K.C., Stoykov, M.E., Triandafilou, K.M., Kamper, D.G. (2014). Training finger individuation with a mechatronic-virtual reality system leads to improved fine motor control post-stroke. Journal of Neuroengineering & Rehabilitation , 11:171.
https://jneuroengrehab.biomedcentral.com/articles/10.1186/1743-0003-11-171

Turolla, A., Dam, M., Ventura, L., Tonin, P., Agostini, M., Zucconi, C., Kiper, P., Cagnin, A., & Piton, L. (2013). Virtual reality for the rehabilitation of the upper limb motor function after stroke: A prospective controlled trial. Journal of NeuroEngineering and Rehabilitation , 10 (1), 85.
http://www.ncbi.nlm.nih.gov/pubmed/23914733

Yin, C.W., Sien, N.Y., Ying, L.A., Chong Man Chung, S.F., & Tan May Leng, D. (2014). Virtual reality for upper extremity rehabilitation in early stroke: a pilot randomized controlled trial. Clinical Rehabilitation , 28(11), 1107-14.
http://www.ncbi.nlm.nih.gov/pubmed/24803644

Excluded Studies

Acosta, A.M., Dewald, H.A., & Dewald, J.P.A (2011). Pilot study to test effectiveness of video game on reaching performance in stroke. Journal of Rehabilitation Research & Development, 48(4), 431-444.
Reason for exclusion: Compares one type of VR with another type of VR, which is outside the scope of this module.

Connelly, L., Jia, Y., Toro, M.L., Stoykov, M.E., Kenyon, R.V., & Kamper, D.G. (2010). A pneumatic glove and immersive virtual reality environment for hand rehabilitative training after stroke. IEEE Transactions On Neural Systems And Rehabilitation Engineering, 18, 5.
Reason for exclusion: Both groups received a type of VR training.

Byl, N.N., Abrams, G.M., Pitsch, E., Fedulow, I., Kim, H., Simkins, M., Nagarajan, S., & Rosen, J. (2013). Chronic stroke survivors achieve comparable outcomes following virtual task specific repetitive training guided by a wearable robotic orthosis (UL-EXO7) and actual task specific repetitive training guided by a physical therapist. Journal of Hand Therapy, 26(4), 343-52.
Reason for exclusion: No intervention of interest (i.e. robotics therapy).

Flynn, S., Palma, P., & Bender, A. (2007). Feasibility of using the Sony PlayStation 2 gaming platform for an individual poststroke: a case report. Journal of Neurologic Physical Therapy, 31, 180–189.
Reason for exclusion: Commercially available gaming therapy.

Hijmans, J.M., Hale, L.A., Satherley, J.A., McMillan, N.J., King, M.J. (2011). Bilateral upper-limb rehabilitation after stroke using a movement-based game controller. J Rehabil Res Dev., 48 (8), 1005-13.
Reason for exclusion: Not an RCT, and all included outcomes available from a RCT.

Holden, M., & Dyar, T. (2002). Virtual environment training: A new tool for neurorehabilitation.Neurology Report, 26(2), 62-71.
Reason for exclusion: Not an RCT, and all included outcomes available from a RCT.

Kim, I.C. & Lee, B.H. (2012). Effects of Augmented Reality with Functional Electric Stimulation on Muscle Strength, Balance and Gait of Stroke Patients. Journal of Physical Therapy Science, 24 (8), 755-62.
Reason for exclusion: No outcome of interest (i.e. lower extremity function only).

Joo, L.Y., Yin, T.S., & Xu, D. (2010). A feasibility study using interactive commercial off-the- shelf computer gaming in upper limb rehabilitation in patients after stroke. Journal of Rehabilitation Medicine, 42, 437–441.
Reason for exclusion: Commercially available gaming therapy.

Laver, K.E., George, S., Thomas, S., Deutsch, J.E., & Crotty, M. (2015). Virtual reality for stroke rehabilitation. Cochrane Database of Systematic Reviews, 2:CD008349.
Reason for exclusion: Review.

Lee, D., Lee, M., Lee, K., & Song, C. (2014). Asymmetric training using virtual reality reflection equipment and the enhancement of upper limb function in stroke patients: A randomized controlled trial. Journal of Stroke and Cerebrovascular Diseases. 23 (6), 1319-1326.
Reason for exclusion: Both groups receive a type of VR.

Lewis, G.N., Woods, C., Rosie, J.A., & McPherson, K.M. (2011). Virtual reality games for rehabilitation of people with stroke: perspectives from the users. Disabil Rehabil Assist Technol, 6 (5), 453-63.
Reason for exclusion: Not an RCT, and all included outcomes available from a RCT.

Lohse. K.R., Hilderman, C.G.E., Cheung, K.L., Tatla, S., Van der Loos, M. (2014). Virtual reality therapy for adults post-stroke: A systematic review and meta-analysis exploring virtual environments and commercial games in therapy. PLoS ONE, 9 (3), e93318.
Reason for exclusion: Review.

McNulty, P.A., Thompson-Butel, A.G., Faux, S.G., Lin, G., Katrak, P.H., Harris, L.R., & Shiner, C.T. (2015). The efficacy of Wii-based Movement therapy for upper limb rehabilitation in the chronic poststroke period: A randomized controlled trial. International Journal of Stroke, 10 (8), 1253-1260.
Reason for exclusion: No intervention of interest (i.e. gaming therapy, not virtual reality).

Mouawad, M.R., Doust, C.G., Max, M.D., & McNulty, P.A. (2011). Wii -based movement therapy to promote improved upper extremity function post-stroke: a pilot study. Journal of Rehabilitation Medicine, 43, 527–533.
Reason for exclusion: Commercially available gaming therapy.

Kang, S.H., Kim, D.K., Seo, K.M., Choi, K.N., Yoo, J.Y., Sung, S.Y. & Park, H.J. (2009). A computerized visual perception rehabilitation programme with interactive computer interface using motion tracking technology – a randomized controlled, single-blinded, pilot clinical trial study. Clinical Rehabilitation, 23, 434–44.
Reason for exclusion: Intention of VR was to improve visual perception, which is outside the scope of this module. This module focuses on VR with the intention of improving the upper extremity.

King, M., Hale, L., Pekkari, A., Persson M., Gregorsson, M., & Nilsson, M. (2010). An affordable, computerised, table-based exercise system for stroke survivors, Disability and Rehabilitation. Assistive Technology, 5(4), 288-293.
Reason for exclusion: Not an RCT, and all included outcomes available from a RCT.

Piron, L., Tonin, P., Piccione, F., Iaia, V., Trivello, E., & Dam, M. (2005). Virtual environment training therapy for arm motor rehabilitation. Presence, 14(6), 732-740.
Reason for exclusion: Not an RCT, and all included outcomes available from a RCT.

Rabin, B.A., Burdea, G.C., Roll, D.T., Hundal, J.S., Damiani, F., & Pollack, S. (2012). Integrative rehabilitation of elderly stroke survivors: the design and evaluation of the BrightArmTM. Disability & Rehabilitation Assistive Technology, 7 (4), 323-35.
Reason for Exclusion: Single subject design (n=5).

Saposnik, G., Teasell, R., Mamdani, M., Hall, J., McIlroy, W., Cheung, D. Stroke Outcome Research Canada (SORCan) Working Group. (2010). Rehabilitation: a pilot randomized clinical trial and proof of principle effectiveness of virtual reality using Wii gaming technology in stroke. Stroke, 41, 1477-1484.
Reason for exclusion: Commercially available gaming therapy.

Sheehy, L., Taillon-Hobson, A., Sveistrup, H., Bilodeau, M., Fergusson, D., Levac, D., & Finestone, H. (2016). Does the addition of virtual reality training to a standard program of inpatient rehabilitation improve sitting balance ability and function after stroke? Protocol for a single-blind randomized controlled trial. BMC Neurology, 16 (1), 42.
Reason for exclusion: Protocol proposal.

Shiri, S., Feintuch, U., Lorber-Haddad, A., Moreh, E., Twito, D., Tuchner-Arieli, M., & Meiner, Z. (2012). Novel virtual reality system integrating online self-face viewing and mirror visual feedback for stroke rehabilitation: Rationale and feasibility. Topics in Stroke Rehabilitation. 19 (4), 277-286.
Reason for exclusion: Single subject design (n=6).

Schuck, S.O., Whetstone, A., Hill, V., Levine, P., & Page, S.J. (2011). Game-based, portable, upper extremity rehabilitation in chronic stroke. Top Stroke Rehabil. 18 (6), 720-7.
Reason for exclusion: Not an RCT, and all included outcomes available from a RCT.

Sin, H.H. & Lee, G.C. (2013). Additional virtual reality training using Xbox Kinect in stroke survivors with hemiplegia. American Journal of Physical Medicine and Rehabilitation, 92, 871–80.
Reason for exclusion: Commercially available gaming therapy.

Standen, P., Brown, D., Battersby, S., Walker, M., Connell, L., Richardson, A., Platts, F., Threapleton, K., & Burton, A. (2011). A study to evaluate a low cost virtual reality system for home based rehabilitation of the upper limb following stroke. International Journal on Disability and Human Development, 10 (4), 337–41.
Reason for exclusion: Protocol proposal.

Standen, P., Threapleton, K., Richardson, A., Connell, L., Brown, D., Battersby, S., Platts, F., Burton, A. (2016). A low cost virtual reality system for home based rehabilitation of the arm following stroke: A randomised controlled feasibility trial. Clinical Rehabilitation, 30. pii: 0269215516640320.
Reason for exclusion: Feasibility trial without between-group analysis.

Standen, P.J., Threapleton, K., Connell, L., Richardson, A., Brown, D.J., Battersby, S., Sutton, C.J., & Platts, F. (2015). Patients’ use of a home-based virtual reality system to provide rehabilitation of the upper limb following stroke. Physical Therapy, 95(3), 350-9.
Reason for exclusion: Prospective cohort with qualitative analysis.

Szturm, T., Peters, J., & Otto, C. (2008). Task-specific rehabilitation of finger-hand function using interactive computer gaming. Archives of Physical Medicine and Rehabilitation, 89, 2213-2217.
Reason for exclusion: Not an RCT, and all included outcomes available from a RCT.

Trobia, J., Gaggioli, A., & Antonietti, A. (2011). Combined use of music and virtual reality to support mental practice in stroke rehabilitation. Journal of CyberTherapy & Rehabilitation; 4 (1), 57.
Reason for exclusion: Case report (n=2).

Viana, R.T., Laurentino, G,E., Souza, R.J., Fonseca, J.B., Silva Filho, E.M., Dias, S.N., Teixeira-Salmela, L.F., & Monte-Silva, K.K. (2014). Effects of the addition of transcranial direct current stimulation to virtual reality therapy after stroke: A pilot randomized controlled trial. NeuroRehabilitation, 34 (3), 437-446.
Reason for exclusion: No intervention of interest (i.e. gaming therapy, not virtual reality).

Yavuzer, G., Senel, A., Atay, M.B., & Stam, H.J. (2007). “Playstation EyeToy games” improve upper extremity-related motor functioning in subacute stroke: a randomized controlled trial. Journal of Rehabilitation Medicine, 44, 237–244.
Reason for exclusion: Commercially available gaming therapy.

Zheng, C., Liao, W., & Xia, W. (2015). Effect of combined low-frequency repetitive transcranial magnetic stimulation and virtual reality training on upper limb function in subacute stroke: a double-blind randomized controlled trail. Journal of Huazhong University of Science and Technology. Medical Sciences, 35(2), 248-54.
Reason for exclusion: Both groups received VR training.

What do you think?