The Landscape of Virtual Reality Use in Mobility Rehabilitation from 2010–2023: A Scoping Review

Abstract

Significant advancements in virtual reality (VR) technology have occurred in the past decade, allowing clinical researchers to take advantage of these reduced barriers to explore the use of VR in patient populations. This scoping review on VR interventions to improve mobility in adults and children focuses on the literature from 2010–2023. A total of 2736 articles were screened and 126 articles met the inclusion criteria. Most of the studies were conducted in inpatient clinical settings (n = 41) and investigated VR interventions to improve balance (n = 118). Less immersive (n = 108) products such as Nintendo Wii or Xbox Kinect were primarily used. Additionally, 37.0% of studies (n = 47) used off-the-shelf programs like Wii Fit Plus and 73.2% of studies (n = 93) found statistically significant improvements in motor outcomes following VR intervention. The articles included in this review suggest that the majority of VR research for physical rehabilitation is being performed in clinical settings. Most studies reported statistically significant improvements in their outcome variables following VR intervention. These observations demonstrate that research in this area is moving beyond proof-of-concept and toward translation to clinical applications.

1. Introduction

Physical rehabilitation provided by a Physical Therapist (PT) or Physical Therapist Assistant (PTA) aims to address functional deficits and optimize independence. While the scope of rehabilitation addressed by PTs and PTAs can vary by state and country, a common goal is to restore healthy movement. Rehabilitation provided by PTs and PTAs can include a focus on the upper and/or lower body, depending on the patient’s needs. Restoring independent mobility is a prevalent therapeutic goal, as it has been shown to be positively associated with quality of life [1]. Therefore, the focus of this scoping review is mobility rehabilitation.

PTs and PTAs create evidence-based treatment protocols to meet a patient’s functional goals by utilizing a variety of therapeutic modalities. One such modality is virtual reality (VR), which has been identified as a promising treatment tool to provide effective and meaningful interventions across a variety of human health applications, including cognitive rehabilitation [2], pulmonary rehabilitation [3], pain management [4], intensive care medicine [5], and improvement of motor outcomes [6,7,8,9]. VR can be defined as a computer-generated environment which contains sensory information that may be interacted with, visualized, and manipulated to allow natural behaviors to emerge as if the environment were real [10,11,12]. VR helps meet a clinical need by providing a training modality that can be challenging to recreate in the real world [13,14] and has been shown to enhance psychological motivation to continue the prescribed treatment [15,16]. Previous research investigating the use of VR in rehabilitation has demonstrated significant improvements in mobility outcomes [6,7,8,9]; however, most of this research has been conducted in controlled laboratory environments and the translation to clinical practice appears to be limited. As suggested by Cano Porras et al. [8], the merit of VR interventions has continued to be supported (see recent reviews by Góra et al. [17] and Bateni et al. [18]); thus, research in this area should move beyond purely efficacy-based questions and begin to answer questions regarding the efficacy of different types of VR hardware compared to each other and the benefits of using particular types of VR in specific clinical settings or with particular populations, as well as investigating the more deliberate designs of VR intervention.

Research interest in the use of VR as a rehabilitation modality has increased since the introduction of low-cost, commercially available VR hardware around 2010 and has continued as technology has further evolved [11]. While relatively small-scale research studies have explored the utility of VR in physical rehabilitation, there have been few large-scale attempts to identify clinically relevant aspects of these studies. To address this gap, the research questions addressed by our scoping review focused on the following: (1) which populations (healthy and clinical) have been included in this type of research, (2) the geographical location where these studies have occurred, (3) the elements of the study design, (4) the type of VR used, and most importantly, (5) the intervention outcomes. In relation to the latter, it should be noted that the infusion of VR in rehabilitation is not universally positive. For example, it has been shown that older adult dropout rates are higher in randomized study designs that use new technology relative to control interventions [19]. This observation aligns with the finding that older adults’ technological acceptance of gamified VR is related to their initial feelings after their first encounter with VR (i.e., was the interaction positive or negative) and their perceived usefulness of the technology [20]. Addressing these research gaps will help to inform research scientists and clinicians of the current evidence for the inclusion of VR in physical rehabilitation.

While VR has been used in several rehabilitation contexts, the purpose of our scoping review was to better understand the research-to-practice landscape of VR for physical rehabilitation interventions to improve mobility. Specifically, we aimed to investigate how VR is being used in physical rehabilitation research to-date, including which clinical populations are being included, the settings and locations of studies, and the types of VR hardware and software being used as a therapeutic tool.

2. Materials and Methods

This scoping review followed the protocol set by the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) updated guidance [21]. The PRISMA-ScR checklist can be found in Table S1 in the Supplementary Materials. While our protocol was not pre-registered, the search terms and methods were agreed upon prior to the search. Additionally, to guide the inclusion criteria, a participant, intervention, comparison, outcomes, and study design (PICOS) model was used [22]. The Zotero citation management system and Excel were used for managing records and data during this review.

A keyword search was performed between 31 July 2023 and 17 August 2023 using the following databases: PubMed, Scopus, and EBSCOhost. Searches were further refined by selecting English-language texts. The keyword searches included all possible combinations of (1) “virtual reality”; (2) “physical therapy”, “Physiotherapy”, or “rehabilitation”; (3) “gait”, “ambulation”, “balance”, “mobility”, “function” or “motor performance”. As previously noted, the year 2010 marked a significant transition in the development, usability, and accessibility of VR hardware. Thus, our publication years for the search were constrained to 2010 through the year the search was conducted (2023).

The study selection was guided by a participant, intervention, comparison, outcomes, and study design (PICOS) model, which can be seen in

Table 1. PICOS inclusion/exclusion criteria.

Participants	Studies were included if they included human (healthy or clinical) populations. Modeling and animal model studies were excluded.
Interventions	Studies were included if they covered treatment interventions using VR that aimed to improve gait, standing balance, and/or upright mobility. Studies were excluded if they used adjunct therapeutic modalities simultaneously with the VR intervention (robot-assisted gait training, transcranial magnetic stimulation, etc.).
Comparisons	Studies were included if they included pre- and post-treatment assessments and/or comparisons between VR and conventional treatment (i.e., balance training, etc.). Studies were excluded if they did not include these comparisons.
Outcomes	Studies were included if they reported functional or clinical outcome measurements of gross motor abilities, or biomechanical measurements related to gait, balance, or upright mobility. Studies with outcomes outside of this description were excluded.
Study Design	Study designs that employed baseline and post-treatment assessments were included. Studies were excluded if they were case studies or had <10 participants in the experimental group.

. The PICOS model was set according to the primary aim of this review, which was to investigate the use of VR as an intervention to improve gait, standing balance, or upright mobility. Using these eligibility criteria, an initial screening of the studies was performed. The full texts of the studies not eliminated during the initial screening were further reviewed for the inclusion and exclusion criteria. The titles, abstracts, and full texts of the articles were assessed independently for eligibility by two reviewers, GP and CH. In cases of disagreement, GP and CH discussed with DF and came to an agreement.

Following the database searches and study selection, data extraction was performed to generate the descriptive statistics of the included studies. Data regarding the publication year, patient population, publication location and setting, study protocols, sample size and age, VR hardware and software used, and intervention outcomes were extracted. Information regarding each intervention was also extracted, such as the intervention protocol, objectives, and outcomes. Following the framework used by Campo-Prieto et al. [9], VR was judged as non-immersive, semi-immersive, or fully immersive. Immersion (semi or full) was operationally defined for our study as the feedback provided when interacting with a virtual reality environment.

3. Results

3.1. Study Selection

Figure 1 illustrates the article selection process. The initial database searches led to a total of 3813 articles for initial screening derived from PubMed (n = 2735), Scopus (n = 613), and EBSCOhost (n = 465). Of these 3813 articles, 1022 were duplicates and 55 were removed because they either did not have associated text or were only abstracts (e.g., conference presentations). After removing these articles, 2736 articles were screened for inclusion through their title and abstract. Of these articles, 2413 articles were excluded for not meeting the PICOS criteria due to being a systematic review, meta-analysis, study protocol, case study, or commentary, or because they used an adjunct modality, studied cognitive or psychological factors, did not investigate a gross motor task or did not use gross motor outcomes, were not a VR intervention, or used an upper extremity task. Following the screening process, 323 full-text articles were sought for retrieval for full-text review, and 75 articles were unable to be retrieved as full-text versions were not available at the time of review. This left 248 articles to be assessed for full-text review. Of these 248 full-text articles, 122 were excluded because they did not meet the PICOS eligibility criteria. The 126 remaining articles are included in this review [23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148]. A summary of the demographics of all included studies is presented in Table S2, followed by a summary of their findings in Table S3—both of which are in the Supplementary Materials.

3.2. Publication Year

This search for this review focused on articles published between 2010 and 2023. Of the included articles, none were published in 2010. Over half (n = 66; 52.0%) of the included articles were published in the second half of the last decade, between 2017 and 2021. A table of the distribution of publications between 2011 and 2023 can be seen in Figure 2.

3.3. Population Studied

The majority of studies (n = 121; 95.3%) used adult populations (mean age: 60.9 ± 13.3 years), with six studies (4.7%) using pediatric participants (mean age: 11.1 ± 2.4 years). Across all 126 included articles, the primary populations studied were stroke survivors (23.6%), Parkinson’s disease patients (20.5%), and older adults (13.4%). Figure 3 shows the distribution of all included populations.

3.4. Location and Setting

The largest number of studies were conducted in South Korea (n = 23; 18.1%), Italy (n = 14; 11.0%), and Brazil (n = 13; 10.2%). Research conducted in the United States (n = 8; 6.3%) and Spain (n = 6; 4.7%) contributed to the included publications. Four publications each came from Poland (3.1%), Switzerland (3.1%), Israel (3.1%), Taiwan (3.1%), China (3.1%), and Turkey (3.1%). Three publications each came from Australia (2.4%), Malaysia (2.4%), and Pakistan (2.4%). Two publications each came from Canada (1.6%), Egypt (1.6%), France (1.6%), and Japan (1.6%). The remaining countries each produced 0.8% (n = 1, each) of the articles included in this review: Austria, Chile, Germany, India, Jordan, Singapore, Slovenia, and Sweden. Four articles (3.1%) did not provide explicit information regarding the geographic location of their research. The locations of the research conducted for each publication included in this review can be viewed in Figure 4. The largest percentage of studies were conducted in inpatient settings of clinical facilities (n = 37; 29.1%), such as acute rehabilitation centers, hospitals, subacute rehabilitation facilities, and nursing homes. This is followed by laboratory settings (n = 27; 21.3%), community centers (n = 9; 7.1%), outpatient clinical facilities (n = 8; 6.3%), home-based settings (n = 5; 3.9%), telerehabilitation settings (n = 4; 3.2%), and research settings (n = 3; 2.4%). Notably, 24 studies (18.9%) did not explicitly state their research setting.

3.5. Study Protocols

Sixty-seven (52.8%) studies used interventions with the primary aim of improving balance. This is followed in descending order by studies with interventions targeting gait only (n = 18; 14.2%), both gait and balance (n = 17; 13.4%), upright mobility (n = 9; 7.1%), balance and upright mobility (n = 3; 2.4%), and a combination of gait, balance, and upright mobility (n = 4; 3.1%). The total sample size for each study ranged widely from 10 to 195 subjects. The average number of subjects used was 37.0 (±25.7 participants). The majority of studies (n = 108; 85.0%) had a total sample size of 50 or less. The distribution of the sample sizes of the included studies can be seen in Figure 5.

Most studies (n = 76; 59.8%) utilized a two-group study design comparing VR to an active control performing some form of conventional rehabilitation (CR). This is followed by studies with a single, VR-only group (n = 28; 22.0%) or a three-group design comparing VR, CR, and VR + CR (n = 11; 8.7%). The distribution of the study design groups can be viewed in Figure 6. More specifically, the figure shows a breakdown of the intervention types (e.g., VR only, VR + CR, etc.) across the number/type of participant groups (e.g., single group, two-group stratified), illustrating the diversity of study designs in this line of research. A total of 70 studies (55.1%) used study designs with a control group. Of these studies, 48 used a randomized method for allocating their participants, while 22 did not randomize group allocation. The studies employed study designs with various dosages of VR interventions. Intervention length varied from 1 to 12 weeks, with most studies using 6 weeks (n = 23; 18.1%) closely followed by 4 weeks (n = 21; 16.5%). Intervention frequency ranged from once per week to two times per day; the total session number ranged from 5 to 60 sessions. Intervention duration ranged from 15 to 60 min per session. Of note, 25 studies (19.7%) provided a rationale for the design of their VR intervention with empirical support, while 102 studies (80.3%) did not provide an evidence-based rationale for their intervention.

3.6. Virtual Reality Hardware and Software

The majority of studies used less-immersive (n = 44; 34.6%), commercially available VR hardware such as the Nintendo Wii and Xbox Kinect. Thirty-four studies (26.8%) used semi-immersive hardware, including the Computer Assisted Rehabilitation Environment (CAREN) system (n = 2), the Gait Real-time Interactive Lab (GRAIL) system (n = 3), and multi-screen interactive setups (n = 6), which provide increased surroundedness and interactivity over less-immersive VR technologies. Of the thirty-four semi-immersive setups, eight were medical grade hardware (CAREN, GRAIL), and six were custom setups which are not off-the-shelf products. Forty-nine studies (38.3%) used immersive technologies with head-mounted displays (HMD), providing a higher level of immersion due to the full-surroundedness of the display. This immersive VR hardware includes the Oculus Rift (n = 9) and other custom setups (n = 18). Figure 7 and Figure 8 show the distribution of the level of immersion (Figure 7) and the grade (medical, commercial, custom-created, Figure 8) of all VR hardware used in the included publications. One study did not report the details of the VR hardware used for their intervention.

The majority of studies (n = 71; 56.3 %) used off-the-shelf commercial software such as Wii Fit Plus or Kinect Adventures. Twenty-one studies (16.7 %) relied on medical- or research-grade packages (e.g., CAREN, GRAIL). Twenty-nine studies (23.0 %) implemented custom-built programs, typically created in platforms like Unity or 3D Studio Max. Five studies (4.0 %) did not report which software they employed. Figure 9 illustrates the full breakdown of VR software types.

3.7. Intervention Outcomes

Most studies (n = 93; 73.2%) reported significant improvements in outcomes related to gait, balance, and upright mobility following their VR intervention. Thirty-six of these studies found a significant improvement in gait, balance, or upright mobility in their VR treatment group compared to the comparison group following VR intervention. Eighteen of the studies (14.2%) that reported between-group differences in favor of the VR group employed study designs in which the VR group also participated in conventional rehabilitation programs. Two studies (1.6%) reported improvements in their VR group; however, these improvements were not statistically significant. None of the included studies reported no change or a decline in function related to gait, balance, or upright mobility following their VR intervention.

4. Discussion

This scoping review aimed to investigate how VR has been used in physical rehabilitation research from 2010 to present. All studies reported improvements in the motor domains of gait, balance, or upright mobility following VR intervention. This is consistent with previous review papers that found that VR interventions provide effective therapeutic outcomes that can be superior to standard care [6,7,8,9]. Despite repeated findings that support the effectiveness of VR interventions to improve motor outcomes, it is important to point out the known publication bias toward statistically significant results over those that accept the null hypothesis [149]. Thus, studies finding a degradation, or no change, in motor skills following VR intervention may not be published, which should be taken into consideration when interpreting the results of this review. However, to this point, 1.5% (n= 2) of the studies did publish non-significant results, reporting that the VR group did exhibit improvements in the functional outcomes of their study, but these improvements were not statistically significant.

Prior reviews and meta-analyses regarding VR for rehabilitation have found that studies have limited methodological validity due to small samples, a lack of controls, or the absence of a clear rationale for the development of their intervention [6,7,8,9]. In this review, the average sample size was 37 participants, with 101 studies having a total sample size of 50 participants or less. Of these 101 studies, most of them had sample sizes between 10 and 20. This is important to consider, as small sample sizes can lead to low statistical power and an overestimation of the significance of the intervention [150]. It should also be noted that an exclusion criterion for this review was studies with less than 10 subjects in the experimental group. Of the 248 articles included for full-text review, 16.4% (n = 41) were excluded due to having an experimental group smaller than 10.

Previous systematic reviews have cited the lack of a comparison group, a lack of treatment or waitlist control groups, and/or a lack of randomization as methodological shortcomings in this area of research [1,7,151]. Regarding the study design for the articles included in this review, 26.5% of the studies used two-group designs with primarily active control groups participating in CR interventions. The use of a two-group design is the cleanest way to test the efficacy of an intervention and provide a performance comparison to a group not participating in the intervention [152]. Further, including an active control group participating in an intervention considered to be standard care provides a more rigorous comparison than a no-treatment control group [7]. Additionally, the use of active control groups helps to circumvent any ethical dilemmas related to providing no treatment to clinical populations [153]. Regarding group allocation, 37.5% of studies used randomization methods for allocating subjects to their study groups. As previously mentioned, small sample size is a common limitation in this area of research, and small sample sizes can reduce statistical power. However, randomization techniques can help boost statistical power that may be affected by both small sample sizes and between-group analyses [150,151,152]. The findings of this review suggest that VR researchers adhered to this methodological quality in their study designs, as most studies had an active comparison group and random allocation of participants to each group. Nevertheless, research continues to lack clear rationale regarding the development of interventions. Only 19.5% of the studies in this review provided a rationale with empirical support for at least part of the design of their VR intervention.

Studies continue to focus primarily on adults and predominantly neurological populations. A majority of studies (93.0%) had adult participants, while only nine included pediatric populations. This disparity between adult and pediatric populations could be due to the general composition of the population, which is projected to see considerable growth in the population over 65 years of age [154]. Additionally, the relative ease of recruiting adult subjects compared to obtaining the parental consent for minors could be a contributing factor. More than half of the studies investigated the role of VR interventions in improving gait, balance, or upright mobility in neurologic populations, including those with brain injury, stroke, multiple sclerosis, spinal cord injury, Parkinson’s disease, or cerebral palsy. This is possibly due to the extent to which neurological deficits impact a person’s overall functional independence and quality of life, making investigating effective treatment interventions a high priority. Injuries to the nervous system can result in a range of deficits, including changes in motor planning and execution, strength, and coordination—all of which can profoundly impact a person’s ability to ambulate, balance, or perform everyday activities of daily living [155,156]. Thus, the primary aims of physical rehabilitation interventions for this population align with the primary aim of this review, which was to investigate VR interventions for gait, balance, and upright mobility.

Additionally, physical rehabilitation to improve motor deficits following neurological impairment involves therapy protocols that use high levels of repetition, which can quickly become disengaging for participants [14,15,16]. To address this barrier, VR has been shown to have psychological benefits for patients in PT, including reducing tension, increasing calmness, easing fatigue, reducing depression, improving motivation, and enhancing the quality of life [156,157,158]. Aside from neurologic populations, older adults are the next most prominently studied group, comprising 19.5% of the publications included in this review. Aging is associated with frailty, decreased independence, and falls, especially in older adults over the age of 65 [159]. Age-related changes in vision and the vestibular and somatosensory systems, as well as changes in the musculoskeletal system, can influence how older adults maintain balance during movement and upright mobility, again, aligning typical physical rehabilitation treatment plans for this population with the primary outcomes of this review [160]. This is also in line with other reviews that have found VR to be effective in improving mobility, balance, and overall physical functioning in older adults [19,161,162,163].

Most of the studies included in this review were conducted in more ecologically valid settings than the small portion (17.1%) of studies performed in sterile laboratory environments. It should be noted that 16 publications did not explicitly report their research setting. Despite this, the findings of this review show that the majority of VR research into physical rehabilitation interventions for gait, balance, and upright mobility was conducted in clinical environments, which is in contrast to a previous review which found that VR research is primarily performed in laboratory settings [6]. This could be due to the fact that the previous review was only investigating VR interventions for upright mobility in neurological populations, creating a narrower review of the literature. Additionally, the article searches for the previous review were performed from 2017 to 2018. In the current review, of the studies performed in non-laboratory settings (n = 62), 43.5% of them were published in 2019 or later.

Most of the studies included in this review were conducted outside of the United States, with most being performed in South Korea, followed closely by Brazil and Italy. Regardless of their location, the studies primarily used less immersive and commercially available VR hardware. Additionally, about half of these studies used off-the-shelf software such as Wii Fit or Kinect Adventures. Less than a quarter of the studies used custom-designed software like 3D Studio Max to produce a VR environment tailored to their intervention. This is consistent with previous reviews which have found that most VR interventions utilize hardware that has associated prefabricated environments, and typically these are off-the-shelf programs such as Wii Fit [6,8,162]. Moreover, this speaks to the criticism that research in this area grossly lacks theory-driven VR intervention development [6,7,8,9]. The majority of research is being conducted to determine the efficacy level of VR at large and, thus, uses low-cost or time-efficient ways to determine efficacy, such as using ready-to-run programs. This is favorable in a sense, as clinicians may be more likely to use VR technology and games that are easier to access. However, in the case of commercially available VR products, these technologies were created for entertainment, without a priori use of theory-based principles to address the desired motor learning principles, motor deficits, or components of exercise prescription and achieve the specific outcomes of an intervention.

As previously discussed, a limited number of studies provided a rationale for the development of their intervention, including study design, VR hardware and software usage, and intervention dosage. Interventions ranged from 1 to 12 weeks, with the two most common intervention lengths being 4 and 6 weeks. Additionally, intervention frequency and duration ranged from 5 to 60 sessions lasting between 15 and 60 min each. The most common intervention dosage used was a duration of 6 weeks, 3 visits per week, and 30 min for each session. This is consistent with a previous review which found that the most common VR intervention implemented 3 sessions per week for 20–40 min over 4–6 weeks [8]. Of the studies that cited evidence-based rationale for their VR intervention development, eleven of them provided support for either the frequency, intensity, or duration. However, only two articles cited evidence to support their intervention prescription that was specifically related to VR interventions [81,111]. The other nine articles cited support based on empirical evidence related to intervention prescriptions for specific patient populations such as multiple sclerosis patients [71], or recommendations for general physical activity prescriptions from national organizations such as the American College of Sports Medicine (ACSM) [92]. Thus, VR research seems to be applying a more systematic and theory-driven approach to the development of VR interventions; however, more work needs to be done to identify VR-specific recommendations regarding dosage. This is supported by other reviews of VR interventions to improve motor outcomes, as currently a wide range of intervention protocols are being employed, and not always with a priori justification specific to VR training [6,8,164].

VR can be defined as an artificial environment containing sensory information that can be interacted with, visualized, or manipulated to allow natural behaviors to emerge as if the environment were real [10,11,12]. In the case of VR usage for rehabilitation in the context of the articles included in this review, the natural behaviors elicited are the gross motor functions of gait, balance, and upright mobility. The variety of functionality and customization of VR makes it a promising tool to provide meaningful interventions to improve motor outcomes.

VR can take many forms and continues to evolve as technology advances. With the development of off-the-shelf, commercially available products, VR has been more widely researched for use in a variety of domains, including physical rehabilitation. As can be seen in the publications included in this review, the majority of studies utilize commercially available VR hardware and software. This is consistent with prior systematic reviews which also found that the majority of research is conducted using low-cost VR technologies. This is likely due to the low cost and easy setup being more conducive to use in clinical or home settings. The clinical implications from our scoping review are as follows: (1) VR has become more accessible for clinicians in terms of cost and usability; (2) VR has been integrated into several different settings for rehabilitation purposes, including inpatient/outpatient settings, community centers, and home environments; and (3) infusing VR into a rehabilitation plan has generally led to positive results.

It should be noted that some limitations of this review exist. First, this review was limited to the outcomes of gait, balance, and upright mobility. Other domains of gross motor function, such as upper extremity rehabilitation, and adjacent domains like pain management and cognitive function were not included in this review. Similarly, research that used an adjunctive modality such as transcranial magnetic stimulation, functional electric stimulation, or robot-assisted gait training was also not included in this review. These domains, as they relate to VR research, may be most appropriate to review separately as they have their own critical mass and may reveal interesting insights into the use of VR in these areas. Moreover, VR technology considerably changed, including the accessibility of the technology, across the included years of 2010–2023, which could have impacted the results. It should also be noted that we used the same search string for each database, which is not a best practice method. Each database has a unique coding system (e.g., PubMed uses MeSH terms, SCOPUS uses a TITLE-ABS-KEY format), so tailoring the search to each database’s coding system would allow a more optimized return of matching articles. Lastly, the articles were limited to three database searches and do not account for eligible studies that could be acquired through additional databases or grey literature. Despite these limitations, this review included a large sample of research (n = 126) covering a wide range of patient populations, research locations, and research settings, giving a diverse understanding of how VR is being researched for use as an intervention to improve gait, balance, and upright mobility.

5. Conclusions

In conclusion, 58.5% of the studies in this review were conducted outside of a research laboratory. This demonstrates that research in this area is moving beyond proof-of-concept and toward translation to clinical applications. Beyond demonstrating that VR can provide significant improvement in gait, balance, and upright mobility, this also indicates that VR has been implemented in clinical, community, and home environments. What still remains unclear is how this empirical evidence is being utilized by clinical providers. Future research may benefit from assessing the usability from the perspective of the clinical providers assisting with the data collection, as many studies reported that PTs supervised their VR interventions. Additionally, as studies find significant improvement in their desired outcomes, studies regarding the knowledge translation or sustainability of an intervention could be useful for implementing actual use of that VR intervention beyond the boundaries of the study protocol.