Home-based virtual reality telerehabilitation for people with Parkinson’s disease: a scoping review of delivery, supervision and implementation

Abstract

Objective

To map the delivery models, supervision approaches and implementation characteristics of home-based virtual reality (VR) telerehabilitation for people with Parkinson’s disease (PD).

Design

This scoping review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews checklist.

Data sources

PubMed, Scopus, CENTRAL and PEDro were searched from inception to 10 January 2026. Searches were restricted to English-language publications.

Eligibility criteria

Studies evaluating immersive or non-immersive VR-based telerehabilitation delivered at home for adults with clinician-diagnosed PD, with a remote support component (synchronous, asynchronous or structured remote monitoring). Observational, qualitative and review designs were excluded.

Data extraction and synthesis

Two reviewers independently screened records, selected studies and charted data on delivery model, technology, onboarding, supervision, monitoring, dose and feasibility/safety. Results were synthesised using descriptive numerical summary and narrative analysis.

Results

18 studies (745 participants) met the inclusion criteria. 16 studies (89%) used fully home-based delivery; structured clinic-to-home transitions were rare. Technologies ranged from commercial gaming consoles (Nintendo Wii, Microsoft Kinect) to custom-built augmented reality, tablet-based and sensor-integrated platforms, with no dominant system. Asynchronous supervision, mainly through scheduled telephone calls, was used in most studies (50%), while only two (11%) employed real-time synchronous supervision. Dose prescriptions varied widely: programme durations spanned 2 weeks to 6 months, session frequencies ranged from two to five times per week and session lengths ranged from approximately 15 to 50 min. Adherence was generally high across studies; however, adverse event monitoring was inconsistently reported. Most evidence came from Europe and Oceania, with limited representation from other world regions.

Conclusion

Home-based VR telerehabilitation for PD is predominantly delivered through fully home-based, asynchronously supervised models, with considerable variability in technologies, dose prescriptions and safety reporting. Standardised dose reporting, structured adverse event monitoring, evidence-based supervision guidelines and geographically diverse trials are needed to support translation into routine clinical care.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist and conducted following established scoping review methodology.

• Two reviewers independently performed screening, study selection and data extraction, reducing the risk of selection and extraction bias.

• Consistent with scoping review methodology, formal critical appraisal or risk-of-bias assessment of the included studies was not conducted.

• The search was restricted to English-language publications and four databases, which may have excluded relevant studies published in other languages or indexed elsewhere.

Introduction

Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterised by the loss of dopaminergic neurons within the basal ganglia, leading to a combination of motor and non-motor impairments.¹ It is the second most common neurodegenerative condition worldwide, and its prevalence continues to rise in parallel with global population ageing.^{2 3} As the number of individuals living with PD increases, the demand for sustainable and accessible long-term rehabilitation strategies becomes increasingly urgent.⁴

The motor manifestations of PD include bradykinesia, rigidity, tremor and postural instability. These symptoms are frequently accompanied by gait disturbance, impaired balance and an elevated risk of falls.^{5 6} These impairments contribute to reduced functional independence and diminished quality of life.⁷ Current clinical guidelines recommend early initiation and ongoing physiotherapy, with exercise programmes targeting mobility, balance and functional performance.⁸ However, the progressive course of PD requires sustained engagement in rehabilitation, and consistent in-person attendance is often hindered by transportation barriers, symptom variability, caregiver burden and financial constraints.⁹ People with PD have also reported being unaware of available allied health services or facing prolonged delays in accessing appropriate care.¹⁰ Taken together, these access barriers highlight the need for delivery models that can extend rehabilitation beyond traditional clinic-based settings.

Telerehabilitation, broadly defined as the remote delivery of rehabilitation services through communication technologies, has emerged as one approach to address limitations in access and continuity of care.¹¹ Although its conceptual foundations predate the COVID-19 pandemic, the rapid expansion of digital health during this period accelerated its adoption across neurological populations.¹² In PD, telerehabilitation has been reported to be feasible and generally safe, with emerging evidence suggesting it may support the maintenance of motor function.^{13 14}

Parallel to these developments, virtual reality (VR) has gained attention as an interactive modality capable of delivering task-oriented, feedback-driven exercise in engaging digital environments.¹⁵ VR-based telerehabilitation combines remote access with immersive or non-immersive platforms that provide real-time feedback and allow progression to be adjusted according to patient performance.¹⁶ Advances in both commercial systems and custom-designed applications have further increased the feasibility of deploying VR technologies within patients’ homes.^{17 18} Nevertheless, translating technological potential into structured, clinically implementable home programmes requires careful consideration of delivery processes. For people with PD specifically, VR-based exercise can provide external visual and auditory cueing, augmented sensory feedback and structured repetitive task practice features that align with established neurorehabilitation principles for addressing gait freezing, impaired balance and reduced movement amplitude.^{15 18} The capacity of VR platforms to deliver these PD relevant therapeutic elements within a motivating, gamified environment makes them a particularly suitable modality for sustained home-based rehabilitation.

A preliminary examination of existing syntheses indicates that studies relevant to home-based VR telerehabilitation in PD are available, yet the manner in which these programmes are operationalised in practice remains insufficiently described. Truijen et al reviewed home-based VR training and telerehabilitation for balance across PD, multiple sclerosis and stroke, but focused primarily on balance-related outcomes and did not differentiate the delivery architecture of interventions within PD specifically.¹⁹ Gallou-Guyot et al examined home-based active video game rehabilitation in PD and, while reporting promising effects on gait and balance, highlighted key uncertainties that remain unresolved: what constitutes an optimal training dose, how supervision should be structured and whether benefits are retained over time.²⁰ More recently, Norouzi-Ghazbi et al reviewed gamification-based telerehabilitation in PD with emphasis on motivational and game design features, but did not systematically differentiate supervision models, delivery structures or implementation processes across home-based programmes.²¹ Taken together, although the primary studies partially overlap across these reviews and the present review, a critical gap remains: none has specifically examined how these programmes are operationally delivered including supervision structures, onboarding processes, dose prescriptions and safety monitoring which are the very elements that determine whether an intervention can be implemented in routine clinical care for this population.

Collectively, these reviews suggest that although home-oriented VR and exergaming interventions have been studied, a PD-specific synthesis of operational delivery characteristics is lacking. Key elements such as whether programmes are fully home-based or hybrid, how supervision is structured (synchronous, asynchronous or minimal), what technologies and platforms are required, how onboarding and home setup are conducted and how monitoring, progression and safety oversight are organised are central to real-world implementation but remain inconsistently reported. From an implementation perspective, understanding these characteristics is essential to inform scalability and translational readiness.

Given this objective, a scoping review methodology was selected to map the breadth and nature of delivery characteristics rather than to evaluate intervention effectiveness, given the anticipated heterogeneity of interventions and the absence of a single quantifiable outcome suitable for meta-analytic synthesis. In this review, a reporting gap refers to a delivery characteristic that was either not reported or not described in enough detail to allow replication. This review aimed to examine how home-based VR telerehabilitation is delivered to people with PD by addressing the following questions:

• What delivery models, supervision approaches and technologies are used in home-based VR telerehabilitation for PD, and how are onboarding, home setup, monitoring, feedback and progression structured?

• How are intervention dose parameters (programme duration, session frequency and session length) reported across studies?

• How are feasibility and safety outcomes reported, and what methodological gaps may influence translational readiness?

Methods

Study design

Using a pre-specified search strategy, we systematically searched the literature to identify studies on home-based VR telerehabilitation for individuals with PD. This scoping review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) checklist.²² A formal protocol was not registered for this scoping review. The review was conducted according to a pre-specified search strategy, with all methodological decisions documented prior to data extraction, and reported in accordance with the PRISMA-ScR checklist²² and established scoping review methodology.²³

Operational definitions

Telerehabilitation refers to the delivery of rehabilitation services remotely using information and communication technologies.^{10 23} VR refers to a technology that generates simulated interactive environments, enabling users to engage with digital content in real time through visual, auditory or haptic modalities^{24 25}. VR systems range along a continuum of immersion, from non-immersive platforms (eg, screen-based exergaming systems such as Nintendo Wii or Microsoft Kinect) to fully immersive head-mounted devices.^{15 18} Therefore, VR-based telerehabilitation combines these two concepts: the delivery of VR interventions, including exergaming that incorporates a simulated interactive environment, in a home or remote setting, supported by a remote component.²⁶ For the purposes of this review, delivery models were categorised as fully home-based (all sessions conducted at home), hybrid (combining clinic-based and home-based phases) or mixed-setting (comparing home and clinic delivery). The terms ‘exergaming’ and ‘exergames’ are used descriptively throughout, as employed by the original study authors, but all included interventions met the criterion of incorporating a simulated interactive environment. Given that this review focused on delivery characteristics, studies were not restricted by specific outcome measures; however, when reported, clinical outcomes commonly used in Parkinson’s rehabilitation, feasibility indicators (adherence, usability) and safety data (adverse events) were charted to contextualise implementation.

Search strategy and study inclusion

Two reviewers (IN and VV) independently conducted a comprehensive search across four electronic databases, PubMed (National Library of Medicine), Scopus (Elsevier), Cochrane CENTRAL (Cochrane Library) and PEDro (Physiotherapy Evidence Database) from database inception to 10 January 2026. All four database searches were conducted on the same date. Searches were restricted to English-language publications only. Search terms combined concepts related to home-based rehabilitation/telerehabilitation, VR/exergaming and PD, using controlled vocabulary (eg, MeSH) and free-text terms as appropriate for each database. The full search strategies are provided in the online supplemental appendix 2. The search strategy was developed iteratively through preliminary searches in PubMed to identify relevant MeSH terms and refine the combination of free-text keywords prior to conducting the final search across all four databases. In addition, reference lists of all included studies were hand-searched to identify any further eligible records; no additional studies were identified through this process.

All retrieved records were exported to Zotero (V.7)²⁷ for data management and duplicate removal. Titles and abstracts were screened independently by two reviewers (IN and VV) against the eligibility criteria. Records identified as potentially relevant by either reviewer proceeded to full-text review, which was also conducted independently by the same two reviewers. Disagreements at either stage were resolved through discussion with the third reviewer (NK).

Studies were included if they (1) were published in English, (2) involved individuals with clinician-diagnosed idiopathic PD and (3) evaluated immersive or non-immersive VR-based telerehabilitation delivered in a home or remote setting and supported by a remote component. The remote component can be provided synchronously (eg, real-time videoconferencing), asynchronously (eg, telephone follow-up, data upload, or platform-based performance review) or through structured remote monitoring mechanisms such as clinician-reviewed training logs or diaries.

Studies were excluded if they (1) did not involve VR-based telerehabilitation; (2) reported only non-motor outcomes; (3) used observational or qualitative designs, reviews or meta-analyses; or (4) were incomplete or preliminary reports.

Data extraction

Two authors (IN and VV) independently extracted and charted relevant data into predefined categories. The extracted information included study details (author, year, country and study design), participant characteristics (sample size, sex and PD stage/severity where reported) and intervention delivery characteristics relevant to home-based VR telerehabilitation (delivery model/type, technology/platform used at home, onboarding and home requirements, remote support/supervision, monitoring and feedback/progression, and prescribed dose). The comparator details and outcomes assessed (including assessment time points) were also recorded, along with feasibility and safety reporting and any study limitations noted by the authors.

Consistent with scoping review methodology,^{22 23} formal critical appraisal or risk of bias assessment of the included studies was not conducted, as the primary aim was to map the breadth of available evidence rather than assess its quality.

Collating, summarising and reporting the findings

This stage followed established scoping review guidance and comprised three components: (1) descriptive analysis of the extracted data (2) structured presentation of the results and (3) narrative synthesis.²⁸ First, a descriptive numerical summary was undertaken to characterise the included studies, including the study design, participant characteristics and key intervention features. Dose parameters (programme duration, session frequency and session length) were summarised using simple minimum–maximum ranges taken directly from the values reported in the included studies, without any pre-defined grouping. Second, the extracted data relating to intervention delivery were organised and categorised using descriptive qualitative techniques to group similar characteristics and identify recurring patterns in the implementation of home-based VR telerehabilitation for individuals with PD. Finally, a narrative synthesis was conducted to examine the mapped findings, identify gaps in reporting and highlight areas requiring further methodological clarification. This involved grouping findings according to the predefined delivery characteristic categories, examining patterns within and across studies and considering the implications of intervention heterogeneity for the overall synthesis, informed by established guidance on the conduct of narrative synthesis.²⁹

Results

Search results and articles selection

A total of 335 records were identified through database searching (PubMed=64; Scopus=85; CENTRAL=122; PEDro=64). After the removal of 47 duplicate records, 288 records remained for title and abstract screening. Of these, 223 records were excluded based on the predefined eligibility criteria. 65 full-text articles were assessed for eligibility. Following full-text review, 47 studies were excluded for reasons detailed in figure 1. 18 studies met the inclusion criteria and were included in the final scoping review. The study selection process is illustrated in figure 1.

Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram showing the identification, screening, eligibility and inclusion process of studies in this scoping review. VR, virtual reality.

Characteristics of source articles

Table 1 summarises the key characteristics of the 18 included studies, which together enrolled 745 participants (sample sizes ranging from 2 to 130). The evidence base was predominantly European (12/18, 67%), with half of the studies (9/18, 50%) published from 2020 onwards, reflecting recent growth in this field. Study designs were mainly randomised controlled trials (10/18, 56%). Most participants had mild-to-moderate PD (Hoehn and Yahr stages 1–3); the H&Y (Hoehn and Yahr) stage was reported in 13 of 18 studies (72%) and not reported in five (28%). Full study-level details are provided in online supplemental table S1).

Table 1. Summary characteristics of included studies (n=18).

Characteristic	Category	n (%)
Study design	Randomised controlled trial	10 (56)
	Quasi-experimental/non-randomised	6 (33)
	Feasibility study	1 (6)
	Case report	1 (6)
Geographical region	Europe	12 (67)
	Oceania	3 (17)
	North America	2 (11)
	Middle East	1 (6)
Publication period	2012–2019	9 (50)
	2020–2024	9 (50)
Sample size	<20	2 (11)
	20–49	11 (61)
	50–99	4 (22)
	≥100	1 (6)
Disease severity (H&Y stage)	Reported (mild-to-moderate)	13 (72)
	Not reported	5 (28)
Total participants		745

H&Y, Hoehn and Yahr.

Synthesis of the results

Delivery models of home-based virtual reality telerehabilitation

16 of the 18 studies delivered the intervention primarily at home, with the home serving as the main site of training rather than as an extension of clinic-based care.^{16 30–44} Designs that bridged the two settings were uncommon: one used a two-phase model in which an in-clinic VR phase preceded home-based telerehabilitation,⁴⁵ one compared home-based delivery directly against outpatient hospital exergaming⁴⁶ and one paired otherwise unsupervised home exergaming with two supervised laboratory sessions.³⁹ In the latter study,³⁸ home was the primary and continuous site of training, with the two laboratory sessions serving only as intermittent adjuncts; it was therefore classified as home-based according to its main delivery setting. The delivery models were therefore predominantly home-based, with little use of structured clinic-to-home transitions or mixed-setting comparisons. This likely reflects a deliberate preference among research groups for fully home-based formats, which sidestep the clinic attendance that is often difficult for a population facing mobility and transportation barriers, though it also means that the graded, supervised-to-independent pathways that might suit a progressive disease remain largely unexplored.

Technologies and platforms used at home

The technologies used across the 18 studies were highly heterogeneous. Commercial Nintendo Wii-based systems were used in four studies,^{16 32 42 43} Kinect-based motion-tracking systems in five^{31 33–35 45} and custom-developed or proprietary platforms in seven,^{30 36 37 39–41 46} including augmented reality glasses, Unity3D-based exergames, step-mat systems, tablet-based VR systems and integrated virtual coach platforms. One study used a treadmill–VR system with motion-capture cameras and remote desktop connectivity,⁴⁴ and another employed VR-enhanced home cycling with heart rate monitoring.³⁵ Detailed technology specifications for each study are provided in online supplemental material table S2.

Onboarding and home setup requirements

Onboarding and home setup procedures were reported in 17 of the 18 included studies, with one study not describing these requirements.³⁶ In-home setup support was the most common approach, typically involving a home visit for device installation and hands-on demonstration,^{30 32 38–44} sometimes supplemented by written materials such as booklets or logbooks.^{32 39 40 43} Six studies instead, or additionally, began with a clinic-based or institute-based familiarisation phase before moving the patient home.^{16 31 33 34 45 46} Several studies also specified home requirements or safety considerations, including caregiver involvement,¹⁶ adequate space and medical clearance,⁴¹ internet connectivity and compatible equipment,³⁶ device positioning for motion capture^{34 35} and safety controls for treadmill training.⁴⁴ Notably, onboarding was rarely self-directed: in most studies, it required a clinician or researcher to attend the home or to conduct an initial clinic visit. This reliance on in-person support has implications for the scalability of home-based delivery and for the resources required to implement it in routine care.

Remote support and supervision approaches

The remote supervision models varied across the included studies. Asynchronous home training with scheduled telephone contact was the most commonly reported approach (9/18, 50%), with call frequency ranging from weekly^{30 32 42 43} to fortnightly,^{38 40 41} and some studies incorporating technical support via help desks or virtual coaching platforms.^{36 45} One study used a threshold-triggered contact model, contacting participants only when adherence fell below a prescribed level.³⁹ Technology-enabled asynchronous monitoring without live supervision was also described, using remote interfaces for post-session data review and therapy planning.^{33 46} A further three studies involved essentially independent home training, one offering optional telephone contact³⁴ and two with no live supervision or telemonitoring at all.^{31 35}

Synchronous real-time supervision was reported in only two studies (2/18, 11%): one delivering real-time audio–visual supervision with remote control of training parameters,⁴⁴ and the other using live videoconferencing for entire sessions.¹⁶ One additional study described a hybrid model combining asynchronous home training with a weekly synchronous physiotherapist session.³⁷ The predominance of asynchronous approaches may reflect practical constraints around clinician time and internet reliability, but it also means that most participants in this evidence base trained without real-time clinical oversight, a point of relevance given the fall risk and symptom variability in PD.

Monitoring, feedback and progression

All 18 studies reported some form of monitoring, feedback and/or progression strategy. Objective monitoring using electronic logs, platform records or uploaded performance data was described in multiple studies,^{30 33 36–39 42 45 46} while participant-completed logs or diaries were used in others.^{16 31 40 41 43} Real-time feedback during gameplay, including visual and auditory cues and end-of-session performance summaries, was widely reported.^{32 34 35 40} Progression, by contrast, was handled far less consistently, ranging from automatic difficulty adaptation based on performance,^{31 33} through structured therapist-led protocols^{16 44} and parameter-based software adjustments,³⁷ to predefined difficulty levels set by the clinician but subsequently advanced by the patient.^{34 39 41} The inconsistency is greatest for progression, which is also the element most closely tied to safety and training effect, making it difficult to determine which approach best balances benefit against risk in unsupervised home-based PD rehabilitation.

Dose and programme structure

The prescribed dose varied substantially across the 18 studies. Programme duration ranged from 2 to 3 weeks^{33 46} to 6 months,³⁸ with most studies prescribing six to twelve weeks of training.^{16 30–32 34 35 37 39–43} Longer programmes of 4 months³⁶ and a two-phase design totalling 4 months⁴⁵ were also reported. The most commonly prescribed frequency was three sessions per week,^{16 30 37 39–42} though higher frequencies of up to five sessions per week^{30 45} and twice-daily sessions³² were also described. Session length was most commonly approximately 30 min,^{30 33 36 42} with several studies prescribing 45–50 min^{16 30 36 44} and some staging session duration upward across the programme.³⁴ No dose recurred often enough to represent even an emerging convention.

Feasibility and safety reporting

Feasibility was reported across the included studies primarily as a descriptor of delivery acceptability rather than as evidence of clinical effectiveness. It was primarily reported through adherence, retention, usability and satisfaction measures. Adherence rates were generally high across studies, with most reporting rates exceeding 80%.^{30 36 41 45} Completion rates were similarly favourable^{33 37 41 44}. Usability and satisfaction were assessed using standardised instruments such as the User Experience Questionnaire³⁰ and System Usability Scale,^{36 45} and participants frequently reported enjoyment and comfort during training.^{32 41} However, these indicators reflect the acceptability of the delivery model rather than clinical benefit and should not be interpreted as evidence of intervention effectiveness.

Safety reporting varied considerably. Several studies reported no serious adverse events,^{16 30 34 35 41} while minor events such as dizziness or headache,³⁰ non-injurious falls⁴⁰ and exacerbation of pre-existing pain^{40 41} were documented in others. One study reported potentially related adverse events in 11% of the intervention group compared with 6% of controls.³⁸ Notably, several studies did not formally report adverse event monitoring,^{31 37 46} representing a significant gap. Technical feasibility issues including difficulty running applications,⁴⁶ the need for setup assistance³³ and connectivity or motivational barriers^{16 36} were also reported. The inconsistency in safety reporting limits the ability to draw conclusions about the risk profile of home-based VR telerehabilitation across different technologies and supervision models. Detailed feasibility and safety data for each study are provided in online supplemental material table S2.

Discussion

Principal findings

This scoping review mapped the delivery models, supervision approaches and implementation characteristics of home-based VR telerehabilitation for people with PD across 18 studies involving 745 participants. Delivery was predominantly fully home-based and asynchronously supervised, most often through scheduled telephone calls, with only two studies using real-time synchronous supervision. Technologies were highly heterogeneous, dose prescriptions varied widely, and while adherence and satisfaction were generally favourable where reported, safety reporting was inconsistent. The evidence base was concentrated in Europe and Oceania and largely reflected people with mild-to-moderate disease.

High technological heterogeneity was observed, ranging from commercial gaming platforms such as the Nintendo Wii and Microsoft Kinect to augmented reality, tablet-based and sensor-integrated systems, with no dominant platform. This complicates cross-study comparison and hinders the development of standardised implementation guidance. Just as importantly, studies varied considerably in how they described the technologies and procedures involved, which makes it difficult for a clinician or researcher wishing to replicate or adapt a programme to know exactly what was delivered and how. Until technologies and delivery processes are described more consistently, the transferability of findings across settings and populations will remain limited.

A related concern is the sustainability of the platforms themselves. Several of the most frequently used systems, notably the Nintendo Wii and Microsoft Kinect, have since been commercially discontinued. This raises questions about the reproducibility of these interventions and the long-term viability of delivery models built around specific consumer hardware. Future programmes may need to prioritise platform-agnostic or software-adaptable approaches that are less tied to the commercial lifecycle of any one device, so that evidence-based interventions remain deliverable as technologies evolve.

Most studies relied on asynchronous supervision, typically scheduled telephone calls, with only two using real-time synchronous supervision throughout sessions. This may reflect practical considerations: synchronous models demand dedicated clinician time, reliable connectivity and sometimes more complex equipment, which may not have been feasible or prioritised. It may also be that many groups were still working out what is viable in home settings rather than choosing asynchronous delivery on principle, though since no study reported its reasons for selecting a supervision model, this remains speculative. What the variation makes clear is that there is no consensus on the level of remote oversight needed to keep patients safe and engaged. This is particularly relevant in PD, where fall risk and symptom fluctuation mean the consequences of inadequate supervision during unsupervised exercise may be more serious than in other populations.^{5 6} The concern deepens with disease progression: as postural instability and cognitive impairment worsen, supervision needs are likely to rise, yet almost all included studies recruited participants with mild-to-moderate disease (H&Y 1–3), a group probably best suited to independent home training. Whether these supervision models would be adequate for people with more advanced disease is unknown and warrants investigation.

Prescribed dose varied on every parameter, with no commonly adopted regimen, and it was not always reported in a consistent format, some studies giving total session counts, others weekly frequency and duration separately, which further limited comparison. This matters because dose–response evidence has begun to identify the VR training parameters associated with better balance and gait outcomes in PD,⁴⁷ with higher session volumes possibly producing larger effects.⁴⁸ For context, Wu and Zhang⁴⁷ identified optimal parameters for balance as sessions of 20 min or less, four to seven times per week over 4–7 weeks, while Hernandez-Martinez et al⁴⁸ reported significant balance gains with programmes exceeding 8 weeks at more than three sessions per week, sessions under 45 min, and more than 20 sessions in total. Most studies in the present review prescribed 30–50-minute sessions two to three times per week, a lower frequency than either meta-analysis suggests is optimal. A precise count of studies meeting, partially meeting, or falling outside these optimal parameters was not attempted, because dose was reported inconsistently across studies, with some specifying total session counts, others weekly frequency and session duration separately, and others weekly minutes. These differences in reporting prevented a direct and reliable comparison of individual studies against those parameters. These parameters were, however, derived largely from clinic-based or supervised VR training, and whether they translate to home-delivered programmes, where session structure and supervision differ, remains to be established, an important question given that the progressive course of PD requires sustained exercise to maintain motor function.⁸

Finally, adverse event monitoring was the clearest reporting gap. Only one trial quantified adverse events systematically, reporting potentially related events in 11% of the intervention group versus 6% of controls; many studies described safety loosely, and several did not report adverse-event monitoring at all.^{31 37 46} Without consistent safety data, it is not possible to tell whether the generally favourable feasibility picture reflects genuinely low risk or simply events that were never captured, and clinicians weighing these programmes are left with little safety evidence to guide them. For a population defined in part by fall risk and fluctuating motor control, that is a significant limitation rather than a minor omission.

Implications for practice and future research

Clinically, the predominance of asynchronous supervision suggests that home-based VR telerehabilitation has been feasibly delivered without continuous real-time clinician involvement, as reflected in the generally high adherence rates reported across studies.^{30 36 41 45} However, as no evidence-based minimum supervision threshold has been established, oversight should currently be individualised according to disease severity, fall risk and technological proficiency. The consistent emphasis on structured onboarding across studies^{16 30 32 38–44} indicates that successful home implementation requires more than equipment provision; adequate setup, technical support and caregiver orientation should be built into service planning.

It should be noted, however, that the included studies were conducted in research settings. Translation into routine clinical services will depend on implementation determinants beyond the delivery characteristics mapped here, including funding and reimbursement structures, clinician workload, integration within existing rehabilitation pathways and regulatory considerations around digital health technologies. The present findings should therefore be read as mapping what has been feasibly delivered in research, rather than what is immediately scalable to clinical practice.

Several priorities for future research follow. Dose parameters remain inconsistently reported; future trials should describe programme duration, session frequency and session length transparently, and align reporting with emerging dose–response evidence.^{47 48} Structured adverse event monitoring is needed in home settings to reduce under-reporting. Head-to-head studies comparing synchronous and asynchronous supervision would help build the supervision guidelines that are currently lacking. Broader geographical representation, particularly from regions outside Europe and Oceania, is needed to improve generalisability. Finally, adequately powered multicentre trials with longer follow-up are required to move from feasibility evidence towards health-system adoption.

Although this review did not formally categorise the type of rehabilitation each intervention targeted, the outcomes assessed suggest that balance-focused and gait-focused interventions predominated, with fewer addressing upper-limb function or combining motor and cognitive training. Future reviews could usefully examine which delivery and supervision models best suit specific rehabilitation goals in PD, to inform more targeted implementation guidance. To address the inconsistent intervention reporting noted above, we further recommend that future studies adopt a reporting framework adapted from the Template for Intervention Description and Replication⁴⁹ and tailored to home-based VR telerehabilitation, specifying at minimum the programme duration, session frequency, session length, supervision model, technology platform, onboarding procedures and adverse-event monitoring protocol.

Strength and limitations

This review has several strengths. It was reported in accordance with the PRISMA-ScR checklist²² and conducted following established scoping review methodology.²³ Two reviewers independently performed screening, study selection and data extraction, reducing the risk of selection bias. The review addresses a clearly defined gap by focusing specifically on how home-based VR telerehabilitation is delivered for people with PD, mapping operational delivery characteristics that prior syntheses^19–21 did not systematically examine.

Several limitations should be acknowledged. Consistent with scoping review methodology,²³ formal critical appraisal or risk of bias assessment was not conducted; the included studies encompassed randomised controlled trials, quasi-experimental designs, feasibility studies and a case report, and the strength of evidence supporting individual findings was therefore not evaluated. The search was restricted to English-language publications and four databases (PubMed, Scopus, CENTRAL and PEDro). These were chosen for broad biomedical, trial-specific and specialist physiotherapy coverage, but including further databases may have identified additional records. The heterogeneity of intervention descriptions also meant that some delivery characteristics were incompletely reported, limiting the depth of mapping in certain domains. Finally, the predominantly European and Oceanian evidence base may not reflect delivery approaches used elsewhere, and this concentration may itself bias the picture of supervision and technology: studies from these regions benefit from relatively advanced digital infrastructure, stable connectivity and greater consumer access to VR devices, factors that could shape both the supervision models adopted and the feasibility of particular platforms in other settings.

Conclusion

This scoping review found that home-based VR telerehabilitation for PD is predominantly delivered through fully home-based models using asynchronous supervision, with considerable variability in technologies, dose prescriptions and safety reporting. The findings highlight the need for standardised dose reporting, structured adverse event monitoring, evidence-based supervision guidelines and geographically diverse trials to support translation into routine clinical care.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.