This review analyzed the effects of Virtual Reality training as a rehabilitation intervention for stroke and Parkinson’s Disease patients. These two common neurological disorders require continuous management in order to maintain functional ability and to improve quality of life through the interventions, targeting multiple outcomes [38, 39]. All studies in this review proposed the VR training to be as effective compared to conventional training. It was found to be advantageous for improvement of the functional abilities in neurological patients, namely upper extremity functional mobility, balance, gait, as well as cognitive, psychoemotional aspects and quality of life. Based on these findings, VR might be suggested for an inclusion in a neurorehabilitation program as a beneficial addition to conventional therapy in patients with mild to moderate conditions. However the treatment effect in most of the trials was only short-term. The findings of trials that included a follow-up assessment [20,21,22, 25, 28, 35, 37] suggest that effect could be maintained if patients continued training for a longer time (i.e. longer than four to eight weeks, which was the most common duration), taking into account the adaptation to new intervention, technology acceptance and motor learning. Based on the characterstics of included trials, no conclusions could be drawn on the optimal frequency of the training or the session duration, suggesting that the dosage and frequency should be tailored to a specific patient’s condition and capacity, with a potential progression. Since VR is emerging within the telerehabilitation field as an activity performed at home, gradual implementation of it to the patient’s home environment may result in continuous adherence of training and potential long-term effects on functional outcomes [28].

The decision to include a broader scope of study population was reasoned by the aim to explore the consistency of findings (and as a result have better generalizability) across different types of participants facing similar functional deterioration over time [18]. In some way it may counterbalance the fact that nearly one third of included RCTs had small sample sizes and lacked generalizability in their conclusions. Despite the rapid growth of telerehabilitation and Virtual Reality research fields, there is still a relatively small amount of quality evidence for PD population. Previously published VR trials and reviews in PD compare the findings to other diagnoses, due to insufficient data within the same population. This review found some consistency in outcomes and test choice among trials in both populations (eg. BBS, TUG were the most common tests among both subgroups) and therefore it is believed that common conclusions could be drawn.

Another important distinction to be discussed, is the Virtual Reality approach – exergames and serious games. Exergames are the commercially available games designed primarily for general healthy population and accessible for anyone having the equipment at home or in clinic. Due to a wide variety of scenarios and intensity levels it can be successfully applied in the home-based telerehabilitation of neurological patients [40]. Serious games however, are specifically designed for rehabilitative purposes for a particular limitation, are less entertaining and more task-oriented. The potential limitation of exergames is that their complex interfaces may not be compatible with postural or mobility constraints, or can be cognitively challenging for a person with a disability [41]. Confusion and disappointment of being unable to use the system can result in non-compliancy. Flexibility and task-orientation of the virtual reality games is therefore imperative, based on a qualitative patient experience review of Lewis and Rosie [42]. It was found that adult neurological patients are seeking more rehabilitative benefits of therapeutically principled design, rather than just playing games [43, 44]. On the other hand, exergames can be perceived as a more enjoyable, fun activity that is distracting from daily problems, since it does not distinguish the patient with special needs from otherwise healthy users. It is also easier to obtain cost- and availability-wise, to vary games and to use it together with caregiver or family members for some social interaction [28, 44].

The type of VR was not a determinant factor for the study inclusion, however interesting tendencies were observed between the subcategories. Coincidentally, the distribution of both approaches among the stroke subcategory trials was even. The VR research in stroke is the largest among other neurological populations, and it might explain the wide variety of approaches investigated. In the PD subcategory, the exergaming as an experimental intervention was slightly more common than serious games (five vs three trials).

It might be assumed that serious games are more valuable for a targeted rehabilitation due to the specifically designed task-oriented scenarios for the patients with limited capacity. On the other hand, the preferrence for serious games might be disadvantageous due to higher costs associated with the design and implementation, as well as the limited availability.

In order to decide whether the intervention is usable for the particular population, three points of interest need to be analyzed – its effectiveness, efficiency and satisfaction, according to the ISO definition of usability of the product or interactive system [16]. An important point to mention, is that the terminology (many synonyms used) and definitions (wide use in relevant aspects/ context) within this complex concept are all interrelated and interdependent and should be analyzed together as a whole. Some aspects that are more commonly addressed or can be quantitatively assessed and evaluated such as response to treatment, satisfaction, adherence and cost-effectiveness are emphasized more in studies. It can be used as a basis for deeper analysis of other aspects that are less commonly addressed or could not be quantitatively evaluated.

Based on the results of the systematically reviewed trials, some conclusions could be drawn on the effectiveness of the intervention (due to the connection between the two aims of this review) and patient satisfaction (possible to evaluate). However the topic of efficiency is left open, which was thought to be particularly important to discuss from a clinical point of view.

The definition of “Efficiency” (see Methods section) includes the factor of human efforts for productivity. Some accompanying deficits and practical barriers resulting in increased human effort and subsequent decreased productivity might affect the successful introduction of this potentially beneficial rehabilitation tool into a neurological patient’s daily routine. Therefore the aim of this review was not only to analyze the effectiveness of VR on functional outcomes, but also to explore whether this innovative intervention is usable for the mentioned neurological populations. There was no separate search query on the aspects of usability due to a limited amount of relevant evidence, therefore the discussion is based on the characteristics of the reviewed trials with a support of the literature.

The positive response to treatment and perceived enjoyment of the VR therapy, leading to the improved quality of life was mentioned in a majority (10 out of 18) of included studies. Several trials that particularly assessed user satisfaction, concluded the significant improvement in mood, motivation and psychoemotional state in favor of VR [22, 28, 36]. Other authors based their conclusion on the maintainability (adherence and low drop-out rates, positive verbal patients’experiences and interest in continuation with VR) after the experiment [24, 25, 29, 30, 33,34,35]. These conclusions are in line with a number of previous reviews covering the topic of VR user satisfaction [3, 4, 40].

A common problem in conventional physical training is the lack of motivation [3, 5]. However it is vital to keep a patient motivated during the long-term rehabilitation to optimize the training outcome, prevent frustration and habituation [45]. Interesting suggestions were made by Perez-Marcos et al. [45], that neurorehabilitation programs should be inspired by Seligman’s „PERMA “theory (i.e. positive emotions, engagement, relationships, meaning, achievement) [46] and create an experience of the patient being comfortabilly challenged and engaged by the task which renders high levels of enjoyment.

According to Lewis and Rosie [42], the key components that make VR enjoyable for patients are the challenging environment, replication of real-life situations, sense of control and success. Engaging, entertaining activities should be incorporated in neurorehabilitation program for a patient to be willingly participating in their own recovery and be responsible for independent training at home, especially in case of life-long rehabilitation [45]. Worth to mention is that a majority of the included trials (15 out of 18) reported low or zero drop-out rates and great adherence, suggesting the positive maintainability and engagement in the intervention. Due to the fact that the effectiveness of both conventional training and VR was similar in a majority of trials, combining these two interventions might result in a better adherence to long-term plans of care [22, 24, 26].

Visual and hearing deficits are very common among the general aging population and can be influenced by a comorbid condition. During the review analysis, it was noted that in a majority of included trials (12 out of 18) the visual or hearing deficits were an exclusion criteria for study participants [21,22,23,24, 26,27,28, 30, 33,34,35,36]. While it might be apparent to not include patients with sensory deficits since it affects the ability to adequately follow instructions and overall performance during the VR training, it raises a concern for the selective accessibility and usability of such intervention. All mentioned studies pointed out this fact as a study limitation affecting the generalizability of the results. Congruent multisensory environments stimulating somatosensory system (eg. via proprioception) besides visuomotor coordination need to be proposed [47]. However, there is an on-going innovative development of VR specifically for visually-impaired persons, with enhanced auditory stimulation, or with haptic feedback through cane controller [48].

Another exclusion criteria in all 18 trials was a severe cognitive deficit that was either tested during the screening process using the assessment tools (MMSE, MoCA) or just mentioned as an eligibility criteria without a test in two studies [21, 24]. Unfortunately, cognitive decline is common among the studied population too, due to general aging and deteriorative changes in the brain due to a disorder [41]. In contrast with the visual deficit that cannot be altered with the use of Virtual Reality, the cognitive training however became a separate branch within the VR field [49, 50]. Traditional cognitive practices are often directed towards the isolated cognitive domains, such as executive functions, attention, visuospatial ability, memory and language. However cognition can be trained multidimensionally through the real-life situation scenarios simulated via Virtual Reality, potentially resulting in a sustained improvement of functional independence in activities of daily living. Thus it may be assumed that VR settings are more ecologically valid while there are no real consequences of failure [50].

The cognitive deficit can negatively impact the recovery of functional ability and quality of life. Therefore, to increase the therapy efficiency, the cognitive training should be incorporated into the motor VR training [45]. The functional training in virtual environment would have an influence on cognition through multitasking, dynamic feedback processing and progressive learning. In this review, some included trials introduced a dual-task activity in the experimental group [20, 22, 28, 30] or assessed the indirect effect of VR on cognition [29, 34]. Dual-task paradigms are a powerful way to evaluate the capacity of divided attention on individual task (cognitive task while performing motor activity) [51]. For example, dual-tasking is associated with gait impairments and results in freezing of gait, decreased stride length and symmetry loss among the general PD population. It does not improve with the dopamine replacement therapy [51], therefore should be addressed by combined physical and cognitive interventions. Serious games using principles of motor learning and neuroplasticity can optimize recovery after brain damage, such as stroke [41].

Technology acceptance has been a topic of attention since the emergence of telemedicine systems and the introduction of the telerehabilitation [8, 52, 53]. Regardless of the benefits and effectiveness of a certain telerehabilitation tool (eg. VR), its feasibility depends on whether the potential users have sufficient access and skills to use the digital technology [10, 54]. It is assumed that acceptance and understanding of VR software and hardware is greater among younger patients, as they might be more familiar with the use of digital technologies [52, 55]. Persons that acquire the disability such as stroke or PD in a relatively older age, might face some challenges with the implementation of digital tehcnologies in their daily life [56, 57]. Decreased learning capacity is augmented by the resistance to change, low level of confidence, lack of skills for using the technologies, and different expectations based on traditional rehabilitation approaches [42, 54, 56]. Sometimes an unsuccessful implementation of VR can be explained by the negative preconceptions of the older person. According to Laver [58], some patients preferred the conventional therapy due to a social interaction with staff, or biased beliefs that VR is too childish and provides only entertainment. However, some publications had positive conclusions on the coping strategies for adoption of digital technologies among older patients regardless their prior level of technological familiarity [53, 54].

The concerns for a technological acceptance described in reviewed research are mostly relevant for the current decade when it was published (i.e.2010–2020), the period of rapid growth of the telerehabilitation field and the introduction of innovative solutions. However it is expected to change dramatically over the next 10–20 years. The percentages of active users are likely to increase together with further development of digital services. The implementation of e-health technologies and its awareness is increasing, subsequently resulting in more technologically literate people among the retired [59]. Current 40–60 years old active users will eventually become the potential users of telemedicine tools, during some transition stage of their life [