Immersive Virtual Reality as Physical and Cognitive Therapy in Acquired Brain Injury: TEVI-DCA Program

Abstract

Acquired brain injury (ABI) is one of the leading causes of disability worldwide. Immersive virtual reality (IVR) is an emerging tool in the field of neurological rehabilitation that has shown promising results, although it has been little studied in patients with ABI. The main objective of this study was to explore the feasibility of a TEVI-DCA program as a rehabilitation tool for people with ABI. In this study, 14 people with ABI were recruited (mean age of 52.43 years (range from 35 to 65 years), 57.1% men) and took part in a twice-weekly IVR therapy program. The intervention was feasible and safe. The participants completed the program with no adverse effects (no symptoms on the Simulator Sickness Questionnaire), and experienced high usability (System Usability Scale > 80%) and outstandingly positive post-game experiences (Game Experience Questionnaire 2.56/4). In addition, the participants significantly improved several of their physical and cognitive capacities, showing increased strength (handgrip p = 0.042), reduced fall risk (Tinetti test p < 0.001), an increase in the physical component of the quality of life (PCS-SF-8 p = 0.006), and improved executive functions (Wisconsin Card Sorting Test p = 0.005). These findings demonstrate that the TEVI-DCA program appears to be a feasible therapeutic tool for people with ABI, as it has shown maximum adherence, with an absence of adverse events, and was shown to lead to improvements in physical–cognitive aspects, although further studies are needed to corroborate the findings of this study.

1. Introduction

Acquired brain injury (ABI) refers to any damage to the brain that is non-congenital or degenerative in origin and occurs after birth [1,2]. It includes both traumatic and non-traumatic brain injury [1,3,4], with stroke, traumatic brain injury, anoxic encephalopathy, brain tumors and infectious diseases of the brain—such as meningitis or encephalitis—being the most common causes [3,5]. This type of damage can affect the areas of the brain that control essential functions such as movement, speech, memory and behavior, leading to functional limitations in the activities of daily life and of social interaction [1].

It is estimated that more than 3.5 million people in the USA develop ABI each year [6], and it is one of the leading causes of disability worldwide [3,7]. In Europe, traumatic injuries alone have rates between 47.3 per 100,000 and 694 per 100,000 population per year [8]. The impact of ABI depends on the location and severity of the damage, and thus the associated symptoms are very varied [1,3,9], from paralysis and loss of motor control to cognitive difficulties such as problems with memory, attention, and problem-solving. Other common symptoms include fatigue, emotional changes, and language disorders such as aphasia. It should be noted that the more severe the condition, the greater the treatment needs are, and the higher the rehabilitation costs [4]. In summary, ABI can affect individuals’ physical, cognitive, and/or psychosocial functions, areas in which rehabilitation will be required [2,10]. In this paper, the physical and functional sphere is the main focus, as this has not been sufficiently studied to date [4,11].

The focus of physical rehabilitation is on restoring as much function as possible and improving the patient’s quality of life [10]. Physical therapy plays a crucial role as it can promote neuroplasticity through challenging, repetitive, task-oriented, and motivational training [9,12]. Physical therapy aims to improve the physical–functional abilities of people with ABI in areas such as strength, mobility, balance, and coordination, in order to achieve functional gains in their quality of life and their independence.

A useful tool in the rehabilitation of people with ABI could be virtual reality (VR). Furthermore, if virtual reality is fully immersive (IVR), it has been shown, due to its multi-stimulation ability, that it can improve attention and reduce distraction [13].

This suggests that it could be a good tool in neuro-rehabilitation. Some recent reviews have confirmed this theory in the case of people who suffered a stroke [14,15,16]. Additionally, immersing the patient in a more stimulating environment and a family environment offers the possibility of involving them in a unique way and improving motivation [17].

In IVR scenarios, the patients are completely immersed in a virtual environment with which they interact as if it were a real environment. This can have certain benefits, as follows [18]: it allows for therapeutic work in an ecological, dynamic, and controlled environment [19]; it can be adapted to the needs of each patient; and it favors adherence to treatment and motivation due to its gamification possibilities [20]. Furthermore, it is an innovative tool that has already demonstrated potential within other groups [2], such as those suffering from neurological and neurodegenerative pathologies [16,21,22,23,24]. However, evidence of its effects in cases of traumatic brain injury (TBI) is limited and contradictory [9,18]. The review by Pietrzak et al. [9] points out that the evidence on the motor and cognitive improvements that a VR treatment can achieve in TBI is very limited since most of the related studies are not randomized clinical trials (RCTs) (only 5/18, and these are pilot RCTs) and are not IVR, and the improvements found in terms of balance or improvements in upper-extremity motor function are not significantly superior compared to traditional therapy. On the other hand, the review by Vilageliu-Jordá et al. [18] does focus on IVR, but of a total of five articles, only one is a RCT; the authors also conclude that, having analyzed the findings on the use of IVR in cognitive rehabilitation, neither its efficacy or its usefulness in the case of ABI were demonstrated. Additionally, in the case of people who have suffered strokes, even if there are improvements in their walking speed, balance, and mobility, it is not clear whether these results are clinically significant [24]. Furthermore, some recent reviews have concluded that it appears that IVR would be a better option compared to non-immersion in the case of upper-limb rehabilitation after stroke [25] or in cases where patients have cognitive impairments or dementia [26].

All these issues point to the need for further clinical evidence [1], especially regarding the use of IVR [27], as most of the studies on the use of virtual reality in physical rehabilitation for ABI took place in a non-immersive environment [2,4,9,24,28,29,30,31,32].

Therefore, the aim of this study is to analyze the feasibility of implementing an IVR program as a rehabilitation tool for people who have suffered an ABI, and, as a secondary objective, to explore its possible effects on their functionality and quality of life.

2. Materials and Methods

2.1. Study Design

A single-group pilot study with a pre-post design (named the TEVI-DCA program) was carried out.

2.2. Participants

The study was carried out with the participation of Red Cross Cerebral Palsy and Brain Injury Center in Castro Riberas de Lea (Lugo, Spain), an organization with whom our research group has a collaborative agreement. All the members of the center were informed about, and invited to participate in, the study at an information session held by the researchers. To this end, the staff members provided an informed consent form so that both the participants and/or their families could read it at home. They were asked to sign and return it within 10 days. No incentives were offered.

The following inclusion criteria were established: 1. ABI due to endogenous and exogenous agents; 2. hemiplegia (right or left); 3. level of functional capacity allowing for unaided standing, ability to walk more than 10 m unaided/unassisted, and a functional cylindrical grip with the hemiplegic or affected hand.

Exclusion criteria were established as follows: 1. presenting cardiovascular, pulmonary or musculo-skeletal problems that would prevent the IVR session from being carried out, 2. inability to respond to the evaluation protocol, and 3. severe visual or hearing impairment, vertigo, epilepsy or uncontrolled psychosis.

Fourteen people with ABI were recruited (mean age of 52.43 years; 57.1% men). Due to the research design, the main objective (feasibility) that the people were volunteers, and considering the sample size of other studies with similar characteristics [33,34], the pilot study was started with a sample size of fourteen people. Registration for this study will be completed in a second phase of development that includes a randomized clinical trial.

The research was carried out after informed consent was signed by all participants and conducted in accordance with the ethical principles of the Declaration of Helsinki [35], as well as the provisions established on Personal Data Protection (Organic Law 3/2018) and on the Guarantee of Digital Rights (Organic Law 3/2018, of 25 May), and after being approved by the University of Vigo Faculty of Physiotherapy Ethics Committee (code 205-2023-2).

2.3. Intervention

The intervention (TEVI-DCA program) was carried out using a head-mounted display (Meta Quest II, Oculus VR, Menlo Park, CA, USA) and HoloFit software (4.7.0.8 version) (available in the library at www.oculus.com, accessed on 25 June 2024). When synchronized with the cyclo-ergometer (Oxycycle III, active and passive; MSD, Roeselare, Belgium), this allowed for the patient to pedal through virtual environments, such as different European cities (London, Paris, Venice, etc.), natural landscapes, or sports competitions, while simultaneously performing small cognitive tasks (making decisions or discovering objects). They performed the activity in two modes of the exergame (explorer and training). All the sessions were supervised by the center’s rehabilitation staff (physiotherapists and/or occupational therapists) (Figure 1).

The intervention lasted 16 weeks, comprising two sessions per week on alternate days, and was divided into three consecutive stages. This structure was established following our previous experiences with this software in the Parkinson’s population [36,37] (Figure 2):

• First stage: One week long and intended to adapt the patients with ABI to the use of the cyclo-ergometer (using either their upper or lower limbs, depending on the patient).
• Second stage: One week long again, with the introduction of IVR. The Explorer mode of the Holofit software was used, with each patient being able to choose between a natural or an urban environment. In this modality, the participant is able to exercise at a free pedaling pace.
• Third stage: Fourteen weeks long and subdivided into two phases of seven weeks each. The first phase is characterized by low-intensity effort (HR: 50–60%)/Borg: 3–4), where the volume of work is progressively increased (2 weeks with 5 min sessions, 3 weeks with 10 min sessions, and the last 2 weeks with 15 min sessions). If the patients’ conditions permitted, the effort was shared between the upper and lower limbs. The second phase of this stage, also lasting 7 weeks, followed the same pattern in relation to the volume of work, but in this case, it was carried out with a moderate-intensity effort (HR: 70–80%/Borg: 5–6). In this stage of the intervention, the Training modality of the Holofit software was used. In this modality, the participant must pedal at the pace set by the exergame. If the pedal rate is too slow or too fast, a message warns the participant to increase or decrease their pedaling frequency.

2.4. Evaluation Tools

Attendance at the scheduled sessions, the number of adverse events, and an assessment of the patients in the weeks before and after the intervention described above were noted. In this assessment, functional capacity, neurocognitive, and quality of life tests were carried out. In addition, an intrinsic assessment of the IVR exposure was also made at the end of the intervention.

Finally, it should be noted that, in each session, the participants’ heart rates were measured, their perception of effort was recorded with the modified Borg scale [38], and the presence or absence of cybersickness was noted.

• The following physical capacity tests were applied:
  • Handgrip Strength Test to quantify maximum isometric hand and forearm strength [39,40,41]. A Jamar^® Smart digital hand dynamometer (J.A. Preston Corporation, Clifton, NJ, USA) was used. The test was applied to both hands with the participant in a sitting position and with the elbow flexed 90 degrees.
  • Five times sit-to-stand test (FTSST) to assess the functional mobility and strength of the lower extremities. The participant sat with his back against the back of the chair. The assessor counted each position out loud so the patient remained oriented. The test stopped (using a stopwatch) when the patient reached the standing position on the fifth repetition [42,43].
  • Timed Up and Go Test (TUG) to assess mobility, dynamic and static balance, and fall risk. The test procedure consists of getting up from a chair, walking 3 m, turning around, returning to the chair, and sitting down. The time required to complete the predetermined route was measured with a stopwatch in seconds. Shorter times indicate better performance [44,45].
  • Tinetti Test to assess gait and balance, as well as to determine the level of early stage fall risk. This consists of a gait subscale and a balance subscale. The maximum possible total score is 28: 12 for the gait subscale and 16 for the balance subscale. The Tinetti score subdivides patients into three groups according to the level of risk of falls: major risk (≤18 points), moderate risk (19–23 points), and minimal risk (≥24 points) [46,47].
• The following quality of life assessments were also applied:
  • Short-Form-8 (SF-8) health survey to assess health-related quality of life. The SF-8 measures the same eight health domains as the SF-36 Health Survey with only eight questions. The scale was Likert type with five points (1–5). The higher the value, the better the quality of life in relation to health. These eight dimensions are grouped into two global components (physical and mental), which result from the sum of the defining dimensions of each component [48,49].
• The following neurocognitive tests were used:
  • Wisconsin Card Sorting Test (WCST) to assess executive functions and identify cognitive flexibility deficits. It consists of 4 stimulus cards and two sets of 64 cards (in the manual version) each, comprising 128 cards in total. The cards are composed of a combination of three types of attributes or characteristics: shape (triangle, star, cross, and circle), color (red, blue, green, and yellow), and number (1, 2, 3, or 4 elements). The task is to distribute the cards that match a particular criterion. When the subject makes 10 consecutive correct answers, it is considered that a category has been completed; from that moment on, the classification criterion is changed without prior notice. If you continue classifying the cards using the criterion of the previous category, you will score perseverative errors. The number of attempts is also recorded [50,51].
  • Clock Drawing Test (CDT) to assess executive function and possible cognitive impairment linked to visuospatial abilities and constructive practice [52,53,54]. This was carried out under two conditions: asking the patients to draw a clock following a verbal command (CDT—verbal) and asking them to copy a drawing of a clock (CDT—copy). The participant was asked to draw a clock face with all the numbers and hands and to say the time shown on the clock. The number 12 must appear at the top (3 points), there must be 12 numbers present (1 point), there must be two distinguishable hands (1 point), and the time must be correctly identified (1 point) to obtain the maximum score.
• Regarding questionnaires on issues intrinsic to IVR exposure, the following were used:
  • Simulator Sickness Questionnaire (SSQ) to assess the safety of IVR exposure by assessing potential associated symptomatology in three broad domains [55,56,57]: 1. oculomotor symptoms; 2. disorientation; 3. nausea. Each item is assessed on a four-point scale (0 = do not feel anything; 1 = a little; 2 = medium; and 3 = a lot) and the total score (maximum of 48 points) is the sum of the scores of the three subscales.
  • System Usability Scale (SUS) to assess the usability of the device/protocol. This consists of ten questions on a Likert-type scale. Each question is scored from 1 to 5 according to the level of agreement or disagreement with each statement, with 5 meaning completely agree and 1 meaning completely disagree. The algorithm that results from these answers creates a score out of a maximum of 100 points [58,59].
  • The post-game module of the Game Experience Questionnaire (GEQ-post game, to assess how players felt after they stopped playing the game. This module is also a Likert-type scale consisting of 17 items, in which responses are graded according to the intensity of the sensations experienced (0 indicates not at all and 4 indicates extremely). These items are, in turn, divided into four components (positive experiences, negative experiences, tiredness, and return to reality), which are scored individually and whose average could result in a maximum score of 4 points [38]. In the absence of a validated version of the questionnaire in Spanish, and so the current study group would have no problems with this questionnaire, a version was used that was translated by the authors and which has been used in previous research [60,61].

2.5. Data Analysis

An analysis of the socio-demographic characteristics of the sample was conducted using measures of central tendency (mean), dispersion (standard deviation, minimum, maximum), and percentages. After verifying the normality of the variables under investigation, Paired Student t-test was applied to assess the effect of the developed program. All analyses were performed using IBM-SPSS version 24.0 software, and a p-value of <0.05 was considered statistically significant.

3. Results

A group of fourteen people with ABI participated, with a mean age of 52.43 years (range of 35–65 years), of which 57.1% were men. The main socio-demographic data are shown in

Table 1. Socio-demographic characteristics of the group.

		Mean/%	SD	Minimum	Maximum
Age (years)		52.43	8.64	35.00	65.00
Gender	Male	57.1%
	Female	42.9%
Diagnosis	Wernicke’s encephalopathy	7.1%
	Stroke	35.7%
	Traumatic brain injury	28.6%
	Frontotemporal dementia	7.1%
	Cerebral palsy	7.1%
	Encephalitis	7.1%
	Meningitis	7.1%
Diagnosis (years)		22.29	17.47	1.00	58.00
Barthel Index	Independent	28.6%
	Mild dependence	64.3%
	Severe dependence	7.1%
Cognitive impairment	Mild	50.0%
	Moderate	35.7%
	Severe	14.3%

.

In relation to the results obtained in the functional capacity tests,

Table 2. Main results regarding functional capacity pre- and post-intervention.

	Pre				Post
	Mean	SD	Minimum	Maximum	Mean	SD	Minimum	Maximum
Handgrip [dominant, Kg]	9.71	2.63	5.00	12.50	11.21	2.08	9.00	13.50
Handgrip [non-dominant, Kg]	7.88	5.09	2.50	20.00	9.50	5.39	2.50	20.00
Five sit to stand (s)	23.25	10.21	10.00	50.00	20.50	7.98	11.00	37.00
Timed up and go (s)	30.27	25.34	9.00	96.00	29.38	22.96	9.00	89.00
Tinetti test—gait	7.06	2.26	0.00	9.00	9.08	2.06	6.00	12.00
Tinetti test—balance	6.81	2.14	2.00	10.00	12.14	3.44	4.00	16.00
Tinetti test—total score	14.47	2.75	10.00	19.00	21.46	4.89	13.00	28.00

shows that there was an increase in strength in both the dominant (1.5 kg) and non-dominant (1.6 kg) hands, an improvement in the FTSST of 2.75 s, an improvement in the TUG of 0.89 s and an improvement in the Tinetti test in terms of both balance (6.05 points) and gait (2.23 points), as well as in the total score (7.67 points).

In terms of health-related quality of life,

Table 3. Results obtained regarding the SF-8 quality of life pre- and post-intervention.

	Pre				Post
	Mean	SD	Minimum	Maximum	Mean	SD	Minimum	Maximum
General Health	2.38	0.96	0.00	4.00	3.29	0.91	1.00	4.00
Physical Functioning	1.69	1.54	0.00	4.00	3.00	1.04	1.00	4.00
Physical Role Functioning	1.69	1.30	0.00	4.00	2.43	1.09	1.00	4.00
Bodily Pain	3.31	0.95	2.00	5.00	4.14	0.66	3.00	5.00
Vitality	2.63	0.89	1.00	4.00	3.07	0.73	2.00	4.00
Social Functioning	2.62	1.15	1.00	4.00	2.71	0.73	2.00	4.00
Emotional Role Functioning	2.75	0.93	1.00	4.00	2.50	0.85	1.00	4.00
Mental Health	2.50	1.15	0.00	4.00	2.79	0.70	2.00	4.00
Physical Component Summary	9.06	3.73	4.00	16.00	12.86	2.63	9.00	17.00
Mental Component Summary	10.50	3.16	5.00	15.00	11.07	2.13	9.00	16.00

shows that, after the intervention, there was an improvement in mental health of 0.57 points and in physical health of 3.8 points.

In relation to the neurocognitive tests, Figure 3 shows a summary of the data from the WCST before and after the intervention, showing a reduction in the number of errors (2.1), an increase in the number of correct answers (9.31), and an increase in the number of attempts (7.32). Figure 4 presents the CDT summary data, showing an improvement of 0.52 points in the CDT—verbal and 0.13 points in the CDT—copy.

Additionally,

Table 4. Inferential analysis of the effects of the TEVI-DCA program in the variables evaluated (functional capacity, quality of life, executive functions, and cognitive impairment) pre- and post-intervention.

	Paired Differences
	Mean	SD	Std. Error Mean	t	p	Cohen’s d	Confidence Interval
Handgrip [dominant, Kg]	−0.70	0.63	0.15	−4.53	<0.001	0.63	−1.69–−0.48
Handgrip [non-dominant, Kg]	−0.38	0.61	0.17	−2.24	0.044	0.62	−1.20–−0.01
Five sit to stand (s)	1.25	4.15	1.03	1.20	0.248	0.15	−0.20–0.79
Timed up and go (s)	−0.60	5.09	1.31	−0.45	0.655	0.09	−0.62–0.39
Tinetti test—gait	−1.69	1.60	0.44	−3.81	0.002	0.60	−1.72–−0.35
Tinetti test—balance	−5.14	1.61	0.43	−11.94	<0.001	0.61	−4.50–−1.86
Tinetti test—total score	−6.92	2.62	0.72	−9.49	<0.001	0.62	−3.79–−1.44
PCS—SF-8	−3.26	3.91	1.01	−3.22	0.006	0.91	−1.41–0.34
MCS—SF-8	−0.66	4.01	1.03	−0.64	0.530	0.02	−0.67–0.34
WCST—total errors	6.00	6.80	1.81	3.29	0.006	0.84	0.26–1.51
WCST—total corrects	7.42	9.34	2.49	2.97	0.011	0.34	0.17–1.38
WCST—number of attempts completed	13.42	14.84	3.96	3.38	0.005	0.80	0.24–1.49
CDT—verbal	0.16	1.58	0.45	0.36	0.723	0.58	−0.46–0.67
CDT—copy	−0.58	1.68	0.48	−1.19	0.257	0.68	−0.92–0.24

shows the comparative pre- and post-intervention data, where significant functional improvements can be seen (in the handgrip test and in the Tinetti test), as well as improvements in health-related quality of life (Physical Component Summary of the SF-8) and in executive functions (in the WCST), after the application of the IVR program for people with ABI.

Finally, in relation to the intrinsic assessment of the IVR program that was carried out,

Table 5. Main results relating to the intrinsic immersive virtual reality exposure post-intervention.

Test	Measured Dimension	Mean	SD	Minimum	Maximum
SSQ	Nausea	0.01/4	0.02	0.00	0.07
	Oculomotor symptoms	0.13/4	0.14	0.00	0.36
	Disorientation	0.10/4	0.13	0.00	0.26
	Total Score	0.24/4	0.26	0.00	0.57
SUS	Usability	81.86/100	10.14	57.50	95.00
GEQ—postgame	Positive experiences	2.56/4	0.60	0.83	3.33
	Negative experiences	0.09/4	0.24	0.00	0.83
	Tiredness	0.71/4	0.31	0.00	1.00
	Return to reality	0.51/4	0.24	0.00	1.00

GEQ: Game Experience Questionnaire; SSQ: Simulator Sickness Questionnaire; SUS: System Usability Scale.

shows the residual values for the SSQ, showing no serious adverse effects, only some mild symptomatology in the fields of tiredness, eye strain, difficulty focusing, blurred vision, and dizziness with eyes open, in some subjects. Usability, as measured with the SUS, showed values above 80%. Finally, in relation to the GEQ post-game, the participants’ perception of the IVR intervention was generally positive, with very low values for negative experiences, fatigue, and problems with returning to reality.

4. Discussion

When examining the results of the intervention, and bearing in mind its objectives, the first thing that stands out is that IVR is a feasible therapeutic tool for people with ABI. As shown in the previous section, the intervention had 100% adherence, with no adverse effects—other than very slight negative symptomatology, which could be considered residual after taking into account the values of negative experiences or problems in returning to reality as measured by the GEQ post-game. Overall, the participants viewed the intervention as a positive experience, as measured by this same questionnaire. These results are in line with those observed in the study by Moraes et al. [62], who also analyzed the feasibility of IVR and applied the SSQ, using a cohort of TBI patients in their investigation. Furthermore, our data are also in line with those produced by studies carried out in other population groups, with the results showing clear effects on physical–functional components, such as multiple sclerosis [21,63,64], Parkinson’s disease [36,65,66], or post-stroke patients [66]. Additionally, Lee et al. [67]—in their study on the use of IVR in the assessment and rehabilitation of the activities of daily life in patients with TBI or stroke, in which they also used the SSQ—pointed out that IVR was a feasible technology to apply in the treatment of this type of patient, as did Prats-Bisbe et al. [68], who applied IVR to facilitate cognitive rehabilitation in their study.

Additionally, the SUS seems to confirm the ease of using IVR in this population group. With values above 80%, the usability of the tool is indicated to be excellent, and although we have not found other studies that analyze usability using the SUS in patients with ABI, there are some studies linked to populations with multiple sclerosis or Parkinson’s [36,64,65] that produced similar data to this one.

In relation to the effects on the physical–functional capacity of patients with ABI, IVR seems to be an appropriate physical rehabilitation tool. In the present study, a 14-week intervention, with two sessions/week, all the analyzed variables improved, with significant improvements in hand grip strength, balance, and gait, as assessed by the Tinetti Test, and the summary physical component, as measured by the Short-Form-8 Health Survey. However, it should be noted that the improvements did not reach statistical significance in lower-limb strength, as measured with the FTSST, in gait, measured with the TUG, or in the summary mental component of the SF-8, although these values did improve after the intervention.

These data, although in line with results from other studies on improvements in gait and balance achieved using virtual reality—such as that of Biffi et al. [10], where the efficacy of VR in improving gait in children with ABI was analyzed—differ from most other studies, which mainly used non-immersive or semi-immersive VR [2,4,9,28,30,31,32,69,70]. The systematic review by Corbetta et al. [24], which analyses the ability of VR versus standard rehabilitation to improve gait speed, balance, and mobility after stroke, concludes that VR is more effective than standard rehabilitation, and that the results obtained from standard rehabilitation are improved if VR is added. Non-immersive or semi-immersive VR also predominated over IVR in the review of Corbetta et al. However, as was pointed out by Kim et al. [29] in their review on the use of VR in upper-extremity motor recovery in stroke patients, future research is needed to confirm the theory that VR therapy is more effective than standard rehabilitation, as well as to determine which type of VR is more effective, IVR or non-immersive VR.

In the review by Garay-Sánchez et al. [71], they concluded that non-immersive virtual reality combined with conventional rehabilitation could be considered as a therapeutic option in this population, although only 2/10 of the selected articles tested IVR. Focusing on the analysis in these two articles, the support of other devices in the immersive virtual program (treadmill) stood out. In our work, we combined IVR and a cycle ergometer, as in our previous research applied to Parkinson’s disease [36,37], opting for a modality that allows the intervention to be carried out while seated, which provides a safer way to carry out exercise-based programs in this group.

Furthermore, future studies should focus on the careful design of IVR interventions based on the established objectives. For example, even a single feature of IVR can have a significant impact on motor performance, as demonstrated in the study of Kim et al. [72], in which the effects of stereoscopic objects during a reaching task in IVR scenarios hampered reaching performance, or in the study of Brazil and Rys [73], in which IVR influenced the time to complete a fine motor task in older adults. In the case of our exergame, the tasks did not require the use of the hands, since the physical tasks were carried out by pedaling and the cognitive tasks were selected with head movements (head-mounted display movements).

On the other hand, although gait and balance—as measured by the Tinetti Test—showed significant improvements in our study, this was not the case with when using TUG. In contrast, the systematic review by Li et al. [28] was able to obtain improvements in mobility, balance, and risk of falls, as measured with the TUG. These differences are perhaps due to the fact that a significant number of the studies included in this review used a treadmill in their interventions, which may have facilitated improvements in this particular test. Furthermore, most of these studies are non-immersive, using traditional consoles such as Wii Fit combined with conventional therapy, and are thus not comparable with the present study.

Turning to hand functionality, grip strength was assessed in this study. In this variable, significant post-intervention gains of 1.5 kg in the dominant hand and 1.6 kg in the non-dominant hand were obtained. Although the data were clinically positive for our patients, it should be noted that the data did not reach what is considered a minimal clinically important difference (MCID), which is established in post-stroke patients at 5 kg in the dominant and 6.2 kg in the non-dominant hand [74]. Other studies have focused on other aspects of hand functionality, such as dexterity or reach [15]. These aspects, although important, were not the subject of study in our proposal, which focuses more on improving the aspects of strength (especially in the upper limbs but also in the lower limbs) that impact gait. However, it is likely that future research will require longer interventions with a higher workload to obtain improvements in MCID. Probably, and in line with a recent meta-analysis [25], the fully immersive approach is the most appropriate, as this work suggests that the rehabilitation of upper-extremity function after stroke should be prioritize these treatments in the following order: head-mounted devices, non-immersive virtual reality systems, Microsoft Kinect, Nintendo Wii, and finally conventional rehabilitation.

In relation to quality of life, significant improvements in the Physical Component Summary of the SF-8 were found. This improvement represented an increase of 3.8 points, which may be an MCID according to the work of Fu et al. [75], who established this as between 1.8 and 3 points for the SF-36. As for the Mental Component Summary, a small improvement was observed here, which, although not significant, was in line with the significant improvements found in the Wisconsin Card Sorting Test and the Clock Drawing Test, tests that indicate the benefits that IVR appears to exert on executive functions, something that has been confirmed by several other reviews [2,9,11,62,68,76,77]. However, exploring the cognitive aspects in depth was not our objective, since we are physical therapists.

These findings open the door to the use of physical rehabilitation programs that include cognitive improvement objectives with measurements with validated scales such as the Montreal Cognitive Assessment or trail-making test, which have been used in similar research to ours [26].

Limitations

Despite all of the above, our research has several limitations. Firstly, the results obtained cannot be extrapolated to the whole population, as the sample is not representative. Given that IVR proved to be a safe and feasible tool in this type of patient, future research should look at a larger sample to allow for the generalization of the data obtained. A second limitation is due to the study design itself. In order to establish cause–effect relationships, future research designs should include randomized clinical trials. Thirdly, to present clinically relevant results, future research should use longer interventions and/or with a greater workload per treatment session. A fourth issue to be modified in future research is the recommendation that a follow-up is carried out after the end of the intervention to investigate how long the positive effects of the therapy are maintained for. As this is an ongoing, long-term, clinical situation, it would be beneficial to establish a balance between maintaining the good health of patients with ABI and the health care costs that this incurs to the system. This lack of follow-up, together with the lack of a real comparison between IVR versus another treatment method, means that the results should be interpreted with caution and should be addressed in future research to corroborate the data shown in our study.

Finally, it should also be pointed out that the study design in this study does not allow us to indicate to what extent IVR improves standard therapy, whether IVR is more effective when used alone or in conjunction with standard therapy, or whether the benefits of IVR are greater than the use of non-immersive VR—which has been more commonly researched in previous publications, as mentioned above. Furthermore, the heterogeneity of the causes of ABI in our sample could have influenced some of the results. Therefore, future research should resolve all of these issues and provide a solution to the limitations of the present study.

5. Conclusions

It appears that the TEVI-DCA program is safe and feasible in patients with ABI, while perhaps simultaneously improving their physical functions, such as gait, balance and hand grip strength, improving the physical component of their quality of life, and improving their cognitive executive functions. However, the design used and the sample size do not allow for the results to be extrapolated to the rest of the population, an aspect that should be considered in future research.