Use of recreational virtual reality for stress reduction in a psychiatric intensive care unit: findings from a mixed-methods pilot clinical trial

Abstract

Objectives

To explore implementation processes, acceptability among participants who engaged with virtual reality (VR), recruitment feasibility and exploratory within-session effects of recreational VR programmes in a closed psychiatric intensive care unit (PICU). The qualitative component focused on patients’ and professionals’ experiences of VR use, while quantitative measures examined inclusion rates, exploratory within-session changes in emotional distress and descriptive trends in coercive events, pro re nata (PRN) medication, perceived stress and length of hospitalisation during a year with versus a year without VR.

Design

Parallel mixed-methods pilot clinical trial with qualitative priority (QUAL-quan).

Setting

A Danish closed PICU.

Participants

All admitted patients who consented to participate during a 2-year period were eligible. Qualitative data included semi-structured individual patient interviews (n=9), a focus group discussion with professionals (n=4) and non-participant observations (8 sessions × 20 min). Quantitative data were collected by ward professionals and via electronic health record extraction. A total of 26 patients were included during the VR year and 9 during the control year, reflecting substantial under-recruitment relative to the planned sample (35/124).

Interventions

VR programmes for stress reduction, mindfulness, entertainment and distraction, including passive and active activities.

Results

Three overarching themes were generated in a qualitative content analysis: (1) Agency in a restrictive environment (categories: Own symptom mastering, A broader clinical toolbox and Connection through a shared third), (2) Escaping and expanding reality (categories: A refuge from distress and A gateway to inaccessible experiences) and (3) Professional readiness and implementation challenges (categories: The know-why gap and The know-how gap). Participants who engaged with VR described experiences of reduced distress, improved mood and greater calmness and perceived VR as a valuable addition to care, supporting autonomy and strengthening therapeutic relationships. However, knowledge and trust gaps, along with insufficient training, limited use. Quantitative measures partly corroborated these findings: an exploratory within-session analysis showed that emotional distress was lower after VR use than before use (95% CI −3.46 to −0.81, p=0.003). However, under-recruitment limited statistical power for between-period comparisons.

Conclusions

Among participants who engaged with VR, the use of VR programmes was generally perceived as acceptable and potentially useful in selected clinical situations. However, recruitment feasibility was not achieved, and the study design does not permit conclusions regarding effects on coercion, PRN medication, perceived stress or length of hospitalisation. Integration into standard care appeared possible under selected clinical circumstances, offering therapeutic value for patients and professionals, but it depended on adequate theoretical understanding and training of professionals.

Trial registration number

NCT05654740.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The qualitative approach allowed rich, in-depth exploration of implementation facilitators and barriers.

• Enhanced trustworthiness through triangulation at three levels and the inclusion of quantitative measures.

• Recruitment challenges and failure to reach the planned sample precluded some quantitative analyses.

• Absence of systematically collected demographic and clinical data of participants substantially limited assessment of representativeness and transferability and prevented meaningful between-period comparisons.

• Consent-related recruitment barriers may have contributed to sample homogeneity; reasons for non-participation were not systematically recorded for the 127 admitted patients who did not consent, limiting interpretation of recruitment barriers.

Introduction

The environment in a psychiatric intensive care unit (PICU) may contribute to emotional distress and behavioural disturbances in vulnerable individuals. Compared with general psychiatric wards, closed PICUs are characterised by higher levels of clinical acuity, violence, aggression and restriction. Patients may experience distress due to interactions with other inpatients, the rules and routines in the ward, feelings of isolation and lack of stimulation.^1–3 This may lead to conflicts, use of pro re nata (PRN) medication (ie, medication that is not taken regularly but as needed), coercive events (eg, physical retention, mechanical restraint and forced sedation)² and a low overall satisfaction with staying in the ward.⁴ Use of coercion has been evidenced to have serious implications for patients, such as trauma/re-traumatisation,⁵ worsening of mental health⁶ or even death.⁷ This evidence emphasises an urgent need to find alternative non-coercive interventions and aligns with the National Institute for Health and Care Excellence guidance emphasising the use of least restrictive interventions, de-escalation strategies and non-pharmacological approaches in inpatient psychiatric settings.³ It has been recommended to offer patients recreational and entertaining interventions to promote relaxation, redirection of thoughts and feelings of being active during the day.⁸ In this regard, research has demonstrated the potential of virtual reality (VR)-based interventions, which have proven efficacious for a variety of psychiatric conditions with minimal side effects.^{9 10} VR-assisted mindfulness has been shown to improve emotional states and management of psychiatric symptoms in individuals with mood or anxiety disorders¹¹ and to alleviate psychotic symptoms.¹² Additionally, evidence indicates that in health settings, VR can operate as a distraction-based method to reduce pain and anxiety¹² and is perceived as fun and enjoyable by patients.^{13 14} Furthermore, some studies in acute psychiatric wards have suggested that VR-based interventions may be associated with decreased violent incidents and restrictive practices.¹⁴ Previous studies have also demonstrated VR-based interventions in PICUs to be acceptable, tolerable and feasible.^{14 15} However, little is known about how recreational VR-based interventions may be implemented and experienced in closed PICUs, including their potential influence on distress, PRN medication use and coercive events. Against this background, we designed Virtual Leisure, a mixed-methods pilot clinical trial aiming to explore implementation processes, acceptability among participants who engaged with VR, recruitment feasibility and exploratory within-session effects of VR programmes comprising stress reduction, mindfulness, entertainment and distraction in a closed PICU. The qualitative part of the study focused on patients’ and professionals’ experiences, perspectives and interactions around VR use. The quantitative part examined inclusion rates and explored within-session changes in emotional distress, while potential changes in coercive events, PRN medication, perceived stress and length of hospitalisation among patients were examined descriptively.

The specific objectives of the Virtual Leisure study were the following:

• To explore patients’ and professionals’ experiences and perspectives on the use of VR for symptom management in a closed PICU.

• To examine patient behaviour and patient-professional interactions in relation to VR use.

• To assess the acceptability of VR among participants who engaged with it, explore recruitment and examine implementation processes in a closed PICU from both patient and professional perspectives.

• To explore inclusion rates and potential effects of VR use in a closed PICU on within-session emotional distress and descriptively examine trends in coercive events, PRN medication, perceived stress and length of hospitalisation.

Methods

Study design

The current study was a parallel mixed-methods pilot clinical trial which enrolled patients hospitalised in a specific Danish closed PICU.¹⁶ Both quantitative and qualitative data were collected over 2 years. In the first year, patients had access to VR programmes. In the second year, VR was not available, and patients received usual care but were assessed with the same quantitative measures. The study was conducted in accordance with a previously published protocol,¹⁶ with deviations described below.

A QUAL-quan design indicated qualitative priority, with an exploratory, descriptive qualitative component and embedded, complementary quantitative measures. Individual interviews were conducted with patients, and a focus group discussion was carried out with professionals. Additionally, an ethnographic approach was applied as non-participant observations were performed. Thus, the qualitative component of the study comprised triangulation of methods (individual interviews, focus group discussion, observations), sources (patients, professionals) and investigators.

Deviations from protocol

Quantitative outcomes, including electronic health record (EHR)-derived measures, were restricted to patients who provided written consent due to regulatory requirements. This deviated from the original protocol,¹⁶ which intended to include ward-level outcomes for all admissions. In the protocol, the primary quantitative outcomes were feasibility and acceptability (operationalised as ≥80% inclusion), while key secondary outcomes included coercive events, PRN medication use and patient satisfaction, with additional exploratory outcomes such as perceived stress, emotional distress, risk of violence and length of hospitalisation. The protocol further anticipated a total sample size of 124 patients (approximately 62 per study period), based on an expected inclusion rate of 80% and power calculations for detecting changes in coercion rates. However, recruitment fell far short of the planned sample size target of 124. Across the 2 consecutive years, we enrolled 35 patients: 26 patients in the experimental VR period and 9 patients in the control period, yielding a small, imbalanced sample and limiting any between-period inference. Accordingly, we deviate from the prespecified analysis plan, which originally included between-period comparisons using interrupted time series and regression-based approaches and instead report within-session change in the Subjective Units of Distress (SUDS)¹⁷ during the VR year and inclusion rates as our primary quantitative outputs. Coercive events, PRN medication administrations, perceived stress measured by the Perceived Stress Scale (PSS-10)¹⁸ and length of hospitalisation are presented descriptively only, without inferential testing due to insufficient statistical power and incomplete data. Exploratory non-confirmatory inferential analyses (Fisher’s exact tests) were additionally conducted for coercive events and PRN medication but should be interpreted with considerable caution given the sequential design, group imbalance, lack of adjustment for patient-days or case-mix and the impact of a serious adverse event during the control period (elaborated in the Discussion). The Client Satisfaction Questionnaire (CSQ)¹⁹ and the Brøset Violence Checklist (BVC)²⁰ data were not systematically collected and are therefore not analysed, primarily due to pragmatic constraints in the clinical setting and data collection procedures. As described in the published protocol, the VR software generated usage logs intended to be reviewed regularly to monitor adherence and implementation only. However, these logs were not retained in a form that allowed retrospective analysis, and attempts to retrieve them were unsuccessful. Consequently, total VR exposure, session duration and linkage between SUDS observations and specific VR programmes could not be established.

Overall, these deviations reflect a shift from the originally planned confirmatory between-period evaluation of clinical outcomes to an exploratory and descriptive analysis focusing on within-session effects and implementation-related indicators. The implications of these deviations are elaborated in the Discussion.

Setting and participants

The study took place in a closed PICU, which is part of the Mental Health Services in the Capital Region of Denmark. The PICU typically treats a diverse group of patients with various psychiatric disorders, including schizophrenia spectrum disorders, affective disorders and personality disorders. It has the capacity to accommodate 13 inpatients, with an average hospitalisation duration of 8 weeks per patient. The treatment and healthcare are based on a multidisciplinary approach, structured according to cognitive and milieu therapeutic principles.^{21 22} The inclusion criterion for all participants was the ability to provide informed consent. Exclusion criteria were refusal of informed consent and, for patients, severely impaired vision and head injury.

For the qualitative data collection, we included 9 patients for semi-structured individual interviews, 4 professionals for a focus group discussion and conducted 8 20-min sessions of non-participant (ie, the observer is not involved in human interaction in the field) observations. For the quantitative data collection, 26 patients gave informed consent during the VR year and 9 patients gave consent during the control year. The two study periods were strictly sequential calendar windows: the VR period ran throughout 2023 and the control period throughout 2024. The 26 patients enrolled during the VR period and the 9 patients enrolled during the control period are reported as separate cohorts. Because identifying information that would allow linkage of admissions across the two periods was not retained for non-consenting admissions, we cannot fully rule out that an individual patient could in principle have been admitted in both periods and consented in each. Furthermore, participants’ demographic and clinical data (eg, age, sex, diagnosis, symptom severity, admission status and baseline coercion or PRN medication history) were not systematically collected for research purposes during the study periods.

Sampling

For the qualitative data collection, patients and professionals were sampled purposefully by combining criterion sampling as all participants had to have some relation to the VR programmes and maximum variation sampling to ensure diverse usage patterns. The 9 patients who participated in semi-structured individual interviews were drawn exclusively from the 26 patients enrolled during the VR period, since the interview focus was the experience of using VR.

Additionally, a convenience sampling approach was employed, whereby a researcher conducted observations during pre-scheduled time slots and included any relevant events that occurred during these sessions. For the quantitative data collection, all patients admitted to the PICU who fulfilled the criteria were asked to participate.

Procedures

For the VR period, healthcare professionals employed in the PICU were trained in applying the VR equipment and educated in being attentive to indications for when to propose it to patients. All inpatients in the PICU were approached by professionals about participating in this study and patients provided written informed consent before study enrolment. Several VR programmes were offered to the consenting patients covering stress reduction, mindfulness, entertainment and distraction. They included both passive (non-interactive activities, eg, breathing exercises, environments with calming music or birdsong) and active (interactive activities such as games, drawing and exploring environments) programmes. This categorisation of VR programmes was intended to describe the range of available programmes rather than to support formal analyses according to programme category. The VR programmes were delivered using Oculus Quest 2 VR headsets, with equipment and software provided by the private company Khora VR (https://khora.com/). Khora VR provided an initial on-site training workshop on the practical use of the VR equipment and software. In addition, designated members of the clinical team received supplementary training and acted as local ‘VR-trainers’, supporting day-to-day implementation in the ward. Beyond initial training and standard technical support, Khora VR had no role in the conduct of the study and did not have access to patient-level or identifiable research data. VR software-generated usage logs contained date, time and type of VR programme, but no patient-level or identifiable information; these logs were intended solely for monitoring adherence and implementation, were not retained in a form that allowed retrospective analysis and were not included in the study analyses. Detailed descriptions of the VR equipment and programmes are provided in online supplemental appendix A. Patients were intended to be offered VR at least daily in situations which instigated stress reduction, mindfulness, distraction or entertainment (eg, when patients experienced distress from psychiatric symptoms or felt restless or bored). Patients could also request to use VR. Additionally, VR was proposed to help de-escalate situations which professionals considered may lead to harmful behaviour, a need for PRN medication or coercive events. Professionals made ongoing clinical assessments regarding whether VR use was appropriate, particularly in situations involving severe agitation, behavioural dysregulation, acute paranoia or concerns regarding symptom exacerbation. If patients were severely agitated, the VR programmes were not offered. Instead, professionals followed standard ward procedures (eg, PRN medication). The VR programmes were available on the unit throughout the 12-month VR period, but individual patients were exposed to VR only during their admission and only when sessions were initiated by the patient or proposed by professionals. Session duration was not systematically recorded.

In the subsequent control period, the VR equipment was not available. The professionals followed regular ward procedures.

Data collection and generation

Qualitative data were collected by multiple researchers from February 2023 to July 2024 and included semi-structured individual interviews, a focus group discussion and non-participant observations. Quantitative data were collected from December 2022 to December 2024 by ward professionals and via EHR extraction. Table 1 presents an overview of the data collection methods, along with themes for the interview guide, the focus group discussion guide and the observation guide.

Table 1. Overview of data collection methods.

Data collection method	Objective	Themes covered	Data source	Time frame
Qualitative methods
Non-participant observation (8 sessions x 20 min.)	Patient behaviour and patient-professional interactions in relation to VR use	Type of VR programme (ie, active/passive), initiation of VR use (ie, patient’s/professional’s), occurrence (what happened?), patient utterances, professional utterances and any informal conversation between observer and patient/professional	Patients and professionals	7 February 2023–25 March 2023
Individual semi-structured interview (n=9)	Patients’ experiences with VR use in relation to symptom management	Initiation of VR use (ie, patient’s/professional’s), experience of VR content, feasibility (sequence of consumption, ease of use, mastering the technical, improvement suggestions), acceptability (affective attitude, burden, ethicality, intervention coherence, opportunity costs, perceived effectiveness, self-efficacy) and VR’s influence on coercion and PRN medication	Patients	7 March 2023–11 November 2023
Focus group discussion (n=4)	Professionals’ experiences, perspectives and attitudes in relation to VR use in closed PICUs	VR in relation to coercion, PRN medication, patient isolation, safety, psychosis/manic episodes, professionals’ resources, ward activities, ease of use and patient-professional relations	Professionals	3 July 2024
Quantitative methods
Questionnaires	Patients’ emotional distress before and after VR (SUDS) and weekly perceived stress (PSS-10)	N/A	Patients	1 December 2022–1 December 2024
EHR extraction	Patients’ PRN medication use, coercive events and length of hospitalisation	N/A	Patients	1 December 2022–1 December 2024

EHR, electronic health record; PICU, psychiatric intensive care unit; PRN, pro re nata; PSS-10, Perceived Stress Scale; SUDS, Subjective Units of Distress; VR, virtual reality.

Qualitative data

Non-participant observations

Non-participant observations (n=8) were conducted in the PICU at distinct time points by researcher GJ, with no prior knowledge of participants. A semi-structured observation guide (online supplemental appendix B) was followed and fieldnotes were taken during and immediately after observations. The objective was to gain first-hand insight into patient behaviour and patient-professional interactions in relation to VR use.

Semi-structured individual interviews

In total, 9 patients, with varying degrees of experience with the VR programmes, were interviewed to explore their experiences of VR in relation to symptom management. To address acceptability, the interviews were carried out in line with the theoretical framework of acceptability. Here, acceptability is understood as a multifaceted concept representing the degree to which individuals receiving or delivering a healthcare intervention perceive it as appropriate, based on prospective or retrospective cognitive and emotional responses to the intervention.²³ Furthermore, the interview guide covered patients’ experiences with feasibility and explored the potential influence of VR use on PRN medication and coercive events (online supplemental appendix C), as well as potential adverse effects associated with VR use. Interviews were conducted by researcher GJ.

Semi-structured focus group discussion

A semi-structured focus group discussion was completed with professionals (n=4) experienced in the VR programmes. They were presented with 10 claims, formulated with inspiration from the fieldnotes and the individual interviews (eg, VR can reduce the use of PRN medication), as a point of departure for discussing their experiences, perspectives and attitudes in relation to VR use in closed PICUs (online supplemental appendix D). Professionals were also encouraged to discuss relevant self-generated claims. The focus group discussion was conducted by researcher JM (moderator) and an external researcher (observer) not otherwise involved in the study’s data collection, generation or analysis.

Quantitative data

Quantitative data on participants’ PRN medication use, coercive events and length of hospitalisation were extracted from the registers at Mental Health Center Glostrup. Furthermore, participants completed SUDS before and after use of VR programmes and PSS-10 weekly. Usage logs generated by the VR software were used for monitoring adherence and implementation only and were not included in statistical analyses.

Data analysis

Qualitative analysis

All interviews were audiotaped and transcribed verbatim. Data were analysed using Graneheim and Lundman’s²⁴ framework for qualitative content analysis. The analysis was conducted by researcher KR, who was not otherwise involved in the study, including study planning and data collection. It was an iterative process and involved identifying content areas in the transcribed data which related to each specific study objective, thus creating a constellation of meaning units all relating to the same meaning. Subsequently, the content areas were condensed and further abstracted, whereby the condensed meaning units were labelled with codes. These codes were sorted into categories, representing the manifest content of the text, which were further collated into themes expressing the latent content of the text.²⁴

Quantitative analysis

For SUDS, we computed session-level differences (post-pre) and tested whether the mean change differed from zero using a two-sided one-sample t-test on the deltas. Only documented sessions with both pre- and post-SUDS values were included. Sessions with missing pre or post values were excluded at the session level (listwise deletion). Total VR exposure could not be reconstructed retrospectively. Session duration, the linkage between SUDS observations and specific VR programmes, and patient-level exposure to specific VR programmes were unavailable for analysis. No imputation was performed given the small sample size and the within-session paired structure of the analysis. As robustness checks, we ran a Wilcoxon signed-rank test on the deltas and a per-patient analysis that first averaged the delta within each participant and then applied a one-sample t-test to those patient means to avoid overweighting frequent users. All participants with at least one valid SUDS session were included; no minimum number of sessions was required. Inclusion rates were calculated per period and overall, with 95% Wilson CIs. PRN medication, coercive events, PSS-10 and length of hospitalisation were summarised descriptively with no inferential testing.

Researcher reflexivity

Multiple researchers with varied clinical and research backgrounds were involved in data collection, generation and analysis, enabling triangulation at three levels: across methods, sources and investigators, thereby enhancing the credibility and trustworthiness of the findings. Furthermore, regular team reflections helped mitigate biases, ensuring that findings reflected participants’ perspectives and experiences rather than researcher presuppositions derived from our previous research experiences in VR use in mental health settings.

Patient and public involvement

None.

Results

In line with the QUAL-quan study design, the qualitative component was primary, while the quantitative was complementary. Accordingly, the results are presented sequentially, with qualitative findings presented thematically, followed by quantitative results.

Qualitative findings

Three themes and seven categories were generated in the qualitative content analysis, which are described below and illustrated in figure 1. Examples of illustrative quotes from patients and professionals, as well as fieldnotes, are provided in table 2. While demographic and clinical characteristics of participants were not systematically collected, online supplemental table S1 provides additional contextual information regarding participants’ relation to VR use and characteristics of the non-participant observations.

Figure 1. Themes and categories generated in the qualitative content analysis.

Table 2. Illustrative sample quotations and fieldnotes according to themes.

Themes	Quotations (samples)	Field notes (samples)
Agency in a restrictive environment	“I’m thinking of a specific patient who benefitted greatly from VR. The patient has a history of escalating to very high levels, and there has been a need for restraint… VR helped that patient. Every time it escalated, we used VR, and the patient was completely neutral and calm again after its use.” (Professional #1, female) “I feel good. I just experience calm, an inner peace, when I watch the waterfalls, the beach, and all those things.” (Patient #4, male) “In relation to the belt-restrained patient, we need tools and strategies the most. We often give a lot of medication but lack activities we can offer them. I would like us to use and consider VR more in these situations. It’s a group we often don’t help much other than just giving medication.” (Professional #3, female) “VR is a tool that enables them to independently regulate affect.” (Professional #1, female) “If we’ve taught a patient to use VR, and it’s the only strategy that works, once discharged they no longer have access to it. When they become irritable or symptoms of mania return, they lack that strategy, which may lead to readmission.” (Professional #3, female) “One of the few positive experiences in the psychiatric system can turn into a feeling of abandonment when that strategy is no longer available.” (Professional #4, male)	Non-participant observation #17 February 2023: The nurse offers it, and the patient accepts and explains that she often benefits a lot from it. “If I can, I would also like to use it at home when I am discharged. It helps me when I feel unwell and restless, and now, while I am psychotic.” Non-participant observation #7, 25 March 2023: The patient approaches the professional and wants to introduce her to the VR headset and show her how VR works. The patient explains that she feels the professional hasn’t familiarised herself with it. The professional is open to being introduced, saying it’s been a very long time since she was introduced to it and hasn’t started using it. The patient smiles: “It’s like the roles are reversed.”
Escaping and expanding reality	“VR gave me a break. The thoughts settled down.” (Patient #9, male) “My focus goes there. I love nature… You can’t always go out to a beach… You can immerse yourself in a situation that is otherwise difficult to experience in real life.” (Patient #4, male) “I think VR could be a really good alternative to distract thoughts.” (Professional #3, female) “There is a lot of medicine that could be discarded… I get a lot of morphine… It relieves the pain.” (Patient #1, male) “Sometimes I can feel uncomfortable because I don't really see what’s going on around me.” (Patient #5, male)	Non-participant observation #8, 25 March 2023: She (the patient) is very willing to share her thoughts about the VR study. The patient says: “I’m the kind of person who likes the real world, so VR is not for me. I’m always treated the same way every time I’m hospitalized. I wish people could accept that I want treatment without medication and be receptive to that. But I don’t think VR can be an alternative to that. I want help through action (…) I’d much rather go to a real forest and take a walk.”
Professional readiness and implementation challenges	“I experienced that when we had patients in the manic phase of bipolar disorder, they really benefited from passive VR programs, where they could go to a calm place and find a bit peace in their chaotic mind.” (Professional #1, female) “They were difficult to engage, because they were often paranoid about the consent process. The need to obtain patients’ signatures was the biggest obstacle. I believe we would have had more patients try it out if it hadn’t been part of a research study.” (Professional #4, male) “This works better for me than PRN medication. I no longer take any PRN medication because of VR.” (Patient #2, female) “If I knew that I was just as confident in using VR, I’d turn to VR as my number one choice instead of PRN medication… The negative thing is the uncertainty around using VR headsets when professionals lack training in using VR and understanding why we should choose this… If I really knew that I would be much more motivated… What is the research behind it? What can it do? Just like we know from our education and professional knowledge.” (Professional #2, female) “I’ve experienced that many of our colleagues needed 4, 5, or even 6 introductory sessions, and still forgot how to use it the following week.” (Professional #4, male)	Non-participant observation #3, 23 February 2023: The chief physician assesses that one patient might benefit from VR. The patient is very distressed, self-harming and loud. The professionals express that they do not believe or cannot imagine that it could help the patient. Non-participant observation #5, 11 March 2023: A patient in a manic state wishes to use the active VR. He is asked whether he feels overstimulated, more energetic or more manic afterwards. He says no, because his distress has decreased. He describes it as beautiful and impressive: “It’s like when I’m psychotic. I mean, the pleasant part.” Non-participant observation #6, 19 March 2023: The iPad isn't working and is plugged into the charger. The patient is informed that it will take an hour before she can use the black headset (ie, the passive VR programmes). After an hour, it is discovered that the headset has also run out of battery. The patient is offered the white headset (ie, the active VR programmes). The patient declines.

PRN, pro re nata; VR, virtual reality.

Agency in a restrictive environment

Own symptom mastering

Patients described having tried the VR programmes several times and stated that they wanted to use them again. VR was highlighted as an effective tool for reducing stress and inducing calmness, with particularly the passive VR programmes being preferred by patients in this regard. For example, one professional (#1) described that VR use “created calmness in a chaotic mind” for a patient with mania. Furthermore, the active VR programmes, characterised by their vibrant content and opportunities for physical engagement, were experienced as particularly mood-elevating and as alleviating boredom and restlessness.

VR was emphasised as a tool that supported patient independence in symptom management, as it most often did not require professionals’ involvement during use, beyond initial introduction. Several patients also expressed a wish for the VR programmes to be available to them after discharge and felt confident in their ability to use VR independently to manage their symptoms. In accordance with this, professionals recounted that VR allowed patients to independently learn emotion regulation, aligning with the PICU treatment goal of increasing patient responsibility. However, professionals also highlighted potential drawbacks, noting that VR being available only during admission could pose a barrier to recovery. They expressed concern that patients might learn strategies they could not use once discharged, potentially leading to re-admission and turning a positive experience within the psychiatric system into a negative one.

Additionally, VR was described as a potential alternative to PRN medication when patients felt restless or agitated, with one patient (#2) reporting it to be more effective than PRN medication. Professionals also shared a case in which VR helped de-escalate a patient with a history of previous forced sedation and mechanical restraint. For this patient, VR was perceived as contributing to avoidance of coercion during the patient’s current admission. They viewed this as a turning point where VR contributed not only to symptom management but also to a more dignified psychiatric care experience.

A broader clinical toolbox

VR was described by professionals as a valuable tool for patients refusing medication or in restrictive conditions, such as belt fixation, as activities for patients, especially in mechanical restraint, were lacking. One professional (#3) explained often feeling inadequate in supporting such patients besides administering medication. VR was also highlighted as a useful option for bedridden patients with depressive symptoms due to its flexible delivery or for patients wanting to be in their own room. Such patients often refused using other ward facilities, such as the sensory room. Professionals described that this expansion in their tool repertoire provided a sense of security and made them feel more equipped to help patients.

Connection through a shared third

Professionals described VR as particularly valuable in interactions with patients in psychotic states or in mechanical restraint, as they often resisted interaction with the professionals. In such instances, VR helped patients calm down and self-regulate without direct contact with professionals. Professionals described a specific case where VR had been a tangible tool to help a patient with severe paranoid symptoms, as their usual methods to approach patients and build a relationship were exhausted and ineffective. In this context, VR served as a shared third, allowing interaction on the patient’s terms without rupturing an established alliance. Additionally, VR appeared to facilitate person-centred conversations about patients’ VR experiences, perceived by professionals as strengthening the patient-professional relationship.

The playful nature of the active VR programmes helped break traditional power dynamics between professionals and patients, fostering “a sense of humanness”. However, professionals cautioned against VR replacing essential patient-professional interaction, emphasising that it should not substitute emotional care when patients need human connection. While the independent characteristic of VR allocated more time to professionals’ other work tasks, professionals stressed that its use should remain purposeful and be followed by support, rather than functioning as a quick fix during busy periods.

Escaping and expanding reality

A refuge from distress

Both patients and professionals recognised VR as effective in providing distraction due to its immersive nature, both from mental distress and from physical pain relief. Patients described the VR headset as blocking out the physical world, making them feel as though they were “in another world” or “a new reality”. They emphasised VR’s authentic and realistic content in this regard. One patient (#1) described VR use as “a break from the hustle and bustle of daily life”, particularly highlighting relief from the physical environment in the PICU. This was appreciated by several patients, and they explained that the PICU environment caused distress by being boring and reminding them that they were hospitalised.

Few patients reported discomfort due to the sensory deprivation and reduced awareness of other people in their physical surroundings. However, they described feeling safe in the physical settings; therefore, these experiences were generally tolerated and did not prevent VR use. Regarding physical adverse effects, most patients did not experience any discomfort from the VR use. A few reported on cybersickness related to the disturbance of sensory integration, but for most it was tolerable. However, one patient (#3) experienced cybersickness while using VR and dizziness lasting for several hours after use, therefore not wanting to use VR again.

A gateway to inaccessible experiences

Patients highlighted VR as an entryway to experiences they felt cut-off from. Described as “magic”, “imagination-fulfilling” and “the perfect world”, VR granted access to environments and activities otherwise inaccessible due to hospitalisation. However, although wanting non-pharmaceutical treatment, one patient (non-participant observation #8, 25 March 2023) preferred real-life activities over VR, believing them to be more effective.

For some patients, VR re-created positive experiences from psychosis or mania. One patient (#6) described feeling that VR replicated a deep focus and connection to nature once felt during mania. Another patient (#2) described psychosis as feeling like being in a movie, and VR provided a similar concept. These associations were embraced by patients, describing it as reducing distress when re-experiencing this.

Professional readiness and implementation challenges

The know-why gap

Both patients and professionals acknowledged VR’s effectiveness and potential to reduce PRN medication use and prevent coercive events. However, this recognition was constrained by a tension between therapeutic effectiveness and trust in relation to its applicability. This dilemma was further reflected when professionals described having limited theoretical knowledge about VR compared with established interventions, such as PRN medication, where education and experience fostered confidence. This was also evident in the non-participant observations where professionals disagreed on the appropriateness of VR in specific clinical situations. Some professionals expressed concern that the active VR programmes might overstimulate patients with manic or psychotic symptoms, potentially exacerbating symptoms and had therefore not offered VR to such patients. Meanwhile, others felt that its controlled nature made it safe. One patient (#2) experienced distressing associations during VR use, although this did not negatively affect the patient’s use or overall perception of VR.

While patients did not believe that VR could worsen symptoms, they recognised that the excitement they felt while using VR mimicked bodily sensations and energy levels felt during mania or psychosis. These experiences contributed to professionals’ cautious approach to determining when VR was clinically appropriate. Some professionals saw this as beneficial, as they experienced patients self-discontinuing medication when recovering from psychosis. They believed that VR might help patients adapt post-recovery by offering a controlled way to simulate positive psychotic experiences.

Additionally, professionals were reluctant to offer VR to highly agitated patients due to safety concerns, including the potential use of the VR headset as a projectile. Some patients also expressed hesitation, with one patient (non-participant observation #4, 11 March 2023) showing reluctance to participate and not wanting to be “a guinea pig”. Furthermore, suspicion and mistrust related to providing consent before study enrolment, particularly among patients with paranoid symptoms, made them hesitant to engage with VR. In this regard, professionals noted that the consent requirement itself acted as a barrier to patients’ motivation and willingness to use VR.

The know-how gap

While patients expressed no difficulties with mastering VR, several professionals described lacking sufficient practical skills to operate the VR equipment. This acted as a barrier to offering VR, and professionals reported often resorting to PRN medication despite preferring VR. One professional (#4) explained this using the concept of “affect contagion”, concerned that the professionals’ insecurities in relation to VR might heighten patient distress.

Quantitative results

Inclusion rates and reduction of emotional distress (SUDS)

Acceptability, operationalised as the inclusion rate among patients, was lower than anticipated. In 2023 (VR year), 26 of 87 admissions consented (29.9%; 95% Wilson CI 21.3% to 40.2%); in 2024 (control year), 9 of 75 (12.0%; 95% CI 6.4% to 21.3%). This decline in inclusion rate is partly attributed to the serious adverse event in the PICU affecting unit operations (also described in the Discussion). Overall, across both periods, 35 of 162 admissions consented (21.6%; 95% CI 16.0% to 28.6%).

Across 23 documented sessions with paired pre-post SUDS values from 10 patients, mean SUDS decreased from 5.04 pre-session to 2.91 post-session, yielding a mean change of −2.13 points (95% CI −3.46 to −0.81; two-sided one-sample t-test of mean change=0, p=0.003). Per-participant session counts and pre-post SUDS values are provided in online supplemental table S2). The standardised within-subject effect size was Hedges’ g_z = −0.67 (95% CI −1.509 to −0.231), indicating a moderate reduction in emotional distress within sessions. Sensitivity analyses supported the direction of the within-session effect: a Wilcoxon signed-rank test on session-level deltas yielded W=32.5, p=0.007, and a per-patient analysis averaging deltas within each participant before testing yielded a mean of patient means of −1.90 (95% CI −3.43 to −0.37; t(9) = −2.80, p=0.021) (online supplemental table S3). Given the limited sample, the magnitude of the within-session distress reduction should be interpreted cautiously. CSQ and BVC data were not systematically collected during the study periods, likely due to factors elaborated on in the Discussion. Consequently, these measures were not analysed.

Coercion, PRN medication, perceived stress and length of hospitalisation

During the VR period, 9 of 26 enrolled patients (34.6%) experienced at least one coercive event, compared with 2 of 9 (22.2%) during the control period (Fisher’s exact OR=1.85, 95% CI 0.26 to 21.6, p=0.69). The total number of coercive events recorded was 18 in the VR period and 2 in the control period (table 3). For PRN medication, 18 of 26 (69.2%) VR-period patients and 6 of 9 (66.7%) control-period patients received at least one PRN medication administration (Fisher’s exact OR=1.13, 95% CI 0.15 to 7.07, p=1.00), with a total of 529 administrations in the VR period and 209 in the control period (table 4). These between-period comparisons should be interpreted with caution given the small and imbalanced sample sizes and the sequential study design. Accordingly, results are presented descriptively, and no inferential conclusions are drawn. Full descriptive data for PSS-10 and length of hospitalisation are presented in table 5 and table 6, respectively.

Table 3. Coercive events during the study periods by type and reason.

Group	Patients in group (n)	Patients with ≥1 coercive event, n (%)	Total events
Control	9	2 (22.2%)	2
VR	26	9 (34.6%)	18

Characteristic	Value
Mean event duration (days)	13.2
Range (days)	0–76
Coercion type - involuntary retention (ward detention)	12 events
Coercion type - forced medication	5 events
Coercion type - straps (limb restraints)	1 event
Coercion type - belt restraint	1 event
Coercion type - physical holding	1 event
Reason - health-related	12 events
Reason - dangerousness	6 events
Reason - agitated state	2 events

Fisher's exact test (any coercive event during admission, patient level): OR = 1.85, 95% CI 0.26 to 21.6, p = 0.69.

Confidence intervals span more than two orders of magnitude. The numerically higher coercion rate in the VR period must not be interpreted causally given the marked imbalance in group sizes, the temporal confounding inherent in the sequential design, the absence of adjustment for length of hospitalisation or case mix and the impact of the serious adverse event during the control year (described in the main text).

VR, virtual reality.

Table 4. PRN medication administration by study period (patient-level and total administrations).

Group	Patients in group (n)	Patients with ≥1 PRN, n (%)	Total administrations	Mean/patient (SD)	Median/patient (IQR)
Control	9	6 (66.7%)	209	34.83 (37.64)	17 (8–61)
VR	26	18 (69.2%)	529	29.39 (20.99)	29 (10–40)

Fisher's exact test (any PRN medication administration during admission, patient-level): OR = 1.13, 95% CI 0.15 to 7.07, p = 1.00.

Per-patient counts are not adjusted for length of hospitalisation; total administrations should not be compared between groups in raw form because of the marked imbalance in patient-days at risk.

PRN, pro re nata; VR, virtual reality.

Table 5. Perceived Stress Scale (PSS-10) scores across study assessments.

Assessments (n)	Participants (n)	Mean (SD)	Median (IQR)	Range
62	22	18.2 (7.5)	19 (13.0–22.8)	3–37

PSS-10 was not analysed inferentially due to insufficient and incomplete data across the two periods.

PRN, pro re nata; VR, virtual reality.

Table 6. Length of hospitalisation (days) by study period.

Group	N	Mean (SD)	Median (IQR)	Range
Control	4	53.25 (36.75)	45.5 (26.25–72.5)	21–101
VR	26	83.00 (86.34)	49 (27–98.25)	5–345

Length of hospitalisation data were available for 4 of 9 control-period patients and all 26 VR-period patients; the substantial imbalance and skew precluded meaningful between-period comparison.

VR, virtual reality.

Discussion

Due to substantial under-recruitment and protocol deviations, the present findings should be interpreted as exploratory rather than definitive. Given the qualitative priority of the study design, particular emphasis is placed on implementation processes and experiential aspects of VR use.

According to the analysis of qualitative data, both patients and professionals perceived the VR programmes to be associated with reduced distress and restlessness, elevated mood and greater calmness. This pattern is partly supported by the SUDS findings from documented sessions with paired values, which showed lower emotional distress after VR use than before. These findings correspond with prior research reporting favourable outcomes of VR use in psychiatric inpatient populations.^{14 15 25 26} Consistent with these studies, most participants reported repeated use of VR and interest in continued use both during admission and after discharge, indicating positive user reception. However, the unavailability of the VR programmes when discharged was pointed out as a factor that could influence patient satisfaction. This indicates that broader implementation, for example, for clinics or supporting housing residences, might be pivotal. Unfortunately, high costs of VR equipment have been identified as a barrier to implementation across various mental health settings.²⁷ Nonetheless, this finding suggests that professionals should discuss with patients which coping strategies they discovered through VR use and whether these can be transferred to crisis management after discharge.

Our quantitative indicator of acceptability, operationalised as the proportion of admitted patients providing consent, was 29% during the VR year, which was well below the predefined threshold of 80% specified in the study protocol. Thus, recruitment feasibility was not achieved as the observed inclusion rate fell substantially below the threshold. Several factors likely contributed. First, limited clinician confidence may have biased recruitment towards patients perceived as more likely to benefit from VR. Second, the consent rate was influenced by the requirement to consent not only to VR use but also to study data collection and therefore did not purely reflect willingness to use VR. Third, many inpatients are admitted in acute crisis, reducing the practical window for introducing a novel technology. Finally, a serious incident in the PICU temporarily limited the unit’s capacity to conduct the study. The serious adverse event involved a critical incident in the PICU unrelated to VR use that substantially affected professionals’ resources and ward operations. These factors may also have contributed to incomplete data collection for some measures, including CSQ and BVC, and may have influenced the observed outcomes. While the inclusion rate fell substantially short of the predefined threshold and recruitment feasibility was not achieved, qualitative findings suggest that among patients who consented and engaged with VR, it was generally perceived as acceptable and beneficial. This distinction indicates that recruitment feasibility does not necessarily reflect feasibility or acceptability of the intervention among those who engaged with it.

A prominent finding was VR’s role in granting access to otherwise inaccessible environments, which patients highly valued. This aligns with existing research on the positive mental health effects of nature exposure.²⁸ Thus, VR offers a unique opportunity to simulate nature experiences in restricted settings, potentially promoting similar positive effects. Furthermore, patients emphasised the authenticity and realism of the VR content as important for successful distraction. This can be understood through the concept of presence (ie, the degree of psychological immersion in the virtual environment, depending on a sense of being there and doing there) which shifts attention away from the immediate physical surroundings.²⁹ Contrary to earlier research demonstrating that interactive VR environments elicit a stronger sense of presence than passive ones,^{30 31} the patients’ experiences of VR’s ability to distract appeared to be indifferent across the active and passive VR programmes in this study. This suggests that when VR is used for distraction, interaction may contribute to immersion but does not appear to be essential. This aligns with the assumption that presence is fundamentally a psychological phenomenon rather than primarily driven by technological features.³² While this interpretation is plausible, the mechanism was not formally assessed, and future research should examine this using direct measures of presence.

Furthermore, VR use appeared to facilitate person-centred conversations and contribute to relationship-building between patients and professionals. This is significant as prior research underscores patients’ appreciation for clinicians who demonstrate humanity, kindness and compassion beyond formal professionalism.^33–35 Given the inherent power dynamics within psychiatric inpatient settings, where power imbalances can trigger conflict and containment,^{2 5} it is a notable finding that VR use appeared to soften hierarchical dynamics between patients and professionals. Interventions aiming to strengthen the patient-professional relationship have been shown to decrease conflict and coercion.^{36 37} Thus, VR’s potential to mitigate power imbalances and support patient-professional alliance may be relevant for enhancing patient satisfaction and potentially reducing coercive events, aligning with internationally implemented strategies such as the Safewards Model²² and the Six Core Strategies.^{38 39} The finding that VR was perceived as promoting patient autonomy and symptom self-management also aligns with the recovery-oriented care approaches in mental health systems internationally.^40–44 Furthermore, VR was valued as an expansion of the clinician toolbox, enhancing professionals’ sense of agency and security while reducing feelings of professional inadequacy. This is important as it addresses a prevalent challenge in high-demand mental health settings.^45–48 Professionals in inpatient settings face increased risk of burnout,⁴⁹ and these environmental conditions have been linked to adverse consequences, not only for professionals themselves but also for the quality of patient care.^{47 50} VR’s perceived ability to allocate more time to professionals and expand their options for optimal care delivery echoes previous benefits highlighted by professionals.^{51 52} This may increase professionals’ willingness to use VR²³ and position it not only as a therapeutic tool for patients but also as a valuable resource supporting professional resilience in challenging environments.^{53 54}

Although most participants described VR as generally tolerable, adverse experiences and safety-related concerns were also identified. Some patients experienced discomfort due to cybersickness or sensory deprivation, not significantly impacting their VR experience. These adverse effects align with those reported in previous studies.^55–57 Some professionals expressed concerns about VR potentially exacerbating psychotic or manic symptoms, which are worries professionals have voiced beforehand.^{51 58} While VR can provoke paranoid ideations in vulnerable individuals^{59 60} and potentially amplify existing paranoid ideations,⁶¹ VR-based interventions for psychosis have been deemed feasible and safe in controlled settings.⁶² Mirroring previous findings,¹⁴ one patient in this study associated VR content with distressing past psychotic experiences. However, this did not appear to lead to exacerbated symptoms. Importantly, VR was not used indiscriminately in acute situations, as professionals continuously evaluated whether patients’ clinical presentation and behavioural state were compatible with safe VR use. It is worth acknowledging that VR was not used in high-risk situations, emphasising the potential of VR as an add-on strategy to already implemented strategies. Because adverse effects were explored qualitatively rather than monitored systematically across all sessions, the study cannot estimate the frequency of adverse experiences.

Furthermore, VR shared characteristics with sensory rooms in promoting autonomy and symptom self-management, along with reducing emotional distress and inducing relaxation.^63–68 Since the efficacy of sensory rooms for reducing seclusion and restraint remains inconclusive,^{63 69} VR may be a promising supplement, as it was perceived by patients and professionals as capable of reducing coercive events. This comparison, however, remains tentative, and direct comparative effectiveness has not been established. Professionals, nevertheless, valued the portability of VR in situations where sensory rooms were inaccessible and experienced it as an adaptable, individualised intervention. The experience of VR as individualising is important as it counters patients’ common experiences of psychiatric settings as dehumanising and neglectful of their subjectivity, causing segregation between professionals and patients marked by feelings of ‘us’ and ‘them’ with power residing in the relationship.^{5 35 70} By fostering more equal patient-professional interactions, VR may strengthen the therapeutic alliance, which meta-analyses have linked to improved treatment adherence and outcomes.^71–73 Moreover, the therapeutic relationship is a key determinant of patient satisfaction in psychiatric care and can mediate the negative impact of coercive events on satisfaction.^{74 75} In this context, VR’s potential to enhance therapeutic alliances and patient satisfaction is clinically relevant, particularly in inpatient settings where coercion may occur.² Another noteworthy finding was VR’s perceived usefulness in situations where patients resisted interaction with professionals. However, professionals cautioned against VR replacing genuine human connection and emotional care, which are closely linked to recovery outcomes.⁷⁶ Given the high prevalence of social withdrawal among psychiatric patients^{77 78} and the negative consequences of affection deprivation,^{79 80} balancing technological interventions with authentic human connection is critical. These insights resonate with the Safewards model, which stresses the importance of professional-patient relationships in reducing conflict and containment, as VR use may help create some of the protective conditions outlined in the model. At the same time, the caution expressed by professionals aligns with Safewards’ emphasis on relational, emotionally attuned care.²

Barriers to implementation included professionals’ uncertainty about VR mechanisms and insufficient training, leading to hesitancy and preference for familiar interventions such as PRN medication. Previous studies have reported similar barriers,^{27 81} which underscores the importance of intervention coherence (ie, the extent to which the individual understands the intervention and its underlying mechanisms) and self-efficacy (ie, the individual’s confidence that they can perform the behaviours demanded by the intervention) within implementation frameworks.²³ Nonetheless, despite skill-related barriers, both patients and professionals generally perceived the VR programmes as meaningful and effective. However, suspicion and mistrust, particularly among patients with paranoid symptoms, limited engagement with VR, reflecting an inherent ethical tension in psychiatric research where procedures intended to protect vulnerable patients may restrict access to potentially beneficial interventions.

Implications for future implementation

To enhance implementation in mental health settings, future efforts should address the knowledge and skill gaps experienced by professionals. These gaps placed professionals in an ambivalent situation: recognising VR’s promise while questioning their ability to use it effectively and safely. Previous research shows that negative attitudes toward VR are a stronger barrier to future use than the presence of positive attitudes are for facilitating use.^{52 82} Variables such as prior experience with VR, openness to new technology and favourable views on exposure therapy are positively associated with acceptance of VR in therapy.^{52 83} These findings highlight the importance of standardised training programmes that not only increase knowledge of VR-based interventions but also offer hands-on experience. Such training may reduce resistance and foster more favourable attitudes, ultimately supporting the integration of VR into routine clinical practice in mental health settings.

Methodological considerations

To enhance trustworthiness, triangulation was employed across methods, sources and investigators alongside complementary quantitative measures.⁸⁴ Across-methods triangulation ensured that conclusions are not just contextually grounded but are cross-verified, thereby fostering credibility and guaranteeing that findings remain dependable across different approaches. Source triangulation further supported credibility, dependability and transferability by demonstrating the consistency of findings across settings and participants. Findings from the different data collection techniques supported each other as patients’ and professionals’ statements generally aligned or reflected one another, indicating a high level of confirmability. Although coding was conducted by a single researcher and no participant validation was undertaken, interpretations were developed and refined through ongoing discussions within the research team. While single-researcher coding may have limited the range of interpretive perspectives brought to the initial analysis, the involvement of multiple researchers in the iterative interpretation of the data contributed to the credibility and depth of the findings. A key methodological limitation was the substantial under-recruitment, which precluded the planned between-period analysis of clinical outcomes and shifted the quantitative component toward within-session and descriptive analyses. In addition, total VR exposure could not be reconstructed retrospectively, and the 23 analysed SUDS observations therefore represent documented sessions rather than necessarily all VR use; session duration, linkage between SUDS observations and specific VR programmes and patient-level exposure to specific VR programmes were unavailable for analysis. It is therefore possible that sessions with paired SUDS values were not fully representative of all VR use, and the within-session findings should be interpreted conservatively. Moreover, suspicion and mistrust surrounding consent procedures may have contributed to sample homogeneity and limited variation in the findings. Selection bias is inherent in pilot designs with informed consent and was further accentuated in this study by restricting analyses to consenting patients rather than the full inpatient population. Future research should examine whether VR benefits extend to patients initially reluctant or ambivalent about its use. However, maximum variation sampling within the qualitative component may have partly mitigated this limitation. A major methodological limitation was the absence of systematically collected demographic and clinical data. This substantially limits assessment of representativeness, transferability and comparability between the two study periods and precludes meaningful interpretation of between-period differences. Accordingly, any apparent differences between the study periods must be interpreted cautiously and not as evidence of intervention effects: CIs spanned more than two orders of magnitude, the groups were markedly imbalanced in size, the count-based comparisons were not adjusted for length of hospitalisation or case-mix and the numerically higher coercion rate in the VR period was incompatible with inferential interpretation given the size imbalance and the serious adverse event during the control year. These results were therefore treated as descriptive only. The sequential design, with VR available in the first year and withdrawn in the second, introduces a risk of temporal confounding and limits causal inference. Future studies should consider a concurrent or active control condition to better isolate intervention effects. Because identifying information for non-consenting admissions was not retained, we cannot fully exclude the possibility that an individual patient was admitted in both study periods, although we consider such overlap unlikely to have materially affected the descriptive findings given the typical admission length (mean approximately 8 weeks) and the relatively low overall consent rate. Reasons for non-participation were not systematically recorded for the 127 admitted patients who did not consent, primarily because regulatory constraints limited documentation for non-consenting individuals. This limits interpretation of recruitment barriers.

Conclusions

This mixed-methods pilot clinical trial with a QUAL-quan design evaluated the acceptability, feasibility, implementation and potential effects of recreational VR programmes in a closed PICU, drawing on patients’ and professionals’ experiences and perspectives, behavioural observations and questionnaire data. Qualitative findings suggest that VR may support symptom relief through distraction, calming environments and a sense of escape and may contribute to improved emotional regulation, as well as enhanced patient-professional interactions. Professionals and patients described situations in which VR was perceived as an alternative to PRN medication or escalation, but the present study design does not allow inference regarding effects on coercion or medication use. Furthermore, recruitment challenges and gaps in knowledge and technological skills were identified as barriers to implementation. Among patients who engaged with VR, it was generally perceived as useful and acceptable. Taken together, these observations suggest that VR may be beneficial in selected situations, while being less feasible during severe distress or periods of high clinical workload. Continued investment in theoretical education and practical training for professionals is essential.

Data availability statement

Data are available upon reasonable request. No data are available.