Following a participatory action research approach (59), our research team involved (n = 28 total) key stakeholders including three LTC staff knowledge users, two LTC organizational manager partners, one industry partner, two patient partners living in LTC, two family partners with caregiving experience for their family members residing in LTC, 14 researchers and/or clinicians, and four trainees. Throughout the research process, team meetings were held monthly to integrate the perspectives and suggestions of the stakeholders described above to improve the study's methodology and materials and to inform our objectives.

This study was part of a larger research program aiming to understand the best ways to effectively implement VR technology with older adults in LTC, as well as to explore the specific feasibility, acceptability, and efficacy of a multi-user VR app (i.e., Toujours Dimanche) in this context. Since 2020, Super Splendide has been developing the Toujours Dimanche VR application, which is also available on mobile devices (Android, iOS) to support VR program delivery by care practitioners. This study examined the use of Version 3 of Toujours Dimanche, which was released in 2023 and was a preliminary version of the application. Super Splendide has since been incorporating significant updates to the application for use in rural and urban LTC homes, run by provincial health organizations in Quebec, Canada. The company released Version 4 of their app in 2024. Version 3 of the Toujours Dimanche platform enabled staff to conduct VR sessions with an unlimited number of virtual users, immersing them in 360-degree videos of natural environments from the province of Quebec. Immersions involving audio-recorded, guided mindfulness meditations linked to the VR videos were also available to users. The application was developed to comply with provincial health and cybersecurity standards (60) and provided facilitators with administrative controls over in-headset content and settings. To date, the Toujours Dimanche application has been deployed in approximately 100 healthcare facilities, primarily in LTC and palliative care settings.

Furthermore, in 2022 the Chaudière-Appalaches public health authority [CISSS de Chaudière-Appalaches (CISSS-CA)] purchased approximately two Meta Quest 2 VR headsets per the 29 LTC homes in their region (61). Super Splendide was hired by the CISSS-CA in 2022 to provide VR training and support for LTC staff through in-person training workshops and an online community of practice. This was the company's first experience deploying services, training, and a community of practice with LTC organizations (during a period spanning from 2022 to 2023), and occurred in a context where approximately 75% of residents have neurocognitive disorders (62).

Our study focused on allied health professionals (e.g., recreation therapists, psychoeducators, physiotherapists), referred to as LTC staff, who work with people living in LTC residences based in the Chaudière-Appalaches region of Quebec, Canada, operated by the CISSS-CA. The residents of LTC homes in Quebec typically consist of older adults too frail to live on their own, who require 24/7 support [MSSS, 62]. The LTC staff targeted for this study work with residents presenting with conditions such as chronic pain, mobility issues, and significant cognitive challenges. They are also responsible for organizing VR-related activities with residents. LTC staff members were recruited for the study using convenience sampling. To be considered eligible, staff members were required to have completed the first version of the VR training delivered by Super Splendide during the 2022–2023 period, and to have used VR with LTC residents. Across the participating sites, approximately 28 LTC staff members met these eligibility criteria.

Initial recruitment was conducted using a network-based (informal referral) approach. Organizational partners in LTC on our research team (e.g., clinical activities specialists) promoted our study to staff via recruitment emails and posters. Recruitment was also conducted through in-person visits by the research team to LTC residences. Research team members then met with interested staff in-person or over the phone to answer their questions about the study.

A sociodemographic questionnaire developed by the research team was used to collect information on participants (see Supplementary Appendix 1 for the sociodemographic questionnaire). The Assessing Determinants of Prospective Take-Up of Virtual Reality questionnaire [ADOPT-VR; 63; 64] assesses health professionals' intentions to use VR and perceived barriers/facilitators. The ADOPT-VR questionnaire has demonstrated strong face and content validity and good internal consistency (Cronbach's α = 0.876) (63). ADOPT-VR was used as a basis for the development of our mixed-methods French-language instrument (i.e., the “VR-use questionnaire”) for LTC staff. The specifically tailored VR-use questionnaire that was used for this study has not yet been psychometrically validated. It combined trichotomous items (yes/no/unsure), Likert-style confidence ratings (1-very uncomfortable; 5-very comfortable), and open-ended questions to assess individual, social, environmental, technology-specific, and system-level factors that influence LTC staff's adoption of VR over time.

Our VR-use questionnaire was field-tested to ensure content validity by asking the knowledge users and organizational partners on our research team for feedback on the clarity of wording, relevance of content, perceived gaps in content, repetitive questions, and general comments prior to being sent to participants. The developer of the original ADOPT-VR questionnaire (Dr. Stephanie Glegg) also provided feedback on the VR-use questionnaire during its development period. Our final questionnaire consisted of 67 items, using 34 trichotomous format items, five 5-point Likert-style items, two 0%–100% confidence scale items and 26 qualitative questions (see Supplementary Appendix 2 for the VR-use questionnaire).

After eligible participants were contacted by the research team to answer their study-related questions, an email was sent to potential participants with a link to the consent form, which they completed through REDCap electronic data capture tools hosted at the Université de Sherbrooke (65). Participants who provided their written, informed consent were then able to access the study questionnaire, which they completed via REDCap. Participants received an email with the anonymous survey link. The link to the study survey was also available as a URL included at the end of the consent form on REDCap. The participants completed the demographics questionnaire and the mixed-methods VR-use questionnaire concurrently. The questionnaire was sent to participants over a period of eight months from July 2024 to February 2025.

The program of research was reviewed and received ethics approval from the Chaudière-Appalaches Integrated Health and Social Services Centre (CISSS-CA) research ethics board, overseen by Quebec's provincial ministry of health (REB #: MP-23-2024-1086). Study participants provided their written informed consent to participate.

The Decomposed Theory of Planned Behaviour [DTPB; 66], an extension of Ajzen's (67) Theory of Planned Behaviour (TPB) combined with the Technology Acceptance Model (68), explains technology adoption through deconstructing the three predictors of the TPB (i.e., attitudes, social norms, and perceived behavioural control) into actionable belief sets (e.g., perceived usefulness, ease of use, and compatibility for attitude; peer and supervisor influence for social norms; self-efficacy for perceived behavioural control). These constructs are modeled to predict behavioural intention and use, providing a structured lens to examine determinants of technology implementation. The DTPB framework guided the creation of the ADOPT-VR2 (63, 64) instrument, which was used to inform the development of our mixed-methods VR-use questionnaire. Our quantitative and qualitative findings were thus guided by the DTPB's framework, to enhance our analysis of the factors impacting LTC staff's adoption of VR. The modified VR-use questionnaire's constructs were grouped into the following composites, based on the DTPB: (1) Attitude: general attitude, perceived usefulness, perceived ease of use, compatibility; (2) Social Norms: client influence, peer influence, supervisor influence; (3) Perceived Behavioural Control: self-efficacy, behavioural intention; and (4) Facilitating Conditions: facilitators and barriers.

The data analysis followed a convergent mixed-methods approach, and integrated findings based on quantitative and qualitative consensus (69). Reporting followed Good Reporting of A Mixed Methods Study (GRAMMS) guidelines (70). Quantitative responses were analyzed using Microsoft Excel. Descriptive statistics were computed for all sociodemographic variables and the quantitative items of the VR-use questionnaire to summarize the data, explore associations, and identify key trends, using frequency counts, means or medians, and graphs, as appropriate. Quantitative results were grouped according to participant characteristics, DTPB domains, and findings specific to VR training quality.

Qualitative reporting followed the Standards for Reporting Qualitative Research (SRQR) (71). The responses to the qualitative questions were analyzed using content-analysis (72, 73), with deductive and inductive coding used to inform the analysis. Guided by the DTPB (66), we structured our analysis using deductive codes derived from its theoretical constructs, which were prepared pre-analysis in a clearly defined codebook, and confirmed for study alignment through team discussions.

Two master's level research trainees (GT and OD), supervised by PhD graduate students trained in qualitative analysis (MGM and IS), then conducted the preliminary data analysis. They started by repeatedly reading through the questionnaire responses to familiarize themselves with the data. Next, they identified meaning units (i.e., parts of the text relating to the research's aim), labeling each meaning unit with a code to briefly describe its content. Codes were either deductive (based on pre-analysis DTPB theoretical constructs), or inductive (developed from the data during the analysis). A coding list was used to clearly define codes and ensure consistency, and to note any changes in codes or evolutions that occurred. To improve reliability, the two research trainees independently coded the same portion of the data (equivalent to approximately 10% of the full dataset). To ensure intercoder reliability, coding decisions were then compared, discrepancies were discussed, and definitions were revised as needed with four additional members of the research team (IS, MGM, MB, CSB). After initial intercoder reliability review and adjustments, the remaining data was split evenly between the two trainees (GT, OD) to be coded. Additional reliability meetings with GT, OD, IS, MGM, MB and CSB were held after coding 30% and 70% of the data.

The coded meaning units were then checked by GT, OD, and IS with the full text to see if any relevant content was missed. These coding modifications were then reviewed and confirmed during a larger team meeting (GT, OD, IS, MGM, MB, CSB). Next, categorization of meaning units was conducted. GT and OD then identified codes describing similar ideas, which were grouped into subcategories. Subcategories were then grouped into broader categories which helped structure findings and were reviewed for accuracy and clarity by IS, MGM, MB and CSB. Next, GT and OD identified themes, which IS, MB, and MGM then confirmed. This process involved reflecting on the categories and developing higher-order concepts that captured underlying meanings (i.e., linking multiple categories at a more interpretive level); results were compiled in a table summarizing illustrative quotes, subcategories, categories, and themes.

Upon completion of both the quantitative and qualitative analyses, a structured, mixed-methods data integration meeting was held with seven research team members: IS, MGM, MB, CSB, GT, OD and CG. This meeting followed a systematic approach to compare whether findings confirmed, expanded, or diverged from each other, thereby enhancing both the meaning and reliability of the overall integrative interpretation (69). Mixed-methods results were synthesized and displayed using tables and illustrative quotations. Given that the questionnaires were completed in French, the quotes presented in this article have been translated into English and verified by bilingual team members.

Malterud and colleague's (74) information power principle guided this study's sample size determination. Prior to recruitment, a target of 16 participants was set based on this principle, which accounts for study aim, sample specificity, theoretical framework, and analytic strategy. This decision was informed by the limited literature on VR adoption in LTC, the theory-driven nature of our approach, and the specificity of our population (LTC staff trained in VR). Recruitment used convenience sampling from a defined pool of LTC staff who had completed VR training and used VR with residents. Organizational partners and the research team promoted the study via emails and posters, and the research team conducted in-person visits to LTC homes to answer questions and invite staff to participate in the study. These steps ensured that participants met eligibility criteria and reflected the intended population. A sample size of 16 participants falls within recommended ranges for qualitative descriptive studies (75–78) and provided sufficient information power for our mixed-methods design. Furthermore, within the specific setting from where we were recruiting, there were 28 eligible LTC staff. A sample size of 16 therefore represents 57.1% of the eligible population, which aligns with best-practice benchmarks for survey research (79, 80).

Of the 20 people initially contacted to participate in the study, 17 provided consent and 16 completed the study questionnaires. The 16 retained participants had a mean age of 40.38 (SD = 8.07) and were white (100%; n = 16), francophone (100%; n = 16), and mostly women (81.3%; n = 13). Participants reported an average of 7.66 years of work experience across CHSLD, MDA, and MDAA settings. LTC staff members primarily identified as recreation practitioners (56.3%; n = 9) and specialized educators (31.3%; n = 5). Specialized educators in this context are psycho-educators and psychosocial workers who support residents' emotional and behavioral well-being and often facilitate individualized activities. For a more detailed overview of participant characteristics, see Table 1.

Regarding participants' prior VR training and experience, most participants (75.0%, n = 12) indicated that they had completed the one-day, 7-hour training offered by Super Splendide in 2022 and 2023, with others indicating that they had participated in the online training webinar offered by Super Splendide in 2023 (25.0%, n = 4). Two staff members (12.5%) reported participating in both the in-person and online training. Additionally, 3 participants (18.8%) indicated receiving supplementary online training delivered via video capsules, as part of Super Splendide's online community of practice. With respect to VR use with LTC residents, staff reported an average of 1.5 VR sessions per month, with a mean usage time of approximately 17.6 min per session. Furthermore, on average, LTC staff reported delivering VR sessions to approximately 3 residents at a time. For more details, see Table 2.

This section presents descriptive statistical findings from the VR-use questionnaire, grouped together by DTPB domains.

A total of 414 meaning units were drawn from participants' verbatim responses and were organized into 97 deductive and inductive code groupings, with their respective subcodes. These codes were then synthesized into nine distinct subcategories, which in turn were consolidated into four higher-order categories, informing the development of the final themes. Overall, coding decisions between coders were highly consistent: only 38 coding decisions were modified, 10 of which were resolved by a third party. The following four themes were generated regarding staff's perspectives on VR use in LTC:

When barriers outweigh promise: Challenges undermining the sustainable integration of VR in LTC.

Across qualitative themes and quantitative results, LTC staff appeared to judge immersive VR as valuable in principle, but routine use seems to be conditional on resident-technology fit and ease-of-use at the point of care. Favorable attitudes coexisted with only moderate confidence, especially with respect to quantitative results on resident selection and troubleshooting. Qualitative accounts tied these weak points to time and workflow pressures, connectivity/interface issues, headset comfort, and a narrow content library. Conversely, individualized sessions delivered in calm spaces often yielded perceived resident engagement, reminiscence, calm, and occasionally reduced BPSD, pointing to potential of VR use in LTC settings, especially when conditions and delivery are optimized. Adoption strengthened where facilitating conditions (e.g., protected time, headset mirroring for real-time guidance, on-unit support, broader culturally relevant content, hardware upgrades) and social infrastructure (e.g., visible leadership backing, peer modeling, multiple trained staff per unit) were in place. These facilitators in turn appear to bolster self-efficacy, potentially helping to normalize VR as “real care”.

The integration meeting also clarified several measurement nuances. First, “adequate space” for VR sessions (endorsed by 93.75%) was interpreted qualitatively as environment quality (quiet, dimmable, private), suggesting that an item assessing space needs with only limited response options may overstate feasibility, unless sensory conditions are specified. Second, several “don’t know” responses (e.g., supervisor support) were interpreted with respect to the LTC context. For example, top-down rollouts of VR technology by regional managers sometimes left ambiguity during unit-level implementation, helping to explain mixed social-influence signals, despite generally positive attitudes reported by participants. Finally, wording such as “as often as necessary”, and limited prior exposure to VR, likely contributed to ambivalence on some behavioural intention items on the questionnaire. The value of tailoring some items by experience level and clarifying definitions (e.g., “session”) was highlighted for future studies by team members.

In summary, staff uptake appeared to rise when compatibility, ease-of-use, enabling conditions, and social support align. Findings also pointed to the potential for ambivalence to resurface and VR use to taper off when any of the aforementioned facilitators were absent. See Table 4 for further details on integration of quantitative and qualitative findings.

Staff endorsed VR's value “in principle’, but routine use appeared to falter at the points of resident selection and technical troubleshooting. These difficulties, specifically selecting appropriate residents for VR activities and troubleshooting technical issues, were evident in the quantitative data, where only 37.5% of participants reported confidence in resident selection and 18.8% in troubleshooting. Qualitative findings further explained these challenges by highlighting time and workflow pressures, connectivity constraints, interface complexity, and headset comfort issues. Our results align with Wong and colleagues (54), who identified conditions under which staff were more likely to integrate VR into routine care. In line with the DTPB, our findings extend this work, and indicate that long-term integration of new technologies in LTC is unlikely when staff do not feel adequately resourced to deliver VR and perceive limited acceptability among residents. Given the potential benefits that VR can offer, our study highlights the need to explore solutions that improve user-friendliness for LTC staff and enhance the experience for residents, based on staff perceptions of what residents found acceptable and engaging. Suggested areas to enhance usability and usefulness in the literature include codeveloping VR implementation toolkits with LTC staff to facilitate troubleshooting, workflows, content selection and headset hygiene (54, 56, 81), mitigating headset discomfort for residents by selecting lighter devices with clear visuals and careful motion design (28, 54, 56, 82), and simplifying session involvement for residents through shorter sessions, seated use, simple headset controls, and staff facilitation (28, 54–56, 82).

Accordingly, our research team has been incorporating these suggestions into practice with the care teams at the CISSS-CA. As part of this process, data from our study highlighted a major issue related to comfort and usability of immersive VR headsets for older adults in the CISSS-CA context. VR participants frequently reported discomfort due to poorly adapted straps and the considerable weight of the devices. For example, the Meta Quest 2 with the Elite Strap can weigh as much as 550 g, proving burdensome for residents with limited cervical strength and endurance, as observations indicated that some residents could not maintain their head upright under this weight load. By contrast, recent devices demonstrate significant progress in miniaturization and weight reduction. The HTC Vive Flow (83) weighs 189 g, and the Bigscreen Beyond 2 (84) currently stands as the lightest immersive VR headset on the market at only 107 g, illustrating the potential for major improvements in ergonomics. As the Quest 2 headsets were already implemented in the CISSS-CA's care environments, our team developed an articulated arm system to mitigate discomfort associated with the weight of the headset.

Integrated findings from this study suggest that VR usefulness is likely conditional upon compatibility, which may be enhanced through individualized content (i.e., aligned with residents' histories, preferences, and sensory needs) delivered in calm spaces. LTC staff perceived these conditions as being linked to resident engagement, reminiscence, relaxation, and occasional reductions in BPSD. Consistent with recommendations by Chaze et al. (56) and Orr et al. (28), to prioritize calming and familiar VR content, such as 360-degree nature, relaxing audio scripts, or travel experiences, our findings indicate that LTC staff observed relaxation and reminiscence benefits for residents during VR activities with nature videos.

Furthermore, our results support that tailoring VR content to residents’ preferences and needs may improve engagement and outcomes. Staff highlighted the importance of fitting technology to residents' current level of functioning and matching content to their past experiences. This finding parallels research emphasizing the important of co-designing VR programs to reflect older adults' preferences and identities, and personalizing scenarios through the use of elements that enhance user immersion and appreciation, such as local environments, culturally-meaningful audio, and family photos (53, 56, 82). This study adds an implementation-focused perspective to the existing literature by showing, in a real-world LTC context, how personalization aligns with DTPB constructs that shape staff uptake (i.e., compatibility, perceived usefulness/ease of use, and self-efficacy) and offering practical targets such as content match, session pacing, and delivery context, to promote sustainable VR use in routine LTC care.

Our findings converged on the perceived necessity of environmental and organizational factors to enable VR to become a routine part of person-centred care. Facilitating conditions such as dependable access to online content, on-site technological help, standardized training with more than one staff member trained per site, adequate support from management, and sufficient time to learn how to use VR were frequently mentioned factors to promote better integration into LTC settings. As per the DTPB, these conditions align with staff's perceived ease of use, self-efficacy, and social norms via supervisor and peer influence, ultimately promoting the conversion of favourable attitudes into intention and use. They also align with implementation guidance from the scientific literature citing the benefits of appointing on-site “champions’, providing ongoing training, planning for Wi-Fi instability, and securing leadership buy-in for protected time and equipment (53, 54, 57, 81).

At the organizational level, staff described structural barriers typical to LTC, such as staffing pressures and lack of dedicated time. This finding reinforces the importance of management support, availability of ongoing training refreshers, and shared VR skill coverage between multiple professionals to prevent the reliance on a single “VR person” and the feelings of isolation that this staff member might feel. Our results are consistent with review articles calling for theory-informed implementation that moves beyond the surface identification of barriers by suggesting explicit strategies to improve VR adoption, such as training modifications, on-site champions, workflow integrations, and technical support (53, 57). Taken together, our findings may help to operationalize VR as a low-burden intervention for pain-relevant care, by making it more feasible for staff to deliver calming, individualized sessions that support relaxation and safe engagement in routine practice.

Ambivalence towards VR use in LTC appeared to be relatively common for staff when ease-of-use and facilitating conditions lagged, even with the presence of initially positive attitudes. Participants noted enthusiasm about VR's potential to engage residents, while concurrently voicing skepticism connected to workload pressures, technological reliability, resident suitability, and uncertainty about VR's contribution beyond usual care. Quantitatively, favourable attitudes were paired with only moderate perceived behavioural control, especially for resident selection and troubleshooting, as previously mentioned. These findings have also been raised in studies focused on VR use for post-stroke recovery (85). This overlap suggests that, although context-specific factors drive implementation, cross-cutting psychological factors (e.g., perceived usefulness of technology) also matter, but may not be enough to sustain VR adoption across diverse care settings. Qualitatively, difficulties were attributed to factors such as time constraints, internet connectivity issues, and technology complexity, leading to increases in perceived barriers and decreases in ease-of-use from staff's point-of-view. These perceived challenges are connected to theoretical and empirical evidence predicting reduced intention to use VR despite otherwise positive attitudes (53, 63, 66). In the broader LTC context, such ambivalence appears to be common due to the prevalence of constraints limiting consistent access to enabling conditions (53, 81).

Staff ambivalence was mitigated when support and visibility were strong. For example, residences with more than one facilitator trained to use VR, improvements to standard operating procedures, real-time sharing of session content (e.g., through projecting content onto screens for multiple viewers to see), and identifiable point-people to support technology use appear to be tied to smoother delivery and more consistent use, in-turn reinforcing self-efficacy and social norms with respect to VR adoption. To mitigate ambivalence, both the literature and our study's participants endorse starting with social and low-risk goals (e.g., building reminiscence, relaxation, small-group activities) to demonstrate value and build momentum (53). They also recommended drawing on cultural-familial resources to select meaningful locations and contributions for VR sessions, factors that have been shown to boost engagement and ease burden (53). Our study advances existing findings by moving beyond descriptive accounts of ambivalence to VR adoption by linking them mechanistically through the DTPB to specific modifiable factors observed in real-world LTC, such as perceived ease-of-use, self-efficacy, social norms, and facilitating conditions. Addressing these factors and co-developing solutions with knowledge users allows VR to become a more feasible, low-burden adjunct intervention to pain-care by enabling the routine delivery of calming, individualized experiences that support relaxation, positive mood, and scaffolded resident involvement.

This study has some limitations. The sample size was modest and situated within a specific LTC context with high neurocognitive disorder prevalence, limiting generalizability. Additionally, our sample involved convenience recruitment, potentially meaning that only LTC staff who showed the most interest in VR agreed to participate in both the training and the study. Furthermore, we were unable to collect information on the characteristics of LTC staff who declined to participate. This information would serve to enrichen the study's findings by building a more accurate profile of who potentially uses VR in LTC, vs. who is actually willing to share about their VR use. However, given the nuanced findings in the responses obtained, it appears that the impact of this sampling technique was relatively minor. It is also possible that because non-participating staff may have avoided the study due to negative experiences, our sample may not capture the full range of views. Including them might have yielded more critical perspectives. Another limitation is that we did not include perspectives from physicians, nurses, or nursing assistants who are typically involved in pain management in LTC. While our focus on frontline staff tasked with delivering VR programs to residents was intentional, future research should incorporate input from diverse clinical care perspectives to better understand how VR might complement pain-care strategies within interdisciplinary teams.

Questionnaire items were self-report and cross-sectional; some contained “unsure” responses, which are consistent with early adoption and variable training recency, but lead to reduced clarity regarding LTC staff's experiences. To address depth limitations inherent to survey-generated data, we complemented the dataset with qualitative analyses and a structured mixed-methods integration meeting, which involved research team members reviewing quantitative and qualitative findings together to confirm, expand, or identify divergences. Additionally, two separate two-hour online focus groups were conducted in April 2025 with ten LTC staff participants (five per group) who had completed the mixed-methods questionnaire analyzed in this study. These focus groups aimed to further test the robustness and transferability of themes and their findings will be compiled for future dissemination to healthcare, industry, and academic audiences. Finally, this study assessed a specific time-period of VR training and use. It does not account for developments in instruction and technology that have since been incorporated and which might have impacted LTC staff's perspectives on VR adoption.

However, this research is not without strengths. Our use of a convergent, mixed-methods design anchored in a theory-informed framework, enabled a more comprehensive interpretation of staff VR adoption alongside integration through the joint display of quantitative and qualitative findings. Second, it was conducted in a real-world LTC setting, enhancing ecological validity for later-life pain care. Third, we incorporated knowledge-user involvement throughout analysis and integration, which supported improved clarity in data collection materials and rapid, context-sensitive implementation actions (e.g., training updates, improvements to standard operating procedures). Fourth, our partnership model involving our neutral academic research team enabled us to link staff realities with organizational constraints and developer capacity to accelerate adoption. Fifth, VR has a favorable safety profile. Immersive sessions carry almost no serious side effects, making them a relatively low-risk intervention for LTC care.

An additional key strength of our scientific approach lies in the way we addressed challenges in collecting questionnaires. Based on recommendations from managers, we conducted in-person visits to meet staff, listen to their needs, provide training, and demonstrate how VR could be used with older adults. This strategy led us to visit 30 care facilities, which facilitated recruitment and encouraged participants to complete the questionnaires. These visits also fostered trust, helped us better understand potential challenges, and enabled the recruitment of additional participants for discussion groups, which will form the basis of future analyses and publications.

Beyond these immediate results, our approach, grounded in direct engagement with LTC facilities through on-site visits, staff training, and collaborative implementation planning, sets the stage for a dynamic and iterative implementation process. The data collected in this study represent a first snapshot that can later be compared to new datasets, in order to evaluate whether modifications and adaptations are having a sustained impact. This continuous cycle of data collection and comparison will allow us to assess how implementation evolves over time and to identify further opportunities for improvement. In future phases, our laboratory aims to extend this comparative approach to include other facilities and care contexts, examining what works, what does not, and how differences across environments shape outcomes. These analyses will also benefit from the company's own datasets collected in parallel, enabling a more comprehensive understanding of barriers and facilitators and fostering a truly collaborative, evolving model of VR integration in care for older adults in LTC.

In line with our iterative approach, preliminary findings from this study were shared with the CISSS-CA and Super Splendide in April 2025. During a similar timeframe, the Toujours Dimanche application was iteratively refined based on feedback gathered by the company since 2022. These refinements included broadening the content catalogue from 64 to over 700 videos, with capacity for personal/family content and site-specific uploads, and implementing technical updates to enhance connectivity and streamline the user interface. Our team's role has been focused on informing staff about these updates and showing them how to access additional free applications (e.g., MetaTV, YouTube VR, Theater Elsewhere) to broaden the range of personalized content available for residents. A group-mode option for VR activities using HDMI cables and tablets has also been taught to staff to support shared viewing with residents, familiarization with different content, and gentle participant activation. For pain care, these adaptations are linked to mechanisms such as positive affect, distraction, presence, relaxation, and graded engagement, that are likely to support comfort and safe movement (86, 87).

Simultaneously, Super Splendide has been updating its training and supports based on feedback collected from staff trained within the CISSS-CA as well as from other training experiences in the multiple establishments where their technology has been deployed. In this context, the developer has been pursuing an iterative improvement process, releasing updated versions of the application and refining adoption supports in response to observed barriers and facilitators. Recent changes include an asynchronous, interactive online training (3–4 h long, delivered in 10–15-minute micro capsules) available on demand to accommodate busy schedules and staff turnover, followed by a comprehensive webinar to consolidate learning. The adoption process has also been revised to address learning needs, practice time, logistics, and change-management requirements, with additional planning support for managers. Following these modifications, an updated training program was reintroduced in summer 2025 and has continued into fall 2025. In parallel, our research team contributed to these efforts by accompanying LTC staff during VR implementation, identifying teams interested in research collaborations, and prioritizing direct, hands-on presence.

This preparatory work has thus laid the foundation for two funded projects to be implemented in the coming months: One project focuses on mindfulness to reduce chronic pain through Version 4 of the Toujours Dimanche app, while the other promotes physical activity using gradual exposure as part of the “Move Without Fear or Pain” program, supported by the VitaMove application developed by our team. Without this prior identification of barriers and facilitators, pilot projects would likely have encountered major feasibility issues.