A mixed methods implementation case study design [17] was adopted to explore implementation and uptake of the Beyond Silence platform across 7 health care organizations. A case study approach provides an effective way to study contemporary phenomena within a real-life context, particularly when it is difficult to separate the case from the context and can lead to new explanations that consider the complexities of place and time [18]. Exploring multiple case study sites enabled cross-case comparison across different workplace contexts. A multistage mixed methods framework was adopted, starting with qualitative interviews to explore the perceptions of organizational champions regarding the context for app implementation [19]. This was followed by the collection of quantitative app usage data. Finally, follow-up interviews were conducted with the organizational champions, during which they were asked about the individual and contextual forces that shaped trends in app usage within the workplace context [20]. Data collection and analysis were informed by CFIR, one of the most commonly used frameworks to assess contextual factors affecting real-world implementation [21].

Ethics approval was obtained through the Hamilton Integrated Research Ethics Board (HiREB number 14740), and all participants provided written consent to participate. In order to maintain privacy and confidentiality, the names of participating organizations and organizational champions are disguised. All app usage data are reported in aggregate; app users were anonymous. Organizational champions received a CAD $100 (US $70.95 as of July 24, 2026) gift card honorarium for participating in each champion interview. Peer supporters in each organization were not paid for their role; however, external peer support providers were provided a stipend of $55 per 3-hour shift. Peer support providers who participated in the focus group discussions were invited to enter a draw for one of 2 CAD $50 (US $36.50) gift cards of their choice.

An advisory team provided input throughout the project, including 4 academic leaders in workplace mental health, mHealth technology, and trauma-informed care, and 6 knowledge user partners representing national associations of nurses, personal support workers (PSWs), social workers, and long-term care workers, a national mental health service provider, and a provincial health and safety association. The team met 3‐4 times a year to provide input at key stages in the project and to facilitate communication within their extensive networks.

Beyond Silence (see Figure 1) is a digital platform (including a mobile app for frontline health care workers and an organizational portal for peer supporters and organizational champions) designed to provide accessible, anonymous, and peer-led mental health support to Canadian health care workers. It builds on a decade of successful research collaborations with health care stakeholders to identify and address their unique mental health and support needs [19,20]. Development was informed by findings from a randomized controlled trial of an in-person workplace mental health training program with over 200 health care workers [21]. The in-person training was effective but resource-intensive and not accessible to all workers; therefore, an mHealth platform was explored as an alternative approach.

Development of the Beyond Silence platform (2017‐2019) was supported by a Canadian Institute of Health Research (CIHR) development grant, and informed by the Knowledge-to-Action Framework, a structured approach for knowledge creation and a 7-phase cycle that moves knowledge into practice [22]. In addition, best practice principles of mHealth design were considered, including engagement (customization and interactivity), functionality (ease of use), aesthetics, information quality, and data security [23]. Additional funding (2022‐2024) from the Public Health Agency of Canada (PHAC) allowed for additional app enhancements. Core functionalities of the app for frontline workers include (1) access to evidence-informed information customized for health care workers (articles, infographics, video, and audio clips) on a range of topics from mindfulness to moral injury; (2) links to self-management tools, including a wellness toolbox, wellness check-ins, and goal-setting feature; and (3) private, secure links via text or phone to a range of health care providers trained to offer peer support. App users can browse the biographies of the available peer supporters within and outside their organization, choose how they want to connect (phone or text), click on who they want to connect, and the time at which they want to connect (either within the next 90 minutes or at a scheduled time that their preferred peer is available). A notification is then sent to the identified peer regarding the anonymous request for support and a secure, private connection is made through the app for the health care worker seeking support.

In addition to the app for frontline health care workers, the platform includes an interconnected web-based portal for trained peer support providers to respond to requests for support. The online portal provides space for peer supporters to upload their pictures and short biographies (visible to the app users), as well as sign up for 3-hour shifts, respond to peer support requests (by text or phone), record digital notes on their encounters, access resources to support their work (eg, crisis protocols and contact details for relevant resources), and connect with a clinical supervisor and/or other peer supporters. Although the peer support providers were visible to frontline app users, the person accessing support was anonymous (digital ID number). A total of 156 health care professionals were trained to offer peer support in the app over the course of the study, including 88 from participating organizations and 68 central peer supporters who were external to the participating organizations and accessible to all. Details of the peer support training are reported in a separate paper [24].

The Beyond Silence platform also included an administrative portal where organizational champions could upload local resources, manage their internal peer support provider list, send notifications, and track overall patterns of app use. It is important to note that the app was available through app stores, but an organizational code was required to download the app, since the app is customized for each organization (including local resources and peer support access). As noted above, there was an onboarding process for the participating organizations and peer supporters, which included training on risk management and data security protocols.

Health care organizations were recruited through established networks of the project advisory team. Interested organizations were asked to complete an application form to assist in screening for diversity across size, geographic location, workplace setting, and type of health care service. To be eligible to participate, organizations required: (1) a project agreement letter signed by organizational leaders, (2) at least one organizational champion designated to liaise with the research team, and (3) at least 2‐3 internal staff members who agreed to be trained peer support providers (numbers varied based on size and resources available within the organization).

The goal was to recruit 6‐8 organizations to explore how implementation varied across workplace contexts. Purposive sampling was conducted to ensure diversity in size (small/medium/large), geographical location, setting (rural/remote and urban), type of service (long-term care, acute care, tertiary care, and primary care), and at least 1 organization where employees spoke French (since the app is available in both French and English). A total of 22 organizations expressed interest in the study; however, 14 of these organizations did not have the capacity to participate. For organizations that declined to participate, the primary reason for declining was documented (typically due to limited resources to support the app implementation and evaluation).

Implementation setting and scope A champion-led, organization-embedded implementation strategy was used to introduce Beyond Silence within each participating organization. Champions were individuals representing participating organizations who approached us to be involved in the study. They were responsible for liaising with the research team, communicating about the app to employees, and acting as the contact point for any questions or concerns among staff. Each site designated one or more organizational champions to serve as local implementation leads and primary points of contact.

Champions represented a range of organizational roles, including senior leaders, health and safety coordinators, and frontline clinicians, including 20 women and 6 men in both clinical (n=13) and nonclinical (n=13) positions. As noted in Table 1, there were typically 1‐3 champions in each organization who agreed to communicate with employees about the app and to be interviewed about their insights at the outset and conclusion of the project. One large multisite organization had 11 champions volunteer to participate in the study.

The target population for implementation was determined by each organization based on local capacity and readiness. Most included all employees, but 3 organizations chose a specific subgroup (Organization 1–nurses, Organization 2A–laboratory technicians, and Organization 4–PSWs). Recruitment relied on existing organizational communication channels and included distribution of standardized promotional materials (posters, email scripts, brochures, and branded materials) containing information about the app, a QR code for download, and a unique organizational access code. Employees were invited to voluntarily download and use the app for the implementation period.

Implementation included organizational onboarding, app deployment to frontline staff, and integration of trained peer support providers accessible through the platform. Uptake, engagement, and variability were tracked across organizational contexts over a 4-month period. Guided by an implementation framework (CFIR), contextual factors, resource requirements, and implementation processes were examined to inform scalability and sustainability.

Data collection focused on prespecified implementation outcomes and used a mixed methods approach combining app usage analytics and qualitative interviews with organizational champions. Evaluation focused on reach, adoption, implementation, and patterns of use rather than effectiveness or health impact.

Quantitative implementation metrics were assessed using real-time, aggregate app analytics captured through the administrative portal. App users remained anonymous and were identified only through a unique user ID. Metrics were selected to characterize adoption and engagement and included the number and proportion of employees who downloaded the app, the number of times the app was opened, mean number of app opens per user, and frequency of use of key features, including content views, goals set, wellness check-ins, and peer support appointments booked. Additional descriptive data captured peer support outreach modality (text or call) and choice of peer supporter. App use data were examined monthly within each organization to describe trends over the 4-month implementation period and to support cross-case comparison; inferential analyses were not conducted.

An update to the app occurred during which onboarding questions were added, asking app users to respond to optional demographic questions regarding age, gender, years of professional experience, and health care role (clinical/nonclinical). Given the timing of this change, only 5 of the 7 organizations included this demographic data. As such, demographic data were not collected or reported for organizations 1, 2A, 2B, and 3.

The study is registered on ClinicalTrials.gov (NCT05514093) as an implementation and evaluation study of a mHealth intervention designed to support the mental health of health care workers in Canada. However, while multiple mental health outcome measures (including standardized pre-post surveys of depression, anxiety, PTSD, burnout, and help-seeking behavior) were included in the study protocol, survey response rates at follow-up were too low to permit meaningful quantitative analysis. As a result, the mental health survey data are not reported, and no conclusions can be drawn regarding changes in mental health outcomes based on these measures.

Qualitative data were collected to explore contextual factors influencing implementation. Interviews with the organizational champions included baseline interviews conducted prior to app implementation (n=19 interviews with 27 individuals) and follow-up interviews conducted at least 3‐4 months following implementation (n=18 interviews with 26 individuals). Interviews were conducted virtually, audio-recorded, and transcribed verbatim, with durations of approximately 60‐90 minutes. Some champions were interviewed individually, but others were interviewed in pairs or groups of 3. Interview guides with the champions were informed by the five dimensions of the CFIR, including questions exploring (1) perceptions of the intervention (characteristics of the app), (2) the outer setting (eg, societal context and policies), (3) the inner setting (eg, organizational culture and access to resources), (4) individuals (role of leaders, peer supporters, and characteristics of frontline workers), and (5) the implementation process [16]. Specific questions within each domain were developed using the CFIR interview guide tool. Since organizational champions were embedded within the organization, their insights enabled identification of critical factors that shaped uptake of the technology.

No predetermined numerical targets for adoption or engagement were specified. Given the exploratory nature of the implementation and the lack of established benchmarks for uptake of this type of workplace mental health and peer support app in comparable health care settings, the study aimed to descriptively characterize implementation outcomes and variability across organizations rather than assess performance against predefined thresholds.

Descriptive data regarding app use included monthly reporting within each organization to examine trends in downloads and use of the different features. Cross-case comparisons were also conducted to review similarities and differences in usage patterns across organizations, but inferential statistics were not calculated.

Organizational champion interview data were transcribed and coded, with initial coding categories informed by the CFIR. Coding was an iterative process that began with reviewing transcripts, categorizing them by the 5 CFIR domains (deductively), and then inductively identifying the key issues within each domain. At least 2 team members coded each transcript, initially coding 2‐3 transcripts independently, then meeting together to compare codes and refine the codebook to ensure consistency. A series of team meetings was held to discuss emerging analytic ideas and revise the codebook to capture key facilitators or barriers to implementation across organizational contexts. Strength and valence ratings of the key forces within each organization were conducted in order to facilitate cross-case comparison [24]. Identification of key forces represented comments across multiple organizations, with quotes selected to illustrate, considering representation across all participants. Dedoose (SocioCultural Research Consultants, LLC) software [25] was used to manage the qualitative interview data. The qualitative data were used to explain the app use trends and facilitate actionable insights to optimize implementation.

Strategies to ensure methodological rigor included triangulation of data sources (app use metrics and interviews with multiple champions over different time points in each organization). Triangulation strengthened in-depth understanding of the organizational context and implementation forces within and across organizations that shaped app use, leading to data saturation across categories and understanding of the contextual forces shaping objective patterns of app use. In order to ensure coding consistency, biweekly meetings were held with the research team to review the coding and to engage in peer debriefing. Meetings between coders were scheduled to calibrate approaches and refine definitions in the codebook until there was at least 85% coding consistency. Reflexive analysis was strengthened through analytic memos that were discussed in the peer debriefing meetings. CFIR was used as a sensitizing framework, but the iterative, inductive analysis process led to nuanced understanding of the key issues shared by participants. Member checking was conducted through sharing preliminary analyses with champions via draft organizational reports, then exploring their reflections on the analysis in follow-up interviews.

The study findings are presented in three parts: (1) an overview of the participating organizations and app users, (2) descriptive metrics regarding patterns of app use, and (3) qualitative analysis of champion interview data regarding factors shaping app use (informed by the CFIR framework), considering the social and organizational context (outer and inner setting), response to the app innovation, influence of organizational stakeholders, and impact of the implementation strategies.

A total of 7 diverse health care organizations participated in the study, joining at different time points. As outlined in Table 1, 2 of the organizations were large (over 10,000 employees), and one was small (n=55 employees). Organization 2 was a large professional association that engaged in the study at different time points, initially with a subgroup of workers (2A/2B) and then later as the full organization (2C). Their participants were members of the association (paying membership dues), rather than employees, and the app was offered as a member benefit. Organizations 1 and 4 chose to focus on a subgroup of health care providers within the organization (nurses and PSWs, respectively), but other organizations provided access to all employees (clinical and nonclinical). Organizations 3, 6, and 7 were multisite facilities, and organizations 1 and 4 provided community care with no on-site services. The number of champions and peer supporters varied across sites. As such, there were various implementation contexts for implementation. The app for frontline workers was made available to a total of 32,753 health care workers across these organizations.

Demographic data were collected from app users in organizations 2C, 4, 5, 6, and 7 via onboarding demographic questions that were added in June 2023. As noted in Table 2, app users in these organizations (n=840) included 77% (n=650) who identified as women, 21% (n=173) as men, and 2% (n=17) as nonbinary or chose not to disclose. Just over half of participants were between the ages of 30 and 49 (n=458, 55%) years. In terms of professional experience, 41% (n=348) had ≥16 years, followed by 15% (n=129) with 6‐10 years. Most participants held clinical roles (n=505, 60%), while 24% (n=205) were nonclinical.

A total of 1066 employees downloaded the app across all organizations over the 4-month implementation trial. As noted in Table 3, the number and proportion of downloads within each organization ranged from <2% (n=108) in the largest organization to >45% (n=25) in the smallest organization. The mean download rate across organizations was approximately 11.6% (SD 15.2) of employees, but the overall downloads represented only 3% of the total number of eligible participants.

User engagement, measured by app opens and feature use, also varied across organizations. As noted in Table 3, the app was opened 3127 times over the course of the study, ranging from a total of 85 opens (45% of employees) in the smallest organization (Organization 5) to 1324 opens (4% of employees) in the second largest organization (Organization 6). Each downloaded app was opened a mean of 3.22 times (range from 2.1 to 4.9).

Given the sample size, normal approximation methods were used to calculate 95% CIs for proportions. A total of 309 app users viewed at least one content item, representing 30% of users (95% CI 27.2%‐32.9%). When viewing the content, users could provide a rating from 1 to 3 stars. There were 105 user-submitted content ratings, with a mean rating of 2.58 (SD 0.68) out of 3, corresponding to 86% of the maximum possible score. The most frequently accessed content by app users was an article titled “6 Mindfulness Exercises,” which received more than twice as many views (n=92) as the next most accessed item (n=35).

Engagement with other app features included (1) wellness check-ins initiated by 25% of app users (95% CI 22.3%‐27.7%), representing between 10% and 22% of the app openings; (2) goal setting initiated by 13% of app users (95% CI 11.0%‐15.2%); and (3) peer support outreach was accessed 94 times (56 calls and 38 texts) by 40 unique app users, with a wide range from 0 (in the smallest organization) to 49 (in the largest organization). Appointments were booked with 38 different peer support providers.

The study findings point to an interesting paradox regarding adoption of the Beyond Silence platform. On the one hand, we heard many positive comments about the need to support frontline health care workers and how the app met many needs for a customized, confidential tool to increase access to relevant information and peer support. On the other hand, the overall app use was relatively low (download rate of <10% in 5 of 7 organizations; average of just over 3 opens per user), which was attributed to a range of barriers to engagement.

Theories about technology adoption can shed light on the findings related to app engagement. The technology acceptance model (TAM), for example, focuses on perceived usefulness and perceived ease of use, and how these impact intention to use the technology [23]. In our study, organizational champions noted that the app was easy to use and that it was very useful in enhancing their current employee support initiatives. These beliefs, however, did not necessarily translate to frontline workers. Ease of use for frontline workers was complicated by poor internet connections in some organizations or inability to use their phones at work, and by lack of time and private space to download and use the app. Furthermore, frontline health care providers may not see the usefulness of the app given that they often ignore their own support needs in their effort to help others, and by their skepticism, lack of energy, and time to invest in new initiatives. As such, they may be more reluctant to use a digital platform to access formal peer support given that it represents a departure from simple, informal interactions with peers and requires time and effort to learn something new.

The unified theory of acceptance and use of technology (UTAUT) is another model, which builds on the factors outlined in the TAM model to expand understanding of forces influencing adoption. In addition to the focus on effort expectancy (perceptions re: ease of use) and performance expectancy (the belief that using the app will increase personal productivity), this model sheds light on the important role of social influence (beliefs that key people expect them to use the app) and facilitating conditions (belief that the necessary organizational and technical infrastructure exists to support usage) on intention to use and actual use of technology [24]. In our study, there were differences between organizations regarding both social influence and facilitating conditions. In the smallest organization, which had the highest proportion of downloads (45% in Organization 5), the organizational champion had many personal connections with their colleagues (social influence), and they set aside time during a staff meeting for employees to learn about the app and try out the different features (facilitating conditions), thereby increasing initial employee engagement. In contrast, the large professional organization with the lowest proportion of downloads (2% in Organization 2) had limited personal connection with members since members worked in diverse organizations across the country. Furthermore, this organization did not have the infrastructure to actively invest in supporting app usage, using largely passive means of communication about the opportunity to participate (initial but no follow-up emails and occasional posts about the app in the electronic membership newsletter). These implementation conditions at an organizational level with limited contact and few active engagement opportunities clearly had an impact on employee engagement.

Finally, the framework outlining NASSS of health and care technology provides additional insights to consider given that the Beyond Silence platform represents a complex intervention that is implemented at an organizational level [14]. According to the NASSS framework, there are 6 main sources of complexity that need to be considered in understanding technology implementation and uptake, namely the condition, the technology, the value proposition, the adopter system (eg, staff and patients), organization implementation, and the wider (institutional and societal) context. In the case of the Beyond Silence platform, mental health “conditions” can be considered complex in that they are often insidious, unpredictable, and stigmatized, and therefore can be difficult to identify and manage. Health care workers may fail to perceive or acknowledge their own psychological distress and often avoid seeking support due to the stigma associated with mental health issues [21,26]. The technology itself was noted to be valuable, but it is available on an “as needed” basis with no requirement for initial or ongoing use. As noted in a systematic review of eHealth interventions with health care professionals, smartphone apps have the advantage of convenience, but they may be either too flexible or too subtle to prompt regular engagement [14]. Another area of complexity relates to the value proposition for frontline workers. Although there was general buy-in from leaders, the value was not necessarily evident for frontline workers given the time, energy, and trust required to adopt something new. Organization implementation was also complex given the challenges in connecting with workers who may work different shifts, across different sites, and have little time available to download or use new technology. Understanding and addressing these dimensions of complexity is critical to creating an effective implementation plan.

The insights provided by these technology implementation theories reveal that low adoption rates may not be a failure of the technology itself, but rather a failure of implementation. There is a growing body of literature on e-mental health interventions in the workplace, including several systematic reviews [27,28], and one umbrella review of 14 review articles published between 2014 and 2023 [13]. The review studies explore a range of web-based and app-based tools, with many rooted in principles of cognitive behavioral therapy (CBT) and mindfulness, including studies that focus specifically on health care workers. Overall, there appears to be a positive impact of these tools on reducing mental distress and improving employee mental health, with a clear value proposition related to accessible, customizable tools that can expand the reach of workplace mental health initiatives [6,13,28]. These studies underscore the value of the technology itself in the context of the workplace.

A notable gap in the literature, however, is in examining implementation processes that shape uptake and impact. Studies of digital mental health interventions note variability in engagement across studies, with a paucity of consistent data to track adoption and engagement, making it difficult to conduct comparisons across different types of technology innovations and workplace contexts [29,30]. As such, this study contributes important data regarding implementation of an app-based peer support intervention, comparing implementation across a range of organizational contexts in the health care sector. The focus on an upstream mHealth intervention, rooted in self-management and peer support, is also unique, given that many studies focus on therapeutic apps (cognitive behavioral therapy and mindfulness).

Based on analysis of the key facilitators and barriers noted in this study, as well as principles of implementation science theory and practice noted above, there are several recommendations to consider in order to optimize future uptake of workplace mental health apps such as the Beyond Silence platform within the context of health care workplaces. The recommendations are also based on a review of several implementation checklists and guidelines regarding strategies to foster implementation [31], including the Expert Recommendations for Implementing Change (ERIC) [32] and the European Platform to Promote Wellbeing and Health in the workplace (EMPOWER) checklist, which specifically examines strategies to foster adoption of occupational e-mental health interventions [33].

This study provides an early examination of mental health app implementation across a diverse range of health care organizations, informed by the CFIR framework and consideration of a range of technology adoption models. There are, however, several limitations, which constrain the strength and generalizability of the findings. A 4-month implementation period provided an opportunity to observe initial short-term patterns, but may not have been sufficient to assess longer-term integration into organizational culture. The follow-up period revealed the challenges associated with maintaining long-term engagement, highlighting the importance of iterative refinement and continued evaluation. Studies of EAP use, for example, report low annual usage rates but significantly higher uptake over a 5-year period, reflecting the time required for integration into organizational culture and workflows [42]. These findings underscore the importance of examining cumulative use over time, with at least a 6‐18 month implementation period. Future research should explore extended longitudinal studies to assess delayed impacts and long-term behavioral changes. As such, the present findings should be interpreted as preliminary, and future studies should use longer observation periods to capture delayed adoption and sustained behavioral change.

The sample of 7 participating organizations also limits generalizability. Although they represented diversity in size, function, and location, all were situated within the Canadian health care system and represented a time when formal peer support and technology to support workplace health were relatively new and a time when the health care system was still recovering from the impact of the pandemic. International contexts may vary in terms of policy environments, workplace cultures, and the maturity of peer support infrastructures.

Data were only collected from organizational champions regarding why they thought that employees downloaded and engaged with (or did not engage with) the app. As nonadopters represented a significant proportion of eligible staff, their perspectives, barriers, and decision-making processes are unknown. This limits understanding of overall receptivity, organizational fit, and determinants of uptake. In addition, the absence of a comparison group restricts the ability to infer causal relationships between implementation strategies and observed outcomes.

The focus of the study was on implementation but did not track mental health or workplace outcomes of the intervention. Future research should incorporate outcome evaluation alongside implementation assessment, include comparison groups, extend follow-up periods, and gather data from both adopters and nonadopters to produce more conclusive evidence regarding effectiveness, reach, and generalizability. There is also a need to track the cost-effectiveness of this intervention approach to help employers to reflect on the value of this type of digital support intervention in their workplace context.

This study provides important insights regarding implementation of the Beyond Silence peer support platform for health care workers. The findings point to a relatively slow adoption of the technology, attributed in part to the stigma and silence surrounding mental health issues, to competing demands on attention and time, and challenges with initial and ongoing engagement of frontline workers. At the same time, the findings also point to strong mission alignment for organizations that are looking for tools to support their frontline workers, and strong support for the opportunity to access high-quality customized information and resources for health care providers and to access tools to facilitate anonymous connections to trained peer supporters. We argue that understanding and addressing issues related to implementation is a necessary precondition to optimize impacts on mental health and workplace outcomes. Inspiring frontline workers to take the time and energy to access resources and peer support via the app requires reflection on potential barriers to change, and innovative approaches to build trust and inspire new ways of responding to mental health issues. Longitudinal research is needed since time is needed to study integration of the technology into workplace policies and practices.