The study was carried out across 2 hospitals in China and the United Kingdom. The first hospital, West China Hospital, is in Chengdu, southwest China. The other is Chelsea and Westminster Hospital in London, UK. Both hospitals are publicly funded, and are among the largest ones in the local regions, providing comprehensive medical care to a high volume of patients, making them representative of large urban health care institutions in their respective countries. Although health care systems in China and the United Kingdom operate differently, some systematic challenges across the 2 hospitals are similar, such as staff shortages, increasing workload, and high burnout levels [27-30].

A volunteer sampling method was used for the survey. The link to the survey was spread to HCPs via email and group chats in WhatsApp and WeChat. To be eligible for the study, participants had to be over 18 years old and employed as staff members (both clinical and administrative) in one of the hospitals where the study was taking place.

Interview participants were recruited via the questionnaire. An option to enter their email address to be contacted for the interview; a separate interview information sheet and consent form were then emailed to participants.

The research has 2 parts in China and the United Kingdom. Studies in the 2 sites were approved by the Research Integrity and Ethics Committee of Imperial College London (22IC7803 and 22IC7585). The Ethics Committee on Biomedical Research of West China Hospital granted local ethics clearance for the research in China. Ethics approval was granted for the study in Chelsea and Westminster Hospital by the Health Research Authority of the National Health Service in the United Kingdom (reference number: 316935). Informed consent was obtained from participants before starting the survey and interviews. Participants were fully informed about the study's objectives, procedures, and potential risks. All data collected in the study was anonymized, and no personally identifiable information was linked to participant responses. No compensation was provided for completing the survey. However, participants completed the interview received a £5 voucher of appreciation for their time.

To our knowledge, this is one of the first studies investigating HCPs’ use and factors impacting their engagement with DMHIs. Based on their feedback, we identified demographical, personal, occupational, and cultural factors that impacted their engagement with DMHIs and provided implications for future research on DMHIs for HCPs. In the following sections, we elaborate on the engagement factors, compare our findings with prior works on DMHI engagement among other populations, and provide design implications.

Furthermore, building on the findings of this study and evidence from previous literature, we developed detailed design heuristics and practical recommendations for DMHIs targeted at HCPs, as outlined in Multimedia Appendix 3.

In this study, we identified substantial differences in HCPs’ engagement with DMHIs across cultures and ethnicities. Specifically, only 6% of HCPs in China had previously used DMHIs, compared with 28% of staff in the United Kingdom. Furthermore, within the UK sample, HCPs belonging to ethnic minority groups, specifically Asian and Black, were significantly less likely to engage with DMHIs. This concerning gap in DMHI engagement among ethnic minorities and people from low- and middle-income countries aligns with findings from previous studies [36,37]. One fundamental cause of such lack of engagement is that most mental health apps and platforms are developed in high-income countries for the majority of the White population [25,38,39]. This poses barriers and challenges for users from underrepresented groups to engage. In our study, HCPs in China struggled to find available DMHIs and were frustrated by the lack of quality of the existing products and platforms (subtheme 3.2). Such inequality in resources, along with language barriers and cultural incompatibility, were major factors identified in previous studies that hinder engagement from underrepresented populations [25,40,41]. More efforts are strongly needed to develop tailored interventions [42] and conduct cultural adaptation on existing DMHIs for underrepresented populations [25,43,44] to overcome this inequality and facilitate engagement among underrepresented groups.

Lack of mental health knowledge and low perceived need emerged as interconnected factors hindering DMHI engagement in both samples. Our results showed that knowledge of burnout interventions and a help-seeking history significantly predicted HCPs’ engagement with DMHIs. Many HCPs, without adequate knowledge and recognition of burnout, had normalized burnout symptoms in their work experiences (subtheme 1.1) and found it less necessary to adopt DMHIs for burnout as it is less severe and urgent (subtheme 1.2). Indeed, without essential literacy and awareness, motivating users to adopt DMHIs can be challenging [45]. This is especially true among HCPs in various stages of burnout. Previous studies have shown that the need to alleviate acute symptoms is a crucial motivation for users to engage with DMHIs [16,21,46]. For HCPs experiencing burnout, however, such motivation is often lacking, as burnout tends to gradually deteriorate over time rather than present acute issues requiring immediate relief [1,3]. Hence, to facilitate early recognition of burnout and motivate HCPs to seek support, systematic psychoeducation for staff is recommended for health care institutions, which was also called for by other studies [47,48]. Moreover, consistent with the design heuristics proposed by Peters [49], we recommend future studies on DMHIs for HCPs to provide adequate information on the necessity and benefits of interventions across various stages of user experiences. This can help form a meaningful rationale among HCPs who are hesitant about engaging or just started using DMHIs, which can effectively motivate them to increase and sustain their engagement [49,50].

As revealed by previous literature, engagement is critical for effective DMHIs [19,51,52]. However, this can be challenging for HCPs as their schedules are intense and unpredictable. HCPs in China and the United Kingdom were reported to undergo issues like staff shortages, increasing workload, and work pressures [29,30]. This is also reflected in this study, HCPs struggled to allocate a set time to engage with DMHIs (subtheme 2.1). For them, structured push notifications, a common engagement strategy applied by prior studies [53-57], may be less effective. However, just-in-time notifications, that is, adaptive and personalized notifications based on user behavior patterns or their locations [58], demonstrated better engagement outcomes [58,59] and can be an ideal engagement facilitator for HCPs. Although HCPs’ schedules vary significantly, certain times (morning brief, between shifts, lunch breaks, etc) can be suitable for specific cohorts. Furthermore, as HCPs’ work is highly location-based, we suggest using context-based reminders to facilitate engagement. For instance, just-in-time notifications that prompt when HCPs enter spaces used for relaxation and on-site visual cues like posters and stickers placed within common spaces could be effective nudges for DMHI engagement. Similar to the design strategies proposed by previous studies that support establishing an engagement routine [60], such context-based reminders could promote the integration of HCPs’ existing routines with DMHI engagement.

Digitization in health care systems is another factor that cannot be ignored. Our quantitative data revealed that a considerable proportion of participants (88% in China and 79% in the United Kingdom) reported using computers frequently (at least once an hour) in their jobs. Being immersed in a highly digitized workplace, participants in both countries reflected on the additional effort and learning curve entailed by systems like the electronic health record. They demonstrated hesitance and reluctance to engage with more digital systems like DMHIs, with a strong need for a “screen break” (as shown in the survey feedback and Subtheme 2.2). This is consistent with recent studies on digital stress and overload among HCPs [61], often caused by digital systems like electronic health records that have been ubiquitously applied in health care settings [62-65]. Concerns of being further overwhelmed and complicated by technology contribute to their hesitation and lack of motivation towards DMHIs. Therefore, we suggest that future work apply best practices for usability and accessibility to avoid causing an extra burden on HCPs’ work experience, a point also advocated by other researchers [7,49,66]. We also recommend that researchers and designers consider HCPs’ digital literacy (especially for older, non-tech-savvy staff, as suggested by prior studies [64,67] and deliver DMHIs via technologies familiar to the targeted cohort of HCPs.

Many participants indicated that DMHIs often lacked relevance to health care workers and their struggle with job burnout. Similar observations have been made in other studies, where users strongly valued personal relevance, highlighting the importance of tailoring to DMHIs [16,17]. To provide relevant content and design for HCPs, we recommend incorporating language and daily scenarios that staff are familiar with. Moreover, using participatory design methodologies [68] to involve HCPs and relevant stakeholders throughout the developing process is also recommended to ensure the final outcome embodies the values and needs of HCPs, a perspective shared by many DMHI researchers [69-71].

Besides, staff also mentioned a need for human-based support, highlighting the importance of feeling understood and related to their experiences with DMHIs (subtheme 3.3). Prior studies also suggested the benefits of providing human-based support or guidance, like expert consultation and coaching, to promote user engagement [36,52,56,72,73]. DMHIs with certain levels of human support generally result in better engagement and clinical outcomes than self-guided DMHIs, as reported by recent reviews [19,73]. However, expert guidance and regular meetings with trained practitioners are less accessible and scalable, especially for large populations like HCPs [19]. We suggest future studies probe the potential of accessible human support and guidance, like text-based, asynchronized human support, or human-like interactions supported by conversational AI. Indeed, DMHIs using chatbots and embodied conversational agents have the potential to facilitate user engagement by providing personalized and empathic interactions [74,75]. Nonetheless, more studies are needed to address the ethical and methodological shortfalls [76,77] and tailor the algorithms for unique cohorts like HCPs.

Cultural differences are influential to HCPs’ engagement with DMHIs, as revealed by our qualitative data. Stigma on burnout and mental health issues was commonly identified among Chinese participants, echoing research evidence on the prevalence of stigma in China [78,79]. The self-stigma among Chinese participants affected their willingness to admit mental health issues or adopt DMHIs for support (subtheme 4.1). Previous studies among culturally diverse populations also emphasized the impact of self-stigma on DMHI acceptability and engagement [21,41]. Furthermore, Chinese HCPs in this study often attributed mental health conditions to personal and moral virtues, demonstrating a moral obligation to remain resilient at work and to do their best to avoid being affected by stress and burnout (subtheme 4.1). Such moral experiences are closely intertwined with culturally rooted mental health stigma and are commonly observed within Chinese communities and other cultural groups [80,81]. Maintaining the ability to work is a crucial coping strategy among Chinese populations against mental health stigma and sociocultural discrimination, as work serves as a key factor in preserving social status and fulfilling familial and moral obligations [28,81]. We suggest future projects for stigmatized populations establish nonjudgmental feedback mechanisms and reframe DMHIs as tools for enhancing well-being, occupational resilience, or improving professional abilities and skills, a strategy also indicated by a recent study on DMHIs [67]. Although researchers suggested shifting focus away from the dichotomous categorization of wellness and health within digital mental health technologies [82], we argue that strategically labeling DMHIs can be beneficial in a stigmatized context.

Peer support and community-based elements in DMHIs could facilitate user engagement by fostering a sense of collective care and support [83]. However, as suggested by our findings, this must be executed cautiously, particularly in environments where mental health is still stigmatized and less discussed openly. Similar to Chinese HCPs in this study, stigmatized users are often reluctant to participate in community support activities [84,85], due to sociocultural factors like public stigma, lack of mental health education, and traditional cultural values [86-88]. We suggest designers build meaningful connections among users while not imposing pressure to share or disclose personal information, which is crucial for nurturing social connectedness and support [49,50]. Maintaining a certain level of anonymity and taking extra caution with user privacy is also advisable.

This study involved 2 major hospitals in key cities in China and the United Kingdom. Given the differences between hospitals of various sizes and geographical locations [89], our findings may not fully represent other medical populations in these countries. Our sample sizes for the survey and interview (around 5% of survey respondents) are relatively small, which may limit the generalizability of our findings. Besides, the gender imbalance between the 2 cohorts (χ²₃=10.907, P=.01) may have also impacted our findings, as gender can be an impactful factor on DMHI engagement [16].

The volunteer sampling method we employed could also introduce a source of response bias. HCPs interested in mental health and DMHIs may have been more likely to take part. This may lead to the study results not being fully generalizable to the broader population of HCPs in China and the United Kingdom. However, the triangulation of surveys and interviews helps somewhat mitigate this potential bias. Furthermore, as a cross-cultural study, we only investigated HCPs in 2 countries. Future studies should broaden the scope to include participants from other countries and regions to enhance the robustness of the evidence.

Participants were asked about DMHIs as a general, nonspecific concept rather than specific digital tools (eg, apps, web-based platforms, mixed-reality). This was intentional, as our focus was on overall engagement with various DMHIs. However, this broad framing may be less effective in helping HCPs navigate different DMHI concepts. More studies on engagement with specific types of technologies could be beneficial.

Another limitation of this study is the use of single-item questions assessing some constructs like engagement with DMHIs and willingness to adopt them. Although such questions were adopted to ensure the time-efficiency of the survey as HCPs are known to have hectic work schedules, they may lack psychometric stability on measures like motivation and willingness. Future research should consider using multi-item scales to enhance the robustness of measurement.

This study used a mixed method approach to explore HCPs’ level of engagement with DMHIs and identified key factors that impact their attitudes and engagement. We found that there is a considerable gap in DMHI engagement among ethnic and cultural minority groups, potentially due to the lack of availability and content quality of DMHIs for minority groups. Lack of time to use DMHIs, lack of mental health literacy, low perceived need, lack of perceived relevance, and lack of human-based support were common factors impeding HCPs’ engagement with DMHIs. Cultural elements like stigma and moral attributions also impacted HCPs’ engagement and preference for DMHIs. Overall, this study contributed to the limited evidence on the experience and engagement with DMHIs among HCPs and shed light on how future efforts can design and deliver DMHIs that are accepted and engaged by HCPs.