Using digital technology to promote physical health in mental healthcare: A sequential mixed‐methods study of clinicians' views

Recent years have seen innovation in ‘mHealth’ tools and health apps for the management/promotion of physical health and fitness across the general population. However, there is limited research on how this could be applied to mental healthcare. Therefore, we examined mental healthcare professionals' current uses and perceived roles of digital lifestyle interventions for promoting healthy lifestyles, physical health and fitness in youth mental healthcare.

People with mental illness have significantly elevated risk for cardiovascular and metabolic diseases, due to a myriad of factors including insufficient access to physical healthcare, side-effects of psychotropic medications, and heightened rates of behavioural factors such as smoking, poor diet and sedentary behaviour (Firth et al., 2019;Solmi et al., 2020).As these behavioural risk factors for cardiometabolic diseases are present in young people from the onset of mental illness (Carney et al., 2016;Stubbs et al., 2016) adolescence/young adulthood presents a critical period for improving health behaviours to prevent cardio-metabolic diseases from arising in order to reduce premature mortality (Firth et al., 2019;Perry et al., 2021).
Currently, there is increasing investment in remote, Digital Lifestyle Interventions (DLIs; defined as interventions to improve health which are delivered digitally, for example via websites, or on mobile applications) in the general population.However, the role of DLIs for people with mental illness is only just beginning to be explored as a novel solution to providing low-cost and scalable interventions for physical health promotion in mental health services (Firth et al., 2019;Torous et al., 2021).This could be especially feasible for Young People with Mental Illness (YPMI), as younger people are perhaps more likely to own and use mHealth technologies (Firth et al., 2016;Firth & Torous, 2015;Naslund & Aschbrenner, 2021;Torous et al., 2021).A recent study found YPMI perceived digital interventions for self-management of mental healthcare as empowering and acceptable, though there were some concerns around access to digital interventions (Berry et al., 2019).Unfortunately, mental healthcare services are often not sufficiently resourced to properly address physical health needs (Bailey et al., 2019).
Whilst previous studies have explored the acceptability and views relevant to mHealth and DLIs in YPMI themselves (Bucci et al., 2018;Dewa et al., 2019;Naslund & Aschbrenner, 2021), the implementation of mHealth-based interventions within youth mental healthcare services will ultimately be dependent on the Mental Health Care Professionals (MHCPs) responsible for delivering them.Although previous studies have also explored MHCPs views on using mHealth in practise (Pierce et al., 2016;Schueller et al., 2016;Sweeney et al., 2021) or to provide self-manage mental illness (Berry et al., 2016;Berry et al., 2017;Bucci et al., 2019).Despite this, there is a dearth of research into MHCPs views on DLIs for promoting physical health amongst YPMI specifically.MHCPs insight into the challenges of using mHealth in mental healthcare settings, will be crucial in the implementation of DLIs.Therefore, this mixed-methods study aimed to examine MHCPs perspectives on their current uses, potential and implementation issues surrounding DLIs in YPMI.

| Study design
This mixed methods study was grounded in a pragmatist approach, which embraces a plurality of qualitative and quantitative methods and focuses on 'what works' to solve research problems above all else (Creswell, 2003;Crotty, 2020;Tashakkori & Teddlie, 2003).An explanatory, sequential design comprising two phases (Figure 1 integration and use (see Supplementary material 2).Interviews were audio-recorded and transcribed verbatim with additional written consent from interviewees (duration: 26 to 79 min).

| Data analysis
Descriptive statistics (counts, percentages, means and standard deviations) were used to describe the baseline characteristics of participants and to report levels of agreement with the quantitative survey items.
An First, author CS familiarized herself with the data by reading three interviews.After familiarization, an interview-coding scheme was developed, which was reviewed and discussed with other members of the research team.Following agreement from the research team, CS familiarized themselves with the remaining transcripts and additional codes were identified.Codes were developed inductively and represented the semantic content of the data.Themes were generated from the codes.The themes were discussed as a team (CS, JF), and any disagreements were discussed until a consensus was agreed.A thematic map was generated, and authors (CS, JF, LH) reviewed the codes.After reviewing the codes, some of the themes and subthemes were recoded to produce the final themes.

| Data triangulation and integration
Survey and interview data were integrated throughout the narrative of the analysis, using a weaving pragmatic approach, where both quantitative and qualitative data is presented together on a theme-by theme approach (Fetters et al., 2013).Therefore, themes exclusive to the interview and survey items which did not fit the themes were not presented in this paper.
The same coding framework developed for the interview analysis was also used to code the survey responses.Codes for survey responses were then grouped into themes by authors CS and JF; themes were then subsequently refined and illustrative quotes were selected from the open-ended questions.NVivo software was used to facilitate analysis.
For the interview data, word clouds were used to visualize the most frequently used words coded to individual subthemes (excluding filler words such as 'and', 'or', or 'the') and to aid comparison with the survey data.

| RESULTS
As not all of the survey findings and qualitative themes could be feasibly covered in one article, this paper focuses on reporting the overlapping findings between quantitative and qualitative components, giving equal weight to each (see Supplementary Table 1).Overall, 114 participants completed the survey and 13 interviews were completed.
The demographic characteristics of the sample are provided in

| Overview of themes
The integrated data analysis yielded five main themes, which are summarized alongside illustrative quotes in (iv) Motivation as the principal barrier; and (v) Practicalities around receiving lifestyle data.

| Digital technology's ability to enhance the physical healthcare
Overall, most participants held positive attitudes towards DLIs being used in mental healthcare (see Table 2, quote 1 and 2).From the survey, no single health behaviour emerged as a priority; all were perceived as relatively equally important targets for improving physical health (see Figure 2a).Amongst interviewees, diet, sleep and physical activity/exercise were the three behaviours most spontaneously mentioned and on reflection highlighted as a priority by most MHCPs, with recognition of the impact of antipsychotic medication (quote 3 and 4).Upon further probing, a few interviewees mentioned other health behaviours such as hydration, cancer screening, dental hygiene, safe sex and substance abuse (see Figure 2b).
All interviewees felt mHealth would enhance physical healthcare (quote 5), although only 19.1% of survey respondents agreed that mHealth would 'reduce/replace access to in person care'.Several MHCPs discussed potential health benefits by monitoring YPMI's behaviours and health outcomes.Self-monitoring was also seen to facilitate better understanding of health and empower YPMI to take more active roles in their healthcare (quote 6).This was supported by survey respondents with around 85% agreeing apps that record service users' behaviour and outcomes would be useful for YPMI.Further, several interviewees mentioned that DLIs, particularly those that monitor health behaviours and/or outcomes, have the potential to save MHCP's time and resource and support provision of tailored health advice (quotes 7 and 8).There were some concerns raised around the appropriateness of monitoring behaviour and/or outcomes for those with eating disorders or severe symptoms such as paranoia (see quote 9).
With regard to specific interventions that could enhance physical healthcare, 81%-94% of survey respondents agreed that apps with health tracking, personal health coach, connecting people to health groups, and/or instructional training classes would be useful for young adults with mental illness.In the interviews, several MHCPs spontane- ously mentioned examples of how technologies could be beneficial for YPMI, including providing instructional videos on exercise or recipes, encouraging peer support, and directing people to external resources and support for promoting physical and mental health.

| Conditions for the acceptability of apps
Amongst survey respondents, 89% agreed or strongly agreed that any apps to be recommended to YPMI should be approved by the NHS.
One interviewee even reported their team had been asked to stop recommending apps not officially endorsed by the NHS (quote 10).
However, after deeper reflection during interviews, mixed thoughts emerged.Whilst NHS approval would give credibility and provide confidence to staff recommending apps, safety, effectiveness, usability and being evidence-based-beyond NHS rubber-stamping-were viewed as more important (quote 11).
Additionally the question of cost and who should pay arose, if the NHS were to approve and recommend apps.Access to phones, Wi-Fi, mobile data and associated costs for service users were all mentioned as barriers in interviews and by a few survey respondents (quotes 12 and 13).Amongst survey respondents, access to mobile data and/or Wi-Fi was seen to be a greater barrier than access to phones (16% vs. 43.6%agreement).Providing hardware (e.g., phones and tablets) or apps that functioned offline and/or with lower data requirements were options suggested to overcome barriers.The notion of the NHS providing service users with an app or a wearable device was broadly acceptable.Some interviewees felt that the NHS should cover costs for apps and/or wearables recommended as part of service users' treatment, with some caveats on usage (quote 14).

| Limitations on staff capability and time
Psychological capability refers to an individual's ability to perform the behaviour, such a knowledge (Michie et al., 2011).Opportunity refers to both environmental and social factors which may impact an individual's ability to perform the behaviour, such as time, resources and social norms (Michie et al., 2011).When asked to rate how comfortable they were using mHealth, survey respondents had a mean score of 68.54 (25.75; range 2-100; 100 = very comfortable) and the majority (74%)-and 11 interviewees-had experience of using mHealth apps themselves for exercise or physical activity, weight loss, 'they're taking responsibility for their own, erm, monitoring of their health as well, you know, we're not just, erm, sending letters out and, and wanting them to come in and do it.I suppose that's taking, you know, a bit more responsibility for themselves as well'.(Interviewee 7).
Quote 7: 'can you imagine give, give everybody a Fitbit, … or one of those watches, and we would have their blood pressure, their pulse, their ECG, you know, all that kind of, the amount of activity they're doing, they wouldn't even need to input anything, you know, we'd have like half of it done' (Interviewee 10) Quote 8: 'anything that's from that, the core components of the physical health check that they can input, would save everybody a lot of time and, erm, effort and money and it would also make it much more up to date' (Interviewee 10) Theme Quote Quote 9: 'I think that these kinds of apps, wearables can be useful for certain people but probably really unhelpful for others (e.g., young people with eating disorders)' (Survey participant id 87) Conditions for the acceptability of apps Quote 10: 'I remember there was that project, where we were allowed to recommend apps and then they sort of pulled it.There's something about NHS they felt like they [apps] needed to be somehow regulated or improved' (Interviewee 3) Quote 11: 'I think that's the most important thing [it's effective] and for it to be an app that people want to use, "cos like an NHS approved apps" are often really clunky things that aren't pleasant to use them, so I'd, I'd, rather the Trust could purchase like a, a membership for that you could encourage people to download with a code or something' (Interviewee 11) Quote 12: 'It is difficult to know the amount of young adults with or without access to smart phones and whether this is in line with the general population or not.There will always be a cohort of people who cannot access smartphones/date/wifi and so forth however many service users who I work with have access to these things, as much as their peers would'.(Survey participant id 55) Quote 13: 'it's making sure everyone has the equipment to do it, not, not, exercise equipment, the digital technology, the laptops, the ipads, the phones.It's making it a time when they all can do it' (Interviewee 4) Quote 14: 'working along the same way social prescribing works that you, you know, even if it's subsidized, 'with SMI …. probably a good point of call would be the care coordinator because they know them.
They work with them' (Interviewee 2) Quote 23: 'I think it needs to be solely with somebody who has, can over see this and can keep track so they would need to be alert of some sort, if you're going to pick up on that information' (Interviewee 1)

Motivation as the principal barrier
Quote 24: 'another barrier as well with CMHTs [Community Mental Health Team]… when you're dealing with people's mental health they got very serious, sort of very chaotic and complex mental health issues, physical health and then sort of things like, apps, they're not on your list of priorities' (Interviewee 6) Quote 25: 'Some service users may struggle to engage, for example, those with severe depression due to reduced motivation, or those with illness affecting cognitive and affective functions (e.g., acute psychosis)'.(Survey participant id 97) diet, smoking, or sleep (see Supplementary Figure 1).Three interviewees and a few survey respondents had experience of using apps  Additional staff barriers included limited time and competing priorities (quote 21).This lead to discussions about which MHCP would be best to approach YPMI about DLIs.There was broad consensus that mental health nurses and 'care co-ordinators' (which can include

| Motivation as the principal barrier
The main barriers for engagement were thought to be both MHCPs (quote 24) and services users' motivation.Both survey respondents and interviewees felt this population had lower motivation and reduced cognitive abilities (partly due to antipsychotic medication) compared with the general population (quotes 25-28).Additionally, interviewees were wary about the severity of service users' symptoms, such as paranoia, and their technological skills or lack of them (quotes 29 and 30).
To improve service users engagement, interviewees highlighted the importance of app design, emphasizing the need to be simple, easy to use and engaging (quote 31).MHCPs also mentioned a variety of other potentially useful Behaviour Change Techniques (BCTs) such as goal setting, social support, graded tasks, provision of information, discrepancy between behaviour and outcome, instruction of how to perform behaviour, demonstration of the behaviour, social comparison, prompts/cues, and rewards, to improve their engagement (quote 31).
Several interviewees commented that service users engagement and motivation, may be influenced by the setting where DLIs are introduced.Early intervention services was perceived as a good opportunity to introduce DLIs, due to the younger population (quote 32).Some MHCPs felt inpatient wards/setting were inappropriate due to YPMIs poorer mental health, and they had concerns it may lead to further harm.
However, this lead some to reflection that the severity of the patient's symptoms rather than the setting was more important to consider.Other felt inpatients setting was still inappropriate due to the restrictions in place and attempts to change behaviour, may not be well received (quote 33).However, others felt inpatient wards could be a good setting for introducing YPMIs to apps, particularly those focused on exercise, as YPMI are often bored with little to do.However, the introduction of these DLIs may need to be handled differently to other MHC settings.

| Practicalities around receiving lifestyle data
The majority of survey respondents (86.5% and 87.7%, respectively) agreed that 'an app which transfers health data to the healthcare team' and 'an app which record physical health side-effects from medication' would be helpful for clinicians (see Supplementary Figure 2).Interviewees, however, were more cautious, with only half agreeing health data should be transferred to electronic health records.Many also doubted feasibility due to perceived technical difficulties with current NHS systems (quote 34).Others raised concerns about consent, privacy, and security of patient data held on apps or devices to be subsequently linked to NHS records (quote 35).Additionally, a few raised safety and resource-related concerns regarding the extra burden of monitoring additional health data; they questioned whose responsibility it would be if a health concern was missed (quote 36 and 37), for example quitting smoking but not adjusting clozapine.Discussions led to the idea YPMIs could share the data with MCHP at their visits, discuss it together, and notify them of any changes or concerns.Alternatively, one person suggested an automated alert system and another person suggested notifications to patients asking them to inform their MHC service of these changes.

| DISCUSSION
In this mixed-methods study, we sought to provide new insights into how DLIs could potentially be implemented to promote healthy Graph from survey question: Who should be responsible for recommending (or 'prescribing') mHealth to service users?(% of survey respondents who selected the role).(b) Word cloud from interview question: Which HCP would be best placed to recommend/ 'prescribe'/provide patients with apps for promoting physical health?
lifestyles and improve physical health in the context of youth mental healthcare.This was achieved through combining quantitative and qualitative feedback from clinical staff working within these services.
The results showed a high level of agreement that mHealth technologies have the potential to enhance the provision of physical healthcare and promote healthier lifestyles amongst YPMI.A wide range of potential health behaviours were identified which could be targeted by DLIs, including nutrition/diet, physical activity/exercise, alcohol and tobacco use, and sleep.Although each of these behaviours are prevalent risk factors for poor outcomes in YPMI (Firth et al., 2019), mental health services are often unable to provide sufficient interventions to address this (Bailey et al., 2019), likely due to inadequate time and resources.Thus, the optimism MHCPs around using DLIs to address this gap, and the myriad of potential approaches for engaging service users, is encouraging.
However, some key considerations affecting the integration and implementation of DLIs were also noted.Concerns around the accessibility of smartphone devices/data and usability of smartphone apps amongst individuals with more severe mental illness were evident amongst MHCPs.Nonetheless, these issues were somewhat diminished in younger service userswho are perhaps more likely to own and use such devices (Firth et al., 2016;Torous et al., 2021).Furthermore, the ownership of mobile devices amongst mental health populations is increasing quite rapidly and latest data from high-income nations shows high rates of ownership by service users with SMI (Firth et al., 2016;Torous et al., 2021).This suggests that any remaining problems around accessibility of DLI technologies in this population may dissipate over time.Nonetheless, it was widely noted that service users' may face additional barriers to making full use of smartphone apps, due to the limited availability and costs of data/WiFi and apps themselves.Recent UK data suggests even those people with SMI own phones at high rates, but lower digital literacy amongst some individuals may be an often unrecognized barrier (Spanakis et al., 2021).MHCPs suggested that apps recommended for use in mental healthcare services should be free for service users, and ideally usable 'offline' to some extent (without constant requirement for internet access to engage with app features).However, even if/when DLIs became ubiquitously accessible, ensuring that these are readily usable and sufficiently engaging for mental health populations will remain paramount (Firth et al., 2016;Firth & Torous, 2015).
Mental health care professionals in this study identified that the cognitive deficits associated with severe mental illnesses and the sedative effects of antipsychotic medications could both present a barrier to the use of self-administered complex interventions delivered via smartphone apps or wearables.However, empirical studies suggest that cognitive impairment may actually not be a barrier for many patients, including those with SMI (Montejo et al., 2021).Still, interventions should be designed to maximize potential engagement, and some service users may benefit from initial support in finding, setting up (i.e., downloading and installing) and using new technologies for their healthcare or self-management of their conditions (Spanakis et al., 2021).Alongside this, MHCPs themselves may also be unfamiliar with these technologies, and lack adequate knowledge for identifying suitable apps and using them in an optimal matter.Indeed, the concept of providing some form of training for MHCPs towards implementing digital interventions for their service users was widely supported by participants in this study.The implementation of new interventions could also be enhanced through specialist roles, for example, a 'digital navigator', within mental health care teams, who could support digital literacy, identify available digital interventions, and support their uptake and usage in services (Wisniewski & Torous, 2020).
Regarding utility, the MHCPs in this study largely felt that mHealth would mostly enhance face-to-face care, rather than interfere with it or replace it.DLIs were perceived as holding certain advantages over non-digital interventions for improving physical health, through providing 'any time anywhere' support, and been remotely administered (not requiring service users to travel to receive the intervention).Indeed, a recent study supports the potential efficacy of using DLIs to provide additional physical health support in mental healthcare, finding that participants with SMI allocated to an online intervention (WebMOVE) lost more weight than the in-person intervention (move-SMI) (Young et al., 2017).However, MHCPs were largely not currently aware/using any effective digital tools for physical health interventions in their practices.
The data collected from DLIs could also bring additional benefits to physical health care for people with mental illness, as previous research has indicated that using remote, real-time data collection in healthcare settings can reduce demands on staff whilst also improve detection of health deterioration and improve service user care (Glenister, 2015).Procedures would need to be in place ensure the secure capture and storage of real-time data.Both phases of this mixed-methods study showed the majority of MHCPs felt that using wearables to passively collect data such as sleep, physical activity and heartrate would be beneficial for gaining a more fine-grain and upto-date understanding of service users physical health.Along with the potential benefits of service users self-monitoring their behavioural health data, such information could also contribute towards personalized remote health coaching.For this, MHCPs, health coaches or even digital navigators could review passively-collected data and feedback to service users either via phone, text, email (Klasnja & Pratt, 2012).

| Limitations and strengths
A major strength of this study is the mixed method design, which allowed for the triangulation of qualitative and quantitative data in order to produce a more nuanced understanding around the issues affecting future implementation of DLIs in mental healthcare.Alongside this, the aggregated findings from the survey and semi-structured interviews identified key ways in which DLIs could already potentially be used in YPMI, whilst also pointing out some of the related barriers to overcome in order to boost the uptake and effectiveness of DLIs in mental health settings.Furthermore, this study has a large and diverse sample of MHCPs (compared to previous studies assessing perceptions around digital technologies in this group).
Limitations of the research pertain to non-responder bias (as only individuals with strong opinions regarding DLIs may have completed the survey), and the potential generalisability of the sample; as findings are limited to MHCPs working in the UK NHS, with the majority from the North West of England.Nonetheless, the rapidly increasing uptake and usage of mHealth technologies across all health services globally means that the results from our initial study here may be applicable to a wide range of mental health settings and services.

| CONCLUSIONS
In conclusion, MHCPs in this mixed-methods study expressed positivity towards using DLI particularly around physical activity, diet and sleep, for YPMI.Furthermore, the results identified a range of ideas for facilitating the uptake/implementation of DLIs in youth mental healthcare, including training for MHCPs, recommendations for apps, and provision of required devices/data for services users.Further research is now required to establish the efficacy and effectiveness of using DLIs to improve physical health behaviours and outcomes in mental health populations.
) was used.In Phase I, a predominantly quantitative, online survey was used to examine MHCP views on using digital technologies for physical health promotion amongst YPMI.In Phase II, in-depth, qualitative, semi-structured interviews, informed by the previous phase, were used to gain a deeper understanding of MHCPs views on DLIs.Finally, survey and interview data were integrated into common themes.the option to also provide their contact details (via a separate form) to take part in the interviews.The 16 individuals who completed the interview recruitment form comprised of a sufficiently diverse group of clinical roles for conducting the interviews, and so were contacted via email to discuss participation.Thirteen of these agreed to take part in the interviews.Online interviews were conducted on Microsoft Teams by two researchers (JF & CS), over the period August to December 2021, 2 to 14 weeks following survey completion.A semi-structured interview schedule was designed to guide discussions, which covered areas including: experience using digital health; how mHealth and DLIs could be used in mental healthcare; barriers to integration and use; and ways to improve engagement and technical factors regarding professionally or recommending apps to patients.Yet skill with technology, and knowledge of available mHealth options were often voiced as barriers for the delivery of DLIs amongst interview respondents (quotes 15-18) and 47% of survey respondents agreed MHCPs were too unfamiliar with mHealth to recommend them.Ninety percent of survey respondents agreed training would be useful to support integration of mHealth into services (quote 19).

Further
detail provided by interviewees indicated that a practical 'pharmaceutical rep' style training programme could be appropriate, where different apps could be demonstrated and staff could trial the apps themselves, which could also increase staff motivation as well as capability.(quote 20).

F
Figure 3a,b and quote 22).The role of GPs was supported by participants in both phases, with interviewees expanding that GPs might engage those not currently receiving treatment from CMHTs, although time pressures were noted as a particular barrier.Many interviewees mentioned it would be useful to have new staff members/roles to facilitate mHealth use in services, especially given time pressures on MHCPs (quote 23).

Table 2
ability to enhance the physical healthcare; (ii) Conditions for the acceptability of apps; (iii) Limitations on staff capability and time;

1
Demographics of MHCPs in the quantitative survey component and interviews.Integrated survey and interview data themes and subthemes, with illustrative quotes.'I have worked in a previous Trust who have used technology to record physical observations on wards and even just this little change and notifying the correct people was beneficial, so I can see how all of these would help' (Survey participant id 45) 'Off the top of my head, look at sleep, hygiene, diet, and exercise.I think that would work really well [in YPMI] and become potentially, I think see some results from that' (Interviewee 1) 'I think the biggest thing for us is, is weight, like, especially in our service, like people's weight tends to go up very quickly.So I think, a way of them monitoring their own weight, er, and maybe, tracking their own diet and giving them like kind of healthy advice about foods' (Interviewee 10) Quote 5:'you know, when you working there's a care coordinator.You're desperate for something to try, and you know to give to help somebody you know, cause you can't be there all the time and you know people need other things.And I think you know to have a.You know, uh, library of stuff at your disposal that you know is valid and useful and you know would be really helpful' (Inter- 'Clinicians not knowing how to use the technology/not feeling confident in their ability to use it meaning it gets forgotten about'.(Survey participant id 59) erm, depending on their financial situation, you would obviously need some feedback that they're using it [apps & wearables]' (Interviewee 6) 'I have heard some, clinicians and service users say they feel, sort of, uncomfortable generally with the, the whole digital health area and feel like it's moving away from what, what people see as the sort of the core function of mental health clinicians and professionals which is around sort of forming meaningful relationships with people' (Interviewee 9) (Continues) T A B L E 2 (Continued) 'It's kind of going back to the old model on when pharmaceuticals used to come in and kind of give out stuff, but now just with technology.cos I think once people see it, get it, understand it" (Interviewee 8) "If I'm really honest this would fall to the bottom of my agenda and to-do list because realistically our time is pressurized in the NHS, [now] more so than ever."(Interviewee 1) 'I do have privacy concerns (laughs) yeah I do have privacy concerns, I think that, it's not why I don't use apps, erm, but I would definitely feel that the privacy element would be really, a really, really, really important thing, erm, the idea, like particularly privacy around that like, what if insur- 'We can't just have data flooding into a system without an escalation or an alert happening.So if they've, say if, er, erm, if they're, erm, say smoking and they suddenly stop, but because it's on the app they think that they've told that practitioner and it goes through, erm, that could have dire consequences'.(Interviewee 5).