Involved, inputting or informing: “Shared” decision making in adult mental health care

Abstract Background A diagnosis of serious mental illness can impact on the whole family. Families informally provide significant amounts of care but are disproportionately at risk of carer burden when compared to those supporting people with other long‐term conditions. Shared decision making (SDM) is an ethical model of health communication associated with positive health outcomes; however, there has been little research to evaluate how routinely family is invited to participate in SDM, or what this looks like in practice. Objective This UK study aimed to better understand how the family caregivers of those diagnosed with SMI are currently involved in decision making, particularly decisions about treatment options including prescribed medication. Objectives were to Explore the extent to which family members wish to be involved in decisions about prescribed medication Determine how and when professionals engage family in these decisions Identify barriers and facilitators associated with the engagement of family in decisions about treatment. Participants Open‐ended questions were sent to professionals and family members to elicit written responses. Qualitative responses were analysed thematically. Results Themes included the definition of involvement and “rules of engagement.” Staff members are gatekeepers for family involvement, and the process is not democratic. Family and staff ascribe practical, rather than recovery‐oriented roles to family, with pre‐occupation around notions of adherence. Conclusions Staff members need support, training and education to apply SDM. Time to exchange information is vital but practically difficult. Negotiated teams, comprising of staff, service users, family, peers as applicable, with ascribed roles and responsibilities could support SDM.

cation which encourages an egalitarian approach to decision making, recognizing the skills and experience offered by all participants involved in treatment decisions. There are ethical, clinical and economic arguments for SDM: only those in receipt of treatment can evaluate trade-offs in relation to side-effects, and service users experience better outcomes when provided with appropriate information and supported to make choices. 14 Charles et al. 15 outline key features of SDM 1. Collaborative working between at least the patient and provider 2. Sharing of information and exploration of health concerns 3. Discussion of treatment options and preferences

Agreed decisions about courses of action and implementation
Positive health outcomes associated with SDM include decreased hospitalization, improved satisfaction with treatment and adherence to medication. 16 Service users diagnosed with SMI value opportunities to collaborate with care professionals and are prepared to engage with SDM within existing patient-professional relationships. 16 However, much of the research conducted to evaluate the impact of SDM to date has focussed on patients with physical conditions, and little work has been conducted to explore SDM amongst people diagnosed with SMI or their families.
Adherence to prescribed medication has been a pre-occupation for those researching the care of people with a SMI. A mean nonadherence rate of 41% has been reported amongst those diagnosed with schizophrenia, 17 and it has been suggested that 75% discontinue their prescribed medication within 18 months. 18 The effects of nonadherence to prescribed medication include reduced treatment efficacy, 19,20 increased risk of relapse 19 and adverse health outcomes. 21 Adherence to antipsychotic medication is an important predictor of illness course 22 and could be a valid precursor to non-medical approaches and longer-term recovery. When considering the reasons for non-adherence, many service users outline concerns about their medication, feel that prescribing decision making is not inclusive 23 and have described feeling disempowered with doctors. 24 Collaborative and trusting relationships between professionals and service users increase the possibility of SDM, 16 enhance satisfaction and could improve adherence to care plans. 19,25 The complexity of decision making within mental health-care commends a model of person-centred care such as SDM to improve experience and concordance.
Despite the potential benefits of SDM, mental health professionals have been criticized for not involving service users or their family members in care planning. 26 Research has found high rates of helplessness experienced by family members. 27 Relatives have been found to have feelings of inferiority to staff which could explain silence from family members, often taken as acquiescence or acceptance by staff. 28 Disparity is evident between professionals and service users in relation to the desired outcomes, with professionals placing greater emphasis on symptom reduction than service users. 6,29 Working alliance is vital to the success of SDM, reinforcing the active role required from professionals and service users. 16 Much less is known about the preferences of family members, and there has been little research conducted to look at the views of family in relation to treatment preferences, their priorities and understandings of recovery, their support or educational needs. Eliacin et al. 16 did not consider the views of family members; however, many of the quotes provided by service users prove illustrative of the central roles that family play in relation to recovery.
The involvement of family members within mental health care has been central to UK policy for 15 years, reflecting an international recognition of the importance of family support. Smith and Birchwood 30 highlighted the "problem of engaging families in a therapeutic programme" as a major national issue. Partners in Care (RCPsych 31 ) highlighted the problems faced by carers of people with different mental health problems, encouraging true partnership between carers, patients and professionals. The Carers Trust launched guidance relating to the "triangle of care" in 2010, updated in 2013. This approach acknowledges that models of engagement appear disconnected and recommends partnership working between service users, carers and organizations to achieve therapeutic alliance. 32 In 2015, the National Involvement Partnership 33 introduced national minimum standards for the involvement of carers in UK mental health services.
Barriers to involving family include unhelpful staff attitudes, unsupportive services, poor communication and inadequate information sharing. 34 Families want to receive information that is tailored to their specific experience and needs, specifically explanations on how to carry out their caring role more effectively. 35 Stigmatized attitudes towards individuals diagnosed with SMI have been found to adversely impact their mental health and well-being. The "Time to Change" initiative was launched to address this stigma, and there have been significant improvements in public attitudes particularly relating to prejudice and exclusion. 36 Despite this, nearly nine of 10 people with mental health difficulties say that stigma and discrimination have a negative impact on their lives. 37 Courtesy stigma refers to the impact of stigma on people who are associated with those diagnosed with stigmatizing health conditions. 38 Little is known about the impact of courtesy or direct stigma on the family members of people with SMI. 38 Research and practice suggest there could be benefits derived from encouraging family members to adopt active coping strategies 5 and that increased contact between health professionals and family members could decrease carer burden. 4 However, there is little research to understand the extent to which family members wish to be involved in decision making, how, by whom and when the notion of involvement is introduced, and the roles adopted by (or assigned to) family in relation to decision making.

| MATERIALS AND METHODS
This UK study was conducted within a large mental health and learning disability NHS organization. b. Capacity to give informed consent Non-prescribing staff were excluded as the study was designed to explore the views of those with prescribing rights towards the role of family within SDM, to consider family input to decisions about medicines (as a first-line treatment). Staff working within specialist dementia or memory clinics were excluded, as were staff working within child services.
Previous work within this region has experienced difficulties with face-to-face methods of data collection due to travel constraints. To overcome this, qualitative feedback was gathered on a written questionnaire, distributed in hard and electronic copies. Questions were designed to give participants the opportunity to write detailed responses, including unprompted entries. Question topics were identi-

Facilitators and barriers to involvement
The areas of questioning did not change post-piloting, but the wording of specific questions was amended to enhance readability.
The study was granted ethical permission from the University ethics committee and given formal approval by the relevant R&I department.

| RESULTS
Carer participants were recruited through snowball and opportunity sampling. A member of the research team (DG) attended four carer meetings and one carer workshop to introduce the study, taking hard copies of the survey tool with a sealable box for return, and the link to the e-survey tool. Posters to promote the study were placed across the organization. Information about the study and the e-survey link was sent to 16 carer groups within the region. This information was also sent to 19 members of staff who identified as leads for carer involvement. Given the wide distribution of invitations to family members, including posters, it is not possible to estimate how many family members considered participating in the study or to provide a response or refusal rate.
Forty-six family member participants completed the survey questions, 30 females (65%) and 16 males (35%). Of these, 31 (67%) were completed on a paper version then entered onto the e-version by the research team. The high proportion of responses completed in hard copy amongst this group reflects in part the benefit of having a research project introduced to participants in person, to provide support with completion or answer questions about the study.
Carer respondents were all family members. The age of participants ranged from 18 to 80 years, with the majority of participants aged between 61 and 65 years (n=9; 20%). Seventeen were caring for a child over 18 years and two for a child under 18 years. Nineteen were caring for a partner/spouse (husband n=7, wife n=6, partner n=4, fiancé n=1 and other n=1). Four participants were caring for a sibling, and two were involved in the care of a parent. Two declined to specify their relationship.
The diagnoses of family members were not always known or disclosed but those outlined included schizophrenia (n=3), bipolar disorder (n=1), autistic spectrum (n=1) and Aspergers (n=1). In addition to these, one of the participants who did not disclose a diagnosis subsequently wrote about their experiences of caring for a daughter with personality disorder. The high number of instances where diagnoses were unknown or not disclosed could suggest a lack of family involvement or knowledge, the use of working diagnoses in practice or reluctance to utilize/share diagnostic labels.
A total of 158 members of staff were identified as eligible for participation and emailed information about the study including a link to the e-survey. A reminder email was sent 2 weeks later. Paper copies of the survey were left in ward offices and with medical secretaries, with sealed envelopes for return. Surveys were taken by the research team to a non-medical prescriber meeting, with sealable envelopes. Fiftyfive members of staff completed the survey (response rate=35%), including 33 doctors (60%) and 22 nurse prescribers (40%). Of these, 19 responses (35%) were completed in hard copy then entered onto the online tool by the research team.
A thematic analysis of the qualitative feedback was undertaken by the two authors. Written comments were analysed using Excel to "hide" the group membership of participants. Once theming was complete, it was possible to reveal group membership and compare themes within and across participant groups. Techniques of thematic analysis were used, 39 including the early identification of concepts from written comments for comparison and contrast across instances. 40 Concepts were grouped together as themes with member checking across both the research team and the project steering group. At the end of the analysis, overarching themes included the following: Defining Involvement and Rules of Engagement.

| Defining involvement
Both participant groups defined family involvement as a process of "opinion-seeking"-to shape decisions about treatment options but also to consolidate decisions proffered by staff  (Family, 19026) Such comments reiterate those highlighted by Rowe. 34 When family members adopt active coping strategies and make a practical contribution, they reduce their levels of distress. 5 The supportive function of involvement was recognized in this study

| Rules of engagement
Interviews highlighted a number of implicit rules at play prior to and during family involvement, which resulted in family being either included in, or excluded from, opportunities for shared decision making. The overarching theme of "Rules of Engagement" makes explicit these rules and contains a number of subthemes; Patient Permission, Presence, Information Exchange, Monitoring and Staff Permission / the Carer "Agenda." The subthemes are explicated below, and relationships between them are highlighted in a conceptual memo (see Figure 1).

| Patient permission
Participants were aware that patient permission had to firstly be navigated, and it was felt, by both groups, that patient views were

| Presence
For staff, the next "rule" was fulfilled by the physical presence of family members at appointments: This suggests staff adopt a passive approach to family engagement, waiting for family to attend appointments, rather than actively encouraging attendance by sharing information or negotiating meeting times.
Without active attempts to inform family, staff are playing into notions of silence as acquiescence or acceptance. 28 Family participants reflected this by highlighting the challenges of "getting-in" to appointments and being "given a chance" by staff. Two of the family participants described how they had to take opportuni- Only families holding relevant information about service policy and practice would be able to "get in" to appointments in this way, reinforcing the important role of staff in sharing such information. In addition, decision making commonly takes place at review meetings which may be 1-1 or conducted with multidisciplinary teams (MDTs). To participate in these meetings, family members needed to be present, but their presence was determined or moderated by staff invitation.

| Information exchange
Both participant groups recognized that information was an essential precursor to family involvement; however, staff time constraints reduced opportunities for information exchange between family and staff, excluding family from possible involvement. In addition, staff participants felt that family had a key role in terms of providing context-specific information about service user response to, and adherence with, prescribed medication. A possible downside of this was that some staff members conceptualized family involvement through a predominantly practical lens-what carers can "do"-rather than a holistic consideration as related to either hopefulness 6 or recovery.
For family, information exchange served a dual purpose-to shape decisions, but also to support them, as carers. 35 The opportunity to provide information to staff, to have a platform in which to share thoughts and views about care, and to be listened to, was a key supportive resource for family participants:-

| Monitoring
Some staff were keen to involve family in decision making as they recognized that family had influence and could adopt reassuring or potentially persuasive roles:- There was a particular emphasis placed by staff, but not family participants, on the role of family as "medication monitors," to overview compliance, assess treatment impact and assert influence For family members, adopting or being assigned a "medication monitor" role could also be problematic. Family reflected on instances whereby sharing their opinion about progress and /or adherence was perceived by service users as an alliance between staff and family, rather than a whole team alliance:

| Staff "permission" and the carer "agenda"
There was a sense from some family participants that staff perceived them in a negative light and actively excluded them It was not possible to identify from this study how such conclusions were drawn, but this is a barrier to shared decision making and family involvement which warrants further exploration in terms of the attitudes of staff towards family. granted, or to discourage staff from discounting family input due to caution about a possible carer "agenda." Staff have a responsibility to prevent family members being subject to "courtesy stigma" 42 or direct stigma from those who may hold the view that pathology is solely rooted in family relationships and dynamics. 43 Named key workers or peer advocates/recovery workers could facilitate family involvement by actively negotiating co-worker roles and ascribing agreed responsibilities.

| DISCUSSION
Making explicit the "rules of engagement" for family input, heightening awareness of the barriers, increasing awareness of policy (including patient confidentiality policies) and disseminating the potential benefits of family input would be important first steps in terms of encouraging staff to further consider family involvement as a core constituent of shared decision making.

| LIMITATIONS
Due to the recruitment strategy for this study, it would be inappropriate to infer that these findings are representative of family experience broadly. Some family participants shared experiences where they had perceived active exclusion from staff just as some acknowledged the important role that staff played in supporting them.
Only staff members with prescribing authority were invited to participate and of these only 35% did participate. This is a relatively low response rate, and the findings should be interpreted with this in mind. Future research will include mental health nurses and care coordinators to encourage a broader discussion of family involvement. The self-selected nature of recruitment could mean that participants holding strong opinions about family involvement (negative or positive) were overrepresented. Despite this, the range of views collected and clustering of themes would suggest that findings are trustworthy. It was not possible to include non-English speakers within the study, but this is acknowledged as an important area for future study.

| CONCLUSIONS
This study suggests that staff have an increasing awareness of the need to inform family and to move towards a model of informed, if not yet shared, decision making. Family has unmet needs in relation to information, which can serve as a supportive and practice resource.
Adherence to medication continues to be a pre-occupation for prescribing staff, who respond by assigning "monitoring" roles to carers.
The prioritization of adherence should be challenged and staff could be encouraged to consider the broader nature of medicines optimization, including the important roles played by whole families when optimizing treatments. Such challenge could also support staff to consider broader roles for family, to negotiate beyond family roles which exclusively focus on and reward, "medication monitoring." In accepting such a challenge, whole teams should consider the difficult position in which family members find themselves, in relation to their caring roles and responsibilities. 44 There are a number of steps prior to family involvement, and subsequent SDM, which include the seeking of service user permission, and timely sharing of information. Both these steps are regulated by staff so it is important to share information with clinical teams about the possibilities of family involvement and to deter service-centred, rather than person-centred, delivery. Staff would benefit from additional training in relation to patient confidentiality, particularly as related to information exchange with family.