The use of carer perspectives and expert consensus to define key components of a biopsychosocial support intervention for stroke carers

ABSTRACT Objective: To identify the key components of a biopsychosocial support intervention to improve mental wellbeing for informal stroke carers within the first year post-stroke based on the combined perspectives of experts in the field of psychological care after stroke and informal stroke carers themselves. Methods: After reviewing the existing literature a cross-sectional mixed-methods design was adopted comprising 1) focus groups with informal stroke carers about their psychological support needs, and 2) nominal group technique with academic and clinical stroke care experts to reach consensus on intervention priorities. Transcripts were thematically analyzed and combined with the ranked priorities from the nominal group to identify key components for intervention content. Results: Key themes for informal stroke carers were associated with: 1) changes in relationships, roles, and dynamics; 2) emotional impact and acceptance; 3) drawing on inner resources; 4) looking for information, solutions, and explanations; 5) support from others. The expert nominal group placed priority on eight ranked areas: 1) acknowledging “normal” emotions; 2) education about the effects of a stroke; 3) reactions to loss and adjustment; 4) recognizing signs and symptoms of not coping; 5) knowing how and when to access practical and emotional support; 6) strategies for taking care of own health; 7) dealing with difficult emotions; and 8) problem solving skills. Conclusions: Themes from the informal carer focus groups, and ranked priority areas will inform the development of a biopsychosocial support intervention for stroke carers to be tested in a feasibility randomized controlled trial.


Introduction
According to figures published in February 2018, there are approximately 1.2 million stroke survivors in the UK and around 100,000 new strokes each year, two-thirds of which are living with a disability 1 and are cared for by family members or friends. An informal carer is defined as, 'a person of any age who provides unpaid help and support to a relative, friend or neighbor who cannot manage to live independently without the carer's help due to frailty, illness, disability or addiction.' 2 Therefore, they play a vital role in the survivor's rehabilitation and long-term care and deal with many practical, personal and emotional care needs. 3 The sudden event of a stroke denies family members the time to prepare themselves for the caring role or for the nature of the care they will need to provide. Some people find the role a positive experience, however the demands of caring for a relative can increase stress and lower levels of mental and physical wellbeing. [4][5][6][7][8] Previous research has reported that the psychological impact of caring can result in increased frequency of anxiety and depressive episodes 9,10 negatively affect relationships both within and external to the family, and reduce the ability to participate in leisure activities or paid employment. 11,12 A wellbeing survey of UK stroke carers revealed that many experience negative psychological effects as a consequence of caring, namely anxiety (79%), frustration (84%), sleeping disturbances (60%), depression (56%), and stress (57%). 13 Given the lack of provision of psychological support services for stroke carers, it is unsurprising that declines in psychological and physical health of the carer can negatively impact the care provided to the survivor. 13,14 However, research to develop effective support interventions for stroke carers has had limited success 15,16 For example, a systematic review of interventions for informal caregivers of stroke survivors found that none of the interventions included showed positive effects on all outcomes of carer wellbeing, which included caregiver burden, depression and anxiety, family functioning and quality of life. The interventions that focussed on providing the caregiver with information about stroke and available resources were least likely to have an impact on caregiver outcomes. However, those that included multiple intervention components such as stroke education, problem-solving, emotional support, and psychological support were more likely to show an early effect on the outcome of carer wellbeing. 17 Therefore, in order to have a positive effect upon carer wellbeing, it is important to consider the multiple sources of support needed by stroke carers when developing such interventions.
The biopsychosocial model of health and illness acknowledges the joint influence of biological (physical health), psychological (thoughts, emotions, and behaviors) and social (relationships and roles) factors upon mental and physical wellbeing and emphasizes that support interventions should focus on reducing negative symptoms and maintaining psychobiological health in order to be effective. 18 However, to ensure that they are fit for purpose in the context of stroke care it has been argued that involvement of the stroke carer perspective using a co-production approach, whereby stroke carer views and perspectives are used to inform the intervention content alongside those of other stakeholders such as clinicians or academics, is needed. 19 Also, involvement of experts in stroke care and stroke research can provide a complimentary perspective on the needs of the stroke carer population.
Therefore, the aim of the present study was to identify the key components of a biopsychosocial support intervention to improve mental wellbeing for stroke carers within the first year post-stroke based on the combined perspectives of experts in the field of psychological care after stroke and stroke carers themselves.

Research design
A mixed-methods cross-sectional design was adopted comprising: (i) Carer focus groups. These were conducted to give a 'lived experience' detailed picture of the support needs and priorities of stroke carers and (ii) a nominal group technique. [20][21][22] This method was used to reach consensus among stroke experts on priorities for intervention content. Written informed consent was obtained from all participants and ethical approval for the study was gained from the local research ethics committee (Nottingham 2 Research Ethics Committee, REC Ref: 14/EM/1264). This manuscript and the reporting of the study findings conform to the COREQ Guidelines 23 (see Supplementary Material). The research team were all female academics with a background in stroke research.
A participatory approach involving stroke carers was adopted as the theoretical framework for the study. Cornwall & Jewkes (1995) described the approach as having four phases: Contractual (whereby participants consent to take part in research), Consultative (participants are consulted on their opinions), Collaborative (participants work with academic researchers on research projects that are controlled by the researchers), and finally Collegiate (participants work alongside the academic researchers). 24 The advantages of the participatory approach are that by working collegiately with the key stakeholder group their experiences and views yield better designed research studies with outcomes that hold greater relevance to their real-world experience. As such the study protocol and research objectives were developed in consultation with an expert group of stroke survivors and carers who advise on local and national research, The University of Nottingham Stroke Research Partnership Group (NSRPG). In line with the collegiate phase of the approach, two members of the group were also involved in data collection procedures, management of the consultation groups and discussion of the study findings.

Data collection & analysis
Carer focus groups Carers were recruited from community stroke support and partnership groups in Nottinghamshire and were included in the study if their experience of being an informal stroke carer was <1 year. We did not exclude based upon stroke characteristic of the survivor, but we did exclude carers from taking part in the study if they were paid to undertake the care of the stroke survivor which would not have defined them as an informal carer. Ethical considerations were taken into account when constructing the sampling framework to ensure that a range of caring experiences were sampled and that the sample was not weighted towards i) one type of stroke, ii) the same type of post-stroke caring responsibilities, iii) family relationship to the survivor, or iv) employment status of the carer. All of which we felt were important sampling characteristics that could affect the experiences of the carer towards their caring role. Another consideration was whether it was ethical to collect data from stroke carers in a focus group format (ie. which would leave the stroke survivor without their carer), however all carers were eager to take part in the groups and arranged for the survivor to be looked after while they attended the group of their choice.
The carers could choose to attend only one out of a choice of three focus groups, which took place on different dates at the University of Nottingham, and were facilitated by the chief investigator (MW) and two other members of the research team (EK, PB) who assisted and took field notes. A topic guide was used to inform the group discussions and was constructed following a review of the literature on stroke caregiver support interventions up to and including the year 2016. The topic guide focussed on the biopsychosocial impact of being a stroke carer as well as practical issues to facilitating a support intervention (see Supplementary Material). Each group lasted for~60 to 90 minutes and discussions were audio-recorded and transcribed verbatim. The data was organized using QSR NVivo 11 (QSR International Pty, Ltd) and inductively thematically analyzed (PB, ST, EK) following the procedure detailed by Braun & Clarke. 25 An interpretive approach was taken to explore contextual, consequential, and hierarchical connections, and any discrepancies were resolved by discussion.

Expert nominal group
The Nominal Group Technique is one of the most common techniques used to obtain views and gain consensus on a given topic. Methodologically it belongs to the same family of techniques as the Delphi method, however unlike a Delphi, it is conducted face-to-face with several 'rounds' of consensus gathering taking place with the same group of participants, in the same place, and at the same time until a final list of consensus priorities are agreed upon 20 In the present study, purposive sampling through stroke research networks and clinical networks was used to identify nationwide research and/or clinical experts. Participants were included if they had experience in psychological care or clinical care after stroke and had experience relating to stroke carers. Participants were excluded if their experience did not relate to stroke or if they had no experience of working with stroke carers. Those experts who were eligible to take part were then invited to participate in the face-to-face expert nominal group with the aim of generating priorities for key components of a group intervention for the psychological care of stroke carers.
The nominal group was facilitated by MW who set out the objectives and processes involved in the expert nominal group at the University of Nottingham and asked participants to individually generate ideas in response to the question 'What elements should be included in a biopsychosocial support intervention for informal carers of stroke survivors?' This was followed by a 'round robin' sharing of ideas. Each participant in turn contributed one idea which was recorded on a flip chart without discussion. This was followed by a group discussion in which the facilitator ensured that all group members had the opportunity to contribute and that all items were valid to the research question. Where appropriate, new items could be added, or amalgamated, but none were eliminated.
The meeting ended with each participant anonymously selecting and ranking their top priorities from the list, which were then collated to form the final list of rankings.

Carer focus groups
A convenience sample of 16 stroke carers (12 female; 8 spouses) took part in the 3 focus groups; 5 carers attended Group 1, 5 carers attended Group 2, and 6 carers attended Group 3. The carers were aged between 47 and 79 years old (mean 57.5 years) and at the time of the study, their time as a carer ranged from 6 months to 10 years (median = 3 years).
The themes outlined below represent recurring topics of discussion between the three focus groups.
Changes in Relationships, Roles and Dynamics: Carers across each group described how there had been quite immediate and pervasive changes in the dynamics of their relationship with the survivor following the stroke. Many described how tensions had arisen between the carer and the stroke survivor due to the consequences of the stroke: I think the problem is … for us, he feels very vulnerable and he feels less of a man because he's got these physical problems … and I think that puts a strain on the relationship. (Focus Group 3, Carer 3, Wife) For some carers changes had also taken place in their relationships with the wider family unit as family members assumed new roles so relationships had to be renegotiated. Carers spoke about how these new dynamics could result in instances of families either pulling together or bringing additional tensions which caused strain on these relationships: Carers talked about the reactions to the survivor's stroke from members of their social circle which elicited feelings of disappointment in close members who had distanced themselves from them and failed to accommodate their change in circumstances. However, they also experienced feelings of surprise when 'peripheral members' of the social circle offered help. In this way the carer's social support network could become restructured with members switching roles and social support often being provided from unexpected sources: … the strange thing is that some people you didn't know that well start getting really interested in you. And other people who you knew really well … kind of move away because they're worried about what they'll find sort of thing. (Focus Group 2, Carer 2, Husband) Emotional Impact and Acceptance: Alongside the challenges of their new role carers were simultaneously having to deal with their own emotional response to the event and aftermath of the stroke as well as managing the emotions of the stroke survivor. This inevitably could cause considerable tensions: you know, she cries all the time and constantly will say … I wish I were dead. I wish this had never happened … and that's really upsetting. (Focus Group 3, Carer 2, Husband) Carers revealed feeling a sense of loss for the previous relationship with the stroke survivor and the life they once had together which was characteristic of the grief response, including anger, frustration, loneliness, despair, and depression: I've been angry a lot of the time … basically that this has happened to us, you know … I used to go downstairs and I used to sob my heart out 'cos it was just … devastating really. This wasn't going to be fixed. Drawing on Inner Resources: Assuming this new role daunted some of the carers and there was a sense of feeling unprepared for the added responsibilities they had to undertake and many carers discussed how they felt a need to be strong and self-reliant. For some, this was out of necessity due to their family situation, for others it was a natural coping mechanism. There were also feelings of protectiveness towards the survivor and some carers were unwilling to accept help instead feeling that they should just be "getting on with it": Looking for Information, Solutions, Explanations: There was a need to gain a better understanding of what had happened from very early on following the stroke, to gather information and look for solutions. Carers felt concern about what had caused the stroke and the chances of recurrence, and sought information about how to access formal care support, and how best to practically support the stroke survivor: … we just thought about ways that we could help dad and mum at home … every single issue that she had, we as a family tried to source information about. (Focus Group 1, Carer 1, Son) Support from others: Practical and emotional support from others, particularly friends and family, was seen by the carers as being key to their wellbeing and to provide relief from their situation. Contact and interaction from other carers who might empathize and give the benefit of their own experiences were raised as something they would value: There were varying degrees of input from health and social care agencies experienced by the carers across all of the groups. Typically, the most input came at the early stages immediately post stroke and then gradually reduced. The withdrawal of input was described as being a difficult period and left some carers feeling abandoned: After hospital pretty much you get on with it and see what you can find out. Ring people that you think can help. I found no help whatsoever and I work for the NHS! I found that really difficult. (Focus group 3, Carer 6, Wife)

Expert nominal group
Ten experts consented to take part in the group; four came from a clinical background, five came from an academic background, and one had a dual clinical academic role. All of the experts had over 15 years of experience in their respective roles (see Table 1) Table 2 shows the long-list of 30 ideas generated by the expert group and the total scores given for each item. This was then reduced to eight ranked priority areas during subsequent rounds of the nominal group technique and comparison with areas of agreement with the focus group findings.
The ranked priorities from the expert nominal group identified the psychological, biological, and social effects of stroke on the carer along with practical considerations for delivering interventions and techniques that can be used within the intervention.
Areas of agreement were identified between the ranked priorities and the carer perspectives namely: acknowledging "normal" emotional reactions to the caring role, learning more about the biological basis of what a stroke was and how it impacts the survivor; developing problem-solving skills to deal with caring situations; dealing with difficult emotions; loss reactions and dealing with adjustment; recognizing the symptoms of not coping; knowing when and how  to access practical and emotional support (see Appendix 1 for conceptual map of areas of agreement). Ranked priorities that were not in agreement with carer perspectives were related to the practicalities of intervention delivery: setting sessional boundaries (eg. session start/finish times, session duration, session frequency); setting tasks between sessions; knowledge and skill level of the facilitator on biological and psychological effects of stroke; and encouraging participants to form a social support network.
Other ranked priorities that were not identified in the carer perspectives related to therapeutic techniques to encourage the stroke carer's self-reflection, increased selfawareness and thinking about the future, namely; using pictorial aides as prompts (ie. the stroke stress thermometer); acknowledging variation in previous life/relationship situations; acknowledge that carer and stroke survivor may have different perspectives on recovery; maintaining the carer's future goals; developing strategies for taking care of their health such as self-monitoring; developing a wellbeing toolbox for the skills that they would learn during the group sessions (Table 3)

Discussion
This study sought to identify the key components for a biopsychosocial support intervention for stroke carers. To our knowledge, this is the first study to use a nominal expert group technique alongside focus groups with stroke carers to identify joint priorities for a biopsychosocial support intervention. The findings of this study have highlighted important areas of agreement between expert priorities and carer perspectives which fall within the domains of the biopsychosocial model of health and illness, namely: learning more about what a stroke was and how it impacts the survivor (biological & psychological); developing problem-solving skills to deal with caring situations (psychological); exploring normal emotional responses to the caring role (psychological); dealing with difficult emotions (psychological); loss reactions and dealing with adjustment (psychological); drawing on inner emotional resources (psychological); recognizing the symptoms of not coping (biological & psychological); knowing when and how to access practical and emotional support (psychological & social).
Educating carers on the etiology of a stroke and the physical, emotional and cognitive effects is in line with recommendations made by previous research. [26][27][28] For example, post-intervention accounts have reported carers gaining confidence by learning about the etiology of a stroke 27 and a systematic review of systematic reviews 16 concluded that carer interventions that include information provision are effective in improving stroke carer quality of life.
Drawing on inner coping resources such as resilience and problem-solving skills emerged from our carer focus groups as a priority for support and have been adopted as a coping strategy by stroke carers in other studies. 26,[29][30][31] Furthermore, a recent systematic review and meta-analysis 31 concluded that the inclusion of problem-solving and stress-coping as intervention components had a positive effect on stroke carers' psychosocial wellbeing and reduced their use of health-care resources. 15 The findings of the present study have also highlighted points of deviation in the expert ranked priorities and the carer perspectives. Practical aspects for intervention delivery were ranked as priorities by the experts but were not highlighted by the carers. The experts also gave priority to techniques that the carers could use to encourage self-reflection and increase self-awareness and self-monitoring of their own emotional health and wellbeing, which also relate to the biological and psychological domains of the biopsychosocial model of health and illness. These may not have been identified as priorities by the carers because they are 'unknown unknowns' for them in terms of techniques they can use to monitor their emotional health and wellbeing because they have not been exposed to them before and so do not know about them, or that they do not know that they can be used as formal techniques.
The process of adjusting to the caring role and the need to explore difficult emotions were reflected in the focus group discussions among the carers in our study. The clinical implications drawn from qualitative studies of carer experiences support the adjustment response and associated emotional difficulties as priorities for the psychological health of stroke carers. 26,32,33 As noted by Backstrom et al 32 ' … intervention strategies need to include awareness of the long-term impact of the changes in the relationship and the losses experienced by the spouses. Even if the partner is still alive, there are losses to grieve and to be understood and it is an important part of the transition in the relationship … ' (p226). The stroke carers in our study didn't report a need to learn techniques to help them cope with this, however they did report often having to deal with difficult emotions but coped with them alone without seeking professional psychological support or having any offered to them as part of standard care. Previous research highlights the negative implications of stroke carers suppressing emotions 34 and the implications on adjusting to the caring role 35 , along with emotional suppression being potentially detrimental to carers' long-term physical and emotional health. 35 This was supported by the accounts described by the carers in the present study.
Overall the joint findings of the stroke carer focus groups and the expert nominal group have identified key component for a support intervention for stroke carers which fits the biopsychosocial model of health and illness. Many of the priorities highlighted by both the carers and the experts related to the biological and psychological domains, and the priority identified relating to knowing when and how to seek support related to the social domain of the model. This resonates with the findings of previous research which has shown that interventions that include elements of peer support can be effective in helping carers cope and have positive effect on their wellbeing 36 This conceptual model will be used to develop the sessional plan of a biopsychosocial support intervention that can be delivered to stroke carers within the first year of caring.

Strengths and limitations
While this study provided data on the support needs of stroke carers, as well as expert consensus of ranked priorities, the findings should be considered in relation to the purpose of the study which was to inform the development of a biopsychosocial intervention. The data therefore underwent validation by two independent researchers to mitigate any interpretation bias. Furthermore, adopting a participatory design approach and conducting the study with stroke carers as co-researchers ensured that the study procedures were acceptable and applicable to its target population. Finally, to enhance the validity of our interpretations they are presented and discussed in the context of previous research in this area.
However, a limitation of our approach was that the overall study findings did not undergo a further collaborative consultation process held simultaneously to include both the carers and experts after the initial individual consultation process. This provides an opportunity for future research on the needs of informal stroke carers to gain enhancement on the interpretation of the findings.

Conclusion
This study combined stroke carer perspectives and ranked priorities of expert stroke researchers and clinicians to identify key components for a biopsychosocial support intervention for stroke carers. Agreement of priorities between stroke carers and experts related to learning more about what a stroke is and how it impacts the survivor; developing problem-solving skills to deal with caring situations; exploring normal emotional responses to the caring role; dealing with difficult emotions; loss reactions and dealing with adjustment; drawing on inner emotional resources; recognizing the symptoms of not coping; knowing when and how to access practical and emotional support. There were also areas of difference, or areas that were highlighted by the experts only, which are also important to include as key components of a biopsychosocial support intervention for stroke carers, such as: defining sessional boundaries (ie. session start/finish times, times, session duration, session frequency), acknowledging variation in previous life/relationship situations, acknowledging differences in perspectives and expectations between the carer and survivor, developing strategies that carers can use to look after their own health and wellbeing, and encouraging carers to form social support networks.