Understanding and using experiences of social care to guide service improvements: translating a co-design approach from health to social care

Challenges for social care

Adult social care in England is distinct from health care and encompasses care and support arrangements for people with (generally) long-term needs and a wide range of characteristics and circumstances. Social care also has different aims of care which are much more aligned to supporting individual well-being, independence and safeguarding, with activities focusing on personal care, domestic routines and supporting people to live full lives.

Care is needs- and means-tested with the underlying core objectives of promoting independence; encouraging and enabling choice, control, and ‘personalisation’; safeguarding; and supporting people to contribute to society. Older people comprise the largest user group and account for a high proportion of social care expenditure, though important groups in terms of incidence, needs and spending include people with physical, learning or sensory disabilities and those with mental ill health. Social care is delivered by many individuals and providers. Skills for Care estimates 17,900 organisations were involved in providing or organising adult social care in England as at 2021–2. Those services were delivered in an estimated 39,0007,8 establishments providing, or involved in organising, adult social care in England, with a somewhat dynamic picture of for-profit private sector providers dominating service markets. Social care is also heavily dependent on family and other unpaid carers, with about 5.4 million in the UK. Its structure is distinct from that of health care and less familiar to the public and those working outside of local authorities or with direct responsibility for commissioned services.

Despite recent cuts, commissioning of care by the public sector remains considerable. There were almost 2 million requests for adult social care support from nearly 1.4 million new service users to local authorities in 2021–2. However, the number of people receiving long-term care has decreased slightly to 818,000. 9 Eligible people receive local council funding for their social care but typically also pay a means-tested contribution. The council funding is either paid directly by the council to the care provider or it can be paid to the person needing care as a budget called a direct payment. Direct payments can be used in a variety of approved ways to meet care needs. If people are not eligible for local council funding, they pay for their care and support themselves and are known as self-funders.

In 2021–2 around 220,000 adults, older people and carers received direct payments, and the total number of direct payment recipients employing staff remained stable (at around 70,000) between 2014–5 and 2021–2. 7,8 However, estimates of a rapidly growing self-funded care sector who are often hidden from view and under-researched are less established, although we do know that self-funders find accessing and arranging social care difficult and emotional. 10,11 The Office for National Statistics estimates that 47% of residents in care homes for older people are self-funding. 7 Copayments (user charges) are common for many public services. Monitoring care quality is possible through contracting, and regulation is the responsibility of the Care Quality Commission.

Need for social care is growing rapidly with population ageing. If today’s care and funding arrangements continue unchanged, public and private expenditure on social care for older people will rise by 162% by 2035. 8 The need to address early identification and prevention agendas to avoid escalation of need is important and urgent in both health and social care contexts. Achieving this effectively would facilitate swifter and better-managed transitions of care and better support the growing self-funder population to access care and manage their resources.

Co-production and experience-based co-design

Co-production is widely held to be an important approach to improving the design and delivery of public sector services, by making them more inclusive and responsive and increasing individual and collective well-being. It involves professionals and members of the public working together; while an evolving concept,12 one definition is: ‘the provision of services through regular, long-term relationships between professionalized service providers (in any sector) and service users or other members of the community, where all parties make substantial resource contributions.’13

Co-production requires things to be done differently to deliver the aims of efficiency, patient or service user empowerment and accountability. 14 It involves a ‘reimagining of traditional health systems and practice trajectories’15 and reflects a political agenda to address inequalities and promote democracy. 16 Co-production has a key place in the values underpinning social care and is enshrined within the statutory guidance supporting the 2014 Care Act.

Co-production can operate at individual, group and collective levels, with the latter viewed as having the most potential to lead to wider, systematic change. 1 Different layers of public value have been identified: user value; value to groups; and social, environmental and political values. 17 In exploring why professionals (in this case hospitals) support co-production, Vennik et al. 18 found a belief that increased and improved patient involvement would improve the quality of care provided, that it helped to maintain the hospital position in a competitive healthcare market and that the approach was aligned to an adherence to organisational goals and ethos.

Osborne et al. 19 argue that co-production in public services may be unintentional and may simply arise from any interaction between professionals and users that shapes the service provided. They therefore use the term co-design to signify a more intentional, active process of working together to improve services.

Experience-based co-design is a well-established co-production approach used in health care. EBCD adopts PAR methods to involve patients and healthcare professionals as equal partners in redesigning services. 20 PAR approaches are underpinned by a collective commitment to investigate and resolve social problems with the building of alliances between researchers and those with direct experience of the issue being considered. 21 EBCD also draws on user-centred design, which again involves a collaborative approach to co-designing solutions to problem areas, and uses learning theory with a focus on reflective practice and narrative-based approaches to change. 5 There are a set of six clearly defined stages to EBCD that have been described in full elsewhere. 6,22 We briefly summarise these in Table 1.

A review of the international spread of EBCD 10 years after its development found more than 80 projects in 6 countries (UK, Australia, New Zealand, Sweden, Canada and Netherlands). 20 The authors note that many positive evaluations have been undertaken which report not only specific changes to services or care pathway redesign but also changes in staff attitude and morale. However, they note several challenges, particularly the time involved in an average EBCD cycle (1 year) and the cost of conducting interviews and observations. These issues have led to often informal and untested adaptations to the method. Only half the studies included in the review used observations, and only just over half filmed interviews with users. These adaptations may reflect not just resource issues but also concerns about anonymity, privacy and unequal power relations. In one inpatient mental health unit, for example, filming was retained but the process of recruiting participants and interviewing was led by a user group with existing relationships within the unit. Rather than a single interview, a repeated cycle of interviewing with an art therapist was adopted, giving people time to prepare and reconstruct the story of their experiences in their own way. 23,24 Mulvale et al. 25 introduced novel elicitation methods with experience maps and co-designing visual ‘prototype’ solutions, alongside the catalyst film.

A case study synthesis of six EBCD projects in Australia covering different areas of health care found that there was a need for guidance on capability development and preparedness for all participants in the project, not only for the project leads. 26 The authors found that the variability in implementation of the approach made it difficult to identify which component parts were essential for sustained service improvement. An evaluation of EBCD in a cancer centre in two large NHS trusts in London found that the engagement in the setting and proactive involvement of key clinical leaders alongside ‘particularly skilled staff who were committed to the philosophy of the approach’ were key ingredients to its successful ‘implementation’. 27 Locock et al. further describe ‘push and pull’ factors:

[T]he ‘push’ of active support and ‘opinion’ leadership at a high level of an organisation; the ‘pull’ of active quality improvement teams on the ‘look out’ for new approaches; and the mediating effects of ‘outward-reaching’ quality improvement leaders to build potential relationships. 6

Ramos et al. 28 explored the barriers and facilitators to the approach through a literature review and small-scale qualitative interview study with EBCD project facilitators. While EBCD has commonly been focused on staff in direct care roles, the authors conclude that active management participation in the process facilitates the implementation of service improvement, and underlines the importance of the project facilitation role, full organisation support, and the relationship between the service and patients. More recently, the use of the approach has been extended in different settings including forensic services29 and domestic violence and abuse services. 30

The importance of the adaptability and flexibility of this approach to different contexts while retaining its core philosophy has been emphasised,6,27 and we now turn to a key adaptation.

Accelerated experience-based co-design

‘Accelerated’ EBCD was developed by Locock et al. ,6 who sought to examine whether efficiencies could be gained by using existing filmed interviews to create a catalyst film, rather than collecting new user interviews locally. The project drew on a secondary analysis of interview data from an existing University of Oxford archive of interviews from studies of people’s experiences of illnesses or health topics. Evaluation suggested this approach led to a similar number and type of changes compared to standard EBCD but at lower cost and time, and the same catalyst film could be used in numerous settings. This model has since been used in different settings, including a hospital-at-home service for patients with chronic obstructive pulmonary disease and for advanced care planning in later life. 4,31,32

Experience-based co-design and its accelerated form are now well-established techniques for service improvement in healthcare settings but have so far not been widely used in social care or to improve care support for people whose lives are affected by social rather than health-related issues. In this study, we produced a data set of nationally conducted interviews focusing on loneliness to explore the use of AEBCD in one local authority, Doncaster. The findings of the interview study and the catalyst film will be published on Socialcaretalk.org and available for use nationally under licence.

Thinking critically about co-production

Co-production is not without criticism, and Steen et al. 33 describe the tendency of researchers towards optimism without engaging with the ‘dark side’ of this approach. Steen et al. highlight the optimism surrounding the approach and an often-presented idealised view by scholars and others, which they argue masks a set of potential problems. Attention should be paid to the potential relinquishing of responsibility by government, local authorities and other public sector bodies. Co-production can lead to a lack of clear responsibility, with boundaries blurred between public, private and voluntary sectors. There is little evidence of sustained change, with innovations often remaining precarious and small,34 and it is important to recognise the costs associated with the approach. 35 Co-production can be dominated by people who are wealthy and well educated, leading to further exclusionary practices and a professionalisation of involvement. 36 There can also be co-destruction rather than co-production if the process is done poorly or as a tick-box exercise; it can cause conflict and lead to misunderstandings, opportunities to change services may be missed, and the approach is open to misuse and manipulation by government. This is not only a waste of resources; it can also generate mistrust. 11,35 While it is important to note that Williams et al. 37 have pointed out flaws in some of these concerns, we agree critical questions should be asked about how situated practices of co-designing, facilitating or participating in co-produced activities can reflect or disrupt power relations at a local level. 14 Tools are available to help navigate this complex terrain. Leach,38 for example, presents a normative framework to assess the efficacy of co-production which involves consideration of inclusiveness, representativeness, impartiality, transparency, deliberativeness, lawfulness and empowerment. Thinking specifically about EBCD, Donetto et al. 39 draw on Bradwell and Marr’s40 working definition of co-design process to frame how the EBCD approach is operationalised. Bradwell and Marr’s framework sets out four key dimensions of co-design process: participation (a collaborative process); development (an evolving, maturing process); ownership and power (a transforming process); and outcomes and intent (a process with practical intent in which unplanned processes and transformations are also likely). Bradwell and Marr acknowledged, as did Leach,38 that such frameworks of principles for co-production or co-design represent ideals which are difficult to achieve fully in practice but can act as a guide to key aspects deserving attention. We return to Bradwell and Marr’s40 framework when reporting on the current study in Chapter 6.

Loneliness

There is an unprecedented rise in loneliness globally, with around a third of people affected by what has been referred to as the ‘widespread disorder of our times’. 41,42 Loneliness is viewed as a serious issue associated with a range of negative health outcomes comparable to smoking and obesity. 43 In 2016, the Local Government Association, Age UK and the Campaign to End Loneliness produced a guide for local authorities on key research about, and practical steps to help tackle, loneliness. 44 This guide was updated in 2018 with additional sections on public health and preventing loneliness. 45 The report advised, among other things, that council leaders should understand local levels of loneliness and set out plans for action. In 2020, loneliness was announced as a government priority in response to the COVID-19 pandemic and government-enforced lockdown conditions. A new public campaign, #LetsTalkLoneliness, was launched alongside a £750 million charity funding package. Despite this focus, the definition of loneliness remains elusive largely because it is a subjective experience. As Fried et al. suggest: ‘We believe loneliness can be defined as a subjective negative experience that results from inadequate meaningful connections, but neither definitions nor assessments of loneliness have achieved wide-scale consensus.’46

Three main theoretical or conceptual frameworks have been offered to understand the phenomenon: existential, social and cognitive/emotional perspectives. 47 Common ground across these approaches is that loneliness is an emotionally negative experience that increases the risk of adverse effects on physical and mental health. The existential perspective conceives loneliness as a natural and necessary component of human existence,48 which involves the perception of being disconnected from others, and emotional aspects which include emptiness and feeling abandoned. 49 The social perspective focuses on the absence of needed relationships, feelings of disconnection, vulnerability, powerlessness and lack of interpersonal affirmation, with roots in attachment theory and childhood. 50 This perspective is commonly used in research, with 70% of studies in a recent review following this approach. 47 The cognitive or emotional perspective describes loneliness as a negative and involuntary experience that arises from the perceived discrepancy between desired and emotional and/or social relationships. 51 Different types of loneliness are distinguished based on chronicity. While transient loneliness may arise due to circumstances (e.g. moving house or retirement) and people tend to adjust to an unfamiliar environment or life situation, chronic loneliness alludes to feelings that last for more than 2 years and are hard to change, or to an intense feeling that is hard to endure. 52

There is controversy around the use of the term ‘chronic’, which suggests a pathologisation of loneliness and has been challenged in favour of terms like ‘persistent’ or ‘prolonged’ loneliness. 53 Research is often framed as though loneliness needs curing or management through health and social care services rather than being a relational and emotional state that can affect us all. A recent report by the Campaign to End Loneliness,54 however, underlined the importance of focusing on chronic loneliness, which affects around 2.6 million people in the UK (around the same number that experienced chronic loneliness pre COVID-19). The authors highlight how chronic loneliness is more likely to be experienced by people with poor health, disabled people, family carers and those receiving care, people living in poverty, and people from minority groups, as well as intersections of these factors. The lack of detail and focus on inequalities in research evidence about loneliness was also highlighted in the evidence synthesis by Mansfield et al. 47

A further limitation of research in this area is that it focuses on generating a quantifiable definition of the term to measure its antecedents and consequences. The prevalence of loneliness in later life, however, has not changed despite considerable attention to this area and a host of interventions across decades. 55 In a paper based on the findings of the first stage of this project (see Chapter 3), we suggest that more qualitative and conceptual/theoretical work is needed to better understand the phenomenon and develop appropriate resources. 56 Studies tend to explore deficits in interpersonal relationships, emphasising individual characteristics such as personality, social skills and physical mobility. Less research explores the role of communities and societal relationships that contribute to loneliness. 57 The word ‘loneliness’ is often used interchangeably with ‘social isolation’, which is measurable and considered an objective state. Unlike loneliness, which can be experienced in the presence of other people, it is more about the lack of ‘meaningful’ relationships than the size of a person’s social network. 58 While a lack of social contact can lead to loneliness, it is possible to live a solitary life and not feel lonely. 59 Indeed, solitude can be connected to feelings of freedom and a sense of comfort with one’s life. 47

There is recent interest in younger people and loneliness,19 in part fuelled by COVID-19, and we need to explore the concept beyond the constraints of age. Advanced age and ageing are often equated with becoming lonely, although a recent synthesis of qualitative studies exploring loneliness indicated there was a fall in the number of studies focusing on ageing to less than half the studies included in the review. 47 Victor et al. 60 found that 14% of older people who were lonely had experienced loneliness in five life stages from childhood to older age. Other qualitative studies explore loneliness in relation to groups such as disabled people or people with mental ill health and identify explicit contextual and person-related factors that contribute to feeling lonely. 61,62

Loneliness interventions

Reviews of intervention studies, for example the work of Victor et al. ,53 draw attention to insufficient separation of related concepts, such as loneliness and isolation, for underpinning the persistent difficulty in establishing ‘what works’ in tackling loneliness, for whom and in what circumstances. Fried et al. highlight how the evidence base for interventions ‘is characterised by poorly constructed trials with small samples, a lack of theoretical frameworks, undefined target groups, heterogeneous measures of loneliness, and short follow-up periods.’46

Although loneliness matters for a prevention agenda, a review of interventions among older people found that mechanisms for reducing social isolation and loneliness differed and the quality of evidence was generally weak. 63 However, local authorities continue to introduce interventions without consideration of the very little existing evidence and few robust evaluations.

Interventions often use ‘asset-based’ approaches focused on maximising personal and social network resources. 64 Asset-based approaches have gained traction in social care, where many people fund their own care and cutbacks contribute to growing unmet need, and have been enthusiastically embraced in preventative agendas around tackling loneliness and isolation within the context of the Care Act 2014, which orientates local authorities to a duty to prevent or delay the need for care.

While many reviews and evaluations have concentrated on the findings from quantitative studies of interventions, qualitative study findings have helped understand some of the nuances of complexity, pointing to why certain interventions may be experienced as more effective. For example, Gardiner et al. 63 found three characteristics common among effective interventions: attention to local context, participants doing productive activities and involving local users in design and implementation of interventions. McMullin1 compared co-production initiatives aimed at older people experiencing loneliness in England and France and identified the importance of how the social problem is framed in terms of values. The Ageing Better programme based in England explicitly linked loneliness to individual problems, while the Monalisa initiative in France was linked to wider social and political values which motivated the selection of activities which are more geared towards successful co-production. Other reviews similarly suggested factors associated with successful interventions, including adaptability, community involvement in the design and development of the intervention, and active rather than passive client engagement. 65–67 Some of the sociodemographic factors considered important when tailoring interventions included age, poverty, and being a carer, as well as aspects of the social environment such as access to transport and driving status. 68 Interventions based on social loneliness and a focus on social connections and community integration are unlikely to work. 47

Where interventions were delivered by multidisciplinary partnerships, colocation of professionals involved was shown to be beneficial, by acting as a reminder to those delivering the intervention of the availability of support from others. 69 Finally, the importance of identifying triggers or events where loneliness and/or social isolation may change – for example, life-course transitions such as retirement, and relationship and health changes – was also shown to be key. 70

The reviews reported fewer factors that hindered the effectiveness of interventions. Of those reported, most concerned digital technologies. Specifically, these were the attitudes of older people towards the required technologies and lack of interest in learning new technological skills; the cost of technology; access to technology (including digital literacy and access to devices, especially in care homes); and concerns over privacy and online safety. 67,71,72 Other non-technological factors hindering the effectiveness of interventions were the workload and lack of interest of those involved in delivery65 and difficulties that carers of people living with dementia had with relaxing at joint events due to concerns about how the person with dementia would react. 73

In general, the reviews report a lack of sufficient evidence on ‘what works’ to guide practice. Reviews note that this is not to deny that some people experience beneficial impacts from interventions intended to help address loneliness or social isolation. Rather, evidence remains weak owing to the complex nature of interventions and research quality issues, which included little theoretical exploration of how interventions can mitigate loneliness and/or social isolation and a lack of methodological rigour in their evaluation. In summary, reviews are increasingly calling for providers to move away from standard, ‘one-size-fits-all’ interventions to more tailored offers and for future research about the effectiveness of interventions to be able to discern what works for whom, how and in what context.

COVID-19 and loneliness

The impact of COVID-19 and associated government policies around social distancing have led to considerable research on loneliness in the last 2 years. This evidence is contradictory in places. Some research suggests the experience of loneliness has remained stable despite social distancing rules74 while other findings suggest the pandemic has resulted in more loneliness. 75,76 It is clear, however, that pandemic-related conditions have had an uneven impact, with some people more affected than others. A recent report from the Campaign to End Loneliness, drawing on data from the Office for National Statistics, showed an additional million people experienced loneliness during the pandemic, and this impacted more heavily on people who already experience disadvantage: ‘[T]he fallout from COVID-19 could further embed economic and health inequalities, with disadvantaged people more likely to be out of work and in ill health, increasing their risk of chronic loneliness.’77

While concerns have been raised about the impact of COVID-19-related responses on older people, with a request for more nuanced and less ageist responses,78 people over the age of 60 years had a greater resilience to loneliness and mental ill health, possibly because older people have more experience of being alone and of dealing with potentially life-threatening medical situations. 79 The pandemic has brought into focus the impact of loneliness on children, young people and people with mental ill health. 80–82 Consideration around women is raised in some studies, although again it is not clear whether this is a continuation of pre-pandemic patterns. 81,83

In summary, tackling loneliness remains a key priority for health and social care services in the UK because the health and well-being impacts of loneliness can be substantial for those experiencing it. The costs to the health and social care sectors of increased health problems can also be considerable. Mansfield et al. suggest that co-production methods would enhance research around loneliness, concluding that: ‘[w]e need to comprehend more clearly who feels lonely, when where and in what context.’47

Why Doncaster?

Doncaster has a high concentration of neighbourhoods at elevated risk of loneliness (Age UK Loneliness Map),84 including several wards (e.g. Mexborough) in the top percentile of risk nationally. Local consultations have highlighted the loss of many social spaces linked to the decline of manufacturing and mining (e.g. community centres, working men’s clubs), and concern that loneliness is worsening (Doncaster Talks). Doncaster’s local authority and public health stakeholders have determined that addressing loneliness and social isolation is a key strategic priority, recognised in the local Health and Wellbeing Board priorities for 2018–21. The Doncaster Cabinet founded a third sector-led Social Isolation and Loneliness Alliance to substantially reduce loneliness in Doncaster.

The local authority and its Alliance have been highly committed to supporting our research. Given the challenges that social care research faces in engaging practice and the time it can take to build and sustain links with local authorities, it is important to have a willing organisation that is interested in research and has a clear commitment to the topic area. Doncaster is also part of the Curiosity Partnership, a 4-year Health and Social Care Delivery Research (HSDR)-funded research capacity-building network which is further developing the use and production of research at the level of local authorities.

In addition to Doncaster Metropolitan Borough Council acting as co-applicants to the research, planning discussions were held with the Head of Strategy and Performance, Head of Programme Delivery, the Assistant Director for Communities, and the Director of Public Health as well as the co-ordinator of the council’s work with the voluntary, community and faith (VCF) sector. They remained particularly interested in engaging with the video-based presentation of qualitative research as a means for enabling evidence to feed into service development and action planning.

The rationale for this project

Research, and more generally the use of data, should play a leading role in supporting commissioning and practice in social care. However, there is a weaker research culture and a lack of quality improvement architecture in the social care sector compared to health, with implications for the involvement of decision-makers, practice and service users. At present, local authorities collect some routine outcomes data but lack (even) the data-analytic capacity of health services. Qualitative data are a more natural format for experiential information, and robustly undertaken, could be more amenable to translation into service improvement. Moreover, narrative forms of data would be more accessible to local communities seeking to engage in service improvements. There is a slight caveat here, in that loneliness as the exemplar sits in local authority prevention work and is in the orbit of public health, which is more research orientated.

Social care policy is set nationally but interpreted, shaped and implemented in 152 locally elected councils. The result is wider variation in social care practice than in health care. There are very few public sector providers, making it harder to get research messages heard or acted upon, and most of the care workforce have no research experience or training. Those who provide services are typically in highly competitive markets and operate with low profit margins.

Staff turnover is high and even those seen as ‘qualified’ social care professionals (e.g. social workers, occupational therapists) receive less research training and continuous professional development than most of their equivalents in health. They are rarely exposed to research evidence, and indeed many do not have access to online research resources during their working hours. Few social care researchers simultaneously combine academic and practice responsibilities.

In short, models of working and the organisation, funding, commissioning and delivery of social care, as well as the research landscape, are quite different from health care, and the generalisability of innovation from health to social care is unlikely to be straightforward or comfortable. This makes it vital to ensure that what seem like promising interventions that support evidence-informed practice (e.g. the work of improvement scientists in the NHS) are tested in social care contexts. We know that there is considerable room for improvement in the provision of social care support in the UK and that innovative and cost-effective approaches are needed both to help prevent the need for social care and to guide the best use of current resources.

Aim

Our aim was to assess whether an effective and efficient co-design approach, AECBD, could be translated from health to social care, using experiences of ‘loneliness’ as an exemplar.

Objectives

1. To understand how loneliness is (1) characterised and experienced by people who are in receipt of social care in England and (2) characterised by social care staff and the voluntary sector.

2. To identify how services might be changed to help tackle the problem of loneliness experienced by users of social care.

3. To explore, with one local authority, whether an approach to service improvement, known to be effective in health care, could be adapted for use in social care.

4. To disseminate all study outputs and publish resources on a newly established Socialcaretalk.org platform for public, family carers, service users, voluntary organisations, researchers, teachers, policy-makers and providers.

The Healthtalk approach

Each Healthtalk project aims to generate a diverse sample in terms of demographic characteristics and types of experiences, which means the number of participants is typically large for a qualitative study. An expert advisory panel made up of patients, third-sector representatives, health professionals and academics offers suggestions about condition-specific experiences to guide the researcher in recruitment and other aspects of the research. Participants are interviewed in depth about their experiences and asked to give permission for their interviews to be used in research, teaching, publication, broadcasting and to form part of the new online resource. They can select whether they would like data extracts to be used in text, audio, or film versions. While Healthtalk.org was originally created to provide support and information for patients and carers going through similar health issues, it has become widely used in medical and health education and the project findings are used to inform health policy, for example, in the development of National Institute for Health and Care Excellence guidelines. 86

While the aim of the current project is to assess whether AEBCD can be translated from health to social care, in this chapter, we present the methods designed to achieve the first part of objective 1, namely, to understand how loneliness is (1) characterised and experienced by people who are in receipt of social care in England and (2) characterised by social care staff and the voluntary sector.

Field review

A brief field review on the topic of experiences of loneliness was conducted to inform the interview topic guides. This review guided the selection of study participants and highlighted issues for inclusion in the interviews. It further helped the team to recognise the limitations of previous studies and find innovative ways to avoid them.

We developed the search through an iterative process, using combinations of keywords and synonyms for loneliness and social care: loneliness; social isolation; social networks; intervention; review; public health; support; lonely; alone; isolated. We searched a range of social care, health, and social science databases, including Social Care Online, Social Science Abstracts, MEDLINE, CINAHL, PsycInfo and Social Science Citation Index. We also screened reference lists of relevant papers. Literature on improvement science was included to identify individuals who write about this area and communities who could be relevant in terms of engagement and dissemination. A literature topic alert was set up for the project’s duration and on the advice of our Study Steering Committee (SSC), the literature review was updated in October 2021 to include studies focused on COVID-19 and loneliness.

The initial field review was completed during months 1–3 of the study by researchers Malli (Oxford) and Maddison (York) when the start of the fieldwork was delayed by COVID-19. From the starting point of the latest synthesis of qualitative studies across the adult life course (144 studies published up to 2018), Malli identified an additional 30 recent articles. Of these 174 studies, 98 focused on loneliness in relation to elderly people and in developed countries.

Maddison examined published research reviews of interventions designed to reduce loneliness and/or social isolation. The purpose was to explore what types of interventions are effective and why. The search was limited to English-language peer-reviewed papers published since 2000, identified via the Cochrane Library or the Social Policy and Practice database. A total of 36 papers were identified, dating from 2003 to 2022; a further 8 were identified by cross-checking the list against a key ‘grey’ report prepared for the Cabinet Office in 2018. 60 The body of research reviewing ‘what works’ in relation to interventions to reduce loneliness and/or social isolation appears recent but is growing rapidly: almost half (18/44) of the reviews identified were published in 2020 or later. The findings of the review are included in Chapter 1.

Recruitment

Participants were recruited using a range of strategies including mental health charity advertisements, local authority newsletters, personal contacts through previous research and the involvement of colleagues, snowballing through existing contacts and social media platforms. More specifically, Facebook (Meta Platforms, Inc., Menlo Park, CA, USA), which has been identified as a useful tool in approaching seldom-heard populations,94, 95 was used. Flyers for the study were posted on relevant public Facebook pages and closed support groups. Due to the latter’s strict access conditions, the study was only advertised after permission was granted by the administrator through private messages providing details of the research. More than 30 closed Facebook groups were approached relating to mental illness, disabilities, drug and alcohol dependencies and lesbian, gay, bisexual, transgender and queer (LGBTQ) issues, of which 3 declined to advertise the study. Participants had to be 18 years old or over and to have experienced loneliness. Although some studies have avoided using the term ‘loneliness’ during recruitment due to negative connotations of the word,96 it was clearly stated in our flyers so as not to mislead potential participants about the topic of our study. Originally, we intended to include only people who used social care services; however, it transpired that people were often not aware of whether groups or services they had interacted with were included, and we widened our sample criteria to people who experienced loneliness.

All participants received a £30 voucher in recognition of their time and contribution to the research.

Interviews

The narrative interviews were conducted by Malli, an experienced qualitative researcher. A ‘warm-up’ discussion preceded the interview to make the participants feel at ease. This is especially pertinent to sensitive topics, such as loneliness. The interview was divided into two sections. An open question (‘Can you tell me a little bit about yourself and how loneliness came into your life?’) was used at the start of the interview to allow participants to highlight what was important to them, their values, meanings, priorities and experiences. The second part of the interview was based on the topic guide grounded in the field review and feedback from our project advisory panel. It covered loneliness and support, identity, loneliness, and the media (see Report Supplementary Material 1). To bridge the two parts of the interview, the researcher started with questions about topics raised by the participant in their primary narrative and then asked the additional questions to understand participants’ experiences of their wider lives including relationships and employment, social care services and other sources of support. We also explored how services could be improved. The interviews lasted between 45 minutes and 2 hours and were audio- and video-recorded with participants’ permission.

Transcript checking

The interviews were transcribed verbatim by professional transcribers, checked for accuracy against the recording, and deidentified. Time codes were added throughout. The transcripts were returned to participants to give them the opportunity to read and remove any sections they would not like disseminated or used in research. At that stage, they were asked to sign a form giving the copyright for their interview to the University of Oxford to enable their interview to be used in online resource production, including a resource to be published on Socialcaretalk.org and the catalyst film. The transcripts were also deposited in a University of Oxford data archive available to other bona fide research teams for secondary analysis under licence.

Data analysis

Analysis was conducted alongside the fieldwork until ‘data saturation’ was reached. We defined this as a point where new data repeated that of previous interviews without bringing new contributions to the understanding of the phenomenon. 97 A multistage inductive thematic analysis was carried out as a systematic method of generating themes and patterns within the data following the Braun and Clarke98 guidelines and using the organisational support of NVivo 12 (QSR International, Warrington, UK). A deviation from the study protocol included the use of grounded theory; this approach was felt to be more flexible, allowing us to analyse the data within the time constraints we were facing due to a delay in gaining ethics approval.

Initially, the data were read carefully by Malli in a process of familiarisation before being coded in NVivo to maintain participants’ meaning as far as possible. Malli and Ryan independently coded the first three interviews and met to discuss the different codes generated. The codes were then reordered into a more formal tree structure allowing the identification of broader categories. Malli continued to code the remaining transcripts. The descriptive codes were organised into categories and these categories were analysed conceptually in what Braun and Clarke98 describe as a process of ‘latent analysis’ to examine underlying ideas, assumptions, connections and links within the data. From this analytic stage, themes were identified, and the data were systematically reviewed to ensure that a name and clear definition for each theme were produced and that these themes worked in relation to the coded extracts. We returned to the original recordings as appropriate to aid this stage of analysis. To increase the integrity and trustworthiness of the study, the first author kept a reflective diary during the interviews, transcription and analysis phase.

Ethics

Ethics approval for the study was obtained from the Health Research Authority Research Ethics Committee (REC), after the submission was moved from Social Care REC, on 2 October 2020 (ethics reference: 20/WM/0223). COVID-19 necessitated further planning to minimise face-to-face interaction with research participants where possible. New operating procedures were introduced for the capture of data suitable for dissemination on a web platform which specifies methods to optimise video and audio quality while adhering to data security policy.

Loneliness as loss

Loss and bereavement strongly featured in participants’ narratives, and these deaths were often unanticipated and accompanied by additional emotional trauma. Participants were deprived of the company and affection of the person and an envisioned future together. P29, who lost her partner abruptly in her late 30s, explained:

I know this woman who lost her husband, but she was in her seventies or eighties. And I always say, ‘You lost a lot of past, but not so much future. I lost a lot of future and not so much past.’ It’s very different to be on your own when you’re, let’s say, in your seventies and maybe you are a grandparent, and you have a completely different role in life than when you are just a young person who had their whole life to go.

P29

Participants described irreplaceable bonds that could not be substituted, which worsened loneliness. P1 talked about the loss she felt after the death of her mother, whom she relied on in different ways:

I think that had a huge impact on how I interact with other people now I just, I don’t feel as connected any more. So she was always kind of like she had, she was a friend, she was a sister, she was a carer as well like, you know, she’d always sort of check up on me and she had so many roles and when that went it just left this huge hole in my life, so I think that, yeah, it sort of affected me quite badly.

P1

The unanticipated loss brought an additional layer of loneliness as participants discussed how they struggled to approach the grief of premature death. There was a lack of the comforting and supportive responses that bereaved people commonly receive. As P40 said in relation to her brother’s death:

When your 27-year-old brother dies from a heroin overdose it’s like there’s no [card company] card for that . . . people also don’t wanna talk about it because it’s a bit like ‘oh, do we bring it up?’ whereas people seem more comfortable bringing it up sort of if a parent or grandparent’s died.

P40

Loneliness as absence

For many participants, loneliness was marked by the absence of a meaningful person to turn to, a relationship that they never experienced. Many wanted a connection with someone they could talk to and confide in. The following extracts capture this absence:

I have no-one to share my life with . . . I don’t have someone to turn to, to tell my problems to, or anything like that.

P37

There’s moments when I’d like to sort of really connect with someone or when I’m feeling low, I want to sort of, because I don’t really have [um] many friends and [um] I think that’s when it does appear like [um] I feel that intense loneliness when I feel quite isolated.

P1

Even participants who perceived themselves as sociable and surrounded by casual friendships missed an intimacy and affection which was absent from their life. For P2: ‘I suppose I would really like my friendship to be like quite close, friendships should be quite close, so I want to feel like, like family’.

For P22 and many participants, the intensity of a friendship rather than the quantity of the social circle combated loneliness: ‘it no longer mattered whether I had other friends because I found another really, person who I could be really close with’.

Being overlooked during childhood

Loneliness was discussed in relation to participants’ family and childhood experiences, including stressful environments, the emotional unavailability of close family members, and other adverse experiences such as abuse and neglect. Looking back at their childhoods, they described how loneliness was about the need to be listened to, believed, supported and protected. The impact of this abandonment could be extreme, as P32 describes, talking about her relationship with her father:

It took years for me to realise that I actually felt annihilated in his presence. I was so unacknowledged, um, that I almost felt like there was nobody reflecting me back and my grasp on my own existence was tenuous. I realised when I was very much an adult, um, that just being in his presence was enough to make me feel suicidal because I felt so unacknowledged, like somebody had rendered me invisible.

P32

The illness of his younger brother, who subsequently died, led P12 to feel neglected by his mother, while P35 felt her parents gave her brother all the attention when he was born 8 years after her:

I was an only child for eight years and then my brother came along. […] So my parents favoured him over me pretty much from the get-go, so I kind of, because everyone was always all over my brother I kind of retreated [um] and just kind of spent a lot of time in my room by myself [um] not really socialising and isolating myself I suppose.

P35

For some participants, loneliness came from keeping experiences of abuse secret. Fear of the consequences of disclosing this and perceptions of self-blame emotionally isolated participants and made them feel that they could not reach out and ask for support. P11 described how:

I was being sexually abused from the age of 9 and, and it was quite a secret thing, [um] it was something I was keeping to myself . . . think it was quite lonely because I suppose it’s a lot to deal with and at the time, I didn’t really have anyone to talk to about it.

P11

The need to keep domestic violence as a private family matter that could not be disclosed and discussed with people outside the household left P37 feeling lonelier:

I was like locked inside of the prison, you know, I couldn’t say a word I couldn’t look for help I literally felt like I’m like decaying inside, like the living dead. You know we are forbidden to talk about this to anyone and it does make you feel lonely.

P37

Being abandoned by society and those who could have helped

For many participants, loneliness was a result of being deprived of community and social care resources due to austerity cuts. Financial cuts were seen as a source of social alienation that made people feel invisible. Social clubs and support groups which for many were an important resource for active networking with ‘similar others’ were closed. P19 talked about the impact of two social clubs closing:

They lost their funding, which was hard because I was enjoying it. I was enjoying the activities they were doing. They were doing like games, cooking in one of them. All different types of stuff . . . It was hard because I was enjoying it. I enjoyed seeing people.

P19

P18 described feeling isolated and abandoned after being diagnosed with multiple sclerosis. He described getting ‘a whole bunch of pamphlets thrown at you and told you have MS, and told we’ll see you in six months for a check-up’.

Other participants discussed not being heard within the mental healthcare services. They described a process of dehumanisation and objectification from healthcare professionals that induced loneliness and social isolation. In relation to mental health care, P6 described:

If I could at least [have] told them about my experiences and what I was going through, I wouldn’t necessarily have wanted an answer, but I’d have wanted someone to listen and if someone’s listening to me, they see me, I’m not lonely any more.

P6

Finally, some participants felt that their social identity was invalidated and undervalued, which made them feel lonely. P3 reflected on her unpaid carer role:

When I’m being sort of selfish, I think maybe I made a mistake and I should have just kept my career because they certainly don’t look after carers, they’re on their own . . . Because people seem to forget that the productivity coming from unpaid carers in this country saves the state about a £132 billion every year so, you know, it’s a bit of a sore point when people say you didn’t go to work.

P3

Loneliness and identity

For many participants, a sense of loneliness came from feeling ‘different’ and was inevitably ingrained in their identity. For example, P1 said ‘I think [um] being autistic [um] I feel lonely quite a lot of the time [um] so I think it’s always there, I think it’s never not a part of who I am’.

For some, being ‘different’ affected their sense of integration and belonging, and emotions were tied to loneliness. P33 said, in relation to mental illness:

I kind of always, I guess, had this like longing to belong or like be a part of other people’s lives where I wasn’t really. Um, so I spent kind of like most of my teen years extremely lonely.

P33

Being ‘different’ sometimes meant being stigmatised and fearing or experiencing rejection. In relation to her mental illness, P39 said:

It’s like damaged goods you don’t ever, nobody ever prioritises damaged goods, you look for an apple you look for a perfect one in the supermarket you don’t look for one that has got lumps and bumps.

P39

A common point was loneliness that stemmed from not being understood by people. Participants seemed to experience a profound sense of loneliness through trying to convey how they felt to people who could not relate to them. In relation to psychosis, P15 commented:

There’s a sense of loneliness that people can’t experience the same things that I’m experiencing, and people say to me, if I’m talking to somebody [um] that other people can’t see, it can be quite frustrating when people are saying, ‘But I can’t see that,’ when it’s very real to me. And that makes you feel lonely because people aren’t experiencing the same world that you’re experiencing.

P15

The absence of people with similar experiences, who could provide emotional support and validation of their identity, creates what Stein99 terms ‘experiential loneliness’:

You feel very much alone and you’re very much aware that the experiences you’re having and the feelings you’re having aren’t the same as the people around you. And as much as they try and are as supportive as they can be, I don’t think that they will ever be able to fully understand. You kind of feel like if you’re trying to communicate something to them about like, something difficult about being like gay or non-binary, they can try to understand but they won’t fully know, and I think that is a form of loneliness.

P33

Loneliness despite the presence of others

For many participants, loneliness persisted even in the presence of others. They felt the experience of emotional isolation amid a group of people, within their family or even their intimate relationship. Togetherness in some cases could make loneliness worse:

I’ve been with people and felt lonelier than when I was on my own. I don’t really know why. I guess, if you can’t relate to people, or like you don’t really feel a part of their group, then it can definitely feel lonelier to be around people.

P25

Loneliness was also experienced in absence of shared understanding, as P8 describes:

Going to the pub and being around people I still felt very isolated because of my HIV. And that people, there weren’t people who got where I was at, and understood [um] the position I was in, and what was going on in my mind. I felt very isolated. Even though I was surrounded by people.

P8

Some participants described an ‘internal’ type of loneliness, that kept them separated from the world and made them feel like an outsider. P37 describes this when referring to the relationship with her then partner:

I used different things in those times to cover over it [loneliness] and deal with it in a different way. But eventually it would come to the surface as the relationship developed [um] . . . but no, I’ve never not felt this crushing emptiness and loneliness.

P37

The vicious cycle of loneliness

Many participants reported feeling trapped in a cycle of loneliness in which the lonelier they became, the more isolated they felt from society. 100 They recognised how they should act to reduce this, while feeling that loneliness had become an integral part of their life in which they felt secure. P32 described the paradox of wanting to ease feelings of loneliness while feeling unwilling to step outside the comfort zone loneliness had become:

That might sound a bit peculiar, but I’ve begun to recognise that when I have the opportunity to socialise it’s almost as though I’ve turned in so much on myself, that [sighs] it’s becoming hard to turn outwards and meet other people again.

P32

It was apparent that loneliness can create a tendency in people to dwell in self-pity, be more sensitive to rejection, withdraw and be less trusting of the people around them. This in turn created worsening feelings of loneliness. The fear of experiencing loneliness in the future could further obstruct the enjoyment of the present.

The silencing of loneliness

A pivotal subtheme here was the notion of stigma, shame and self-blame which restricted any constructive discussion about loneliness. Many participants discussed the ‘archetype of the loner’ who is depicted as socially ‘defective’, inept and incapable of forming key relationships to combat isolation and loneliness. P39 highlighted: ‘If you’re a loner there’s something wrong with you . . . It is not a positive word to put on someone . . . there must be something, because why would someone choose to be alone?

This stereotype, which carries negative and derogatory connotations, views the person through a lens of personal and social failing and intrinsic character flaw. Subsequently loneliness is pathologised and viewed as an experience outside the boundaries of what is considered to be normal.

I think there was a lack of talking about loneliness as a thing that can happen to just normal people. It was always this person visibly has no social skills and you know is really profoundly like struggling in society, not this is another well-adjusted person who has got some confidence issues and is struggling to make friends in a new context.

P22

The ‘tyranny of positive attitude’, which is saturated with the view that we must think positive thoughts and block out negative emotions and avoid any difficult but nevertheless authentic emotions, further contributes to the silencing of the discussion of loneliness. P10 expressed ‘We’re meant to be these very successful, bubbly, happy-go-lucky people who do these great things every weekend and you know, are never stressed, are never behind at work, you know, are never struggling’.

Indeed, the stereotype and trope of the lonely elderly prevail in the collective consciousness, leaving little room for other populations to share their experiences of loneliness. By being excluded from the loneliness discourse, many participants resorted to self-blame for failing wider expectations:

They always say it’s the older people that feel lonely . . . They never talk about the younger generation. I think because society expects people at that age to be married, have children, and have their lives sorted so to speak, so they don’t think about people that might be gay, they don’t think about people that might be single, they don’t think about people that might be trans-sexual, they just think about the mainstream. And then sometimes that makes me feel guilty about being lonely or feel weird about being lonely, because I think I shouldn’t be lonely at my age, I should have lots of friends, I should have a partner. And then I think ‘what’s wrong with me? Why do I not have these things?’

P17

The stigma attached to loneliness dissuaded many participants from seeking support from services. Indeed, the intolerance towards individuals who experience loneliness acted as a barrier to help-seeking. In reference to accessing help for loneliness, P5 mentioned: ‘I would have really benefited from it at the time [support] I don’t know if I would have accepted it if anybody had said it because I also think there’s that kind of, you don’t wanna be stigmatised.'

The trivialisation of loneliness

Many participants said that the seriousness of loneliness can be diminished through a process of trivialisation. The consequences it might have for quality of life and mental health are downplayed and the reality, nuances and causes of the experience are oversimplified and even ignored. Moreover, the legitimacy and severity of loneliness are questioned and there is an implication that it may be the person’s fault. People are blamed for not being proactive, for not taking steps to make positive changes. Thus, individual agency was presented to participants as an effective approach to managing loneliness. For example:

People just go, ‘Oh, loneliness, you just need a fellow’, or ‘Loneliness, oh, get a dog’, or, it’s just people minimalise it so that it’s . . . loneliness leads to so much more than just being lonely . . . It can lead to all sorts of things that people don’t necessarily think would start from loneliness.

P13

Since the essence of loneliness is oversimplified, the interventions and advice offered by experts are unsophisticated, and, as discussed in Chapter 1, only superficially approach the problem. P5 shared her experience of being given ‘by the book’ advice from a health visitor as an early mother that was not tailored to her circumstances or needs:

You get all the health visitors come round and all that kind of stuff and I always felt like they were just saying OK, yeah but you can just fix that, you know, can you do this? Can you that? Can’t you do such things like it’s easy to, to kind of just get on with things and not feel isolated.

P5

Indicatively, participants discussed support from services that aimed to broaden people’s social networks and therefore to tackle social isolation (quantity), while they felt that they lacked a close intimate attachment to another person (quality). Thus, strategies between social isolation and emotional loneliness were used interchangeably, indicating the insufficient understanding of the differences between the two concepts. Many participants suggested that genuine attachments could not be formed in loneliness support groups since they are not based on common interests between people. For P2:

I just feel like in that [support group for loneliness] group other people will just be maybe desperate for friendships . . . like some people are lonely, they’re just happy to be friends with anyone [um] and obviously that’s not what I want. Either I want friendships to be genuine . . . you need to have a rapport, a rapport and [um] genuine fondness for each other and a likeness with [um] appreciation for each other’s company.

P2

Understanding loneliness and people’s experiences of being lonely

Local staff (who were predominantly community-based or working in the VCF sector) recognised that there was more to learn from people with lived experience (PWLE) of loneliness. Some participants thought local authority adult social care staff less likely to understand loneliness owing to having fewer opportunities for ongoing, community-based contact with PWLE. Learning from PWLE was discussed as key for understanding how people negotiated loneliness and discovering what people valued from loneliness support. This greater understanding could lead to the co-design and development of new activities or adaptations to existing support.

Local staff talked in terms of population groups or characteristics. Most commented that the focus of loneliness support was on older people, echoing the research literature and our interview study findings, with older people’s loneliness prompted by situations common in their life stage, such as bereavement, living alone and loss of independence. Participants were aware of other population groups experiencing loneliness, including younger people, people with poor physical health, people with mental health issues, people from minority ethnic groups and people living in poverty. Staff recognised that younger people experiencing loneliness received less attention and were often less visible owing to the associated stigma of being ‘a bit of a loser’ or being pitied for having no friends. Staff also spoke about people from minority ethnic communities feeling disconnected from their local community, for example due to language barriers, or feeling undervalued because of their cultural identity.

Participants commented on how loneliness could be interwoven with changes in life circumstances such as loss of employment, marital breakdown, retirement or taking on a new role such as that of an unpaid carer. Also highlighted were situations such as domestic abuse, which might necessitate people moving away from their local area, and substance misuse resulting in family members breaking off ties.

There was acknowledgement that anyone could experience loneliness. Some staff talked about their own experiences, or those of people close to them, as available to draw upon when crafting and delivering support. COVID-19 was highlighted as having increased awareness of loneliness and what it can feel like. However, staff also spoke about loneliness as nuanced. For example, some people may not be aware that they were experiencing loneliness because they had lived that way for so long that it had become the norm. Similarly, people who may appear to be sociable, with many connections with family or friends, could also feel lonely if lacking the special relationships they want.

‘Hallmarks’ of good loneliness support

For the purposes of prompting discussion in the co-design workshop, ‘hallmarks’ of good support were identified from the analysis.

A key hallmark was identifying people in the community who are lonely. Doncaster participants discussed including attention to loneliness as part of their Making Every Contact Count (MECC) approach. MECC encourages staff who have routine interactions with people in the community to initiate conversations related to how they might improve their health or well-being. 107

Related to MECC, participants stressed the importance of looking at a person’s situation holistically when they presented to staff for something other than loneliness, and finding out whether and what they wanted to change. Key to this work was having time to develop trusted relationships that could support people to help themselves.

A further hallmark of good support was the personalised matching of support with the person’s needs. Local staff said they felt inspired by the range and volume of loneliness support on offer in Doncaster; however, it could be challenging to be aware of everything and to keep information up to date, despite a locality model in which staff had strong contacts with communities.

Physical accessibility of activities or groups was another hallmark. The affordability of public transport, and for some people their willingness or ease of using it, could act as a key barrier.

Accessibility was also considered in terms of the lack of confidence people had to access groups and the support needed to improve this. Participants were proud of their in-depth knowledge of the activities available, sometimes attending new activities themselves to offer people a good match with their interests. Having first-hand experience of an activity/group meant they could answer people’s questions, so they could feel more confident about trying something out. A related issue was the value of offering graduated support, more intense initially (e.g. accompanying someone to their first session of a new activity) then tapering as the person gained confidence. Staff felt that better connections between different providers of loneliness support would enable their different offers to be used in a joined-up, stepwise manner as people’s confidence grew.

In turn, staff felt a further hallmark of good support was its function as a conduit to help people make their own meaningful and authentic relationships. Local staff were pleased with the way support services and activities actively engineered these initial opportunities for social contact which facilitated people to find and grow their own meaningful connections. In carrying out this work, staff pointed out the importance of recognising that some people might only want limited contact with others.

Aim

To assess whether an effective and efficient co-design approach, AEBCD, can be translated from health to social care.

Objective

To explore, with one local authority, whether an approach to service improvement, known to be effective in health care, could be adapted for use in social care.

The intervention: adapted accelerated experience-based co-design

The adapted AEBCD intervention was carried out in Doncaster, where the local authority had identified loneliness support as a priority and commissioned the creation of the Alliance of VCF support providers. Doncaster Council’s Public Health Lead for Working Age and Healthy Lives was the designated ‘Project Lead’ responsible overall for delivery of the intervention. She worked closely with the Alliance’s Programme Manager, who provided a range of day-to-day support for the local co-design process, including initial recruitment of staff and PWLE, a role described in this report as ‘Co-design Supporter’.

Steps in the co-design intervention

Overview

The intervention in Doncaster followed the two-stage AEBCD process of a series of workshops followed by small group co-design work. The first stage involved three workshops: one workshop for staff and one for PWLE; and a third, joint workshop that brought staff and PWLE together to agree priority areas for improving local loneliness support to be worked on in the second stage of the process. In this stage, smaller co-design groups (comprising both staff and PWLE) focused on one priority area each to co-design suggestions for improvements. A celebration event was held at the end of the intervention phase.

Figure 2 shows a flow chart of the AEBCD process in Doncaster with the timings of events.

FIGURE 2.

Flow chart of the AEBCD process in Doncaster.

An external, professional facilitator, expert in social care co-design approaches, facilitated the workshops. The facilitator worked closely with the Project Lead and Co-design Supporter, assisted by the evaluator, to plan the content and format of the three half-day workshops. While there was uncertainty about whether COVID-19 rules would permit in-person workshops (and a REC amendment was secured to permit remote workshops), in the end, all three workshops took place as in-person events. See Report Supplementary Material 3 for an example agenda (joint workshop).

Staff workshop

The staff workshop was held in November 2021. Nine people took part: seven from VCF organisations, one from a housing association, and one from a private sector home care provider. Seven participants were interviewed in the discovery phase. Following the AEBCD process, the catalyst film was not shown in this workshop, to give the opportunity for staff to first consider their understanding of loneliness and share related support issues from their provider perspective. To help prepare participants to build a broad coalition around improving loneliness support and generate a list of priority areas for improvement, the Project Lead spoke at the workshop about the history of local loneliness support in Doncaster and how it became a local priority. The evaluator presented a thematic summary of the findings from the local discovery phase staff interviews, which prompted discussion around staff understanding of loneliness, hallmarks of good local support and potential areas for improvement. The remainder of the workshop involved staff thinking individually about ideas for improving support, then sharing these ideas in small groups and finally agreeing between groups an overall ‘top five’ improvement list (not ranked):

• Information – to improve information about loneliness support and services available and connect partner organisations

• Group activities – to develop more tailored activities to better reflect diversity in the local population, for example specific groups for people from ethnic minority communities

• One-to-one support – to improve this offer for people needing support to access group activities

• Transport – to improve this and other practical barriers to accessing loneliness support, especially for people on low incomes

• Extending the ‘Circles’ model of support – currently in place locally for older people but seen as having potential to expand to other people experiencing loneliness (a circle of support is a group of family, friends and supportive workers who come together to give support and friendship to a person)

People with lived experience workshop

The PWLE workshop was held in January 2022. Seven people took part, two of whom were also volunteer providers of support. The aim was to bring together PWLE to generate and prioritise ideas for improving loneliness support to share with staff in the third, joint workshop. The Co-design Supporter also shared information about Doncaster’s Alliance of 50+ VCF organisations offering loneliness support.

The workshop began with participants sharing their experiences and feelings around loneliness. They then viewed the catalyst film developed during the discovery phase (see Chapter 2). The film was used to prompt discussion around what constituted good and less good practice in loneliness support. Then, mirroring the staff workshop, participants spent time thinking individually about their ideas for improving support, then sharing these ideas with other participants, finally agreeing an overall ‘top five’ of improvement priorities. A live scribe created a visual representation of the workshop discussions (Figure 3).

FIGURE 3.

Live scribe representation of the PWLE workshop discussion. Reproduced with permission from Stephen Lee Hodgkins.

The top five improvement priorities were summarised by the facilitator (not ranked):

• Information – better information about ‘what’s out there’ in Doncaster for people experiencing loneliness.

• Groups – providing more groups for people to attend where they can meet other people, share experiences, support each other, and get involved in fun and meaningful activities such as cooking, walking, drama, sports, singing and being with animals.

• One-to-one help and support – it can be difficult to come to a group for the first time, so people may need support to do so. Everyone needs help of some kind. People want a chance to receive help and help others.

• Transport – the costs of transport and the need for assistance because of disability, mental health issues or long-term health conditions can create obstacles around public transport.

• Attention to other barriers – it is important that people are treated fairly and not excluded from getting involved in groups or activities. This means paying attention to the various aspects which can make it hard for people to get involved, such as disability, mental health issues, age, race, gender, sexuality and not having English as their first language.

Members of the group volunteered to recap the main points on each topic in the third workshop. To help members prepare, the facilitator circulated a summary of the conversation on each topic.

Joint workshop

The joint workshop was held in February 2022, bringing together staff and PWLE for the first time. Thirteen people took part: six staff and seven PWLE (three of whom were also volunteer providers of support). Eight people had attended a previous workshop: four staff and four PWLE (including one person who was also a volunteer provider). The workshop aims were for staff and PWLE to build a coalition around improving loneliness support, to agree joint improvement priorities and to form smaller co-design groups to take these priorities forward to develop more practical steps.

Participants first shared their reasons for wanting to come together and carried out a practical group task designed to prompt discussion around co-production principles. The catalyst film was then shown – the first time staff had seen it – followed by an open discussion on issues raised by its content. Staff felt the themes in the film resonated with their own work, the main area they discussed being differing views on the extent to which it was appropriate for staff to share aspects of their personal selves to people drawing on their support. This discussion was prompted by a clip in the film in which a person cited this sharing from staff as a valued aspect of support.

Staff and PWLE then shared the five improvement priorities identified at their respective workshops, four of which were revealed as common to both: information, transport, groups and one-to-one support. The fifth priority identified by staff – to extend the Circles model of support – was left aside, since it was known that work was already under way on this topic. The fifth priority identified in the PWLE workshop – attention to other barriers – was also left aside after the agreement that attention to diversity and inclusivity should underpin all improvement suggestions.

With the focus on the four common priorities, participants first decided that they had capacity to create two co-design groups. After a facilitated process involving general discussion and a ranking exercise, it was decided to leave aside the ‘information’ priority. Although information is crucial to finding support, Doncaster Council was known to be repopulating its website information on all local activities post COVID-19 and it was thought more appropriate to review information once that initial work had been completed. The ‘transport’ improvement priority was selected for one co-design group and ‘support’ for the other (covering both group activities and one-to-one support to engage in groups). Before leaving the workshop, participants volunteered to take part in one or other of these co-design groups. A live scribe attended virtually and created a visual representation of the workshop discussions (Figure 4).

FIGURE 4.

Live scribe representation of the joint workshop discussion. Reproduced with permission from Stephen Lee Hodgkins.

Co-design groups

Both the Transport and Support groups held seven 2-hour monthly meetings between March and October 2022, working up improvement ideas for practical implementation. Seventeen people took part in this stage. A few participants withdrew during the 7 months, for personal reasons. Thirteen people remained in the groups throughout (six in one group, seven in the other). Both groups had two PWLE members; other participants were paid and unpaid staff.

Each group had a ‘coordinator’, a role volunteered for by a group member and which was designed to facilitate group discussions, note key meeting points and actions (in a meeting log) and maintain the group’s momentum. The Co-design Supporter role in this part of the process was to support the groups, primarily by carrying out associated practical tasks (agreeing meeting dates, arranging venues, booking refreshments, and administering travel expenses and vouchers for eligible participants) but also being open to supporting the groups in other ways that might emerge as needed during the process (including nurturing individual involvement), given this new and different way of working.

Celebration event

At the end of the period set aside for the co-design small group work, an in-person half-day celebration event was held in October 2022. Although common in previous AEBCD work, this event was not included in the original study proposal. The Project Lead, Co-design Supporter and evaluator discussed the potential value of a celebration event. Three areas were identified: (1) to bring participants back together to share and celebrate their work; (2) to consider what might happen locally to take forward and build on what the process had achieved; and (3) to reflect on the experience of using the AEBCD process.

The workshop facilitator collaborated with the Project Lead, Co-Design Supporter and evaluator on the planning and agenda. Seven participants attended: three PWLE and four staff (of whom two were in paid roles and two unpaid). Other attendees were from Doncaster Council, the Project Lead and two public health colleagues; the Co-design Supporter; the Alliance lead; principal investigator (PI) and evaluator.

A live scribe created a visual representation of the event discussions (Figure 5).

FIGURE 5.

Live scribe representation of the celebration event discussion. (Image drawn by Cara Holland of Graphic Change). Reproduced with permission.

Evaluation

Implications of the change in setting for the set-up and operation of co-design groups

Unlike the use of AEBCD in NHS organisations, where staff and patients are drawn from a single ‘host’ service/organisation, the Doncaster co-design work was undertaken by staff and PWLE involved in varied types of loneliness support offered by a range of (predominantly VCF) organisations in the local community. This had implications for the set-up and operation of the co-design group work.

The main implication was that the co-design groups had a broad remit to work across as many or as few organisations and/or areas of support as necessary.

A second implication of working across multiple services and organisations was anticipation that routes to implementation of co-designed solutions would be longer and more complex than for comparable work within single organisations. The timescale for the small group work (7 months: March–September 2022) was defined by the evaluation period. Groups were free to decide on area(s) of focus, asked only to be in a position at the end of the evaluation period to describe their proposals or make recommendations in sufficient detail to be taken forward locally.

Third, the appointment of co-design group facilitators was also complicated by the number of organisations involved. In NHS AEBCD work, there is the option to draw on in-house staff from the ‘host’ organisation to facilitate the groups. In the current study, without a single ‘host’, the model of seeking a volunteer from each group as its Co-ordinator was used.

For both groups, a staff member volunteered for this role. In one group (Transport), the Co-ordinator ran a voluntary organisation, a role which brought valuable experience and skills, including consensus-seeking and ensuring all voices were heard. In the Support group, settling on a Co-ordinator proved more difficult. The role was initially taken by a staff member skilled in one-to-one listening and sourcing support. At the end of the first meeting, another staff member, experienced in PAR methods, offered to work alongside the Co-ordinator in a support role, for example recording key points as discussions flowed. However, for personal reasons, the original Co-ordinator withdrew from the group, after which two other staff members agreed to rotate the Co-ordinator function flexibly, according to who was able to attend. The member who had taken on the support role continued throughout, although was unable to attend all sessions. When present, their support in helping to facilitate and noting key discussion points on a flipchart to retain for subsequent meetings proved invaluable, both for giving members visual recall and for encouraging momentum to move on from discussion to identifying action points.

Asking the Co-ordinators to complete meeting logs proved unsuccessful owing to capacity pressures and no protected time being available for this task. As a result, co-design group meetings began to lose efficiency, each needing to begin with recall/recap of the previous meeting, and actions promised between meetings being forgotten. In response to this problem, the Co-design Supporter gradually extended her role. She took notes in each meeting, e-mailing participants a summary of key points discussed and any actions agreed. She reshared the e-mail with members before the next meeting as a reminder and contacted Co-ordinators to help shape a plan for the session. This liaison proved more challenging for the Co-design Supporter in the case of the Support group, given their arrangement for two rotating Co-ordinators. The evaluator, who attended all co-design meetings, took meeting notes as a safety net for the Co-design Supporter.

Question 1: what kinds of improvements did the co-design groups make?

This section describes how each co-design group decided on the main support improvements they wanted to influence and what they did.

Question 2: was the accelerated experience-based co-design process acceptable to staff and people with experience of loneliness support?

A key focus of the evaluation was whether the AEBCD approach was acceptable in social care, in this case to the community setting of loneliness support. Overall, the AEBCD process as evaluated in the community setting of loneliness support in Doncaster was experienced positively and considered by participants as worthwhile.

Question 3: what suggestions might participants have for improving the process?

This section builds on challenges outlined in the findings reported above. In addition to data from participants, suggestions from the Doncaster project team and the workshop facilitator are included. It is important to preface these suggestions by noting that this type of feedback from any local use of a process such as AEBCD will necessarily reflect and blend components of the process itself with how (and how well) these were adapted for local use (for the current study, described in Chapter 5, Methods), along with contextual factors such as, in the Doncaster study, recruitment being lower than planned owing to COVID-19.

Overall, there was strong support for the AEBCD process. Improvements were suggested in the following areas: clarifying and reinforcing the process and expectations; ground rules; co-design group facilitation and support; co-design group membership; routes to implementing co-designed improvements; and dealing with endings.

Question 4: what adaptations might be needed for future use of accelerated experience-based co-design in social care?

Experience-based co-design and its accelerated form AEBCD, used in this study, were originally developed as an approach for co-designing improvements in patient healthcare experiences. Transferring the approach to community-based social care identified adaptations to consider in three key areas: the characteristics of the setting; the scope of the co-design work; and how staff and PWLE experiences are framed. These areas are considered below.

Some participants felt using AEBCD in a more analogous single social care service/organisation setting (e.g. a needs assessment process or a care home) may have proved more straightforward than in community-based loneliness support, where provision is delivered by multiple providers. This is not to say that the approach was seen as inappropriate in this context, and indeed a good deal of the evaluation evidence suggests the contrary, but rather it was a stiffer exploration of the approach.

Furthermore, reflecting on the experience of running the workshops, the external facilitator summed up some benefits of loneliness support as an exemplar in not being attached to a single organisation:

I think the nice thing about this was it was almost quite a pure form of co-production [and] it’s very unusual that you are in that situation really, you know, [more often] . . . you’re kind of trying to work out a strategy of how you can introduce co-production into an organisation or into a project or something that’s already got its sort of own momentum, whereas what was nice about this was that the starting point was you had this topic but you were just sort of saying ‘Well what would help, you know, to improve people’s isolation and loneliness?’ in a really open way . . . and so I think that was a really genuine place to start.

Workshop facilitator

Transferring accelerated experience-based co-design to social care, using the exemplar of loneliness support

The evaluation demonstrated the potential for transferring a co-design service improvement approach called AEBCD from its original healthcare setting to social care, using the exemplar of loneliness support. The process generated co-designed recommendations for improving local support for people experiencing loneliness. Participants reported broad acceptability of the approach. Some challenges, such as the quality of group facilitation, capacity constraints on staff to taking part, and complexity around implementation of co-designed improvements, feature in previous studies (e.g. Adams et al. ,27 Locock et al. ,6 Tanay et al. 111).

Donetto et al. 39 note considerable variation in how the concept of co-design is understood and practised in the UK; it can often include limited, one-off feedback or consultation. They use the four dimensions in Bradwell and Marr’s40 definition of co-design to set out how EBCD is operationalised. These four dimensions are summarised as:

• Participation – co-design is a collaborative process in which as many stakeholders as possible have input.

• Development – co-design evolves as a process, maturing and adapting as it takes place: ‘[I]t involves the exchange of information and expertise relating to both the subject of the design process and the process itself. In this sense, co-design teaches co-design.’40

• Ownership and power – co-design shifts power to the process, seeking to transform usual hierarchical power relations between staff and ‘client’ stakeholders and generate collective ownership.

• Outcomes and intent – co-design has a practical focus, although unplanned processes and transformations are likely to occur as side effects of the process.

Bradwell and Marr40 acknowledge, as have others offering frameworks for co-design (e.g. Leach38), that these principles represent ideals which can be difficult to achieve fully in practice. Nevertheless, they serve as a guide to key aspects of co-design that deserve attention.

In relation to the current study, the evaluation found evidence of both achievement and challenges across all four domains. Study recruitment was affected by COVID-19, with capacity issues and competing pressures on local authority social care and public health staff particularly severe. This in turn affected staff recruitment of PWLE. Nevertheless, a sufficiently diverse range of participants, including external stakeholders, provided input to the process. The co-design groups evolved and matured over time as participants built positive relationships and learnt more about the process and how to work as a group. While ownership and power relations differed between the co-design groups, factors in this social care setting suggest a less hierarchical starting point than assumed in health care. This is discussed in more detail below. Finally, the co-design group focus was practical, but with the brief being open to outcomes and intent, the process was exposed to more unplanned developments than anticipated in the fourth dimension of co-design.

When considering how to adapt AEBCD for social care, evaluation findings suggest it is important to reflect on its original healthcare context. Writing about the scope and challenges of co-production in health care, McMullin and Needham14 draw on Bransden and Honingh’s112 differentiation between co-production in ‘core services’ and in ‘complementary tasks’ surrounding core services. For McMullin and Needham,14 this is a useful distinction in health care, where patients cannot co-produce clinical practice or treatments due to the medical expertise required, but their lived experience can help shape its delivery. This recognition of the limits of co-design is reflected in the AEBCD approach, developed for co-designing improvements in the quality of patient experiences along their journey following a defined clinical pathway or setting.

The social care context is different to health, as discussed in Chapter 1. Social care can happen in a range of different settings across a community, with a diverse set of activities and tasks. The evidence base around many interventions is also less developed, and what constitutes best practice is less bounded by professional expertise. This difference widens the potential scope of co-design in social care to include improvements to provision itself. In the current study, this is indeed what happened; one co-design group chose to focus on filling a gap in one-to-one support to help people re-engage with group activities to tackle loneliness post COVID-19. For social care, the potential scope for co-design work to include gaps in existing provision raises issues around how and by whom the scope of the co-design work should be narrowed or, if not, improvement suggestions taken forward and funded.

McMullin and Needham14 identified the hierarchical nature of expertise and knowledge in the healthcare field – fostered by the highly professionalised character of the workforce, which typically requires many years of training – as a particular challenge for co-production. This hierarchy can lead to personal and cultural resistance to the value of patient expertise and a reluctance to engage fully in co-design. 113 By contrast, in community-based preventative social care, such as loneliness support, provision is dominated by VCF sector staff, paid and unpaid. In common with many areas of social care, the relationship between practitioner and client is different to that between doctor and patient; decisions about social care support are often shaped by strengths, or asset-based approaches, which, as McMullin and Needham14 point out, focus on people’s capabilities, rather than a deficit or impairment focus which can prevail in healthcare contexts.

It follows that AEBCD was developed with awareness of aspects of health care which can run counter to co-design principles. This context helps explain why a hallmark of the approach is the catalyst film of patient experiences along their clinical journey. This generates often emotionally charged ‘touch points’ for potential quality improvements. When used in health care, the use of the catalyst film has been reported as having an eye-opening, positive impact on professionals’ commitment to the co-design process. 114

By comparison, in the current study, staff’s emotional reactions to the catalyst film were low key. Several reasons may account for this difference. First, loneliness as a universal experience was observed as enabling staff and PWLE to connect on a human level. This is less likely when co-designing improvements in, say, the patient experience of intensive care, or a hospital cancer pathway, where few, if any, participating staff may have had a corresponding patient experience. Secondly, with multiple providers involved in loneliness support and no single pathway or journey, the current study did not use the film to identify or elaborate on emotional touch points along a journey but instead served to prompt discussion more generally around support. These differences in the use and impact of the catalyst film call into question whether it is always needed, especially given the effort required to produce it. However, its prime aim to prompt discussion was successful, and retention of the film has been urged as a distinctive feature by Donetto et al. following an international survey of EBCD studies:

The film is an important catalyst in the co-design process as the visualisation of patient experiences helps (re)connect people with similar experiences and stories, whether users or providers of a service, and offers an emotionally and cognitively powerful starting point for the co-design process. 20

The authors further suggest that using film makes the process distinct ‘from other consultative formats in which anonymity and circumspection can hinder rather than enable quality improvement’. 20 Publishing the catalyst film on Socialcaretalk.org will make it available for use in similar co-design work elsewhere. For future use of AEBCD in other areas of social care, questions arise around how similar resources might best be created, given that healthcare AEBCD benefits from a longer-standing and more extensive web video archive of patient experiences (https://healthtalk.org) which can be drawn upon for developing films on new AEBCD topics.

When considering the adaptations suggested for using AEBCD in social care, it is important to remember that the study exemplar of loneliness support falls under prevention in social care, in a community setting with multiple providers who draw on a variety of funding sources. This setting is noteworthy for two reasons. First, it does not reflect all of social care, some of which – for instance, the work of local authority social work teams – may have more similarities with healthcare teams, reducing the need for some adaptations when using the AEBCD approach. For example, such single-service/-organisation social care teams may be more comparable to those in health care, in terms of having clearer lines of authority and accountability for taking forward co-designed improvement suggestions. Second, despite differences between the new setting of loneliness support and healthcare settings where AEBCD has been used, the approach produced positive results. Therefore, some adaptations suggested by the evaluation are more a question of degree or nuance than departure from the original model. For example, regardless of setting, time will be needed to build staff rapport, but more time may be needed in a multiprovider setting to build a staff coalition/consensus at the start than in a single-provider setting where staff will have some common understandings.

Transferring AEBCD to a social care setting does nevertheless throw up new challenges. Challenges stem from extending the scope of co-design work into gaps in provision and, in community settings, the lack of a clear host organisation for routes to, and funding for, implementation of co-designed improvements.

It is, however, important to acknowledge that previous studies demonstrate that implementing AEBCD in health care is not necessarily straightforward. Barriers include staff culture being resistant to change and uncertainties around how to take forward improvement suggestions which require commitment from other services or organisations. 6

The commentary at the end-of-project event was striking in terms of how people described the impact on them. For example, one staff member said, ‘we learned from ourselves too’, and she was able to use this learning in her voluntary work with a Chinese Elders group. A PWLE member joined a church-related group due to finding the co-design experience rewarding. Group members underlined the importance of listening to people’s stories and how empowering this could be, not least by learning that they were not alone in experiencing loneliness. It was clear from contributions that the two groups were supportive and encouraging and that people felt a strong sense of achievement and commitment.

The discussion about the project findings more broadly involved robust challenge about the process, its component parts and how it might fit within social care. Ryan received e-mails from attendees following the event which added additional points for consideration. These included that AEBCD could be conceived of as a project to tackle the loneliness of project members regardless of the project’s focus. Concern was raised about the tension between creating a supportive group environment and that group being necessarily short-lived, potentially leaving members with a sense of loss.

The event ended with one attendee reiterating the point that the exemplar of loneliness was a challenging focus for exploring the use of AEBCD in social care. One attendee added that the findings of this project should make it easier for people to use the approach in future since we had, in effect, tested AEBCD ‘to the max’.

Socialcaretalk resource

Ryan held meetings with Monks, Menan Mirza and Vaid to obtain their views on what should be included in the Socialcaretalk resource as it was in development. Monks and Vaid provided useful suggestions for ways of presenting users’ stories on Socialcaretalk, such as including poignant extracts from a diversity of participants to highlight the fact that loneliness is a global problem and to ensure there is a greater chance that people will connect with the resource. Malli also held meetings with five participants from the discovery phase interviews to discuss how their experiences are presented on Socialcaretalk to ensure the resource is accessible. This group felt the example resources shown from Healthtalk.org (sister site of Socialcaretalk) were written for an audience with too high an educational attainment; considerable effort was taken to ensure the new Socialcaretalk resource was written in very plain English.

The resource consists of 21 summaries covering different aspects of participants’ experiences of loneliness illustrated with video, audio and text extracts from the interviews. Individual summaries were shared with our PPI and advisory panel members to make sure that they were accessible, resonated with readers and had no glaring omissions. The feedback was positive, with people reporting that the summaries were powerful, accessible and wide-ranging.

The various meetings have fed into a revised, less text-heavy and more accessible output for publication on Socialcaretalk.org. Furthermore, a recent formative evaluation of Socialcaretalk led by co-applicant Ziebland with Ryan, Martin and SSC PPI member for this study Bostock engaged with potential users of the platform to develop recommendations for how material is best presented, and these were followed in development of the loneliness resource.

Catalyst film for service improvement

Bebbington (supported by Davidson) was involved in the development of the catalyst film during its production. Bebbington and Davidson highlighted areas that were too fast-paced, and requested an easy-read version of the film. Feedback led to the addition of subtitles, a voiceover for the title of each section of the film and an easy-read summary to explain the main points of the film.

Lambat reviewed the catalyst film and felt social care commissioners could learn from it as participants made their points extremely well.

Academic papers

Advisory panel member Kalpa Kharicha, who was the Chief Executive Officer of the Campaign to End Loneliness, was a co-author on the first project paper. 56

End-of-project stakeholder event

All PPI members attended the end-of-project stakeholder event, which was held in Manchester in January 2023. Bebbington introduced the event and briefly talked about her involvement in the project.

Discovery phase

The experience of loneliness and lack of emotional support experienced by people who are unexpectedly bereaved is a neglected topic, including the social needs of bereaved people who do not fit commonly held stereotypes. The findings further highlight loneliness in childhood, relating to specific exclusionary contexts, an area identified by Mansfield et al. 47 Research should identify government mechanisms and social infrastructures that isolate people, rather than viewing loneliness as a deficit located within the person. Our findings support other work that indicates loneliness is socially stigmatised, and it is important to explore factors that contribute to this and ensure interventions do not unintentionally further perpetuate shame and misconceptions around it.

Co-design phase

Areas for future research are: the costs and opportunity costs of the approach compared to more ‘top-down’ initiatives; exploring the purpose and focus of the catalyst film; the impact of AEBCD as an intervention for people who use social care services – what this might mean to participants, and the potential of the approach to generate service improvements; the adaptation of the approach to enable greater inclusion and accessibility; and exploring whether using AEBCD in a more clearly defined area avoids some of the challenges identified in this study.

The availability of the catalyst film based on a national sample of interviews and learning from the current study might help encourage the replication of co-designing improvements in loneliness support in other geographical areas. Such work could lead to refinements and/or additional suggested adaptations of AEBCD.

Socialcaretalk.org includes new sections about social care topics with video and audio extracts of people talking about their experiences. Recent and planned topics include self-funding social care for older people115 [funded by NIHR Research for Social Care (RfSC)], the experiences of social care leavers (funded by NIHR RfSC), and the experiences of carers of people with dementia and people with early-stage dementia (funded by NIHR SSCR). Catalyst films for service improvement are a potential output of these projects.

Recommendations specific to using accelerated experience-based co-design in social care

• Identify people or organisations who potentially could have responsibility for implementing improvements, including finding relevant funding.

• Identify an appropriate sample of staff and PWLE, taking time to fill gaps in representation of provision, knowledge and people’s characteristics, and consider whether staff and PWLE have distinct or shared experiences and how to build on these.

• Allow the necessary time for coalition-building, developing trusted relationships and understanding different perspectives.

• Ensure information about the research and co-design phase is clear and accessible and reshare this at the start of each meeting to highlight what stage the project is at and what is left to do.

• Consider whether PWLE and staff participants have pre-existing relationships or should be selected on account of these, and the impact of having or not having such relationships.

• Consider opportunities for co-design group members to continue contributing their expertise.

General recommendations

• Ensure good facilitation of the workshops and the co-design group work. Facilitation is especially important for agreeing ground rules and creating the right atmosphere for the honest sharing, open discussion and mutual respect required for the process.

• Ensure that paid staff involved in the co-design – whether as participants or supporting the process itself – have protected time for the work involved.

• Be clear about processes, aims and roles from the outset. This includes taking time to explain the EBCD process of small groups working over an extended period and being clear about the time commitment involved. In small group work, ensure expectations, goals and roles are discussed and agreed. Remember to discuss how and when the co-design process will end.

• Ensure that groups are large enough to represent all relevant parties. Not all members will be able to attend all meetings, so an over-representation of PWLE can help ensure individuals do not become a lone voice for PWLE in some meetings.

• Keep the first two workshops separate – one for staff and one for PWLE. This helps to create safe spaces where participants can build trust and think about issues from their perspective.

• Consider aspects of the process which may exclude some people, such as the workshop format, and any adaptations to accommodate these, for example to enable deaf people to participate.

• Ensure co-design group participants know that they can seek outside views and bring in external experts, as necessary.