Identifying the processes of change and engagement from using a social network intervention for people with long‐term conditions. A qualitative study

Abstract Background Personal and community networks are recognized as influencing and shaping self‐management activities and practices. An acceptable intervention which facilitates self‐management by mobilizing network support and improves network engagement has a positive impact on health and quality of life. This study aims to identify the processes through which such changes and engagement take place. Methods The study was conducted in the south of England in 2016‐2017 and adopted a longitudinal case study of networks design. Purposive sample of respondents with long‐term conditions (n = 15) was recruited from local groups. Barriers and facilitators to implementation were explored in interviews with key stakeholders (5). Results Intervention engagement leads to a deepening of relationships within networks, adding new links and achieving personal objectives relevant for improving the health and well‐being of users and network members. Such changes are supported through two pathways: the mobilization of network capabilities and by acting as a nudge. The first is a gradual process where potentially relevant changes are further contemplated by forefronting immediate concerns and negotiating acceptable means for achieving change, prioritizing objective over subjective valuations of support provided by network members and rehearsing justifications for keeping the status quo or adopting change. The second pathway changes are enacted through the availability of a potential fit between individual, network and environmental conditions of readiness. Conclusions The two pathways of network mobilization identified in this study illuminate the individual, network and environmental level processes involved in moving from cognitive engagement with the intervention to adopting changes in existing practice.


| BACKG ROU N D
There is a recognition that providing person-centred care and understanding what people with long-term conditions value in relation to self-management requires exploring the contexts and ways in which social ties and resources shape everyday interactions and mechanisms through which changes in existing practice are negotiated. 1,2 Social network interventions designed to mobilize resources have to compete alongside pre-existing practices and manage interactions between people and their contexts to ensure the acceptability, workability and integration of new ways of doing things in everyday life. 3,4 Two ideas underlie the development, deployment and successful implementation of a social network intervention Genie (generating engagement in network involvement). Firstly, self-directed support for managing health can be accessed through people's social networks and engagement and is predicated on the wide range of connections available to people in open settings (family, friends, groups, acquaintances and pets). The latter provide opportunities for connectivity reciprocity and accessing resources amongst network members for support. 5 In terms of living and managing well with a long-term condition (LTC) this means realizing and sustaining valued activities and participating in social, cultural and group activities 6,7 and maintaining and developing valued reciprocal relationships with others within proximate communities. 8,9 The social network intervention considered here is facilitated and includes mapping and reflecting on the composition of personal networks, eliciting preferences, and considering options for engaging with local and online resources, groups, people and organizations. 4 It is predicated on the notion that people with long-term conditions are more likely to engage with relationships, things and activities they choose and value. 8 When delivered by trained facilitators in a community setting (supporting people with diabetes and early stage CKD), Genie led to an increase in diversity of participants' networks, greater engagement with community activities and had a positive impact on blood pressure, health-related quality of life and lower health-care utilization. 4,10 However, uncertainty remains about the processes through which these changes occur and how network engagement activated by the intervention interacts with the relevant contextual, network and individual level factors within people's everyday lives. Here, we are interested in developing a better understanding of how this process is shaped by the structure of people's networks and the immediate environments within which they are located.

| Recruitment and data collection
The study was conducted in the south of England in 2016-2017 and adopted a longitudinal case study of networks recruiting a purposive sample of respondents who were over 18 years old and living with long-term conditions (n = 15). Local voluntary and community groups that supported this population were visited in person by a researcher or a PPI representative or were contacted via online support networks. Respondents included people of different ages (45-84), and varied by gender, income, employment and marital status, and number of network members (Table 1).
Each participant met with a facilitator face-to-face at two time points, with a 3-month interval in-between. The baseline meeting lasted 45-90 minutes and was followed by a qualitative interview with a researcher, lasting approximately 60 minutes. The 3 months follow-up focused on the network mapping stage and lasted 30-40 minutes. Facilitators came from a range of backgrounds including care navigators, community navigator, local area co-ordinator, PPI representative, public health practitioner and applied health researchers.
We collected qualitative data about the processes of implementation and the outcomes of engagement or non-engagement with personal networks and online and off-line resources. We used observation and in-depth interviews at two points in order to elucidate the complexities of social practice and multiple actors over time. 11 A researcher observed intervention delivery using note-taking and video recording, and focused on user-facilitator interaction, and contextual, individual and network factors of potential relevance for users in adopting changes in practice. Following each observed case study, the researcher interviewed the participant and wrote field notes including impressions of how the intervention was used and accepted. Three months after the intervention all respondents were interviewed again in order to explore changes in the structure of personal networks, engagement with social network support, and accessing services and devices relevant for self-management support. The follow-up interviews included a "think aloud" method 12 where the interviewees were asked to comment on the challenges they experienced in using the resources discussed at baseline. We were interested in how users approached, accessed, navigated and engaged networks and resources of support as informed by previous evaluations of e-health and SMS tools. 13,14 In order to explore how the social and physical environments shaped network activation, practice change, and to identify barriers and facilitators to the implementation and long-term sustainability, we set up a working group, which included health trainers, representatives of adult services, public health, representatives of voluntary and community organizations (n = 15). We kept extensive notes of working group meetings and informal discussions with key local decision makers and interviewed five members of the WG involved with different aspects of the implementation process (managers and intervention facilitators from voluntary organizations and local service providers).

| Data analysis
The analysis drew on normalization process theory and focused on understanding how coherence and cognitive engagement developed during the intervention 4  Lower than average members in adopting changes in their everyday practice and the reflexive monitoring of this process over time. 15 A coding and analysis framework described the extent and nature of changes made by users over three months; the relevant factors, the types of work done by network members, and the processes involved in making these changes; the selective engagement of network members (navigation) and the process of reshaping existing relationships (negotiation) in making new connections, improving capacity to enact healthy behaviours, improving well-being, reducing isolation. 16 The coding framework was agreed collaboratively by members of the research team. Any coding differences were discussed at regular meetings in order to reach agreement. In analysing the data, we used comparisons and drew out new improvements and benefits specific to individual circumstances.

| FINDING S
Our findings related to processes and change in personal networks.
Most users reported increased number and frequency of network contact identifying additional members of personal communities who they thought were important to them, but who had not been previously identified ( Table 2). The intervention was effective in extending user networks by adding new groups and activities (eg walking group and Parkinson's support), tools (eg pedometer, weightwatcher points converter, laptop and mobility scooter) and engagement online (Table 3). Users with small and family or friend-centred

| Building capacity for articulating, reframing and re-orientating relationships and capabilities
Respondents found that visually mapping their network and discussing this with the facilitator opened up space where they felt listened to, "had the opportunity to express feelings," it was like "a warm comforting exercise" (ID7) that allowed "time for myself" (ID1). It was apparent that discussion opportunities where one did not feel "categorised, stigmatised" were valued by respondents but not al- The discussion at T1 and T2 made it apparent that in some cases there was lack of fit between opportunities for engagement, network capacity and personal priorities. Using Genie supported a process of articulating and engaging with personally defined objectives and personal community members. Cognitive engagement offered a set of reference points for reframing self-management support in network terms and for identifying potentially relevant changes to existing practices. However, these needed further thinking through in terms of identifying the rationale for making changes and identifying alternative activities that might lead to more substantive change. This process included negotiating objectives and engagement with network members, forefronting the items of most preference and value and rehearsing justifications for these.

| Negotiating objectives and engagement with network members
The main initial focus was on engaging network members and aligning users to local preferred activities which the participant had not previously tried. However, the option of immediate engagement was not always possible if the options were seen to be currently unachievable due to incapacity, or required yet to be negotiated access, resources time and effort.
Yes, I still want to join the W.I. which is one of the things that I want to do but it will be a few weeks until I feel well enough to walk up there because that's the For some users, engagement with the intervention failed to deepen or extend network engagement, but brought about an enhanced awareness of the value they put on maintaining existing activities and the individual and network resources that these required (ID3).

| Forefronting evaluations of network support
The work that different network members do to manage things and the value of this to respondents was sometimes presented in procedural terms with clearly defined boundaries and responsibilities.
However, in some instances respondents started identifying potential tensions between subjective and objective valuation. daughters. He talked about changing his will to reflect the loss of relationship, which troubled him as they were "his blood" yet "they weren't interested." He contrasted that with the supportive relationship with his son-in-law and stepbrother who, even though living in the United States, came over and stayed with him when his wife died. At T2, the respondent put many members of his US family on the diagram and had regular FaceTime conversations with them.
Recognizing the value of some of the less intimate (weak) ties was in some instances subsequently accompanied by the extending and deepening of such relationships. Thus, ID6 thought her volunteering work "is a lifeline" that offered her a respite from the difficult relationship with her partner, and at T2, she was able to increase the time spent there.
Another respondent (ID11) felt that he improved his skills and deepened his involvement with the walking group he was attending when he started playing the guitar with one of the group members.

| Rehearsing justifications in engaging others
Renegotiating relationships and roles, and mobilizing network engagement involved developing justifications for change to support arrangements that were acceptable for respondents and appeared so for members of their personal community. The initial Genie discussion led to revising and rehearsing changes to views and positions about individuals within their networks. For example, although ID10, who had MS, experienced financial difficulties he found it difficult to accept that he might need to apply for carer's allowance to which he was entitled.
He felt that this was morally wrong, a view shared by his mother as she thought "he gets enough already" and did not need to accept additional financial support. Although shifting this view was difficult for the respondent, he resolved it by arguing that the money would be spent on getting "nice things" for his parents and going on holidays "to the cottage in Cornwall" that "we all love." Additionally, this was justified because his mother was doing "huge amount of voluntary work for other people and deserves some acknowledgement." But this money would also make it possible to help financially his partner and stepdaughter in Argentina. At T2, he took a decision to ask the Genie facilitator to help him with "doing the forms" and claiming the allowance.

| Nudging a link to enabling environments and activated networks
For some participants, the intervention coincided with the contingencies of a fortuitous combination of an activated personal community and a supportive environment. In this context, the intervention acted as a tipping point towards changes that were already part of an ongoing discussion within people's personal communities.
For example, the wife and daughter of one respondent were in the process of looking for someone to help him get up and dressed in the morning, as his wife was finding it increasingly difficult to help him physically. The respondent was concerned how he would cope as "I wake at different times" and that if he got different carers he would "have to teach them my routines" although he recognized that "…my daughter is anxious that I shouldn't wear my wife out" (ID5).
At T2, the personal care has been arranged, fitting in with a neighbour who had the same carer so that "we would probably fit in around her. So, if she is seen say at 9 am, she'd come here at 9.30… probably once a week." Although the respondent still felt "a bit ambivalent because I've never had that kind of support before," he and the members of his family were able to make this change more acceptable by likening it to them employing a weekly cleaner who has now "become more like a friend" and "a ray of sunshine." In other cases, participating in the intervention created a "nudge" 18 This respondent was able to extend her walking activity by arranging to walk with her daughter "two or three nights a week" and by linking up with her friend with who she used to walk in the past.
In explaining this change, a narrative link to other contextual and personal factors was made: "had my knee done," "got over the op and had the stitches out," and the "summer came and the lighter evenings came and we went out to different things," "different garden centres on the island." Similarly, a nudge might be made towards reorganizing network support in a new context. For example, one participant, with multiple mental and physical health problems who lived alone, realized she was quite isolated and that most of her con- So, she is walking more in part "because I'm worried […] because they put me on that Clexane to prevent thrombosis and DVT and so obviously I need to be mobile." For all respondents, engagement with new activities tended to fit with familiar activities, such as joining walking groups or starting walks with a network member, while more complex and unfamiliar changes were less likely to materialize as they required more time for engagement and additional support from members of their network.

| Environmental fragilities in engaging and sustaining practice change
Participant engagement with the intervention highlighted differences in how sustainable engagement with new activities was co-shaped by the type of groups accessed, the availability of longerterm facilitator support and the structure of personal communities.
Some of the organizations that users linked with had existed for a long time, had stable funding structure and opened possibilities for user engagement that were self-organizing and entirely focused on the evolving user preferences and needs.
Oh yes, this last couple of weeks there is this one guy who has a massive allotment and he's been bringing runner beans, tomatoes and loads of veg and he puts them there and you take what you want and just make donation to the club. Things like that which is nice.
You are building up a social group, aren't you and the of course there are the activities they put on, trips out, there is a variety she on, they have quizzes, …I put quizzes, together, this is something I enjoy doing… We are trying to start up a pétanque club.
Such organic growth and engagement with network members although narrower in scope was discussed in relation to engagement with self-organizing groups of colleagues, the "banter club," or church groups (ID5).
However, other groups were small, poorly funded and their continued existence depended on the ongoing support of the users.
ID11, for example, relies heavily on support from a staff member at one of the resource centres he attends. They have been sorting a lot of household/domestic issues together, and at T2, the respondent was using "we" rather than "I" to denote a feeling of support. Uncertainty about the remit of services, roles and responsibilities of link workers, and long-term funding commitment were also recognized by the local stakeholders as having an impact on the implementation of Genie and on maximizing its effectiveness in supporting network activation and change.
[people with complex circumstances] need more support than just identifying there is a group at the end of the road, they might not actually be able to get to the end of the road so they need more support with finding a volunteer who can potentially pick them up for example to take them to that group.
This also reflected a broader systemic problem: …engagement I think, wider than just health and social care would be great because ultimately if we are looking at things holistically that would be great and I think at times it's been very health and social care orientated as most things on the Island tend to be. I think that would help to support it and getting that wider network. Again, in a wider system barriers which is hard for Genie to be able to get over that because ac-   justifications, developing narratives and rehearsing the sequencing of potential changes in practice. This is likely to be a gradual and reflexive process. By contrast, the nudges towards realignment of support were seemingly made possible through the availability of a potential fit between individual, network and environmental conditions of readiness. In such cases, engagement with Genie acts as a steer towards readjustment within conditions that already exist and only require minimal change. For example, engagement with weak ties within personal communities could potentially act as a nudge towards change by providing a missing link or type of support that makes everything else fit (eg acting as a companion for walks, where starting walks is already an immediate priority due to professional advice about taking a medication, where there is easy access to a safe and walkable area, and past but discontinuous experience of going for walks).

| D ISCUSS I ON
Our findings suggest that the mobilization of network capabilities might be seen as a useful pathway to supporting changes to individual circumstances because it highlights a process of engagement with the current concerns of individuals and their network members.
Navigating and negotiating relations within personal communities is a condition for engagement with network-based interventions such as the one reported here with indications that it enhances existing capacity for long-term condition management work. It may also indicate the building of individual and collective resilience and flexibility in adapting to the changing needs of people with LTCs in terms of managing everyday life. 25,26 In this regard, access to different types of ties which make up a personal community is likely to be relevant through the properties of interaction. Thus, weak ties can act as a counter to strong tie connections by avoiding the need to make changes in relations that are both valued and difficult to change, avoiding or reducing the burden on strong ties, providing a wider range of options. 27 This study indicates that people with limited resources, smaller networks and lower levels of community connections are more likely to be supported through network engagement and negotiation.

| CON CLUS I ON S AND P OLI C Y IMPLIC ATIONS
The Genie intervention appears to be effective in bridging the gap between cognitive engagement with a network framed understanding of self-management support through network mapping and preference elicitation, and its activation in the context of people's everyday life.
The two pathways of network mobilization towards adopting practice changes identified illuminate interdependencies between individual, network and environmental level processes and highlight potential challenges for its future use as a scalable intervention for supporting long-term condition management. The impact of Genie in activating networks and supporting behaviour change is likely to be enhanced by the availability of local resources enabling people to live well. 9

ACK N OWLED G EM ENTS
We would like to thank Ms Sandy Ciccognani, the Local Area