Implementing community participation via interdisciplinary teams in primary care: An Irish case study in practice

Abstract Background Community participation in primary care is enshrined in international and Irish health policy. However, there is a lack of evidence about how stakeholders work collectively to implement community participation within interdisciplinary teams; community perspectives are rarely captured, and a theoretical underpinning for implementation of community participation in primary care is absent. Objective To conduct a theoretically informed, multiperspectival empirical analysis of the implementation of community participation via primary care teams (PCTs) in Ireland. Methods/Design/Participants Participatory learning and action (PLA) focus groups and interviews were held with 39 participants across four case study sites within a nationally funded programme designed to enable disadvantaged communities to participate in primary care. Normalization process theory (NPT) informed data generation and analysis of how diverse stakeholder groups worked together to implement community participation via PCTs. Results The various stakeholders had a shared understanding of the value of community participation on PCTs. Motivations to get involved in this work varied, but were strong overall. Challenges to enacting community participation on PCTs included problems with the functioning of PCTs and a lack of clarity and confidence in the role of community representatives at PCT meetings. Informal appraisals were positive, but formal appraisal was limited. Discussion and Conclusion The implementation and sustainability of community participation on PCTs in Ireland will be limited unless (i) the functioning of PCTs is strong, (ii) there is increased confidence and clarity on community representatives’ roles among all health‐care professionals, and (iii) more sophisticated methods for formal appraisal are used.


| INTRODUC TI ON
Community participation in primary care has its origins in the Alma-Ata Declaration of 1978, which stated that "people have the right and duty to participate individually and collectively in the planning and implementation of their health care." It is defined as: a process by which people are enabled to become actively and genuinely involved in defining the issues of concern to them, in making decisions about factors that affect their lives, in formulating and implementing policies, in planning, developing and delivering services and in taking action to achieve change 1 (p. 10) Since then, the concept of involving patients and the public in health-care planning has gained acceptance and is enshrined in health policy across a range of international settings including the UK, 2,3 Scotland, 4,5 Wales, 6,7 Canada 8,9 and New Zealand. 10 There are examples internationally of individual and collective processes to implement community participation in primary care. [11][12][13][14][15][16][17] In recent years, collective processes have been adopted in several countries: for example, citizens juries 14 and patient participation groups 15,16 in the UK, citizen juries 13 and community representatives on health service committees 18 in Australia, dialogue sessions 19 in Canada, mixed advisory committees (MACs) 20 in Italy and community participation in primary health-care organizations 21,22 in New Zealand.
In this study, we focus on collective participation in primary care, which can overcome the reductive individualistic approach to health-care participation 20 and create a more efficient and effective health-care system. [23][24][25][26][27] It has also been shown to enhance the delivery and uptake of health interventions to address health inequalities, [28][29][30][31][32] and increase community cohesion and leadership. [31][32][33][34] Despite this international policy context and efforts to implement community participation in primary care, there are major gaps in our understanding of its purpose, processes and outcomes. 32 There are limited data across the multiplicity of stakeholder perspectives on implementing community participation in primary care in practice, and community perspectives are rarely captured. 27 Furthermore, there is a lack of evidence for how the various stakeholders work together in a primary care setting to implement community participation within interdisciplinary teams. Lack of clarity and agreement between stakeholder groups about the roles of community representatives remains a major obstacle to effective community participation. 18,[35][36][37] Although theory has been used to understand how far patient and public involvement (PPI) was embedded within health-care research in certain areas, 38 there has been no use of theory to study community participation in practice despite the call for theoretically informed, empirical analysis of implementation to generate insights and transferrable lessons for community participation in primary care across settings. 39 This is a priority for research, policy and practice. 32 In Ireland, community participation in primary care became enshrined in health policy with the launch of the 2001 primary care strategy. This strategy sought to transfer most health-care provision into the community to be delivered by interdisciplinary primary care teams (PCTs). 40 PCTs were encouraged to ensure community participation in service planning and delivery. A greater input from the community and voluntary sector was proposed to enhance the advocacy role of PCTs. 40 Despite this, and other interim measures such as the national strategy for service user involvement, 41 involvement of patients and communities in the development and running of PCTs is not routine practice across the country, 39,42,43 is hard to achieve, 44 and is generally not regarded by service providers as an important resource for PCTs. 45 Therefore, the aforementioned gaps in international literature are also relevant to the Irish context. 43,46

| R ATI ONALE FOR THIS S TUDY
The aim of this study was to address these international and national gaps in knowledge and to conduct a theoretically informed, multiperspectival empirical analysis of the implementation work that has taken place in Ireland to embed a programme of community participation in primary care (known as the Joint Initiative). This study focuses on the implementation of community participation on PCTs.

| Study context
This study took place within the Irish primary health-care context following the end of a nationwide funded initiative-the Joint Initiative (JI)-to support community participation in primary care.
As a function of the JI, a range of community participation activities were developed including community needs assessment, health promotion and mental health awareness programmes, and community representation in the development of local primary care services 42,47 As mentioned above, the focus of this study was on collective community participation processes on PCTs in Ireland.

| Study design
The analysis in this study is drawn from a larger qualitative retrospective case study (2011-2014) of the JI programme. The design of the study was in accordance with Yin's recommendation for use of case studies to explore a phenomenon within its real-life context. 48

| Sampling and recruitment
Following the principles of purposeful sampling, 49  experience with community participation, the various populations involved, and the "successful" and "less successful" interactions with PCTs (see Table 1).
The research participants (n = 39) were identified and invited to participate in the study via gatekeepers at the four case study sites.
Gatekeepers were paid project coordinators at each site who communicated with community representatives and health service employees and managers about the study and extended the invitation to them to participate in focus groups or interviews.
Participants were categorized as follows: 1. Community representatives* who had been involved in the JI demonstration projects and had some experience of interacting with PCTs within this context (n = 27).

| Ethical approval
The Irish College of General Practitioners (ICGP) in Ireland provided ethical approval for this study.

| Data generation
We employed normalization process theory (NPT) to inform data generation and analysis. See Box 1.
Participants were contacted via gatekeepers and chose their preferred method of data generation (ie semi-structured interviews or participatory learning and action (PLA) focus groups). PLA focus groups and data generation methods 50,51 were used with community representative groups where possible. PLA focus groups involve the use of PLA techniques with inherent visual and analytic techniques.
They were valuable because they allowed community representatives' perspectives to be shared across and between participants and for preliminary data analysis to be conducted in a collaborative and participatory fashion. 52,53 see Box 2. These techniques have been previously used with migrants and people with aphasia. [53][54][55][56][57] Community representatives chose focus groups as their preferred method of data generation as these research sessions were held to coincide with their usual scheduled meetings, which was convenient and time-efficient. Community representatives also indicated that it was a welcome means to reflect together on their community participation practices and their shared experiences of interacting with the PCT.
Interviews were favoured by health-care professionals, GPs and HSE service planners and policymakers, allowing the participants to *Community representatives in this context were defined as: "individuals, who are 'representing', representative, and/or 'consultative' of one or more populations or affinity groups. They can be stakeholders, opinion leaders, organisers and advocates. They serve as a platform and channel for information and voices of community, communicating ideas and concepts between community and health and social services and who hold people and processes accountable" (p. 34). 41  speak within their own conceptualization of the phenomenon of community participation in primary health care and to make this explicit. 58 They were more convenient for this cohort of participants as the interviews were scheduled at a time and location suited to the individual and did not interrupt their busy schedules of work.
Gathering data from both focus groups and interviews provided rich narrative accounts, which were analysed for shared and differential perspectives and experiences between and among the participant groups, and across and between case study sites (Table 1).

| Data analysis
All interviews and focus groups were recorded and fully transcribed for analysis. Participants chose a pseudonym to maintain anonymity.
Two researchers were involved in the focus groups, ET and RME.
ET undertook all interviews. Data analysis for the wider project pertaining to community participation in primary health care was led by ET and deliberated in data analysis meetings with AMF and RME.
Data analysis for this study specifically focused on data pertaining to community participation on PCTs and was led by ET. Analysis was then discussed and developed with AMF and AH.
Deductive data analysis 59 was informed by normalization process theory (NPT) using NVivo. While there were different data generation methods used, with implications for group reflection (focus groups) versus individual conceptualization (interviews), data from both methods resonated with the four constructs of NPT. This indicates that the data generation methods did not impact on the conceptual nature of the results.
Findings per construct were analysed asking "how strong is the implementation of community participation in PCTs?" There is no recognized system for this layer of NPT analysis. Therefore, a working definition for strong implementation was developed by the research team (see Box 1) and was used as a benchmark to classify the implementation as strong, medium or weak (

| Quality and rigour
Several steps were taken to increase the quality and rigour of our results. 60 These included the following: recording of reflective notes during fieldwork, regular data analysis clinics for NPT analysis, member checking with participants via email and face-to-face meetings as well as feedback sessions with participants. NVivo 10 software was used to facilitate data coding and analysis and sharing data across the research team. These steps were continued until there was sufficient, thick description in the data, that is until data saturation had been reached. 60

| Coherence: Can stakeholders make sense of community participation on PCTs as a new way of working?
All participants in the study considered that community participation on PCTs was about meaningful reciprocal relationships between stakeholders to represent the voice of the community in primary health-care delivery generally and at the PCT meetings more specifically. For GPs, the introduction of community representatives on PCTs elicited a fear that they would lose control of their work, and this was a concern for them at the start of the process.

| Cognitive participation: Will stakeholders engage with/"buy into" community participation on PCTs?
Buy-in to this way of working for community representatives happened because they were invited to be a representative on the PCT by "champions of the JI" who were known to them. This was usually a community development worker, project coordinator or PCT work TA B L E 2 Levers and barriers to community participation on primary care teams (PCTs) using normalization process theory (NPT) constructs to evaluate implementation; synthesis of findings across research sites [61][62][63][64][65]  Cognitive participation was strong. This means that stakeholders from all groups bought into this way of working because they were invited by champions, and existing relationships supported the work.
There were also complementary, differential motivations for community and professional participants that fuelled interest and responsiveness to invitations to get involved in the JI.

| Collective action: What do stakeholders need to enact community participation on PCTs in daily practice?
All participants talked about the importance of a paid role to coordinate this work. Where community representatives did get to participate in PCT meetings, there was a lack of clarity among some health professionals about the precise role of the community representatives at those meetings. There were misunderstandings about issues such as loss of confidentiality at meetings, and what the community representatives were trying to achieve.

I don't think it would happen unless somebody specifi
But we were trying to kind of get across the idea that the community reps weren't here to discuss specific clients, they were here to discuss broader issues and they could bring stuff to us and we could advise them of things that they could share with the community, but the team wasn't ready, that's the reality. Collective action was moderate. This means that available resources and training were important levers for enacting community participation on PCTs. However, the PCTs were not sufficiently developed for community participation to operate effectively. This impacted on relationships in the team, and community representatives did not feel that they were viewed as equal partners at the PCT meetings.

| Reflexive monitoring: Can stakeholders formally or informally appraise the impact of community participation on PCTs?
All participants agreed that community participation on PCTs is hard to evaluate or measure. Reflexive monitoring was weak. This means that informal appraisals of community participation on PCTs were quite positive, but it was hard to formally evaluate or measure. The scope for sustaining the work and transferring lessons learned to other sites was considered to be very poor, particularly in the context of the economic recession that decimated resources.

| Summary of key findings
There was a shared understanding about the idea and potential value of community participation on PCTs among stakeholder groups involved in the JI across roles and case study sites, but this did not hold across the wider network of stakeholders on PCTs and community.
Stakeholders across groups bought into this way of working because they were invited by passionate and convincing "champions." Existing relationships and complementary motivations also fuelled buy-in.
There were positive examples of enacting community participation on PCTs, supported by available resources and training.
However, it was challenging because it is time-consuming work for those in management roles. Furthermore, it was taking place against the background of poorly functioning PCTs as well as confusion and concern about community representatives' role at PCT meetings.
This thwarted health-care professionals' confidence in the work and inhibited meaningful engagement experiences for community representatives.
There were informal, positive appraisals of impact from most stakeholder groups. There was also consensus that impacts are difficult to capture formally and that sustaining and transferring the work that had started was going to be very difficult in the context of the economic recession.

| Comparison with literature
Similar to findings about PPI in research, 38  there was general enthusiasm for community participation in planning primary health care via PCTs across stakeholders in this Irish study. 18 The potential benefits of community participation on PCTs, such as improved service delivery and increased awareness, were recognized. [24][25][26][27]69 Visionary leaders who are committed to working with communities were an essential ingredient of encouraging buy-in and commitment to community participation. Community workers acted as what have been identified elsewhere in the literature as "boundary spanners," 70,71 which means that local people were drawn into the process and, with increased confidence, became advocates and translated and mediated between local people and professionals. 21,70 However, despite a considerable investment of resources through the JI to build capacity for this work, clarity and agreement between different stakeholder groups about the roles of community representatives was problematic, as cited elsewhere, [35][36][37] and GP concerns about the potential for negative impact on their practices was reported. 21 From an NPT perspective, this lack of clarity and confidence will undermine the workability of community participation in PCTs in practice. The challenge seems to be in reaching the full network of relevant stakeholders to enhance understanding, engagement and readiness for community participation on PCTs.

| Implementation and enactment
It is not possible to consider community participation outside a political context. 72 This analysis has highlighted that there were two political innovations at play In Ireland at the time of this study: the introduction of primary care teams via the primary care strategy and the introduction of community participation on PCTs via the joint initiative. The problems with full implementation of interdisciplinary team working are not unique to Ireland. 73 From an NPT perspective, in this analysis, while this dual interplay did not seem to impact so much on sense-making or engagement processes, it clearly impacted on the readiness of PCTs to enact community participation on PCTs. Put simply, community participation on PCTs, without a proper PCT structure, is hard to enact.
Participants in this study were adamant that PCTs should be fully resourced and running effectively before community participation is introduced.
The implementation and sustainability of community participation in PCTs in Ireland will be limited unless the functioning of PCTs themselves is stronger, there is increased confidence and clarity on community representatives' roles among all health-care professionals, and more sophisticated methods for formal appraisal are employed. Future research could investigate how training in methods to enact community participation on PCTs could enable shared understanding to be achieved and clarity of roles to be developed.
Evaluation strategies could be built into team processes early on to investigate impact and outcomes on PCT activities. Evaluative frameworks that capture a range of outcomes including unforeseen ones should also be developed.

ACK N OWLED G EM ENTS
We acknowledge the contribution of the following to the wider study and thank them for their support: Jane Pillinger, Rachael King, participants and gatekeepers.

CO N FLI C T O F I NTE R E S T S
None to declare.