Community participation to design rural primary healthcare services

This paper explores the outcomes from inviting community members to participate in designing primary healthcare services for remote rural places. In our study, healthcare models were designed that address communities’ priorities and are affordable within existing budgets. The community participation process we used involves local people in service decision-making as is desired within contemporary healthcare policy.

Internationally, remote and rural places are changing due to conditions of global capitalism [1]. One result is that younger, working age people become concentrated in metropolitan and commuter areas, leaving concentrations of older people in smaller, more peripheral, remote rural settlements [2]. With regard to services, larger regional centres tend to have a range of primary healthcare and a general hospital, while smaller towns might have a hub of more limited primary healthcare services and a community hospital. Suitable service arrangements for small remote settlements are hard to define [3] and remote places are vulnerable to small changes in population and healthcare providers.

Contemporary advice on how to design primary healthcare services spans a range from providing rational, algorithmic models that suggest workforce based on accessibility to services and amenities [4] or need, combined with evidence of effectiveness [5] through to suggesting population health planning founded on discussing local priorities in relation to need and social determinants [6]. While there is strong direction from governments internationally to use community participation in local service planning [7, 8], there is little clarity about what to do, what outcomes to expect and how to incorporate evidence about health innovation [9].

This paper shows outcomes from deploying a community participation process to design healthcare for remote communities. It draws on findings from a study ‘Remote Service Futures’ (RSF), the primary purpose of which was to devise a feasible methodology for remote community members’ participation in health service reconfiguration [10]. In 2008–10, within the action research process to devise the methodology, residents of four remote Scottish Highland communities were invited to participate in planning to identify their local health priorities and design ways to address these. This paper outlines the community participation methodology derived and the healthcare designs produced.

The Scottish Government Urban–rural Classification defines the settlements included in this study as “very remote rural areas” [11]. For shorthand, we call them remote. The Scottish Government defines very remote rural areas as having fewer than 3,000 inhabitants and being over an hour from a settlement of 10,000 or more. There are different categorisations and understandings of the concepts of remote and rural depending on countries’ size, population and geography, but we suggest that, internationally, remote areas share features of sparse population, distance from and therefore inaccessibility to, specialised services [12] and choice of services, and identification as geographically peripheral in the national psyche.

Relative to other UK states, Scotland is well resourced with general practitioners (GPs) and nurses [13] and these are more evenly distributed throughout the country than in other nations [14]. Remote health services are provided by a mix of GPs (either employed as National Health Service (NHS) independent contractors or salaried), community staff employed by regional health authorities, Scottish Ambulance Service personnel, council and voluntary/non-profit organisation workers. NHS24 is a national first response phone triage service and there are after-hours service schemes to see a primary healthcare practitioner. About 6.4% of Scotland’s 5.3 million population live in remote areas [15]. Remote and rural areas are distinguished by higher suicide rates, incidence of alcohol related disease, numbers of accidents and palliative care workload [16]. Compared with its urban areas, remote Scotland does not have severe socio-economic disadvantage, but there are pockets of disadvantage [17].

Internationally, problems of remote and rural healthcare include centralisation, lack of chronic condition care, health worker shortage, failure to adequately address prevention and lack of infrastructure for co-ordinated, integrated care [18]. Decades of policy to incentivise rural workforce and introduce tele-health have had very small impacts in the face of systemic and societal disincentives. Those concerned with redesigning healthcare for remote areas have had to formulate creative ideas about service provision and planning. At the time the RSF study started, a policy document Delivering for Remote and Rural Healthcare (2008) [19] had recently been produced and its goal was to provide a standard framework for how Scottish rural health services would be delivered into the future. It was the culmination of several years of discussion, some of it based on evidence from the NHS Remote and Rural Areas Resources Initiative which had been established in 2000 to address the problems of recruiting and retaining rural health professionals.

There is a spectrum of approaches to deciding what health services a rural community should have. One way is to provide standard service designs for types of places with certain packages of objective characteristics. In Australia, a rural and remote primary healthcare typology has been established, with authors concluding that “ a critical minimum population base of about … 2,000-3,000 people for remote communities is necessary to support a comprehensive and sustainable range of … services” [3]. How to service communities of less than 2,000 people is not addressed. The Index of Rural Access applies a fine-grained approach to remote and rural healthcare planning [4], including service availability and proximity, population health needs and mobility and claims to be sensitive in describing accessibility deficits. Another algorithmic approach combines need and best practice evidence to design community workforce.

Using chronic disease as an exemplar, it identifies need for condition sub-populations, effective interventions and then calculates the competencies required for the local population [5].

An alternative to standardised models is to customise services to local context within a population health planning philosophy [20], incorporating community participation. Community participation has been described as “…social interactions to influence and localise outcomes” [21]. Variations are community engagement or involvement, which we consider are all about desire to include the views of local people in service planning. Keleher [6] describes population health planning as requiring stakeholder input, being predicated on a social determinants understanding and informed by data about care delivery, illness prevention, health promotion, resourcing and effective design and implementation. Community participation has been described as a ‘social process’ and ‘an ideal’ [22]. It has also been discussed as a state on a continuum between community readiness and community empowerment [23]. Experts [21, 24] have commented on the need to distinguish between clients and citizens in participation, with clients’ participation based on self-interest as consumers, while citizen participation asks people to reach beyond their own concerns, to what is good for the community as a collective. The way we conceptualise community participation is as an intervention that could extend into a philosophy of working, with the key idea being that it endeavours to bring together the voices of those who have an interest in the community’s health – i.e. it’s a space in which local citizens, health and service practitioners and managers can discuss local health issues and how to address them.

Diverse motivations have been ascribed to why international governments promote community participation. The Scottish Government [25] describes need for:

“…a relationship where patients and the public are affirmed as partners rather than recipients of care… where we think of the people of Scotland not just as consumers – with only rights – but as owners – with both rights and responsibilities”.

Here, the Scottish Government is invoking both client/consumer (individual) and citizen/community (for the good of society), perspectives [21, 24]. A review found over 100 methods for public engagement [26], including focus groups, participatory appraisal, Planning for Real, citizen’s juries and future visioning. With imprecision about what it is and many methods offered for doing it, it is unsurprising that health managers wrestle with understanding what to do about community participation. While there is general evidence about beneficial effects, few studies evaluate outcomes affecting health service decision-making [27].

Remote and rural settings have been suggested as prime sites for community participation due to a history of rural community development [28]. The OECD [2] has promoted community participation within the “new rural paradigm”, suggesting its centrality to harnessing rural communities’ internal resources. Rural places have been shown to have high social capital [29] and volunteering [30].

While community participation is an established policy concept, it has been suggested that little power or creative input into design has actually shifted to citizens [31]. Some in rural health think that the public would have difficulty making “realistic” decisions [32], presumably because understanding the health system is complex. Given the lack of studies showing whether or how community participation can affect service design, this paper offers new insights and starts to address a gap in knowledge.

Findings show that using community participation can lead to designing new service models that fit within existing budgets and address local aspirations and healthcare priorities.

Participant communities were similar on dimensions of rurality, health status, aspirations and health priorities. They proposed a consistent set of requirements for remote Scottish primary healthcare. These were: resident practitioner(s), expert emergency triage, monitoring and anticipatory care of vulnerable people, community volunteering for health improvement, and leadership of community (health) volunteering.

Despite community similarities, community participation led to different service designs for three communities and one community did not participate at the final service design stage. These diverse outcomes suggest the influence of local contextual factors. This implies that different communities do want different things so suggesting the same standard model for all, apparently alike, rural communities is likely to be unsatisfactory and could lead to community disaffection. Community participation appears to allow for the customisation of local service models.

The two innovator communities (C and D) included new types of practitioners and community members as volunteers in their service designs, showing that some communities will be quite bold in a community participation process. Community B did not innovate and Community A informally withdrew. While reasons for this behaviour were not explicitly investigated, the literature and previous research suggest a range of explanatory ideas. Previous study has noted that communities are at various stages of receptivity for participation and that some maybe prone to hegemonic power, with prominent local people exerting influence over others [40]. For Community A, there was evidence of community leaders spreading advice not to participate as this would indicate compliance with health service change.

While Community B participated, the outcome replicated their current GP-led model, despite much discussion at their workshops about the value of different types of workers. Actions of A and B could be interpreted as forms of protest or at least non-compliance with, the potential for change.

Communities A and B had ‘rich’ existing service delivery models, each having a resident GP and a nurse despite small populations. A and B are island communities and thus may feel strong insecurity, particularly they might perceive threats to health and community sustainability if they lost their GP. Conversely, C and D may have been more receptive to change due to circumstances. Residents of C were dissatisfied with having no locally resident health professionals and were keen to advocate for improvement in local provision. Community D was about to lose the resident nurse and people were anxious about local service depletion. These circumstances may have made Communities C and D readier to envision innovative services, particularly as their designs highlight local presence as a priority. Community receptivity for change can be understood to arise from different origins. Portes [41] reflects on relationships between local social capital and change, with strong bonding capital (i.e. relationship ties between neighbours that are alike in social group and status, and embedded in the local community) less associated with innovation, compared with the existence of strong local bridging capital (i.e. relationship ties between people in the community and people with access to external resources) which is more associated with innovation [41]. Although we lack evidence about length of time community participants had lived locally, observation in all of the communities indicates that they combined diverse long-term locals and incomers and this was reflected in workshop attendance. This questions the role of social capital as underpinning different responses. Rather, we suspect evidence from a previous study about how remote health services incrementally adapt over time, in relation to their context, might be relevant [42]. Findings of that study indicated that rural health services developed incrementally in relation to inter-linked demand and supply factors of what skills and roles were available locally and how local ‘consumers’ adapted to these. We suspect community members need to be able to envisage and therefore accept, change to aspects of service delivery that are akin to current or known services, as opposed to being asked to envision radical change. That is, Community C wanted moves back to a previous model of a resident community nurse because they remembered this and nurses still lived in the community (though now working peripatetically). Community D saw nurses and paramedics already in their community and thus could envisage a role that combined these two sets of skills. Although we suggest this as a theory, why communities design different models is an issue worthy of future, in-depth exploration.

The composition of the participant group is likely to impact both on choices made, and acceptance of these by the wider community. For all communities, only small proportions of the population attended workshops and outcomes depended on their views. Thus, if the models designed were to become more than hypothetical – and moved to implementation, fellow citizens might protest that they had not participated in decision-making. This issue was raised by a health manager who questioned the status of the designed models and community reactions if the health service moved to implementation? “When and how is a community decision made?”, she asked. Inclusion is discussed in community participation literature, with questioning about the credibility of processes involving small numbers. Others suggest that all are offered the opportunity to participate and that not participating is also a choice [43]. Taylor et al. [21] note that community participation implies collective involvement and Alford [24] that true citizen involvement requires representative participants. In practical terms, methods such as Renn and colleague’s citizen panels [44], seem to come closest to an ideal inclusion method. They applied random selection to lists of community residents to invite citizens to participate in community decision-making forums. They say this is a way to ensure marginalised people are invited to participate, but there is still the problem of whether such people feel sufficiently comfortable to turn up in a public setting.

In redesigning rural services, standard models devised for communities with shared characteristics avoid these complications of involving communities that we have described. Their basis is to produce equity for all – “…not ad hoc responses…” [3]. If community participation results in different service designs for apparently similar communities, this could result in some communities being overworked compared with others. An example would be if local people provided lay first response in some communities, but not in others. Hanlon and Halseth [45] have commented on the exhaustion in rural Canadian communities due to service withdrawal and community members having to co-produce services.

But while the acceptability of diversity, with potentially different impacts for local residents, presents a philosophical debate, in fact there are uneven models already present in Scottish remote healthcare, and likely internationally. Taking the example of first response, some communities have already established volunteer first responder schemes, while others have not [46]. Community service models vary as do the skills and experience of individual health professionals [47]. Some recent initiatives in rural health endorse the perspective that it is realistic to draw on existing assets, local needs and rural propensity for adaptation to circumstances, when redesigning local services [48]. As co-production, resilience and resourcefulness become increasingly promoted in government policy, it seems communities will have to increasingly become self-reliant co-producers, with the implications for diverse service accessibility and potentially for outcomes, that brings [49]. The challenges of community participation may make the application of standard models seem attractive when redesigning services; however, there is evidence that rural communities protest if externally conceived plans, into which they have had no input, are inflicted upon them [50, 51].

The overall RSF study was about producing a community participation process for remote places that could be used by health authority managers (i.e. was straightforward to understand, produced plans for service redesign, and was relatively cheap). It thus collected limited, mainly observational and informal, data about participants, their motivations and experiences. As engaging people in a comfortable and ‘normal’ process was important, most data were not recorded verbatim, meaning that the words of participants cannot be shared here. Since the process was collaborative, it was necessary to ‘go with the community’ , rather than being overly directive.

This produced messy evidence, but Tritter and McCallum [52] note that “users must have agency and the ability to shape the methods used for their involvement”. We do not have formal data about why people participated or not. This would have been useful regarding the apparent withdrawal of Community A at the design stage. As the models derived by communities were hypothetical, we do not know what would happen if the health authority chose to work with the communities to implement the designs. However, in enacting an actual process of community participation to produce designs for future services, a number of issues about rural community participation have been raised.

This paper illustrates that community participation can be used to design rural primary healthcare services, but outcomes may vary from innovative models to passive protest, depending on community receptiveness. That communities produced different responses to apparently similar circumstances and priorities, suggests that aspects of local context affect the choices communities make and therefore that engaging community members should add value when designing acceptable local services. Population health planning incorporates community participation while top-down standard models appear to neglect a community perspective, even though they are underpinned by a desire for fairness. There maybe a role for standard models as part of community participation discussions as they are informed by evidence and provide abstractions that community members can use as a basis for discussing adaptations or additions that they deem important given local requirements.

Inviting communities to participate in decision-making produces messy and unpredictable outcomes and this is insufficiently acknowledged by policymakers. More needs to be written on the productive and/or messy aspects of community participation in system change.