Sample and recruitment

Interviews were undertaken with older people and focus groups with health professionals. Ethical approval was obtained from the University of East Anglia Faculty of Medicine and Health Sciences Research Ethics Committee (reference 20142015–43) before the research began and for each protocol amendment. Written informed consent was obtained prior to any data collection. Findings are reported in accordance with COREQ guidelines [15].

Participants who lived in a rural area, were over 65 years old and relied on financial support, such as means-tested pension credit, housing benefit or council tax support, were recruited for semi-structured interviews. We specifically did not use the term “benefits” in recruitment material because of socio-cultural connotations. Rurality was defined, using the Office for National Statistics definition, as settlements with less than 10,000 residents [16]. Those in residential care or unable to consent were excluded. We purposively decided to recruit participants through the community rather than primary care organisations to avoid any perception that the research was directly associated with their GP surgery. Our sampling framework aimed at avoiding oversampling of women (criterion: include at least three men), younger participants (criterion: include at least two participants over 80 years old) and people from the same practice (criterion: include participants from four different GP surgeries). Recruitment and data collection took place between August 2015 and January 2016.

The recruitment strategy evolved throughout the research because of low participation. In the first round of recruitment, three electoral wards in the English rural county of Norfolk, given its large population size and geographic area, were selected based on rurality, household deprivation and pension credit claimants. Posters and/or invitations were displayed in post offices, churches, shops, social clubs and parish newsletters, and this led to recruitment of one participant. For the second round of recruitment, we introduced a £20 shopping voucher, increased the number of electoral wards, amended the invitations and hand delivered approximately 300 invitations; leading to four additional participants. For the third round, the lead researcher (JAF) was interviewed on local radio and discussed the research at local Age UK and older people’s forums; leading to an additional two participants. In the fourth round, over four weeks across six pharmacies, about 300 invitations were placed in the pharmacy bags of patients over 65 years old and having their medications delivered; this generated a further eight participants at which data saturation was reached.

For four focus groups, rural health professionals were recruited through the NIHR Clinical Research Network (CRN). The CRN invited general practitioners (GPs) and practice mangers from research-active practices in Norfolk and the research lead of Norfolk Community Health Care to identify community health professionals. Health professionals, or their employing organisations, were reimbursed for their time and travel. None of the participants were known to the research team prior to participation, except for one GP who was employed part-time at the University of East Anglia. No participants who contacted the research team subsequently refused to participate or withdrew.

Data collection

Participants chose to be interviewed at home (n = 12) or the University of East Anglia (n = 3) while the focus groups were held at the university. All focus groups and interviews followed individual topic guides (see supplementary online material). Interviews lasted about one hour, and focus groups two. Interviews and focus groups were audio-recorded and transcribed verbatim by an experienced transcriber at the University of East Anglia. Field notes were audio-recorded immediately after the interviews and focus groups and subsequently transcribed. No interviews were repeated.

The interviews and focus groups were conducted by a single male researcher trained in qualitative methods (JAF, qualifications MBChB MSc) whilst a second qualitative researcher (TP) observed two of the focus groups. JAF is a public health doctor, undertaking a PhD, but introduced himself as a researcher looking at access to primary care, not revealing his clinical background to participants unless specifically asked, which happened in one interview.

Data analysis

We analysed the data using thematic analysis, based on the steps outlined by Braun and Clarke [17]. Briefly, this started with familiarisation of the data through repeatedly listening to the interviews and focus groups and reading the transcripts (JAF and RT). Codes were generated by identifying and labelling key sections of text which contained data on access to primary care by two researchers independently (JAF and RT). Similar codes were then independently grouped to form initial candidate themes by the same two researchers (JAF and RT). Themes were refined iteratively through repeated discussions of the data between the two coding researchers, the wider research team, colleagues within the department, and two patient representatives. QRS NVivo was used to code the data [18]. JAF has previously undertaken a realist review in this area [10] and had substantial knowledge of the topic; therefore a constant comparison approach [19] was used to avoid confirmation bias.

To ensure rigour and trustworthiness, we implemented the following measures: 1) coding and generation of initial themes was undertaken independently by two members of the research team (JAF and RT), 2) transcriptions were checked for accuracy by the researcher undertaking the interviews (JAF) and 3) themes were interrogated by the research team and patient representatives. We did not undertake member checking, so as to leave the data as it was captured during the interview.

Older people’s perspective

The health professional’s perspective

Summary of principal findings

We found that rural, socio-economically disadvantaged older people’s experience of access to primary care can be understood within the context of a social contract–an individual is careful not to bother the doctor in return for additional goodwill when they became unwell. However when participants accessed primary care, most found it difficult due to engaged telephone lines, availability of appointments, interactions with receptionists; breaching their perceived social contract. This left some feeling unwelcome, worthless or marginalised, especially those with high expectations of this social contract or limited resources, skills and/or desire to adapt to service changes. From a health professionals’ perspective, barriers included rising demands and expectations, necessitating service development of traditional models of care with less resource. Service developments included fewer home visits, more telephone consultations, triaging calls and modifying the appointment system.

Strengths and limitations

One strength of our design was collecting data from both service users and health professionals allowing different perspectives of access to primary care. However, we experienced challenges in recruiting this group of hard to reach older people. Recruitment was particularly challenging because we used a community recruitment strategy, compared to recruiting via GP surgeries. We did this so that participants would be less likely to perceive that the research was directly associated with their own GP surgery. However it made recruitment more difficult and to overcome this challenge we modified our recruitment strategy on three occasions. Our final method, which resulted in most recruits, was to use delivered pharmacy bags to recruit vulnerable older people. This recruitment method that may be applicable to a range of other studies. However, participants were recruited using this method may have been more likely to be engaged with primary care and housebound.

We included both men and women and anyone over 65 years old, but did not have sufficient data to explore the difference between older men and women [20, 21]. The age of participants ranged from 67 to 87 years old. Those belonging to the ‘baby boomer’ generation (52 to 71 years old) and ‘silent generation’ (72 to 93 years old) may have different life experiences and attitudes, but similarly we did not have the data to explore this.

Our main inclusion criteria to identify socio-economically disadvantaged older people was relying on financial support, acknowledging that financial deprivation is only one aspect of disadvantage. Despite receiving financial support, at the interview stage two participants did not appear to fit this group, and their data was interpreted with caution.

Furthermore we only undertook four focus groups of health professionals and while similar themes began to emerge in the third and fourth focus groups we were may not have reached full saturation.

Whilst our data came from one area of the UK, we do not have any reason to think that the concept of a social contract is generalizable to other parts of the UK, although the terms of the reciprocal relationship may be different. Other studies including older people identified similar themes of not wanting to bother the doctor [22] and there is no suggestion that our findings are unique only to our participants. Although areas with low population and GP workforce mobility, such as Norfolk, may facilitate a social contract more than inner city areas. Further research is needed to understand if a social contract exists in other health care systems, such as the US. Previous research from the US has found similar barriers, such as transport, lack of resources, limited health care supply and lack of understanding by doctors [23–27]. However cost of medical treatment was more important in the US compared to the UK because of different health care systems. The health professionals’ perspective is likely to be generalisable across high income countries because most health care systems are trying to address the challenges of an ageing population, rising demands, increasing costs and financial constraints.

Comparison with other studies

Our findings support the previously described concepts relating to access to services by vulnerable groups of candidacy (“the ways in which patient's eligibility for medical attention and intervention is jointly negotiated between individuals and health services” [28]) and permeability (“the ease with which people can use services” [28]). Health professionals and older people both alluded to the permeability of services, especially concerning availability of appointments, telephone access and receptionists. We found a two-way, dynamic relationship, supported by previous literature [29], between older people’s service use affecting practice procedures, and practices informing older people how to best use the service.

The concept of a social contract has been applied to health care previously, but at a macro level [30, 31]. Cruess and Cruess argued that members of society expect the medical profession to provide the services of a healer in return for autonomy, trust, status, self-regulation and a functioning health care system [30]. Our findings support the existence of a social contract, but enacted at a relational level and for a specific population group.

Bentley undertook a qualitative study using mini-ethnography to explore access to health care for older people living in an English village [32]. The study found that culture, legitimacy and, as suggested in previous research [21, 33], a retained hierarchical medical model of health care are barriers to access for older people. The authors found that older people did not identify themselves as consumers. We did not find data to support a consumer model either, and our findings support a hierarchical model in which older people do not bother the doctor for minor ailments. Building on this, we found that in return for this hierarchical system older people expect the GP surgery to make special provision when they are ill.

Goodridge and colleagues looked at the experiences of patients with chronic respiratory illness in rural areas of Canada using semi-structured interviews [34]. The authors found several important facilitators to health care, such as distance, relationship with their family doctor, supportive local community and health knowledge. We found that the GP relationship and local social networks were important, but older people in our study highlight the importance of suitable transport options rather than the physical distance to the surgery.

Meaning of findings and policy implications

In England, Clinical Commissioning Groups have a legal obligation to address health inequalities [35]. Access to services is a key policy area to address inequalities [36]. Policy decisions in relation to access to primary care are challenging because most health services operate within a fixed budget, and improving access may lead to increased patient expectations or unnecessary use. One-size-fits-all policies may increase access for the whole population, leading to increased pressure on the system and counterproductively worse access for marginalised groups—a targeted approach is hence needed [12, 13].

Our findings suggest that health professionals support a targeted approach to improving access when vulnerable older people use primary care services, but not necessarily a proactive approach which may, for example, involve outreach activities in the community. We identified various personal and structural resources needed to access primary care, with the most important being transport, social networks and personal skills. An individual’s access to structural and personal resources were linked with their ability to manage service changes. For example, some older people are unfamiliar with changing staff roles, such as the changing role of a receptionist to include triaging, as well as booking appointments.

A tension exists in the health professional’s views–on one side ageing is being medicalised, but conversely patients do not consult because they attribute their health problems to ageing. It is likely that the medicalisation of ageing, in part, relates to conditions with incentivised or nationally recommended policies, such as raised cholesterol or blood pressure. It may that the some patients are being contacted to meet incentivised prevention targets, which in turn leads to fewer appointments to fulfil the social contract.

To address the significant financial and workforce challenges within the NHS, there is a drive to deliver primary care within ever larger practices, such as networks, federations or super-partnerships [37]. The effect of these changes are unknown. Our findings suggest a tension between the expectations within a social contract and experiences of older people which were compounded by recent service changes. Reconfiguring services to deliver primary care at scale may lead to socio-economically disadvantaged older people in rural areas feeling more disconnected from primary care, because, if a social contract is present, the older person will continue to carefully use the service, but are unlikely to be shown the goodwill they expect in return.

The interface between patient and practice is crucial and many practices have been striving to improve this for a long time. The drive to deliver primary care at scale is likely to make this interface, in the absence of careful planning, less personal. Representative patient groups could help practices to improve the experience of older people. Simple measures may make a significant difference, such as helping practice staff understand the perspective of different patient groups to facilitate a more sensitive approach to the needs of vulnerable groups. This may mean, for some patients, receptionists not asking about a patient’s health problem, or having a lower threshold for home visits. A major obstacle is time and resources within primary care; without giving reception staff more time to answer calls and appointments to allocate, older people are likely to continue to have negative experiences.

More community support is needed in rural areas to help with transport, such as community car schemes. Closer collaboration between commissioners and local authorities may help to improve public transport to health care. Furthermore it may help to protect some appointments which coincide with public transport timetables or when there is limited taxi availability, such as during school drop off and pick up.

Conclusion

We found that rural, socio-economically disadvantaged older people’s experience of access to primary care can be understood within the context of a social contract–an individual is careful not to bother the doctor in return for additional goodwill when they became unwell. However the perceived social contract was often breached when participants accessed primary care due to engaged telephone lines, availability of appointments, interactions with receptionists. This left some feeling unwelcome, worthless or marginalised, especially those with high expectations of this social contract or limited resources, skills and/or desire to adapt to service changes. From a health professional perspective, barriers included rising demands and expectations leading to the need to reconfigure services with less resource. As primary care is re-organised to reduce costs, commissioners and practitioners must not lose sight of the perceived social contract and models of care that form the basis of how many older people interact with the service.