Do we all agree what “good health care” looks like? Views from those who are “seldom heard” in health research, policy and service improvement

Abstract Context The aim of this study was to ask whether there are shared ideas about what good health care looks like that apply across different populations and conditions. Do priorities among “seldom heard” groups differ from mainstream views and, if so, how might we understand these differences? Design Focus groups were recruited with the help of our study patient representatives. Participants discussed and prioritized a set of eight “core components” of good care. We recorded and transcribed the data for thematic analysis. Setting and participants We recruited people who are seldom heard in health and policy research for separate focus group discussions (one each with illegal drug users, Irish Travellers, migrant workers, young men and learning disabled people). We also ran a reference group of educated, older adults and an online group with people with long‐term conditions. Results There were few differences in what participants thought was important in health care but considerable differences in their expectations that they might personally receive good care. Differences related to participants' previous experiences. The drug users group reported particularly poor experiences and low expectations of good care. Discussion Differences in what is regarded as an entitlement or privilege in health care underline the persistence of structural and relational differences in how services are experienced. While we can be reassured that core aspects of care are similarly prioritized across different patient groups, including those who are seldom heard, a more intractable challenge remains: how to provide equitable health care for marginalized groups in an unequal society.

whether these definitions and standards encompass what matters to "seldom heard" groups, that is sections of the population who are typically left out of research. 5 Are priorities different when care is viewed from less familiar, typically less privileged, perspectives? Given the enduring health inequalities in England (despite a National Health Service free at the point of delivery 6  The turn to public involvement is increasingly familiar and justified through accountability and as well potential to improve decision making by involving wider perspectives 10 (although the evidence for this remains under developed 11 ). The approach used by NICE, inviting contribution and observation from registered stakeholder organizations, conforms to expectations about public and patient engagement in policy, but concerns are regularly raised about whether those members of the public who are involved are the "right ones" and whether they are (or should be) representative. 12,13 In health research, learning disabled people, illegal drug users, homeless or Traveller populations are often left out of mainstream studies; they tend to be included when they are the sole focus of a study, or when the focus of the study is on deficit. 14 These omissions leave our understanding of patients' experiences incomplete as "seldom heard" groups may have different views about what matters in health care. Indeed, Lemert 15 argues that those living at the margins of society can, by their very position, view mainstream life with a clearer lens than those at the centre. Enduring health inequalities within the UK, whether yet to be explained by behavioural, material or psychosocial orientations, 16 underline the importance of understanding the health experiences of seldom heard groups. 17 Therefore, the aim of this study was to ask whether there are core components of good care that apply across different populations and conditions. Do priorities among "seldom heard" groups differ from mainstream views and, if so, how might we understand these differences?

| METHODS
Between January and April 2013, we ran six face-to-face and one online focus group to examine whether how and why ideas about "good care" might differ when discussed with participants from social groups who rarely appear in mainstream research. We drew on published health-care quality frameworks [18][19][20] and an earlier qualitative secondary analysis of narrative interviews 21,22 to identify a set of "core components" of good health care to use in the focus group discussions.
We deliberately focused on good rather than basic (such as, my doctor is qualified to treat me) in selecting the components.

| Sampling
Participants were recruited through our project Patient and Public Involvement (PPI) contacts, personal relationships in the voluntary sector and social media and held in four different parts of England: Liverpool, West Midlands, Oxford and South London. We wanted four different locations for the groups and these areas were selected pragmatically because they were where the different groups were based.
Separate focus group discussions were arranged with migrant workers (n6), Irish Travellers (n8), young men (n10), illegal drug users (n13), people with long-term conditions (n11) and learning disabled people (n5) as well as a reference group of older people (n9) who meet for educational and cultural activities through the University of the Third Age (U3A). Most of the participants were from pre-existing groups and were familiar with each other.

| Informed consent
We took particular care to adapt our consent procedures for the variety of language, literacy and comprehension skills between groups. explaining what was involved in the study and inviting people to participate. Several of the Irish Traveller group had poor reading and writing skills and we went prepared to audio record consent, and with several local helpers, who were well known to the participants, available to support literacy throughout the session.

| FOCUS GROUP METHODS AND PROCEDURE
We used focus groups to explore the relevance of the candidate core components with people whose views may be under-represented in the literature and in the archive we used for secondary analysis.
Focus groups are an established method for using with "seldom heard" groups, 23 allowing researchers to pay attention to those who have "little or no societal voice". 24 Being among similar others, in a supportive and reassuring environment, can encourage participants to talk openly. Kitzinger 25 argues that focus group work "ensures that priority is given to the respondents' hierarchy of importance, their language and concepts, their frameworks for understanding the world." A skilled group facilitator can shepherd the group through discussions of personal or private issues and possible contradictions between accounts. 26 Focus groups can also bring difficulties less common in an interview. 27 The cognitive and emotional demands of reflecting on other people's arguments and engaging opposing views can be particularly challenging for people with learning disabilities. 28 A group in which some members need frequent explanations or translations to fully participate requires particularly alert and sensitive facilitation. Flexibility in the structure and comportment of groups is particularly important when working with participants who may not usually take part in research.

| Focus group structure
Each group was held in a venue convenient to participants, agreed with our local contact. The groups lasted between 90-120 minutes facilitated by at least two members from the research team and, in two cases, the local PPI member in attendance. Introductions and a warm-up were followed with discussion about what makes good (and not so good) health care. Participants were invited to consider the eight pre-prepared candidate components of "good care" and order these and any additional priorities. Finally, the groups discussed why they had ordered the priorities in the way they had.

| Online discussion forum
We also held an online discussion forum 29 with 11 patients with longterm conditions. The conditions (some comorbid) were as follows: Chiari malformation, postural orthostatic tachycardia, Ehlers-Danlos syndrome, systemic onset juvenile idiopathic arthritis, cerebral palsy, gender dysphoria, chronic pain, endometriosis. MS, epilepsy, clinical depression, asthma, polycystic ovaries, sickle cell, chronic heart failure, atrial fibrillation, osteoarthritis, cataracts and spina bifida.
A webspace "Goodhealthcare" was set up using Ning, a free online platform for creating social networks. A forum was created for each of the core components. The details of the forum were shared on social media, and eleven people were recruited. We invited participants to read the statements, watch a short illustrative clip from the HERG data archive and then contribute to a discussion about the importance of the statement. Participants were encouraged to return to the webspace to respond to comments left by others. The method allowed us to hear the views of those we could not reach through the focus groups, either because of the severity of a long-term condition or that of someone they cared for. The facilitator, SR, responded to comments and invited further response which generated richer detail. The site remained open for comments for a six-week period and then the responses were collated and analysed alongside the other six groups.
All participants were given £30 shop vouchers for their time as well as travel expenses if applicable. Compensation is routinely offered in focus groups. 30 It helps to increase participation rates and there is some evidence that it does not affect responses. 31

| Analysis
All of the discussions were audiorecorded; the research assistant and facilitators also made notes. The research team met for an analysis workshop in which we listened to the recordings to capture the richness of the interaction between participants. Recordings were also transcribed to enable us to further analyse the content of what was said. An interpretative, thematic analysis was conducted while listening to the recordings and further developed using the transcripts. Our analysis sought to examine and conceptualize the limitations and reach of components of good care and how they may vary between groups. We charted the ordering of the "core components" from the groups which helped to highlight differences and similarities. It also enabled us to explore how participants, referring to their own experiences, regarded the components as good, basic or aspirational.

| FINDINGS
In this section, we compare perceptions and priorities about the constituents of good care between the focus groups. We also consider why priorities and expectations differ, drawing examples from the sorting exercise and transcribed discussions. Table 1 shows the rankings of the eight components from six of the focus groups. There are evident similarities, especially in the most highly rated components. Four of the six groups selected "taking time to answer my questions and explain things well" as their top priority, while the "drug users'" group chose "guiding me through difficult conversations" and the learning disability group chose "letting me see the same health professional." Before further discussion of these data, we should note that some of the participants (particularly in the older adults, Irish Travellers and young men's groups) thought that all of these components could be regarded as essential rather than "good" care. As one of the Irish Travellers put it, when considering priorities But they're all things that should be there anyway ….

| Differences and similarities in priorities between groups
While most participants agreed that staff "having a friendly and caring attitude" was important, it was apparent from discussions that this included different elements such as displays of empathy (warmth, eye contact, smiling, remembering personal details), behaving with respect, willingness to listen, being careful to check understanding and taking concerns seriously. The migrant workers explained that language difficulties could be overcome if the health professional had a genuine desire to understand. The older adults warned that offence could be caused if staff equated friendliness with overfamiliarity, for example if they used first names without permission.
"Guiding me through difficult conversations" was rated as particularly important by the migrant workers, drug users and the young men, while signposting to further support was rated highly by the older adults and Irish Travellers. Learning disabled people, migrant workers and some of the Irish Travellers said that receiving leaflets or website addresses for additional information would be of little use to them. However, personal advice and information was appreciated.
The young men and people with long-term conditions welcomed additional information from doctors as long as this was not seen as a way to avoid engagement during the consultation. One young man recounted a consultation about a skin problem where the GP hardly spoke and just printed a page from a website for him. People with long-term conditions said it was important for health professionals to consider broader needs for information and support and wanted details of support and advocacy organizations. The drug users group had different views on this topic depending on whether they were already integrated into support networks, or had not (yet) accessed support from social care or voluntary agencies.
Efficient sharing of information across services was considered important but did not attract any "top 3" ratings. Several participants said T A B L E 1 Focus group exercise, ordering the components (from most to least importance)

| Other examples of good care
At the start of the focus groups, we asked participants to tell us about what they saw as indicative of good care. This was illuminating. The drug users, who were conscious that health professionals sometimes reacted with hostility and suspicion when they saw their history of drug use, stressed the importance of health professionals listening to their views and attending to their feedback about treatment effects.
The Irish Travellers said they would like to see colour-coded medicines for people who do not read, a coordinator or broker to act as a go-between with services, and access to a familiar doctor for out of hours care. The older people group emphasized prompt referral to specialists, use of consistent evidence-based approaches to treatment, and recognition that relatives and friends might also need information.
Some of the young men who had been in hospitals with friends or family commented on the bright lights, hard reflective surfaces and alarmingly loud noises; they suggested that hospitals would be healthier places if they were quieter and less threatening. Migrant workers said they valued staff who behaved professionally and respected confidentiality-one told a story about a GP's receptionist in her home country who had disclosed a patient's use of antidepressants to a family member. The group of learning disabled people raised a concern that "efficient sharing of information across services" might involve professionals gossiping inappropriately about their health and circumstances.
These concerns were not raised by any of the other groups, perhaps because they regard confidentiality as a fundamental of care.

| The good, the basic and the unlikely
There was also variation between groups about what could and should be expected from health care. An aspect of care that was described as basic or routine in one group might be seen as good, desirable or unattainable in another. To illustrate this point, we consider two examples: "involvement in decisions about care" and professionals "having a friendly and caring attitude."

| Involvement in decisions about care
The older adults group described this as a basic part of health care; their expectations were broadly in line with a shared decision-making model 32

| A friendly and caring attitude
Participants gave many examples demonstrating why a friendly attitude, from reception staff as well as clinicians, really matters in health care if the patient is to feel able to discuss sensitive health issues. A friendly caring and warm attitude was commonly presented as an inte-

Box 1
I remember an incident of being on the maternity suite at the (local hospital) with someone who was giving birth and they, they were on a Subutex script, so they hadn't used illicit drugs in I think maybe ten years and you know I thought we'd better tell the midwife that this person's on a Subutex script. Now they had no idea what Subutex were, they went away and obviously they went away to find out and when they found out their attitude when they came back was appalling. I mean this person hadn't touched drugs in…illicit drugs, in ten years. They put someone at the end of the bed to watch, watch her from then on and it was stupid, it was very stressful, it was upsetting, it was awful. I complained and, and we, we got some kind of I don't really know, an excuse of an apology but it was just the lack of knowledge and ignorance with the, with the midwife, the maternity suite staff. And they, they should be aware of these, these, these sort of things, I thought it was, it was disgusting.
There were some differences between groups in the ordering of priorities which could often be understood through participants' prior Population characteristics, including familiarity with health and care services, which influence experiences may be hard to discern through conventional socio-demographic data.
Codesign approaches, in which patients and members of the public are invited to help staff identify problems and plan service improvements, have gained popularity. 38 Our study suggests that such initiatives may sometimes benefit from involving people with relatively little experience of health services, as well as seasoned patients. Members of the public who bring a fresh eye to services can challenge assumptions and bring new insights. The young men group, with relatively little experience of hospital care, raised issues that may be relatively invisible to those who have been working in, or using, a service for many years. For example, noise may be taken for granted in hospitals, but is not inevitable and its impact could be reduced. 39 The confidentiality of the consultation may be taken for granted by those who have never had reason to expect it would be otherwise. However, there would be losses if only people who can turn a fresh eye to health service problems were involved-it is sometimes only when people have considerable experience of health care that they start to realize how the system works or become aware of variation in skills and service provision.
This work contributes to a field where there has been little evidence about whether (how and why) patients ideas about what constitute good care may differ for those in "seldom heard" groups. Our study was designed to look at the components of "good care" rather than "basic care"-participants also expected their health care to be safe and that health professionals would be well trained and aware of best treatments.
Exploring the relevance of the core components of good care in the focus groups found there were some differences in what is seen as important about health care in these "seldom heard" groups, and certainly no suggestion of a completely different value system, yet experiences and expectations of these prioritized aspects of care were very different underlining the persistence of structural and relational differences in how services are experienced. Thus, while we can be reassured that the reach of existing outcome measures and experience survey instruments may be robust across different groups of patients, including those who are seldom heard, a more intractable challenge remains: how to provide equitable health care in an unequal society.