Findings

This is a full report of findings reported in Toye and colleagues42 (Reproduced from Toye et al.42 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/) and Toye and colleagues43 (Reproduced from Toye et al.43 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/).

Search results

The results of the systematic search are shown in Figure 2. We screened 954 potentially relevant studies and excluded 770 after screening the abstracts. We retrieved 184 full-text articles and excluded 101 for the following reasons:

• not about chronic pain12,44–87
• did not include HCP voice88–100
• not qualitative, or limited qualitative data101–116
• out of scope117–143

FIGURE 2 Systematic search results. (Figure not reproduced here; see the original article.)

Of the 83 potential studies, we unanimously excluded six on the grounds of methodological report.144–149 We included 77 published studies reporting the experiences of > 1551 HCPs.11,150–225 Table 3 provides the author, year of publication, professional/topic group, geographical context, number of participants, data collection and analytical methods for each study. HCPs included a diversity of doctors, nurses and allied health professionals in various contexts and geographical locations. Not all of the studies reported the number of participants from specific professional groups, which meant that it was not possible to give the exact sample number from each profession. The majority of studies were from the USA, the UK, Canada and Sweden. The sample size from the studies ranged from 6 to 103 (average 22). One focus group study165 and three ethnographic studies156,200,223 did not report their sample size.

Quality assessment and inclusion

Table 4 provides the outcome of appraisal: the CASP score from each reviewer [from 10 (low) to 30 (high)]; the difference in CASP score between reviewers; the seven themes developed from a previous meta-ethnography;8 the global quality score (key, satisfactory, uncertain, irrelevant or fatal flaw); and reviewers’ assessment of potential value based on CASP question 10 (‘how valuable is the research?’). The difference in CASP score between two reviewers ranged from –4 to 2 (a possible score of 10–30). They did not agree about the inclusion of four studies,172,178,196,201 which were subsequently included by a third reviewer. Two reviewers agreed that 89% of primary authors had reported their study aim and 84% had described their sample. They agreed that 89% of the authors had not reported their perspective and its potential influence on findings and 69% had not reported methods for challenging their interpretation. Reviewers also agreed that only 65% of authors had provided clear examples to demonstrate that their findings were drawn from the data and only 55% of authors had clearly described all their findings. Twenty-eight studies11,150,151,153,154,156,157,159,165–167,174,176,179,187,190,195,204,207–209,211–214,221–223 were reported as having ‘changed the thinking’ of at least one reviewer. However, reviewers’ appraisal comments (see Appendix 1) suggested that, even if they did not change their thinking, the primary studies encouraged reviewers to think. Two reviewers unanimously appraised five studies150,151,214,222,223 as ‘key papers’ and 72 as ‘satisfactory’ (see Table 3). At least one reviewer appraised 26 studies11,150,151,154,156,157,160,170,171,176,187,190,191,195,198,204,205,210,211,214,219–223,225 as potentially making a ‘valuable’ contribution to the analysis. They agreed on 21 of 26 of the valuable studies. To allow readers to evaluate the transferability of findings, Table 5 shows the reviewers’ assessment of relevance (direct, indirect, partial or uncertain) and aim of each study. We rated 60 studies11,150,152–164,168–171,173–178,180,181,183–188,190–193,195,196,198,200–211,214,215,217,218,221,222,224,225 as directly relevant, seven151,165,166,197,219,220,223 as indirectly relevant, nine172,179,182,189,194,199,212,213,216 as partially relevant and one167 as uncertain.

Coding and conceptual categories

We coded batches of studies in the following order according to topic/professional grouping:

1. Ten studies reported the experiences of primary care physicians/GPs.11,158,164,192,195,196,204,205,209,222
2. Four studies explored the experiences of a mixed group of HCPs.172,194,203,220
3. Three studies explored the experiences of physiotherapists.170,201,214
4. Three studies explored the experiences of physiotherapists specialising in chronic pain management.151,154,210
5. Five studies explored the experiences of a mixed group of HCPs providing health care to people with fibromyalgia.152,161,181,183,206
6. Eleven studies explored the experiences of a mixed group of HCPs working in specialist chronic pain services.150,156,163,186,200,202,215,216,219,221,225
7. Five studies explored the experiences of a mixed group of HCPs working in pain management related to employment.167,168,179,182,224
8. Twelve studies explored the experiences of a mixed group of HCPs prescribing opioids to patients with chronic pain.153,155,157,174,175,177,190,191,198,211,217,218
9. Six studies explored the experiences of a mixed group of HCPs utilising guidelines for chronic pain.169,173,180,207,212,223
10. Three studies explored the experiences of a mixed group of HCPs working with older adults.160,162,208
11. Thirteen studies explored the experiences of a mixed group of HCPs working with older adults in long-term care facilities.165,166,171,176,178,184,185,187–189,193,197,199
12. Two studies explored nurses’ experiences of providing health care to people with chronic non-malignant pain.159,213

Appendix 2 provides a coding report of included and excluded concepts. Two reviewers identified 492 potential concepts in 77 primary studies (Figure 3). They excluded 115 potential concepts for the following reasons: inadequate conceptualisation (31 concepts in 17 studies11,150,155,160,162,169,173,178,181,188,192,196,197,206,208,219); did not explore HCP experience (18 concepts in seven studies168,174,198,202,203,215,223); and explored a topic with limited transferability beyond a specific context (five concepts in five studies166,167,197,208,223). We agreed that for some topics there were insufficient concepts to allow us to develop robust conceptual categories: return to work (seven concepts in two studies167,182), ethnicity (11 concepts in five studies171,204,205,215,216), gender (seven concepts in one study206) and older people (21 concepts in nine studies159,160,162,165,176,187,194,199,208). Fifteen potential concepts did not fit our developing conceptual analysis (Table 6).

FIGURE 3 Concepts that were included and excluded. (Figure not reproduced here; see the original article.)

All reviewers organised the remaining 377 concepts into 42 conceptual categories. Table 7 gives an example of one of the conceptual categories (‘is the pain real?’) and its included concepts. Our description of the concepts that formed the raw data of analysis were a combination of the primary author’s description of the second-order construct (in which they briefly and clearly described the construct), and our interpretation of the construct (if the original was unclear or lengthy). In some cases we found that there was a section of narrative exemplar provided by the original authors that adequately described the essence of the concept.

Once the reviewers had agreed on a description of each conceptual category, Fran Toye wrote a statement of this finding in the first person.

For example:

This conceptual category described how endless paperwork eats into HCPs’ limited patient time

became:

This endless paperwork eats into my limited patient time.

We have found that writing concepts in the first person is a powerful way for reviewers and their advisory group members to fully engage in the meaning and sentiment of each concept. It also facilitates the use of accessible language for a diverse audience in both analysis and dissemination.

Table 8 provides a description of each of the 42 conceptual categories written in the first person. The reviewers worked with a research fellow and a project advisory group that included patients and HCP members to further abstract these 42 conceptual categories, printed on postcards, into six overarching themes that underpin HCPs’ experience of providing health care to people with chronic non-malignant pain (Figure 4). Table 9 provides a list of the conceptual categories underpinning each of the six final themes. Table 10 shows the number of studies and concepts for each theme organised by topic/professional group. It also indicates the global appraisal score given for individual studies supporting the theme.

FIGURE 4 Illustration of conceptual analysis with advisory group. (Figure not reproduced here; see the original article.)

Assessment of confidence in findings (GRADE-CERQual)

Indicators of confidence in each review finding are shown in Table 11 [including the number of studies rated as key/valuable or satisfactory (methodological limitations); the number of concepts (adequacy); the number of studies out of 77 (coherence); an assessment of study relevance; and our overall assessment of confidence]. We rated our confidence in the review finding as high when it was supported by more than half of the studies. However, there is currently no agreed way of assessing confidence.

Conceptual themes

We illustrate the six themes with narrative exemplars. The film output ‘Struggling to support people to live a valued life with chronic pain’ can be watched on YouTube (see Report Supplementary Material 1; URL: www.journalslibrary.nihr.ac.uk/programmes/hsdr/1419807/#/documentation).

A sceptical cultural lens and siren song of diagnosis

This theme describes a cultural lens that provides a sceptical view of chronic pain. With this view, clinical work focuses on determining whether pain is ‘something’ (biomedical) or ‘nothing’. Although some acknowledged the dangers of ‘judging a book by its cover’, HCPs viewed the world through this sceptical lens and at times found themselves making non-clinical judgements about whether a patient’s pain was real or imagined. HCPs described how they could experience a dissonance between what they saw and what the patient said:

Sometimes I could have a patient sitting there and saying that they are hurting, 10 out of 10, and they are sitting like you and I.

Some people say ‘This is the worst pain I’ve had in my whole life’ without any real sort of physical signs of pain so it’s really tough; we have a complex job in assessing that.

The patients tell me with a smile on their lips that they are suffering immensely from all kinds of bodily disorders. How on earth can they look so terribly healthy?

However, HCPs recognised the dangers of their scepticism and acknowledged a need to not make hasty judgements. Although they knew that patients did not always look like they were in pain, HCPs remained ‘on guard’ against being exploited by fraudulent claims of pain:

You’d have to be on guard this man isn’t laying it on.

It is not clear to me why he is the way he is . . . this catastrophic pain and what he is telling himself about it . . . but there is always a little bit . . . of concern; am I being manipulated, is this really real?

Such people . . . ones whose wishes you cannot fathom – provoke anger and frustration because at some point, you don’t always know how to verify their complaints. You feel somewhat exploited. It is a very unpleasant feeling.

Health-care professionals thus engaged in moral ‘boundary work’ in order to second-guess patients’ claims and determine credibility. The process of boundary work hinged on a multiplicity of dualities superposed on a polarity of ‘good’ and ‘bad’ (Figure 5).

FIGURE 5 Dualities of moral boundary work. (Figure not reproduced here; see the original article.)

In contrast, some HCPs recognised that moral boundary work was flawed and advocated trust as the basis of a therapeutic relationship:

Sometimes we say ‘oh she came in with back pain but I don’t think she’s really in pain’ . . . but really even if somebody is in pain and distress, [it] doesn’t always have to be in how they present themselves . . . that doesn’t mean she is not in pain.

I hate to say it . . . but I used to be one of the people that used to say, ‘Oh, well, they are probably just wanting attention’. But I’ve changed in that matter. People are in pain, and it’s not just to get attention.

Pain is so subjective and so that’s where the difficulty lies . . . I find it hard to say how someone’s pain can be judged by someone else . . . You have to show a patient you’re empathetic to him. There is a pain. Pain is real.

Navigating juxtaposed models of medicine

This described HCPs’ challenge of navigating juxtaposed models of medicine: the biomedical and the biopsychosocial. HCPs described a culturally entrenched pull towards the biomedical ‘siren song of diagnosis’.214 Clinical work, therefore, hinged on proving (or disproving) the presence (or absence) of something ‘real’:

Being able to track something gives me more comfort than going by what you’re telling me . . . because I like to see proof . . . You [want to] . . . be convinced that you’re treating something and that what you’re treating is real.

I will listen to their story, I will examine them and I always say you have got to exclude the physical first that is your job . . . I think as doctors we have an obligation to exclude the physical first and not jump into [psychosocial explanations] because it reduces the patient to being an un-necessary complainer and I don’t believe that they really are.

Unexplained symptoms such as chronic pain were felt to obscure ‘real’ health problems such as diabetes or heart conditions:

They don’t seem to worry about issues that might be real . . . his cholesterol is high, he’s not worried . . . There are some other issues that he needs to attend to, and he’s not worried. His father died when he was 52. He’s not worried. He was 53.

Even those HCPs who advocated a biopsychosocial model could make an abrupt shift towards psychosocial explanations only after they had excluded something biomedical. After this shift, clinical work moved away from diagnosis towards persuading patients that psychosocial factors influenced pain. This abrupt shift, indicating a dualistic biopsychosocial model, could threaten the therapeutic relationship:

The terminology . . . psychiatric and psychological . . . have a stigma attached to them that is not intended . . . we accept that patients with long term pain will have a psychological component to it but actually labelling it as that.

Health-care professionals recognised the potential tension and, therefore, would default to biomedical explanations or use ‘bridging’ strategies to smooth, or even conceal, the shift from biomedical to psychosocial explanations. These explanations still tended to focus on the physicality of the body:

If you present [the pain explanation] as a completely airy fairy psychological . . . then they are going to go away dissatisfied . . . you have got to lead them in gently . . . there is an element of mind and body that contribute.

If you start from the body and if you ask a little carefully how things work when you are physically like that, then it is not threatening, and you can approach things, like, through the body.

Some HCPs felt that offering a diagnosis helped patients to move forward, or gave a sense of relief. Some might even ‘feign diagnostic certainty’222 to achieve this goal:

The fact that they tell you that you have a problem that’s not just to do with your nerves and that there’s something wrong physically . . . Just that gives you a certain sense of relief.

Once I give [a diagnosis] . . . I am ready to say . . . let’s stop doing tests. Let’s stop sending you to lots of different doctors looking for a cure . . . let’s try and make you better . . . [they] have more tests than is good for them.

Well, if she seems to have use for a diagnosis to achieve her goal, I’ll gladly help her.

Others voiced ethical concerns about the deception of feigning diagnostic certainty:

I am very happy for a person to treat the whole person and we do it all the time . . . But I think giving it a label that actually has no justification I think is misleading to the patient and I actually feel quite strongly about that.

Not all HCPs utilised a dualistic biopsychosocial model. Some did not make an abrupt shift towards psychosocial explanations but, rather, utilised an embodied biopsychosocial approach with ‘no breaking point where the physical becomes psychological’.156 In these cases, there was a sense that pain is multidimensional and that the ‘physician gaze’226 is multifocal. Clinical work involved understanding a person’s suffering from the outset of health care, alongside attempts to define pain in biomedical diagnostic terms. These HCPs described an embodied approach to health care underpinned by the aim of understanding that a person is suffering from ‘something’:

If there isn’t a physiological problem, it doesn’t mean that there isn’t an illness, and if the patient is suffering then we should look at the problem and how we can help.

While we talked . . . many losses came up and I began myself to think about what all this was about in fact, what is this pain? Where it came out that there was a lot of disappointment, where there was divorce and . . . yes, it can’t be purely physiological.

Once a person’s life has fallen apart it’s not so much about the pain and the injury anymore. It’s about all these other things in their life and it’s all these other things that need to be addressed in order to get them better and get them back on track.

Those who utilised an embodied psychosocial model recognised that sitting alongside and supporting patients, rather than trying to ‘fix’ them, could be rewarding for both the patient and their HCP. There was also a sense that sitting alongside a person on a healing journey, rather than holding on to the professional desire to ‘fix’, could take away the sense of professional failure and be rewarding for both the patient and the professional:

[It is] a journey rather than ‘this is my one chance . . . and I have to get it all done in one go’ . . . It's about moving away from that place of stuckness and creating a little bit of momentum . . . you know when you're pushing a car.

[The] traditional model . . . doesn’t allow people to express how pain has affected their whole life, it is very homed in to the particular area of the body and trying to fix it . . . [I] find it more satisfying to work in a way that acknowledges and discusses the impact.

Some described how time restrictions and demands in the health-care system could encourage HCPs to focus on the physical body:

The biggest problem in the whole thing is lack of time. Typically these are complex people with multiple problems, and you really could spend the whole appointment, more than one whole appointment, just talking about this.

We are limited by the amount of time with the patient. I know this sounds bad, but [talking about pain] opens a can of worms.

I think as a clinician [you] focus straightaway on . . . a biological type approach to it. I think some of the psychological feelings get more brushed over perhaps . . . very often there is not the space in the consultation.

Navigating the patient–clinician borderland

This theme describes the complexity of navigating the borderland between patient and HCP. First, there is a need for give and take in order to maintain a therapeutic relationship. Second, there is a need to negotiate effective control of the therapeutic decisions. Integral to a negotiated relationship is the need to find common ground. Tension could arise when the patient and the HCP did not share an agenda, in particular when the patient’s expectations of diagnosis, treatment and cure were not met. The challenge for HCPs became to find this common ground or lose the patient’s trust:

People feel let down by their doctors . . . The degree of satisfaction is very low . . . basically because we don’t solve their problem . . . They go from one to the other, they find a doctor who gives them hope and they go to him.

We kept having a difference of opinion: we think that your back isn’t damaged; we think the pain is caused by a false danger message from your brain while there is no tissue damage in your body. He said: I can’t understand that . . .

To effectively maintain an effective therapeutic relationship, HCPs sometimes made clinical choices that they might regard to be of doubtful medical utility because this was what the patient wanted (e.g. referring for an X-ray, prescribing painkillers or referring to secondary health care). Maintaining a therapeutic relationship was described as a matter of give and take and attempts to balance long- and short-term gains were therefore described as integral to clinical work. Concessions were sometimes necessary:

Sometimes patients refuse to believe that their condition cannot be treated . . . and insistently ask for a series of medical investigations that you, as a doctor, would not perform. In such cases, a medical investigation can work as a therapy because it . . . shows that you listened to them.

There definitely have been times when I’ve issued a [sick] note just purely because it’s became so antagonistic in a consultation that’s what I’ve done.

Health-care professionals described the challenge of negotiating control of therapeutic decision-making and of deciphering the borderline between professional expertise and patient empowerment. HCPs were on guard against being ‘bullied’ into clinical decisions, but at the same time they wanted the patient to be involved in clinical decisions. HCPs described times that they found it difficult to stand back and let patients make a decision that they considered to be the wrong one:

I recognise that . . . we are trying to promote learning by giving choice and allowing people to get it wrong . . . get that, although it is still hard . . . not to give advice when I see . . . that the advice can be really helpful.

Trying to allow myself to listen objectively and to . . . sit with the fact that actually [the patient] might want to do something which is wholly unsensible, but allowing that to happen if that truly is what they want.

When opinions conflicted, the HCP ‘short-circuit’ tended to take control.221 This navigational strategy utilised a model of un-negotiated HCP ownership over clinical decision-making.214 Here, professional knowledge is paramount and part of clinical work is to persuade patients to follow their recommendation. Patients were allowed to participate in decisions, but only up to a point. Clinical work was to tell the patient what to do and hope that they did what you wanted:

How do you get them to do what you want?

[Some say] ‘I know so much more about my illness than you do and I’m going to tell you what to do.’ . . . You want to have a team effort . . . but you also don’t want to be bullied . . . they can be the toughest patients to deal with.

However, there was a sense that trying to enforce control or ‘adherence’ was not effective. An alternative strategy involved enquiry and negotiation, rather than persuasion or enforcement and, here, clinical work incorporated ‘wondering together with the patient’:221

If you share your knowledge, not by telling that you are right, but really share what you know about the drugs and the effect . . . then you offer the patient an opportunity to think and decide by himself.

There’s something magical in this. If I’ve understood what the patient wants from me, and he understands what I’m going to do with him, then I have a good chance of succeeding.

Patients have to embrace our suggestion because they are convinced that it is the right one and not because we want them to choose a particular option. If you propose something that is inconsistent with their experience or knowledge, there is a risk that they will not listen to you.

This theme was not supported by studies that explored the experience of HCPs in care facilities for older adults and this might indicate a difference in experience in these situations.

The challenge of dual advocacy

This theme describes the HCP as simultaneously an advocate of the patient and the health-care system. Although representing the patients’ interests, they are also a cog in a larger health-care system. At times it could feel like cogs in the machine were working against the patient and each other. HCPs could find themselves in a position of mediation or gatekeeping within the health and social care system. At times, this might involve undoing the work of other HCPs, and some HCPs felt like a linchpin keeping the ‘wheels’ of the system in place:

I mean it’s a dual role for us obviously we’re responsible for someone’s physical health and this is the second role that we have as a sort of gatekeeper to, you know getting incapacity benefits and the two sometimes don’t sit very comfortably.

It all ends up on our doorstep. It is not only we who face the system – we are mediators of sorts between the patient and the system. Not only must we work with the patient against the system, but with the system as well.

I think it’s good to take ownership . . . as soon as someone gets sort of uncomfortable they will shift to a different prescriber . . . And I honestly think it’s like a ship without a rudder and it’s just going round and round in circles.

At times it felt difficult to access effective specialist services:

There is a really big access issue with the pain clinics right now . . . while I can refer them, their likelihood of getting an appointment, even with strong advocacy from me, is very low.

Many cases are being referred to rheumatology and rehabilitation, to specialist doctors in rehabilitation and some are sent to psychiatry . . . but not all patients are given this opportunity, it’s not open to everyone because there just aren’t the resources to offer these services to everyone.

There was a sense of mismatch between what primary care practitioners expected from secondary care and what they received:

Every time I send somebody to chronic pain [clinic] they come out with more medication, or injections.

Often I find that they are not accomplishing any more than I was and [patients] are often sent back to me with them [pain specialists] essentially saying, ‘we did our best’. It’s very frustrating, because if they were easy patients they wouldn’t have been seeing them.

Others described a working health-care system in which the cogs work smoothly together. This system incorporated the benefits of reciprocity, mutuality and collaboration between HCPs and the benefits of being able to break down or cross professional boundaries and hierarchies. Being confident in other professionals’ capabilities, reciprocal respect, team intelligence and shared knowledge could underpin an effective chronic pain service. Geographical proximity and continuity of staff were described as facilitating teamwork, whereas hierarchical relations could impede effective communication and pain management:

We get a lot of mileage about slapping each other on the back a little bit. And increasing other members of the team’s confidence by respecting other members of the team, their profile is improved, I think.

There’s no sort of hierarchy. There’s the team, then there’s the clinical practice lead and then there’s management. But amongst the team there’s no real hierarchy.

If the team sort of echoes the same message and provides richness in terms of their different perspective on it but it all sort of amounts to the same message, then I think there’s less confusion for the poor clients.

Contrary to this, in residential care facilities, unqualified front-line staff described how they could feel disempowered by qualified professionals:

We’re at the bottom of the ladder, but we do the most important job . . . I mean we’re the ones in the wards and spend the bulk of the time with the residents.

They act like the only thing we’re good for is to go clean butts . . . they really need to realise that we spent all this time with these residents . . . and we do know a little bit about what’s going on . . . they need to give us a little bit more respect.

I’ve charted many times . . . the patient screaming, yelling, pulling hair, blah, blah, blah, and right underneath my charting the doctor writes, ‘Stable’.

Personal costs

This theme described the emotional costs of providing health care to people with chronic non-malignant pain. On the one hand, the biomedical model could create a sense of professional failure for not being able to fix the problem. On the other hand, an embodied approach to pain management could incur a personal cost. As chronic pain did not comfortably fit the biomedical model, some HCPs felt challenged because they could not locate or define pain biomedically:

I’m a scientist, at least I like to have objective things, numbers, data, pictures, lab results. I actually don’t like the challenge of trying to read a person’s mind.

Unpleasant emotions, they explained, can cause as much suffering as physical pain. What, then, is pain? And can it be quantified or even identified as a pure sensation?

The problem is that there is no objective test to diagnose these patients. I don’t have a test that enables me to say: ‘this patient has fibromyalgia or doesn’t have fibromyalgia’.

Being unable to diagnose a problem and offer a cure instilled a strong feeling of professional failure. HCPs found it professionally challenging to be unable to ‘fix’ people despite a strong desire to fix things. HCPs became frustrated and demoralised, and at times they could feel ‘on a hiding to nothing’; ‘how did we fail them?’:213

It’s awful, and I think it’s demoralizing when you leave people in pain. That’s just so disrespectful. I mean you’re supposed to be a doctor, you’re supposed to relieve pain and suffering, and you ignore the pain.

You become a doctor not to tell people I can’t do anything, I can’t find anything, you have this perception of yourself as well that you’re going to sort it out and if you can’t sort it out, it’s frustrating. What’s the point of you being there?

No matter what I do, I can use prescription pad and pen and do sleight-of-hand feats and stand on my knees and perform conjuring tricks! Nothing helps!

An embodied approach to pain management involved recognising the overwhelming loss of a fellow human being and this could come with a personal cost:

We forget how much chronic pain affects the patient. They lose their jobs, they have emotional stress and depression and the depression itself is a big loss of productivity to the patient but also to the entire family and to the community.

What they need is somebody to sit and talk with them and rub their back or they need company and they need understanding and compassion and you can’t give them that in a pill.

Health-care professionals recognised a need to balance the emotional burden of familiarity:

Trying to listen to the person . . . sort of empathise . . . [but] almost protected professionally . . . trying to see where that person was coming from but not letting it become too personal . . . I’ve used the phrase detached empathy.

The craft of pain management

The final theme described clinical work as an experience-based competence or ‘craft’.170 This craft was gained from experience and patient mileage rather than through didactic education or research. At times HCPs felt underskilled in chronic pain management, particularly in relation to psychological therapies:

The problem is, we don’t know how to treat pain. And so everybody is telling me I’m not treating pain well, but nobody is helping me figure out how to treat the pain.

I am not a psychologist . . . Someone bringing out a lot about their past . . . we don’t want to say the wrong thing and it be to someone’s detriment . . . you don’t want to open this can of worms.

It’s fine saying [to a patient] yes well you know that is part of the pain . . . you feel depressed with it . . . I don’t really feel I’m at all competent in knowing what to say to try and help them round that.

Personal experience and maturity, patient mileage and learning from more-skilled professionals (apprenticeship) were described as integral to craft knowledge:

One becomes more stable as a person [with age], and does not really have the same demands and does not believe that one can do everything, that one is able to solve everything . . . Young doctors can have in them, that they believe that they will solve everything.

New grads can’t learn all of this, they need a certain number of years, you can’t teach them all of this, there’s so much they’ve got to learn.

If you have a patient who’s not responding the way that you expect, it’s nice to have someone there who can have a look and see . . . you can see and hear what everybody else is doing and that is how you learn . . . it’s like osmotic learning.

Pragmatic use of guidelines

There was a sense that evidence-based guidelines could challenge or constrain craft knowledge. Although some HCPs felt that evidence guidelines supported a holistic approach,180,207 there was a stronger feeling that guidelines did not facilitate individualised care:

Treatment has to be tailored to patients’ needs and prescriptive guidelines promoting ‘one size fits all’ is not acceptable.

If you work according to the guidelines, you are constrained in your performance . . . what would be left of your independence, your own competence, your own practical experience . . . Am I to conclude then that my training was useless?

Most pain specialists spend most of their time treating people with low back pain that’s what we do. And a guideline comes out that says, you’re not needed. Full stop. It can be managed without any reference to you whatsoever . . . no need for doctors and nurses.

Health-care professionals showed limited attention to guidelines in clinical decision-making. Clinical work was about the craft of learning from doing things:

My experience is that therapists say they adhere to the guidelines, although they still all work in different ways.

Such a huge heap, such a bundle of paper, such a bundle of characters . . . we have been educated to do things. So if you give this group a pile of papers, who will read them? I think nobody will.

Health-care professionals framed guidelines within the remit of their own professional knowledge and used them pragmatically and flexibly:

But of course, I’m free to take or leave these things, to look at whether they suit my own ideas of how to approach my patients.

Of course, it is not necessary to follow the guidelines exactly; it is more like: this is roughly the approach, regardless of the background you have.

For example, some used the evidence guidelines as a tool to convince patients or colleagues that their professional decision was the right one:

In the case of disagreement . . . about the treatment policy, you can always turn to the guidelines, and you can argue while showing them these national guidelines.

Health-care professionals described the decision to refer a patient for clinical testing as a craft. This highlights the multiplicities of clinical decision-making whereby HCPs face the challenge of balancing varied competing interests in order to achieve an optimum outcome (Figure 6).

FIGURE 6 Multiplicity of clinical decision-making: do I refer for a test? (Figure not reproduced here; see the original article.)

First, a test or investigation might be used to confirm or to rule out serious pathology. This might be to reassure patients or HCPs, or alternatively to convince a patient or HCP that there was no pathology:

‘Wear and tear’ is, in a way, quite nice to have. Most are satisfied with that . . . nothing dangerous, and nothing that needs surgery. It’s a short version of an explanation.

At times, there might be a need to rule out pathology as a defence against complaints of malpractice.173 However, there was a sense of ambivalence regarding the usefulness of tests in clinical decision-making:

What are we afraid of? After all, it occurs to every doctor sitting here that he may just miss something. That is how we were taught, we were told watch out, you may miss something. How much did we miss? Almost nothing.

X-rays could also be used strategically to support patients’ credibility and to maintain trust. Referring a for an investigation was felt to be a strong statement that patient I hear you, believe you and I am even prepared to bend the rules for you. Some ordered X-rays because the patient asked for one, or as a compromise for a preferred or less accessible service:

[An X-ray referral] is interpreted as a definite signal that the physician thinks it is something physical. It means . . . [the patient] can come home and say, ‘I had an X-ray’, and then everybody will realise I have pain in my back.

Investigations could also be used as a means of biding time or buying time. It was sometimes quicker just to order an X-ray. It might also save time en route to a different service:

I buy time by ordering an imaging test, even though I know it will not be helpful . . . It buys time . . . It’s a kind of therapy. It reassures the patient that they’re being cared for.

Sometimes I found myself referring a person for an X-ray in order to clear the waiting room and allow myself 2 minutes of breathing time . . . the patient keeps quiet while I write out his referral. Sometimes you find yourself doing this and it goes against any reasoning or logic.

Some erred on the side of caution in ordering investigations and considered the potential repercussions of securing a false positive:

He’s got a scan of his disc, it’s there [a disc bulge], but I think he’s had it for years . . . but he’s now fixated on a disc.

It come back at me like a boomerang: ‘now I’m worn out, I can’t work any more, I’ll go over to social security benefit’.

A word of caution given was to think very carefully about the person for whom the test was being done:

[Be] very self-aware when you're thinking about doing tests . . . sometimes in chronic pain doctors are doing the tests for themselves not for the patient . . . that uncomfortableness with not being able to fix or do something . . .

Conceptual model

The final phase of meta-ethnographic analysis is to develop a conceptual model that is abstracted from, but more than, the sum of its themes. The reviewers developed a model that helps us to understand the experience of providing health care to people with chronic non-malignant pain (Figure 7). The model is underpinned by a series of tensions: (1) between a dualistic biomedical model and an embodied psychosocial model; (2) between professional distance and proximity; (3) between professional expertise and patient empowerment; (4) between a need to make concessions in order to maintain relationships and known biomedical utility; and (5) between patient advocacy and health-care system advocacy.

FIGURE 7 Conceptual model: complexity of providing health care to people with chronic non-malignant pain. (Figure not reproduced here; see the original article.)

We conceptualised these tensions, on a mixing console, as underpinning the craft of chronic non-malignant pain management. The poles are neither inherently good nor bad; just as bass and treble are neither inherently good nor bad. It is the correct mix within a context that contributes to the quality of music. Our console also incorporates the pitch or level of loss, both professional and personal, that can contribute to the harmony or dissonance of a therapeutic encounter.

Opioid prescription

The reviewers also identified 76 concepts in 17153,155,157,161,174,175,177,187,189–191,198,208,211,213,217,218 of the 77 studies that uniquely explored the experience of prescribing opioids to patients with chronic non-malignant pain. The reviewers discussed and organised the 76 concepts (see Appendix 3) into 19 conceptual categories (Table 12), and then into six themes that underpin HCPs’ experience of prescribing opioids to patients with chronic non-malignant pain:

1. should I, shouldn’t I?
2. pain is pain
3. walking a fine line
4. social guardianship
5. moral boundary work
6. regulations and guidelines.

A core concept overarching all six themes was a sense of ambiguity surrounding opioid prescribing. We illustrate the six themes with narrative exemplars. Indicators of confidence in our review findings as recommended in the GRADE-CERQual framework37 are given in Table 13.

Should I, shouldn’t I?

The theme ‘should I, shouldn’t I?’155,157,161,174,175,177,191,198,217 described uncertainty about when to prescribe opioids, and a feeling of ambiguity about the effects of medication:

Depending on who I run into, I may get into all kinds of problems . . . ‘You’re just being unfair because he’s a drug addict . . .’. But if I run into somebody else, they might say ‘No . . . that was a good clinical decision‘.

The aetiology of the disease is not really known and you have few means of knowing what you’re doing. You’re treating the pain and you don’t know why there is no response . . . No matter what you give them, the pain doesn’t go away.

Some felt that clinical education did not prepare them to make these decisions:

We took an advanced pharm[acology] class, and we discussed it in one lecture, but that was it. Isn’t that ridiculous considering how many people we see in pain?

Nobody here knows how to treat pain in anybody who has a history [of addiction] and already on something like methadone. Nobody knows how to treat them . . . Most of us in primary care end up [doing it] by default. But that’s not good.

Uncertainty was compounded by the sense that specialist referrals were either restricted or unproductive:

Often I find that they are not accomplishing any more than I was and [patients] are often sent back to me with them [pain specialists] essentially saying, ‘we did our best.’ It’s very frustrating, because if they were easy patients they wouldn’t have been seeing them.

Pain is pain

The theme ‘pain is pain’153,177,187,189,211,217 was underpinned by the concept that if a person is in pain then the primary aim of the HCP should be to try to relieve their pain. If a person is in pain, even drug addiction should not be a barrier to opioid prescription:

At the end of the day, if someone’s got chronic pain it doesn’t matter if they’re addicted to painkillers if it sorts out their quality of life.

I had a guy last week who’d been stabbed . . . and he had to discharge himself because they wouldn’t give him any pain control . . . he wasn’t even getting his prescribed dose of methadone . . . there’s a protocol . . . they choose not to know about it and it’s just pure stigma.

However, HCPs felt that the decision to prescribe opioids was complex and that, although in theory pain is pain, in practice it is different. HCPs described differences in practice that would make them less likely to prescribe opioids. For example, in the case of chronic non-malignant pain, they would need to carefully balance the risks and benefits of long-term opioid prescription:

With [malignant pain] . . . your aim always is to get complete relief of pain . . . For chronic pain . . . you’ve got to weigh up . . . the potential side effects . . . you are not necessarily going to get them pain free because they’ve got the rest of their lives to live.

Health-care professionals explicitly compared prescribing practices for palliative care and chronic non-malignant pain:

We tend to focus too much on pain control for palliation as opposed to just everyday clients. Certainly nobody wants to die in pain, but nobody wants to live in pain either.

It might be that you are . . . just getting them through the last few weeks . . . so maybe we feel more comfortable, whereas the patient with chronic back pain you don’t how long you’re going to be treating . . . I think it’s a risk–benefit issue.

Walking a fine line

The theme ‘walking a fine line’153,157,162,177,187,188,190,194,208,211,217 described the need for HCPs to carefully balance the benefits and adverse effects of opioids. On the one hand, emphasising adverse effects might lead to unnecessary pain, but, on the other hand, emphasis on pain control might lead to harm or abuse:

There are two mistakes you make . . . the mistake of undertreating or of giving medicines that end up being sold or used for unintended purposes. You’re going to make errors both ways, and I think it’s generally better to risk opiates being misused versus not treating someone’s pain.

This sense of walking a fine line was heightened when prescribing opioids for older adults, and HCPs felt that specialised knowledge was necessary to effectively prescribe opioids to older people with chronic pain. Although in theory age should not affect decisions, there was a sense that in practice it did:

But there are safety issues, and at the end of the day if they came to grief . . . you’d feel . . . excessive side effects . . . that contributed to some major event . . . So we walk a fine line sometimes between giving adequate pain relief and giving safe treatment.

Older people metabolise medication differently . . . so you don’t want to give them medication that’s going to impair their ability to function . . . You don’t want to interfere with their ability to make judgements and so on, so I don’t like using opioids in elderly people at all.

I just have a hard time prescribing opioids in my older patients. I get frightened with 80+ year olds; how are they going to respond? Am I going to absolutely drop them to the floor even with a small dose?

Social guardianship

The theme ‘social guardianship’153,155,157,174,175,177,190,191,208,211,217 described a culture hostile to opioid use and the professional taboo of prescribing opioids for chronic non-malignant pain. HCPs compared their own practise with that of their colleagues and were worried that they would be harshly judged by their peers for prescribing opioids. Some felt a personal responsibility to protect society from the consequences of opioid misuse and viewed certain patients with suspicion, particularly those who requested opioids. To protect society, some implemented strategies to control patients’ behaviour (e.g. bottle checks, opioid contracts and background checks). HCPs were also concerned that opioid prescriptions might be diverted to others:

I am a naysayer on opiates . . . too much of my day is spent policing how many [opioids] have been prescribed and how many times a patient is a return patient and how often they visited requesting opiate prescriptions.

If you prescribe to a population where you think diversion is going on, you definitely have a responsibility. I also worry about who is getting the drug, is it my son? I mean, we are members of society after all.

Some HCPs discussed indicators of potential abuse (e.g. lost prescriptions, early requests for medication and frequent attendance), whereas others acknowledged that these might actually indicate poorly managed pain and, therefore, a greater need for pain relief:

If people are taking it genuinely for pain they tend to stick to the prescribed dosage . . . addicts tend to be the ones who are always ordering early . . . you don’t lose your tablets if you are . . . getting great benefit from them for pain.

The concern would be is this pain real, or is it just put on to obtain opioid? . . . I mean, an assessment of the pain and whether I think it’s genuine or not. I think it’s very difficult; it’s something I’m currently dealing with at the moment, and not very successfully.

Moral boundary work

The theme ‘moral boundary work’153,155,157,174,175,177,190,191,198,208,211,213,217,218 was underpinned by the clinical work of deciding whose pain is ‘real’ and thus who should be prescribed opioids. There was a sense of an underlying unwillingness to prescribe opioids for pain with no clear biomedical cause, or if symptoms were vague. Suspicion arose when there seemed to be dissonance between the patient’s report of pain and the symptoms observed by the professional:

A lot of patients . . . they really need it . . . based on their underlying pathology . . . a patient who has a cancer or a real anatomic foundation . . . you have to count on more observation, combined with other clinical data . . . you pretty much know who is abusing and who is not.

I think for patients who have chronic pain it’s more challenging and I think that’s the place where I’m constantly rethinking my practice . . . You’re always on the fence: am I doing the right thing for my patient?

Health-care professionals described good patients (e.g. those who made appropriate demands, took advice and did not cause trouble) and ‘difficult’ patients (the demanding, non-adherent and trouble-making):

For those patients that have a legitimate reason for wanting to take it and if I can trust them . . . most of these are older patients of mine. They never request early refills, they don’t go to the [emergency room] in between visits to get them . . .

Non-clinical moral judgements or gut feelings contributed to prescribing decisions. HCPs recalled episodes when they had made a mistake by trusting ‘the wrong’ patient. Over time they felt that they had become better at making the right decision. Examples were given of HCPs getting their fingers burned when prescribing opioids, along with an understanding that lessons had to be learnt from mistakes:

At times there was a feeling of being ‘deceived’ by patients so that you prescribed opioids:

I’ve had trust in people, and it’s been betrayed – sometimes the trust is betrayed multiple times. I find I’m not always that great a judge of who to trust and who not to trust, so, I don’t trust my own judgement on trust all the time . . . I think people feel like they’ve been violated, you know, cheated, like they’ve been taken advantage of. I feel some of that, too. Ultimately you feel you’ve made a poor judgement, and you get mad at yourself . . . My impression was that he had a true ankle problem. Then you find out it was all lies, but you know, at that time and at that moment, that was my assessment, and I did it. You’re allowed to make mistakes.

Prescribing . . . is a sort of rather woolly, nebulous product . . . I’ve been moulded by the successes and the failures . . . we all learn on the hoof, don’t we? . . . I think everybody’s fingers get burnt with people who you give the opioids to with a more trusting attitude.

However, some sensed the dangers of judging a book by its cover and acknowledged that basing clinical decisions on their gut feeling might not be fair or accurate:

There’s a disconnect . . . even if it’s the sweetest little 85-year-old woman who looks like your grandmother, versus, you know, some guy from the ghetto wearing his pants down at his knees . . . it shouldn’t really matter.

Regulations and guidelines

The theme ‘regulations and guidelines’155,157,175,177,190,191,217,218 described HCPs’ views about external regulation of opioid prescription (specifically guidelines, opioid agreements and drug screening). Some described a negative view of prescribing guidelines and felt that they interfered with professional autonomy, an example of the ‘legislature practising medicine without a licence’:190

You’re there to help them and they can tell you their deepest, darkest secrets, but yet you’re policing them . . . I like to see the person as a person . . . You can’t do your job when you are thinking about these things.

There could be negative implications to that if patients are actually leaving the emergency department because of the way they interpret that [regulations] poster . . . there’s potential that sick patients could actually leave your emergency department when they need help.

Health-care professionals feared legislative reprimand if they prescribed outside guidelines:

My name is on that bottle . . . I had a patient die. He took the entire bottle, and the police came to see me because they found him dead with the empty bottle with my name on it . . . I won’t accept that burden.

[GPs] are scared of being in front of the coroner . . . with a load of angry family and relatives wagging a finger at us, when all we’re trying to do is really help . . . if it works, brilliant. If it doesn’t work . . . people are very quick to criticise.

Others described a negative view of opioid prescribing agreements and drug screening as striking a blow at the heart of a patient–clinician relationship by creating mistrust and hostility:

It can really strike a major blow to trust in the doctor patient relationship when you ask someone to sign a piece of paper . . . if there is already mistrust between the patient and the doctor, it could heighten that mistrust . . .

Some used guidelines pragmatically as leverage or to justify decisions and thus help them to deal with ‘challenging’ patients:

I tell them this is standard protocol. I’m not singling you out. I’m not picking on you. I’m not treating you like an addict. This would happen to anybody. If you take our chronic pain meds long enough, anybody will become physically dependent on them.

[An agreement] gives me leverage or comfort in discontinuing the medication . . . because we’ve kind of laid it out from the beginning that those behaviours were not okay . . . it made my life a little easier, but I’m not sure it did the patients a giant service.

Others described a more positive attitude to regulation. For example, opioid agreements could be useful in establishing boundaries and opening up honest discussion:

I think it improves the care, because you are able to then have more open and frank discussions around their pain . . . and [about] other things going on in their life . . . In the best of circumstances it actually will make for a deeper more trusting relationship.

Conceptual framework: prescribing opioids

Our conceptual model (Figure 8) hinges on the HCP’s need to decipher the ambiguity surrounding opioid prescription for chronic non-malignant pain (should I, shouldn’t I?). Social suspicion and hostility towards opioids (social guardianship) will tip the balance for prescribing towards the negative, whereas the pre-eminence of pain (pain is pain) might tip the balance towards prescribing. The decision is not clear-cut. First, HCPs might make non-clinical judgements about the person (moral boundary work). Second, they must determine the balance of positive and adverse effects for each individual (walking a fine line). Last, there is a sense of professional ambivalence towards prescribing guidelines (regulations and guidelines). This conceptual model demonstrates the complexity of making a decision to prescribe opioids to someone with non-malignant pain. It also demonstrates that the decision is influenced by intra- and interpersonal factors and broader external concerns.

FIGURE 8 Conceptual model: the ambiguity of prescribing opioids to patients with chronic non-malignant pain. (Figure not reproduced here; see the original article.)

Discussion

We aimed to undertake a QES using the methods of meta-ethnography reported by Toye and colleagues8 to increase our understanding of what it is like for HCPs to provide health care to people with chronic non-malignant pain and thus to inform improvements in the experience and quality of health care. At the outset of this study, we had intended to include HCPs’ experience of treating chronic non-malignant musculoskeletal pain in order to mirror a previous QES of patients’ experience of chronic non-malignant musculoskeletal pain. However, our preliminary reading indicated that HCPs’ experiences of treating chronic non-malignant pain were not boundaried to a particular body system, but were a summative experience that cut across conditions. After consultation and agreement with our advisory group, we therefore sought and attained permission from the funders to remove ‘musculoskeletal’ from our study title. Our search strategy then focused generically on chronic non-malignant pain. Further research might focus on specific diagnoses (such as neuropathic, visceral, pelvic or phantom pain or arthritis) to explore potential similarities and differences in HCPs’ experiences of treating these conditions.

This is the first time that such a synthesis has been undertaken on this topic. Already we know that, from the patient perspective, the experience of health care can be adversarial.8 Patients with chronic pain struggle to affirm their sense of self; their present and future appears unpredictable; they search for a credible explanation for their pain; they do not always feel heard, believed or valued by HCPs; and they struggle to prove themselves in the face of scepticism.8 Our findings can help us to consider this experience of health care from the perspective of HCPs. We identified six themes from 77 studies that help to explain the HCP experience of providing health care to people with chronic non-malignant pain and six themes specific to the experience of prescribing opioids to this group of patients. The innovation of our conceptual model is to propose a series of tensions that are integral to the experience of providing health care to people with chronic non-malignant pain. These tensions were between a dualistic biomedical model and an embodied psychosocial model; professional distance and proximity; professional expertise and patient empowerment; the need to make concessions to maintain therapeutic relationships and the need for evidence-based utility; and patient and health-care system advocacy. This model may be transferable to other chronic conditions.

A sceptical cultural lens and siren song of diagnosis

From the HCP perspective, the findings highlight a deep-seated cultural scepticism towards people with chronic non-malignant pain. This scepticism influences a powerful gravitational pull towards the biomedical model and siren song of medical diagnosis. The biomedical model takes disease to be an objective biomedical category not influenced by psychosocial factors or judgements about moral worthiness.227 Alternatively, the biopsychosocial model focuses on the embodied experience of illness.227 HCPs describe the challenge of navigating juxtaposed biomedical and biopsychosocial models. Although many HCPs may feel that they adhere to a biopsychosocial model, our findings indicate a dualistic approach whereby HCPs make a sudden shift to biopsychosocial explanations after exhausting attempts to decipher a diagnosis. In essence, psychosocial aspects are ‘grafted onto the somatic’ core (p. 44).228 Our interpretation supports the need for a more embodied biopsychosocial model that focuses on the personal meaning of pain. Embodiment theory has its roots in the writings of Merleau-Ponty,229 who breaks down the dualism of mind and body and focuses on the personal meaning of suffering. Personal meaning always accompanies the experience of pain.228,230 From a patient perspective, both covert and overt scepticism from HCPs, along with a cultural pull towards biomedical explanations, may help to explain why patients with chronic pain can experience a strong sense of not being believed.8 An abrupt shift in explanation towards the psychosocial by HCPs may exacerbate this loss of credibility, particularly in the context of persisting cultural dualisms (real/unreal, physical/mental, medical/psychological and body/mind). We know that people in chronic pain can feel a profound sense of shame and stigma because they do not fit the dominant medical discourse and that they struggle to prove to others that they are a credible and ‘good’ person.8 A more embodied non-dualistic biopsychosocial approach at the outset might help HCPs to support patients with chronic pain. These findings highlight that both HCPs and their patients share a struggle to negotiate dual explanatory models for pain. However, although the patient might persevere in efforts to find a medical diagnosis in order to gain credibility, the HCP might make an abrupt shift towards the biopsychosocial, and this is a potential source of tension.

Navigating the patient–clinician health-care interface

From the HCP perspective, our findings demonstrate the complexity of navigating the interface between themselves, patients and the health-care system. Navigating the patient–clinician borderland, the challenge of dual advocacy and the craft of pain management all indicate the challenges of managing this interface. The finding that HCPs sometimes make concessions that are not evidence based in order to maintain effective relationships may have implications for policy and practice. For example, it might help to explain why an increasing number of HCPs are prescribing opioids despite very limited evidence for long-term opioid therapy for chronic pain before the end of life,231–233 or why a doctor may refer a patient for an investigation when its efficacy is not well supported. The findings demonstrate that HCPs can struggle to find a balance between providing professional expertise and empowering their patients. At times, this can mean that they feel pressed into making decisions of limited clinical utility. The short circuit can be to take control of the clinical encounter, thus alienating the patient from any sense of control.

Our findings indicate that HCPs can find it a challenge to simultaneously represent the patient and the health-care system. They also demonstrate the benefits of mutual respect and shared knowledge for an effective chronic pain service, and suggest that recognising mutual skills and combined knowledge of colleagues rather than focusing on differences might benefit health care. From the patient perspective, we know that the person with chronic pain experiences an ambivalent relationship with the health-care system and can feel trapped in a system that is not meeting their needs.8 The craft of pain management describes clinical work as an experience-based competence that at times could feel constrained by external guidelines. At times HCPs felt underskilled in chronic pain management, particularly in relation to psychological therapies.

Personal costs

The findings demonstrate that from a HCP perspective there are emotional costs involved in providing health care to people with chronic non-malignant pain. HCPs can feel that they have failed in their professional duty to find and fix a medical problem. On the other hand, those who take time to understand their patients’ personal losses can feel ‘bombarded by despair’.220 We know that from a patient perspective there are also profound personal losses. The person is struggling with an overwhelming loss of present and future self and does not always feel valued as a person in health care. They look back nostalgically to the past self and feel a sense of great loss. They describe a fundamental need for their HCP to believe and bear witness to their suffering.8 Our findings have clinical implications: how do we support HCPs to sit alongside patients and yet not be able to fix them, without experiencing the losses of failure? It might be useful for clinical educators to consider overlaps in training need between palliative care, which by definition does not aim at cure, and chronic pain management. HCPs included in this review did not discuss their own personal life context, which may, at times, contribute to and make HCPs less resilient to personal or professional losses. Our findings suggest that it may be useful to frame a more embodied approach as one that has advantages for both patients and HCPs. Although seeing the patient as a fellow human can impose a personal burden, it can also provide a mutual and positive human experience.234 Sitting alongside a patient frames the clinician as an advocate, rather than adversary. Charon235 demonstrates that sitting alongside a patient, although emotional, can have a positive impact on work satisfaction by allowing clinicians to interact on a human level. This may take the pressure off clinicians to fix their patients.

Using the conceptual model in clinical education

We present our conceptual model as a mixing console that can help us to understand, think about and modify our experience of providing health care to people with chronic non-malignant pain. The model is underpinned by a series of tensions:

• biomedical model versus an embodied psychosocial model
• professional distance versus professional proximity
• professional expertise versus patient empowerment
• making concessions to maintain an effective relationship versus biomedical utility
• patient advocate and health-care system advocate.

Health-care professionals can use this mixing console to help them to think about their encounters with patients with chronic pain, for example:

• Am I making a sudden shift to psychosocial explanations?
• Am I considering psychosocial factors at the outset?
• Do I understand this patient’s experience or am I too distant?
• Am I trying to enforce my decisions?
• Am I making a concession and for whose benefit?
• How is my patient experiencing this dual advocacy?
• What personal impact is this having on me?
• Am I feeling like I have failed?

Our console will allow HCPs to consider their individual mix and contemplate a re-mix if necessary in order to successfully support people with chronic pain, or to individualise their approach for different patients. Using the mixing console could potentially be used to help appreciate and alter the tensions during a consultation with a patient. This fits in with Mead and Bower’s15 framework for patient-centred care, which recognises both the importance of the therapeutic alliance and the influence that HCPs’ personal characteristics and responses can have on care. Further research to explore the usefulness of this console within a therapeutic encounter would add benefit to this QES.

Opioid prescription

Despite recent guidelines suggesting that patients might not benefit, there has been a significant rise in opioid prescription for chronic non-malignant pain.231,232 The findings do not indicate that the limited efficacy of opioids for chronic non-malignant pain is the primary barrier to prescribing. Our findings indicate ambiguity surrounding the prescription of opioids to patients with chronic pain. Although some were aware of the adverse effects of opioids, they were concerned that choosing not to prescribe would lead to unnecessary suffering. Although HCPs discussed the potential harm from opioids, there was also a feeling of personal responsibility to police and protect society from opioid misuse. A more positive and collaborative frame for making the decision to prescribe, or not, might be useful, for example emphasising the limited efficacy and serious adverse effects of opioids.231,232 The findings demonstrate ambivalence towards the external regulation of opioid prescription. There was a sense that regulation limited professional autonomy and that there should be freedom to prescribe. Some felt that regulation could lead to mistrust and hostility. Others felt that opioid agreements could help to open up honest discussion. Others used them to justify difficult or unpopular decisions. In view of the rise of opioid prescription and recent guidance, further research to explore HCPs’ experience and view of guidelines would be useful. Although we did not set out to explore the experience of prescribing opioids for chronic non-malignant pain, the findings indicate that opioid prescription is an important facet of the HCP experience. In qualitative research methodologies, unexpected findings can demonstrate that the reviewers have not been constrained by a priori concepts. Research to explore patients’ experience of deciding to take opioids for chronic pain is also timely. This would help us to understand the motivations and experiences of those with chronic pain and the HCPs trying to manage that pain.

Film outputs from qualitative research

We produced a film output to present themes from this QES. Findings from a previous meta-ethnography of patients’ experience of living with chronic pain have received > 29,000 views. YouTube comments suggest that it resonates with experience and that watching the film can make people think. The film has also been utilised in pain management education.220 Qualitative research aims to understand a variety of perspectives and it can be challenging to demonstrate specific impacts on health care. Its value includes encouraging us to think outside our own experience and challenge our values and practices.10 Parsons and colleagues236 argue that the impact of visual representations of qualitative research might be a ‘subtle shift in viewers’ perspectives’. Visual media lend themselves to an interactive, or dialectic, style of learning,237–245 which can evoke, provoke and stimulate ideas.239 They can be powerful because they facilitate emotional engagement beyond that from reading reports.236,246 Performative methods have been used in clinical education to facilitate learning through dialogue246–249 and to develop empathetic understanding.241,242,250,251 Through film, viewers can access different perspectives in a safe environment and explore their own clinical practice. Although there are diverse ways of knowing,252 evidence-based medicine has a strong strand of objective modes of knowledge, or episteme. Greenhalgh253 invites us to challenge accepted ways of knowing and incorporate other forms of knowledge, knowledge that can be conceptualised as a dynamic process occurring at the interface between ideas and audience. Future research to explore the impact of this film on relevant stakeholders would help us to understand the impact of qualitative research films on health-care experience.

Methodological issues

Are we constrained by a priori concepts?

The findings of qualitative research will inevitably be the authors’ interpretation. It is impossible to conceptualise, or think, without existing categories in mind, and researchers will always bring existing ideas and points of view into their thinking and analysis. This should not be seen as a weakness in qualitative methodologies; on the contrary, useful qualitative analysis hinges on utilising existing concepts. This resonates with Blumer’s sensitising rather than definitive concepts: ‘definitive concepts provide prescriptions of what to see, sensitising concepts merely suggest directions along which to look’ (p. 7).254 It is more useful to consider qualitative analysis not as a linear, but as a dialectic process in which tension between existing and new concepts can create innovative ways of thinking.255 Researchers contemplate their data with an open mind but not an empty head.256 Existing concepts become a limitation only if you are reluctant to change them. This is why it is necessary to collaboratively challenge developing concepts in a team environment in which reviewers feel safe to challenge each other. The strength of our team was that members felt free to agree, disagree or change their mind within the safety of the group.

How much should we search for?

Campbell and colleagues,27 in their meta-ethnographies of 25 and 37 studies, suggest that ‘40 papers is realistic [in order] . . . to maintain sufficient familiarity’. Toye and colleagues28 suggest that it is possible to include a larger number of studies in reviews. The size of a QES will ultimately depend on the aims and resources available (both financial and published). Researchers and funders should consider whether or not an exhaustive search of the literature will add value. A growing volume of published qualitative research means that there will inevitably be an increase in the potential number of studies for qualitative syntheses. Reviewers therefore need to consider what they include. However, there are issues to consider when deciding what not to include. Fundamentally, how do we determine which are the most valuable studies?

We made the decision to systematically search for, and include, all papers that we felt were ‘good enough’. Some qualitative researchers suggest a more targeted approach to sampling data. For example, you could stop searching for new data when ‘theoretical saturation’ is reached (i.e. when collecting additional data seems to add no more insight).41 This is more comparable with the sampling strategies commonly used in qualitative research. One option might be to start your search with a single database and expand as analysis proceeds.23 However, there are some important issues to consider when deciding what to include or exclude from QES. Importantly, reviewers do not necessarily find the conceptually rich papers first. This means that reviewers may need to read and consider several ‘thin’ studies before finding rich and key concepts. Sometimes gravitational concepts develop from a small number of studies; yet at other times it may take a larger number. We also need to consider recent development that aims to evaluate confidence in QES, specifically GRADE-CERQual for determining confidence in reviews. GRADE-CERQual considers adequacy of data (overall determination of the degree of richness and quantity of data supporting a review finding) and coherence (whether or not the finding is well grounded in the primary studies and whether or not reviewers have considered variations across studies), as important determinants of confidence. We feel that a systematic exploration of the available research can contribute to confidence in findings. It is also important to consider that if review findings are to be used to influence policy and practice we may be more confident in the adequacy of data if the finding is drawn from a large pool. Although it can be time-consuming, it can add confidence if we systematically search for all relevant studies.

How can we determine confidence in review findings?

There have been calls to standardise reporting of qualitative syntheses,29,257,258 but there is currently no agreed way of making an assessment of confidence for QES. As qualitative analysis is underpinned by an interpretive framework, efforts to regulate it will always be complicated by varying opinions vis-à-vis quality.34,36 Establishing confidence in the output of qualitative research synthesis (ConQual)259 and GRADE-CERQual37 offer overlapping but distinct approaches for determining confidence in QES.

‘Methodological limitations’ (‘dependability’)

Both ConQual259 and GRADE-CERQual37 consider the quality of included primary studies to be an important determinant of confidence in a particular review finding. GRADE-CERQual refers to this as methodological limitations and ConQual refers to this as dependability. However, determining quality in qualitative research is complex and unresolved. Although there are a growing number of frameworks for appraising the quality of qualitative research,27,34 and a growing number of reviewers appraising primary studies,24 a significant number of qualitative reviewers take a legitimate choice not to appraise for the purpose of QES.27 This choice does not imply low quality. Indeed, although quality appraisal might help us to recognise methodological flaws, it does not necessarily help us to appraise the value of findings.36

We found that the CASP was a useful way of framing discussions on quality and papers that we considered to be ‘key’ did gain higher CASP scores. However, this threshold was not clear-cut and it is unlikely to be useful as a threshold score for determining inclusion in and exclusion from qualitative syntheses. There was a much higher level of agreement on CASP score35 than in our previous study.8 This may be because the reviewers have developed their view on what constitutes quality over the process of working together for several years. ‘Satisfactory’ papers all scored ≥ 20 on CASP, similar to a previous study in which satisfactory papers scored ≥ 19. However, two studies scoring > 20 were still excluded as ‘fatally flawed’. The reviewers found it challenging to decipher the concepts in these studies.

We also aimed to test the usefulness of a list of themes developed from a qualitative study embedded in a previous meta-ethnography funded by the HSDR programme (see Table 2).36 Unlike CASP, these themes were developed specifically for meta-ethnography. Findings from this appraisal confirmed that qualitative studies are very likely to report their rationale, aim and sample adequately. However, there are areas for improvement, specifically:

1. Qualitative authors do not always give readers adequate insight into the researcher’s perspective.
2. They do not always report how they have challenged their own interpretation.
3. They do not always use appropriate exemplification to show that their interpretation has come from the data.
4. They do not always articulate their concepts with adequate clarity.

Fundamental to conceptual qualitative syntheses is the capacity of included studies to make the reviewers think and develop concepts. Studies reported as having ‘changed the thinking’ of at least one reviewer tended to score higher. Appraisal comments (see Appendix 1) suggest that even if it did not change thinking, the primary studies did encourage reviewers to think.

Finally, although reviewers agreed that if studies were ‘key’, ‘fatally flawed’ or ‘irrelevant’,23 the large majority of studies were judged to be ‘satisfactory’. This confirms findings from our previous study.8 It would be useful to find ways in which we can be more discerning about what to include. We felt that the CASP question 10 (how valuable is this study?) had face value in providing a particular reviewer’s appraisal of value. Further research to explore ways of deciphering the potential value of particular qualitative findings would be useful, particularly in the context of research proliferation.

Reviewers and funders need to consider why we are spending a lot of time appraising studies and whether or not this has any impact on review findings. Does it make any significant difference if we include studies with a poor methodological report27,260 (particularly if our unit of analysis is a concept)? Fundamentally, some concepts will exert a strong influence on concept development and other concepts will not. The reasons for this are complex and will depend on factors that include a priori knowledge of the reviewer and the persuasiveness of the written primary findings. We think that, for the purposes of determining confidence of review findings for GRADE-CERQual, an appraisal of either ‘satisfactory’ or ‘not satisfactory’ would therefore be fit for purpose. QES is a time-consuming process and time might be better spent abstracting concepts rather than pettifogging over fine details of appraisal, particularly when there is no agreed method for determining what good quality is.

What is a satisfactory study for qualitative evidence synthesis? Suggested themes for future research

For future QES, the questions shown in Table 14 might help reviewers to consider whether or not a study is satisfactory for inclusion into meta-ethnographies. This list of questions is intended as food for thought, rather than as an appraisal checklist. Future research could explore the utility of these questions for conceptual QES. Importantly, reviewers need to consider whether or not their chosen appraisal criteria have any real impact on their decision to include studies. Although we agree that it is vital to exclude studies that are methodologically fatally flawed as this will have a direct impact on the truth claims (e.g. if the study is unethical), we would argue that as long as studies are satisfactory, then appraisal for conceptual reviews hinges on being able to extract clear concepts. Studies that lack conceptual insight will not influence analytical decisions and, therefore, are unlikely to have an impact on conceptual development; they will therefore do no harm.

Relevance

We agree that studies included in a review should be relevant to the topic under review. However, studies that we rated as ’partially’, ’indirectly’ or ’uncertainly’ relevant also contained useful concepts (i.e. although a study may be partially relevant, it may contain some concepts relevant to the topic under exploration). We think that for the purposes of determining confidence of review findings for GRADE-CERQual reviewers should include only concepts that are relevant to the topic, but these concepts may be present in studies that are partially or indirectly relevant.

Adequacy

Adequacy regards the depth of data supporting each finding. Lewin and colleagues37 define it as ‘the ‘overall determination of the degree of richness and quantity of data supporting a review finding’. However, how can reviewers/readers determine the degree of richness or quantity of data that is adequate to support a finding? It may be useful to conceptualise adequacy in terms of gravity and gravitational pull. One could argue that as data contribute to the development of a concept, they begin to create their own gravitational pull that moulds them into shape. At this point the number of data is adequate. Concepts develop as part of a dialectic process and it would therefore be impossible to define the point at which the concept is pulled into a coherent shape. Thus, determination of adequacy is complex and subjective. Qualitative researchers should demonstrate that their interpretations have gravitational pull. This can be challenging as an appraisal of gravitational pull is qualitative. However, it can help the reader if reviewers provide clear information about the source of their concepts, along with clear and appropriate exemplification of this concept. If the reader feels that the narrative exemplar does not illuminate the reviewers’ concept then this will decrease the gravity of that concept. This resonates with Munn and colleagues’259 concept of ‘credibility’ as the ‘the congruency between the author’s interpretation and the supporting data’ (p. 4259).

It also important to consider that a single concept can exert a strong gravitational pull and that gravity has a qualitative component. We advocate caution regarding making truth claims purely on sheer weight of data; the tale of the Emperor’s New Clothes has taught us the validity of a small voice. Single concepts can stimulate a new way of thinking or highlight possible areas in which research is timely. We need to consider our own views about the purpose of qualitative research:

• Do we aim to stimulate thinking and understanding through qualitative research?
• Do we aim to increase our understanding of different experiences of living in the world?
• Do we aim to develop theories that underpin complex processes?

In all these cases, size may not matter. Indeed, we would want our syntheses to include some unique concepts that challenge our thinking. However, qualitative reviewers cannot ignore the issue of size; they face the same questions as primary researchers grappling with the questions – how many qualitative interviews is enough?261 A large review can appear, and may arguably be, more ‘adequate’ than a small review. If 40 studies support a concept, then this is in some way different from only five studies supporting it. If the aim is to stimulate ideas and contribute to a particular field of enquiry, then ‘adequacy’ may not be our only concern. If our aim is to develop robust gravitational concepts, we should consider both the economy and diseconomy of scale. For our purposes, we have gained confidence from including a large body of qualitative studies. We cannot know whether or not these findings would be qualitatively different had we included fewer studies. For the purposes of determining confidence of review findings for GRADE-CERQual reviewers, it is useful to provide a tally of the concepts that support each finding so that the reader can make their own assessment about adequacy for purpose. The power of concepts to make us think, however, is not based on quantity of data included. However, qualitative health researchers live in a research culture that values numbers and economies of scale and there is a real danger that potential stakeholders continue to undervalue the contribution that qualitative research can make towards improving people’s experiences of health care.

Coherence

Coherence considers the consistency (and difference) across studies. We agree that it is important to demonstrate to readers that concepts have not been cherry-picked and that it is therefore important to present concepts that do not fit the analysis. However, we do not feel that inconsistency necessarily negates confidence in a particular finding. An insightful concept might appear in a single study or it may represent a point of view that is untapped in other studies. We agree that it is useful to provide information about consistency for the same reasons that it is important to provide information about adequacy; readers should know where the concepts come from. We think that for the purposes of determining confidence of review findings for GRADE-CERQual it may be useful to provide a tally of individual studies supporting the finding and, importantly, to present concepts that do not fit the conceptual analysis. However, as we have said:

• Qualitative research does not aim to be representative or true in a positivist sense, but aims to generate useful ideas.
• The tale of the Emperor’s New Clothes has taught us the validity of a small voice.

It is important to consider these issues when utilising tools such as ConQual259 and GRADE-CERQual37 as well as reporting guidelines, such as the imminent meta-ethnography reporting guidelines (eMERGe; www.stir.ac.uk/health-sciences-sport/research/groups/emerge/; accessed 22 January 2018). Although these tools can definitely encourage reviewers and readers to think, the issue of quality in qualitative research is complex and these tools cannot guarantee confidence in findings. Thus, access to a GRADE-CERQual chart in an appendix does not remove the need for intellectual scrutiny on the part of reviewer or reader. Indeed, the tools were not intended to encourage a tick-box exercise, but to allow the reviewers and readers to consider issues of quality and utility. We need to consider that readers working outside an interpretive methodological tradition might not be familiar with the complexity of quality in qualitative research. For example, although both ConQual259 and GRADE-CERQual37 advocate quality appraisal, we cannot ignore the fact that a significant number of reputable qualitative reviewers have pro-actively chosen not to appraise.27 This is not surprising in the light of the fact that there is no agreed means of determining quality. We need to consider the dangers of yielding to methodological monopolies and continue to challenge our ideas about what is good.

Credibility

Munn and colleagues259 propose an additional criterion, credibility, for establishing confidence in the findings of qualitative research syntheses. Credibility is a global evaluation of ‘fit’ (‘equivocal’, ‘unequivocal’ or ‘unsupported’) between the primary data and the reviewers’ interpretations as demonstrated by adequate exemplars. This resonates with concept–indicator fit.262 We feel that concept–indicator fit underpins good-quality qualitative research and would therefore consider it fundamental to a quality assessment.36 However, deciding whether or not the author has provided ‘an illustration that is beyond reasonable doubt’36 is the reader’s personal judgement.

Conceptual insight

Conceptual insight is fundamental to meta-ethnographic aims and, therefore, primary authors need to provide an adequate and clear description of their concept. In addition to clarity, a concept should allow reviewers to think with it. Neither conceptual clarity nor strength is considered in either GRADE-CERQual or ConQual. There were 18 studies11,150,155,160,162,169,173,178,181,187,188,192,196,197,202,206,208,219 that provided inadequate conceptualisation in at least one finding (see Appendix 2) and one study188 from which we felt unable to include a single concept into our analysis.

Overall confidence in review

This was a large QES and we rated our confidence as high when more than half of the reviews (n ≥ 39) supported the finding. However, there is no agreed way to determine how many data provide sufficient gravity to support the validity of a concept. One concept may be weightier than 10 other concepts. Attempts to rate overall confidence raise important issues for reviewers. Importantly, are we in danger of slipping into a positivist fallacy that qualitative research can help us to find ‘the’ right answer. Qualitative research encourages us to think outside our own box and develop a questioning approach that can be utilised in policy and practice decision-making. It cannot provide the answer, but can provide concepts to think with and encourage us to see things from a different vantage point and to challenge our thinking and practice.