We present our findings in a thematic structure that represents clinician reflections on navigating towards a diagnosis of endometriosis. We show how this process is complex and associated with nuanced considerations about diagnostic reasoning and processes. Finally, we show that these considerations and processes are not always emotionally inert for healthcare professionals.
The value(s) of diagnosis
Conceptually, clinicians regarded disease-based diagnoses as valuable in clinical practice. A major component of this was their recognition of the potential importance and benefits of an endometriosis diagnosis for patients. These included access to specialist care, social and occupational validation of symptoms, membership of support communities, and accessing information and support.
The advantages for patients of having a diagnosis to guide and inform journeys through evidence-based care were also experienced as beneficial for clinicians, enabling them to be more effective at offering evidence-based treatment and managing symptoms. These were conceptualised as shared benefits experienced in clinician–patient relationships and consultations:
‘[An] advantage in the diagnosis would be, you’d be probably more effective at managing their symptoms, you’d probably help them with the psychosocial component of their illness because [of] the validation stuff that happens, and also being able to provide targeted treatment appropriately. As opposed to the managing the unknown type symptom, and the advantage also is allowing the patient to make an informed decision.’ GP32
Diagnoses, once attained, could be embedded into a biomedical model of care, where they functioned as a mechanism that could make clinicians’ jobs more straightforward. This included knowing which evidence-based treatments to offer, accessing specialist services and treatments (for example, gonadotrophin-releasing hormone [GnRH] analogue therapy or surgery), informing shared decision making, and anticipating and responding to potential future health needs, including those related to pain or fertility.
As one GP explained, managing future presentations without a diagnosis was akin to working ‘in the dark’ (GP7).
Some GPs reflected that having a diagnosis helped them to feel more ‘confident’ (GP14) or ‘comfortable’ (GP31). Seeking diagnosis was, for some, conceptualised as an intrinsic part of their role as a doctor:
‘[T]he gut reaction is you, you’ve got to have a diagnosis, but that’s probably just my medical model head. Having been firmly trained for the last […] well thirty years, so I’d be uncomfortable not having a diagnosis, or at least not even trying to have a diagnosis. I think early stages you’d want to try and have a diagnosis, but you do it, you know, in the expectation that endometriosis is not an easy thing to diagnose without being some, you know, you can get up to the invasive end of the spectrum. Do you really want to do that? […] I’m sitting there as a doctor, I want a diagnosis. Because it makes it easier to try and manage, you know it makes my job easier if I’ve got a diagnosis. So, if we know for sure its diagnosis, then we also know that it’s not, you know, anything more sinister.’ GP31
Diagnoses could confer permission to recognise the limits of what GPs could (or should) do, and allow them to seek advice or pass care on to another team. The contribution of protection from potential personal risk, for example, in avoiding complaints or litigation, could also be an important contributory driver towards seeking a diagnosis, although this could also be complex and associated with uncertainty:
‘… at the end of it, five years down the line saying, “I was worried about my fertility and you did nothing about it.” … then why make it even more stressful and say, “You know, you might be infertile as a result.” And then if they, say, find out it’s not endometriosis, then you’ve got a guaranteed complaint coming your way saying, “You freaked me out unnecessarily.” You know, so. You know I’ve seen it too many times, so no, I’m not going to bring that up.’ GP2
‘I guess that probably reflects a little bit of my own lack of confidence in the early management of endometriosis. I suppose we want them to get that concrete diagnosis from the specialists, and then you can then sort out the management, but yeah, I’d involve them probably quite early on […] And then the medicolegal defensive part of me thinks well, in five years’ time I don’t want a solicitor’s letter pitch up saying, “You delayed referral.” GP15
Diagnoses could make the clinician’s job ‘nicer’ and ‘easier’ (GP22) because they knew what they were ‘dealing’ with (GP7, GP26).
Having a disease-based diagnosis was contrasted against the challenge of managing patients who had undergone specialist assessment and testing, with no resulting unifying diagnosis but with impactful and ongoing symptoms. Chronic pelvic pain was not always conceptualised as a diagnosis, with the implication that an absence of identified pathology represents diagnostic uncertainty, rather than a symptom-based diagnosis of equal parity. Supporting people with pain and no unifying or identified diagnosis could be difficult or challenging; this was potentially compounded when support and management for undiagnosed symptoms fell to primary care clinicians when patients were discharged from secondary care:
‘Often you’re either not finding anything or what you’re finding isn’t necessarily something that patients are accepting of what you can do about it. I think pelvic pain we do often find quite challenging, and then even when you’ve got the diagnosis, you’re left with the pain, aren’t you, and that can be obviously very difficult to manage, and pain in general’s difficult to manage often.’ PC9, GP
The drive for a diagnosis meant that patients could ‘kind of fall between people’ (GP30) or ‘end up shopping for [a] diagnosis’ (PC18, GP); this could be difficult for both the patient and the clinician:
‘The expectation from the patient is often quite high, the importance that they place on getting to the bottom of this and getting it fixed is quite high, and together with the mismatch that a large proportion of chronic pelvic pain we don’t ever find a cause for can be really difficult.’ PC8, GP
Complexities arising when diagnosis (rather than treatment) is an aim of care
Although they valued the role of a disease-based diagnosis, clinicians held a parallel awareness of the need to offer care to those without a diagnosis. Their accounts depict how they work to ensure that patients without diagnoses are not disadvantaged. This reflected the context of general practice, where clinicians continue to hold care and responsibility whatever the outcome of referrals or diagnostic tests.
Clinicians explained how they actively worked to manage patient expectations throughout diagnostic journeys and provide ongoing support, whatever the outcome. The core role of primary care was conceptualised as offering holistic support over time, throughout patient journeys and through healthcare services. This role was, ideally, embedded in ongoing relationships, and characterised by rapport and trust so that the clinician kept the patient ‘onboard’ (PC8 GP), which they actively sought to achieve and sustain:
‘You know the discussion of yes, you may find an obvious cause and diagnosis for your pain; however, there’s [a] significant amount of women with cyclical pelvic pain, but you don’t find an absolute answer for it. It’s about trying to treat the symptoms. So, I’d try and be honest and say, “Yes I agree it would be nice to get a diagnosis, but prepare yourself that we might not.” In which case, we have to think about how we go about, deal[ing] with that in terms of symptom control.’ GP39
As well as having uncertain outcomes, clinicians knew that diagnostic and referral pathways took time, describing increasing waiting lists for scans and gynaecology appointments. This meant that, where possible, they wanted to start symptomatic treatment at the same time as referring. While knowing that this might not be possible for fertility concerns, they knew it could be helpful for pain and bleeding.
Whether patients had a diagnosis or not, clinicians recognised that those with ongoing symptoms would likely be returning to primary care for treatment and support. They were aware that patients could rotate around the system, and that this could adversely impact on experiences and diagnostic delays:
‘Pelvic pain is vague because you obviously explore the menstrual sort of history, sexual history and all sorts, and it doesn’t really sort of yield anything at all, so you end up having to do blood tests and scans, and everything comes back as normal, then you end up referring them on to a secondary care [specialist] who say[s] they’re not interested because there’s nothing structurally wrong with the scans or there’s nothing wrong in terms of blood, so the patient’s left in limbo, which is quite a difficult thing to sort of manage because the gynaecologist will say, “Well, actually it’s not gynae”, but then they won’t suggest what it could be, and you then go down the route of maybe a gastroenterological sort of opinion and they say, “No, it’s not that”, and then the patient’s just left frustrated, and then you feel a bit uncomfortable [about] how to manage those.’ PC18, GP
Knowing who to refer, and when, were pervasive concerns balanced against structural limitations, including pressure to not refer and an awareness of over-burdened secondary care services. The sense that ‘can’t refer everyone’ (GP15) could be uncomfortable for clinicians, who had to navigate a precarious balance between over- and underreferral — and therefore, by implication, diagnosis:
‘There’s a lot of uncertainty and a sort of and, um, you know a lot of difficulty round, um, you know predicting how it’s going to go in the future for patients, um, cos obviously some of them can be referred so then it’s, you know, the specialist manages it, it’s easier in some ways, cos it’s shared isn’t it, the, er, I mean I, myself, I would rather refer everyone because I think it’s quite complicated. But I know with all the, um, you know the pressures of not referring …’ GP34
These broader considerations were amplified by the uncertainties and unknowns about endometriosis as a condition. Clinicians knew that a diagnosis of endometriosis might not change first-line medical management, and that the first-line treatments upon diagnosis are often the same as empirical treatments trialled before referral. This could jar with the perspective that tests should only be undertaken if they would — or likely would — significantly change management. However, clinicians were aware of specialist treatments available, such as GnRH analogues or surgical treatment, which would (or could) not normally be instigated in primary care. This meant that, for patients with ongoing symptoms, referral for these treatments (and/or diagnosis) was clear cut and important.
Clinicians were aware that disease-based diagnoses could facilitate identifying and responding to emerging symptoms, priorities, or concerns. However, they were also aware that there might not be interventions that could reduce or mitigate the likelihood of potential future complications or impacts:
‘I’m not aware of any of the evidence, but I don’t know whether there [is] any prognostic value to diagnosing endometriosis early and being referred for laparoscopic diathermies and things, that they might do to get rid of it.’ GP8
Finally, clinicians reflected that the heterogeneity of endometriosis made it difficult to predict future impacts or needs. Many reflected on the difficulty, even with a confirmed diagnosis, of offering ‘good advice, apart from platitudes in terms of prognosis’ (GP27) because of a lack of evidence about interventions and outcomes. For patients living with chronic pain or seeking referral for difficulty conceiving, the decision to refer for specialist care was clear cut. But, knowing how to advise other patients about potential future impacts, including where endometriosis was identified incidentally, was a real concern. The unpredictability of whether a diagnosis would change treatment or enable effective interventions that could reduce future health impacts, and limitations in being able to meaningfully prognosticate for patients, were balanced against the potential risks of the diagnostic process, notably laparoscopy. This could cause equivocation about the value of diagnosis:
‘Having a diagnosis of endometriosis is on top of my brain, but what difference is that going to make for her if the treatment I’m doing is helping her with the symptoms? It might give you the satisfaction that, “Okay I, I do know that I have endometriosis”, but if that is being treated whether you’ve been branded with it or without the brand how much I honestly do not know.’ GP41
The consequence of these uncertainties was that decisions to refer could be complex, comprising potential tensions and uncertainties (Figure 1):
‘I think what’s difficult, cos you have got to think diagnosis is a thing out there, and it’s a scientific thing, and you can peg everything on it, but it isn’t really. You know it’s a lot more woolly than that. That’s the problem we’ve got.’ GP34
