Untimely illness: When diagnosis does not match age‐related expectations

Abstract Background We explore the concept of “untimely diagnosis,” where the onset of a long‐term condition occurs at a life stage which does not conform to traditional expectations, focusing on two conditions (asthma and arthritis) typically associated with a particular life stage (childhood and older adulthood, respectively). Previous literature has focused on the meaning of chronic illness in terms of life history, and the biographical lens has been used in various ways to make sense of the experience. Less attention has been paid to the condition onset when it seems dissonant with chronological age. Methods Secondary analysis of two qualitative data sets (total 58 interviews) exploring the experiences of people with adult‐onset asthma and young people diagnosed with arthritis. Data from the original interview transcripts relating to diagnosis and symptom recognition were re‐analysed using a “candidacy” framework to examine how age and diagnosis intersect. Results People did not always assert their candidacy for either condition because of pre‐conceived expectations around age. Similarly, health professionals sometimes failed to recognize patients’ candidacy, instead pursuing “age‐plausible” possibilities. In some cases, participants were proactive in suggesting a diagnosis to the health professional where diagnosis was delayed. Conclusion The diagnosis of adult‐onset asthma, and arthritis in young people, may be regarded as “untimely.” We suggest that being diagnosed with what is perceived to be a “childhood” condition in adulthood, or “an older person's” condition in childhood, may be viewed as a “biographical paradox” and an “untimely breach” to the expected order.


| INTRODUCTION
There has been a growing interest in the sociology of diagnosis in recent years. Here, we focus on narratives of asthma in adulthood, and arthritis in young people, to explore the idea of "untimely diagnosis," where a diagnosis occurs at a life stage which does not conform to individual or societal expectations of that condition.
Diagnosis is increasingly recognized as a process involving a series of interactions with health-care systems and staff, rather than strictly as a "diagnostic moment" at which the health professional communicates a medical label to a patient. 1,2 The individual and their family also play a key role in puzzling out a diagnosis, seeking to make sense of symptoms and clues from health professionals' behaviour, and drawing on their common sense stock of knowledge. 3 Of relevance here is the concept of "candidacy," redefined by Dixon-Woods et al's 5 critical synthesis as "the ways in which people's eligibility for medical attention and intervention is jointly negotiated between individuals and health services." The focus of the critical synthesis was access to health care by vulnerable groups, and the authors found that candidacy is a continually negotiated property of people, subject to various influences including social contexts, aspects of self, situated activity and resource allocation. The core strands of the new theoretical conceptualization of access to health care are as follows: the identification of candidacy, navigation (of services and support), the permeability of services (the ease with which people can use them), appearance, adjudication, resistance and local conditions. Three factors are identified that contribute to the downgrading of warning signs of illness for vulnerable groups: the lack of a positive conceptualization of health; the normalization of symptoms within deprived communities; and the fear of being blamed by health professionals.

This work has been extended by various studies including
MacDonald et al 6 who found hazy and indistinct boundaries between these components of candidacy. Their paper further highlights how age and gender can compromise candidacy and the authors conclude "It is important that we do not underestimate the ways in which our shared understandings of the signs of illness influences patients' experiences of care, and importantly, access to care" (p. 109).
Davison et al, 7 who introduced the notion of candidacy, remind us that we "tend not to invent completely fresh explanations [for the misfortunes which befall ourselves or other people]; rather we employ the knowledge and lore which we have received from the wider society during our formation and development as individuals" (p. 5).
In this study, we explore how people experience and make sense of what both they themselves, and health professionals with whom they interact, may perceive as an "untimely diagnosis." We use the candidacy framework to consider how knowledge and lore may mask a "curveball" such as an untimely diagnosis and disrupt typical help-seeking behaviour. We explore what may happen when possible diagnoses seem to defy the "common sense" stock of knowledge by occurring at a time of life which does not "fit," using adult-onset asthma and arthritis in young people as exemplars.
We considered the candidacy framework to be more appropriate than, for example, the theory of illness representations 8 which more commonly focuses on people's beliefs and expectations about their illness after diagnosis. However, Bishop and Converse's study 9 of how people invoke "prototyped conceptions" to make sense of a particular set of symptoms is an interesting use of illness representations relevant to this pre-diagnostic phase.

| The importance of age in making sense of chronic illness
A diagnosis of long-term illness is unlikely to be welcome at any stage of life but may still be understood as part of a "normal" life trajectory.
Following Bury's original exploration of chronic illness as biographical disruption, 10,11 it has been argued that it may instead be accepted as biographically anticipated "normal illness," and has been described using terms such as "biographical flow," "biographical anticipation" or "biographical continuity". 11 The meaning of chronic illness to people has been found to be dependent partly on their age. Thus, contrasting the perspectives of young people and older adults provides a particularly good exemplar to illustrate the importance of age as a mediating factor in the experience of chronic illness.
Sanders, Donovan and Dieppe 12 for example, found that older people interpreted the pain and impaired mobility of osteoarthritis as a normal part of ageing, consistent with their expected biography rather than disruptive of it (even if at the same time it disrupted practical daily living). In explaining their findings, they drew on Bury's 13 distinction between "meanings as significance" (in this case, the significance-or not-of the condition for one's sense of self) and "meaning as consequence" (in this case, the practical activity restriction and social disadvantage resulting from the condition).
The meaning of illness may also depend on comorbidity and wider social factors such as class, poverty and poor housing, as well as the type of illness and social perceptions of it. Faircloth et al 14 in a study of experiences of stroke amongst white, Hispanic and African Americans argue that their findings "suggest a biographical flow more than a biographical disruption to specific chronic illnesses once certain social indicators such as age, other health concerns and previous knowledge of the illness experience, are taken into account" (p. 242).
Pound et al, 15 in a study of older working class people in London, suggest they may have "lower expectations of health and may anticipate illness as inevitable in old age, or meet it with a greater sense of acceptance" (p. 502). The participants in this study described witnessing death at a relatively early age amongst family and friends, alongside experience of other hardships. In this context, Pound et al encourage us to pay attention to "the straightforward possibility offered by the interviewees themselves" that "chronic illness may be anticipated and experienced by some older people as normal" (p. 502).
Bury and Holme recognized this as a valuable qualification of the original idea, as they point out, most people operate a "social clock" that guides expectations of events. 16  When it comes to the experiences of young people, the relevance of "biographical disruption" cannot be taken for granted either. This is particularly true in cases of congenital illness, such as cystic fibrosis, where there is no prior period of wellness or perceived normality. 17 Similarly, the relevance of "biographical disruption" is also unclear when the experiences and meanings of chronic illness take into account that a major task of adolescent development, healthy or otherwise, is the development of one's identity or "self-concept". 18 In this context, chronic illness can be seen more as "biographical contingency" than as "biographical disruption", 19 with young people attempting to reconcile their pre-illness identity with developing an altered identity that incorporates chronic illness as an accepted component of life. 20 However, other researchers such as Grinyer 21 have suggested that adolescent development in fact exacerbates biographical disruption, given that at this stage, identity is particularly fragile and key developmental goals such as increased autonomy and independence could be compromised.
Age is thus clearly one major factor affecting the likelihood that a diagnosis is seen as normal or expected, and much of the literature has focused on examples where illness is congruent with age-related expectations. There has been less attention paid to situations where the onset of chronic illness is seen as dissonant with chronological age. For example, it is important to acknowledge the inherent difficulty young people with chronic conditions face in navigating the fundamental tension between the very concept of youth, typically portrayed as a time of health, and being diagnosed with a chronic illness at a young age. 22,23

| Chronic illness as an age-dissonant life event:
adult-onset asthma and young people with arthritis Some conditions are commonly associated in the popular imagination with particular life phases. This may include "childhood" infections such as chickenpox or measles; hypertension as a condition of older adults; type 1 diabetes as a condition of childhood or adolescence, contrasted with type 2 typically associated with late middle age; or dementia as a condition of old age. Popular representations in the media and online images for diseases such as these tend to reinforce these understandings. In all these cases, however, the condition may be diagnosed at a different and unexpected time of life. For example, Khanolkar et al 24 note increases in diagnoses of type 2 diabetes in adolescence particularly in ethnic minority populations. Similarly, Higginbottom 25 found that being diagnosed with hypertension at a young age was perceived to be stigmatizing for young people, with hypertension commonly assumed to be a condition that only older people experience as part of the ageing process.
In this study, we focus on two conditions that are closely associated with a particular time of life: asthma, which is commonly thought of as a disease of children and young people, and arthritis, which is typically represented as a disease of older adulthood. 26 In both cases, we explore the perspectives of those whose diagnosis does not match such age-related expectations Asthma is a chronic and potentially life-threatening disease characterized by recurrent attacks of breathlessness and wheezing which currently is estimated to affect 335 million people worldwide and is the most common chronic disease amongst children. 27 In children and young people, asthma is most often associated with allergies (atopic asthma). Asthma that comes on later in life is less obviously allergic and is more common in women and smokers. A number of risk factors that may be associated with the onset of asthma in adulthood have been recognized, including respiratory infections, environmental factors, hormones, obesity and stress. 28 The diagnosis of asthma in older adults can present something of a challenge for various reasons. Many older people, particularly with late-onset asthma, fail to identify breathing difficulties as asthma initially. 29 They, or their doctor, may first attribute symptoms to other causes such as bronchitis, respiratory tract infection, cardiac disease, ageing or lack of fitness. [30][31][32] Arthritis is an umbrella term covering joint pain and inflammation arising from a range of causes. In popular discourse, "arthritis" is often taken to mean osteoarthritis, and in turn considered to be age-related wear and tear of the cartilage in joints which results in painful rubbing of bone on bone; although recent research is revealing a more complex condition now considered to be a disease characterized by "tear, flare and repair". 33 In young people, however, arthritis is more likely to be a form of inflammatory arthritis, which is an autoimmune condition.
Young people with inflammatory arthritis live with chronic or recurrent pain and disability, particularly if there has been a delay in starting effective treatment, which can limit their ability to complete daily physical tasks and participate in school and social activities. 32,34,35 In over a third of young people, the disease remains active into adulthood requiring drug therapy. 36 The diagnosis depends on the demonstration of clinical signs, rather than laboratory tests, with no other defined diagnosis being evident. 37 In the UK, all patients both children and adults present with their symptoms to a General Practitioner (family doctor) in the first instance. For patients where an inflammatory arthritis is suspected, GPs should refer to a specialist rheumatologist to confirm the diagnosis and initiate disease-modifying treatment. Treatment is often complex and challenging and involves ongoing medications, monitoring, physical therapy and surgery in some cases. 34,35

| METHODS
We used secondary analysis methods to explore how people make sense of a diagnosis of a condition that may not "fit" with popular expectations, particularly in relation to age. Our secondary analysis involved two data sets from two separate primary studies. Both primary studies were qualitative, narrative interview studies, addressing the following research question: "what are the experiences and information and support needs of people with a (particular) health condition? 38 Both were conducted using the overarching methods and Both studies used a purposive maximum variation sampling approach 39 to include variation across types of experience (such as time since diagnosis and degree of disease severity/progression) and demographic variables (such as gender, age, ethnicity, socioeconomic group and region). A range of recruitment avenues were used including through GPs, specialist nurses and hospital clinics; voluntary support groups; media advertising; word of mouth and snowballing.
Our secondary analysis was based on a total of 58 interviews, including a subset of 18 people diagnosed with adult-onset asthma from the first primary study (Table 1); and 40 people diagnosed with arthritis in childhood or adolescence and 9 carers from the second primary study ( Table 2).
Interviews for each typically lasted between 1 and 2 hours. They were conducted in the participants' own home or elsewhere if they preferred and were video or audio-recorded. In both studies, interviews started with an open-ended invitation to tell the story of the individual's experience. Following this unstructured narrative, semistructured prompting was used to elaborate further. The topic guide in each case included many consistent areas (eg initial symptoms; path to diagnosis; referral and secondary care experiences; treatment; living with the condition; information and support needs) but also conditionspecific prompts derived from the literature and advice from an expert panel including patient representatives.
Consent was sought on the day for the initial interview; participants were later sent a verbatim transcript of their interview to review before final consent and copyright was agreed for publication of ex-

| Findings
Both conditions were typically seen as "belonging" to a certain age or stage of life. Stereotypical images of arthritis as a condition typically affecting older people were frequently invoked: As we go on to explore in the next section, participants' own sensemaking around early signs and symptoms were often mirrored in their interactions with health professionals when they attended medical consultations prior to diagnosis.

| Health professionals and candidacy
Age-related explanations also seemed to mediate interactions with health professionals, who tended to favour other age-plausible reasons for symptoms, such as "growing pains":

| Negotiating candidacy with health professionals
People with adult-onset asthma often presented with a persistent cough or recurring chest infections initially. Anita felt progressively more unwell over several months whilst other possible causes, such as the possibility of heart problems, were explored before asthma was considered. In fact, despite a family history of asthma, her doctor did not consider it as possibility until she herself suggested it:

| DISCUSSION AND CONCLUSION
We have analysed our findings using the concept of "candidacy" de- So whilst people might have asserted a generic claim of candidacy as an ill person, or on behalf of an ill child, they typically did not assert candidacy as a person with asthma or arthritis. Exceptionally, some people eventually put two and two together themselves and attempted to negotiate their candidacy more proactively by suggesting a diagnosis to the GP. However, this had mixed results. For example, Anita felt she successfully persuaded her GP to think of asthma as a possibility, whereas Kay's parents reported that the GP "just said no." Although the narratives specifically provide patient perspectives, these included their accounts and interpretations of how GPs responded to reported symptoms in similar ways. The narratives provide interesting illustrations of candidacy as a product of the interplay between patient, GP and the wider health system. 44 Just as individuals may be inclined to seek age-plausible explanations first, according to these accounts, so too did many GPs. Two of our co-authors, HS (a GP), and JM (a paediatrician) certainly recognized the picture that our patient narratives paint of such interactions, and the difficulties that can arise for GPs in recognizing and diagnosing these conditions.
Our findings suggest patients perceived GPs in some cases to be minimizing their reported symptoms as insignificant or un-concerning.
Even where symptoms were regarded as worthy of investigation, other diagnoses were often pursued first such as bronchitis, COPD or heart problems in the case of asthma, or muscle and joint trauma in arthritis. Occasionally, a psychological explanation such as attentionseeking or psychosomatic presentation was regarded as a more ageplausible explanation in young people than a physical cause. We found a certain resonance here with Evans et al 42  suggests there may be gaps in training in paediatric rheumatology, and sometimes young people are seen initially by clinicians who are not optimally trained or resourced to deliver best practice [46][47][48][49] As we have seen in some of the examples above, continuity of care may also affect how long it takes to reach a diagnosis. When a patient sees a different doctor each time they attend, there is a tendency for each doctor to start again with the thread of clinical reasoning, and overcoming the perception of age implausibility may take longer overall. Conversely, it is possible for a problem to remain unsolved and just be accepted as normal for that patient after various investigations have revealed no cause for a symptom, and no medications have helped. In this situation, a new doctor bringing fresh eyes and different expertise may be required to make the "untimely" diagnosis.
The diagnosis of both adult-onset asthma and arthritis in young people may be regarded as an untimely breach of the expected order of things. We have thrown light on the way that patients, and doctors, sometimes draw on age-related beliefs including "common sense" or "popular" discourses and repertoires as way to make sense, or otherwise, of puzzling symptoms, which can result in delays in diagnosis. We argue here that the experience of being diagnosed with what may be perceived to be a "childhood" condition in adulthood, or an older person's condition in childhood, might be viewed as a "biographical paradox" in that such a diagnosis contradicts common beliefs in relation to what might be seen as usual for a particular age group, and therefore may not be considered initially as an explanation.
Our paper contributes to the sociology of diagnosis, in that we pay particular attention to the diagnosis of chronic illness at a time of life that is seen as dissonant with chronological age. As our findings illustrate, this can have an impact on help-seeking behaviour, as well as health professionals' readiness to consider a diagnosis.

| Strengths and limitations of the study
The analysis presented here benefited from the knowledge and insights from the original study researchers, as well as clinicians who have experience of diagnosing both conditions. We have used two conditions as exemplars to explore where age-related expectations have an impact on recognition of symptoms and diagnosis, from the patient perspective, including their experiences of interactions with GPs leading to delays in diagnosis. Interviews with GPs about their experiences of "untimely diagnosis" in these, and potentially other similar conditions such as childhood cancer, would provide valuable additional insights. We have focussed our attention on age-related barriers, but acknowledge that there are a range of other factors that impact upon health-seeking behaviours, such as illness awareness and perceptions, and knowledge of the prevalence of these types of conditions.