‘Delicate diagnosis’: avoiding harms in difficult, disputed, and desired diagnoses

Margaret McCartney is an Honorary Senior Lecturer at the School of Medicine, University of St Andrews.

Natalie Armstrong is a Professor of Healthcare Improvement Research at the Department of Health Sciences, University of Leicester.

Graham Martin is Director of Research at The Healthcare Improvement Studies Institute, University of Cambridge.

David Nunan is a Senior Research Fellow at the Nuffield Department of Primary Care Health Sciences, University of Oxford.

Owen Richards is Chair of the Patient and Carer Partnership Group, Royal College of General Practitioners.

Frank Sullivan is a Professor of Primary Care Medicine at the School of Medicine, University of St Andrews.

Medical diagnoses can be subject to ambiguity, flux, subjectivity, and inherent uncertainty. This is particularly true in primary care, where many reported symptoms do not conform to a clear diagnosis. Thresholds around normality are often unclear.

Symptoms are commonly experienced and described as an ’iceberg’. Over a third of otherwise well people without a chronic condition have felt tired or run down, or had a headache in the previous 2 weeks, and over a quarter have had back or joint pain.1 Distinguishing conditions that would benefit from diagnosis and earlier intervention from those that are temporary, self-limiting, and prone to harmful medicalisation, remains challenging.

While overdiagnosis is a diagnosis of a condition that, if unrecognised, would not cause symptoms or harm, contested diagnoses are symptoms given a diagnosis, but where the diagnosis, as an explanation for them, is controversial. Opponents consider contested diagnoses ‘wrong’ not because of an error in history taking, clinical examination, or interpretation of a test result (which would risk misdiagnosis), but because the diagnosis itself, for example, chronic Lyme disease or multiple chemical sensitivity, is invalid.2,3

Contested diagnosis is bordered by ‘pseudoscience’, for example, adrenal fatigue, leaky gut syndrome, and chronic candidiasis, which attempt to explain symptoms with ‘science’ that is demonstrably incorrect (Box 1).4

Box 1. Type of low benefit or harmful diagnosis to individuals
Definition Examples Potential benefits of diagnosis Hazards of diagnosis
Overdiagnosis Diagnosis of a condition that, if unrecognised, would not cause symptoms or harm a patient during his or her lifetime – Ductal carcinoma in situ;

– screening for thyroid cancer; and

– pre-diabetes

– Minority may benefit through earlier diagnosis – Majority do not benefit from diagnosis;

– may be harmed via side effects of low value interventions; and

– burden of patienthood

disputed diagnosis
‘Characterised by a lack of known biological cause or abnormality, and uncertain or ill-defined treatment paradigms … often result in a disputed disease status for patients’5 – Gulf War syndrome;

– chronic fatigue syndrome;

– multiple chemical sensitivities; and

– fibromyalgia

– Validation/
recognition;– community support; and– trust in healthcare professionals
– Diagnostic overshadowing;

– ineffective/
harmful medical intervention;

– over-
medicalisation; and

– burden of patienthood

Misdiagnosis Incorrect diagnosis of a condition – Indigestion versus myocardial infarction;

– neurological conditions versus functional conditions; and

– placebo/
nocebo responses

– None/
negative; and– potential legal redress (but with trauma in the process)
– Late diagnosis potentially worsening prognosis;

– incorrect treatment; and

– loss of trust in professionals

Pseudoscience Demonstrably false explanation for symptoms – Leaky gut syndrome;

– Morgellens syndrome;

– total body candida; and

– adrenal fatigue

– Validation of symptom recognition; and

– placebo response

– Diagnostic overshadowing; and

– ineffective/
harmful medical intervention

Underdiagnosis People with recognised condition, where diagnosis leads to useful improved outcomes, are not diagnosed – Multiple sclerosis; and

– endometriosis

– Earlier effective treatment; and

– reduction of suffering

– Diagnostic inaccuracy potentially greater at earlier stage

Many contested diagnoses relate to symptoms that are not clear-cut, meaning that some people are given a diagnosis that does not confer benefit. Others are primarily unexplained ‘yet’. For example, Gulf War syndrome was treated as a contested illness but is now recognised as caused by exposure to sarin gas.6 Other recognised conditions are prone to a variety of influences such that diagnosis risks causing harm, through a variable mixture of misdiagnosis, overdiagnosis, pseudoscience, or overtreatment.

A framework for considering some conditions as ‘delicate diagnoses’ may help to recognise and reduce the harms of medicine in these circumstances.

Delicate diagnosis

A variety of models have been used to describe and explain how diagnoses are reached by doctors. Heuristics — rules-of-thumb — discounts less important information to pragmatically reach likely probabilities, while responding to environmental circumstances. Bayes’ rule means that small pieces of information can have a profound impact, with prior probabilities of disease rapidly shaping possibilities.

GPs work mainly from ‘mindlines’,7 ‘collectively reinforced, internalised, tacit guidelines … informed by brief reading, but mainly by their interactions with each other and with opinion leaders, patients, and pharmaceutical representatives’. Similar heuristic styles may be true for patients. This gives rise to potential for bias, and the associated harms of poor-value care, overdiagnosis, and overtreatment.

“A framework to consider some diagnoses as ‘delicate’ allows for systematic identification and analysis of the influences surrounding them.”

Finding good-quality information is often difficult. For example, searching for ‘Do I have ADHD?’ on Google brings up, in the first few hits, a trivia quiz site with an unvalidated questionnaire, a questionnaire hosted by a site sponsored by a manufacturer of drugs for ADHD, and only then, NHS information. The top hits for information about ‘cow’s milk protein allergy’ includes baby milk manufacturers and patient organisations sponsored by them. Campaigns regarding prostate cancer have been run by men’s organisations and consultants in private practice.

Nor have GPs in the UK been able to approach certain diagnoses neutrally. They have been previously financially incentivised under their contract to seek to make diagnoses ‘early’ that they would not otherwise have considered clinically useful.8 Despite continuity of care being desired by patients,9 it is in decline.10 Patients seek internet advice for ease,11 but doctors for professional skill — yet this may not be readily available given the stress in primary care.

In 2005, the Health Select Committee (HSC) review of the pharmaceutical industry argued that ‘The Department of Health has not only to promote the interests of the pharmaceutical industry but also the health of the public and the effectiveness of the NHS … The Secretary of State for Health cannot serve two masters’.12

However, in 2013, the government launched Academic Health Science Networks (AHSN), seeking to increase the speed of the adoption of innovations within the NHS and ‘an entirely new relationship with industry based on partnership, not just transactions’.13 These delivered a programme — screening for atrial fibrillation — that the UK’s own independent committee on screening has recommended against.

In this context, full of competing interests, challenges in finding definitive evidence, and with gaps in NHS care, some diagnoses, and the circumstances in which they are likely to occur, may be considered ‘delicate’, meriting special care to navigate them.

Features (Box 2), many overlapping, include fuzzy margins, unsettled science, and cultural differences, together with subjectivity of symptoms and risks of transactional elements (for example, payment to access a sought-after condition and associated sequelae).

Box 2. Features suggesting a ‘delicate diagnosis’ 
Diagnosis is subject to Why this creates ‘delicacy’ Pressures exerted via Examples
Fuzzy margins No clear demarcation between normal/abnormal Guideline committees, industry, and awareness campaigns Pre-diabetes, pre-hypertension, depression, and anxiety
Boundary fault line May be social rather than medical phenomena Cultural influences including the media/internet, lobby groups, and private practice Gender dysphoria in young people, ADHD, reward deficiency syndrome, and intermittent explosive disorder
Conflicts of interest Increased diagnosis/treatment rates promoted by those who also benefit from it Industry via professionals, lobby/patient groups, and APPGs HRT pre-menopause, testosterone use in HRT, and prostate cancer screening
Unsettled science Controversy in mainstream medical literature about fundamentals of diagnosis/intervention Industry, individual healthcare professionals, guidelines, and research publications Screening for atrial fibrillation, treatment for gender dysphoria in young people, and long Covid
Private practice Single issue clinic predominance where holistic assessment may not be prioritised Individuals or groups working in private practice PSA screening, health checks, and ADHD diagnosis in adults
Cultural differences Markedly different rates of diagnosis in different settings and between patients and clinicians Governments, religious beliefs, and historical/sociological factors Heavy legs (France) and transsexuality (Iran)
Subjective reliance Little or no ability to externally validate findings Lobby groups, industry, media, and universities Chronic Lyme disease, long Covid, chronic pain, and chronic fatigue
Overdiagnosis and overtreatment risk Indications for valid diagnosis and treatments is expanded beyond benefit Government policy, industry, and online interventions ‘Low T’, HRT, ADHD, and treatment for thyroid/breast cancer found at screening
Transactional element Entry point of testing dissociated from joined up healthcare system Single issue private services, elements of GP contract, and ‘tick box’ diagnosis PSA screening in sports clubs and private screening blood tests
Single issue lobby groups/pressure groups/campaigns Awareness raising/exclusive concern Industry via media and lobby groups Sepsis and dementia screening
Multiple possible symptoms Wide diagnostic capture with risk of wrong and overdiagnosis Industry and media Cow’s milk protein allergy, menopause, and low T
Screening outwith UK NSC criteria14 Unlikely to be cost-effective or high value, and harms to patient Industry via professionals, lobby/patient groups, and APPGs PSA and atrial fibrillation screening
Pseudoscience May be desired diagnosis with high value to patient and pressure on clinician to agree social media, influencers in social media, and private clinics electromagnetic hypersensitivity, Morgellens syndrome, and chronic candidiasis
ADHD = attention deficit hyperactivity disorder. APPG = All-Party Parliamentary Group. HRT = hormone replacement therapy. Low T = low testosterone. NSC = National Screening Committee. PSA = prostatic specific antigen.

These features — not intended to be exhaustive or definitive — may be especially important when considering diagnosis without the context of experienced heuristics or multiple means of sense checking via mindlines — or when these are systematically biased or based on poor information. The combination of sensitivity of area, expectations, and the difficulty of achieving certainty make them particularly prone to influence, including seeking diagnosis, and the threshold at which it is made.

A shared understanding between diagnosticians, citizens, and patients

It may not be possible, or desirable, to make ‘delicate’ diagnoses more robust. Instead, a tolerance of uncertainty, and willingness to adopt a questioning, tentative, or provisional diagnosis may be appropriate. This has to be set against the potential for harm where there may be value in diagnosis and consequent treatment. However, urgent intervention in these cases is rarely needed, and rapid diagnosis has to be set against the harms of diagnostic overshadowing, low-value, and harmful medicine, and the burden of ‘patienthood’.

When the prevailing environment reflects a bias, work may be required to achieve and maintain a more neutral equipoise (Box 3). Computer coding systems may reflect a potential or uncertain diagnosis, and may require expanding to reflect uncertainty. Nor should the value of diagnosis be underplayed: the ability to name a disorder gives the ability to develop robust definitions, fairly test, and develop effective interventions.

Box 3. Challenges of a ‘delicate diagnosis’
Type Why Features Strategies
Difficult – Subjective, fluctuating symptoms may be variations of normal or common to multiple other conditions; and

– diagnosis itself may be sensitive

– Online questionnaires;

– poor quality evidence for interventions;

– lack of control groups;

– low pretest probabilities; and

– other more likely explanation

– Use of time as a diagnostic tool;

– express chances of benefiting from diagnosis in absolute (not relative) terms;

– honesty about the risk of inaccuracy and hazards of diagnosis;

– use of/generate data about control groups; and

– computer coding that recognises uncertainty

Disputed – No broad medical consensus or strong challenge to diagnostic validity exists – Disputes may arise between: lobby groups; professionals; industry; guideline authors; activists; regulators; private sector; insurance companies; and the NHS and its staff – Clear statements of conflicts of interest and mitigation strategies, including what conflicts are not manageable and should lead to exclusion;

– agreed, official ‘cross party’ explanations recommending independent advice; and

– efforts to achieve a multiparty consensus

Desired – Single issue pressure groups/clinics seeking increased diagnosis rates, awareness campaigns, contractual targets, and media campaigns;

– patient preference for rapid diagnosis; and

– high expectations of value of diagnosis and consequent intervention

– Strong media presence;

– industry influence via third sector;

– charitable or pressure organisations;

– few/no downsides presented;

– claims of underdiagnosis; and

– social pressure put on patients/citizens

– Seek independent information that balances benefits and risks of diagnosis;

– consider that ‘single issue’ diagnostic routes create a bias within themselves; and

– advocate for holistic assessment considering wider possibilities for cause of symptoms

The challenge for clinicians and patients is to find benefits while avoiding harms. This may not be easy for either party, especially given the attractiveness of certainty and privileges associated with a particular diagnosis. Minimising harm and maximising benefit requires partnership between doctor and patient, especially in an environment replete with vested interests.


As medicine shifts in the era of COVID-19, new challenges are likely to impact on how and why doctors and patients seek diagnosis and the culture in which they do it. Continuity and relationship-based care are valued but are becoming less common. A panoply of environmental pressures, some visible and some not, are prevalent. New and emerging information can be difficult to critically appraise and use, often because of biases and misinformation. A framework to consider some diagnoses as ‘delicate’ allows for systematic identification and analysis of the influences surrounding them. This may help organise appropriate mitigations against bias, help patients and doctors avoid harm, and inform research and policy.

Key points
– Some diagnoses are particularly prone to ambiguity, overdiagnosis, overtreatment, and associated harm.

– Systematic factors include bias, conflicts of interest, and variability in information quality for patients and clinicians.

– Considering certain diagnoses as ‘delicate’ allows identification and potential mitigation of these features.

– Tolerance of uncertainty and willingness to adopt tentative or provisional diagnosis may be an appropriate means of balancing risks.

Competing interests
Margaret McCartney has written about overdiagnosis and overtreatment in the past and some has been paid work. Full declarations of interest can be found at:

Margaret McCartney is an NHS Research Scotland clinical academic research fellow. Graham Martin is based in The Healthcare Improvement Studies Institute (THIS Institute), University of Cambridge. THIS Institute is supported by the Health Foundation, an independent charity committed to bringing about better health and health care for people in the UK.

1. McAteer A, Elliott AM, Hannaford PC. Ascertaining the size of the symptom iceberg in a UK-wide community-based survey. Br J Gen Pract 2011; DOI:
2. Lantos PM, Wormser GP. Chronic coinfections in patients diagnosed with chronic lyme disease: a systematic review. Am J Med 2014; 127(11): 1105–1110.
3. Das-Munshi J, Rubin GJ, Wessely S. Multiple chemical sensitivities: a systematic review of provocation studies. J Allergy Clin Immunol 2006; 118(6): 1257–1264.
4. Shapiro ED, Baker PJ, Wormser GP. False and misleading information about Lyme disease. Am J Med 2017; 130(7): 771–772.
5. Rebman AW, Aucott JN, Weinstein ER, et al. Living in limbo: contested narratives of patients with chronic symptoms following Lyme disease. Qual Health Res 2017; 27(4): 534–546.
6. Haley RW, Kramer G, Xiao J, et al. Evaluation of a gene–environment interaction of PON1 and low-level nerve agent exposure with Gulf War illness: a prevalence case–control study drawn from the U.S. Military Health Survey’s national population sample. Environ Health Perspect 2022; 130(5): 57001.
7. Gabbay J, le May A. Evidence based guidelines or collectively constructed “mindlines?” Ethnographic study of knowledge management in primary care. BMJ 2004; 329(7473): 1013.
8. Brunet MD, McCartney M, Heath I, et al. There is no evidence base for proposed dementia screening. BMJ 2012; 345: e8588.
9. Aboulghate A, Abel G, Elliott MN, et al. Do English patients want continuity of care, and do they receive it? Br J Gen Pract 2012; DOI:
10. Tammes P, Morris RW, Murphy M, Salisbury C. Is continuity of primary care declining in England? Practice-level longitudinal study from 2012 to 2017. Br J Gen Pract 2021; DOI:
11. Clarke MA, Moore JL, Steege LM, et al. Health information needs, sources, and barriers of primary care patients to achieve patient-centered care: a literature review. Health Informatics J 2016; 22(4): 992–1016.
12. House of Commons Health Committee. The influence of the pharmaceutical industry. Fourth report of session 2004–05. Volume 1. 2005. (accessed 22 Sep 2022).
13. Department of Health. Innovation, health and wealth: accelerating adoption and diffusion in the NHS. 2011. (accessed 21 Sep 2022).
14. UK National Screening Committee. Adult screening programme: atrial fibrillation. 2019. (accessed 21 Sep 2022).

Featured photo by visuals on Unsplash.


Notify of

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Newest Most Voted
Inline Feedbacks
View all comments

Fascinating article, thank you. I think that as trust in the medical profession dwindles, due to factors like the loss of continuity of care, the rise in use of social media with the associated echo-chambers and special interest groups, and the rate at which knowledge availability outstrips wisdom and experience, then the growth of delicate and disputed diagnoses will continue. The medical profession are also to blame with the use of more sensitive but less specific investigations and more – as you note.

Other than reinstating continuity of care with a trusted general practice team, I cannot think of any intervention that is likely to slow down this trend. It’s only in a trusting relationship that a ‘delicate diagnosis’, ‘time as a diagnostic tool’ and ambiguity and uncertainty can be sustained. Most of what GPs do is complex-holding work, and this is where patients with these problems should be

Previous Story

Burnout, patient and physician safety

Next Story

Book review: Digital development: stories of hope from health and social development

Latest from BJGP Long Read

Would love your thoughts, please comment.x