Clinicians’ Insights

Debate 3: Clinician’s Insights

Denis Pereira Gray worked as a general practitioner and Chairman of Council and later President of the Royal College of General Practitioners. He is the only general practitioner ever elected Chairman of the Academy of Medical Royal Colleges.

Glenn Roberts is retired consultant psychiatrist working at the Devon Partnership NHS Trust rehabilitation service in Exeter.

Exploring Diagnosis: ‘What sort of social factors, in your opinion, might influence diagnosis of autism- or come to that any condition?’


It’s a broad and complex topic isn’t it…

As a starter we’re invited to consider, ‘What sort of social factors, in your opinion, might influence diagnosis of autism- or come to that any condition?’

I’ve no particular expertise in autism but I have wrestled with the issue of diagnosis over a professional lifetime in adult psychiatry, mostly working with people experiencing psychosis.



A discussion between a general practitioner and a psychiatrist will have a special flavour as we are in the only two branches of medicine where the patient’s feelings are central to the doctor’s agenda. Furthermore, we use pathology and x-ray investigations to make or support diagnoses less than colleagues in other specialties. I too have no special interest in autism and in addition I am now clinically retired, so I am writing from previous experience.



There are of course social factors present in the origins and arising of difficulties, aetiology, and disputes about social factors in causation does have major implications for some people who are caught in the tension over whether they have something wrong with them or if their presentation is because something has been done to them … but within diagnostic systems e.g. ICD 10, certainly in psychiatry there is for the most part no aetiological implication attached to diagnosis – they are patterns which can arise from various causes.

So … as I see it … diagnosis is a process and diagnoses are for the most part names, conclusions, often tentative, given in medicine to different forms of suffering, disability and dysfunction. Classificatory systems are then encyclopaedias of such named entities, ordered in some reasonable fashion, with associated principles, rules, for application and applying such diagnoses in practice.

All of which is amenable to social factors … in that it has all arisen from complex argument and discussion between people identified by society as having legitimacy in determining what should be the guiding concepts and appropriate names with which to describe human suffering and to discern one ‘type’ from another. So I’d see diagnosis as a social process and diagnoses as socially constructed … and although there’s a dominant view that determines what our institutionally endorsed descriptors are … there are always minority views that see it differently and thus have difficulty even accepting the names and concepts. Many internationally accepted psychiatric diagnoses are contested by those they are applied to and within mental health the ‘service user movement’ is most often characterised as an activist or protest movement. Lacking objective biological markers, most psychiatric diagnoses are a matter of judgement and some recipients do feel ‘judged’ … this may be very different in areas of human suffering more amenable to objective testing and evaluation.


Denis :

General practice is round terms about one third physical illness, one third mental/emotional problems, and one third social factors affecting health.

General factors in diagnoses

In general practice most diagnoses are to some extent provisional.

We too are conscious that a diagnosis is in some conditions can be stigmatising.



In mental health there are also different social processes at work between diagnosticians of different professional backgrounds and often rivalrous tribes … so doctors are often characterised, I’d say caricatured, as making … giving … single word diagnoses … ‘you have …’ whereas psychologists would typically offer a diagnostic formulation putting the conceptualisation of ‘type’ of suffering in a personal and biographical context with an assumption of describing how present difficulties arose from past problems and what are the variables at work that maintain it … to be fair, psychiatrists are taught the same … but these professional disputes and rivalries are another social dimension that set a context for how diagnoses play out in practice.

The (social) risk of single word diagnoses … ‘you have bipolar disorder’ is that they easily become conflated with identity … ‘you are a bipolar’ in contrast with a more socially informed formulation which upholds the fact that you are a person with a problem … of course this collapses and confuses again when thinking of personality disorder!

The diagnostic process is a social interaction … a conversation between people characterised as experts and recipients and thus the diagnoses offered or pronounced are subject to social variables occurring within and between them. This leads into considerations of how people come to understand and feel about what is said to them i.e. if they feel introduced to a useful way of understanding their problem that indexes help or if they feel labelled and dismissed. I’m aware of colleagues working in streamlined memory clinics for dementia care who feel very troubled by only being ‘allowed’ a few minutes to ‘give someone the diagnosis’ … this is a socially determined process with many forms of social implications. If the point of all this is to relieve human suffering, some diagnostic practices may add to it …

And then onto ‘having a diagnosis’ and the implications … there are many social factors at work … some see this as ‘at last I know what’s wrong with me’ and feel relief, others as being accused or condemned by mistaken authorities … a lot depends on the social implications of specific diagnoses and the social context into which a diagnosis is made … if there is support and understanding it may well play out very differently.

Your experience of the value of diagnosis may also be determined by whether or not you’re accepted by your own social context, family, friends, neighbours, workplace with ‘it’ and if you can find or are introduced to socially supportive peers … for example HIV, substance misuse and I guess autism. In a wider sense stigma, prejudice and discrimination are socially determined responses that may carry experiential implications more impactful than the diagnosed disorder itself. And we’re starting this conversation at the time of the London Marathon and the Royals promoting the value of talking about mental health problems to reduce stigma and enhance acceptance and understanding, which is a social interaction with mental health diagnoses at a national level.

I guess I’m also illustrating that social factors in diagnosis is a complex and interweaving topic that plays out at multiple levels, all of which have implications and can be helpful or hindering for the person with the problem and those close to them and are equally amenable to educational and other social intervention to enable such factors to be beneficial rather than additionally harmful … which for us then informs ‘social medicine’…



 Social factors

  1. The patient as a person

The pattern of consulting doctors is relevant. For example a man of, say 50,  who has not consulted for ten years and who then presents with a vague symptom needs to be taken very seriously.

  1. Family pattern

Our alternative name is ‘family doctor’ and ‘family physician’ is the commonest name for generalist doctors world-wide.

A high number of the conditions that GPs see are familial, so knowing the pattern of illness in a family is important and can greatly change diagnostic possibilities. There are two forms of familial influence-physical i.e. genetic and family culture.  Many common diseases run in families, so knowing this helps GPs suspect such conditions earlier and patients often point out the family pattern themselves.

Secondly, children learn patterns of living including patterns of consulting doctors from their parents, especially girls from their mothers.

A third family influence on diagnosis is the parent’s judgment of a problem in a child. The parent knows the child better than anyone and a worried parent is always significant. That does not mean the parent(s) are always right. Also, some parents are great worriers and a few present their children in effect as proxies for their own anxieties. The general rule in general practice is to take the parent’s opinion seriously and assume it is correct until there is evidence to the contrary.

  1. Locality/environment

The environment can give important social pointers to diagnosis.

For example, I once noticed on several home visits that a house in a poor part of Exeter had no garden and my patient, a little girl, played in a side road which never got any sun. Then I observed her legs were bending. So, two and two made four and I referred her with a diagnosis of scurvy, which was confirmed. I had never seen scurvy before.

  1. The patient-doctor relationship

The patient-doctor relationship is a fascinating part of general practice and has been continually studied since Balint’s (a Hungarian psychiatrist) book in 1957. It can be seen as a social phenomenon. Its relevance to diagnosis is integral, as a good relationship liberates patients to confide more, especially symptoms or problems which are embarrassing.   This makes it easier for general practitioners to understand more and so make better diagnoses or make them earlier.

  1. Socio-economic factors

Socio-economic factors, often shortened to ‘social class’  or in the words of an editorial Lancet this week “social rank” is a huge overriding factor in medicine (Lancet, 2017).

Essentially poorer people have more illness and die younger.  The difference is as much as 12 years of life between the most socially deprived districts and the least deprived. This is the backdrop to all medical services, in all countries.

For general practice, poorer people consult more often than average and so see GPs proportionally more. The opportunity for making diagnoses is therefore greater in poorer people.  A major recent advance is the addition of a numerical marker of social deprivation (from the Index of Multiple Deprivation) to GP computer systems, making it easier for GPs to see the level quickly and take it into account. General practice is the only branch of medicine which has been shown to be significantly associated with reducing health inequalities.

  1. Ageing of the population

An over-riding social feature of the UK population is that it is ageing progressively. The over 85s, the most vulnerable subgroup, has doubled in the last 25 years and will double again in the next 25. A big number of diseases like cancer (1,000 new diagnoses a day) dementia, degenerative conditions in many parts of the body, diabetes, and Parkinson’s Disease  all are age-related and increasingly  occur clustered together, so called multi-morbidity. Interestingly social factors apply to the diagnoses of many of these.

  1. Life style

The way people live greatly alters their chances of getting many diseases. GPs by seeing the same patients repeatedly over years (NHS average = 3 face-to face consultations with a GP each year, median duration of registration in St Leonard’s = 7.3 years) get to know the life style of many patients. Obesity is obvious and has doubled in prevalence in the Practice in 24 years. Smoking is the biggest single risk factor for health and is commonly discussed.

In the St Leonard’s Practice, 80% of people with type- 2 diabetes are now diagnosed before they report a single symptom of the disease. This can be achieved clinically as the commonest risk factors are known to the personal doctors. This disease is virtually diagnosed on social and family factors.



It looks as though you too see the overarching domain of ‘social factors and diagnosis’ as wide ranging and touching on many levels of consideration.

I would agree that there’s a common interest between general practice and psychiatry in being focused on ‘feelings’ – human experience … and not just biopathology, which gives us both an overarching social orientation. You mention we’re the only branches of medicine where that’s the case and I would have tended to agree with you, but I’ve spent the last month or so in the company of 3 surgeons … or at least their substantially autobiographical writings (refs below) and to my surprise find each of their ‘internationally bestselling’ books to be deeply insightful of the human predicament and of psychosocial dimensions of ‘illness, medicine and what matters in the end’ (to borrow the subtitle of Gawande’s ‘Being Mortal’). For me this is all the more surprising as 2 are neurosurgeons (Marsh, 2014, Kalanithi, 2016) whom one imagines relate to brains but not people and it’s perhaps notable that they are both pausing for thought at major life transitions, one on the companionable cusp of retirement and the other in the context of terminal illness and his own approaching death, the epilogue being written by his surviving spouse. The fact that these surgical expositions and memoires have achieved such popularity interests me and I wonder if they are contributors to a wider societal process that is hungry to rehumanise both illness and medical practice … perhaps part of a much belated counter swing critical of ‘mechanistic medicine’ in recognition that when it comes down to it … it’s all about people in relationships … illness and medicine, life and death are fundamentally social processes …. ?

So, against the broad sweep of these background considerations we’re asked to consider in particular, ‘what social factors might influence whether a diagnosis is given or not’ and my starting position is again in thinking that it’s all about what influences the social dynamics in the relationship between people involved in exchanging this ‘gift’.

The diagnostic process at its simplest may go as follow… ‘I, the doctor, conclude from your description and my understanding of your personal and family background, examination and test results that your difficulties conform to a recognisable pattern such that they can be named as a particular problem, and in response to you asking me, ‘what’s wrong with me Dr?’ I can tell you … I ‘name the silence’ (Hauerwas, 2004) and give you a diagnosis.’

And in my view there are social variables attached to each element and at work in the interactions between them.


Gawande, A. (2015) Being mortal

Hauerwas, S. (2004) Naming the silences

Kalanithi, P. (2016) When breath becomes air

Marsh, H. (2014) Do no Harm



Generally I think it’s clear that we are in broad agreement with the main issues and I appreciate the time you have taken and the length of your comment.

I still think that we are likely to be different from many doctors because GPs and psychiatrists are the only two branches of medicine where the patient’s emotions are central to the doctor’s agenda.

I’ll comment now not because I fundamentally disagree but because I think there are some aspects of what you have written which are somewhat different in general practice.


Whilst I share your view of the books that you refer to, I’m not sure one can judge all surgeons by excellent books by three of their leading writers. My experience is that people who get to the top of the profession usually have much broader views than the average, wherever they have come from.

My experience of listening to thousands of patients whom I had referred to surgeons over 38 years leads me to believe that many of the consultations were relatively mechanical.



The doctor as diagnostician

Diagnosticians are people too … and despite the hopes and intentions of their qualifying organisations different Dr’s go about this in different ways. Some may be ‘early diagnosers’ more prone to identify and name a problem early, others will want to only come to this later in the process … some are quite sure that diagnostic systems carry truths others more inclined to see them as descriptive shapes. As you mentioned Denis, GP often deals with early or ambiguous presentations and psychiatry has very broad concepts under which there’s a huge range of human experience … I suspect that although our disciplines have views on ‘how to make a diagnosis’ that practitioners differ in how they respond to uncertainty, ambiguity, conflicting or incomplete data. The same patient would have a different experience with different diagnosticians.

Over and under diagnosis may also co-vary with the implications for the diagnostician… it has sometimes been suggested that the low pick up of mental health problems in community services could co-vary with the preparedness of those services to then respond to what they’ve identified as needing help. It may be a lot easier to make a diagnosis for something you have a good response for.

In terms of psychiatric or psychosocial diagnoses there’s also a counter current arguing against diagnosing at all … i.e. resisting the medicalisation of everyday life and the avoidance of personal responsibility e.g.

Degrees of confidence in the diagnosticians findings

I think the phrase is that we ‘make a diagnosis’ … it may seem that we recognise a natural form and name it rather like bird spotting but actually we are constructing a meaning out of the information available and have varying degrees of diagnostic confidence at the end of that … and different diagnosticians will behave differently when faced with variable levels of confidence in the applicability of a name for a problem … some would prefer to keep it as near as possible to a description rather than jump the gap and ‘give’ a more conclusive response. And what is it we then give? Do we give a name, several names of possibilities or hand back a description gathered from the person of not only what has arisen but how it arose and why … this is the difference between a diagnosis, a differential diagnosis and a diagnostic formulation. In psychiatry we often talk of a ‘working diagnosis’ with a differential diagnosis kept in mind … i.e. ‘it looks a bit like this, so let’s respond (treat) accordingly, but keep an open eye for changes that suggest something different’. Patients often despair over the variability of psychiatric diagnoses and the more long term and experienced customers may reel off, with dismay and dark humour, the list of things they’ve been ‘accused’ of having. This can be a basis for consumer rejection of diagnosis at all but the reality is that conditions change over time, wax and wane … the diagnostic error may have been in ‘giving it’ too firmly in the first place.



Social dynamics

I very much agree with you that the process of making a diagnosis, or indeed providing medical care, is part of a social dynamic in which the patient and doctor are heavily influenced by social factors. I think much of this is missed in medical schools.

Individual variation of doctors

I also agree that all doctors in all branches of medicine vary much more than patients realise and that this is particularly true in general practice, because of the weighting needed for physical, psychological, and social factors. In other words I think GPs vary more than specialists.

As for medicalisation this is a hot subject and getting hotter! Ginny and I and one of our shared students did an article on this last year in the Journal of the Royal Society of Medicine!



Given to whom?

It would be natural to assume that the issue in play is what affects the Dr giving a diagnosis to a patient … but the social orbit around the doctor as diagnostician is more complex and there may be social factors that impact on who else Dr’s give or withhold diagnoses from. The list of players is possibly quite long including, the patient, carers, friends, the work place, other medical colleagues, the Dr’s employers, the patient’s employers, commissioners, governmental agencies. If the patient is severely learning disabled, very ill or otherwise incapable of ‘receiving’ the diagnosis it may fall entirely to others to be informed. In other circumstances parents or careers may strongly wish to ‘know what’s wrong’ but the person wishes to withhold … in my world issues around abuse, psychosis, substance misuse etc. strongly colour these wider communications – which are then cloaked in often confusing and misunderstood frameworks of ‘confidentiality’. The ‘system’ may need to know … in terms of funding and planning services and Drs may or may not care or be incentivised to respond … if there’s no repercussions we can be lazy or indifferent … at one time it was observed by the records dept in DPT that more than 40% of inpatients had no recorded diagnosis at discharge at all. If my income depended on it I’m sure I’d be keen to declare 100% of diagnoses and in pay-for services you may only gain funding for declared diagnoses. Some services are commissioned to be more directly accountable so early intervention services for first episode psychosis are particularly attentive to this and there have been recent skirmishes around financial incentives for diagnosing dementia. There are all kinds of issues in relation to employers and people around the individual that can modify whether a diagnosis is shared or not … mostly in relation to whether or not they are likely to respond supportively or negatively and I recently heard of a colleague in forensic and prison care who was sufficiently concerned by the adverse reactions some attracted through their diagnoses to have developed a practice based on negotiation with the person, ‘what shall we call it?’

 The implications of names

Having come to some kind of conclusion, preliminary or not, that someone’s difficulties can be named, diagnosed, as a particular entity we may then have all kinds of feelings ourselves and variations in practice concerning whether and what we ‘hand over’ … some of these relate to the diagnosis itself. It used to be that people would not be told they had cancer … and conversations were hedged about in euphemisms, but probably no longer. People want and value some diagnoses, ME, ADHD, PTSD, perhaps autism, certainly the now defunct ‘Asperger’s’ and possibly ‘bipolar’ … some have a fairly neutral value and relatively low social repercussions e.g. Phobia, depression, anxiety and perhaps OCD and others are highly aversive such as ‘schizophrenia’ … and this has an impact on the diagnostician and the conversation in which such words, concepts, names are transacted.

These social implications and processes can change in the light of shifting implications for the name as well … so it used to be that ‘borderline personality disorder’ was criticised as the medical way of saying ‘I don’t like you’ and resisted / rejected by patients whilst being often used between practitioners and sometimes valued by families and partners … then therapeutic advances meant that having BPD rendered you eligible for a high quality, high cost and helpful treatment package … and people sought and valued it as a diagnosis.

I don’t know if financial benefits can be seen as influential social processes but I’m also reminded that diagnosis is often the key to resources so the use and misuse of diagnoses can often be a route to care, support, specific help and money all of which the person may value and the push and pull around giving a diagnosis may be influenced by what else you get with it. A funny example would be that of a young man I worked with over some years who was insistent that there was nothing wrong with him at all and following discharge after his fifth compulsory admission for episodes medically diagnosed as ‘schizophrenia’ filled in a DLA form to get additional money … and under ‘diagnosis’ put ‘schizophrenia’ … when I asked him why he’d written that when he’d been stridently rejecting it for the previous 5 years he said, ‘ well … I still don’t believe it but if you’re going to accuse me of that I may as well get paid’ …  At a more sophisticated level the give and take of diagnosis may be mediated by wilful collusion according to the balance of accompanying advantages. Malingering and factitious disorders take these considerations even further where the perceived advantages of gaining a diagnosis are deceptively manipulated from diagnosticians and services.

The willingness to receive and the capacity to understand

So the diagnosis itself may impact on how easy or difficult the diagnostician finds it to ‘give it’ but how easy is it for the recipient to accept and unwrap it? The condition, stability, mental state and resourcefulness of the person are all influential. How prepared are they to hear, how well supported are they to handle the implications, how is their memory and concentration, how able are they to understand etc.  all of this mediates the social interactions in the diagnostic relationship. Taking the extreme example of someone in a psychotic crisis and brought to hospital under the mental health act whose mental disorder is appropriately diagnosed by admitting clinicians as ‘paranoid schizophrenia’ … it may be weeks or months before it is possible to have a conversation with the person concerning ‘how we understand what’s wrong and why we are giving you this treatment’ i.e. even trying to ‘give the diagnosis’ in the sense of developing shared understanding. And in practice it may be longer or even never that the person themselves are interested, able or willing to accept what is given … I may ‘make the diagnosis’ and offer it but the person will not receive it … so it’s not given … except to other interested parties such as hospital authorities, review tribunals, possibly DVLA and in complex negotiation with family and other carers. And there may be other non-medical diagnostic concepts that cover the issues but carry different implications and are more acceptable to persons themselves e.g. being understood as ‘hearing voices’ rather than schizophrenia, suffering from ‘extreme states’ rather than ‘psychosis’, having ‘eating distress’ rather than anorexia and having ‘complex PTSD’ or being ‘an adult survivor of childhood sexual abuse’ rather than personality disorder etc … and these words and meanings are socially defined and transacted in social contexts with different social implications e.g. for joining groups and finding peer support.

I’ve tried to give my views on the social processes in play that influence whether a diagnosis is given or not and of course there are then others at work affecting whether that ‘gift’ is unwrapped, kept, valued and used … but all of that’s another story.




I think I have to expanded on incentives both overt and hidden because they are so much a feature of general practitioners being in the front line of the Health Service. GPs more than any other doctor faced repeated decisions about whether or not to give a certificate, when a medical certificate is in effect legalisation of the sick role, and a cheque on the Government as well. Some patients have strong incentives to go off sick.

Some patients are sent by employers and some patients are in effect in a war with their employers and using the GP as one of their weapons. Other patients have strong incentives not to go off sick, and the GP has to read between the lines.

A huge complication is when the state intervenes by altering incentives and doctors and GPs again have been more subject to this than any other branch of the profession. The GP Quality and Outcome Framework has been called in the New England Journal of Medicine the biggest pay for performance system for doctors in the world.

There are inevitable distortions if doctors are paid to make certain diagnoses and we had a most uncomfortable example last year when the NHS wanted to pay GPs for diagnosing dementia, with GPs resisting the principle vigorously.

General practice and the specialist services

One point we haven’t yet covered is the distortion and pressures on GPs when support from specialists is either removed or reduced. Currently about one in four of referrals by GPs to child psychiatry is rejected. This of course puts huge pressure on GPs not just because they are then forced to manage patients and families for whom they want help, but it will also alter what diagnoses they mention to patients, as if they mention a diagnosis that normally needs a specialist and they can’t provide a specialist, they make their patients’ and their own lives even more difficult.

I entirely agree with you but about the use of symptoms as alternatives to medical labels and increasingly in my later years used them in my medical notes, partly to remind myself, partly to brief partners, and partly because medical notes are now available to patients.

Making a diagnosis or a note which is controversial to the family can cause GPs quite a bit of hassle, and I remember a difficult consultation when a man demanded that I remove from my notes ‘family history of schizophrenia’ from his record, even though it was completely accurate.

Screening and presymptomatic diagnosis

We haven’t mentioned this because it is essentially general practice issue. General practice increasingly is concerned with risk factors in people who have no symptoms. In the UK there are about 7 million people now taking statins for example, most  of whom have no symptoms and who are being treated or “diagnosed” because of a blood test reading.

This then extends the role of the doctor in society and moves it well away from the traditional patient going to the doctor for help with a problem. Now the doctor is giving the patient a problem they didn’t know about!

This is widespread and increasing and has considerable social ramifications.


Lastly I think we should just note that there is a problem with the current television image and language of medical scientists which is all about precision medicine, tailoring treatment precisely to diagnoses, which is all perfectly true, but for a very limited number of medical conditions.

The realities in clinical practice are very different where there is huge uncertainty and a high degree of error in all clinical branches of medicine. It is thought that 8% of GP diagnoses are wrong (Singh et al., 2017), 8% of hospital prescriptions contain an error, and there are reports of 15% differences between consultant radiologists reporting on the same x-rays (Fitzgerald et al., 2000).

Personally I believe we have to help our patients and ourselves accept clinical reality and realise that we are doing our best in very uncertain circumstances and that diagnosis in clinical medicine is inherently uncertain and often is simply a working plan, continually open to revision.