Targets for dementia diagnoses will lead to overdiagnosisBMJ 2014; 348 doi: https://doi.org/10.1136/bmj.g2224 (Published 01 April 2014) Cite this as: BMJ 2014;348:g2224
All rapid responses
Burns, Hughes and Rasmussen, 15 April responded to Brunet. Their response, lengthy though it was, failed to impress the other discussants. I asked Brunet and colleagues a few simple questions. They remain silent.
Under the circumstances, the bodies they represent lose credibility in my eyes. Accordingly, I will not, financially or otherwise, support their cause.
Competing interests: I like to extract facts. I question the morality of case-finding if you are not going to offer care and treatment NOW.
In my last post, I wrongly typed Brunet. I meant Burns et al.
This slip of mine does not let Burns et al off the hook.
Competing interests: I erred. I sinned.
How rich a seam has been revealed by these responses! Dr Coeburgh provides a comprehensive account of one systematic disadvantage of ‘targeting’. While an intense concentration on one facet of a problem is often rewarding, it comes at the price of ignoring most others.
The same is true of any preliminary diagnosis, which at the bedside typically colours and prejudices all subsequent deliberation. In the case of ‘dementia’ the medical shorthand tends to amalgamate the symptoms of a syndrome with a disease category, causing even more mischief. In practice, some clumsy over-diagnosis is bound to be balanced by under-recognition due to the inevitable inefficiencies of human endeavour and ascertainment, the other systematic problem with aiming at, rather than to achieve, a target1,2. Is there a Bayesian treatment of making the diagnosis of dementia available, as a complement to the initiative?
1. Will E. Targets and Targeting. Am J Kid Dis 2001(2);38:411-414
2. Kay, J. Obliquity. Profile Books 2010
Competing interests: No competing interests
As Dr Brunet writes, the diagnosis belongs to the patient. And that is the problem in dementia when the patient can have little insight and often does not complain of the symptoms. This drive for GPs to increase diagnosis rates is admirable but there are many unknowns and potential dangers.
Most patients referred to my general or cognitive neurology clinic with ? dementia don't have it but a mixture of worried well, depression, anxiety, side effects of medication, previous brain injury, and sleep apnoea, all very treatable and easily missed if an established diagnosis of dementia is in place. In the last year despite writing about (Pract Neurol doi:10.1136/practneurol-2013-000796 How to do it ‘Undiagnosing’ neurological disease: how to do it, and when not to. Jan A Coebergh, Damian R Wren, Colin J Mumford) I have not been successful in removing the label of dementia in patients previously diagnosed by others who clearly did not have it (but some of the above). Patients are not happy at all not to have dementia since the label allows much support. This is uncharted territory but will probably become common with this drive for early non-specialist diagnosis (specialists get it wrong also!).
Diagnosis in those with language barriers, chronic mental illness, epilepsy, hearing and visual loss, previous substance abuse, illiteracy or a combination of these also can make it a very difficult diagnosis and the risk exists that the dementia (mis-/over)diagnosis leads to shortcuts in assessing patients' need for other investigations. These are predominantly elderly people and the focus on one diagnosis excludes a way of thinking about co-morbidity that is essential to good individualised care (? chemotherapy for your cancer; no, you have dementia).
The focus on memory symptoms of course also narrows the focus on Alzheimer's and misses the symptoms of Progressive Supranuclear Palsy, Frontotemporal Dementia and Lewy Body Dementia for example, which together constitute a large part of all dementias. The failure of all amnestic MCI to convert to Alzheimer and the behavioural phenocopy variant of Frontotemporal dementia is well documented so good follow up is needed in many difficult cases.
Empowering patients' family and friends to express concerns, standing up for the vulnerable who can't stand for themselves and assessing diagnostic difficulties and effect of co-morbidity can improve care; solely increasing diagnostic target rates is unlikely to.
Competing interests: I have been struggling to undiagnose people with a dementia diagnosis but no dementia
May I please request Prof Burns and his coauthors to reply briefly to my rapid response of 24 April?
May I request the Editor of the BMJ to encourage (there can be no coercion) Burns et al to reply?
Competing interests: Aged.
I have read the Personal View of Dr Brunet and all the responses so far (24 April).
Dr Brunet makes complete sense.
Professor Burns, Mr Hughes, Ms Rasmussen leave me with the following thoughts.
1. Prof Burns and colleagues should spend their existing time and existing resources on a) treating, b) caring for those patients who are already known to their organisations.
2. Please do NOT look for hitherto unknown "cases" if you are not able to care for and treat the known patients.
3. If you engage in screening when you are unable to adequately care for and treat the known cases, then your screening is an immoral, unethical activity.
Thank you, Dr Brunet, for trying to awaken a slumbering public.
Competing interests: Aged
Editor's note: This response has been edited since original publication. In its original version, the response complained of the absence of an author's competing interests, which has since been traced back to clerical rather than respondent error.
I am grateful to Prof Burns and colleagues for responding to my article, and would like to reply in turn.
It is incorrect to say that I am of the opinion that the ‘current interest in dementia will lead to harm and over-diagnosis.’ Indeed I welcome the much needed attention dementia care has received in recent years; it is specific policies that I am concerned about, and in particular the entirely novel concept that we should set targets for diagnosis rates. I am disappointed that Prof Burns chose not to address this central theme of the article, or indeed acknowledge that there might be both ethical considerations and potential harms in such a policy.
Prof Burns feels that, as the estimates of dementia prevalence suggest that nearly half the people with dementia are as yet undiagnosed, this is not the time to consider the possibility of over-diagnosis. This seems complacent since not only are the estimates outdated and in dispute, but the presence of undiagnosed persons in the community is of no consequence to an individual who has been over-diagnosed – the presence of one does not exclude the possibility of the other. Furthermore, we have to consider the effects of this policy many years down the line when diagnosis rates may well be much higher. The fact that there is to be an independent reassessment of the prevalence data is to be welcomed, but the old figures continue to be used while this is awaited. The launch of the latest version of the calculator means that each practice has been given new figures; my own practice has had its diagnosis rate changed from 126.7% to 59.6% - does this not demonstrate how meaningless these figures are at the level of a single practice?
The case of Mrs Jones is clearly fictional, but the influence of policy on the diagnostic process is not. One of my own patients was diagnosed with Alzheimer’s disease on the basis of an MRI scan alone, in the absence of supportive cognitive testing. When I challenged the psychiatry ST4 who made the diagnosis he seemed unconcerned that he might be wrong, and stated that the emphasis was to diagnose dementia early anyway. He had been influenced by the prevailing view in dementia care and neglected to give proper consideration to the person in front of him, whose diagnosis has since been refuted. The greatest risk of over-diagnosis is not the occasional human inaccuracy, but an ethos or policy which involves unrecognised harms.
It is disingenuous to imply that I have singled out dementia care as part of a wider attack on financial incentives in primary care. What is concerning about dementia policy is that financial incentives have been applied for the first time to the making of a diagnosis. I agree that ‘appropriate and timely diagnosis’ is important in achieving support for patients with dementia, but in my experience as a GP ,support frequently does not follow diagnosis; my most recently diagnosed patient, for example, was deemed to be unsuitable for anticholinesterase medication, and so discharged after initial assessment without any follow-up or support whatsoever. Policies that improve support for patients and their carers would have my full backing.
Competing interests: I am the author of the article, my declaration can be found with the article and also on www.whopaysthisdoctor.org
It was so refreshing to read Martin Brunet's article on Targets for Dementia diagnoses, but alarming to discover that they may well be the hidden or not so hidden financial agenda behind a GP's questioning of their elderly patients. If I am asked to name the Prime Minister, I am now primed to reply 'Disraeli', and see what happens!
I recently had a hearing aid assessment. Part of this was a question as to whether I felt isolated. It felt as if this was part of an algorithm check list, not a sensitive personal enquiry. By the end, the questioning, done almost by rote, left me feeling like a child. There had been no apparent attempt to intuitively sense whether this approach was appropriate in my particular case.
Martin Brunet's article was a most welcome warning to be on the alert for intrusive target-questioning when visiting a GP for those precious (to the patient) ten minutes of his time.
Competing interests: I am aged 77 and therefore at risk of an unsolicited GP Dementia assessment.
Martin Brunet highlights the pressure on Herefordshire GPs to increase the number of diagnoses of dementia . As the local memory clinic feels the strain, Herefordshire CCG is now trying to ease the bottleneck by encouraging GPs to make the diagnosis themselves. A recent newsletter stated that “adding a diagnosis of dementia does not require a referral to memory clinic, if the diagnosis is clear and treatment would not be beneficial” . This is contrary to both the NICE Guidelines  and the specifications for the “Facilitating Timely Diagnosis” Enhanced Service  which make it clear that patients with suspected dementia should be referred for a diagnosis to be made and to have a specialist assessment of the subtype, in order to guide possible treatment and access the support and benefits that early recognition is claimed to provide.
The culture of targets ignores the frequent difficulty and uncertainty in making a diagnosis of dementia, even by specialists in the memory clinic. Treatments are controversial, and it is often not clear who might benefit. Other chronic neurodegenerative conditions such as Parkinson’s Disease can be difficult to diagnose, have major prognostic implications and a range of potential treatments. We accept the need for initial specialist assessment for these conditions, so why should dementia be any different?
 1.Brunet M. Targets for dementia diagnosis will lead to overdiagnosis. BMJ 2014;348:g2224
 Herefordshire CCG Mental Health Steering Group Spring Newsletter Spring 2014
 CG 42 Dementia : Nice Guideline 22 November 2006 http://guidance.nice.org.uk/CG42/NICEGuidance/pdf/English
Enhanced Service Specification: Facilitating Timely Diagnosis and Support for people with Dementia. NHS Commissioning Board 2013
Competing interests: No competing interests
Dementia’s deadly twin: delirium
Brunet provides a welcome corrective to the troublingly rigid consensus that, when diagnosing dementia, early equates to timely.1 Dementia has now become the preserve of specialists sitting within prescription-oriented memory clinics.2 However, for most sufferers from dementia, it is a geriatric syndrome of multifactorial aetiology that demands a multicomponent approach by family, social and medical care teams rather than a complex neuropsychological approach 3. Recognising vulnerability to cognitive decline within general practice by ascertaining risk factors such as age, vascular morbidity, episodes of delirium and functional limitations may provide opportunities for safeguards against future cognitive decline. In this regard, a “Stop Delirium” model of care (optimising prescriptions, social stimulation, nutrition, hydration, mobility, pain control, sensory functions) may be helpful in formulating care plans to protect those at risk of dementia and those who have crossed thresholds for dementia already 4. The potentially catastrophic impact of delirium in the very elderly (including a 9-fold risk of dementia and 3-fold risk of accelerated functional decline in those with a pre-existing dementia) 5 means that militating against delirium must not be forgotten amidst the pressures to identify people in need of a diagnosis of dementia.
1.Brunet M. Targets for dementia diagnosis will lead to overdiagnosis. BMJ 2014;348:g2224
2 Pelosi AJ, McNulty SV, Jackson GA. Role of cholinesterase inhibitors in dementia care needs rethinking. BMJ 2006;333:491-93.
3 Brayne C, Davis D. Making Alzheimer's and dementia research fit for populations. Lancet 2012; 380: 1441-1443
4 Siddiqi N et al. Stop Delirium! A complex intervention to prevent delirium in care homes: a mixed-methods feasibility study . Age and Ageing 2011; 40 (1): 90-98.
5 Davis, D et al. Delirium is a strong risk factor for dementia in the oldest-old: a population-based cohort study. Brain 2012; 135(Pt 9):2809-16.
Competing interests: I have accepted speaker fees from Lundbeck and Pfizer (who market antialzheimer drugs)