The scandal of poor diagnosis in dementia that’s not

New research is a ‘wake up call’ for GPs- at least, according to Professor Steven Field, who is quoted today in the Telegraph as saying that doctors are needing more training in recognising dementia symptoms. The paper is in the BMJ, here, and I am rather amazed at the conclusions that both Field and the researchers themselves have drawn.

The research is this:

The researchers identified 22 529 patients who’s GP had coded as having dementia.

(This means that the GP had put a disease code onto the patient’s notes to say that they had dementia.)

Next, the researchers identified 5 patients in the same practice who had not been diagnosed with dementia.

Then they looked to see what happened next – specifically, who died first.

They found that people with a diagnosis of dementia had an increased risk of death; higher at that time of diagnosis, and which gradually fell, compared to the people who had not been diagnosed with dementia.

The abstract conclusion is really quite startling, considering what the research was; it says

“Median survival was much lower than in screened populations. These clinically relevant estimates can assist patients and carers, clinicians, and policy makers when planning support for this population…. Late recording of diagnoses of dementia in primary care may result in missed opportunities for potential early interventions.”

I think this is rather a brave, if not incorrect interpretation.

Firstly, the researchers did not classify patients as having dementia if the GP had written a code like ‘memory impairment’. To the GP, this might mean a suspicion of dementia. There may be all kinds of reasons as to why further more diagnositic tests are not done – the patient may not want them done, or the symptoms may be attributable to something else going on and meriting follow up rather than tests. GPs vary in who codes when – I like to be pretty certain before I code someone, whereas other GPs may use coding more as an aide memorie to remember current problems.

Second, it is hardly surprising that people with dementia had a higher mortality rate intially. These people were likely to have been diagnosed when they were unwell with something else. It may well have been that ‘something else’ which resulted in a higher mortality.

There was no ‘screening’ in this study: the researchers did not actively try to find cases of dementia in their comparison group. Comparisons with screened groups were from other studies, not patients within this study. The researchers cite a study of screening for dementia from the US as proof that screening is useful, but do not seem to taken cognisence of the findings, from the US Preventative Medicines Task Force, which showed that the accuracy of screening – the positive predictive value – was less than 50%, and that ‘ The effect of cholinesterase inhibitors or other treatments on persons with dementia detected by screening is uncertain.’ In other words, case finding by screening to find very early dementia is not necessarily going to give rise to better treatments for people any earlier.

So why the cries for earlier diagnosis? This study looks like it is a bad idea to be diagnosed with dementia earlier. You die younger – that’s what they found.

The Telegraph article goes on

“Dr Simon Ridley, from the Alzheimer’s Research Trust, backed up the call for better training of GPs and more research to make diagnosis easier.

He said: “Diagnosing dementia early is vital; it is when future treatments are most likely to be effective and it also gives people time to plan and prepare for the future.”

Well, we don’t have ‘future treatments’ yet, but I concur on the future planning – not just for people with memory problems, but for all the other difficult to forsee health problems we might have. At present, NICE say that pharmacological treatments for dementia are likely to help and are recommended for moderate problems only – not the early stage which screening looks for.

The bottom line is that we have not shown that early diagnosis of dementia will help people remain well for longer. Nor have we considered that early diagnosis may have side effects – false positives certainly, but also an impairment of quality of life from the moment of diagnosis onwards. People with a formal diagnosis of dementia may even be less likely to be offered scarce resources such as intensive care. This research asks more questions than it answers.

4 Responses to “The scandal of poor diagnosis in dementia that’s not”

  1. NI GP August 11, 2010 at 3:12 pm #

    Quite a few of the patients I suspect of having early dementia refuse referral to our memory clinic because they fear that they will lose their driving licence.

    If I code a patient as having “dementia” this triggers all the QOF requirements so I only code this diagnosis when it has been confirmed by a Psychogeriatrician. As you point out this often follows a new comorbidity (stroke, admission with acute febrile illness etc) or an exacerbation of an existing comorbidity. So as you say it would be surprising if my “dementia” patients did not have a higher initial mortality rate than my possibly dementing patients driving round in their cars.

  2. Margaret McCartney
    margaretmccartney August 11, 2010 at 10:02 pm #

    thanks – I’ve just been reading about all the research databases which rely on accurate GP coding. It’s rather alarming to think that this is seen as being so reliable that it can be used as such. The other thing about coding someone is that life insurance/holiday insurance becomes often difficult – and getting rid of a code after it’s been put on is almost unheard of.

  3. Gloria Crispino August 23, 2010 at 2:22 pm #


    I often read your column in the FT magazine. The issues you comment on are always interesting and novel. For completeness (and usefulness) could you include the full reference to the published articles you mention? As you know there is a monumental amount of published information and it is not always easy to find your way around it. Much appreciated. Gloria
    PS: By the way, the whole FT Weekend edition is a fantastic read. We always look forward to our Saturday morning when we walk to the newsagent to buy the paper 🙂

  4. Margaret McCartney
    margaretmccartney August 24, 2010 at 9:15 pm #

    thanks for this: the weekend section used to put all the references online so that they were easy to access. I’ll see if we can reinstate that either there or here, as it seemed to work well. thanks, mgt