Diagnosis: Scandal

    27 September 2014

    Seen your doctor recently? You may have been asked about your memory. This would be quite usual if you had memory symptoms — like forgetting phone numbers, or wandering into rooms unable to recall what you were looking for. But for people who went to their doctor to consult for something different, being questioned about memory is quite new.

    As a newly qualified GP in the late 1990s, I was bemused by the system used for paying us. It was chaotic, and priced via the ‘red book’. GPs were paid for individual ‘items of service’. This meant pennies and pounds every time we prescribed, say, contraception. Each prescription, vaccination and smear test had to be documented on bits of paper sent to the Health Board to trigger a payment. The paperwork piled high. In 2004, after a ballot of GPs, the ‘red book’ was replaced by the gleaming new GP contract. But the contractual thesis remains the same. We continue to be paid piecemeal for bits and pieces of our work. Crucially, however, the contracted hoops we are to jump through are replaced, narrowed and heightened every year.

    So, in 2012, the Department of Health negotiated a new hoop. Jeremy Hunt had taken on dementia. He told parliament on 23 October that ‘tackling dementia — particularly the shockingly low diagnosis rates — is a key priority for me and the Prime Minister’. A month later he wrote, ‘For far too long people with dementia and their carers have not received the care and support they deserve. With our ageing population that must change. More early diagnosis, better research and better support for carers are essential if the NHS is to offer decent, humane support in line with its founding values.’ This is all true; but the crucial phrase buried within is ‘early diagnosis’.

    In January 2013 Hunt kept up the pressure, writing in the Telegraph, ‘We must start with better access to, and attitudes towards, early diagnosis. It can be a total nightmare getting a diagnosis — and the result is that, shockingly, only 46 per cent of all dementia cases are identified. Yet with access to the right drugs and support for a partner, someone can live happily and healthily at home for much longer… too many health and care professionals are not aware of the symptoms. Some even believe that without an effective cure there’s no point putting people through the anxiety of a memory test — even though drugs can help stave off the condition for several years.’

    The outcome of this political priority has been to push for ‘early diagnosis’ of dementia in the GP surgery. The determination of Hunt to ‘improve’ diagnosis rates of dementia has meant that commissioning groups are offering a ‘bonus payment’ of £200 to ‘any practice that increases the diagnosis rate by 5 per cent’. The average practice taking this work on can expect to receive 37p per patient for doing so. Yes, it feels grubby to me too.

    Missing from the political dementia drive has been the medical evidence to support it. It sounds logical and sensible to ‘catch it early’. A stitch in time; better sooner than later; don’t put your head in the sand! And certainly, if someone has symptoms that would suggest dementia as a possibility — difficulty doing familiar tasks, disorientation, problems with remembering words — seeking help is useful. Other conditions can be excluded, tests can be offered, support can be given. Neither will this be a one-off appointment, but a series of interactions over time — for time itself can be a diagnostic tool.

    But the new hoop of the GP contract is nothing to do with such symptomatic people. Instead, the political fervour for ‘early diagnosis’ means that questions about memory are aimed at a different group of people, the asymptomatic. These are the people who did not come to their doctors to discuss their memory.

    In scientific terms, this changes the game — it becomes health screening. Screening science is complex and poorly understood, especially by many politicians, yet dementia screening not only doesn’t work, but also does harm. This, according to research studies, is the bottom line. Take 100 people, six of whom had dementia. Current evidence would suggest that screening for dementia would pick up four of those people, missing two. But it would also pick up 23 people as having dementia when in reality they did not. So 27 out of 100 people are told they have dementia — but only four do. In other words, if you tested positive from the screening test, only 15 per cent of the time would you really have dementia. In screening science terms, this is a poor test.

    Additionally, screening tests often pick up abnormalities whose meaning isn’t clear. Dementia screening will often pick up ‘mild cognitive impairment’ — minor memory changes. What does this mean? When people given this label are followed up, the results are surprising. Most people with this diagnosis will never develop Alzheimer’s disease. Between 5 and 10 per cent of people will progress to dementia each year, and some people who have been told they have ‘mild cognitive impairment’ will find that their memory powers improve over time rather than deteriorate. So what does this mean for the person who has come to see his doctor about something quite different? It means the GP is incentivised to ask questions about that person’s memory. It means that a quarter of the time ‘abnormalities’ on screening will be found when the diagnosis isn’t dementia. That doesn’t sound to me like a good deal.

    But what of Hunt’s idea that dementia screening can find patients earlier who are eligible for medication which will, he supposes, enable people to live a higher quality of life at home for longer? The scientific fact is somewhat sobering. The British Association for Psychopharmacology produced a statement in 2011 saying, ‘There is no clear evidence that any intervention can prevent or delay the onset of dementia.’ In other words, prophylaxis using medication will not stop dementia occurring. As for the drugs available for the treatment of dementia, the benefits are frequently overstated. Reviews suggest that the cholinesterase inhibitor class of drugs — donepezil, galantamine and rivastigmine — are capable of producing an improvement in cognitive testing of 2.7 points in a 70-point memory testing scale. This, frankly, is small. As for quality of life, the drugs do make a difference to the ability to perform ‘functional tasks’ like washing and dressing, but in the words of the review ‘none of these treatment effects are very large’. And, since the side effects of the medication can include nausea, vomiting and diarrhoea, as well as an increased risk of fainting, hip fracture, and of being fitted with a pacemaker, we can hardly suggest that the medication we have is a panacea against dementia.

    What outrages me most about the dementia screening contract — although it’s been officially called ‘case finding’, as if that makes it scientifically more valid — is the blindsiding of patients. For a government who said ‘no decision about me, without me’, being tramlined into screening for dementia when the patient seeks help from their GP about a completely different matter alters the dynamic between patients and doctors. And it is ethically questionable to spend resources to administer the dementia screening questions when people with dementia and their families often struggle to get the bare minimum of care. As for the suggestion that many people have dementia but not a diagnosis — recent research has suggested that the amount of people with dementia in the community has been greatly overestimated.

    Nor is dementia screening the only problematic screening test pushed by the government. The right of every over-40-year-old to have a five-yearly ‘Health Check’ is now enshrined in law. The problem is that evidence for its blood pressure, cholesterol and kidney function testing working is distinctly lacking. Notice a pattern? The original ‘evidence’ as provided by the NHS claimed ‘health checks’ could save 2,000 lives a year — by 2013, the claim was for 650 lives. When divided into the 15 million or so people eligible, this is not a great return. While Public Health England has claimed that the programme is a ‘fantastic opportunity’ to prevent suffering and detect early disease, it also notes that there is an ‘absence of direct randomised controlled trials to guide it’. In other words, we don’t actually know whether the millions being poured into Health Checks will make any difference at all.

    The first problem is the ‘healthy attender’ effect. When invitations are issued, the most healthy, rather than the most unhealthy, are most likely to turn up; but of course they have the least to gain. The second problem is ‘lead time bias’, which can make screening sound more powerful than it actually is. For example, take a person diagnosed with cancer through screening in 2010 and dying from it in 2015. Without screening, that same person might have been diagnosed in 2013 but still died in 2015. It would look as though screening led to a longer survival, when in fact, it only led to earlier detection — and without clinical trials to compare one group with the other, this bias will continue to play out.

    In fact, the evidence against Health Checks is quite clear — they don’t work. A high-quality, randomised trial from the UK, published in the Lancet in 2012, found that screening high-risk people for diabetes made no improvement to death rates, heart attacks or diabetes-related deaths. What, then, do they achieve? So far, it seems that Health Checks seem to be a route to ‘statination’ — the process of putting a vast number of people on to statin tablets. Before a Health Check, 14 per cent of patients were prescribed statins; after, 61 per cent. This might all seem very well — if they work. There is a big difference between their usefulness after a heart attack or stroke, say, and in people who are invited to the checks who have no such history and are otherwise well. There is a dispute over whether statins can prevent deaths in this group, especially in women. Taking 100 people taking statins for five years, between one and two heart attacks or strokes will be prevented. However, the side effects will include two people developing diabetes, others will get muscle aches, and less commonly, kidney failure.

    If it’s bad in the NHS, it’s far worse in the private sector. Private clinics will take several thousand pounds to do an MRI scan of your body with the tagline ‘prevention is better than cure’. Yet 7.2 per cent of entirely ‘normal’ people will have ‘abnormalities’ on their brain MRI — and the cost of dealing with this is usually, but immorally for the private sector, picked up by the NHS. So common are these problems that the acronym Vomit — victims of modern imaging technology — has been coined to describe the fallout from too many tests.

    We have to wake up to the fact that screening turns well people unnecessarily into patients. We should stop throwing our time, energy and money into screening with its many illusory benefits. Instead we should put our resources where it matters — looking after people who are ill, who have dementia, or even just people who want ten minutes to talk about something that’s bothering them — and not spending our time and efforts to fulfil bad political policy.