By Jerome Burne
It’s no secret that there are serious problems with the practice of scientific evidence based medicine (EBM). It’s obviously a good idea to have a system for ensuring treatments are safe and effective. But as a defence against dangerous or poor drugs, the working of our current one makes the pre-crash banking regulation look rigorous.
Even EBM’s most energetic defenders are eloquent about its failings. Recently Lancet editor Richard Horton wrote an editorial about a symposium held in London on the ‘reproducibility and reliability of biomedical research’.
“The case against science is straightforward,’ writes Horton ‘much of the scientific literature, perhaps half, may simply be untrue. Afflicted by studies with small sample sizes, tiny effects, invalid exploratory analyses, and flagrant conflicts of interest … it has taken a turn towards darkness.”
But there is a more serious flaw. Even if tightened regulations could improve the testing, that does nothing to address another growing problem. More drugs are not the solution to the rise in chronic diseases that’s threatening to bankrupt health services. However, demanding the same large randomised controlled trials (RCTs) used for drugs, to assess the benefits of lifestyle changes will simply reinforce the drug-driven status quo.
Why better RCTs are not the answer
Ben Goldacre energetically catalogued the failings of evidence based medicine as it related to drugs companies in his book Bad Pharma in 2012. Three years on there is still major problems. Last month he wrote in the BMJ: “Well documented problems exist in the funding and prioritisation of research, the conduct of trials, the withholding of results, the dissemination of evidence, and its implementation with patients.” (Link behind a firewall)
Goldacre among many others believes that fixing this litany of failings would deliver a true scientific system that could do the job. But the elements of heathy lifestyles that need researching don’t carry anything like the same risks and are not dominated by the marketing pressures that distort drug research. Relying on very expensive RCTs to test them is both unnecessary and impractical since only drug companies have the resources to pay for them so they won’t be done.
Even if charities or the government dug deep into their pockets and began to run many more RCT’s on lifestyle changes, they are the wrong tool to use. The lifestyle approach we need to integrate much more effectively into medicine doesn’t involve just changing one thing – drug or no drug – it involves doing lots of things at once – for example: different diets and more exercise combined with psychological techniques such as stress reduction. RCTs have difficulties with such multiple interventions. Yet when they are tested they often turn out more effective than drugs.
Recently I visited the surgery of a pioneering GP and had a glimpse of how patients’ experience could be transformed when a doctor spots some of EBM’s absurdities and begins taking lifestyle changes seriously. There was a big difference between the protocols demanded by EBM and his approach. He gets his patients involved in their health by finding out what life style changes they feel they need to make.
Disenchanted with evidence based medicine
Dr David Unwin is a GP in the Liverpool area who has become increasingly disenchanted with what he calls the “religion” of EBM. ‘When it was first arrived I bought right into it,’ he says. ‘Thought it was absolutely what was needed to prevent drug disasters like the claim that HRT protected against heart disease rather than causing it.’
But increasingly he feels it has less and less relevance to what is actually going on in his surgery. GPs prescribing options are constrained by the supposedly EBM guidelines which tell doctors which RCT-approved drugs to prescribe for each condition, often when a biomarker reaches a certain level.
But this number is an average and takes no account of the individual differences. ‘If your blood pressure goes above 120/80, I’m supposed to put you on hypertension pills,’ says Unwin. ‘But I can measure your blood pressure on three different occasions and get three different values. Which one should I treat? And it gets even less clear. A placebo study found that blood pressure could vary a lot depending entirely on whether the patient had been told his level was fine and healthy or too high and he needed treating. What’s also needed is clinical judgement.’
It’s worth pointing out that Unwin is no isolated maverick. He’s an advisor to the Royal College of GPs on setting guidelines and last year his surgery was shortlisted for BMJ awards for both Diabetes Team of the Year and Primary care Team of the Year.
Drugs tested on the wrong people
Another factor that makes the guidelines far from scientifically precise is that RCT trials are usually done on people who only have one disease. But as Unwin says: ‘Almost none of my older patients have just heart disease or just diabetes they often have both and some more. So how are those trials relevant to the patients I see every day?’
And this brings up the ironic fact that the logical result of applying evidence based guidelines to patients as they get older and sicker – the group that consume most of the drugs – is that they are soon being treated in a way that has no evidence underpinning it at all. The term for this is polypharmacy. No trials have ever or will ever be done on people with four conditions getting a total of 12 or more drugs.
What’s missing from the stack of guidelines, Unwin suggests, is one for ‘de-prescribing’ – how and when to take patents off some of their multiple drugs. ‘But who’s going to run large RCTs on giving fewer drugs?’
But the problem is not just that EBM doesn’t deliver what it promises. Rather than being a scientific way to discover what benefits patients, whatever the form it comes in, EBM has resulted in drugs becoming the main focus of medical research, with RCTs the tool for licencing them.
For years researchers and doctors when asked what research they would like to see say more investigation and testing of non-drugs treatments. This has simply been ignored. Not because there is something unscientific about it, but because non-drug treatments rarely make as much money.
Trials don’t test for things patients want to know
A recent study found that over the last decade 86 percent of commercial trials registered in the UK involved drugs. When doctors and clinicians formed partnerships, a project organised by the James Lind Alliance, just 18 percent of the research they initiated was drug based.
In a genuinely scientific system the lack of good trial evidence for nutritional and other lifestyle approaches to prevention would be a cause for outrage not a justification for usually paying them lip service before issuing the drugs. The idea it is safe to assume it can be largely ignored is absurd
Dr Unwin’s growing scepticism about guidelines meant he was increasingly prepared to trust himself. ‘I realised that if chronic diseases such as heart disease diabetes and cancer are the greatest health problem facing us, then handing out pills according to guidelines was increasingly irrelevant. Chronic diseases, as everyone acknowledges, are largely caused by unhealthy lifestyles. We don’t need more pills we need to get a lot smarter about the way we go about changing people’s behaviour.’
A couple of years ago he began to apply these ideas to the treatment of diabetics. Having become impressed by the evidence that a high fat low carb, Atkins-type diet was a more scientifically based way of treating diabetes than the official advice to eat little fat and lots of carbs, he faced the challenge of persuading his diabetic patients to switch and start eating lots of the long demonised saturated fat.
A level playing field between doctor and patient
‘I realised there had to be more of a level playing field between me and my patients. We had to balance evidence based medicine – you come with a problem; I give you a solution – with evidence based practise. That means drawing on my years of clinical experience, rather than just relying on guidelines, and applying it to patient’s own experience. They are the expert on their lives, what they need and what works for them. Without taking that into account you are not going to change anything.’
More details of Dr Unwin’s project can be found here.
The results were remarkable. Patients who had been following the evidence based guidelines for years and had remained obese and diabetic not only lost a lot of weight but also either significantly cut their drug intake or stopped taking them altogether.
A crucial element in the project was a psychological technique called ‘Solution Focused Practice’. Essentially this means rather than concentrating on the patient’s problems and then offering a solution – usually a pill – the doctor asks the patient what would make the situation better.
This then opens the way to talk about what they could do to achieve that goal. That was combined with the option of joining a group for weight loss and or diabetes held in the surgery. ‘I used to go along with some of the other surgery staff,’ says Unwin.
‘I was surprised at how powerfully effective it was. Until then I had totally given up on trying to get patients to lose weight. It never worked. Now I have a way that does. I’ve used it on 100 patients who are obese and have at least one other condition. The average weight loss over a year is 9 kg.’
So thinking outside the EBM tick box achieved impressive results in the treatment of obese and diabetic patients using by using multiple interventions: offering a diet that was the opposite of the recommended one, finding out how they wanted to change their lives and then supporting them to make them.
Cancer drugs get an easy ride
Does Dr Unwin’s radical approach, which just looks like good medicine, really need large RCT trials to prove it works? And even if there were and they did, how widely would it be used? Trials of the sort demanded by EBM show that the Mediterranean diet protects the heart better than one that cuts out fat, cognitive behaviour therapy is more effective for depression than anti-depressant drugs and better and safer for insomnia than sleeping pills. Yet which treatments are more widely used?
Under EBM’s current regime RCTs are not often run on non-drug treatments and when they are and are shown to be safe and effective, they are commonly ignored. So when patients, worried at the prospect of a life-time on drugs, ask if there is anything else that works they will be told – not any evidence based treatments. That’s why EBM in its present form is not fit for purpose, which should be to test any treatments that look safe and effective.
Instead what it does very well is keep them out by making a distinction between supposedly scientifically tested drug medicine and unscientific non-drug treatments which can then be dismissed as “quackery”, relying on anecdote and the placebo for any apparent effects.
And nowhere is the scientific/quackery boundary policed more vigorously than in the field of cancer where RCTs costing £100 million-plus provides the stamp of scientific approval for a new drug. But just how rigorously scientific is the testing? A lot less than drugs in other fields according to a recent BMJ article titled: ‘Why do cancer drugs get such an easy ride?’.
EBM not delivering
A review that looked at nearly 9000 trials of cancer drugs, run between 2007 and 2010 found that compared with drugs for other conditions, they were nearly 3 times more likely not to be randomised, 2.6 times more likely not to be compared with any other treatment and 1.8 times more likely not to be blinded.
‘The result,’ writes the author Donald Light, professor of comparative health care and an economic and organizational sociologist at Rowan University School of Osteopathic Medicine in New Jersey, ‘are cancer drugs that offer few significant benefits for patients.’ Between 1995 and 2005 the 71 drugs licensed for solid tumours in Europe improved survival by a mean and median of 1.5 and 1.2 months respectively.’ That is the kind of results that EBM is delivering.
Now of course showing that it’s drug companies, rather than patients, who benefit hugely from this light touch regulation does not show that cancer treatments without RCTs are effective and safe. But what it does suggest is that dividing cancer and other treatments into scientific and non-scientific is misleading and not beneficial to patients since the system ensures that however promising the ‘non-scientific’ ones (which also happen to be the ones lacking the commercial potential needed to trigger a large trial) stay on the shelf.
We need ways to test treatments but EBM in its present form isn’t delivering. One final example: Statin drugs have been used for over 20 years, they are the most widely prescribed drugs ever, they have been subject to a vast number of RCTs and yet last month the Chief Medical Officer called for a committee to be set up to decide if they were safe and effective or not.
Twenty years and still no answer?
The James Lind research shows people want something different, so does the massive use of the internet to track down information on health. What’s needed is a move away from the sterile debate: scientific vs. not-scientific, greater non- commercial involvement in testing and investigations to discover techniques besides RCTs that can deliver reliable information about lifestyle change.