Read the latest HealthWatch newsletter:  Newsletter 116, Summer 2021

Dr Margaret McCartney for her promotion of evidence-based medicine in general practice

Evidence-based general practice

Margaret McCartney is the 2008 winner of the HealthWatch Award. She is a GP in Glasgow who writes a stimulating and often provocative weekly medical column in the Financial Times Weekend section. She received her award at October’s Annual General Meeting. Here is the presentation she made on acceptance.

I am most certainly ‘for’ evidence based general practice. But evidence based general practice is not necessarily good general practice, and has brought a lot of new problems. What we surely want is not just any old EBM, but thoughtfully applied, professionally delivered, evidence based general practice. And while I do try and do my best to achieve this, I am not sure that I manage to. Bearing that in mind, I want to try and explain the problems and issues that EBM have led to in general practice, especially given the set of circumstances that medicine is now practised in.

I have a few personal guiding principles. More in medicine does not always mean better. Second, the most potent things about medical practice are not easily measurable. And last, harm is always possible.

‘Patient choice’ has been decided to be a good thing by politicians, and GPs are meant to offer choice to patients, for example, in where to be referred to for hospital care. GPs now spend large amounts of time discussing what clinic to send patients to, when perhaps what is of more importance lies elsewhere. The National Institute for Clinical Excellence offers to assess the evidence for treatment interventions and to try and measure its worth in terms of cost effectiveness. Yet NICE is often regarded as an enemy preventing people getting access to expensive interventions.

We have also seen the rise of the guideline. Guideline based medicine usually offers a distillation of evidence, but they are often used with a similar distillation of the limitations and uncertainties of that evidence. Guidelines have become goalposts, not necessarily a good thing.

GPs are usually independent contractors to the NHS. The GP contract essentially measures ‘Quality and Outcomes Framework points’. All sorts of things are measured. How many times blood pressures have been checked, how often people have been told to stop smoking, how many peak flows have been measured annually in people with asthma. Some have some evidence behind them—for example, there is a detectable decline in smoking where doctors offer help to quit. However, even here, it is still possible to do harm. For example, there are people who avoid their doctor because they are afraid that they will be chastised for being smokers still.

And of course doctors themselves are changing. GPs, and not just the female ones, are increasingly wishing to work part time. ‘Cradle to grave’ care is evaporating. Doing evening and weekends is in most places optional.

It is with these circumstances as the backdrop that a lady comes to see me. Statination is a word I use to describe the mechanism by which GPs currently seek to turn most people over the age of 50 into patients via prescribing a cholesterol lowering drug. Since her husband died three years ago, this lady has had episodes of anxiety and low mood which have been treated. She is an ex-smoker. Her high blood pressure is now in a satisfactory range. And she took up an offer to have her cholesterol checked for free at a high street chemist. The pharmacist was concerned. By the time my lady came to see me, she was agitated and worried. I went through her notes, and found that she had wanted to try a statin five years ago for a cholesterol of 6.9. There was a note about our discussion on lack of evidence for treating women in these circumstances. She had muscle pains and fatigue on taking the statin which resolved when she stopped it. We had decided then to concentrate on lowering cardiovascular risk overall via diet and exercise. This lady was now worrying that a heart attack was imminent. Have we really done her any favours?

As we know, there is good evidence and there is evidence which we can—and should—pick holes in. But even where evidence is favourable, it does not mean that we should necessarily start using the intervention. Statistical significance does not imply clinical significance.

What about the harms from something as commonplace as cervical screening? Every week or two I get a phone call from a lady who is distressed at being recalled for colposcopy. In fact the chances of being recalled unnecessarily—false positive—are far more likely than having a true positive. We should not underestimate the damage that this can do, both psychologically and in its wider effect on a woman’s family.

In recent years GP’s have been finding that smear results may include the additional information that the smear has shown ‘features consistent with HPV infection’. HPV has been in the news a lot recently because the new vaccination against it is now available. But when women have their smear test, the issue that signs of HPV might be reported in the smear result is not raised. When the result is available, GPs have a duty to share the information with the woman. But what then? The reason why we don’t know about the stress and damage caused by being told you have an untreatable infection which may—may—cause cervical cancer is because we have not adequately addressed the prospect of causing harm by our intervention. Yet such harm is witnessed in primary care and is most certainly done.

There are people in the audience who know much more about breast screening than I do. We have all seen on buses and billboards that the chances of getting breast cancer are ‘one in nine’. However this is not the case; this is only the risk if a woman is over 85, and most UK women don’t live that long. But this statistic wasn’t just used by breast cancer charities which, while they may have a useful role, also have an interest in getting media attention. I interviewed Julietta Patnick, the head of NHS screening for a piece in the Financial Times recently [1]. I was asking her why the publicity for the NHS breast screening programme also uses the one in nine figure, and this is what she said. “I’ll tell you what happened: we were debating what to put on the leaflet and I was going to work, following a bus. And there it was, ‘1 in 9’, on an advert on a bus. I realized that if we didn’t put it on, we were going to confuse like mad.”

We know that women overestimate their chances of getting breast cancer, underestimate the benefits of treatment, and do not realize that the risk of breast cancer increases with age. It could be said therefore that all the breast cancer ‘awareness’ months as well as the NHS information have been associated with damage rather than benefit. When staff from the breast screening service published research last year showing that women are not aware that the risk of breast cancer increases with age, they did not suggest that their information should be improved—only that GPs should be more aware of the lack of knowledge.

Breastfeeding is another example of damage being done by well intentioned doctors. GPs are meant to encourage breast feeding, even at the pre-conception stage. If staff in the practice all agree that they will encourage women to breast feed then you can get a special certificate to put up on the wall.Yet we know that breastfeeding rates, despite the enormous publicity, haven’t actually changed that much in the UK over the last decade. And we have not looked adequately for evidence of harm that these campaigns have done. I see many women who are struggling to breast feed and feeling desperately unhappy if they stop. I have also been concerned that women who have had problems and stopped breastfeeding become reluctant to seek medical advice about other matters because they are afraid of being scolded; and I am concerned that ‘failure’ to breastfeed could fuel postnatal depression. But apart from my anecdotes, I have no definitive evidence—because no one has looked for the harms of the relentless drive to push women towards breast feeding.
Even the benefits of breast feeding may have been overemphasized by health professionals. There is a good Cochrane review showing that the benefits from breastfeeding are overall small gains. These are important—but we should also be aware of the harms our enthusiasm can do.

“Awash in information, patients face a lonely, uncertain future”. “We’re in the outer reaches of medical knowledge,” he said, “and none of us knows what you should do. So you have to make the decision, based on your values.” Ms Gaines, bald, tumour ridden, and exhausted from chemotherapy, was reeling. “I’m not a doctor—I’m a criminal defence lawyer. How am I supposed to know?”

It is the word ‘lonely’ that gets to me. Of course doctors must acknowledge and attempt to reduce uncertainty—even the GMC says so—and doctors must be able to explain why they have done certain things. But ‘evidence’, or lack of evidence, is sometimes used as a way of abdicating professional responsibility. ‘No evidence’ can say, ‘no evidence for what I have to offer being of benefit’ but it is also used as meaning, ‘not my job’. This is increasingly seen where protocols are used to decide who might benefit from a particular intervention. But if you don’t fit the protocol, who is going to look after you, and help you make difficult decisions steeped in uncertainty? I think it is telling that there are ‘advocacy services’ now who seek to get better medical care for patients. ‘Advocacy’ used to be firmly in the doctor’s job description. Sadly, this is not now the case.
Neither are doctors seen as being ‘in charge’ of informing people about health and disease. To a degree this is a welcome destruction of medical paternalism. But the void has been filled by other sources who are arguably even more ‘interested’.

A prostate cancer charity produces a leaflet aimed at women, encouraging them to persuade their man into having a ‘prostate check up’. It suggests that they, “leave information leaflets lying around in the bathroom, remote control or car seat” and has a guide to how to, “tug at his heartstrings”. I’m not even going to start on the misinformation there. Or we have a suggestion in a problem page that, “GPs are not trained to deal with skin problems” which is news to me; however it seems that the private clinics listed are. Then we have another company offering ‘MOTs’ via CT scanning, with nothing about the risks of radiation, or false positives or negatives, but which they tell us ‘could save your life’.
So who do we trust to support us through illness, to advise us fairly about preventive health services, and to give balanced information with due regard for risk as well as benefit? This should be the job of professional healthcare workers, but to do this, we have to put our own house in order. General practice is in a mess, because we are conflicted thanks to the GP contract. We are still in servitude to the pharmaceutical industry with our reliance on them for postgraduate medical education—pharma estimates it provides half. Physicians with private rooms may be guilty of using the media to stoke a market. Our correct enthusiasm for independent and unbiased evidence based medicine has not been accompanied by a similar need to be independent and unbiased ourselves.

Healthcare workers need to ensure that evidence based medicine is applied professionally as well as honestly and fairly. General practice can do more than provide fair information and advocacy. Everyone here will know about the benefits of the placebo effect, but also that using placebos are riddled with ethical conundrums. However, if we are to regard the psychologically mediated effects of good healthcare as ‘placebo like’ effects, there are lots which we can gainfully and ethically use. For example, in general practice, continuity of care, longer consultations, and confident professionals all improve measurable health outcomes. One intervention for COPD—pulmonary rehabilitation—uses peer support and graduated exercise to improve quality of life and reduce hospital bed days. It seems to me that these kind of ‘evidence based’ interventions are the ones which are underutilized and under acknowledged and yet which bridge the gap between a professional health service which cares about individuals and the unafraid application of evidence based practice. I think there is hope for evidence based general practice yet, but if all patients are going to gain from it, we need to be heightened in our sensitivity to harms, and to deliver that medicine with professionalism, compassion and care.

Margaret McCartney



1. McCartney M. Reality check on breast cancer. Financial Times, 27th September 2008.
View online at: 0a-893c-11dd-8371-0000779fd18c.html?nclick_check=1


Are you for science and integrity in healthcare?



Already a member?


If you'd like to help us, please consider a donation.


yt logo rgb light

Mailing lists

If you are a member of the media or connected with a medical school or medical student organisation, please subscribe to our email list to receive occasional emails about press releases, our annual student competition and other news items we think might be of interest to you.
I agree with the Terms and Conditions