Interview with Margaret McCartney. The waste and harm of overdiagnosis

Puedes leer esta entrevista en castellano aquí.

Margaret McCartneyMargaret McCartney is a General Practitioner (GP) and undergraduate tutor at the School of Medicine in the University of Glasgow. She is also a columnist for the British Medical Journal. Award-winning writer and regular broadcaster for Radio 4’s Inside Health, she started writing for the press after reading in a newspaper that CT body scans were the way to stay healthy. She frequently blogs and tweets.

Her main interests focus on evidence, screening, risk communication, medicalization and professionalism. Some reflections on these and other concerns are included in her first book The Patient Paradox – Why Sexed up Medicine is Bad for Your Health, a brave analysis on how the healthy are turned into patients while we do not provide enough care for the sick. Her next book, Death Matters, Caring and Compassion at Life’s End, will be published later in 2014.

In this interview, Margaret McCartney shares with us her views on screening and health checks, the evidence that supports them and their impact.

Laura Diego (L.D.): What principles guide your practice?

-Margaret McCartney (M.M.): There is no doubt that general practice is very easy to get badly wrong. We have to place the patient at the heart and soul of all we do, but there are so many diversions that make it easy to be distracted from what the patient needs and listening to what they say. But I am no better than anyone else; I often think I could have done better. The political and structural context we work in means that it’s so important that we remember that more in medicine does not always mean better. The things that show quality in medical practice are not easily measurable. And last, harm, even with good intentions, is always possible.

L.D.: Can really screening harm?

-M.M.: Yes, it can. There is evidence showing that screening is not always straightforward and comes with dilemmas and side effects. The first words from the book, Screening Research and Practice, by Muir Gray and Angela Raffle are: “All screening does harm. Some does good as well, and of these, some do more good than harm at a reasonable cost”.

-L.D.: …but people usually don’t perceive it like this, what’s wrong then?

-M.M: The big problem is that we have not always shared this with patients. We don’t often explain screening well enough – meaning that it is the last reserve of unethical practice. However, we know that when people use decision aids and they make more informed choices about screening, more people decide not to have them.

L.D.: Are healthcare professionals looking at screening numbers right?

-M.M.: Studies suggest we are not, it’s very easy to get it wrong. Let’s take this theoretical example, from Gerd Gigerenzer. 1% of the female population have breast cancer. Mammography is 90% accurate. The false positive result -women testing positive is 9%- when they don’t have cancer.  So what’s your risk of having the disease if you test positive? Given that information, most people say there is a 90% chance of having breast cancer. There have been several studies running this kind of test with different health care professionals, and we get it wrong quite a lot too. The actual chance is about 10%.

L.D.: So the right answer would be…

-M.M.: The sums are quite easy when done this way. Take 1.000 people, 10, 1% will have the disease. The test is 90% accurate, so will pick up 9 of them. But it has a false positive rate of 9% for the 990 women who don’t have breast cancer. So it will pick 89 women with a false positive test. So there will be a total of 89+9 positive tests but only 9 of them are true positives. This means the chance of having the disease and testing positive are less than 10%. Of course we could look at this in a more complex way, but as a basic way of looking at the numbers, this method –from Gerd Gigerenzer´s work on Reckoning with Risk– is, I think, brilliant.


-L.D.: Is it then that we don´t hear about the negatives as often?

-M.M.: Exactly. Screening has side effects, but it is popular and often has a political will behind it – yet the uncertainties and problems it creates rarely get a fair hearing.

L.D.: However people still feel screening saved their lives…

-M.M.: This is the “popularity paradox”. The worse a screening test is, the more false positives there are; the more false positives there are, the more people are led to believe that a screening test saved their life. So the paradox is that the poor screening test becomes more popular, as people are led to believe that they have been saved, not harmed, by it.

L.D.: And what about health checks?

-M.M.: In the UK we now have the launch of ‘free NHS health checks’ which are claimed to reduce heart disease, stroke, diabetes and kidney disease. It is now law that every 40-75 year old should be offered a check every 5 years: smoking advice, body mass index, blood pressure, cholesterol, eGFR for renal function, and glucose if any risk factors are found. About a third of people who are invited will go for a health check – and it will miss around a third of the high-risk individuals. This programme is being implemented in absence of direct randomised controlled evidence to guide it. There has not been, and there are no plans, to subject this intervention to a randomised controlled trial. However the leaflet inviting people offer no drawbacks or mention of overtreatment…

-L.D.: And do we know what does and doesn’t work?

-M.M.: There are few things we know that work. We know smoking advice from professionals increases the quit rate from 3% to 5-6%. Physical activity advice works in the short but not long term, and 12 people have to be advised for one person to make the change to more exercise. Alcohol intake advice has a little effect; one study suggested that men, rather than women, could decrease the average of consumed units from 36 to 31 units a week, still excessive. Cholesterol and blood pressure lowering for primary prevention remains contentious for women, and for men the chance of benefit varies according to the risk to start with, a typical NNT, number needed to treat with a statin, is 18 per 1000 over 5 years. And we have good RCT evidence that screening for diabetes doesn’t cut death rates.

L.D.:  So then… health checks are not good at all?

-M.M.: Health checks are very good for example at putting people onto statins – 17% before, 60% afterwards. This is what I call “statination”. The new NICE guidelines say that we should offer statins to people who are at 10% cardiovascular 10-year risk using the Q-risk calculator in people aged over 40. This means that millions more people will be offered statins – 56% of women over 60 and 83% of men over 50. However the lower the risk, the lower an individual’s chance of benefiting from it.

-L.D.: Any other example you can think of?

-M.M.: GPs in the UK are incentivised under their contract with the NHS to diagnose and treat patients with high blood pressure. We are paid according to how well we do. We are told to treat high blood pressure where the levels are 140/90 or over or where ambulatory blood pressure is 135/85 or over; our target is to get blood pressure under 150/90. Points mean prizes – money. However, we know from a Cochrane review in 2012 that there is no evidence that treating a blood pressure, as a single risk factor, of under 159/99, does result in a benefit for the patient. In other words, we are being paid to treat patients but not to benefit them.

-L.D.: Is this “the patient paradox”?

-M.M.: Exactly. Too much ineffective treatment for the well and not enough for the sick. Look at the advertising for ‘health’ products people simply don’t need and the overselling of many medical interventions. We give lots of care and attention to people who will not benefit from the drugs we give them. We make it hard for sick people to get decent care, while pushing screening tests at the well. Everybody loses.

L.D.: So…

-M.M.: Preventing disease may be sometimes possible through medical action, but the biggest improvements to healthcare will be social and political. Reducing inequalities is where the biggest gains in health are to be made.

-L.D.: Is it time to let patients decide?

-M.M.: Absolutely. Doctors’ role should be to assist and guide patients and to challenge politicians about policies which make patients sicker.

-L.D.: Are you optimistic about changes in the future?

-M.M.: I hope that more patient engagement will mean more criticism of the way we work and better-informed patients who are unafraid to challenge medical dogma.

Margaret McCartney has recently visited Barcelona where she gave the Opening Lecture at the 20th Annual Meeting of the Societat Catalana de Farmàcia Clínica.

Interview by Laura Diego.

2 pensamientos en “Interview with Margaret McCartney. The waste and harm of overdiagnosis

  1. Pingback: Post d’estiu sobre sobrediagnòstic i sobretractament. D’on venim – AQuAS Blog


Introduce tus datos o haz clic en un icono para iniciar sesión:

Logo de

Estás comentando usando tu cuenta de Cerrar sesión /  Cambiar )

Google photo

Estás comentando usando tu cuenta de Google. Cerrar sesión /  Cambiar )

Imagen de Twitter

Estás comentando usando tu cuenta de Twitter. Cerrar sesión /  Cambiar )

Foto de Facebook

Estás comentando usando tu cuenta de Facebook. Cerrar sesión /  Cambiar )

Conectando a %s