Sunday, 29 January 2012
Prevention is better than cure: the media coverage of our recent paper in the BMJ
The front page headline for the Daily Mail last Thursday was ‘Heart attack deaths halve in 8 years’. This was a report of our paper published that day in the British Medical Journal (BMJ) and to be found here.
The paper also made the front pages in the Telegraph and Express and into the Guardian, the Independent, the Mirror, the Huffington post and onto the BBC News (radio and television), the Today Programme and even into a newspaper in Winnipeg (according to Kate Smolina – the first author of the paper and until recently my DPhil student (jointly supervised with the co-authors of the paper: Michael Goldacre and Lucy Wright)). Unfortunately the paper doesn’t seem to have made it into the Daily Mash yet as one of our other papers recently did.
Our paper presents several ‘findings’ but the main finding is, in my view, that just over half of the decline in heart attack deaths between 2002 and 201 in England was caused by fewer people having heart attacks and just under half by more people who had heart attacks surviving. We knew already that the number of deaths from heart attack was declining and suspected that this was partly due to fewer people having attacks and partly due to more people surviving attacks but we didn’t know for certain which was more important and now we do.
I therefore find the headline of Jeremy Laurance’s piece in the Independent ‘The curious case of the vanishing killer’ and its opening lines particularly irritating. Laurance says ‘It's a good news story that has [not’ had’ note] medical researchers baffled: in the past decade, deaths from heart disease in the UK have fallen spectacularly, and no one knows why…’. Has he not read our paper? Does he not get the point? But what was the point of our paper as opposed to its main finding? I’ll try and explain that later but meanwhile what did the media reports think was the point.
The reporting of our paper has been extensive but also extremely variable with the report by NHS Choices being one of the best and the report by Sky News being one of the worst.
Many of the reporters don’t seem to have bothered to try and contact the authors of the paper for their views on the pointy of the paper or even what the paper says but have relied instead on comments from others – notably the British Heart Foundation. The best report I have seen – from NHS Choices – does not rely on comments from others but draws its own conclusions. I don’t entirely agree with all of its conclusions as I do not entirely agree with everything that is said in the paper. (A paper written by multiple authors is always a compromise.)
In particular – as NHS Choices says - ‘It should be noted that although the study shows a fall in heart attacks and also in death rates among heart attack patients, it does not tell us the precise causes of either.’ It’s probably because the paper is so clear in its main finding – and sorts out an important question – that speculation on both fronts has been so rife. For my own speculations see the news story from the University of Oxford University
Some of this speculation is quite frankly ‘wild’. There is some wild speculation in our paper itself such as: ‘The increase in acute myocardial infarction event rate in London between 2007 and 2009 may be a result of the financial crisis that peaked in 2008 and greatly affected the London financial district.’ How on earth did this get into the paper I am now thinking? This rather small, foolish bit of speculation seems to have been reproduced in several articles: as if the financial crisis just affects the ‘London financial district’. See for example the Telegraph.
So what are the real causes of the a) decline in the number of heart attacks b) the death rates among heart attack patients? These are many and varied and could well be displayed in the sort of map that the Government’s Foresight Programme developed to ‘explain’ the causes of obesity and which I have discussed previously on this blog here.
I was tempted in my one (rather ill fated) media appearance in relation to the paper to suggest that improvements in food labelling had played a major role in the decline in the number of heart attacks. Implausible? Equally plausible, in my view, as the suggestion that the financial crisis had important effects on the trends.
There is actually some research into (rather than speculation on) the causes of the decline in deaths from heart attacks in England - admittedly from an earlier time period i.e. from 1981 - approximately the point at which the decline in deaths from heart attacks began - and 2000. This research - from Simon Capewell and his colleagues at Liverpool University - suggests that 58% of the decline can be attributed to ‘population risk factor reductions’ (principally smoking, 48%; blood pressure, 9.5%; and cholesterol, 9.5%). Given that the annual decline in heart disease deaths between 1981 and 2000 was roughly the same as between 2002and 2010 it is not unreasonable to suggest that this research should be used to interpret our results. Did any of the journalists do this? No of course not. We did mention this paper in our paper but perhaps we should have given it a bigger billing.
Of course interpretation in science is largely value driven rather than evidence driven. There are two interpretations of our results which I like. The first comes from Andy Burham - Labour's Shadow Health Secretary.
Any Burham seems to be suggesting that the introduction of the National Service Framework for Coronary Heart Disease – introduced by the last Labour Government – was a cause of the fall in heart disease deaths. He says ‘In Government, Labour launched a national drive to cut deaths from heart attacks through the national service framework on coronary heart disease. It succeeded because of the strength of the NHS structure - which Mr Cameron is now to break apart.’
Andy Burnham’s contention that the National Service Framework was a major cause of the decline in heart disease deaths is a tad speculative - particularly as the decline started in around 1980 long before the National Service Framework was introduced (in 2000) but nevertheless it is true that our study ‘calls into questions whether the NHS needs to change radically in the way it does things.’ And Andy Burnham’s point the ‘if the NHS isn’t broke why fix it?’ is appropriate – at least with regard to the treatment of heart disease which, it should be noted, is one of the causes of ill-health in England, if not the biggest. Note that both David Cameron and Andrew Lansley have used the argument that England’s high rate of heart disease (compared with say France) justifies their proposed reforms.
The other interpretation I like I haven’t found anywhere yet. This is that our study shows that ‘prevention is better than cure’. Note that more than 50% of the decline in heart attacks deaths is due to the decline in the actual number of attacks and less than 50% of the decline is due to improvements in survival after an attack. I.e. the prevention of attacks is more important than what goes on after the attack. Note too that it is possible that survival after attacks has only a little to do with improvements in the treatment of heart attacks (after they have occurred) and more to do with hearts getting stronger (from preventive activities prior to the attack).
Of course the decline in the number of attacks may be partly due to improvements in the drugs given to people at high risk of an attack (statins for raised blood cholesterol, anti-hypertensives for raised blood pressure, etc) and even improved forms of surgical treatment (such a ‘stents’) for people who have serious heart disease but given these drugs and new forms of surgical treatment came into use only over the last decade or so – long after the decline in deaths began (around 1980) it seems unlikely that they have made a major contribution (see too Simon Capewell’s findings mentioned above) This leaves us with lifestyle change as the mostly likely main cause of the fall in heart attack deaths (again see Simon Capewell’s findings). The decline in smoking is probably the most important lifestyle change but improvements in diet such as the switch from animal fat to vegetable fat which began about the time heart attack deaths began to fall, may also have contributed.
Whether the Health Service (much as I love it) has had anything much to do with these lifestyle changes seems unlikely. The switch from animal fat to vegetable fat was largely due to the increasing availability of cheap vegetable oil (sun-flower, rape seed, etc.) due to agriculture subsidies under the Common Agriculture Policy etc. When did your doctor last talk to you about saturated fat and did you head his/her advice?
Interpretation of the results of papers is, in the end, value driven. My interpretation of the results of our paper (which I am not sure is even entirely shared by my co-authors) is that it means that prevention is better than cure. This interpretation is – like Andy Burnham’s –based, not just on the findings of the paper – but on speculation. Though I do have some empirical support from other research such as that of Simon Capewell and his colleagues. I also have some logic to support my case. It is value driven: because I am a preventionist rather than a treatist.
The problem with the media coverage of our paper – from my (preventionst) point of view was that it got hijacked by too many treatists. OK perhaps I should feel grateful for the amount of coverage our paper got and that at least prevention sometimes got a mention. (The paper does to some extent speak for itself without the need for too much interpretation.)
Several people have congratulated me on the extent of the media coverage for the paper. But I really do not care about coverage. All publicity is not good publicity in my book. I am sorry that the coverage didn’t do more to help Andy Burnham’s cause - hat the NHS isn’t broke so why fix it - or mine that ‘prevention is better than cure’. In saying this I should say that I am not blaming anyone but myself.