By Jerome Burne
Last week a massive piece of research extolling the benefits of statins was published in the Lancet. Its headline message was that the benefits of statins are hugely underestimated and far outweigh any harm. It had found that only 2 people per hundred suffer side effects while 15 people per hundred avoid strokes and heart attacks.
It came garlanded with the full majesty of the medical profession. Thirty pages long with 28 authors, 300 references and a declaration of interests that filled 44 lines of small type and mentioned more than 20 of the largest pharmaceutical companies. This juggernaut was designed to bury the widely publicised doubts about the effectiveness and safety of these drugs already prescribed to 6 million people.
‘This study will give patients and doctors the confidence in statins as a life-saving medication,’ declared the Chief Executive of the charity Heart UK.
So that’s that then. All those worries about large numbers stopping taking the drugs because of side-effects, links with diabetes and cataracts, studies showing that hundreds of healthy people with a small risk of heart disease have to take them for one to benefit, all laid to rest. Or are they?
Bizarre history of statins for all
This hasn’t come out of nowhere; it arrives with a long and frequently bizarre history. The lead researcher – Professor Sir Rory Collins – has been publishing papers saying much the same thing for decades. Six years ago one of his papers, also published in the Lancet, was the basis for a change to NICE guidelines which recommended a doubling of the number of people eligible for statins because they were so safe and effective.
But this latest recitation of the benefits is unlikely to settle the many doubts about statins because, like all the others, it totally ignores the reasons why many researchers and clinicians don’t trust Professor Collins’ research. It’s all based on data from trials run by pharmaceutical companies which no researchers other than he and his team in Oxford are allowed to see.
Everyone now agrees that keeping the full results from trials secret makes them unscientific and unreliable. Dodgy statistics and unfavourable results that have been concealed regularly emerge whenever hidden data is finally examined. Despite repeated requests, the institute Collins heads – the CCT in Oxford – resolutely refuses to let the light in on the dark matter in his vaults. Instead he repeatedly accuses his critics of killing people by casting doubts on statin’s effectiveness.
Alarmingly this ‘killing patients’ defence seems to have been taken up by the Lancet, whose editorial makes the case that research challenging medical orthodoxy should be checked especially carefully and possibly not published. Authoritarian support for the status quo is common in politics and religion. It’s not supposed to be part of science. A far stronger case could be made for not publishing research based on hidden data.
Effectiveness of statins not evidence based
If you are looking for a short explanation of why the latest Lancet paper, despite its heavyweight support, is not the final word, you could stop now. But if you’d like to know more about the other side of the statin propaganda picture, please read on.
An important criticism that deserves a full response – rather than special attention because it is rocking the boat – comes from the well-respected statin critic Dr Michel de Lorgeril of TIMC-IMAG CNRS UMR 5525, Laboratoire Cœur et Nutrition, Université Joseph Fourier, Grenoble, France.
Last year he carried out a review of statin trials which, he concluded: ‘Strongly suggests that statins are not effective for cardiovascular prevention……a complete reassessment is mandatory. Until then, physicians should be aware that the present claims about the efficacy and safety of statins are not evidence based.’ (Read the article)
To come to this conclusion, he divided statin trials into two groups – those done before and after 2005. That’s because in 2005 new, stricter regulations were introduced in America and the EU changing the way randomised controlled trials (RCTs) were to be done. (All the big commercial trials supporting statin use and held by CTT are RCTs)
The change had been prompted by the discovery that a range of published trials, such as those of the painkilling drugs Vioxx and Celebrex, had downplayed or concealed dangerous side-effects. ‘The implication (of the changed regulations) was that previous drug assessment procedures were not satisfactory,’ writes Lorgeril.
Diabetes link with statins took 30 years to emerge
So his team looked at the statins trials done since the new requirements for greater transparency. There were four of them, all involving rosuvastatin, and they ‘unambiguously showed that the statin was not effective in secondary prevention, while the results are highly debatable in primary prevention,’ he writes in the Journal of Controversies in Biomedical Research in a paper entitled ‘Can We Evaluate the Real Efficacy and Safety of Cholesterol-Lowering with Statins?’
One of these trials – presented as a great success at the time – was known as Jupiter and it revealed, for the first time, that statins raised the risk for diabetes. Lorgeril pointed out that, the failure of the large number of statin trials to pick up this serious risk for 30 years clearly revealed ‘a high level of bias in reporting harmful outcomes in (pre-2005) commercial trials.’ So what other omissions or distortions are lurking there?
Lorgeril concludes that ‘the only possible interpretation of the data is that ‘statins do not protect diabetics’, and that ‘there is no question about their diabetogenic effect’. The great bulk of the trials held by CTT were done before the 2005 regulatory upgrade.
But in addition to serious doubts about the reliability of CTT’s data, there is also well-informed disagreement about how many people (usually out of a hundred) have to take a statin for one to benefit or to be harmed. (Known as numbers needed to treat (NNT) and numbers needed to harm (NNH).
The latest Lancet trial claims a very favourable balance between the NNTs and NNHs. A far less rosy picture emerges from a recent analysis available on a website called theNNT.com, run by Professor David Newman, director of clinical research at Mt. Sinai School of Medicine in New York, together with doctors at St. Luke’s-Roosevelt Hospital Center in New York City.
Greater chance of developing diabetes than avoiding CHD
This calculated the NNT and NNH of statins taken for five years by people who hadn’t had a heart attack but who were at high risk – most had conditions such as high blood pressure and/or diabetes or smoked. [Link] Their risk of a heart attack (25% over the next 10 years) was much higher than the risk level that NICE decided should make people eligible (10% over ten years). That means that you would expect them to show up as getting more benefit than the lower risk NICE group. This is not what they found.
- NNT: Lives saved – 0
- Preventing a heart attack – 1/104
- Preventing a stroke – 1/154
- NNH: Developed diabetes – 1/100
- Muscle damage – 1/10.
In other words, you are more likely to develop diabetes if you take a statin for five years than you are to avoid a heart attack, while your risk of muscle damage is far greater than the Lancet paper claims
Neither Newman nor Lorgeril would claim their studies are the final word but they are clearly serious, informed and thoughtful criticisms of the Lancet’s heavyweight claim of infallibility. Until they are properly taken into account and until Sir Rory massive collection of secret trial data has been independently assessed, the confusion and uncertainty around statins will remain.
Collins, however, has made it very clear that he belongs to the ‘never apologise, never explain’ school of argument. When he attacked the BMJ two years ago for running two articles critical of his claims about the low-level of side effects, he made no attempt to answer them. Instead he persistently demanded they be withdrawn. The committee set up to decide on the matter found that there was no basis for withdrawal. In fact, the panel was critical of Collins for refusing to submit a published response to the articles:
Not interested in scientific dialogue
‘The panel noted with concern that despite the Editor’s repeated requests that Rory Collins should put his criticisms in writing as a rapid response, a letter to the editor or as a stand-alone article, all his submissions were clearly marked ‘Not for Publication’. The panel considered this unlikely to promote open scientific dialogue in the tradition of the BMJ.’ Two years on Collins is still not promoting dialogue, preferring to debate by fiat.
As for apologies, his behaviour over side-effects would have drawn an abject one from most people but not Collins. Some months after the committee’s decision it emerged that, amazingly, he and his team had never actually conducted an analysis of his secret data to find the level of side-effects they had reported. Rather than apologies for failing to gather evidence for a claim that is absolutely central to widespread statin use, he offered to run a belated trial himself.
Why on earth has the Lancet, long a champion of evidence based medicine, chosen to ally itself with the idiosyncratic form of eminence based medicine that relies on hidden data and a refusal to engage in normal scientific dialogue? It’s hard to see how this latest report gives patients or doctors confidence about anything to do with statins.
As if that wasn’t enough the one question that every patient wants to know about statins is never even mentioned in the Lancet paper.’ ‘If I take these drugs for five years doc, how much longer will live? Statin research claims to ‘save lives’ but as long-term statin critic Dr Malcolm Kendrick points out, since we all die eventually, all any drug can do is to prolong it.
Live 5 to 20 days longer taking statins
A recent study in the BMJ provided an answer. (doi:10.1136/bmjopen-2014-007118) Death was postponed between 5 and 19 days in primary prevention trials and between 10 and 27 days in secondary prevention trials!
So the Lancet paper certainly hasn’t made all the uncertainty disappear but it has revealed a nasty iron fist clenched beneath its evidence based glove. The editorial refers to: research papers that are ‘more high risk to public health than others’ which therefore ‘deserve extra vigilance ….and should be ‘subjected to rigorous and extensive challenge … If publication is agreed, it should be managed with exquisite care.’ In other words, it might be better if some papers that risk rocking the orthodox boat were quietly disappeared. For more on this see Dr Kendrick’s latest blog.
Such an overtly authoritarian line only makes sense if the criticism is wrong and the public policy is soundly based. But who decides that? Massive analyses like the latest Lancet have become so complex that only a few people can properly analyse them. This is not a problem unique to medicine. It’s an inevitable part of the growth of ‘big data’ which has been analysed in a new book by mathematician Cathy O’Neil: Weapons of Math Destruction: How big data increases inequality and threatens democracy. The tools for analysing this data gain their authority from appearing objective and scientific but, as O’Neil points out, they are based on choices made by fallible humans.
‘Like gods these mathematical models are opaque,’ she writes ‘their workings invisible to all but the highest priests in their domain: mathematicians and computer scientists.’ O’Neil was talking specifically about the sort of algorithms increasingly used in industry and financial services. She might have added ‘bio-medical statisticians’.
Just how comfortable are you having the very high statistical priest Sir Rory Collins holding the health of your heart in his calculator?