GPs should consider offering statins to all patients aged over 75, researchers sayBMJ 2019; 364 doi: https://doi.org/10.1136/bmj.l522 (Published 01 February 2019) Cite this as: BMJ 2019;364:l522
All rapid responses
I have several concerns about this paper.
Firstly, it appears to be a re-analysis of trials to look for associations for which the original trials were not designed.
Secondly, although repeated requests have been made for the raw data from some of them to be made available for independent analysis, this has never been done and I am puzzled how the promoters of statins can hold a monopoly.
Thirdly, only 14 of the 28 trials appear to be eligible for inclusion in the re-analysis.
Fourthly, there seem to be some errors in calculation; the reported percentages for deaths do not match the quoted absolute numbers of subjects.
Fifthly, the description of the statistical methodology seems to be designed to obfuscate and confuse and it is unscientific to reappraise trials in this way.
And, finally, the data that are presented anyway make a very weak case for statins. The overall absolute risk reduction for mortality – which is shown in the paper – is just 0.1%. Failing to quote absolute reduction (ARR) and trumpeting a large relative risk reduction (RRR) is misleading and grossly overstates benefit. If the RRR is 50% (which sounds impressive) but is based on an initial absolute risk of 5%, the ARR is only 2.5% (which is not impressive at all).
In any event the current algorithm for statin prescription will result in all patients over 75 being eligible for statins, regardless of their past cardiac history or cholesterol level. So the whole thing is a nonsense. Statins do work, but not very well, and it is highly likely that their effects are due to their anti-inflammatory properties and not to the fact that they lower cholesterol. If it was then one would expect a drug that has a dramatic effect on cholesterol, such as a PCSK-9 inhibitor, to have a dramatic effect on cardiovascular mortality. But the one major trial showed nothing of the sort, and indeed was terminated early. There are other observations which serve as Black Swans that destroy the cholesterol-heart hypothesis (which anyway is based on flawed research) and it should be abandoned.
Competing interests: I am a member of the International Network of Cholesterol Skeptics (THINCS)
This article reports that Colin Baigent said the following at the press conference, on January 30th, to launch The Lancet paper:
“Only a third of the 5.5 million over 75s in the UK take statins and up to 8000 deaths per year could be prevented if all took them.”
Someone at the press conference should have asked: How can this be correct when Figure 5 reported no statistically significant result for vascular deaths in over 75s (even after excluding trials where statins had “not been shown to be effective”) and Figure 4 reported no statistically significant result for major vascular events in over 75s without vascular disease?
Competing interests: I have written and published content about cholesterol and statins.
I would like ask Colin Baigent one question on this study - at this time. He claims that the Lancet study was a meta-analysis of twenty eight RCTs. The study was called. 'Efficacy and safety of statin therapy in older people: a meta-analysis of individual participant data from 28 randomised controlled trials.'
However, in the Appendix to the Lancet paper it is made clear that five of the studies are a comparison of high dose vs. low dose statins. PROVE-IT, A to Z, TNT, IDEAL and SEARCH. They cannot be used to test the hypothesis that statins are beneficial in the over 75s vs. placebo, as they were not done to answer this question.
Also, in nine of the RCTs used in the meta-analysis there were 0% participants over the age of 75 at the start of the study. These were 4S, WOSCOPS, CARE, Post CABG, AFCAPS/TexCaps, ALERT, LIPID, ASPEN and MEGA.
Which means that five of the studies could not address the question of statins vs placebo in the over 75s, and nine of the studies had no participants over the age of 75, which leaves fourteen studies that would be relevant to the issue of prescribing statins in the over 75s.
My question is, why did you call this study 'Efficacy and safety of statin therapy in older people: a meta-analysis of individual participant data from 28 randomised controlled trials.'
Competing interests: I have written two books on statins. The Great Cholesterol Con and A Statin Nation..
GPs should indeed consider offering a statin to people aged over 75 years. They should also ensure they enter a code for “statin declined” when patients refuse to take a statin to make clear they have offered a statin but that is was the patient’s decision to refuse to take one.
Competing interests: I am a part-time general practitioner in an NHS general practice.