Re: Adverse effects of statins
I am in favour of the BMJ’s decision not to retract these two articles. The magnitude of the error (18% versus 20%) is so tiny that it should have deserved only a letter to the authors. Any way, there was a much bigger reason to criticize both the analysis of Abramson and colleagues and the meta-analysis of CTT. Indeed, the first did not say a word about the MEGA study (Lancet 2006 and Circulation 2008 for post-hoc analysis in women), and the second have incorporated the results of the MEGA study in their meta-analysis without saying that it was the only study conducted in a population (the Japanese) characterized by a very low risk of death from ischemic heart disease and stroke (Within the OECD countries the lowest rates are observed in Japan, Korea and France). It is not necessary to do a meta-analysis mixing so heterogeneous data when a trial done with the least powerful statin on LDL-cholesterol (pravastatin) at its lowest dosage (10 mg daily) yields such impressive results (relative risk reduction after 5 years follow-up were 32% for all cause mortality, 48 % for myocardial infraction, 30% for strokes).
Competing interests: No competing interests