More Smoke and Mirrors on Statins, Brought to You by Pfizer

An esteemed colleague called my attention to:
http://content.healthaffairs.org/content/31/10/2276.full
--in which a team of authors, who either work for Pfizer or who mostly have extensive financial relationships with the drug industry, explain to us that prescribing statin drugs is a huge economic plus for society. According to their calculations, prescribing statins aggressively even for people at low risk for cardiovascular disease costs the US each year about $305B and yields health benefits worth $1.25T. So naturally they recommend revising guidelines to recommend statins for more people and in even higher doses, and more vigorous use of pay-for-performance incentives for docs to get on the ball and prescribe more statins.

Sounds like a no-brainer, right?

Of course in this blog I've been harping on all the accumulated evidence that statins have been way overhyped and are probably of little use in all groups except those with already-diagnosed cardiovascular disease. So what's the deal?

It turns out that a careful reading of the article shows that Exhibit A for the huge success of statins, according to which the authors then calculated their economic benefits, was the CTT meta-analysis which I have previously alluded to: http://brodyhooked.blogspot.com/2012/05/statins-in-water-supply-continued-why.html

In that earlier post I took advantage of the expertise of David Newman to explain why the CTT analysis basically answered the wrong question--not, what happens if you put people in various risk categories on different doses of statins; but what happens if you prescribe statins and the result is that a patient's LDL level drops by a certain number of points. If you do an analysis the latter way, then statins appear to be marvelously useful, but as explained in the earlier post, that doesn't answer the doc's question about whether it's good to start the statin in the first place. If you answer the question in the former manner you get the more pessimistic numbers that show no real benefit except in the highest-risk groups.

Here's the other kicker in the more recent economic analysis, this amazingly bald statement near the begimnning of the article: "Importantly, we do not account for possible side effects in our social value calculations below." Come again? If statins make half the people who take them sicker than dogs, that doesn't count as a social cost--so your number-crunching simply assumes that statins are completely free of side effects?

The authors try to finesse this by noting that statins cause low levels of adverse reactions in many of the large-scale studies of efficacy--studies for the most part sponsored by the drug industry. This ignores the accumulating evidence of serious adverse reactions, even though most of them are not deadly, thank goodness. Consider for example a research letter by Golomb and colleagues (subscription required), noting from randomized trial data that as many as 4 in 10 patients might have worsening of fatigue and energy loss when taking statins; or other recent evidence raising the question of memory impairment as a consequence of statin use.

So what are we to conclude? First, this recent economic "analysis" is incomplete for omitting adverse reactions, and is probably deeply flawed by being based on a flawed meta-analysis of trial data that asks the wrong question. Second, this is how industry marketing works. First, get a study that's methodologically weak, but that contains the message your marketers wish to convey, out there in a prestigious journal (in the case of CTT, The Lancet). Second, keep churning out more and more articles conveying the same marketing message, referring back in each case to the first article, and hope that the journal's prestige will prevent any readers from questioning the methodology closely. Third, you can prove anything you wish in a cost-benefit analysis, just be sure to make the assumptions that tip the scale your way.

Golomb BA, Evans MA, Dimsdale JE, White HL. Effects of statins on energy and fatigue with exertion: results from a randomized controlled trial. Archives of Internal Medicine 172:1180-82, August 13/27, 2012.