Wednesday, July 15, 2009

More on Statins II: What's the Message; Where's the Harm?

In response to my earlier post on the latest statin meta-analysis in BMJ:

--thoughtful and faithful reader Marilyn Mann added some challenging comments which I suggest that you review. She alluded to the PROVE IT study as possible evidence that some statins may be better than others and generally supporting the cholesterol-lowering hypothesis of how statins work, which I had been calling into question.

This offers an opportunity to do two things--first, explain better what I believe that my intended message is when I keep harping on statins; and second, share a bit of information that happened to come my way today about the sorts of harm that unbalanced drug industry marketing about heart disease prevention causes.

What I think Marilyn may have been responding to in part was what she perceived to be my going too far out on a limb to argue a contrarian case re: statins. So let me back up and state as clearly as I can what I think I am and am not trying to say.

Things I do not mean to say:
  • Patients who have some risk factors for heart disease, and have high cholesterol, should not take statins.
  • We have proven that statins are no good for primary prevention of heart disease.
  • We have proven that the cholesterol-lowering hypothesis of why statins work is bogus.

Things I do mean to say:
  • The contrarian view, that maybe statins do not work as well as they are cracked up to be, and that their efficacy depends on an effect separate from their cholesterol-lowering effect, is a view that should be taken seriously under consideration (the people saying those things are not crazy).
  • The amount of evidence favoring statin use for primary prevention, and the value of cholesterol screening and adjusting statin dose carefully to achieve a target cholesterol level, may have been seriously overstated in the portion of the literature influenced by drug industry marketing, and in the associated practice guidelines.
  • Before patients are placed on statins for primary prevention, they deserve at least to be told the number needed to treat (NNT) statistics as well as the costs and adverse effects of statins. Told the NNT, as Marilyn indicates, a number of patients would no doubt elect to take statins--but at least some, and perhaps many, would not. (Like many other aspects of medicine, the right answer is shared decision-making.)
So I apologize if I have been overly enthusiastic in urging the contrarian view; the excess enthusiasm is a measure of how effective the industry marketing juggernaut has been in selling the viewpoint that pushes the most drugs.

Now, you might ask, where's the harm, since fortunately serious adverse reactions from statins remain quite rare? Thanks to my friends Rick Bukata and Jerry Hoffman and their "Primary Care Medical Abstracts" program, I learned today of a recent editorial by Bethell et al. decrying the widespread failure to use cardiac rehabilitation programs post-heart-attack, despite the mountains of evidence supporting the use of such programs. In turn Bethell and colleagues cite a meta-analysis published by Taylor et al. in 2004. Taylor looked at 48 randomized trials of cardiac rehab involving 8940 subjects. Now, Taylor and company reported their findings primarily by odds ratios and not by NNT, so you have to do a couple of fishy things from a statistical point of view to generate an estimated NNT for the mass of studies as opposed to any individual study. Looking at their summed numbers for outcomes in the rehab vs. control groups, I calculated the following estimates: NNT is 62 to prevent one death from any cause; 45 to prevent 1 cardiac death; and 143 to prevent one heart attack. (Sharp-eyed readers will ask: you need to treat that many patients for how long to prevent one bad outcome? and the answer is that it varies as the studies had different lengths of follow-up; but most formal cardiac rehab programs run only for a certain number of weeks, so they are not like pills that you have to take for the rest of your life. The benefits achieved from the cardiac rehab programs tend to substantially outlast the formal length of the program itself.)

Why do I bring this up at all? Cardiac rehab seems to me to be one of many poster children for the "anti-reps"-- the reps that do not show up at the average doc's office trying to sell that treatment and offering the staff free food in the bargain. So what is true of cardiac rehab in patients who have had heart attacks or similar coronary events?

  • The efficacy is demonstrated in a large number of studies
  • The cost is minimal to moderate, and is reimbursed routinely by Medicare
  • There are hardly any adverse effects
  • The NNTs, as a measure of the effect size, generally look much better than NNTs for statins--certainly better than the NNTs seen in the PROVE IT trial to be specific.

Our rush to add statins to the water supply, at the same time as we routinely forget to order cardiac rehab for patients whose lives might be saved by that well-proven but nonpharmacologic intervention, highlights how heavily skewed medical practice has become under the influence of Pharma marketing.

Bethell HJN, Lewin RJP, Dalal HM. Cardiac rehabilitation: it works so why isn't it done? British Journal of General Practice 58:677-79, 2008.

Cannon CP, Braunwald E, McCabe CH, et al. Intensive vs. moderate lipid lowering with statins after acute coronary syndromes. New England Journal of Medicine 350:1495-1504, 2004. [PROVE IT trial]

Taylor RS, Brown A, Ebrahim S, et al. Exercise-based rehabilitation for patients with coronary heart disease: systematic review and meta-analysis of randomized controlled studies. American Journal of Medicine 116:682-92, 2004.

No comments: