Monday, December 16, 2013

New cholesterol guidelines: a brave and wise departure from current practice?

Well, that's the claim made in a recent BMJ editorial on the new guidelines (HT to DB's Medical Rants). To recap, the emphasis of the new guidelines was to treat with fixed doses of statins based on patients' risk rather than titrating to goal. While this represents a shift from the old recommendations it is not a radical departure from current practice. The “set it and forget it” approach offered by the new guidelines was, I suspect, already common in clinical practice. Moreover, based on accumulating research evidence it is more common nowadays for patients suffering acute vascular events to be discharged from the hospital on high statin doses, without much attention to their baseline lipids, similar to those recommended in the new guidelines. What the document did, though, was codify this practice which, one would hope, will cause it to be more uniformly and systematically applied. So this will not result in the “tectonic shift” that the BMJ author has claimed.

Then he goes on to say this:

A spate of studies over the past few years contributed to an insight that had previously eluded the field. Cholesterol plays a key role in atherosclerosis, but its modification by drugs does not always produce the expected result. Drugs have thousands of effects and their influence on cholesterol concentrations does not convey their net effect on patient risk. Trials showed that lowering LDL-C and raising high density lipoprotein cholesterol did not necessarily lower risk.8 9 10 11

Those trials did not ask the simple question of whether pharmacologic alteration in lipid levels would reduce events. With the exception of one study which may be an outlier because it involved a novel class of drugs these trials essentially addressed whether adding a second drug would address residual risk after statin therapy. Older studies which did ask the simple question supported the lipid hypothesis and provided evidence that altering LDL and HDL cholesterol levels leads to a reduction in important clinical events. I reviewed those studies in yesterday's post.

I did find some good points in the editorial. For example the author makes this point which is fundamental in the application of clinical trials to the principles of evidence based medicine:

Also, the risk thresholds for treatment should be understood as recommendations and not dictums. For any individual, the decision about the worth of a drug treatment depends on how that person feels about potential benefits, burdens, and harms—something that no writing group can determine. Ultimately, the decision depends on our ability to provide patients with the knowledge and guidance needed to make high quality decisions about their treatment.

Then he makes this point about performance measures:

The new guidelines are also a cautionary tale about the premature translation of medical guidelines into performance measures. They highlight that the push to chase targets was based on speculation, not on direct trial evidence, and opened the door to the use of drugs that had not been fully tested.

Performance measures, while they appeal to guidelines and research evidence, do so without appropriate critical analysis and expertise which is one reason they ultimately fail.

No comments: