Cholesterol & Heart Disease: How Much Evidence Do You Need?

For the longest time, the medical profession was not able to determine an(y) association between cholesterol and heart disease, much less cause & effect, at least not until the Framingham study was published.  Since then, multiple studies have been published demonstrating not only a link between high cholesterol and heart disease, but also that actively lowering cholesterol can decrease heart disease risk.  In fact, we now use the Framingham risk calculator in its various guises to help assess heart disease risk and determine LDL (bad) cholesterol goal.

Thus, having swallowed the Kool-Aid figuratively, even so far as to go back & ask for more, it never ceases to amaze me when patients ask for the latest cutting edge, nay, bleeding edge, testing to determine their risk. It's never clear whether they don't believe the Framingham data or whether they're looking for more enticement/rationale/proof before they change their lifestyle and/or start their medication.  They ask for Berkeley Heart Lab, VAP panel, NMR Lipoprofile, PLAC test, etc, by name thanks to fantastic advertising.

I'm one of the first to tell my residents to only order tests that will make a difference in their decision making. If the result won't change the decision, then why waste the money?  And since most, if not all, of the patients asking for these bleeding edge tests (none of which have randomized controlled trials demonstrating effectiveness, just observational studies demonstrating association & correlation) have yet to reach their LDL goal, I typically attempt to dissuade them from wasting their own money.  More importantly, what's the solution to an abnormal test?  More (powerful) statin and therapeutic lifestyle changes.  How is that any different from what we'd already recommended?

In a study published in last week's JAMA, the authors followed 165,544 patients w/o baseline cardiovascular disease for 10+ years.  The addition of apolipoprotein B & A1, lipoprotein(a) & lipoprotein-associated phospholipase A2 to traditional lipid panel led to net reclassification from intermediate risk to high risk of heart disease of less than 1% of those assessed warranting pharmacologic management.  In essence, 4,500 patients w/o known heart disease would need to be tested for additional cardiovascular markers (at great expense) and agree to take a statin for 10 years in order to prevent one extra event.  So while the study may prove these additional bleeding edge tests are statistically significant, I question whether they are truly clinically significant.

I suppose if your crystal ball was in working order and you knew that it would be you who'd be having a vascular event, you'd want the additional testing and you'd be willing to pay for it.  But it doesn't work that way.  We don't know who's going to benefit.  And we all to have to pitch in to an ever increasingly expensive health care system.  Like the law of diminishing returns that we learned about in Economics 101, at some point we have to look at the numbers and declare, "that dog don't hunt".

Follow @alvinblin
Tweet