Hi-
I agree with Matthew’s point that we need tax dollars go toward unbiased research focused on the question of (a) what target LDL to shoot for, and (b) how aggressively we should modify treatment to achieve those targets.

I want to underscore that part of the challenge here is a near-universal failure to appreciate the difference between 2 enterprises, each requiring distinct intellectual tools.

One is the research that helps us to determine IF a drug is effective (i.e. drug effectiveness research, typically funded by drug companies). The other is research and the related analyses that help us determine how to frame guidelines for care and benchmarks for success in care.

A guideline or a benchmark is an intervention that changes health care practice and health systems. Studies that prove a drug is better than another one may help to inform how we think about guidelines and benchmarks, but they rarely offer a direct test of whether a particular guideliine or quality benchmark is ideal. They also do not test the full range of consequences, favorable and unfavorable, of devising a new performance indicator to judge the quality of a doctor’s care.

The major problem I obsesrve is that the federal, nonprofit and for-profit bodies charged with devising quality standards typically do not recognize the distinct nature of these 2 intellectual enterprises. They throw all their eggs in the drug trial basket, and rely on subspecialist experts who are very oriented to tweaking individual organ systems. This is an easy trap to fall into, in part because drug company-funded specialists are easily found (and easily funded; the two are related), and they often speak in simply, declarative sentences, like “this drug works” and “lower is better.”

So, unfortunately, many guidelines wind up being devised by an enthusiastic and well-paid cadre of specialists who conduct trials, and who have clinical trial experience but not much health system or health services expertise.

This is how one arrives at the weighty and to my reading, not-terribly-credible 2004 statement of the National Cholesterol Education Program. The actual language of that Statement included some appropriate intellectual caveats, but of course the tone of the Statement and the press coverage (and the NCEP’s press release) were aggressive.

I found it remarkable that on the very same day that the NCEP Statement came out (7/12/2004), the Archives of Internal Medicien published a major review that reached a diametrically opposite conclusion. A review of 25 randomized controlled trials of cholesterol reduction for secondary prevention stated “There were no data to determine whether lowering the LDL-C level to less than 100 mg/dL ( 2.59 mmol/L) was superior to lowering it to 100 to 130 mg/dL (2.59-3.36 mmol/L).”
(Wilt et al, 2004;164:1427-1436).

How can it be that a federal panel buoyantly endorses ever-lower cholesterol targets while a major review, published on the very same day, says that there is NO DATA?

That can only happen when there is a fundamental disagreement about what counts as legitimate data to use when proposing a new performance target for clinical practice. I favor a conservative standard of pretty strict scrutiny. The fact that a medicine is “good at reducing cholesterol” should not lead us to endorse a new, broad-based performance standard for health care. The protection against that kind of misstep probably should include funding new kinds of research, and a greater level of scrutiny for the process by which guidelines and peformance measures are developed.

Stefan