[timr]

You should actually read the article. In particular:

> Fourteen trials recruited patients with specific conditions (raised lipids, diabetes, hypertension, microalbuminuria). All‐cause mortality was reduced by statins (OR 0.86, 95% CI 0.79 to 0.94); as was combined fatal and non‐fatal CVD RR 0.75 (95% CI 0.70 to 0.81), combined fatal and non‐fatal CHD events RR 0.73 (95% CI 0.67 to 0.80) and combined fatal and non‐fatal stroke (RR 0.78, 95% CI 0.68 to 0.89). Reduction of revascularisation rates (RR 0.62, 95% CI 0.54 to 0.72) was also seen.

So the evidence base is a collection of studies where most of the participants had at least one prior indicator of CVD or diabetes, and their outcome is a relatively weak benefit to all-cause mortality, CVD, CHD and stroke. For primary prevention, what you really want is a strong outcome in a study of people without any prior indication of disease [1].

I think the article posted by parent is exaggerating, but even the Cochrane review is pulling its punches here, saying specifically "cost-effective in primary prevention", instead of the stronger claim. Common jokes about putting statins in the water supply aside, there's not a ton of evidence for giving them to, say, otherwise healthy 20-somethings.

[1] Imagine the following, not-uncommon scenario: you have an otherwise healthy patient who is both pre-diabetic, as well as presenting with elevated cholesterol. Statins have a tendency to elevate blood glucose. So which risk do you choose?

The available evidence provides poor guidance.

[bluGill]

Careful. You are correct at what we want for primary prevent. However for primary prevention we need much larger sample sizes and thus data is much harder to get.

Lack of data doesn't mean the treatment won't work. There is plenty of reason to think statins work for primary prevention even though it hasn't been proved yet. For most the side effects are acceptable, and the cost is low. Thus for most it is worth trying as primary prevention even if we don't have data to show it works. Remember you are playing with your own life here, and the best evidence we have is on the side of stains for primary prevention - this may change in the future when we get data of course.

[timr]

> Lack of data doesn't mean the treatment won't work

In drug development, that is the default presumption, and rightfully so: almost nothing ever works.

> There is plenty of reason to think statins work for primary prevention even though it hasn't been proved yet.

Define "primary prevention" -- do you propose giving this to a healthy 20 year old with no other signs of illness? Younger? Should we "put it in the water", as they say? How about older patients? How old? Or, do you mean someone with symptoms? If so, then what about the case I cited (which is quite common in "primary prevention") where you have multiple things in tension?

The evidence provides no guidance here, and anyone who tells you otherwise is guessing. For what it's worth, though, we agree completely on the need for larger data. I think what drives me most batty about the "appeal to consensus" is that it's almost invariably used as a highbrow-lowbrow way of beating up on people who want to ask the question, which is the first step toward getting the answer!