To understand the medical claims that statins reduce cardiovascular disease risk by, say, 20% or 30% and side effects (a.k.a. adverse effects) are supposedly not statistically significant, you have to understand the way risk (or benefit) is publicized. Here is a simple illustrative example using hypothetical numbers..
Benefits:
Suppose two groups of people are monitored for 5 years. 500 people are given a placebo and 500 people are given statins. In the placebo group 4 people have heart attacks by the end of 5 years. In the statin group only 3 have heart attacks. This is reported as a 25% reduction in heart attack risk (4 reduced to 3). This is called relative risk reduction (benefit). Sounds great! But it is one person who benefited out of 500. That is a 1/500 = 0.2% absolute risk reduction.
Adverse Effects.
Lets look at the risk of developing diabetes, for example, from statins (a documented side effect). Lets run the same observation. In the 500 person placebo group 3 people may get diabetes in 5 years. In the 500 person statin group 4 people might get diabetes. If you question your doctor, she would probably say that your additional risk of developing diabetes in 5 years by taking a statin is only 1/500. That is just 0.2% absolute risk increase. Why worry?
But that is the same 1 person statistical difference in both cases. If the diabetes risk were explained in relative risk terms that would an increase from 3 to 4. That is actually a 33% increase in relative risk of getting diabetes. Wow!
—
The point is that benefits are often measured one way and risks are measured another way. Anyone with integrity must provide data using the same method for both (absolute risk is best, in my view).
return to the welcome page