The Illusion of Certainty – Part II.

Last time I introduced the term illusion of certainty and our need for easily understood information before making decisions.  The following example should help—hopefully.

In a study published in 1995 about cholesterol-lowering drugs, it was reported that Pravastatin [Pravachol] reduced your risk of dying by 22%.  This 22% is known as the Relative Risk Reduction and sounds impressive but what’s it really mean?

Most people interpreted this to mean that 220 people out of 1000 taking the drug would avoid having a fatal heart attack.  Wrong!  The actual findings showed that out of 1000 people taking pravastatin, 32 died, compared with 41 who died taking a placebo—a difference of only 9 per 1000, or 0.9%.  This 0.9% is called the Absolute Risk Reduction.

Ignore the statistic Relative Risk Reduction [RRR].  It is typically reported in journal & newspaper articles but is misleading because it amplifies small differences and makes the insignificant appear significant!  In the study above, 22% is inaccurate.

Look for the Absolute Risk Reduction [ARR], which measures the effect of treatment in terms of the absolute number of people improved; and better yet, ask for the Number Needed to Treat [NTT], which refers to the number of people who need to be treated for one to be improved.  The pravastatin study showed a NNT of 111—in other words, 111 have to be treated to prevent one death.  The smaller the NNT the better.  Authorities suggest a NNT less than 20-50 is needed for a treatment to be cost-effective and beneficial.

Bottom line:  the benefit of taking pravastatin is not as great as you are being (mis)led to believe.  Which means you could focus on more natural treatments, such as high fiber diet, almonds, oatmeal, and exercise, to achieve similar benefits without the cost and risk of side-effects from the drug.

For trustworthy stats and information, go to Oxford University’s Patient Decision Aids (;  my website which also has a link to it [].