An excellent review of how statistical deception was used by the pharmaceutical industry to make statin drugs appear safe and effective was published by Dr. David Diamond and Dr. Uffe Ravnskov in the journal “Expert Review of Clinical Pharmacology.” Dr. Diamond serves on the Medical Research Service in Tampa Fla., and the Center for Preclinical and Clinical Research on PTSD at the University of South Florida. Dr. Ravnskov works with the Department of Molecular Pharmacology and Physiology, also at the University of South Florida.

To understand how statistics are used to deceive, you need to learn a few terms. The first is absolute risk reduction (ARR). This represents the real difference between a treatment under study and patients exposed to a placebo (an inert substance).

For example, say half of the patients in a study are given a statin and the other half are given a placebo. If 2 percent of those taking the placebo suffer a heart attack and 1 percent of those taking the statin suffer  a heart attack, we say that the ARR is 1 percent  (2 minus 1).

That is, the statin is effective in 1 percent of the patients treated.

The next term you need to know is called number needed to treat (NNT). In the example above, with an ARR of 1 percent, you would need to treat 100 people to prevent one person from having a heart attack. (In the other 99, the drug would have no effect.) Therefore, the NNT is 100.

Finally, there is the deceptive method used for citing benefits. This is called the relative risk reduction (RRR). Doctors have been brainwashed during their training to only consider this number, and professional statisticians understand how it can be used to deceive even the heads of major university departments.

Representatives from institutions as prestigious as both Harvard and Yale medical schools have been fooled by this devious method.

Taking the same example above, instead of looking at the ARR — that is, 1 percent — to get the RRR we take the percentage difference between 2 and 1, which would be a 50 percent variance.

Therefore, using the RRR a researcher could say there was a 50 percent reduction in heart attacks among the patients taking statins.

Of course, that makes the drug sound a lot more effective than claiming a 1 percent absolute reduction in heart attack rates.