In an article published in the Cochrane Review, researchers have shown that people don't really understand statistics. That's not news, but now we have proof that our poor grasp of stats can be bad for our health. Statistics have a huge influence on a how effective people think drugs are - except, of course, it's the wrong influence. For example, in a case where a drug lowers the risk of contracting a disease, there are four ways to describe its effects. According to the researchers, these are:
•Relative Risk Reduction (RRR): there's a 50% reduction in risk
•Absolute Risk Reduction (ARR), the risk has fallen from 1% to 0.5%
•Number Needed to Treat (NNT), 200 people need to be treated to prevent one occurrence
•Natural Frequency, 1 out of 200 people will be helped by this drug
In a meta-analysis of 35 studies, the one description that both consumers and doctors most reliably understood was frequency, and the least accurately comprehended was the relative risk reduction, which usually lead to individuals thinking a drug was far more effective than it actually was.
The researchers are calling for results to be published as a frequency rather than a probability, but I somehow doubt it'll have much influence on drug marketing.