An Introduction to NNT

The key to understanding the potential benefit of any preventative drug or procedure is to focus, not on the widely publicized measures of relative risk reduction, but on the less well known measures of absolute risk reduction.

The difference can be striking. A recent study of the benefits of an anti-cholesterol drug, for example, indicated close to a 30% reduction in heart attacks experienced by those using the drug daily over the multi-year study period.  The 30% is a relative risk measure which, on the surface, appears substantial.

But there’s another way to interpret the study’s result. For every 100 people taking or not taking the drug, 93 wouldn’t have had any heart attacks.  Of the remaining seven, five would have had a heart attack whether they were taking the drug or not.  Effectively, that meant that the absolute reduction in heart attack risk for the drug’s users was two people out of a hundred.

Put another way, for every 50 people taking the drug, one less person would have had a heart attack.  In other words, 49 out of 50 people would be paying for an expensive drug (and subjecting themselves to possible side effects), but would not statistically benefit.

Over twenty years ago, epidemiologists developed a simple measure of absolute risk reduction known as the “Number Needed to Treat,” or “NNT.”  It represents the total number of people who must receive a given drug or procedure for any one individual to benefit.  In the example above, the NNT would be 50.

Unfortunately, NNT has received little attention.  More commonly, we see full page magazine drug advertisements, complete with a smiling models, extolling the relative benefit of various drugs.  The same ads typically include one or more pages of small print warnings and disclaimers — but no NNT measure of the drugs’ absolute effectiveness.

This Web site is dedicated to promoting a better understanding of NNT with the goal of: