Healthy Skepticism Library item: 16181

Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.

Publication type: Journal Article

Sprenger M.
Statistics for health: Let’s start with reporting
BMJ. 2009 Jul 28; 339:
http://www.bmj.com/cgi/content/extract/339/jul28_3/b3035

Abstract:

Clinical epidemiology is now well established. Therefore no article should be published without mentioning the number needed to treat. This is especially important when dealing with issues around medical prevention, such as statins.1

Gigerenzer et al recommend using frequency statements instead of single event probabilities, absolute risks instead of relative risks, death rates instead of survival rates, and natural frequencies instead of conditional probabilities.2

They also say: “reporting relative risk reductions without clearly specifying the base rates is bad practice because it leads readers to overestimate the magnitude of the benefit. Consider one medication that lowers risk of disease from 20% to 10% and another that lowers it from 0.0002% to 0.0001%. Both yield a 50% relative risk reduction, yet they differ dramatically in clinical importance.”

The communication of risks or therapeutic effects is a critical and responsible task. So why is a meta-analysis published without the number needed to . . .