The (relative) risks of using odds ratios.

Odds ratios are commonly presented in the medical literature, including dermatology journals. Even when used appropriately, odds ratios are often difficult to interpret. This article illustrates this problem using an example from the recent dermatology literature. It then reviews the definitions of odds and odds ratio, as well as how odds and odds ratio relate to probability and relative risk. The divergence of odds ratios from relative risks when events are common (occurring in > or =10% of a sample) is explained. Methods to convert odds ratios to relative risks (and the reasons why conversion should be considered) are discussed.