Relative risk

Relative risk, also called the risk ratio, compares how often an outcome develops in two groups over the same period. It is found by dividing the risk in the exposed group, meaning the proportion who develop the outcome, by the risk in the unexposed or comparison group. For example, if 20 of 100 smokers and 5 of 100 nonsmokers develop a disease, the relative risk is 0.20 divided by 0.05, or 4: the smokers were four times as likely to develop it.

A relative risk of 1 means the risk is the same in both groups. A value above 1 points to a higher risk with exposure, suggesting a risk factor, and a value below 1 points to a lower risk, suggesting protection, as with a vaccine. Results are reported with a confidence interval, and if the interval includes 1, chance alone could explain the difference.

Because it needs true risks, relative risk comes from cohort studies and randomized trials, which follow people forward from exposure. Case–control studies use the odds ratio instead, which approximates the relative risk when the disease is rare. A large relative risk can still mean few extra cases if the outcome is uncommon, so it is often reported alongside the absolute difference in risk.