Relative risk reduction

Relative risk reduction expresses treatment effect as a percentage of the starting risk, calculated as one minus the relative risk, or as the absolute risk reduction divided by the control event rate. It is the number most likely to appear in press releases because it stays large even when the underlying benefit is small. That is the reason to treat it cautiously. A drug can honestly claim a 50 percent relative risk reduction whether it cuts risk from 2 in 100 to 1 in 100 or from 40 in 100 to 20 in 100. When reading a study, never accept a relative risk reduction alone; find the baseline risk and convert it to an absolute figure before deciding whether the effect is meaningful. Ask whether the writer reported the relative number precisely because the absolute one looked unimpressive. The same value applies across different baseline risks, so it travels poorly between high-risk and low-risk populations.