hazard ratio

A hazard ratio compares how often a particular event happens in one group with how often it happens in another, over time. In cancer trials the event is usually death, or the cancer growing again, and the comparison is between people given a new treatment and a control group given the standard treatment or a placebo. The US National Cancer Institute defines it as a measure of survival at any point in time in one group against another.

The number is read against one. A hazard ratio of one means there was no difference in survival between the two groups. A figure above or below one means survival was better in one of them. A hazard ratio of 0.40 for a new drug means that at any given moment in the trial, people taking it were dying at 40 percent of the rate of people on the comparison treatment, which is usually reported as a 60 percent reduction in the risk of death.

It is worth being precise about what that does not mean. A 60 percent reduction in the risk of death is not a promise that anyone lives 60 percent longer, and it is not a cure rate. It describes the rate at which the event occurred across the trial period, which is why hazard ratios are almost always reported alongside the median survival times in each group, and alongside a confidence interval showing how precisely the figure was pinned down.