Risk Reduction

Absolute risk reduction is the event rate (number of deaths in the trial period) among the cases minus the event rate (number of deaths in the trial period) in the controls.

Relative risk is the event rate (number of deaths in the trial period) among the cases divided by the event rate (number of deaths in the trial period) in the controls.

Benefits and drawbacks of both absolute risk reduction and relative risk:

Relative risks tend to be larger numbers and are consequently more impressive. They are popular with drug reps and marketing organisations. Impressive sounding changes in relative risk can mask much smaller reductions in absolute risk. Absolute risk reductions tend be smaller and less impressive numbers. Absolute risk reduction has two main advantages over relative risk reduction as a summary statistic from a trial. Absolute risk reductions show the impact of the intervention while considering the overall risks to the population. The Number Needed to Treat (NNT) to prevent one outcome can be calculated from the absolute risk reduction, but not the relative risk.

Imagine if a treatment cut your risk of death from 20% to 10%. The relative risk is 50%; the absolute risk reduction is 10%. Then in another trial, the treatment cut the risk of death from 2% to 1%. The relative risk is still 50%, but the absolute risk reduction is 1%.