Estimating a treatment effect: Choosing between relative and absolute measures

Abstract

The size of a treatment effect in clinical trials can be expressed in relative or absolute terms. Commonly used relative treatment effect measures are relative risks, odds ratios, and hazard ratios, while absolute estimate of treatment effect are absolute differences and numbers needed to treat. When making indirect comparisons of treatment effects, which is common in multiple sclerosis (MS), having now many drugs tested in independent trials, we can have different figures if we use relative or absolute measures, and a frequently asked question by clinicians is which approach should be used. In this report, we will try to define these measures, to give numerical examples of their calculation and specify their meaning and their context of use.