Relative vs. absolute measures of benefit and risk: what’s the difference?

OBJECTIVE: When appraising evidence clinicians are confronted with two types of comparisons: ratios, such as relative risk, and absolute differences, such as number needed to treat (NNT) or number needed to harm (NNH). METHOD: A review of the definition, calculation and interpretation of relative measures such as relative risk, odds ratio and the hazard ratio, and how they are different from absolute measures such as NNT and NNH. RESULTS: Relative and absolute measures provide different perspectives. Ratios can be misleading and exaggerate clinical differences, but NNT can appear to trivialize the risk of potentially important adverse events. CONCLUSION: There is a need to understand both relative and absolute differences in order to make informed decisions.