Quantifying treatment effect: relative and absolute measures in clinical research

Quantifying the effect of treatment is crucial for clinical decision-making. Effect measures can be expressed in absolute terms, such as risk difference, number needed to treat, or number needed to harm, or in relative terms, including risk ratio, odds ratio, and incidence rate ratio. Although all are derived from the same data, they provide different perspectives: relative measures capture proportional changes, while absolute measures translate these into clinically meaningful terms. Confidence intervals (CIs) are essential to assess the precision of these estimates and to evaluate their statistical significance and potential clinical relevance. In this methodological note, we illustrate the use of these measures with examples from studies on venous thromboembolism, a clinical field where both benefits and risks of anticoagulant therapy (i.e., thrombosis and bleeding) must be carefully balanced. Clinicians and researchers should be aware of the strengths and limitations of each measure. A balanced interpretation, integrating absolute and relative metrics together with their CIs, is central to properly assessing treatment effects and to support evidence-based decision-making.