- Published
- Updated
A relative effect describes a ratio, while an absolute effect describes a difference on a specified scale. Similar relative changes can produce different absolute differences at different baseline risks; neither measure should replace the other without context.
Define the question and assumptions
Explain numerators, denominators and assessment times. Distinguish risk ratios, rate ratios and hazard ratios as different quantities. When deriving an absolute difference, retain the population and time associated with the underlying risk. Discuss limitations before transferring an estimate directly to a population that was not compared.
Hypothetical study example
In a hypothetical report, relative reduction appears large but events are uncommon and the absolute difference is small. The team presents both measures over the same time horizon. Confidence intervals and data sources accompany the numbers rather than turning them into an exaggerated statement of success.
Design review checks
- Confirm that absolute and relative summaries use matching populations and times.
- Make denominator changes and unequal follow-up durations visible to readers.
- Name the ratio correctly rather than using different ratio measures interchangeably.
Record the decision and its limits
Accepted presentation uses consistent direction and numerical precision. Rounding should not erase or inflate small differences. Do not choose an effect scale merely because it appears more impressive. Discuss clinical meaning and uncertainty without turning a trial estimate into an individual treatment promise; generalisability needs separate assessment.
Frequently asked questions
- Is relative reduction sufficient on its own?
- No. Baseline risk, time horizon and absolute difference provide essential context.
- Is a hazard ratio a risk ratio?
- No. Their definitions and interpretations differ and should not be interchanged.