Measures of effect
Framing
- Treatment effect can be expressed through several related but distinct measures - relative risk/odds ratio (proportional effect), absolute risk reduction and number needed to treat/harm (absolute effect, often more clinically meaningful), hazard ratio (time-to-event effect), confidence intervals (precision/uncertainty of the estimate), and p-values (statistical significance) - correctly interpreting a trial result requires understanding which measure is being reported and what it actually conveys
- Relative risk measures can make a small absolute benefit appear more impressive than it clinically is, while absolute measures (number needed to treat) often give a more directly clinically useful picture of real-world impact
1 more section, plus exam facts
Premium unlocks every note across every specialty, and the full exam fact library behind it.
Get premium access