A striking hazard ratio can sit on a handful of events. Before acting on a result, look at the absolute numbers in each arm, the width of the confidence interval and whether the primary outcome is a composite whose parts moved differently.
A ratio built on a few events can change substantially if two or three patients are reclassified, and a trial that is open-label with soft endpoints is more prone to bias than one with hard, adjudicated outcomes.
Ask three questions of any headline result: how many events, how wide is the interval, and what does the absolute difference mean per hundred patients treated.
- Convert relative risk into absolute difference per 100 patients before counselling.
- A wide confidence interval means the true effect could be much smaller or larger than the point estimate.
- Check what is inside a composite endpoint.
- Weigh open-label design when outcomes involve judgement.
Why it matters
Appraisal habits protect against acting on a result that looks larger than the data under it.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for cardiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free