Research methods guide
Odds ratio, risk ratio, and hazard ratio explained
A risk ratio compares probabilities, an odds ratio compares event-to-nonevent odds, and a hazard ratio compares instantaneous event rates during follow-up. State the groups and time frame.
In a synthetic example, 20/100 in group A and 10/100 in group B have the event. These counts give a risk ratio and an odds ratio. A hazard ratio requires time-to-event and censoring information.
Calculate RR, OR, and risk difference from counts
A is compared with reference group B over the same observation period.
| Item | Worked example | Verification |
|---|---|---|
| RR | (20/100) ÷ (10/100) = 2.00. | State the population and time point, and describe the risk as twice the reference group’s risk. |
| OR | (20/80) ÷ (10/90) = 2.25. | Use nonevent counts in the denominators and label the result as an odds ratio. |
| Risk difference and HR | Risk difference: 20%−10%=10 percentage points. An HR cannot be estimated from this table alone. | An HR uses event times and risk sets; document the model and assumptions. |
Apply the method to your materials
Reconcile effect scales
Record group, event count, denominator, follow-up, scale, and interval. Verify whether counts are participants or episodes, and keep the reference direction consistent.
Trace adjusted estimates to their models
A crude 2×2 table does not reproduce adjusted ORs or HRs. Identify model, covariates, and analysis set and state which estimate is reported.
Questions that arise in practice
Can OR and RR labels be interchanged for rare outcomes?
Discuss an approximation where its conditions apply and retain the original scale in formal reporting. Absolute risk also needs a baseline and time horizon.
Verified sources
The official sources below were checked on 2026-11-12. Access conditions and update dates are rechecked before publication.
RigorMed workspace
Start with your research materials
Bring your research question, available materials, and the steps you want to check.
