Overview
Relative risk compares how often an outcome happens in an exposed group with how often it happens in a control group. It's widely used in medical and public health studies.
Enter cases and totals for both groups.
Worked examples
| Inputs | Result |
|---|---|
| Exposed Cases42Exposed Total180Control Cases21Control Total200 | Exposed risk: 23.33%Control risk: 10.5%Relative risk: 2.22Absolute risk difference: 12.83% |
Show 2 more examples
| Exposed Cases10Exposed Total1,000Control Cases5Control Total1,000 | Exposed risk: 1%Control risk: 0.5%Relative risk: 2Absolute risk difference: 0.5% |
| Exposed Cases30Exposed Total100Control Cases40Control Total100 | Exposed risk: 30%Control risk: 40%Relative risk: 0.75Absolute risk difference: -10% |
Exposed risk by exposed total
With the calculator's starting values, the exposed risk is 23.33%. The table shows how it changes with the exposed total.
| Exposed Total | Exposed risk |
|---|---|
| 100 | 42% |
| 125 | 33.6% |
| 150 | 28% |
| 175 | 24% |
| 180starting value | 23.33% |
| 200 | 21% |
| 225 | 18.67% |
| 250 | 16.8% |
Other inputs: Exposed Cases: 42 · Control Cases: 21 · Control Total: 200. Change them in the calculator above for your own numbers.
How it works
Each group's risk is cases ÷ total. Relative risk is the exposed risk ÷ the control risk. 1 means no difference, above 1 means higher risk with exposure, and below 1 means lower. The absolute risk difference shows how many extra cases there are per 100 people.
The formula
Practical tips
- 42 of 180 exposed against 21 of 200 controls is a relative risk of 2.22, with a 12.83-point difference.
- A doubled risk sounds dramatic, but if the base risk is 1 in 10,000 the absolute change is tiny.
- Number needed to harm is 1 ÷ the absolute risk difference, here about 8.
Frequently asked questions
How is relative risk different from the odds ratio?
They're similar for rare outcomes, but differ when outcomes are common.
Does relative risk prove causation?
No. Confounding factors can create an association.
What study designs use it?
Cohort studies and randomized trials, where you follow groups forward in time.