How to calculate
Divide events by the number of people in each group, then divide the first risk by the second. The confidence interval applies a normal critical value to the standard error of the log risk ratio before exponentiating the bounds.
Worked example
Use Events in exposed group 80 · Exposed group total 1000 · Events in comparison group 10 · Comparison group total 1000 · Confidence level 95%. The result is 8.
Input and result notes
The risks must describe comparable follow-up periods. The interval uses the large-sample log method and is unavailable if either group has zero events.