Compare event rates across groups
Build comparable event counts and time at risk, estimate a crude rate ratio, and check whether differences in population composition call for a standardized comparison.
Worked decision
Suppose group A records 12 events over 600 person-years and group B records 9 over 300. The rates are 2 and 3 per 100 person-years. The crude rate ratio is 0.67. Check follow-up definitions, recurrent events and group differences before interpreting the contrast.
Record your decision
Which comparison answers the study question? Use a crude ratio to compare the observed group rates. If composition matters, choose common reference weights for direct standardization or external rates for an observed-to-expected comparison.
Work through the evidence
- Define events, time at risk and rate units
- Check the assumptions behind a Poisson rate ratio
- Calculate the crude rate ratio and approximate interval
- Apply reference weights to stratum-specific rates
- Compare observed events with an external reference
- Download the rate-comparison worksheet (TXT)
- Compare interval methods for independent Poisson rates