We are reporting a study looking at mortality difference between two groups in an observational cohort and have reported absolute difference in % and odds ratio (95% CI), with P. However it has been suggested we use mean difference instead of odds ratio. What if any advantages does this confer?
James Leigh
The mortality difference may be more relevant if you are assessing the impact of a preventive measure.
You should not need odds ratios as you can directly measure rate ratios in a cohort study.
rate ratio would be the better measure for strength of association with a risk factor.
2 votes
2 thanks
- Badges
- Science topic
- Similar topics
- Education
- Higher Education
- Conferences
Similar questions and discussions