Recently, I have started to analyse the results of a case-control study. The main idea of the project was to evaluate the association of SNPs of some gene products with risk of lung cancer among tobacco smokers.I have noticed that's most people use OR to describe the association, but for some reasons, I am wondering whether it is better to describe the results in term of RR instead of OR?
Depending on the method of control sampling (density or cumulative), and/or the rare disease assumption ,odds ratio estimates or equals the relative risk. So often one sees in reporting case control study results "OR (approx RR)" or "OR (RR)" or sometmes just RR.
Relative risk is easier to explain to a general audience and is more in line with a cohort.study result.
Thanks for the useful comment, I was not a ware about the method of control sampling.
I simply would present both, and give the raw data in supplementary material.
W
Sometime putting too much information may confuse readers.
I think putting the raw data in supplementary material is a good idea
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning