0
$\begingroup$

I have a regular case-control study where I can calculate the proportion of patients infected when exposed to say, UV vs regular light. For hypothesis testing, how do I decide which is a better measure - relative risk or absolute risk? I found this answer with comparison of the two approaches - How do you explain the difference between relative risk and absolute risk?

However, there were no guidelines on how to choose the right approach. Could you please help?

Improve this question
$\endgroup$
5
  • 2
    $\begingroup$ You cannot calculate either measure in a case-control study; you can only calculate an odds ratio. So please clarify. $\endgroup$
    – Frank Harrell
    Commented Nov 10, 2016 at 20:54
  • $\begingroup$ What do you mean? I don't come from a stats background. And my problem is not in epidemiology. But for each group, I have exact information about the number of subjects infected, out of total number of subjects in the study. I want to compare the two groups. I can calculate the ratio and difference of the proportion of subjects infected. Maybe there is another name for this problem. Please guide. $\endgroup$
    – Mr K
    Commented Nov 10, 2016 at 21:00
  • $\begingroup$ Study the different study designs. Case-control designs involve outcome-dependent sampling. With such retrospective designs you can't calculate certain quantities. $\endgroup$
    – Frank Harrell
    Commented Nov 10, 2016 at 23:57
  • $\begingroup$ @FrankHarrell Thanks for the advice, Frank. However, I would be grateful for a little more details. Also, if you could recommend some reading - books or research papers, that will be helpful. $\endgroup$
    – Mr K
    Commented Nov 11, 2016 at 0:00
  • $\begingroup$ See for example en.wikipedia.org/wiki/Odds_ratio $\endgroup$
    – Frank Harrell
    Commented Nov 11, 2016 at 0:06

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.