1
$\begingroup$

I am studying letality in cardio-vascular events. My outcome is binary (death at 1 month) for each event. My patients can have multiple cardiovascular events but obvioulsy only one of them can lead to death.

So for exemple a patient can have ten cardiovascular event the last one leading to death.

I have explaining variables at event, patient and county levels.

Therefore I used a mixed model logistic regression.

However I wonder if the fact that among my repeated events only one can lead to death (you obviously can only die once) is an issue.

Improve this question
$\endgroup$
3
  • 1
    $\begingroup$ Two questions: 1) What do you mean by multiple events? Do you mean that you measure death at multiple time points? 2) What do you mean by "among my repeated measures, only one can be positive?" I'm guessing you are referring to by-subject random effects, but its not clear from your question how you have modeled your mixed effects regression. It would be better to edit your question to clarify these points as well as explicitly lay your the syntax of your model. $\endgroup$
    – Shawn Hemelstrand
    Commented May 10, 2023 at 18:12
  • $\begingroup$ The survival , cox-model and kaplan-meier tags may be of interest to you. $\endgroup$
    – mkt
    Commented May 11, 2023 at 5:33
  • $\begingroup$ Thanks for the question, I clarified in the text of the question directly. Briefly my question is in a logistic mixed model, with more than one observation per individual using an individual random intercept is it problématic if the observation have this specific patern of never having more than on event per individual. $\endgroup$
    – Jason Shourick
    Commented May 17, 2023 at 14:22

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.

Browse other questions tagged or ask your own question.