Hello, I am working on a question involving survival time in the presence of competing risks, which is something I haven't worked with before.
I am looking for a quick sanity check, as in my study the outcome event of interest IS the failure variable. In this case once the data is stset, the failure variable as an outcome has already been considered as the "event" and should not be included in the regression model, correct?
Thefore, if running a regression for the effect of subject sex on the occuence of the event, the code should look like:
Instead of:
As the interpretation of these outputs would suggest the risk/effect event==1 has on the occurence of event==1, correct?
Sorry if this seems obvious, just hopinh to confirm before I get too deep one way or another.
Thank you!
I am looking for a quick sanity check, as in my study the outcome event of interest IS the failure variable. In this case once the data is stset, the failure variable as an outcome has already been considered as the "event" and should not be included in the regression model, correct?
Thefore, if running a regression for the effect of subject sex on the occuence of the event, the code should look like:
Code:
*Cox stset time, fail(event==1) id(id) scale(365.25) stcox sex *Fine & Grey stset time, fail(event==1) id(id) scale(365.25) stcrreg sex, compet(event==2)
Code:
*Cox stset time, fail(event==1) id(id) scale(365.25) stcox ev sex *Fine & Grey stset time, fail(event==1) id(id) scale(365.25) stcrreg ex sex, compet(event==2)
Sorry if this seems obvious, just hopinh to confirm before I get too deep one way or another.
Thank you!
Comment