Dear stata users,
I am using Stata 17 to run competing risk survival analysis using Fine and Gray (FG) as well as flexible paramteric models on my breast cancer data. I want to perform the anaylses using „age as time scale“, instead of „time on study/survival time“ as a time scale.
I can run the models if I use „time on study/survival time“, but an error saying „initial values not feasible“ pops out when I try to run the FG model (also the flexible parametric model using „„age as time scale“. The following codes show how I stset the survival time (st) for the FG model:
stset iibc2age if lokseite!=3, fail(iibc==1) id (PAT_ID) enter(ageatdx)
Note: [iibc2age refers aged at exit (" iibc2age" refers to age at diagnosis of ipsilaterla invasive breast cancer (iibc), censoring or end of follow-up; "ageatdx" refers age at entry (diagnosis of ductal cell carcinoma insitu, and iibc==1 captures info about the occurence of iibc]
The stset is okay.
But when I run the multivariable FG model, then the error „initial values not feasible“ along with „r(1400);“ displays and no result appears in the result window. Please see the output below!
The same problem apprears for the flexible parametric model.
I would appreciate if anyone could explain me why, and suggest a solution for this problem.
Thank you!
Semaw Abera
I am using Stata 17 to run competing risk survival analysis using Fine and Gray (FG) as well as flexible paramteric models on my breast cancer data. I want to perform the anaylses using „age as time scale“, instead of „time on study/survival time“ as a time scale.
I can run the models if I use „time on study/survival time“, but an error saying „initial values not feasible“ pops out when I try to run the FG model (also the flexible parametric model using „„age as time scale“. The following codes show how I stset the survival time (st) for the FG model:
stset iibc2age if lokseite!=3, fail(iibc==1) id (PAT_ID) enter(ageatdx)
Note: [iibc2age refers aged at exit (" iibc2age" refers to age at diagnosis of ipsilaterla invasive breast cancer (iibc), censoring or end of follow-up; "ageatdx" refers age at entry (diagnosis of ductal cell carcinoma insitu, and iibc==1 captures info about the occurence of iibc]
The stset is okay.
But when I run the multivariable FG model, then the error „initial values not feasible“ along with „r(1400);“ displays and no result appears in the result window. Please see the output below!
The same problem apprears for the flexible parametric model.
I would appreciate if anyone could explain me why, and suggest a solution for this problem.
Thank you!
Semaw Abera
