I am implementing a Fine & Gray competing risk model. The model includes covariates that are static for each person (e.g. location) as well as covariates that change over time (such as local unemployment rates).
The issue I have is: how do you generate cumulative incidence curves, assuming covariates that change over time? It is straightforward to use the STATA command "stcurve" and generate a CI curve that assumes fixed values for covariates (e.g. covariate value X for all times), but there is no functionality to generate a CI curve assuming covariate value X in time t, and then covariate value Y in time t + 1.
Is there a paper that discusses this, or a package in R/SAS/STATA that lets you generate curves assuming different values of covariates over time? Thanks very much for any help!
