I am doing competing risk regression. I have three possible functions and one of them produces completely different results than other two. In my dataset I have time to event, status (censored=0, event of interest=1, death=2) and three variables which are all continuous.

The first one is riskRegression().

crr.1<-riskRegression(Hist(f_time, status) ~ var1+var2+var3, data = dataset, cause = 1,link="prop")

The second one is comp.risk().

crr.2<-comp.risk(Event(f_time, status)~
const(var1)+const(var2)+const(var3),data= dataset,

And the final is crr().

crrMat <- model.matrix(object = ~ var1 + var2 + var3, data = dataset)
crrMat <- crrMat[,-1]

crr.3 <- crr(ftime=dataset$f_time, fstatus=dataset$status,cov1=crrMat, failcode = 1,cencode  = 0)

The problems is that results from crr.1 and crr.2 are similar but result from crr.3 is totally different. They should be all doing the same thing?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.