life table for cumulative incidence rate

[Eun Um]

I tried to create life table for cumulative incidence rate with NRM (non-relapse mortality) by using the following code, I got the error message:
ltable nrmyear nrm, cif
option cif not allowed
(option is invalid or you need to specify graph)
r(198);

How can I fix it? Thanks.

[Matt Warkentin]

Hi Eun,

The answer is in the error Stata provided you. The argument 'cif' is not an acceptable one for the command ltable. In the future, I recommend consulting the help documentation for the command you are interested in (help ltable). Most issues of this kind can be solved that way.

The option that you want to specify is failure, for example...

webuse rat
ltable t died, failure

[Eun Um]

Thank you for your response. I have a question. Does this command taking into consideration of competing risks? Just failure = 1 - survival, that would not equal to cumulative incidence?

[Eun Um]

Failure indicates cumulative failure probability. Does it indicate Cumulative incidence?

[Matt Warkentin]

No, --ltable-- does not account for competing risks. You can use the --stcrreg-- and --stcurve, cif-- commands to plot cumulative incidence functions if you have competing risks data.

[Eun Um]

Thank you for explaining this to me. I really appreciate it.

I was able to generate the CIF of NRM without covariate and then with KPS by using the following code.

. stcrreg kps, compete(nrm==2)
stcurve, cif
. stcrreg kps, compete(nrm==2)
. stcurve, cif at1( kps=0 ) at2( kps=1 )


Now the question is: how do I generate CIF of NRM at 2-yr with 95% confidence interval after I have the curve? In other words, how to generate a life table for this curve?

[Marco Cerrano]

Hi,
I faced the same problem. In medical research estimating X-year or month Cum incidence of relapse, NRM, etc.. is often absolutly necessary, and without this option we are forced to use other programs (if we are able to!). Is there any way to solve this problem?
Thank You!

[arcita pramudita]

I tried to create life table for cumulative incidence rate with NRM (non-relapse mortality) by using the following code, I got the error message:
ltable nrmyear nrm, cif
option cif not allowed
(option is invalid or you need to specify graph)
r(198);

How can I fix it? Thanks.

Hello Eun Um, i face the same problem. Did you find a way to have confidence interval for CIF in stcrreg? Thanks