Hi there,

I have a question regarding survival data

My data is as follows:

Code:
* Example generated by -dataex-. To install: ssc install dataex
clear
input byte ID int weeks long Gender byte VeinPat
1   0 1 .
1   2 1 1
1   4 1 1
1  12 1 1
1  24 1 1
1  48 1 0
1  96 1 0
1 144 1 0
1 192 1 0
1 240 1 0
2   0 1 .
2   2 1 1
2   4 1 1
2  12 1 0
2  24 1 0
2  48 1 2
2  96 1 2
2 144 1 2
2 192 1 2
2 240 1 2
3   0 2 .
3   2 2 1
3   4 2 1
3  12 2 1
3  24 2 1
3  48 2 1
3  96 2 1
3 144 2 1
3 192 2 1
3 240 2 2
4   0 2 .
4   2 2 1
4   4 2 1
4  12 2 1
4  24 2 1
4  48 2 1
4  96 2 2
4 144 2 2
4 192 2 2
4 240 2 2
end
label values Gender GenderNew
label def GenderNew 1 "F", modify
label def GenderNew 2 "M", modify
Time variable is weeks; the event is "VeinPat" and it can be either 0 which means failure; 1 which means success; and 2 which means unknows at this week.
At timepoint 0 (baseline) I set it to missing because it is always 0 at the beginning and then converted to 1 with the treatment.


How can Stata differentiate between failure and lost/unknown? I get a preliminary success with

Code:
stset weeks, id(ID) failure(VeinPat==0,2)
But that doesn't properly value the lost/unknows.

Thanks a lot!
Michael