Hello everyone!
I have a (potentially) straightforward issue. My objective is to propensity match subjects from two different locations (this is the treatment variable, 'CN', 1= from CN, 0 = not from CN) and get certain outcomes. I have managed to do pretty well using the -psmatch2- command and recovered data on several outcomes. It was easy enough to compare means before and after matching.
Now I have to do a survival analysis. Say, I need 30-day outcomes in terms of Stroke, my time-to-event variable (in days) is 'timetostroke' and failure variable is 'strokepostop', and run:
followed by
, I believe I get outcomes on the entire sample.
How would you suggest I run this analysis on specfically the treated and the matched controls?
Thank you for your advice as always!
I have a (potentially) straightforward issue. My objective is to propensity match subjects from two different locations (this is the treatment variable, 'CN', 1= from CN, 0 = not from CN) and get certain outcomes. I have managed to do pretty well using the -psmatch2- command and recovered data on several outcomes. It was easy enough to compare means before and after matching.
Now I have to do a survival analysis. Say, I need 30-day outcomes in terms of Stroke, my time-to-event variable (in days) is 'timetostroke' and failure variable is 'strokepostop', and run:
:
stset timetostroke if timetostroke<=30
:
sts graph, by(CN)
How would you suggest I run this analysis on specfically the treated and the matched controls?
Thank you for your advice as always!