competing risk

[rykwong]

Hi,

I have a survival analysis to do that involves competing risk.  Dataset contains approximately 2500 pts, a subset of which died from either cardiac, cancer, or unknown (death=1,2,3, respectively, 0 if not dead at end of study followup).  I need to built a model (forward selection strategy with entry and stay criterion both set at p=0.05) using proc phreg accounting for the competing risk from non-cardiac deaths.  Then I need to add a new variable omega (serum omega-fish oil level) to see if omega level has association with death incremental to the variables in the model (e.g. var1, var2, var3).  The death of interests is cardiac death (death=1) and I had previously censored all other deaths, which obviously did not account for the competing risk.  

I found many macros in the web but so far have trouble running them.  Unfortunately my skills in using macro is limited.   Would anyone be willing to share a solution.  Would you be so kind to give me detail instructions if this is a macro you are sharing with me?

much appreciated

[Reeza]

What version of SAS do you have? If you have SAS 9.4+ PROC PHREG can now do competing risk analysis. Or are you looking for customized graphs or tests?


And if you want to outsource it feel free to contact me 😉