Stratified Cox Regression

[#1]

Hello,

I am trying to get hazard ratios for assessing effect modification by strata of another variable with 3 categoroes. Would this procedure be correct?

PROC PHREG DATA=NEW;

class rank_score (ref="1") education (ref='0') income (ref='3') ;

MODEL time*exposure(0)=rank_score age educationincome/ rl;

strata lifestyle;

run;

Using this I dont really get startified hazard ratios by lifestyle so I am thinking how to modify it please?

Thank you very much.