Follow- up question: While using extended Cox model, where exposure is time varying and has more than two arms, how to handle that to get two different HR. For instance, see below small data and Proc Phreg code. For data tumor01 - exposure is "0" or "1"; so I will get one HR. For data tumor012 - exposure is "0", "1" or "2"; so I should get two HR but I am getting only one HR. I tried to include that in class statement but it gives an error. SAS CODES: data Tumor01; /*P1 - P15 variable has value of "0" and "1", say drug_A and drug_B*/ infile datalines missover; input ID Time Dead Dose P1-P15; label ID='Subject ID'; datalines; 1 47 1 1.0 0 1 . 1 1 1 1 1 . . . . . . . 2 71 1 1.0 0 . 0 0 0 0 0 0 1 1 1 1 1 1 1 3 81 0 1.0 0 . 1 1 1 1 1 1 1 1 1 1 1 1 1 4 81 0 1.0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 5 81 0 1.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 65 1 1.0 0 0 0 1 . . 1 1 1 1 1 1 1 . . 7 71 0 4.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 69 0 4.0 . . 0 0 0 0 0 0 0 0 0 0 0 0 0 9 67 1 4.0 . . 1 1 1 1 1 1 1 1 1 1 1 1 . 10 81 1 4.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ; run; data Tumor012; /*P1 - P15 variable has value of "0", "1" and "2", say drug_A, drug_B and drug_c*/ infile datalines missover; input ID Time Dead Dose P1-P15; label ID='Subject ID'; datalines; 1 47 1 1.0 0 1 . 1 1 1 1 1 . . . . . . . 2 71 1 1.0 0 . 0 0 0 0 0 0 1 1 1 1 1 1 1 3 81 0 1.0 0 . 1 1 1 1 1 1 1 1 1 1 1 1 1 4 81 0 1.0 0 0 0 0 0 2 2 2 2 2 2 2 2 2 2 5 81 0 1.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 65 1 1.0 0 0 0 1 . . 1 1 1 1 1 1 1 . . 7 71 0 4.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 69 0 4.0 . . 0 0 0 0 0 0 0 0 0 0 0 0 0 9 67 1 4.0 . . 1 1 1 1 1 1 1 1 1 1 1 1 . 10 81 1 4.0 0 0 0 0 0 2 2 2 2 2 2 2 2 2 2 ; run; proc phreg data=Tumor01; *TRY data = tumor012; model Time*Dead(0)= NPap / rl; array pp{*} P1-P15; do i = 1 to 15; if pp ne . then do; npap = pp ; end; end; run; /* QUESTION: Why I am not getting two hazard ratios for data tumor012 Shouldn't I get HR for 1 vs 0 and for 2 vs 0 */
... View more
