I got a problem with how to get P for trend in cox regression even after searching by google.
I have a DDD_1 variables which is a factor variable, containing 0,1,2.
res.cox = coxph(Surv(t$death_365days_duration, t$death_365days==1)~
DDD_1+variable 1+variable 2, data=t)
res.cox showed:
exp(coef) exp(-coef) lower .95 upper .95
DDD_11 0.6342 1.5768 0.5885 0.6835
DDD_12 0.5888 1.6985 0.5479 0.6327
coef exp(coef) se(coef) z Pr(>|z|)
DDD_11 -4.554e-01 6.342e-01 3.818e-02 -11.929 < 2e-16 ***
DDD_12 -5.297e-01 5.888e-01 3.668e-02 -14.442 < 2e-16 ***
But I don't know how to get P for trend of DDD_1 variable.
Is that:
res.cox = coxph(Surv(t$death_365days_duration, t$death_365days==1)~
as.numeric(DDD_1)+variable 1+variable 2, data=t)
res.cox showed
coef exp(coef) se(coef) z Pr(>|z|)
as.numeric(DDD_1) -2.935e-01 7.457e-01 1.784e-02 -16.451 < 2e-16 ***
Is this P the P for trend?
