[R] standard error of exp(coef) from coxph

[David Winsemius]

On Jul 23, 2011, at 1:41 AM, Ehsan Karim wrote:

>
> Dear List,
> Must be a silly question, but I was wondering whether there is a  
> direct way of calculating "standard error of a HR or exp(coef)" from  
> coxph objects
> x <- coxph(Surv(time, status) ~ age + inst, lung)> x        coef  
> exp(coef) se(coef)     z    page   0.0190      1.02  0.00925  2.06  
> 0.04inst -0.0104      0.99  0.01028 -1.01 0.31
>

?predict.coxph

(The third example is precisely what you are requesting, at least  
under the assumption that you are considering a discrete predictor or  
want the HR for a one unit change in a continuous one.)


> cheers,
> Ehsan


David Winsemius, MD
West Hartford, CT