Re: [R] Trend test for hazard ratios

2014-04-16 Thread David Winsemius
On Apr 16, 2014, at 4:07 AM, Göran Broström wrote: > On 04/15/2014 10:51 PM, David Winsemius wrote: >> >> On Apr 15, 2014, at 6:32 AM, Therneau, Terry M., Ph.D. wrote: >> >>> You can do statistical tests within a single model, for whether >>> portions of it fit or do not fit. But one cannot ta

Re: [R] Trend test for hazard ratios

2014-04-16 Thread Göran Broström
On 04/15/2014 10:51 PM, David Winsemius wrote: On Apr 15, 2014, at 6:32 AM, Therneau, Terry M., Ph.D. wrote: You can do statistical tests within a single model, for whether portions of it fit or do not fit. But one cannot take three separate fits and compare them. The program needs context t

Re: [R] Trend test for hazard ratios

2014-04-15 Thread David Winsemius
On Apr 15, 2014, at 6:32 AM, Therneau, Terry M., Ph.D. wrote: > You can do statistical tests within a single model, for whether portions of > it fit or do not fit. But one cannot take three separate fits and compare > them. The program needs context to know how the three relate to one another

Re: [R] Trend test for hazard ratios

2014-04-15 Thread mkleber74
Hi Terry, thank you for your suggestion. I tried to do it the way that you described. I fitted my Cox model with strata and I get coefficients and P values for my ZlnGalectin in the different strata (see summary below). So am not so familiar with R so please excuse my perhaps stupid questions but

Re: [R] Trend test for hazard ratios

2014-04-15 Thread Göran Broström
On 04/15/2014 05:00 AM, r-help-requ...@r-project.org wrote: Hello, I have the following problem. I stratified my patient cohort into three ordered groups and performed multivariate adjusted Cox regression analysis on each group separately. Now I would like to calculate a p for trend across the h

Re: [R] Trend test for hazard ratios

2014-04-15 Thread Therneau, Terry M., Ph.D.
You can do statistical tests within a single model, for whether portions of it fit or do not fit. But one cannot take three separate fits and compare them. The program needs context to know how the three relate to one another. Say that "group" is your strata variable, trt the variable of inte

[R] Trend test for hazard ratios

2014-04-14 Thread mkleber74
Hello, I have the following problem. I stratified my patient cohort into three ordered groups and performed multivariate adjusted Cox regression analysis on each group separately. Now I would like to calculate a p for trend across the hazard ratios that I got for the three groups. How can I do tha