R help - Apr 2008 - How to do survival analysis with time-related IVs?

Hello folks,

 

I am wondering how to do survival analysis with time-related IVs in R. For
example,

 
> > If we have time-related variables, such as the Overall Condition of
1990, 1991 etc., how can we include these variables in coxph model? 
> > 
> > If we can not use coxph model, do we need to rearrange the dataset to
make it something like:
> >      ID time age drug censor     Year   Overall Condition 
> > 1     1    5  46    1      0     1990              XXX
> > 2     1    5  35    1      0     1991              XXX1
> > 3     1    6  35    1      1     1992              XXX2
> > 4     2    5  46    1      0     1990              XXX
> > 5     2    6  35    1      1     1991              XXX1
> > 6     3    6  35    1      1     1990              XXX2
 

And then run logistic regression using "censor" as the DV? 

 

 

 

Tianxu Chen 

101 N. 33rd Street
Academic Building, 3FL
Drexel University
Philadelphia, PA 19104
Phone: 1-215-895-2884(O)
Email: tc332@drexel.edu  

 


	[[alternative HTML version deleted]]

"Tianxu" <tc332 at drexel.edu> wrote in
news:20080421012323.F06A311662F at smtp.mail.drexel.edu: 

> I am wondering how to do survival analysis with time-related IVs in
> R. For example,
See section 4 of Fox's contribution:

<http://cran.r-project.org/doc/contrib/Fox-Companion/appendix-cox-regression.pdf>

<http://cran.r-project.org/doc/contrib/Fox-Companion/cox-regression.txt>


<snip example>
 
> 
> And then run logistic regression using "censor" as the DV? 
> 
I was initially unclear about why you were running logistic regression on 
"censor", but perhaps you are trying to assess the non-informative 
censoring assumption? The R function that immediately comes to mind 
is glm( ...., family=binomial). Hmisc package also has lrm(...) and 
associated summary and diagnostic functions.

-- 
David Winsemius

Thanks. The first two links are helpful.

For the log-reg, I was initially thinking that we need to re-arrange the
dataset to make ID-Year observations (I think some books are suggesting
this), and then run regression on "hazard rate". I did not know
whether we
should write code to calculate the hazard rate by ourselves. However, the
paper solved my puzzle. Thanks!



Tianxu wrote:
> 
> Hello folks,
> 
>  
> 
> I am wondering how to do survival analysis with time-related IVs in R. For
> example,
> 
>  
> 
>> > If we have time-related variables, such as the Overall Condition
of
> 1990, 1991 etc., how can we include these variables in coxph model? 
> 
>> > 
> 
>> > If we can not use coxph model, do we need to rearrange the dataset
to
> make it something like:
> 
>> >      ID time age drug censor     Year   Overall Condition 
> 
>> > 1     1    5  46    1      0     1990              XXX
> 
>> > 2     1    5  35    1      0     1991              XXX1
> 
>> > 3     1    6  35    1      1     1992              XXX2
> 
>> > 4     2    5  46    1      0     1990              XXX
> 
>> > 5     2    6  35    1      1     1991              XXX1
> 
>> > 6     3    6  35    1      1     1990              XXX2
> 
>  
> 
> And then run logistic regression using "censor" as the DV? 
> 
>  
> 
>  
> 
>  
> 
> Tianxu Chen 
> 
> 101 N. 33rd Street
> Academic Building, 3FL
> Drexel University
> Philadelphia, PA 19104
> Phone: 1-215-895-2884(O)
> Email: tc332 at drexel.edu  
> 
>  
> 
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 
> 
-- 
View this message in context:
http://www.nabble.com/How-to-do-survival-analysis-with-time-related-IVs--tp16801272p16812343.html
Sent from the R help mailing list archive at Nabble.com.