[R] cch() and coxph() for case-cohort

Terry Therneau therneau at mayo.edu
Mon Jun 16 21:21:21 CEST 2008


--------- begin included message ---------

 I tried to compare if cch() and coxph() can generate same result for
same case cohort data

Use the standard data in cch(): nwtco

Since in cch contains the cohort size=4028, while ccoh.data size =1154
after selection, but coxph does not contain info of cohort size=4028.

The rough estimate between coxph() and cch() is same, but the lower
and upper CI and P-value are a little different.  Can we exactly use
coxph() to repeat cch() using with appropriate configuration in
coxph()?  Is SAS a better way(PHREG,CASECOH.SAS) to implement
time-dependent case-cohort?

<output deleted>

----------end included message --------------------


I showed you how to do time-dependent case cohort analysis.  Stop complaining 
and use the advice.
 
 No, SAS will not be better.  It will be different due to defaults (different 
convergence criteria, different default ties method).  Not better, not worse, 
just different.  
 
  The code I sent results in one variance estimate type, the default cch code is 
another.  They are asymptotically equivalent, and in your sample are essentially 
identical: one CI was (1.497, 2.91) the other (1.491, 2.925).  Neither is 
'better' or 'correct', both are good.  If you feel that you must have the other 
estimate, then read the paper upon which cch is based (which I wrote); you can 
find it in the references for the cch help page.
  
  The cch output has some default information not in the coxph output and vice 
versa.  So what?  
  
  	Terry Therneau



More information about the R-help mailing list