R help - Jan 2011 - Function comparable to cutpt.coxph from "Survival Analysis using S"

Dear Mrs Rachel Pearce,
I am looking for a function "cutpt-coxph" in R - like you did some
years ago.
How have you solved the problem? Have you found it or a similar function?

thank you, Sincerely, Friederike


"The title says it all really; I am looking for a function along the lines
of
cutpt.coxph as described in "Survival Analysis Using S" (Tableman and
Kim), Chapter 6. As may be guessed, the function optimises the
cutpoint of a continuous variable for cox proportional hazard
modelling. I can't find it, or any similar function, on CRAN.

Alternatively, perhaps there is a way of extracting the likelihoods
from the output of coxph."


Dr. med. Friederike Schneider 
Assistenz?rztin 
Klinikum der LMU 
Campus Grosshadern 
Marchioninistr. 15 
81377 M?nchen 
Tel.: 089-7099-425
Friederike.Schneider at med.uni-muenchen.de

It is very uncommon for the assumptions underlying this method to be
satisfied.  These assumptions include (1) the relationship between X and log
relative hazard is discontinuous at X=c and only X=c; (2) c is correctly
found as the cutpoint; (3) X vs log hazard is flat to the left of c; (4) X
vs log hazard is flat to the right of c; (5) the 'optimal' cutpoint does
not
depend on the values of other predictors.

These relationships rarely occur in nature unless X=time.  Failure to have
these assumptions satisfied will result in (1) great error in estimating c
(because c doesn't exist); (2) low predictive accuracy; (3) serious lack of
fit; (4) residual confounding; and (5) overestimation of effects of
remaining variables.

This non-existence of cutpoints is why in medical research no two
investigators seem to find the same cutpoint for the same predictor in
different datasets.

Frank


-----
Frank Harrell
Department of Biostatistics, Vanderbilt University
-- 
View this message in context:
http://r.789695.n4.nabble.com/Function-comparable-to-cutpt-coxph-from-Survival-Analysis-using-S-tp3229420p3229704.html
Sent from the R help mailing list archive at Nabble.com.

Quoting Frank H: "These relationships rarely occur in nature..."  I
agree; I have seen cutpoint relationships only a handful of times in 25
years of medical work.
  A better approach is to look at the data using a smoothing spline:
     options(na.action=na.exclude)
     fit <- coxph(Surv(time,status) ~ age + weight + pspline(x))
     temp <- predict(fit, type='terms')
     plot(x, temp[,3])
My primary goal has always been to learn rather than to test.
It is common to see upper or lower thresholds, but not cutpoints.

Terry T.

Hi everyone,
Does anyone know if there is a package to run Latent Transitional 
Analysis using R?

Regards!

-- 
Sebasti?n Daza
sebastian.daza at gmail.com