R help - Sep 2011 - How to understand the plotting of the cox.zph function

I have a coxph model which gives me HR of about 2.9 for presence of factor B
(factors can be A, B, C, A as baseline in the model), with 95% CI 1.8-4.8 ,
p<0.001.

When checking the proportionality assumption there is significant evidence
that there is a violation

On the link is the results of the plot(cox.zph) of the model for factor A.

http://img31.imageshack.us/img31/7213/coxzph.jpg

My question is how should I understand the smoothing line of the graph, and
what is its relation (and the relation of the values on the y-axis) to the
beta estimate the coxph function gives me (2.9 for the above example)

IF there was no violation and the line of the cox.zph plot was straight,
would the y-value of the line be (in this example) log(2.9)=1.06? If there
is no violation of the proportionality assumption, does the
"intercept" of
the line equal the log of the HR that the coxph outputs? Or is the intercept
the delta of the beta?

Thank you very much 
JT

--
View this message in context:
http://r.789695.n4.nabble.com/How-to-understand-the-plotting-of-the-cox-zph-function-tp3788886p3788886.html
Sent from the R help mailing list archive at Nabble.com.

Under the assumption of PH, the graph would be a horizontal line at ylog(2.9) =
coefficient of the model.  That is, a constant hazard ratio.

Your plot shows that there is no effect early (HR of 1) but there is an
impact later.

Terry T

Thank you very much for your answer.

I would like to construct for presentation purposes the HR(t), not the
beta(t). How can I perform this? 

I obtained the x-y values of the cox.zph plot 

time= as.numeric(as.character(rownames(cox.zph.object$y)))
HR=exp(cox.zph.object$y[,1])

However when I plot the plot(time, HR) and fit a lowess line (to time, HR),
the HR seems to jump over proportionally high (higher that what the cox.zph
plot suggests)

Am I making any error in my thinking?

Many thanks
JT

--
View this message in context:
http://r.789695.n4.nabble.com/How-to-understand-the-plotting-of-the-cox-zph-function-tp3788886p3793651.html
Sent from the R help mailing list archive at Nabble.com.

Any insight maybe (sorry for the bump, need this info), JT.

--
View this message in context:
http://r.789695.n4.nabble.com/How-to-understand-the-plotting-of-the-cox-zph-function-tp3788886p3814881.html
Sent from the R help mailing list archive at Nabble.com.