# R CCH CI Univariable Analysis

Using the survival package in R I am trying to calculate 95% confidence interval (CI) of the proportional hazards model coefficient. Normally this is pretty simple to do with the survival package.

> confint(cch(Surv(time, event) ~ a + b + c, ...))

2.5 %    97.5 %
a   -0.4216482 0.1225872
b    0.2624281 1.3339170
c    0.3344311 1.2224415


However when I try to do this in a univariable setting, the confint function fails to return anything, except a matrix with no rows.

> confint(cch(Surv(time, event) ~ a, ...))

2.5 %    97.5 %


Is this a bug in R or am I missing something else? If this is a bug how can I calculate these CIs 'by hand'?

-
Does this happen for all cases of one variable? Or is it just this case? And if just this case, have you looked at a? –  Peter Flom Aug 15 '12 at 0:52
@PeterFlom This happens with any variable, but only in the univariable setting. In the multivariable setting everything works as expected. –  Nixuz Aug 15 '12 at 4:24