[R] Multilevel survival model

[Kingsford Jones]

On Thu, Jul 22, 2010 at 8:23 AM, Christopher David Desjardins
<desja004 at umn.edu> wrote:
> * Please cc me if you reply as I am a digest subscriber *
>
> Hi,
> I am wondering how I can run a multilevel survival model in R? Below is
> some of my data.
>
>> head(bi0.test)
>  childid famid lifedxm     sex      age delta
> 1   22.02    22 CONTROL   MALES 21.36893     0
> 2   13.02    13   MAJOR   MALES 21.18001     0
> 3   64.02    64 CONTROL   MALES 20.09377     0
> 4    5.02     5 CONTROL FEMALES 19.94524     0
> 5  183.02   183 BIPOLAR FEMALES 19.87406     0
> 6   37.02    37 BIPOLAR   MALES 19.83847     0
>
> Where age is my time variable and delta is development of thought
> problems (0 = No; 1 = Yes). My covariate of principle interest is
> mother's diagnosis (control, major depression, and bipolar) and I am
> principle interested in examining how the survival (not developing
> thought problems) differs by this covariate. However, I have siblings in
> my data set (famid) and I would like to account for this shared
> variability. How might I do this in R? Can I do it with the survival
> library? If possible, I'd also like to control for sex.

as usual, R has an embarrassment of riches

see <http://glmm.wikidot.com/faq> under the heading 'Mixed Modeling
Packages' - particularly under the subheading 'extensions'

If you want to stick with the survival package the frailty function
may be of interest.

The package bayesSurv would be worth looking at as well.

As an aside: I assume many (most?) of your groups are singletons -- I
wonder what impact that will have on estimators if you are including
famid as a random effect.

Kingsford



>
> Thanks,
> Chris
>
> --
> Christopher David Desjardins
> PhD student, Quantitative Methods in Education
> MS student, Statistics
> University of Minnesota
> 192 Education Sciences Building
> http://cddesjardins.wordpress.com
>
> ______________________________________________
> 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.
>