R help - Jul 2012 - Stats question: Comparison of the same individuals during two exposure times

Hi,

I'm hoping that someone will be able to help.  I would like to compare how
covariates associate with the risk of a binary outcome during two periods. 
Period 1 will be non-exposure to a treatment and period 2 will be exposure
to a treatment.  The same individuals will be examined in each group but I
want to be able to compare the association of certain covariates between the
two groups to see if there is a treatment interaction.  I've looked at
case-crossover designs and time series analysis and don't think that they
are suitable.  The cohort has longitudinal data so individuals will go onto
treatment at different times and the effect of the treatment needs to be
administered for a while before it has an effect.  The reason why I cannot
just go ahead with an exposed vs unexposed design is that most individuals
in the cohort end up on the treatment eventually and the unexposed group is
very small and lacks power for a meaningful comparison.  

Is there anyway to compare the same individuals during different exposure
times and to look at the effect of different covariates under the exposed
and unexposed conditions?

Thanks for you help,
Natalie

-----
Natalie Van Zuydam

PhD Student
University of Dundee
nvanzuydam at dundee.ac.uk
--
View this message in context:
http://r.789695.n4.nabble.com/Stats-question-Comparison-of-the-same-individuals-during-two-exposure-times-tp4636732.html
Sent from the R help mailing list archive at Nabble.com.

I would STRONGLY recommend that you talk to your local statistician.

Further:

1. R-help is not a statistical consulting forum
2. Remote statistical consulting is very risky due to the inherent
difficulty in communicating all essential context of the problem and data.

You have been warned!

-- Bert

On Tue, Jul 17, 2012 at 2:10 AM, natalie.vanzuydam
<nvanzuydam@gmail.com>wrote:
> Hi,
>
> I'm hoping that someone will be able to help.  I would like to compare
how
> covariates associate with the risk of a binary outcome during two periods.
> Period 1 will be non-exposure to a treatment and period 2 will be exposure
> to a treatment.  The same individuals will be examined in each group but I
> want to be able to compare the association of certain covariates between
> the
> two groups to see if there is a treatment interaction.  I've looked at
> case-crossover designs and time series analysis and don't think that
they
> are suitable.  The cohort has longitudinal data so individuals will go onto
> treatment at different times and the effect of the treatment needs to be
> administered for a while before it has an effect.  The reason why I cannot
> just go ahead with an exposed vs unexposed design is that most individuals
> in the cohort end up on the treatment eventually and the unexposed group is
> very small and lacks power for a meaningful comparison.
>
> Is there anyway to compare the same individuals during different exposure
> times and to look at the effect of different covariates under the exposed
> and unexposed conditions?
>
> Thanks for you help,
> Natalie
>
> -----
> Natalie Van Zuydam
>
> PhD Student
> University of Dundee
> nvanzuydam@dundee.ac.uk
> --
> View this message in context:
>
http://r.789695.n4.nabble.com/Stats-question-Comparison-of-the-same-individuals-during-two-exposure-times-tp4636732.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> R-help@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.
>


-- 

Bert Gunter
Genentech Nonclinical Biostatistics

Internal Contact Info:
Phone: 467-7374
Website:
http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm

	[[alternative HTML version deleted]]