I'd like to know whether anyone has had experience using the statistical
method known as generalized estimating equations in the analysis of
repeat viral diseases. A friend of mine is thinking of using the method
to analyze repeat seroconversions among children exposed to Norwalk
virus from a contaminated water source. As I understand it, the advantage
of GEE is that one can take into account repeated exposures to the virus
and account for previous immunity. The analytical procedure is relatively
new and I'm looking for examples of how it's been applied, its perceived
strengths and weaknesses, and ease of use.
I suspect this may be a topic more suitable for a statistical newsgroup
but I thought I'd try here first in case any virologists also get into
the analysis of disease outbreaks.
--Jeff Johnston (ilva at email.unc.edu), UNC-Chapel Hill, Epidemiology