Hi, I am looking for some guidance on the command and command structure to look at the relationship of infections at two anatomic sites. Each site can have up to 15 infection genotype. I want to do a logistic regression on how one site predicts infection at another site, controlling for other factors. That is, for each person, I will need 15 dummy variables. This would be combine into a single logistic regression model, the unit of analysis changes from person-level to infection-level (15 entries for each type with a common id and covariates). What is the GEE command and command structure to account for person-level clustering and perhaps include infection genotypes as dummy variables so each infection genotype could have a different intercept? This would yield one set of association for for all infections in general. Thanks for any guidance.
-
Login or Register
- Log in with
