Hi,
i'm looking for a method to calculate the population attributable fraction for independent variables in a conditional logistic regression model built using Stata's "clogit".
The stats are explained in Bruzzi et al, 1985 (http://www.ncbi.nlm.nih.gov/pubmed/4050778).
If someone could point me to a Stata implementation (eg., ado-file) I'd be grateful. The user written command "aflogit" does what I want, but only for "un"conditional logistic regression analyses.
Thanks,
Mirko
Have you looked at Roger Newson's punafcc command (findit punafcc)? It is a very flexible command.
Hi Mirco,
I am facing the same issue as yours. I hope by now you have solved it.
Any advice on how to calculate PAF in conditional logistic regression models using STATA?
Hi Danah,
i went with the suggestion of Ariel Linden and the command punafcc (see above). The command is contained in a STATA package you need to install using eg., the findit command.
When it's ready to use, you first perform your conditional logreg. Then you run the punafcc command using additional parameters for your matching variable, the variable you would like to calculate the PAF for etc.
punafcc gets most of the data and parameters from your last calculated model, that's why you need to run your logreg before punafcc.
Hope that helps.
Best,
Mirko
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression