Any stata code for calculating adjusted number needed to be exposed (NNE)?

[Zihan Dong]

Hello,

I was wondering is there any Stata code available for calculating adjusted number needed to be exposed (NNE)?

thanks!

[Rich Goldstein]

not familiar with it - citation please (with description of what it is if you want)

[Zihan Dong]

not familiar with it - citation please (with description of what it is if you want)

hi Rich,

The NNE can be calculated as follows, it means the number of *exposed individuals* in the population for one additional case due to exposure:

where π0 is the risk of disease in the unexposed; RR is the relative risk.

I found -bcii- could calculate NEE, but not adjusted NEE. As I want adjusted NEE, so I guess maybe there is any option or coding following lostic regression to achieve this?

[Rich Goldstein]

my guess is that you want poisson regression (with robust SE's), which will give you the RR directly (but inefficiently though your formula won't care about that) rather than logistic regression which will give you OR's

added in edit: forgot to include that if you use -poisson-, be sure to use the irr option

[Paul Dickman]

How about using predictnl after fitting your model?