Using mi estimate with menbreg in Stata 13.1-possible?

[Heather D'Angelo]

I am using Stata 13.1 and the menbreg command to run multilevel negative binomial regression models. However, I now have an imputed variable that I need to include as an independent variable. I am able to run the menbreg model with the mi estimate prefix, even though the manual says that it is not available with menbreg. Does anyone have any experience with this? Should I choose another type of model that the manual lists as being available with the mi estimate command? Thank you!
Heather

[Carlo Lazzaro]

Heather:
I am not clear whether your seeming problem refers to -mi impute - or - mi estimate -.
That said, as a general rule, you can increase the likelihhod of obtainng helpful replies if you post exactly what you typed and what Stata gave you back (as per FAQ, that you are requested to read before joining this Forum).
Moreover, please provide fully referenced sources of statements such as

...the manual says that it is not available with menbreg.

or

...the manual lists as being available with the mi estimate command?

(which section(s)? which page(s)? of Stata 13.1 .pdf manual).

Kind regards,
Carlo

[Richard Williams]

I don't know what the manual says, but the first two lines of the program are

*! version 1.0.6  19aug2013
program menbreg, eclass byable(onecall) prop(irr mi)

Since mi is listed in the properties, that means that multiple imputation can be used with the command.