Dear Statalist

I would like to calculate adjusted absolute risk differences after using the xtgee command in stata

I am using xtgee to model the association between characteristics of children admitted to hospital and the odds of being transferred to intensive care (binary outcome).

Each row in the data is one "admission to hospital" - and one child can have multiple admissions. Each child has one id number.

Transfer to Intensive care is coded as a binary outcome (0 or 1) associated with each admission

The exposures are sex, age, previous chronic conditions, as binary variables. These are all constant within one child, across n admissions

The code I am using is

xtgee intensive_care i.sex i.age. i.chronic_condition, family(binomial) link(logit) corr(exchangeable) robust i(id_number) eform

This gives odds ratios of transfer to intensive care for admissions to hospital where the child has a chronic condition vs not having a chronic condition, adjusted for age, sex and clustering of repeated admissions within each child

Some of the odds ratios are large - but the absolute risk of admission to intensive care is small in both groups, so I would like to describe absolute risk difference, using this model. Is there a way of doing this?

I have looked at adjrr - but am not sure I will be able to add additional packages as I am using the data on a remote secure server. I am using Stata 16.

Any advise would be really appreciated!

Best Wishes

Joe