'Sensitivity and Specificity using GEE for correlated data in R?

I am trying to calculate the sensitivity and specificity of a test for correlated data using the approach described in this article: https://iovs.arvojournals.org/article.aspx?articleid=2770806

The article contains SAS code but I would like to translate this to R, see Appendix 2 here.

My model takes the general format:

geeglm(disease ~ test, id = id, data = data, family = binomial('logit'), corstr = 'independence' )

where disease and test are coded as 0/1

How do I then get the estimates for sensitivity and specificity?

Thanks

Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Solution	Source