Weighted AUCs using survey data


Has anyone calculated weighted AUCs for survey data in STATA or SAS? STATA does not seem to incorporate the weights in calculating the associated 95% CI.

What makes AUC applicable to your situation?

Hi Dr. Harrell,

Thank you. We are using survey data to look at the diagnostic performance of certain biomarkers to identify undiagnosed diabetes.

Thanks. I don’t know of software that does weighted AUC but I don’t recommend AUC. Take a look here.

This is helpful, thank you.