I am weighting survey data with pweight and have successfully estimated sensitivity and specificity for the outcome of interest. I have used the ci command using weighted Ns to estimate 95% CIs around the sensitivity and specificity estimates. The my biostat collaborator would like to see 95% CIs estimated with robust SEs for these estimates. I am not finding a clear solution to this. Any suggestions for estimating robust SEs for weighted senspec estimates? Thanks so much.