Understanding Society User Support

I am using Youth data to see how their drinking behavious, using Stata. I understand we have to use svy command to account for the complex survey structure. However, standard errors are not calcuated because 'Note: missing standard errors because of stratum with single
sampling unit.' I tried to create new strata using adjasent strata, but it failed. Results from svydes showed there are multiple single strata cases and I suspect this was the reason. I also think this is due to youth data being the subpopulation data from the household data.
I am wondering if we can ignore the structure of the survey design, but account for household clustering.
Or should we use 'single unit' option in the svyset command?
I appreciate your suggestion.