I am unable to use svychisq() on categorical variables?

I am using svychisq() to compute the significance in the difference between proportions in my survey data but I keep getting an error that the approximation may be incorrect. My input and output are both categorical data.

obj <- svydesign(ids = ~1,
         strata = NULL,
         weights = survey_weights,
         data = survey_data)

svychisq(~input, ~output, obj) 

Error message I am getting is "Chi-squared approximation may be incorrect". I am not sure why I am getting this error and am wondering if anybody knows why I might be getting this. Should I be doing a different test?