That is a shocking stat, although I see that the source article only looked at the 20 most populous counties in the US: https://pmc.ncbi.nlm.nih.gov/articles/PMC8325358/

I wonder if that causes some selection bias (e.g. density correlating with poverty).