But US is also a non uniform country, so IMO best statistics should be done on clusters, even small countries have different regions with different stats, a go

But US is also a non uniform country, so IMO best statistics should be done on clusters, even small countries have different regions with different stats, a good comparison would also be on the larger cities and see how the percentages compare, if they differ a lot analyze what could cause the difference and if you figureout the reason apply it to reduce deaths.