I'd imagine if instead of per 100000 inhabitants, we used per million kilometers driven, we'd see more comparable numbers across countries. US has far more vehicles per capita than Denmark where I'm living. I'd also imagine people spend longer time in traffic in US.
So I don't think it's that driving is more dangerous there, rather it's that you're more likely to be in a car at any given time if you're in USA. If you measured bicycle accidents per 100000 inhabitants probably the stats would be switched for DK/US.