A couple of notes:
1. Hopefully, natural selection over the next few decades will improve the driving population in those states
2. From a cycling perspective the particular statistic doesn't take into account the amount of traffic. It would be interesting to estimate the number of vehicular encounters per hour and then factor in drive behavior. the good news is that while we're not completely masters of our own destiny, we can greatly affect the most important statistic with our own behavior, our own survival probability.