I think the study is likely detecting "hybrids are disproportionately used in big cities and blue states", where pedestrian/bicycle traffic is disproportionately likely to be close to cars and thus gets hit more often (versus, say, rural Kansas where cars generally hit stationary objects or each other).

You would probably also find that hybrids strike proportionately more vegetarians than non-hybrid cars. This does not mean meat scares Priuses away.