Risk per km is the appropriate measure for comparing the safety of making a journey by different modes. So for example you typically are deciding whether to walk or cycle to the shops. You are generally not thinking of walking to the local shops or cycling to shops the other side of town that are the same travel time away. On that basis, even just taking accidents involving a vehicle, the fatality risk per km is about 30% higher for a pedestrian than a cyclist. Taking all accidents in the roadway the risk for pedestrians is about four times higher per km for serious injuries.
So the figures for serious injuries per bn km for 2009 are:
Pedestrian: 2,001
Cycling: 517
or just under 4x higher for pedestrians. If you factor in the higher head injury rate for pedestrians, the risk of a head injury is 4.8 times higher for a pedestrian than a cyclist per km
so , basically, if i travel the same distance by foot, then its 4* more dangerous than cycling.. even with my wimpy amount of cycling, that still makes it more likely i will have an accident on the bike than walking... whats more useful is risk per journey surely ?
After all, otherwise I'd travel to the shops via the motorway or via a plane, which is patently ludicrous!
I could accept the comparison of journey types and rates, except the data we quote isn't setup like that ( well the link I had wasn't). it would have to only count journeys where it was okay to do either type..
Some spooky info on the web about rate analysis though, you made the point about 'whats an accident if its just one bike is it an RTA?' but even then, theres concern the police dont report all accidents correctly and try to compare with hospital admission data. I mean, you'd think this info was kinda accurate, but it seems far from it!
http://www2.dft.gov.uk/pgr/roadsafety/research/rsrr/theme5/roadaccidentscasualtiescomp.pdf
we sometimes worry about more and more cyclists dying on the road... but the motorbike rates are depressing reading in comparison...