These were injured drivers seen in hospital after a crash, so there may be a different population. These are from about a decade ago. We compared toxicology testing done in hospital—alcohol levels—with their subsequent driver records. We found that of drivers with a BAC between .08% and .16%, 4.7 % were ultimately convicted of impaired driving. For drivers with a higher BAC, over .16%—more than twice the legal limit—it was 13.6%.
I would add that a similar study was done in Alberta—similar population and similar methods—and their conviction rate was slightly higher, but not much. I think it was in the range of 15%.
Of these particular populations of drivers, very few of them are convicted. We can speculate on the reasons for that, but the numbers are low.