Per se limits are difficult if you're going to say a given limit is equal to a given amount of impairment. With alcohol, we could do that, based on the studies from the beginning of time—I think it was in 1969—when we brought forward the legislation in Canada.
In the case of marijuana, it's not clear. Typically, one would look at a limit that is low, and in the case of these testers it's now down to five, which is the lowest detectable limit in oral fluid. If, as in some states, it's zero tolerance, what is zero? Mathematically, you can't measure zero. It has to be something positive. What do you take as the first positive measurement? Perhaps two nanograms per millilitre is indicative of marijuana present in the body at a level that should have some impairing affects. Dr. Huestis et al. show that this is the case.