It's certainly something you could propose. I think it works for alcohol, and it works to the extent that it's enforced in a public fashion. Among other things it works because a large portion of the driving population may have been drinking, or might drink in the future if they have not had the experience of realizing, “Gosh, the police are out there. I'm going to get stopped, and I'm going to get screened.” If that situation applies to drug use, then I think it's perhaps something we might consider down the road. I don't think there are any evaluations of that kind of approach in the literature as of yet, so we don't know what kind of an outcome there would be, in contrast to what we understand about the effects of mandatory alcohol screening.
On September 18th, 2017. See this statement in context.