I’m understanding here that cannabis is linked to a higher RISK of crash fatalities not linked to crash fatalities themselves? That’s a very important distinction.
No it’s really not, the same thing is true of alcohol use and car crashes because sober people still crash or lung cancer from smoking etc.
Suppose you have a study of 10k people split 5k:5k between smokers and non smokers. In the non smokers group you get say 10 cases of lung cancer and the smokers group you get 200. It’s really tempting to say the individual smokers got long cancer from smoking, but the base rate isn’t 0 so you don’t know who was screwed either way. Perhaps if nobody smoked in that total population there would have been 19 cases of lung cancer or perhaps 22.
Thus only thing you can say is smoking increases RISK of lung cancer not link it to specific cases.
