That's what I mean about abusing statistics you oaf. Those figures are meaningless unless it's per user, because you can't otherwise get a comparative measure of 'risk' (for want of a better term).
If one person took a substance, and was the only user of that substance, and died as a result, that would make it 100% deadly. According to you, because only one person died, it's less deadly than alcohol, or the others.
Since I suspect that there's more than 30 times the number of people who drink than use heroin, or 10x the number who use cocaine, your figures support MY point, NOT yours.