I’m only intimately familiar with the hospital reporting numbers and unfortunately I no longer know of a publicly available source on the side-effect incidence rate of in-hospital usage (for many depressingly obvious reasons, thank you Donald Trump). However benzos are extremely well documented as having diverse reactions, and the label is extremely clear that even low therapeutic doses see severely adverse psychological reactions at rates of around 1% in the population (delerium, hallucination, self-harm, etc.) (page 17). This is predictably exacerbated by being in a clinical setting and, coupled with a tendency for the side effects of benzos to have superlinear increases in severity as dosages go up, is a very worrying trend on it’s own.

It’s a disparity from my (admittedly reductive) 30% claim yes, but between the datasets I am basing that claim on getting purged and the shocking lack of independent public studies on the topic (wtf??) I’m at a bit of a loss to provide concrete sources for it being that high (how convenient for me…). From the data available though you can see it’s closer to 5% that has a “severe” psychological reaction outside a clinical setting, if you’d prefer to assume that number is correct and that I was exaggerating for effect I sure wouldn’t blame you. IMO that’s still unbelievably high for a medication that has a risk of becoming even more popular in recreational usage.