[dahart]

Your question is conflating multiple kinds of risk. Drugs and procedures that are known to work still do have risks of complications or side effects, but statistically speaking, they are known to work and have known levels of risk.

Fake and/or investigational drugs carry an additional and independent risk of unknown efficacy and unknown potential harms. That additional risk layers on top of the first kind of risk of known complications. Drugs with no proven efficacy can carry additional financial risks too.

This is all why we have drug regulations - to hopefully understand the risks, minimize chances of severe harm and/or death, give consumers informed choice and allow them to assess the risk and weigh the risks against the known benefits. Unregulated drugs carry unknown benefits, and unknown & potentially very high levels of risk that patients have no hope of knowing or evaluating rationally, and that is what the comment you replied to is referring to.

Risk isn’t binary like your first comment implied. We don’t just have risk or no risk, we have risks that are known and risks that are unknown, and we have risks with different levels of harm vs benefit, and we have risks with varying levels of likelihood. In medicine it’s important for the risks to be known, for the harm levels to be low, and for the likelihood to be low. That doesn’t always happen, but if harm is likely or high then it’s important for the benefits to be high (to outweight the risks) and its important for anyone accepting the risk to have informed consent.