How do we get to understand if a drug has: long term safety for most Rare side effects for some Efficacy for all, or Efficacy for some based on what
without controlled clinical trials and long term use?
Once we have sufficiently varied and large enough natural history cohorts in approved databases for many different diseases great. We don’t yet!
Until then hard to see how we could do without randomised controlled trials and not least in rare diseases.
Just letting biotech and pharma companies loose e.g. selling their wares under the auspices of say P4 trials or card blanche provisional approvals would be a free for all snake oil market.
And no we are far from understanding human biology and biochemistry well enough and with complex and powerful enough compute power to simply simulate our way to new perfect drugs in silico.