[TechnicalVault]

If you've ever had to organise a scientific study you'll understand why. Privacy and ethics are important but they make recruitment of a large and representative sample for any medical study a nightmare. First you'll need to run your protocol past an ethics board and do back and forth until it's agreed. Once you have that done you'll send out an invite (often blind) in hopes of getting people who fit your study aims and you will get who you get. There are often costs associated with this.

So now you have some willing participants you have to screen your initial group to filter it down to people who actually meet your recruiting criteria and consent them for your study. Finally you actually get to gather your data, if you need people to come in for sampling and have to cover their expenses.

Next you'll look at your data and realise there's some confounding effect which reduces your powers to infer anything (e.g. somehow you overrecruited a particular group and they turn out to do something which correlates with the thing you're studying). You'll cry a little and realise you need to recruit more people to have any statistical power to draw a conclusion.

tldr; medical and scientific studies are hard if you want them to actually have any validity.

[AdamN]

It seems like we really need to move forward to 'indefinite/active studies'. I believe some researchers are trying to get onto this pattern but of course there are major privacy/quality of care concerns.

By 'indefinite/active studies' I mean that the studies never stop - data just keeps going back into the system as a flywheel of the drug distribution process. I take a second-generation drug for CML (leukemia) and none of my health data goes back into research unless my doctor decides to elevate my data into a paper or something (which I don't think has happened yet).

[markus92]

These kind of platform trials are getting more common luckily! Famous example is the COVID Recovery trial.

[pas]

wouldn't this side-effect become visible in Phase4 data? (and there should be at least some hints of this in the previous ones)

so, what I'm trying to say is that for this particular drug we "just need to wait", no?