Prediction of clinical trial outcomes?


#1

Clinical trials are based around an initial investment into a trial which will test whether a drug is safe and effective. They are VERY expensive, in fact the health R&D sector is being disrupted because the old model of trials are no longer feasible. Having the ability to accurately predict which drugs will be successful would a be incredibly valuable to world health productivity.
Also as trials are in progress, the certainty of outcome will tend to become stronger as unsuccessful drugs are dropped so the remaining ones have a higher chance of success.
If we could give scientists and physicians the power to 'vote' on which they think will succeed in a capital context, that would actually be highly beneficial to health.

Your system may be able to be adopted similarly, but forgive me for being un-knowledgeable of the technicals. So specifically, can you explain how your system verifies the outcome without a human component?

I may be include this in a paper I am writing for my University, about new methods for clinical trial design.


#2

Hello @seeriktus - this sounds like an extremely interesting usecase for prediction markets.
So I guess the expensive study would be needed anyways but you would have a real time mechanism to know in what direction to go or what study to end early because the projected chances for success are to low. And this number would be the aggregated knowledge of all experts who are confident to participate in a forecast.

For determining the outcome we would need in this case a human component - we call it an oracle. Some entity that brings the result (e.g. the FDA approval) in a machine readable way on the blockchain. But that should not be a big issue. If you interested we can do a call to discuss more.