If you view clinical trials as primarily being about science, then sharing data is a no brainer.
But to the extent that we should also view clinical trials as part of a political and public relations process, then it’s not clear that transparency is always best. Sunshine laws in politics often reduce compromise by disincentivizing effective but unflattering-looking backroom haggling.
Is there such a thing as too much transparency here? How worried are you about the publication of negative results leading to public backlash against companies or particular medial methodologies?
I personally think that unfortunately yes, there is such a thing as "too much transparency". An ideal situation would have:
1. access restricted to some extent to researchers.
2. safe harbours for companies in terms of being able to measure things that don't get released etc. The concept of safe harbours is generally important.
If you view clinical trials as primarily being about science, then sharing data is a no brainer.
But to the extent that we should also view clinical trials as part of a political and public relations process, then it’s not clear that transparency is always best. Sunshine laws in politics often reduce compromise by disincentivizing effective but unflattering-looking backroom haggling.
Is there such a thing as too much transparency here? How worried are you about the publication of negative results leading to public backlash against companies or particular medial methodologies?
This is a very good question.
I personally think that unfortunately yes, there is such a thing as "too much transparency". An ideal situation would have:
1. access restricted to some extent to researchers.
2. safe harbours for companies in terms of being able to measure things that don't get released etc. The concept of safe harbours is generally important.
Ah yes, that’s fairly simple in theory actually. Open to researchers / relevant institutions. Not to everyone
From what I can tell, neither Vivli nor CSDR offer any compensation for sharing the data. This looks like an incentive problem!
How can one have any confidence in research if the data is not made available to enable independent validation?
One cannot