[NetBehaviour] Fwd: Bitcoin tech applied to clinical trial documents

[Edward Picot]

Ruth -

I think the problem with the system is how/whether it will be policed. 
As I understand it, the idea is to use BitCoin as a means of putting a 
kind of unique ID onto the protocol at the beginning of a process of 
research - the 'protocol' being a statement on the part of the 
researchers about what they intend to research and how they intend to 
research it. If they subsequently change their minds and rewrite the 
protocol, because the results aren't quite what they were hoping for, 
then the unique ID will be invalidated, which means that everybody will 
be able to tell there has been an alteration.

Let's say you were a pharmaceutical company and decided to research into 
the efficacy of Viagra as a cure for male-pattern baldness. As your 
research progressed, you discovered to your dismay that it wasn't 
effective against male-pattern baldness at all - but then you realised 
that your statistics were showing much more favourable results in the 
area of erectile dysfunction. So you rewrite your protocol to make out 
that erectile dysfunction is what what you intended to research all 
along, and lo and behold, instead of your results looking like a failure 
they suddenly look like a resounding success.

The problem is, that when you go to publish your results in one of the 
medical journals, they don't check back to see what your original 
intentions were - they only look at the end product. There's a 'peer 
reviewing' process, which means that each piece of research is meant to 
be looked at by people who know all about that particular field, but it 
obviously isn't completely watertight, as shown by the example given in 
the article, Study 329. Researchers were supposed to be investigating 
whether an antidepressant (paroxetine) was effective for adolescents. 
When their initial findings were unfavourable, they kept moving the 
goalposts until eventually they found criteria which did produce 
favourable results, and then published their results, purporting to show 
that it was a good treatment. Subsequent studies have shown an 
association between paroxetine and suicidal behaviour and thinking in 
children and adolescents. GlaxoSmithKline were fined $3 billion in 2012 
for misrepresenting the product, and it's no longer recommended for the 
under-18s.

You could avoid this kind of protocol-tweaking via the BitCoin process. 
But you could also avoid it if you had a central repository where 
protocols had to be lodged at the outset of research, and against which 
research was always checked by the medical journals before they 
published any results. The real question is not how to authenticate the 
protocols, but how to get everyone to agree to a transparent system and 
how to make sure it's properly enforced.

- Edward