setting the record straight, page-16

Andrew - are you sure you have this quite right.

You said:
"Regarding failure to publish: It is not acceptable to use unpublished data which has not been peer reviewed as evidence of a treatment's efficacy."

This is simply not true. The FDA does rely on unpublished data that has not been peer reviewed as evidence for efficacy.

It would take too long to explain why and how this situation is as it is. And there is no doubt it causes problems. Have a read of this BMJ article to get a flavor of the problem.

http://www.bmj.com/content/344/bmj.d7202?view=long&pmid=22214754

You confusing what is required to get a drug approved with the scientific process. Two completely different things.

Your statement:
"Furthermore, excusing an unexplained positive endpoint produced by data mining as an acceptable means by which to say a clinical trial has been successful is blatantly wrong."

Seems to reflect a misunderstanding of different phases in clinical trials. The analyst could have explained it better but the gist is essentially correct. Just as an aside look at where the FDA is headed following ObarmaCare - trying to better match patient characteristics to interventions. Obarma has put in about $1bn to develop methods to do exactly what you are saying shouldn't be done. Do a search on "Comparative effectiveness research"

The slight irony in all this is that it is the humble debt collector with:
"...you can publish and peer review all you want but its the FDA which decides if you go to market and its the big Pharmas that open their wallet and put 1.7 bil on the table AND PAY FOR THE TRIALS".

that is spot on. All imo of-course.

Regards Southoz.