And it was done on decisions between 2008-2015. That is the way...

And it was done on decisions between 2008-2015. That is the way research articles of this nature are done and have to be done. You need to wait a significant amount of time after the end date to make sure that all of the information that is going to come out has come out. It isn't only companies with disclosure obligations that develop drugs and apply to the FDA. Other companies do to and you can't assume that a drug that didn't receive approval, received a CRL (complete response letter). The FDA also does not always provide a response within a stated time. Once you have collated the data and analyzed, say a year. Written the paper, 3 months to 15 months, submitted the article, had it peer viewed and published, up to 18 months, given the relatively scarce number of journals for this type of research, those are about the most timely figures you are going to get. Anyway, the bias is toward under-estimating discordance, then over-estimating it, because panel votes and drug approvals are made public by the FDA. CRLs are not. Moreover, there does not appear to have been a change in agency policy or policies that would be expected to dramatically increase or decrease discordance over the last five years or more. There is no mystical database that the FDA allows people access to so that they can get precise information. Does more accurate data exist? Most likely. But it will be proprietary, not public. You will need to be employed and in an appropriate position at one of the investment firms that keep track of the data to get a chance to see it. Could the true rate of discordance be 18%, certainly, but it could just as likely be 24%. Just because a paper is 12 months old, doesn't mean it isn't accurate. Go through a presentation or two from a respectable biopharma company and discard every statement that is made based on a paper that is more than a year old. I bet you will lose support for 95% percent of the referenced based statements they make or more. BTW, if you were really interested in finding out the truth, rather than just ramping an investment. You would go to PubMed and find the actual paper as I did. You would also find that it is free for download and that you could do some actual real research.