Someone made a good point over at Stockhouse.com.  I am cutting & pasting his/her comments below:

That does not sound right. Since the trials are double blind, how would anyone be able to determine efficacy? 

It seems to me that the earlier DSMB statements were made in error...an error that has now been corrected by someone in the latest DSMB, as we come to the close of the trial.
Read more at http://www.stockhouse.com/companies/bullboard?symbol=t.rvx&postid=29137410#5P3yG0hCAAbxFdoj.99