Precision Medicine is all about data mining to get the strongest correlations from and for response(responders)/non-response (non-responders) to a specific treatment for a specific ailment.
Much, much more data to be mined from an experiment where a specific patient is treated for a specific ailment with a placebo and then SWITCHED over to an active drug. (2 controls - same patient and same disease state, 1 variable - without drug, with drug. Great science - this will enhance, enrich, boggle the mind with stat sig data correlations - extremely AI friendly environment).
Now, you do the rest..
OK, Mr. Smartguy riddle me this. In AI world if not fail-extension is set equal to pass (redefine) then what values do you put on limits. PM will emphasize early fail. You have then to better define what is next once we get results that show evidence of possible efficacy. Since all this CNS disease have never been effectively treated this is a gray area. We only can define fail. We are now buying time using an old clock.
AVXL staff surely know what is going on. Patients will eventually make this obvious. This cannot go on for long, IMO. Interim results are everywhere for sure. I always thought that some RETT mom would let us know.
So, we assume that NOT FAIL is set equal to extension. You must Write the inference logic for the rest of it. Somehow the investment calendar and the science calendar have to get in synch. I hear Dr.M. saying, not yet or have I not read it correctly?. The only question we have to ask is, if they had poor results would they let us know? How? When? Some new clock must be developed which synchs up the PM results clock w/all the rest of us.