Thank you Inquirg for educating me and let me see if I got
it right.
We had three placebo groups in this trial (with 3, what can go wrong?).
1. A very large number of external controls.
2. A group of 64 placebos who after progression were treated with
DCVax-L and for some mysterious reason displayed an unusually
long mOS.
3. A group of 35 placebos of whom 6-7 dropped out of the trial whereas
the remaining 28-29 never received DCVax-L because most of them
progressed rapidly, needed salvage treatment and had a shorter than
average mOS.
Now if I am doing this correctly, I should entirely discard the OS results of the external
placebos and instead consider the combined mOS of the 64 crossovers and 28-29
permanent placebos, as the true blue placebo group whose combined mOS is (whoopie),
somewhat higher than the mOS of the treatment group.
For all of the above calisthenics, I now give myself an E for effort.
Market Data 
Markets 
AI
