That’s not what I meant. I’m asking for yours or anybody’s guess on why, in the pool of people that received DCVax-l, had only 4.7 percent unidentified mgmt, whilst 63% of the small group of noncrossovers had 63% unidentified mgmt?
Doesn’t really make sense though. Flippers question is why would there be so many in that tiny group that was on placebo and did not cross over? Logic dictates that there would be an even distribution, at the very least between the original placebos and treatment arms. Your analysis suggests that most were in placebo and that for whatever reason they did not crossover.
Market Data 
Markets 
AI
