I never said that I can explain the discrepancy between 2:1 randomization and the ultimate 232:99 ratio. I too have not heard a sound hypothesis that doesn’t mean that I bought into the most optimistic of the bunch. In any case, we will maybe find out soon enough.
Stealth crossover would also help explain improving outcomes seen between 2017 and 2018 readouts and cause FDA to work extra hard to find ways to support analysis that works around crossover confoundment. Best wishes.
Market Data 
Markets 
AI
