jq's reply to you was correct, but the point of the conversation has drifted from what whent wrong in the trial.
It was not that the control arm outperformed, it was that bavi did not perform better when added to what should have been an identical arm.
If bavi was supposed to have a 25% longer mos, and then if the control arm came in at 12 months, why did bavi not come in at 15 months (and the trial would be ongoing)?
There are 3 possible explanations:
1) Bad luck. Even with an effective drug and well planned trial it can happen.
2) A trial failure that led to a biased effect. Blame it on changing conditions, poor design or sabatoge. Does not matter.
3) Bavi failed.
Any or all of those played out to some extent.
Technical note before jq corrects me. Both arms showing the same improvement over trial design can reduce the power of the trial, and by itself be the cause of a failure. That was not the case here with the halt this early.
Market Data 
Markets 
AI
