Yep, the key point is that information lies in "events" (deaths), not patient counts. At 20%/10%, not significant two-sided. There has been a lot of discussion about what to assume here - the fact that we are running the "trial of the sickest" is important.