I was referring to publication in a peer reviewed journal. In order for them to break out critical they must have done analysis on the full trial including severe. If the full trial or just severe hit stat sig than it would have been published. From what was published there was probably a low percentage of deaths in severe (maybe 25% or less). The lower that percentage the more likely that there was less difference in the two arms. I am still surprised at the low number of critical patients. I expected at least a third to be critical.