I keep thinking there should be some way to combine the evidence from the mild/mod and sev/crit trials.
If they had exactly the same measurement at the same date, a statistical technique called meta-analysis would be appropriate. Not sure if you could just treat them as one pool, but certainly the statistical power would be greater than either alone.
Alas, I could not find an identical metric at 14 days in both trials.
Dr. Doughty, any ideas on some sort of pooled analysis?