Of course the FDA will want all sorts of data incl subsets
My point is that the bogus cherry-picked subsets in the spring 2018 PR's from IPIX re B OM were done by IPIX in order to suggest better efficacy than initial data
And that such statistical shell games are not worthy of reporting. There was lots of discussion of this at the time the data came out(as you well recall)
reminiscent of crummy data presentation from P2a, lapped up by the mindless and heedless