[michaelmrose]

So if we kill another thousand people administering it differently how likely are we to discover it now works instead of substantially increasing mortality?

[matthewtoast]

I don't know. We should follow whatever procedures we follow when testing any potential new treatment.