[jschwartzi]

Is there sufficient evidence in favor of rejecting the null hypothesis, that 57% of patients in the HCQ study group would have gone on to spontaneously recover without HCQ? Because that's the really interesting question. It's not a question of the study group versus control group, it's a question of whether the study group and control group had enough characteristics in common to make them statistically equivalent samples of a population that vary only in whether they were given HCQ. And I think the consensus among people who are clear on that is that the populations have too little in common.

[Judson]

A fair criticism of the study, but does not make the case that “no patient on HCQ alone was virus free by day 6”, since that is objectively false by my reading.