> So, maybe hcq has no effect on outcomes on mild cases. This does not mean it couldn't have an effect on severe/hospitalized patients. Sure, but that's the ex

> So, maybe hcq has no effect on outcomes on mild cases. This does not mean it couldn't have an effect on severe/hospitalized patients. Sure, but that's the exact opposite of the earlier narrative when the trials in Brazil were called off. (paraphrasing) "Sure, it doesn't help people who are critically ill, but you should use it as a prophylactic to prevent milder cases from becoming more severe". So... it doesn't work in patients with a mild, moderate or severe case; (per the earlier advice) it must be prescribed after infection but before symptoms develop; and it must be prescribed in doses that are known to cause heart problems in a significant portion of the population (so significant, in fact, that patients prescribed much lower doses of HCQ are normally required to be regularly monitored for the onset of said problems). At what point do you admit that any positive effect of HCQ is just cherry-picking/p-value-hacking?