The new remdesivir data from Tony Fauci from N >1,000 placebo-controlled trial (unpublished, Oval Office citation) does not show statistically significant reduction in mortality (8% vs 11%, in the right direction) but a clear-cut benefit in time to recovery 11 vs 15 days, P<0.001
It's been a very confusing story. First a positive signal from compassionate use @NEJM, then @UCHicagoMed leak encouraging. Then @Lancet today clearly negative trial, underpowered. Then this.
The good news: it looks like we now have a drug w/ some efficacy. And that's very good.
The good news: it looks like we now have a drug w/ some efficacy. And that's very good.
I'm feeling better that the signal I called weeks ago from the 1st report panned out to be true, despite many who challenged the call
To add to the mix (and mix-up) there is another trial reported today (by press release) comparing 5-days vs 10-days of remdesivir. It was underpowered, used a composite outcome. Here is the Table, you can't say 5 days is as good (or better, or worse). Adds to the safety assurance 
The P value for mortality reduction (11.0 -> 8.6%) from the big trial is 0.059.
P<0.05 threshold is not absolute
That's very helpful and adds support to the conclusion that with remdesivir is an efficacious drug beyond recovery time (but not a miracle, or extremely potent one).
P<0.05 threshold is not absolute
That's very helpful and adds support to the conclusion that with remdesivir is an efficacious drug beyond recovery time (but not a miracle, or extremely potent one).
That's 11.6 reduced to 8.0%, adding to my list of infamous typos :-(
