For years, we have had mixed data on whether steroids were beneficial or not in ARDS. Now, there is evidence — not yet in paper form — from a well respected group in an RCT w/ a rigorous protocol that dexamethasone reduces all-cause mortality in severe COVID-19 (!) ...
Obviously we're all drooling for the paper itself. But I'm so surprised by the outpouring of skepticism. For months, completely non-evidenced based therapies have been used for COVID-19 ARDS in hospitals. For instance:
(1) Hydroxychloroquine, with basically zero evidence;
(2) "Intermediate dose anticoagulation", with zero randomized evidence (e.g. microthombi, high d-dimer);
(3) Some have liberalized tidal volume goals on weak evidence that COVID-19 ARDS has better static compliance ....
... and remdesivir only recently attained reasonable RCT evidence.

So here's thing thing ...
I'm a zealot on evidence-based medicine, but I sense an odd inconsistency. Yes, the dexamethasone results were published in press release form — although recruitment literally only stopped *this* month. But whatever uncertainty that should give us, from a Bayesian perspective,
the evidence for dexamethasone in severe COVID-19 is , at this moment, superior to that of convalescent plasma, more-than-prophylactic-dose-anticoagulation, and far greater than hydroxychloroquine was even for a hot second.
If people give dexamethasone tomorrow, and it turns out that the press release inaccurately represented the findings of the study, which (in this *specific case*) I find highly improbable, they will still have acted appropriately given the existing body of evidence.
Particularly in light of the fact that this is not the first RCT to show benefit from steroids in ARDS, although admittedly the evidence is mixed. Given how heterogeneous ARDS is as a diagnosis, it’s not surprising that the efficacy of steroids might vary depending on etiology.
You can follow @awgaffney.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: