Thread by @drjohnm

Paper from Prof. Ioannidis et al. <preprint>
They study the risk of #COVID19 death in younger vs older age groups from hotspot areas across the world.
I like this paper b/c it’s data rather than anecdotes.
medrxiv.org/content/10.110…
Thread

The authors take current data -- imperfect as it is-- and use it to compare death rates across geographically separate cities, states and countries. They divide COVID deaths by population to get absolute death rates. (low for all ages.) New York data interesting.

They then use the data to compare relative risks of death in old and young and when possible use co-existing conditions.
A lot of it is with personal communication. Not ideal but it's a fluid situation.

This paper confirms what is known: much higher risk for older patients with co-existing disease. That bears on mitigation strategies. The authors discuss this.

The value in this paper is that instead of presenting anecdotes --which scare the crap out of you-- it presents data. In medicine and in policy, decision making is better with the latter rather than the former.

RE decisions: Consider stents in pts w a severe blockage. If you show a picture of a blockage to a person (doc, nurse, patient) they get scared and want a stent They remember an anecdote.
But the data (oodles of it) say stents don't reduce the chance of dying compared to meds

I am sensing a bit more data informing the #COVID19 crisis. And that is good. B/c uncertainty is scary.
Finally, Prof Ioannidis recently gained lots of public criticism. Good on him for responding with evidence.