Applying machine learning to identify unrecognized Covid-19 deaths in the US
Comments
umvi
simonw
The paper touches on that:
> The ARR shows the extent by which total predicted COVID-19 deaths exceeded officially reported COVID-19 deaths during the period. A limited number of counties had ARRs < 1, which suggests that there were more officially reported COVID-19 deaths than total predicted COVID-19 deaths. One reason that a county could have an ARR < 1 is if death certifiers recorded people as dying from COVID-19 when they had COVID-19 but actually died from another unrelated cause.
bulbar
Later on, many who made that claim went on reversing that statement, arguing that a lot of died "of covid vaccine" instead of "with covid vaccine".
kennywinker
Afaik that was a story that spread around, but has very little connection with reality. As i recall this was mostly down to people who don’t know how to read a death certificate misunderstanding what goes in the cause of death field. I.e. something along the lines of it would list the cause of death as “organ failure”, because that was what caused death - but covid caused the organ failure.
Open to a good source that says otherwise, of course
whatsupdog
Exactly.
peyton
> Our machine learning approach relies on different assumptions—(i) that there was no measurement error for COVID-19 deaths in hospitals and (ii) that a model trained on in-hospital deaths is transportable to out-of-hospital settings.
Whether these assumptions are valid is highly debatable given various incentive structures applied in in-hospital settings over the period studied.
WalterGR
> given various incentive structures applied in in-hospital settings over the period studied.
My vague memory is that that was disinformation commonly spread by COVID deniers that has been thoroughly debunked.
Could you point me to a good source about it being true?
I thought we needed a ML model to deflate Covid deaths. In the US at least, a lot of died "with Covid" were counted as died "of Covid"