News for hackers. Stuff that matters.

A comment on Infection fatality rate of Covid-19 inferred from seroprevalence data [pdf]. Parent comment

Re:Infection fatality rate of Covid-19 inferred from seropre... by petertodd on Sunday October 18, 2020 @12:24AM (#24815536)
...but why would excess lockdown deaths happen in a uniformly distributed way? Lockdown itself wasn't applied in a uniform fashion: hospitals figured out pretty quickly that lack of care was killing people and modified policies to mitigate that damage. My own advice to friends and family was "probably best not to go to the hospital for that" in March, and back to normal in April when I realized that the numbers weren't as bad as portrayed. Anecdotally, all of the deaths I mentioned happened close to the beginning of lockdown in their respective countries.

Again, this isn't idle speculation: genuine mainstream health authorities believe lockdown has killed significant numbers of people. This is not a controversial position.

[ Parent | on HN ]
Re:Infection fatality rate of Covid-19 inferred from seropre... by notahacker on Sunday October 18, 2020 @08:04AM (#24817220)
> ...but why would excess lockdown deaths happen in a uniformly distributed way?

No, the question is why does the excess deaths distribution perfectly align with the expected and recorded COVID death spike and drop so sharply afterwards when lockdown was still in place.

It's uncontroversial that lockdown has killed and saved significant numbers of people for reasons other than COVID, but similarly it is entirely uncontroversial that the aggregate increase in excess deaths was caused by COVID. The idea that the inflection point COVID-time-to-death days after the start of lockdown is better explained by unannounced changes in policy or your personal advice to friends and family, on the other hand is about as scientifically credible as blaming 5G.

[ Parent | on HN ]