Hacker News new | ask | show | jobs
by hammock 1281 days ago
>it coincides closely with the spikes in COVID cases we've seen.

Aren't there about 100 other confounders that could be in the mix there

2 comments

If it was one general noisy correlation, I would agree. But there are multiple spikes that go up and down and match almost precisely with the multiple waves of covid, persisting through many shifts in other factors (like downstream effects of different levels of lockdown) with almost no change in the trends for all other deaths beside it. It's about as close of a correlation as you can ever hope for with real-world data.
Which of them do you propose would cause the American public (and others worldwide; we see excess deaths increased everywhere, whether they required masking or closed schools or did very little to mitigate) to suddenly start dying in significantly higher numbers - again, of any cause - for the last two years - and especially so at the same time large COVID waves are occurring?
> Which of them do you propose would cause the American public (and others worldwide; we see excess deaths increased everywhere, whether they required masking or closed schools or did very little to mitigate) to suddenly start dying in significantly higher numbers - again, of any cause - for the last two years - and especially so at the same time large COVID waves are occurring?

I can give one example, which happened to my dad...

"Sorry, we have to reschedule that surgery to remove the tumor because we're anticipating a covid spike and need the bed availability."

Repeat until he died due to cancer spreading.

That sort of thing is relatively trivial to tease out of the stats, because in addition to excess deaths we have per-disease statistics as well.

See the chart about 2/3 down on https://www.healthsystemtracker.org/brief/covid-19-continues..., titled "Average daily deaths in the United States, by cause (2020 - Present)". We don't see a massive spike in other causes of death (including cancer, which is a flat line at ~1,600/day throughout), and no one's likely to misidentify a cancer or heart attack death on a death certificate.

There are certainly cases like your dad's, where delayed screenings and treatment had individual impacts, but none of the stats available to us show this explaining a meaningful amount of the 1.3M excess deaths observed.

Does it matter which?
It does if you can’t think of any that would demonstrably cause 1.3M excess deaths in the US, yes.