Read on Twitter
We've got excess death data again!
Regardless of what method of error-correction your prefer, US deaths are rising again and could very quickly return to first-wave peaks.
Regardless of what method of error-correction your prefer, US deaths are rising again and could very quickly return to first-wave peaks.
For everyone who's going to reply "WELL TECHNICALLY THIS IS THE FIRST WAVE"
e hump
two humps
humps are waves
second wave
e hump
two humps
humps are waves
second wave
We can look at this by age, so here it is.
Social distancing does still appear to be saving child lives. Deaths of minors are waaaaaay down. The decline in minor deaths is about 2x what can be explained even if you assume traffic and related deaths fell to zero.
Social distancing does still appear to be saving child lives. Deaths of minors are waaaaaay down. The decline in minor deaths is about 2x what can be explained even if you assume traffic and related deaths fell to zero.
Even if traffic deaths fell a ton and neonatal deaths fell due to fewer births, the decline in child deaths is simply too big for those kinds of explanations.
You can only account for it by assuming that deaths due to negligence, inattention, and/or infectious diseases fell.
You can only account for it by assuming that deaths due to negligence, inattention, and/or infectious diseases fell.
In all cases those would be changes we could clearly attribute to parents and kids staying home more.
Meanwhile, deaths for adults are way up.
Meanwhile, deaths for adults are way up.
lol I forgot to attach the by-age chart. Here it is.
I know it's stylish to argue that parents being overwhelmed is a big problem but actually the vast majority of parents are quite competent and able to prevent their children from literally dying inside the house. https://twitter.com/MamaMashaBear/status/1289273106385776646
It's very vogue and all to be like "Parenting is a complicated task and requires very specialized skills and if parents are stressed it's a big problem!" but actually parenting is not extremely complicated and when your outcome measure is "not dying" even less so.
Also y'all I'm just gonna put it out there that what we're seeing is we could appreciably increase life expectancies by shifting social norms towards parents and kids spending more time together at home.
Anyways. On with the thread.
Anyways. On with the thread.
Vaccines are good and surveys show that child vaccination rates may actually be rising during COVID due to increased salience of infectious disease but whatever anti-vaxxers gonna anti-vax. https://twitter.com/MamaMashaBear/status/1289273772252508161
As every week I need to briefly discuss how the sausage is made. https://twitter.com/DWAnderson1/status/1289274148649394176
I have 3 ways to get at true excess deaths:
1. Take the raw data, take the typical observed revision patterns from past data releases, and *add* the typical revisions to the released data
2. Same, but instead take the observed revision pattern *ratio* and *multiply*
1. Take the raw data, take the typical observed revision patterns from past data releases, and *add* the typical revisions to the released data
2. Same, but instead take the observed revision pattern *ratio* and *multiply*
3. Take the historic ratio between official COVID deaths and excess deaths and apply it to recently-reported COVID deaths and add that to a seasonal normal trend.
Which of these is best depends on the state the time and a whole lot of randomness. When they disagree appreciably we just have to admit a fair amount of uncertainty. But in this case they all show deaths rising and above normal!
One interesting and surprising thing is that despite massively increased testing capacity, under-measurement of COVID deaths appears to be about as big or even larger than it was during the first wave.
This is pretty interesting because it suggests that our testing regime is still pretty woefully incomplete, indeed that our testing regime may not be doing any better than it was several months ago.
By the way, my estimate is that about 8-15% of the country has gotten COVID-19 at this point. We could totally have a third, fourth, fifth, sixth wave and still not have hit herd immunity depending on how the actual underlying infection dynamics work.
In terms of cumulative excess deaths, we're way up and going wayer upper. Also everybody being like, "well but deaths will fall a lot afterwards" well, still waiting.
The good news is that R values seem to have peaked and are now falling. Dunno if it's policy shifts, more distancing, masks, or Mystery Phenomena, but for whatever reason the leading indcator for R values is saying we're back below R of 1 now.
I just want to make clear, this is absolutely bogus fearmongering with no basis in facts at all. https://twitter.com/zoyd/status/1289279366506008577
The mortality data I use is part of a long-standing system CDC has been using for many years to track mortality. While it's possible that the data handling has changed in some way, the reality is that it's actually state-reported data through a system they share with CDC.
If CDC fudged their figures, states would know almost immediately and complain. None have. Which suggests that the system is still operating as normal. The narrative that the US data is somehow unreliable is absurd and silly.
Getting lots of people pulling a "Monty Python," "It's only a model"
I mean, I'm using raw deaths of all causes as they come in each week. That's not a model.
The adjustments are indeed a model, but an extremely straightforward and simple one with a long track record.
I mean, I'm using raw deaths of all causes as they come in each week. That's not a model.
The adjustments are indeed a model, but an extremely straightforward and simple one with a long track record.
So I mean if you really want to ignore it fine but posting the unrevised CDC weekly death totals is idiotic; that data is known to be incomplete, with a relatively predictable degree of incompletion.
Anyways, you may wonder how the US compares to peer countries!
Here's all deaths in the US and 14 European from the start of the pandemic to now, versus deaths in the same period in 2019. The US and western Europe have had approximately identical mortality spikes in total.
Here's all deaths in the US and 14 European from the start of the pandemic to now, versus deaths in the same period in 2019. The US and western Europe have had approximately identical mortality spikes in total.
The US second wave is a lot bigger than the European second wave thus far.
But the European first wave was a lot bigger than the US first wave.
So far, it looks like the US and western Europe have had about the same cumulative mortality spike.
But the European first wave was a lot bigger than the US first wave.
So far, it looks like the US and western Europe have had about the same cumulative mortality spike.
So for all the folks being like "AMERICA IS THE WORST UGH WHY ARE AMERICANS SO STUPID LITERALLY NOWHERE IS WORSE THAN US"
your beloved Europeans have had it *exactly* as bad.
your beloved Europeans have had it *exactly* as bad.
Life expectancies in the US and western Europe are quite similar. https://twitter.com/moultano/status/1289285552995364864
Let's do states!
Here's Texas. Deaths high and rising. However, if official case counts and estimated R values are to be believed, then Texas' deaths should begin to come down next week.
Here's Texas. Deaths high and rising. However, if official case counts and estimated R values are to be believed, then Texas' deaths should begin to come down next week.
Oh I'm a total doofus.
That's Florida, not Texas.
HERE'S Texas. It's bad. But, again, estimated R values have fallen below 1, so deaths should peak soon.
That's Florida, not Texas.
HERE'S Texas. It's bad. But, again, estimated R values have fallen below 1, so deaths should peak soon.
Here's California. Deaths definitely elevated but not seeing the huge spike like FL and TX.
Of course, New York is hard to beat. But deaths in New York have mostly returned to normal now, though there remains potential for a recurrence.
Here's Georgia. Georgia looks a lot like California to be honest.
Arizona is pretty bad though to be honest Texas may end up worse.
Illinois graaaadually approaching normalcy.
If lockdowns kill, why aren't deaths spiking in Maine, which had a lockdown?
Maybe because there aren't many COVID infections in Maine and the excess deaths really are COVID deaths.
Maybe because there aren't many COVID infections in Maine and the excess deaths really are COVID deaths.
More generally, here's official COVID-19 deaths per million in each state vs. estimated excess deaths per million in each state. WHat you can see is excess deaths are EXTREMELY well correlated with official COVID deaths: it's not lockdowns killing people! It's COVID!
It really is COVID y'all! COVID deaths are being undercounted! Excess deaths which aren't officially COVID deaths do NOT spike in response to lockdowns, they spike in response to COVID DEATHS, BECAUSE THEY ARE COVID DEATHS BEING MISSED.
The main reason they are missed is *probably* either 1) they are deaths at home and not getting tested or 2) they have no obvious COVID symptoms and die of a non-obvious symptom like heart failure, which was actually unbeknownst to them brought on my an "asymptomatic" COVID case.
Anyways. here's peak and cumulative excess deaths by state.
NY and NJ remain the uncontested leaders in terms of tons of people dying.
NY and NJ remain the uncontested leaders in terms of tons of people dying.
No, the point is lockdowns don't actually cause increases in those deaths. That's literally what the chart showed. https://twitter.com/RichardHair20/status/1289293960238624768
I use data from all 14 countries for which I could get it: Norway, Denmark, Finland, Sweden, Germany, Austria, Switzerland, Belgium, Netherlands, the UK, France, Italy, Spain, Portugal. I don't think that's cherry-picking??? https://twitter.com/R_H_Ebright/status/1289293688997154816
This is a possibility. Lockdowns could definitely have altered the composition of non-COVID deaths. But evidently not enough to actually increase total non-COVID deaths. https://twitter.com/s_r_tarnmoor/status/1289294659890499584
Okay. I'm sitting in the yard at my sister-in-law's house not playing with my kid and her cousins because I'm talking to you wretches on this hellsite. So, I'm out! Hope you all stay healthy and well, even those of you engaged in lowkey manslaughter (i.e. not masking up).
