Read on Twitter
Here's the raw reported Florida P&I mortality by week back to 2012-13 from CDC.
The tweets circulating come from people comparing to monthly data downloaded from CDC WONDER, which uses a different classification system for cause of death than the mortality-surveillance data.
The tweets circulating come from people comparing to monthly data downloaded from CDC WONDER, which uses a different classification system for cause of death than the mortality-surveillance data.
Here's another way of looking at it.
Florida's "excess all cause mortality" and its excess "pneumonia and influenza mortality" are *smaller* than its reported COVID deaths, suggesting that Florida's COVID outbreak is SMALLER than official data suggests.
Florida's "excess all cause mortality" and its excess "pneumonia and influenza mortality" are *smaller* than its reported COVID deaths, suggesting that Florida's COVID outbreak is SMALLER than official data suggests.
Latest weeks will be revised upwards. https://twitter.com/complexifire/status/1266261542674161665
The point is that Florida's reported increase in pneumonia deaths or deaths of *any* cause is similar to or smaller than its reported COVID deaths, meaning the story about massive underreporting is FALSE. https://twitter.com/coreysdavis/status/1266261704003825664
It's not "misclassification" to label a COVID death as a pneumonia death.
COVID kills you *via pneumonia*. Pneumonia deaths *are* COVID deaths, and vice versa.
And pneumonia deaths < COVID deaths! https://twitter.com/coreysdavis/status/1266269493891747845
COVID kills you *via pneumonia*. Pneumonia deaths *are* COVID deaths, and vice versa.
And pneumonia deaths < COVID deaths! https://twitter.com/coreysdavis/status/1266269493891747845
To be even clearer:
Folks, deaths can be classified as pneumonia AND COVID. Here's CDC's own page which explicitly flags deaths which are both COVID AND pneumonia. https://www.cdc.gov/nchs/nvss/vsrr/COVID19/
Folks, deaths can be classified as pneumonia AND COVID. Here's CDC's own page which explicitly flags deaths which are both COVID AND pneumonia. https://www.cdc.gov/nchs/nvss/vsrr/COVID19/
Yet another way to show this:
Here's FluView reported P&I deaths vs. CDC WONDER's count for ALL acute respiratory deaths.
You can see that FluView routinely runs several TIMES the CDC WONDER data. Different classification systems!
Here's FluView reported P&I deaths vs. CDC WONDER's count for ALL acute respiratory deaths.
You can see that FluView routinely runs several TIMES the CDC WONDER data. Different classification systems!
Okay. Enough Florida hysterics. Let's do our weekly EXCESS DEATHS ROUNDUP!!!!!
Okay nevermind it's gonna be a while, being conscripted into an exotic torture ritual, aka "exercise"
Okay, NOW ready for excess deaths!
Folks, we've got some very good stuff. ANd by good stuff I mean some actually good stuff and also some very bad stuff, like a lot of deaths. Which is not good.
Folks, we've got some very good stuff. ANd by good stuff I mean some actually good stuff and also some very bad stuff, like a lot of deaths. Which is not good.
I'm going to start in Latin America again. Here's a bunch of charts of excess deaths, implied infections, and implied R values for Peru.
Peru's lockdown completely failed to do SQUAT.
Peru's lockdown completely failed to do SQUAT.
Here's the thing.
No matter your metric, whether you count backwards from excess deaths, COVID confirmed deaths, or official COVID cases, before Peru's lockdown, they *did not have enough cases to reliably compute an R value*.
And yet, to show it again:
No matter your metric, whether you count backwards from excess deaths, COVID confirmed deaths, or official COVID cases, before Peru's lockdown, they *did not have enough cases to reliably compute an R value*.
And yet, to show it again:
One theory on why R did not fall quickly, or according to some metrics even ROSE after the lockdown began, is that Peruvians didn't REALLY socially distance that much. For example, we know that markets stayed open, and news reports confirm crazy high seropositivity there.
But I have some skepticism about this argument. The cell phone based data suggests Peru did a LOT of social distancing.
But there's a rebuttal to that: maybe a lot of Peruvians don't have cell service!
Okay. Maybe. Soo....
But there's a rebuttal to that: maybe a lot of Peruvians don't have cell service!
Okay. Maybe. Soo....
Let's compare to Peru to another country with a similar level of eletrificiation, cell use, income level, etc, and which we KNOW had an EXTREMELY intense lockdown: The Philippines!
Peru and the Philippines look identical.
Peru and the Philippines look identical.
Now, you could still argue that the cell phone data coverage is an issue: maybe non-tracked Peruvians behaved very differently than non-tracked Filipinos.
Totally possible!
Totally possible!
But that explanation is pretty much evidence-free. Rather, my working theory on why the lockdown in the Philippines worked is.... wait for it...
The Philippines is an archipelago. Lockdowns and ferry cancellations achieved near-perfect isolation of many areas.
The Philippines is an archipelago. Lockdowns and ferry cancellations achieved near-perfect isolation of many areas.
Whereas this was not the case in Peru. We are still seeing new regions get hit as the geographic extent of COVID within Peru expands.
This contributes to my longstanding argument that the most important macro-tool we have is geographic isolation: you have to nearly-totally sever travel between regions. This is as or MORE important than social distancing.
See my thread here for some more elaboration and US examples of the importance of TRAVEL RESTRICTIONS. https://twitter.com/lymanstoneky/status/1265919107771232258?s=20
Okay. So. In Peru, lockdowns, and evidently social distancing generally, completely failed to prevent massive spread of COVID-19.
Today, Lima is probably the worst-hit large city on Earth. Lombardy or New York are peanuts next to the scale of death in Lima.
Today, Lima is probably the worst-hit large city on Earth. Lombardy or New York are peanuts next to the scale of death in Lima.
Relative scale, I should say; dunno about *absolute*.
Anyways, on to the next country!
Anyways, on to the next country!
Sorry, paused for a second. My research assistant needed her diaper changed.
Okay, on to Brazil!
Brazil also has a big outbreak with R still above 1 despite lots of social distancing and many states having declared lockdowns.
Brazil also has a big outbreak with R still above 1 despite lots of social distancing and many states having declared lockdowns.
In Brazil we can see that R values remained high, or even INCREASED, after various measures were put in place.
Note that my COVID death data for Brazil uses Brazil's date-and-cause file, not the JHU file, and I haven't adjusted it like I have all-cause, recent values are ????
Note that my COVID death data for Brazil uses Brazil's date-and-cause file, not the JHU file, and I haven't adjusted it like I have all-cause, recent values are ????
But look you can sort of retcon a story where Lockdowns Are Super Important but they just didn't work in Peru or Brazil because of (insert idiosyncratic reason here).
Or.... maybe this strategy just doesn't work as well as people think!
Or.... maybe this strategy just doesn't work as well as people think!
ESPECIALLY in very dense areas where people are packed together, like the urban environments of Peru and Brazil, or, say, NYC, I don't think we should assume that locking everything down is necessarily optimal.
It may be that we'd be better off with a gajillion roadblocks around the country at every county line and letting people do whatever they want within their county. That's basically what the Philippines did!
And also India!
And in both cases it may have led to some starvation deaths which is terrible! But it may have led to lower overall deaths than a lockdown strategy.
And in both cases it may have led to some starvation deaths which is terrible! But it may have led to lower overall deaths than a lockdown strategy.
Kinda. I'm an advocate of MUCH STRICTER rules on some things: travel, isolation of contacts, school closures, but MUCH LOOSER rules on other things: business closers, assemblies, sheltering-in-place, etc. https://twitter.com/mattyglesias/status/1266369928174526477
People also often forget that this was China's strategy too. Photos famously circulated of villagers with polearms guarding makeshift roadblocks.
That was a pretty okay strategy!
That was a pretty okay strategy!
And besides, it's always fun to set up barricades!
Before we leave South America, let me mention a few things.
I have GOOD data for Peru and Brazil. Not because I'm smart or clever. I'm not that smart or clever. It's because some awesome followers sent me links.
I have some crap data for Chile. Would like better.
I have GOOD data for Peru and Brazil. Not because I'm smart or clever. I'm not that smart or clever. It's because some awesome followers sent me links.
I have some crap data for Chile. Would like better.
And I suspect decent quality data probably exists for Uruguay, Argentina, and Colombia. I'd also love to see data from a Caribbean country or two. If any followers know where to get this (I've looked!) pleeeeeaaaase let me kno!
Okay, on to southern Iberia!
A few notes on southern Europe. Greece allegedly has data of good enough quality that EuroMoMO, but I haven't been able to find it. Any Greeks out there, I'd love an assist!
A few notes on southern Europe. Greece allegedly has data of good enough quality that EuroMoMO, but I haven't been able to find it. Any Greeks out there, I'd love an assist!
EuroMoMO also reports weekly Italian data, but, again, I can't find the source material. The lack of Italian data beyond early April is pretty irritating. And if anybody has Slovenia, who I suspect have good data, that'd be sweet too.
Okay, on to Iberia!
Portugal is getting hit by a second wave. Notably, Portugal began easing up on its lockdown around May 4-18.
So if you want an example of "Lockdowns are super important," Portugal is your guy!
Portugal is getting hit by a second wave. Notably, Portugal began easing up on its lockdown around May 4-18.
So if you want an example of "Lockdowns are super important," Portugal is your guy!
However, if you believe this argument, then we're in for a nightmarish next year. Before guys, Portugal's out-of-the-house activities are still down like 40%.
Portugal's latest data still shows Portugal doing MORE social distancing than Korea, Japan, Hong Kong, Taiwan.... so maybe removing the lockdown led to a second wave.
Or, maybe ***not having a contact isolation program*** is the actual problem here.
Or, maybe ***not having a contact isolation program*** is the actual problem here.
That is to say, *maybe* lockdowns are useful in the first 1-6 weeks while governments set up contact isolation.
But if you don't have contact isolation, lockdowns are nowhere CLOSE to a guarantee of disease suppression!
But if you don't have contact isolation, lockdowns are nowhere CLOSE to a guarantee of disease suppression!
Maybe you get lucky and your lockdown actually breaks ALL the transmission chains.
But if you're not lucky, if ANY remain, and if you DON'T have an extensive contact isolation program, IT WILL COME ROARING BACK.
But if you're not lucky, if ANY remain, and if you DON'T have an extensive contact isolation program, IT WILL COME ROARING BACK.
Y'all.
Test
Trace
Isolate
Or, just, like, trace and isolate, testing is pretty optional TBH.
Test
Trace
Isolate
Or, just, like, trace and isolate, testing is pretty optional TBH.
So, how about Spain? How are things going there?
Well, Spain's latest revision to their data GREATLY increased their retrospective estimates for prior weeks they'd previously said were finalized, which was irritating from an analytic perspective.
Well, Spain's latest revision to their data GREATLY increased their retrospective estimates for prior weeks they'd previously said were finalized, which was irritating from an analytic perspective.
But whatever, NBD.
Here's Spain. The latest data has a lot of error potential in it due to asjutment uncertainty, so I'd say a 2nd wave is *possible*.... but I've been saying that for weeks and keep revising the latest deaths downwards.
Here's Spain. The latest data has a lot of error potential in it due to asjutment uncertainty, so I'd say a 2nd wave is *possible*.... but I've been saying that for weeks and keep revising the latest deaths downwards.
Here's Switzerland. Life returning to normal. Is it the lockdown? Maybe, but Switzerland's death peak came too early for that, and Switzerland's official tracking data shows R falling well BEFORE the lockdown.
Just gonna bury in here to troll coronatruthers:
We should force contacts of infected people into isolation
We should set up roadblocks between states
Testing should be compulsory
School should remain closed for a long time
I ain't no COVID-anarchist!
We should force contacts of infected people into isolation
We should set up roadblocks between states
Testing should be compulsory
School should remain closed for a long time
I ain't no COVID-anarchist!
Okay, so, what about France? France didn't publish new data this week, sadly. But here's last week's data.
