Did Sweden end up taking the best approach?

260,931 Views | 1675 Replies | Last: 6 mo ago by Enzomatic
GAC06
How long do you want to ignore this user?
AG
Do you think that has to do with the recent study about T cell response and lack of antibodies?
HotardAg07
How long do you want to ignore this user?
AG
Keegan99 said:

So Sweden has 26% "excess mortality" but somehow is right in line with their population-adjusted all-cause mortality for the last three decades?

That whoever produces those charts don't make clear precisely what figure is +26% above normal is a sign that that whoever produced those charts is deliberately attempting to mislead.

Keegan, candidly, you question every data that doesn't agree with your pre-conceived notions and you post some absolute junk data from twitter on here yourself to defend your preconceived notions (some el gato loco guy on twitter with some hastily made excel graphs with way more egregious issues is one common tact I've seen you take). You were out here definitively saying that the deaths for COVID would definitely be less than 60,000, despite the obvious and well-known issues with the IHME model that was a geometric curve-fitting model based on China and Italy data made only to predict peak deaths and had uncertainty going DOWN in the future months to zero in 2 months. How could you not question that? When I pointed out the issues to you, and pointed you to superior models who did a better job of analyzing the uncertainty in the future months, you ignored them and continued to push the narrative that the virus would soon dissipate to zero with far less deaths than anybody predicted.

What I posted was a graph from a tweet of a Financial Times writer. He has been posting those graphs regularly on his Twitter for weeks now, along with them being available on the Financial Times website (which is like the UK WSJ). He has discussed in great detail what his methodology is already and he also openly discusses issues he sees with the various data sources he works with. He makes graphs like this, which display the different ways to view the excess mortality data:


I am as skeptical of his analysis as I am of the ones you post, because he is a JOURNALIST so he is trying to use his data visualization to tell a STORY which may bias his approach, but I don't think you are applying the same rigor candidly and it's frustrating to me because I know you're a really smart person with strong quantitative skills who I and a lot of other people look up to for good information.
Keegan99
How long do you want to ignore this user?
AG

Quote:

You were out here definitively saying that the deaths for COVID would definitely be less than 60,000


Yep. IHME used fraudulent Chinese data to predict. My bad for trusting the "experts".

tremble
How long do you want to ignore this user?
AG
The fact that we have to deal with three different graphs so the "narrative" can be set is disgusting in the first place.
Keegan99
How long do you want to ignore this user?
AG
"Source: FT analysis of mortality data"

Tautological citations!


The graph I posted has data from http://www.statistikdatabasen.scb.se/pxweb/sv/ssd/START__BE_BE0101__BE101G/ManadFoddDod/
HotardAg07
How long do you want to ignore this user?
AG
Keegan,

Two seconds of research. Financial Times CV analysis is here:https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06441

Found at the bottom of the page:

Please use the sharing tools found via the share button at the top or side of articles. Copying articles to share with others is a breach of FT.com T&Cs and Copyright Policy. Email to buy additional rights. Subscribers may share up to 10 or 20 articles per month using the gift article service. More information can be found here.
https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06441

SOURCES Unless otherwise stated, national-level case and deaths data comes from the European Centre for Disease Prevention and Control. Data for the US its territories comes from the Covid Tracking Project. Data for UK deaths data comes from the Department of Health and Social Care. Data for Spain comes from the Institute of Health Carlos III . Previous versions of this page used data from the Johns Hopkins University Center for Systems Science and Engineering and Worldometer. Regional-level case and deaths data comes from official sources or verified local aggregation projects: the Covid Tracking Project (for US states), Montera34 (Spain), the Italian Department of Civil Protection, Public Health France, Jan-Philip Gehrcke (Germany), Canton of Zurich Statistical Office (Switzerland and Liechtenstein), the Public Health Agency of Sweden, the Brazilian Ministry of Health, the National Health Commission of China, and Tom E. White (UK). The full excess mortality dataset used for this analysis is freely available for download on Github. It is compiled from data originally produced by: Statistics Austria, Sciensano (Belgium), the Civil Registry of Brazil, the Chilean Ministry of Science, Technology, Knowledge, and Innovation, Statistics Denmark, the Civil Registry of Ecuador, Statistics Finland, the French National Institute of Statistics and Economic Studies (current and historic data), the German Federal Statistical Office, Statistics Iceland, Jakarta Provincial Park and Forest Service, the Israeli Ministry of Health, the Italian National Institute of Statistics, Statistics Netherlands, Statistics Norway, the Peruvian Ministry of Health, the Portuguese Directorate-General for Health, the Russian Federal State Statistic Service, the City of Moscow, the St Petersburg Civil Registry Committee, the South African Medical Research Council, the Spanish Institute of Health Carlos III, Statistics Sweden, the Swiss Federal Statistics Office, Istanbul Metropolitan Municipality, the UK Office for National Statistics (current and historic data for England and Wales), National Records of Scotland, the Northern Ireland Statistics and Research Agency, and the US National Center for Health Statistics. Help us improve these charts: Please email coronavirus-data@ft.com with feedback, requests or tips about additional sources of national or municipal all-cause mortality data. Thank you to the many readers who have already helped us with feedback and suggestions. We continue to incorporate your suggestions and data every day. We will respond to as many people as possible.
tysker
How long do you want to ignore this user?
AG
Keegan99 said:

"Source: FT analysis of mortality data"

Tautological citations!


The graph I posted has data from http://www.statistikdatabasen.scb.se/pxweb/sv/ssd/START__BE_BE0101__BE101G/ManadFoddDod/
Dont forget this project is the self-proclaimed 'best measure for international comparisons of Covid deaths' even though there are known inconsistencies with COVID-death tracking within hospitals, by doctors and among cities and states within the US alone. I'm not sure how one can easily compare data between countries when there are different definitions of "COVID death" between people, institutions and countries. I'm not exactly sure what the end-goal of this project is supposed to be.
HotardAg07
How long do you want to ignore this user?
AG
The FT reporting does not rely on COVID coding, it is reporting on changes in all-causes death.
tysker
How long do you want to ignore this user?
AG
HotardAg07 said:

The FT reporting does not rely on COVID coding, it is reporting on changes in all-causes death.
Maybe I'm just being dense but I dont think that's accurate at all. The author even admits as much in the twitter feed:
Quote:

We're tracking Covid deaths, in this particular case using excess deaths as a proxy. If total deaths fall due to e.g fewer road deaths & homicides as in South Africa, that's great news, but it doesn't mean they had negative Covid deaths.

The issue I have is that COVID deaths are inconsistency measures across locations but also they are not taking into account what others refer to as 'negative deaths' (lives saved due to COVID-related shutdowns). It just appears inconsistent in its measurement from the start. Also wouldn't we like to see per capita figures? So again I'm not sure what the end point of the exercise is and don't get me wrong, I think it's kinda cool but I'm not sure what value it adds.


editing to add: I think I have a better grasp of the excess death proposition the FT is claiming and again I think it's cool but I'm not sure what exactly it's telling us. COVID is obviously pulling forward deaths. As has been asked numerous times (I think by Keegan and others) how many of COVID deaths would have died were going to die anyways over the next 6-24 months? If the number is different than a random walk, why?

Also I'm not sure why they show 65,700 excessive deaths in the UK and only 43,000 confirmed deaths. Why the large disparity. The difference is only 7,000 (122k vs. 115k) in the US. There does seems to be a positive bias toward excessive deaths versus confirmed deaths but they aren't reducing lives saved as negative deaths so something else has to be going on in the figures. Are COVID deaths being over-reported or under-reported?
DadHammer
How long do you want to ignore this user?
AG
Deaths in Sweden have dropped big time. Are they actually near herd immunity?

https://www.worldometers.info/coronavirus/country/sweden/
GAC06
How long do you want to ignore this user?
AG
The curve in the death graph looks familiar.
Complete Idiot
How long do you want to ignore this user?
DadHammer said:

Deaths in Sweden have dropped big time. Are they actually near herd immunity?

https://www.worldometers.info/coronavirus/country/sweden/
They had their highest number of new cases recorded on June 24th, recently, but I have to assume that is because they are testing way more and/or testing a smartly targeted group of people.

I do think deaths is the ultimate measurement since it is death we fear, not getting sick, and the deaths/day curve looks very similar in many places with outbreaks. AS long as deaths have been recorded with the same criteria throughout the pandemic, that should mean either it spread way more quickly than we could detect and the at risk population was hit early and those at most risk died - but with them no longer in the population, death rate drops. OR it means we are treating it much better than early on, early on it was heavy ventilator usage and other treatments and maybe current treatments are more effective. OR it means the virus has become less deadly than it was early on, which I've read isn't uncommon in novel virus outbreaks.

But yes, based on that deaths per day curve, it seems they have reached herd immunity or one of the other explanations above has applied, not matter the explanation it means they are way past the worst of this Covid outbreak in their country. (Unless they did social distance way more than I realize and if they change that behavior it might spike up a little again, but that doesn't sound like it is true)
Keegan99
How long do you want to ignore this user?
AG
DadHammer said:

Deaths in Sweden have dropped big time. Are they actually near herd immunity?

https://www.worldometers.info/coronavirus/country/sweden/


"Burnout" at ~20%

Complete Idiot
How long do you want to ignore this user?
TImeline of Diamond Princess: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7156812/

(just for data and analysis, no commentary)

There was a German ski village that reported 42% infected last week - or I should say, 42% who tested positive for antibodies. But if it is felt we can't distinguish between antibodies for COvid 19 from antibodies for other coronaviruses, perhaps that number has no meaning. HAving said that, I don't think we should necessarily believe the 20% burnout statement, at least not as it is stated. I do think that when the deaths start dropping greatly is when we THINK 20% have been infected, but it could be a higher percentage that what we think. Not that it matters necessarily, based on deaths per day the virus seems to burn through a population in 4-10 weeks.
BiochemAg97
How long do you want to ignore this user?
AG
Complete Idiot said:

TImeline of Diamond Princess: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7156812/

(just for data and analysis, no commentary)

There was a German ski village that reported 42% infected last week - or I should say, 42% who tested positive for antibodies. But if it is felt we can't distinguish between antibodies for COvid 19 from antibodies for other coronaviruses, perhaps that number has no meaning. HAving said that, I don't think we should necessarily believe the 20% burnout statement, at least not as it is stated. I do think that when the deaths start dropping greatly is when we THINK 20% have been infected, but it could be a higher percentage that what we think. Not that it matters necessarily, based on deaths per day the virus seems to burn through a population in 4-10 weeks.
The thing about antibodies is it doesn't really matter how you acquired them, they still do the job. The test uses a piece of the virus and the antibodies bind to that piece. If the antibodies bind to that piece, then they will recognize the virus in the body and do what antibodies do and fight off the virus. So, if people have antibodies because of some previous exposure to HuCoV, then they may not have ever gotten COVID but they would still be protected.


Not all antibodies are the same. Some are going to be more effective. Some some will bind in such a way as to prevent the virus from doing its thing and others will bind only to identify the virus to the immune system. And some will bind tightly and other not so much. Although, depending on the test design, there is usually a way to separate the antibodies that don't bind strong enough.


Bottom line, if 42% of the village was positive for antibodies, they may not have had it but should not be susceptible to it, so are effectively part of the herd immunity group.

Also, I think how quickly it goes through a population would have an effect on when the threshold is met. You don't suddenly stop at 20% (or whatever the magic number is) but if you have a high velocity then the overshoot can be large. For example, say the village had 1000 people, and the 20% is 200. Now mid cycle when about 10% are infected, you have a few of large weddings or gatherings that end up as super spreader events and you infect 150-200 people all at once, then you have overshot the 200 by a good margin.

It is also interesting to note that ~20% has been the number for other pandemics.
deadbq03
How long do you want to ignore this user?
AG
I found a well-balanced article a few days ago that highlights a key distinction about Sweden's response:
https://www.bloomberg.com/news/articles/2020-06-28/sweden-s-covid-expert-says-the-world-still-doesn-t-understand

Quote:


William Hanage, associate professor of epidemiology at Harvard's School of Public Health in Boston, said "Sweden's policy is unusual in that it took a much less stringent approach to preventing transmission, but interestingly it implemented those measures at a very early stage in the pandemic, before large amounts of community spread had occurred."
[url=https://www.bloomberg.com/prognosis?in_source=postr_index][/url]
"Sweden's approach may be sustainable in ways other countries' have not proven to be," according to Hanage. "It should be noted that lockdowns are a response to an imminent surge into health care, with the goal of stopping as many transmission chains as quickly as possible. The full accounting will only be possible after the pandemic."


Basically the case he's making is that Sweden was a very early adopter in adopting small measures, whereas UK, US, etc did nothing and then had to overreact. And the numbers show that the results from these opposite responses are about the same.

And to that degree, I agree wholeheartedly with the premise of the thread. I'd rather have taken Sweden's approach than our approach if we were going to end up with similar numbers. It would've been much better to see society take this seriously early but no shutdown rather than do nothing for weeks and then have a complete shutdown.

But it's important to remember that we have a much more individualistic culture than Sweden... the same type of person who's praising Sweden for this action of lesser government intervention would be appalled at the rest of the politics of Sweden. In my opinion the two are somewhat related; that is, it's easier for Sweden to accept more healthcare risk since their healthcare is public and universal. I'm definitely not trying to debate the merits of that system, but during a public health crisis, it does give them greater control over what the medical response is and a greater certainty of who has access to resources.

On the flip side, the numbers also clearly show that an earlier, stricter response does better than Sweden and US. You simple can't deny that Sweden's neighbors have some of the best Covid numbers in the world. The ethicists and economists will debate which response is "better" for society but there's no denying that taking a firm, early response saved lives.
fig96
How long do you want to ignore this user?
AG
Quote:

In my opinion the two are somewhat related; that is, it's easier for Sweden to accept more healthcare risk since their healthcare is public and universal. I'm definitely not trying to debate the merits of that system, but during a public health crisis, it does give them greater control over what the medical response is and a greater certainty of who has access to resources.
That's an interesting observation and I think a good point. When your system is controlled holistically it's a lot easier to be able to do everything from allocate resources to ensure people that they'll be taken care of.

And not to dive deeply into the political side of it, and I certainly don't have "the" answer to how to fix healthcare, but I find it hard to believe that folks of any party can look at 12%+ unemployment and people losing their healthcare in a pandemic, people going bankrupt if they get cancer or another long term disease, etc., and not say "we can find better ways to do this".
DadHammer
How long do you want to ignore this user?
AG
Don't forget, Sweden is counting more deaths than its neighbors.

So hard to compare.
ETFan
How long do you want to ignore this user?
DadHammer said:

Don't forget, Sweden is counting more deaths than its neighbors.

So hard to compare.
Source?
DadHammer
How long do you want to ignore this user?
AG
Go through the thread there are references all over this thread.
Complete Idiot
How long do you want to ignore this user?
DadHammer said:

Go through the thread there are references all over this thread.
You repeatedly saying they count deaths differently in multiple posts would not count as unique references.

To the person who asked the question - I don't know the answer, but maybe it's in here somewhere: https://ourworldindata.org/covid-excess-mortality (sorry I don't have time to look)
Keegan99
How long do you want to ignore this user?
AG
Sweden is pretty aggressive with counting COVID fatalities.

https://www.barrons.com/news/sweden-coronavirus-deaths-pass-4-000-health-authorities-01590408903

Quote:


State epidemiologist Anders Tegnell of the Public Health Agency stressed countries' death tolls should be compared with caution.

"In Sweden, anybody who has the diagnosis of COVID-19 and dies within 30 days after that is called a COVID-19 case, irrespective of the actual cause of death. And we know that in many other countries there are other ways of counting that are used," he told AFP.


Keegan99
How long do you want to ignore this user?
AG


Thread discussing some of the Nordic data, from an all-cause fatality perspective.
goodAg80
How long do you want to ignore this user?
AG



deadbq03
How long do you want to ignore this user?
AG
DadHammer said:

Don't forget, Sweden is counting more deaths than its neighbors.

So hard to compare.
I don't doubt that, but I do doubt they're inflating numbers enough to have per capita deaths over 5 times higher.
Keegan99
How long do you want to ignore this user?
AG
DadHammer
How long do you want to ignore this user?
AG
It has been answered multiple times earlier in the thread.
DadHammer
How long do you want to ignore this user?
AG
Thank you, I don't feel a obligated to keep reposting things already pointed out way back in the discussion. People need to read through the discussions themselves and not expect everyone to go back for them.
Keegan99
How long do you want to ignore this user?
AG
https://threadreaderapp.com/thread/1280383235604860928.html

Great thread analysing Sweden in a historical context. Lots of visuals and explanations.

Bottom line? While Sweden has had a very little more excess death in comparison to recent years, which were quite low, in no way is it far outside of the norm. (The true outlier in the data set is 2018-2019, which was especially low.)




Also, the IHME projections for Sweden were GARBAGE. Look at this one from early May, when deaths for Sweden were already on the decline and IHME still predicted looming disaster. You don't get to be THIS wrong when the trend was already clear and still claim any semblance of credibility. This is just fanciful wishcasting based on nothing more than the doctrinal belief that lockdowns are effective.


DadHammer
How long do you want to ignore this user?
AG
https://www.worldometers.info/coronavirus/country/sweden/

Looking at new infections - they spiked in June but are coming down again.

The death rate 7 day average is down to about 3.

Just looking at the graphs covid in Sweden is either over or just about over. Took about 12 weeks to build and then die off. So lots of hope the rest of the world sees the same approximately 12 week burn and then it dissipates. Tough 12 weeks for people though.

Still praying for a treatment breakthrough to end this sooner.
amercer
How long do you want to ignore this user?
AG
It's hard to think of a place less like Sweden than Houston TX.

Keegan99
How long do you want to ignore this user?
AG
Yes, Houston has a density of 3k per square mile. Stockholm is 18k.

Helps explain why Houston is likely to fare a bit better.

Already we are seeing this, as Texas all-cause death rate is below expectation for February through June, at 98% of expected per the CDC.

(See table 2 here: https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm)

But in fairness, we're not really out of the ordinary in that respect. Most of the country is at or below all-cause death expectations.
deadbq03
How long do you want to ignore this user?
AG
To be fair, if people are heeding measures to reduce Covid, they're also going to reduce exposure to other forms of death. The plethora of insurance companies giving token rebates on auto insurance speaks to how much fewer auto accidents are occurring. I'd wager there will also be fewer deaths from other transmissible causes (flu, for example) this year.

On the flip side, I'd bet stress-related deaths are increasing (suicide is the obvious one; stress is also linked to increased risk for heart attack and stroke).

It's gonna take years for "real" numbers to get sorted out on this. And even then people will manipulate them to fit whatever agenda they have.
Keegan99
How long do you want to ignore this user?
AG
Doesn't look like the "offset by fewer automobile deaths" theory holds much water.

https://www.wsj.com/articles/the-roads-are-quieter-due-to-coronavirus-but-there-are-more-fatal-car-crashes-11588152600

http://www.wbur.org/bostonomix/2020/05/04/massachusetts-roadway-deaths-coronavirus

https://www.npr.org/sections/coronavirus-live-updates/2020/05/20/859829779/as-states-locked-down-in-march-motor-vehicle-fatality-rate-spiked-by-14

https://abc7.com/traffic-fatalities-california-highway-patrol-speeding-covid-19/6198603/


The absolute number of traffic accidents did fall, reducing claims, but fatalities didn't budge that much, if at all. And in many locations roadway fatalities increased.

deadbq03
How long do you want to ignore this user?
AG
Interesting. Maybe we can add that to stress-related when it's all said and done.

Come to think of it, I remember early in the lockdown having to drive to Fedex before they closed to get a package I missed. Coming home at just after 8:00 PM I-45 was the damn autobahn. I always drive pretty fast (5-10 over limit) and dudes were blowing by. I chalked it up to bored kids not having school the next day.
 
×
subscribe Verify your student status
See Subscription Benefits
Trial only available to users who have never subscribed or participated in a previous trial.