> With 39.26 deaths per 100,000, Sweden's mortality rate is not only higher than that of the U.S. (29.87 deaths per 100,000) but also exponentially higher than those of its neighbors Norway (4.42 per 100,000) and Finland (5.56 per 100,000)
Small thing, but the usage of 'exponentially' in this context, comparing several discrete datums, really does bug me. If humanity's greatest weakness really is our failure to understand the exponential function, then we don't help ourselves by using the term as a clumsy proxy for 'substantially bigger'.
That, but also a misunderstanding of the strategy: herd immunity works on the thesis that the total number of victims is independent of speed of spreading. So naturally, you'd expect more dead/hospitalized sooner under this (lack of) regime, but also a shorter active period. I.e.: a sharper peak, but roughly same area under the curve.
The reason most countries chose not to do this, is because they expect the peak to be too sharp to handle all the hospitalizations.
Which is of course why choosing a strategy is and will always be an informed guess. We won't know which strategy works best, which assumptions turned out to be true of flawed, until after the show.
Sweden provides us with an upper bound. I think they're showing it's perhaps not as bad as we once thought it might be. Which is why many countries are now relaxing or looking to relax rules to regimes more similar to Sweden.
It provides us with an upper bound in Scandinavian countries with populations that behave like Swedes.
How does that generalize to other countries with different densities, distancing compliance, healthcare, etc?
Another way to read this is that Sweden is quite a lot worse than it's neighbors, but people there still followed the rules so it's actually a lower bound?
> If humanity's greatest weakness really is our failure to understand the exponential function
If this is in regards to dismissing coronavirus concerns, I don't think that was the problem - more a case of people assuming the exponential growth would be magically stopped earlier by some governmental force or whatever.
When the first case in the city I live in was announced, I remember thinking to myself, it's just one case, they're now quarantined and they've probably done proper contact tracing and the country is preventing people coming in from infected regions abroad. I was perfectly aware how quick exponential growth happens, but to some extent had faith that "the system" would prevent it from going out of hand.
It's quite a famous quote from some notable smart person who I don't remember. I think they were talking about population growth (perhaps ironically). My intention was not to make a sly point about reactions to coronavirus.
Oh I never imagined it could be contained. It's not the 1950's where folks lived and died within 2 miles of the spot they were born.
People travel promiscuously today. Even if only 1 in 100 did, it'd defeat 'containment'. And far more travel than that.
Heck we are more careful with pets and cattle, than with people. Because, disease. People were somehow exempt because it's inconvenient to quarantine - takes time. And now we're reaping the result of this.
About "being more careful with pets and cattle", I have an example that's more literal than you think: In Germany there have been several Covid-19 outbreaks between workers at meat processing plants, which highlight their bad living conditions. One satirical TV show promptly came up with the idea of also having food labels for the living conditions of the workers, besides the labels for the conditions in which the animals are kept...
Summer had very little to do with it, IMO. SARS spread almost exclusively in a temperature controlled environment, hospitals. Hospitals around the world learned their lessons from the mistakes made in Toronto, applied them effectively, and stopped the disease.
The US epidemic did get a bit bigger, into the low hundreds, but they were spread across half of the states. That still looks a lot more like containment than an outbreak which happened to die off.
I see that now. Interestingly, no deaths in the USA though worldwide the rate was ~10%.
To me, spreading to 29 states then disappearing seems more like dying off than containment. In fact, it seems more like "uncontrolled spread" followed by a lowered R-factor in the summer. Which is what the doctors in my daughter's medical school initially suggested would happen to Corona-virus, "just like SARS". Unfortunately this one has a much higher R-factor.
Indeed, this was also one of my hypotheses. The problem is that there was much more asymptomatic transmission that it was initially assumed.
Even later, when the exponential got out of control, I expected the states to step in with much more drastic measures than the lockdowns we have witnessed. I thought there will be tanks on the streets, I thought one day we will wake up and see the military everywhere à la 1968 Prague.
While I have no problem criticizing some of their choices and (IMHO) lack of transparency, I do take some issue with the comparisons (not yours) in the text you quoted.
It's true that the death rate is higher, but an often ignored factor is the (differing) chosen qualifier for covid deaths between countries [0].
Compared to UK/Italy/Spain/France/US/etc, Sweden doesn't have many "unaccounted" deaths during the pandemic [1].
This is a fair point, although I tend to think (based on nothing other than intuition) that Sweden likely has reasonably similar metrics to its Nordic neighbours?
I'm a technologist but somehow have found that my friend group is mostly English professors, and they are not idiots. This conversation comes up every time we hang out, with them stating that "exponential just means a lot."
Examples help quite a bit. Tell a story, a gamma ray causes a mutation in the very room you are in and causes a common virus to be virulent and deadly. It takes 2 days to make you sick enough to spread and is virulent enough for each person to spread it to 8 people on average.
So 8^11 is about the people on the planet, so in 11*2 days everyone on the planet will have it. Sure one a noticeable fraction of the planet have it, the rate of spread will slow, that's called heard immunity.
Then maybe discuss how with such quick growth that some countries drop the ball. What looks like a over reaction while it's happening looked like a under reaction in hind sight. Mention specifically Wuhan, NYC, and Italy as examples.
Might also be worth mentioning that in the 1918 pandemic the first wave broke records, made the news, encourages changes like social distancing and masks. But it was the second wave that was worse, caused by opening up things to quickly.
Take it as a synonym for "order of magnitude" and make your peace with the term? In scientific (aka exponential) representation, the exponent is actually higher in this case...
I think it's worth trying to be prescriptivist here. Exponentiality is a specific, important phenomenon, and it frustrates me to see language diluted in this way. I wouldn't care in the context of a reddit post, but NPR should know better.
NPR should know better and only use the rigorous mathematical definition of the term exponentiality?
If they would’ve said “an order of magnitude higher”, you would be bitching that 8x higher isn’t even a single order of magnitude. You just want to complain.
Words have flexible definitions outside of academics and science, let it go. It saddens me when pedantry is the top comment on an HN post.
In common English, I'm not sure "exponentially" has ever meant 2^x. I think it may have moved straight into English from mathematics already wrecked. I thin the majority of the time I see it, it's being used for x^2 or similar polynomials.
You won't be able to "solve" this until the general population is aware of the differences between those two functions, which is probably pretty hopeless.
Nit pick. There is another imprecision lurking in the quoted statement. Epidemics are subject to power laws, making it misleading to take averages over large populations. US mortality rate may be 29.87/100k, but that is an average weighted by population between 50 states with different geographies, demographies and policies, for example NY's 120/100k vs. CA's 10/100k.
> we don't help ourselves by using the term as a clumsy proxy for 'substantially bigger'.
I thought about it too and I noticed more unfortunate conflicts in colloquial use. For example "weight" when we mean mass and "phenomenon" when we refer to remarkable events or people. I recognize languages evolve, so I'm not upset about it.
Well, that would make sense when describing the relationship between two functions, but not between two datapoints, right? Since exponentially pertains to growth, not magnitude.
I think we should call two points that are separated by more distance than the first point squared. For something to be exponentially away from something else, it would have to be at least itself squared. So 26 is exponentially far from 5 (5^2).
“Exponentially higher” can refer to magnitudes, not growth. If I have $1000 (10^3) today and $100 (10^2) some time ago, I have exponentially more dollars now than then even if that value grew linearly over that time period and continues to grow linearly. I won’t have exponentially more over the next time period, because it’s not growing exponentially, but it’s still correct to say at that point my bank account is exponentially higher.
Lines are infinite, so you can pick two points on any line with a nonzero slope where the second point is exponentially higher than the first. That doesn’t mean the line is an exponential. Both of these views only look at the magnitudes of the two points with respect to one another and not the trend of the growth of the curve from which they were sampled.
in literature "exponentially" can be used in a lot of ways that are appropriate but not mathematically correct. you can use it as a euphemism, as a metaphor, as an adjective, like a loose comparator, etc... everything is valid.
the same applies to a lot of other mathematical concepts (i.e. the double negative, etc...)
I don't think it's asking for an unreasonable amount of rigour to require that terms like 'exponentially' not be used to describe discrete differences in magnitude? I mean, you might as well say 'additively', since for any given (x, y) pair of positive numbers there is some n such that x+n=y, just as there is some n such that x^n=y.
yes, I feel the same way, especially in data/scientific oriented articles. Unfortunately, most data journalists are writers first and scientists second, which kind of biases the way things are presented in written form.
I'm Swedish but I live in Scotland and I've been watching the Swedish approach to COVID with great interest. I remain fairly sceptical but I also thought this interview[0] with Johan Giesecke who was the previous state epidemiologist and is advising both WHO and the current state epidemiologist interesting. In particular it seems like Sweden messed up the response in elderly care which is something that everyone is admiting to.
I do think it's too early to say whether Sweden's approach is as bad as it currently looks, likely this is not something that can be measured and compared to other nations for years.
About "messing up the response in elderly care": I wonder what the strategy was there - if you want to protect the elderly, you also have to limit the freedom of movement of the employees that care for them and of the visitors, otherwise one of them will inevitably bring the infection in. But even if you do that, these persons can be infected by the people they live with - where do you draw the line? So I don't think it's about messing up in elderly care, it's more that the whole plan is unrealistic, it's just in elderly care that it shows...
You can only accomplish so much, but many places didn't even do the basics. I don't know what Sweden did specifically, but there are stories from many countries of elder care homes being required to accept coronavirus patients.
I don't condem their approach on this situation, not having a lockdown. What I don't aggree with is not making wearing a mask mandatory. At least this should've been made a must for everyone
We don't talk enough about the goals for wearing masks. My mask may not be helping me because I'm not disciplined enough about wearing it, washing it, touching it. But if I cough or sneeze into a mask, the people around me are a lot more protected.
The reasoning I heard was - you breathe on the mask it gets moist and has some mucus on it maybe, maybe its enough for the virus to multiply, making a nice little culture dish, particularly if you wear it all day. I have no idea, this is just the reasons that I've heard put up.
Bacteria and fungus could reproduce inside a web mask, moisture and a source of food (spit, sugar, mucus, protein, or similar) is enough. Virus however need targeted cells to invade and take over the cell mechanisms to reproduce the virus. Said target cells for COVID19 are unlikely to be reproducing in your mask.
So buy (or make) 2 masks, use them on alternate days, and make sure they dry out completely between uses. If particularly scared/careful wash and dry them between uses.
(from below) Yah, think I misremembered that - I think it was because they were worried people wouldn't maintain the masks well and bacteria would grow.
You may find this hard to believe that but there is very little evidence that masks work for this particular disease,
The group think is strong with the masks - but in reality, the scientific evidence is scant and inconclusive at best. This is why CDC never recommended them in the first place.
There are a large number of people that put all their faith in masks, because what else is there? They fervently hope and really want masks to work. And I understand that I wish masks worked, but do they really? The scientific evidence is just not there.
This post will be heavily downvoted for the same reason. It is a psychological thing. When they put on the mask people feel like they are doing something.
People strongly believe that masks HAVE TO WORK it just MAKES SENSE THAT THEY WORK, HOW COULD MASKS NOT WORK?, in reality, the evidence is not there. Masks feel to me like a superstition or flat earth belief. Where is the evidence?
Questioning the efficiency of masks is heresy, you'll get banned from Youtube or Facebook for questioning masks, but not for questioning vaccines or for promoting vitamin C as a cure for cancer ...
Why is that? Because there is no evidence, if there were you'd be pointed to that.
Do you have sources or articles? I’ve read quite a few and I’m unconvinced to say the least.
The study showing it doesn’t stick long to to surfaces lends some support to touching your face but I’d rather be safe than gamble on those probabilities and a longer term accumulation.
The worst and most common argument I’ve heard is that “it only prevents you from infecting other people, not them from infecting you”, as if that isn’t a major piece of the puzzle.
I know most infections happen with extended close contact and masks are pretty useless outdoors - which is why I only wear it in close quarters indoors. But I can’t imagine it doesn’t help at some level with the way it limits the velocity of your exhales and protects touching, even if it’s 10-30% safer than it’s worth it IMO.
I take elevators and live in a building with thousands of other people sharing the same public spaces. There seems plenty of opportunity.
First, coughing absolutely is a coronavirus symptom even if sneezing isn't.
Second, I sneeze every day, sick or not. And I don't even suffer from allergies. You're conflating it not specifically being a symptom with it not happening. Sneezing happens for different reasons.
Cyclists should be familiar with this debate, as it is closely mirrors the bicycle helmet debate.
Seems like it should work, and there is some weak evidence that it is better than nothing, but mostly we wear it as a talisman to ward off evil car-spirits. Despite weak evidence (because of it?) opinions run strong, and some cycling forums have to ban the subject in the interest of civility.
Maybe it's worse than the bike helmet debate, because the general left/right debate has been applied to it. So you get people berating those in masks, or castigating the President when he doesn't wear one.
I am as much for science based thinking as anyone but the facts that there is little proof that they work is silly to point out, when there is very little proof that they don't work. The lift of people wearing masks in public is so small that on even a tiny percentage of "working" it's worth it. People pointing out that there is little actual proof, as if there has or has been the proper amount of time and resources for proper peer-reviewed studies in a variety of environments, are likely actively harming people by encouraging more risky (even if it's a small percentage of chance of it being risky) behavior.
Stop and think about why people don't like masks. Think at a deeper level, don't jump to superficial explanations like people are lazy or selfish.
People don't wear masks because it is dehumanizing.
You start to treat your fellow citizen as a bag of viruses. Everyone is a suspect, and everyone is just a non-distinguishable face.
This is the reason, in my opinion, why people don't want to wear them. It is a lot more to it than an inconvenience.
And that is why it is only regularly worn in societies where dehumanization is already common, where there are too many people and none caring about each other. It is also common in societies where freedoms and innovation are scant and it is a tool to suppress society.
At that point it does make a difference if you should be compelled to wear something that does not actually work.
I understand that many people feel masks are dehumanizing, and I respectfully suggest that they need to get over it. In cultures where mask use is common (which now includes many areas of the US, so your sociological explanations don't make a ton of sense), masks don't feel dehumanizing; they feel like dress shoes or a tie, a bit uncomfortable but entirely normal and required in certain situations. I've never felt pressured to treat my fellow citizen as a bag of virus, nor have I been treated like one myself.
I think you are wrong about mask usage and especially theories on places where it's common. My understanding, which could very well be wrong, is that high mask usage is common in cultures were being a burden to others is a faux pau. Masks don't protect you from others, they protect others from you. This is why in some Asian countries it's common to wear a mask when you have symptoms of a cold out of respect for others. Rather than being a sign of a dehumanized society it seems to be a sign of an emphatic society where people don't want to burdern others.
There's something uniquely western about the focus on the individual and wearing masks to protect oneself rather than protecting the collective from the individual. My understanding of the research on the topic is that mask are good at preventing spread by the wearer but not good at preventing infection of the wearer by others.
> People don't wear masks because it is dehumanizing.
There is nothing more dehumanizing then dying of a disease that could be prevented by someone wearing ~$0.50 worth of cotton across your face.
And evidence suggests it does work, just not hardcore, beyond-the-shadow-of-a-doubt conclusively. Given that they're stupid cheap and have zero long-term complications -- chloroquine will damage your organs; a mask should only leave gentle crease lines that go away soon after -- it's crazy to not expect people to wear masks.
Put on your damn mask and go about your business -- it's working fine in Taiwan and Korea.
The one thing that can be said about lockdowns is that it's not some obvious clear cut "it definitely helps a ton" type of thing. If that were true we should see nations and US states with strict lockdowns do much better than those without.
Instead some places with more strict measures (say the UK and Belgium) are doing worse than Sweden, and Sweden's curve is also flattening. The number of confounding variables just seem too large to make anything close to definitive statements about anything at the moment.
In Denmark, which took a different strategy similar to most countries, we never quite had a full lock-down as seen in many other countries.
Majority of retail continued throughout the "lock-down", but with an abundance of care taken in terms of keeping a 2m distance, limited number of customers in a shop (queue stretching out of stores is a common sight), and alcohol sanitizers everywhere.
The majority of businesses and institutions have reopened (schools and kindergardens being the first ones), and restaurants, café and cinemas have just recently opened (with appropriate restrictions).
What is interesting is that Sweden has seen the roughly the same level of economic contraction as Denmark, despite the vastly different policies.
Closing schools is an over reaction though. Denmark has significantly lower life expectancy than Sweden and will have it this year as well. Why? Largely alcohol and tobacco. Sure, Denmark having less alcohol and tobacco measures than Sweden is great for your economy, but limiting them like Sweden does would save way more lives and have way smaller effects on peoples lives than your current Corona measures.
All of these articles pointing at Sweden as if it was a failure really has didn't consider how minor effect it had in Sweden, I'd say that the other Scandinavian countries over reacted instead. Maybe not Norway since they seem to value life higher, but both Finland and Denmark will still have lower life expectancy than Sweden and if they cared about life they would put resources on other measures than Corona.
> A risk factor that we want to highlight, however, is the low vitamin D levels found in the Swedish-Somali population. Vitamin D status is strongly related to low sun exposure and dark skin. In two different studies, the great majority of Swedish women of Somali origin had very low levels of S-25(OH)-D (< 25 nmol/l).[3,4] In Finland, Somali women required more than twice the amount of vitamin D in order to maintain recommended vitamin D status. [5] In addition, vitamin D deficiency was twice as common, regardless of gender, in immigrants from Africa compared with those from the Middle East.[6]
Sweden only messed up if you consider the variable to optimise to be the number of deaths. If you attempt to minimise sum decrease in quality of life as a consequence of the pandemic, then as callous as it may sound, the severe lockdown strategies come out of the equation quite poorly.
Strictly, Sweden messed up if you consider the variable to optimize to be the number of short-term coronavirus deaths. If no vaccine or effective treatment comes available soon enough, then other countries may pay the same cost in mortality after reopening, just later. Theirs could even be worse, if e.g. they accidentally push their second wave into winter and the weather effect turns out to be significant.
And that's before considering the second-order cost in mortality. Society has existing treatments for various diseases that we know statistically would save lives, but that we don't pay for because they're not cost-effective, often defined around $100k per quality-adjusted life year (QALY) in the West. Assuming about 10 QALY lost per coronavirus death, the coronavirus has so far cost Sweden the same mortality as a $4B cut in health spending, about 1% of their GDP.
So if you favored an intervention that would have cut that mortality to zero but cost Sweden an extra 2% of GDP, then in a meaningful statistical sense, you're killing people--you're proposing to allocate limited resources from a place where they save more lives to a place where they save fewer. The only difference is that we see coronavirus victims dying now, but the future deaths due to an under-resourced medical system will be lost in the noise.
None of this is to say that Sweden's response is obviously right. For example, if a treatment is discovered next month that dramatically cuts the IFR, then they'll regret not delaying cases until after that. To criticize Sweden's response solely on the basis of their current death count is not useful, though.
Better optimize then for haircuts. How many haircuts can we trade for one death? Haircuts are so much more important that we should accept some number of deaths so that people can get their haircuts. What's this number to you? Callousness doesn't even begin to address the stupidity of this argument.
Depends on who's dying, but somewhere between 10k and 500k haircuts per death. (A teenager or child is "worth" more than someone in a nursing home - and if you go by life expectancy, it's more than 50x higher)
So 10k @$20/haircut is $200k, near the top of the statistical value of a year of life. (And the life expectancy of someone entering a nursing home in the US is less than a year) 500k @$20/haircut is $10m, or at the top end of what federal agencies put as the value of a life in the US.
At either conversion rate opening up for haircuts alone is remarkably silly.
If you think the debate is about haircuts, I'd wager you have a work from home friendly job with direct deposit where you don't have to decide between "staying home & going broke" vs "working & feeding my family."
I'd wager you don't get sarcasm. I'd also wager the people who are out of work aren't protesting with signs demanding haircuts. Then again, I might lose this second wager if these protestors are that stupid, dumber than even I previously thought.
What tangible things is the US not doing that others are?
I don’t see much difference from here in Canada and people seem happy with how it’s going and our politicians.
The main difference I see is the US has a higher obesity problem (30% US vs 20% Canada last I checked) and cities like Toronto are far less densely populated than NYC. Although Montreal is more dense and is having a worse time than Toronto.
I wish we closed our border faster instead of spending a month attacking the US for doing it.
The lack of masks and temperature screening seems to be common among all western countries. And something that seems like a obvious thing to me that could be changed. Our top doctor only started formally recommending masks last week.
The US has a spectrum of responses. San Francisco and the entire state of California shut down early and hard and therefore is doing pretty well both in absolute cases and per capita.
New York shut down much later and actually sent covid-19 positive patients to nursing homes and New York City kept the subway open throughout, so they represent roughly 25-33% of the US's cases and deaths on its own. If you add adjacent states (New Jersey and Pennsylvania), it gets over 50%.
Iowa never had to shut down because people aren't on top of each other.
Dig into the reports and media that you're reading and watching and pay attention to how many are about New York versus everywhere else.. NY goofed.
I agree that NY's response was bad; but how much of the difference between NY and CA is attributable to their response, and how much to pre-existing behavioral, genetic, environmental, or other factors? California (where I am) has lower population density than NYC, lower reliance on public transit even in SF or LA, an Asian population accustomed to wearing masks, fewer black people (who seem worse hit even adjusting for socioeconomic status, for a yet-unknown reason), warmer weather. How do you know what fraction of the difference in mortality is explained by those attributes vs. the government response?
I say this not to excuse sending infectious patients into nursing homes, but with the goal of recognizing the tremendous uncertainty in almost all aspects of society's management of this disease. For example, Japan seems to be doing great, with no major lockdown and testing that's missing ~80% of cases (assuming 1% IFR, their 830 deaths imply 83k cases; but they've found only 17k). I strongly suspect that the Japanese response would have brought disaster in NYC or Lombardy, but I'm not sure anyone can confidently say why.
> *Edit: To the people giving me shit, I am pointing out that this commenter is using a strategy (I don't know if there's a formal name) to tangentially push an agenda. 1. Sweden did the right thing, and 2. Brown people are 'invading' Sweden.
And this is libel. I’ll be seeking legal action if you don’t remove it (and if you are subject to US tort law). Also the term you’re looking for is “dog whistle”.
Blaming it on Vitamin D instead of socioeconomic policies of the ruling government is deflection and dismissal of more substantial class and race issues in responses to Covid, especially Sweden's which is broadly considered to be especially fucked up.
"Nope, Sweden didn't mess up, it's just that they have more brown people and uhhh vitamin D". The hell prior do you have that you think such a conjecture makes more sense than "Sweden's COVID response not only sucked, it sucked especially for poor and marginalized people"?
I don't want to jump in on the topic of whether Sweden's response was correct, but there is data to indicate that there is likely some other factor, possibly vitamin D deficiency, at play in addition to social factors. In the UK there's roughly a 4.2 times higher overall fatality rate for blacks than whites, and 1.9 after controls. That isn't enough to explain Sweden's rates, but it does look like it would plays a substantial role alongside socio-economic issues.
> Areas with large proportions of Black/African American residents are at markedly higher risk that is not fully explained by characteristics of the environment and pre-existing conditions in the population.
> There are now seven observational preprints (Abstracts and links below), based on COVID-19 positive patient data, indicating vitamin D deficiency and insufficiency, do indeed factor significantly in COVID-19 severity and mortality – for example, and thought provokingly, 10 to 20 times increases were seen in ICU and mortality in Asian studies.[5, 6] One further USA based study observes vitamin D may factor in COVID-19 infection.[7] Whilst differing in, size and exact approach, results of all preprints point clearly in the same direction.
> Concerningly, vitamin D is often low in high-risk COVID-19-groups, such as, BAME including African Americans, elderly in care-homes, and the obese. Low vitamin D is also increasingly common in young.[8] Studies suggest low ‘D’ factors in Kawasaki disease[9, 10, 11] ; occurrences are increasingly being linked with COVID-19.[12,13]
Yes there is a correlation between Vitamin D deficiency and COVID mortality. But you are not just saying that. You are saying that the effect size coming from Vitamin D explains a significant portion of the excess mortality in Sweden. That's just a guess and it's weird that you would make that conjecture when there's the wildly spectacular policy difference between Sweden and everyone else. Neither of the studies you linked support that.
I don't have time to read the studies you linked. If you did, what is the relative impact of socioeconomic status, the fact that the most polluted zip codes in NYC are predominantly black neighborhoods, and vitamin D deficiency?
The superficial bashing by foreign journalists of the Swedish strategy will become a case-study in confirmation bias. It's remarkable how eager a large variety of people are to justify their own policies by sloppy and superficial conclusions like for example deaths per capita, like lockdown or not is the only single variable that goes into that.
Small thing, but the usage of 'exponentially' in this context, comparing several discrete datums, really does bug me. If humanity's greatest weakness really is our failure to understand the exponential function, then we don't help ourselves by using the term as a clumsy proxy for 'substantially bigger'.