Population is supposed to be representative of LA county. Tests are from Premier Biotech with some preliminary data on false negative/ false positive rates shown. I suspect assumptions related to the test itself, rather than population, may be a bigger factor for any inaccuracies. There's also the question of whether antibody response implies immunity for SARS-COV2
Another interesting aspect of the result was 6% of men showed antibodies but only 2% of women tested.
Also:
--2.4% of people between the ages of 18 and 34 had antibodies to the coronavirus
--5.6% of people between 35 and 54 had antibodies
--4.3% of between 55 and older had antibodies
Edit: The key assumption that really, really needs to be triply checked here is the specificity of the test. They're assuming 99.5%, similar to the Stanford study I believe. If it's closer to 98% then prevalence estimate confidence intervals should include 0%, I believe. Also should be noted that the same PI on the Stanford study, Jay Bhattacharya, is also involved in this. Same criticisms likely apply here - more caution needed!
Second edit: On further reflection, I think these studies could do serious damage based on how the results are being announced by press-release - in this case, afaik there's no pre-print. In our current environment this isn't just science. Policy is being enacted and people may die because of poorly framing results. Everyone involved needs to take a step back and think through this carefully. Also - did anyone figure out why a hedge fund guy was an author on the Stanford pre-print? And why that same guy then published a Wall Street Journal op-ed about the study without identifying himself as a co-author? Shady stuff.
I’m very skeptical. Why were they so few hospitalizations in LA vs NYC if such a large % in LA contracted Covid? Is this suggesting that a much, much higher % in NYC have it (and the number of those with any serious symptom and thus need for hospitalization was higher)?
This virus is much more transmissible than the viruses we usually deal with in humans.
If a virus has a low fatality rate, but is extremely contagious and prevalent, hospitals will get overwhelmed. A small percentage of a big enough number is itself a big number.
The subway and high-utilization of buses are likely factors in NYC vs. LA. LA is much more representative of the typical US city.
Frankly, I'm not sure why so many on HN are surprised by this. The scientific data on asymptomatic carriers coming out of Italy, Germany, South Korea, and Iceland has been very clear on a majority asymptomatic frame for this virus. I think that maybe authorities and responsible people are concerned that this information could lead to people flouting lockdowns more.
I mean NYC publishes the number of positive tests each day and it’s dropped from 60% to 30%. So majority of people proactively getting a test because they are symptomatic are showing up negative. Hard to believe a very high percentage of all people have had Covid, then. https://twitter.com/natesilver538/status/1252317247785271297...
Isn't that the virus test, not the anti-body test? Maybe a fairly low percentage have the virus but a high percentage have been exposed and have the anti-body.
The optimistic take on this is: that means soon enough will be exposed that their immunity will stop the spread.
The pessimistic take on this is: we also don't have strong evidence that the previously exposed are immune, ie, have a strong enough immune response to stop the virus.
Those 30% numbers are still very high, the similar rate in Italy has dropped bellow 10% for 2 weeks now (it's the line-chart on this page [1] bellow "Tamponi giornalieri e contagiati"), with the latest week's figures somewhere between 5% and 7%.
Norway's number of positive tests was 5% at the current peak of the epidemic, for reference. Testing and isolation measures have been very effective here, so we have not had an overwhelming wave of infections (yet, knock on wood).
You are conflating a test for active virus versus an antibody test. my suspicion is that a very large percentage of people in certain neighborhoods of New York City are going to show up having been exposed and never presented symptoms. I just don't compute how else you can explain this with a virus that is so transmissible. we should also keep in mind that the number of deaths isn't necessarily indicative by itself of number of infections since deaths can be caused by hospitals being overrun.
Already 0.1% of NYC's population has died from COVID-19 - 10k deaths of 8.54 million people. That's... quite a bit higher than LA's current 576 of 10 million.
So either everyone in New York City was infected, or most probably 200-400k infected in LA is an order of magnitude of wishful thinking (e.g. ~2% false positive test rate.)
There's the theory of Vitamin-D deficiency. It would make some sense given NYC's climate and time of year. In LA, people likely are exposed to the sun year round. It would be interesting to see if other respiratory illnesses exhibit a similar pattern between these two cities, as I understand Vitamin-D levels have an affect those as well.
But also, NYC is considerably more dense, so diseases spread far more rapidly. Especially so considering heavy public transit use versus, what I would say is, extreme overuse of personal vehicles.
So like the fact that most of NYC, like most of Italy, and Spain is infected because they are a more dense city than LA is lying about the data because this is backed up by data in other countries like Korea, Germany and everywhere else with similar data.
Honestly, it seems more likely that California's much maligned car culture has served us well in this case - ensuring many of us maintain social distance which is impossible on NYC's subway and bus system. All the positive press about governor Cuomo's leadership and handling of the crisis seems a little weird given his failure to shut-down the subways (or at least curtail a but the absolutely most urgent ridership).
> Honestly, it seems more likely that California's much maligned car culture has served us well in this case - ensuring many of us maintain social distance which is impossible on NYC's subway and bus system.
California local and state government being ahead of NY in issuing shelter-in-place orders even though NY was being harder hit even before the divergent response also makes a difference. But, yeah, the hyperdensity of the NYC metro area is a big factor.
Previous studies suggested that the strain circulating in NYC is of European origin, whereas the California strain, arriving earlier, was of Asian origin.
It's always possible that the NYC one is more fatal. We don't have data, but it is possible.
Thing is, though, at the peak, NYC was seeing a 60% positive rate on tests and is still seeing something near 30%. That pretty much guaranteed that the vast, vast majority of NY cases were not detected, and still probably are not.
(Italy is back around 5-6% again, for example)
It must be the case that a dramatic percentage of NYC is infected. Certainly not more than 100% as some oversimplified maths suggest, and likely not 80%+ that would be very comforting -- but quite likely between 30% and 45%, given the Stanford and LA studies, and the NYC testing data.
Seems so obvious that this was the flu but far more contagious from the beginning because of the lack of herd immunity and vaccines. We are just getting verification.
It's not opinion is the logical conclusion from literally every random sampling study. Pretty much consistently 50x as many people have been infected that we thought. And if we made the current strategy with limited data why aren't we updating what we are doing with more data.
"Setting: Diamond Princess cruise ship, Yokohama, Japan (n=-634 tested positive).
Proportion: 18%"
Also:
"23 Residents of a Long-Term Care Nursing Facility King County, Washington
10 (43%) had symptoms, and 13 (57%) were asymptomatic."
But then:
"Seven days after testing, 10 of 13 asymptomatics developed symptoms"
That means that just 3 of 23 remained asymptomatic there too, i.e. 13%.
It is estimated that the people mostly spread this virus exactly in the phase where they still haven't developed symptoms, and the virus can be detected. But that doesn't mean they will remain without the symptoms.
Thanks, I actually didn't realize the number drifted so much over time. Good to know! I'm curious how many of them were actually seriously afflicted vs. a stuffy nose. I'd argue almost 20% actually asymptomatic is still huge in context.
Especially as we know the disease affects these demographics up to 100X more than the young.
For instance consider: "Emerging evidence suggests many more people are infected than tested. In Vo Italy, at the time the first symptomatic case was diagnosed, about 3%, had already been infected – most were completely asymptomatic." [1]
The most common symptoms of COVID-19 are fever, tiredness, and dry cough. Some patients may have aches and pains, NASAL CONGESTION, runny nose, sore throat or diarrhea.
>Frankly, I'm not sure why so many on HN are surprised by this. The scientific data on asymptomatic carriers coming out of Italy, Germany, South Korea, and Iceland has been very clear on a majority asymptomatic frame for this virus. I think that maybe authorities and responsible people are concerned that this information could lead to people flouting lockdowns more.
Because people on here panicked without thinking and don't want to admit it.
No one wants to say "Oh yeah, I ignored second year calculus and never bothered plotting logistic growth curves with the full spread of expected fatalities. Because of this I supported legislation that's started the next Great Depression. But I meant well!"
Are you claiming that everything is normal and would benefit from a complete reopen? Are you also claiming that the economy is in otherwise healthy shape regardless of the interest rate levels and global buyback programs that have been ongoing prior to the outbreak?
A complete reopen, no - crowded stadiums and amusement parks probably shouldn't happen. Things would never have been normal right now, and there's no silver bullet to make them normal again.
I can't speak for the original commenter, but I think the evidence increasingly suggests that we would benefit from allowing people to visit small groups of friends and run non-essential businesses.
It's rough because it's simultaneously very mild in the vast majority of people, but a horrific, suffocating death in a tiny minority, and highly contagious.
I don't know if there is a "right" policy answer on what to do with something like that.
As someone else said, the lockdowns are probably illegal, you can't just throw away freedom of association over the flu. But if we didn't have them, we might have seen another 100k dead..
". . . the lockdowns are probably illegal, you can't just throw away freedom of association over the flu."
Where did you study constutional law? Rights granted in the constitution are subject to many restrictions. There is no explicit "freedom of association" in the Constitution, although it has been held to be implicit in freedom of speech. Free speech doctrine includes a "time, place, and manner" test that's applied to restrictions on speech when the government has a "significant interest" in the restriction[1] (e.g., national emergency for a viral pandemic). I don't know how the lockdowns (which are mostly not full lockdowns) would fare if Supreme Court were to apply the time, place, manner test, but I wouldn't presume to offer an offhand opinion that they're "probably illegal".
Not op. I am not a lawyer and I’ve not studied constitutional law. What I recall from high school is that anything not expressly stated as illegal is legal for citizens and any power not expressly given to a state or federal government is unavailable to them.
A law has to exist that gives states and the federal government the ability to do what they are doing and that law needs to not violate special protections built into the Bill of Rights.
With a state of emergency declared, some things can temporarily be enacted that under normal circumstances would not be legal.
I don’t know if all the lock down rules are legal, but I wouldn’t be surprised to find out that governments have overstepped their bounds.
You may be remembering something, but you seem to be remembering it almost exactly backwards. It sounds like you are remembering something about the 10th Amendment, which reserves to the states any power that is not granted to the federal government in the constitution:
https://en.wikipedia.org/wiki/Tenth_Amendment_to_the_United_...
It is C that is pretty close to true. Laws don't give states their powers. It is in fact the states that make their own laws, subject to some restrictions from the U.S. Constitution.
In part because all the headline numbers we've seen are CFR numbers which suffer from huge adverse selection bias -- and aren't directly comparable between regions for a lot of reasons.
As I've mentioned before, using CFR to determine fatality rate of the disease is like trying to determine the fatality rate of skydiving by measuring the ratio of people who go to hospital with skydiving injuries over those who recover. Obviously you'd think skydiving kills basically everyone who does it, but it just doesn't.
It's starting to become clear that the infection fatality rate of COVID is in the lower quartile of the range 0.25-1%.
And yeah, a huge portion of New York likely has it. It's super contagious.
Yeah, trying to do the basic math on the implications of these figures...[Not claiming they're true, that's a different question]
The article says 221,000 infected at a 2.8% rate, in "early April". So they're assuming a 7,892,857 population. Currently, we have 601 Deaths total in LA[1], which we can assume are a result of early april event on April 20. So a 0.002% "fatality rate" relative to total exposures, "Woohoo" (NOTE, this is an optimistic figure implied by this study, the standard 2% [edited] previously claim but still not equivalent to the flu, which has 0.2% [edited] relative people who get sick). Now, with weekly doubling: 221,0008 = 1,768,000 infection now = 22.4% - > 4808 deaths. However, past 10%, we can ask whether we start reaching saturation/diminishing returns. But that is a key question.
-- Let's look at NYC. There we have 14,500 Death. Using the article's implied figure, 14,500/ 0.002 = 7,250,000 infected on April 1, out of 8,398,748. So this implies an infection rate of 7,250,000/8,398,748 = 86.32239%, which, in fact, seems very implausible and which implies; a higher death rate among "vulnerable people". Maybe we're about to hit full saturation with NYC and see an immediate decline.
Now, the other bad* scenario is the possibility that a lot of sickness and death NOW is being driven by reinfection, that exposure isn't giving much immunity at all and instead of the unexposed dying, what we're seeing is the re-exposed dying and with this dynamic, the death rate settle into a steady and unacceptably high level. Note that Italy, which has seen a high death rate, hasn't a strong drop-off at all, just gradual plateauing. IE, guess what, one possibility is "herd immunity" just doesn't exist for this virus.
Anyway, time will tell, literally a week will tell if NYC has enough fatalities to make this 0.002 exposure-to-death rate numerically impossible.
There is none, it is just speculation/hysteria like so much else about this virus.
There's been only a very few cases globally where a person who previously tested negative later tested positive, for which a test giving a false negative is a much more likely explanation - especially as these cases were early on when tests were more unreliable. Otherwise - everybody else who tests negative does so consistently.
The virus is very similar to the original SARS (genome about 80% the same which is huge). It is a relatively stable by virus standards not mutating massively. In fact, if you had SARS your are likely immune from covid19: https://marlin-prod.literatumonline.com/pb-assets/journals/r...
So the general scientific consensus I have seen is that there's no evidence to suggest covid19 is some sort of supervirus that will reinfect people and defeat your immune system.
Yes, reinfection is speculation and limits the virus to infect the population is speculation. Beyond the situation of the virus growing exponentially, there are few certainties.
I add this point because most of speculative arguments are used to claim allowing the virus infect a large portion of population won't cause much harm.
Yeah, 2% is the standard and more plausible in many ways than 0.002% [edit, WRONG ,missed a decimal point, should be 2% & .2%]. My comments above are about what this study implies, this something new.
Italy has had 24114 deaths. If take those as being in Northern Italy with a 27,000,000 population, that's 0.8931 % of the entire population. Even adjusting for age, that can't squared with 0.002% death rate [wrong wrong, it's 0.08931%, which is possible but with a huge infection rate, which maybe].
I think you've misunderstood the comment by zucker52: your post is using 0.002 and 0.002% interchangably. It's quite hard to figure out what you mean when the numbers are off by two factors of magnitude in places.
For example:
> So a 0.002% "fatality rate" relative to total exposures,
You appear to be deriving this from the computation 601/221000. That gives .27%, not 0.002%.
> Italy has had 24114 deaths. If take those as being in Northern Italy with a 27,000,000 population, that's 0.8931% of the entire population
Yeah, 0.08931% of the population of Italy experienced fatality.
If we assume .2% fatality rate, that implies 50% of the population was infected (which might make sense with an extremely contagious virus but that's more than even the cruise ships).
The issue is you're not adjusting for the exponential fatality risk for old folks, and that those in Lombardy are the oldest in Europe. It's hundreds of times worse for old folks than for young folks.
The US doesn't have a hundred times less old people than Italy. The whole "don't worry about old people dying" attitude of those eager to "reopen the economy" is kind of despicable.
Florida has 20% of it's population over 65 and it's 21 million people might similarly be rather hard hit.
> The US doesn't have a hundred times less old people than Italy. The whole "don't worry about old people dying" attitude of those eager to "reopen the economy" is kind of despicable.
It's totally despicable, and it's not my position.
I propose we do what the UK proposed, and what Sweden is doing: keep all vulnerable folks inside, provide them with services such as groceries and any healthcare they may need on-site, and release the low risk to build a protective barrier of herd immunity around them. That should have been evident when I said "old people were hundreds of times worse off."
That's totally orthogonal to my point, which is that you can't project the mortality rate of a population that's overwhelmingly disproportionately affected, scale it linearly onto the world population and assume that'll be at all representative of the reality on the ground.
And yes, Florida may be affected very badly, and the plan may need to be adjusted accordingly.
"Now, the other bad* scenario is the possibility that a lot "of sickness and death NOW is being driven by reinfection, that exposure isn't giving much immunity at all and instead of the unexposed dying"
This would explain: NYC vs LA vs Italy. Why isn't this theory brought up more?
"Wang, whose full name has not been disclosed for privacy reasons, is one of more than 100 reported cases of Chinese patients who have been released from hospitals as survivors of the new coronavirus — only to test positive for it a second time in the bewildering math of this mysterious illness."
Because if reinfection so close to recovery were possible, it would make this virus really unique, and it doesn't really fit the available data very well. It makes much more sense that the virus takes a long time to clear after symptoms abate, and the tests have a really lousy false negative rate.
The common cold is infamous for mutating at a rate where you can catch the virus again two months after infection. And moreover, the common cold is also a coronavirus.
Moreover, if a lot of Covid infections are in the area and the virus is mutating quickly, a person might be exposed to a different strain soon after recovery and wind up infected, if the SAR-COV-19 is similar to the common cold. This would be hard to notice, since it would tend to happen as the virus was spreading rapidly anyway. It would become a problem only if people expected the virus to abate once everyone had been exposed.
No. The "common cold" is a collection of 200-something distinct viruses with the same/similar symptoms, only like 10-20% of which are coronaviruses. Most of them are rhinoviruses.
If what you were saying were true then we would be seeing a huge number of symptomatic re-infections and deaths in China and Italy. As far as I know the ‘reinfections’ have been asymptomatic. That suggests that the virus takes a long time to clear, not that it’s mutating.
There’s no evidence yet that COVID is mutating in a way that would cause reinfection. There haven’t been any documented confirmed cases, the speculative cases are much more likely false negative test results followed by actual positive test results.
The main reason should be that speculating on both this and the exposed-and-maybe-immune isn't that relevant for the more responsible approaches to managing the situation - isolation, quarantine, safety equipment, testing and contact tracing.
This kind of speculation is relevant to the more irresponsible approaches; "developing herd immunity" and "immunity certificates" but it mostly should be a reminder to "don't do that because you really don't know what you're getting into".
It’s not clear this is the responsible solution as there’s no evidence the entire process wouldn’t start again the minute restrictions are lifted due to the sheer infectiousness of the disease and the large number of people with no symptoms or limited symptoms.
"In epidemiology, a case fatality rate (CFR) — sometimes called case fatality risk — is the proportion of deaths from a certain disease compared to the total number of people diagnosed with the disease for a certain period of time."
As compared to the IFR: "The term infection fatality rate (IFR) also applies to infectious disease outbreaks, and represents the proportion of deaths among all the infected individuals. It is closely related to the CFR, but attempts to additionally account for all asymptomatic and undiagnosed infections." [1]
One's the risk that given you show up at a hospital, you die. The other is the risk given that you get infected, you die.
Case Fatality Rate; the rate of people who are confirmed to have COVID-19, and die. The problem is that the people most likely to be tested are those in the worst condition.
According to the New York data, the fatality rate is about 0.5% for tested persons without underlying conditions. Considering the gross selection bias it likely means actual death rates in the healthy population close to that of the flu. It is this data that shows the lockdowns were a complete illegal overreaction. Locking down nursing homes would have fixed most of it.
I don't understand what calculation you're doing here.
The worst flu season we saw in a decade killed (~80k / 300M) = .02% of the US population. NYC has seen ~14000 deaths, or .18% (~14k / ~8M) death rate for the entire population, and while the absolute worst may be over, people are still dying at a high rate in NYC.
The only way I can make sense of this is that you're somehow mixing figures relating to deaths in healthy populations (COVID) vs. overall death rates in flu (including older/unhealthy populations).
What exactly do you mean by worst-case here? The actual worst case would be if nobody took any precautions, which almost certainly would result in more than 0.18% of the total US population dying.
Let's see. NY has 965 death per million. Next state is New Jersey (which is basically "New York by another name" in this situation) with 493. Next is Connecticut (see above) with 372. Next is Louisiana - which had a grand idea of proceeding with Mardi Gras - with 285. Do you think things will get three times as bad in Louisiana than they are now, while not getting any worse in New York? Or which state would you say would be the next worst case, not today but in the future? I'd certainly bet NY (with adjacent NJ and CT) will keep this sad championship.
Depends on what you're comparing. I agree we can't assume that the rest of the country will see similar infection rates. However, it's less clear to me that fatality rates would be incorrect.
On the other hand, there's one very specific way that considering NYC gives us data we can't (yet) get elsewhere. While it's possible that we're underestimating disease prevalence by 50x elsewhere, it's basically impossible in NYC (almost 2% of people already have tested positive, with less than 10% of the population being tested).
Note that the test NYC is using only tells you whether the person has Covid-19 at that point in time, so it's entirely possible to test even 100% of the population and still substantially underestimate disease prevalence if you tested most of them at the wrong time (which you probably would!)
> However, it's less clear to me that fatality rates would be incorrect.
Depends on whether mortality rates can be variant on environmental conditions, such as viral load, population density, access to healthcare/ICUs, population socio-demographic profile, etc.
These are good points, and I think I ended up claiming a much stronger claim than I should've. There's plenty of reasons NYC could have a higher death rate than elsewhere. I don't know how precisely you could estimate the potential death rate from just NYC's data.
What I would still say is that it's reasonable to say "NYC has .18% of its population dead. It's extremely implausible that this disease is no more lethal than the flu is, when that's 10x the overall death rate from the flu. If you present me evidence suggesting that from elsewhere, it's going to have to be quite strong to overcome the evidence from NYC."
the flu would be way more deadly than it is if there wasn't a seasonal flu shot that gives herd immunity.
edit: it shouldn't be controversial to say that a thing that kills 30-80k per year with active measures of mitigation already something that would be more deadly without a vaccine. seriously reflect on that. the corona virus is something we need to take 100% seriously but also acknowledge that the flu is also very deadly... even more so without any mitigation (thankfully we have for the common flu)
The flu probably wouldn't be dramatically worse if we didn't have a vaccine. The vaccine isn't particularly effective all things considered, ranging from 10% to 60% depending on the year and how well scientists were able to guess which strain would be predominant.
The flu as-is causes 45,000,000 sicknesses every year in the US alone.
Nah, my calculus is basically as follows. The disease is not bad for people who aren't old, and who do not have pre-existing conditions. The data is pretty unequivocal there. The number of deaths under 20 worldwide rounds to zero. It's incredibly contagious (and contagious while no symptoms are shown), and a vaccine is 12-18 months away at a minimum. There's no way we're going to be able to contain everyone indoors, with no jobs, for 18 months while we await a vaccine that may not arrive. And that's just here in the US -- the weakest link dominates.
If we don't develop herd immunity ASAP, and instead pursue a course of lockdowns, as soon as we lift the lockdowns (either voluntarily or because people just walk out -- see the midwest), we'll immediately start playing rolling lockdown whack-a-mole as China is. The first contagious person who flies in from a foreign country (or domestically) without perfect lockdowns will re-ignite the wildfire.
We should do exactly what Sweden is doing and what the UK proposed: isolate and provide support services to those who are at risk, and let out those who aren't. I think it speaks volumes that Sweden's new infection rate stabilized at the same time as the rest of the world but without lockdowns.
It's really the only path forward. Is it perfect? Of course not. People will die. However, there's no world in which about 70% of the population won't get the disease before the vaccine arrives, so we need to control who gets it, when, and in what order, to minimize harm.
While I agree mostly with your assessment, it's worth mentioning that the lack of official lockdown in Sweden doesn't mean there isn't a lockdown. According to Google's mobility data for Sweden, their lockdown activity looks pretty similar to what we see in the US. People are voluntarily staying home.
Since Sweden doesn't have a lockdown, this statement is meaningless. A major objection to lockdowns is their involuntary nature. In general, there is a big difference between choosing to stay home, and being coerced (with the threat of force) to do so. One is "choosing to do what you believe is best", the other is "prison".
And there are big differences in social cohesion between countries. Voluntary efforts appear to be sufficient in Sweden precisely because there is a high level of compliance.
In the meantime, the coastal town that I live in is getting swarmed with out of town visitors every weekend. This is in violation of the shelter-in-place orders. Other towns have stepped up enforcement to combat this type of behavior, but our police department is small and does not have the resources. We shut down the parking lots, but now they just park in the neighborhoods. And to add insult to injury, we have a large senior housing complex at the entrance point to my neighborhood.
I have little faith that the US could achieve the type of distancing and isolation necessary on a voluntary basis. There is a sharp vein of individualism that runs through our society that works strongly against us in these types of situations.
For what it's worth, Sweden's goal is to get 70% of the country infected to achieve herd immunity. They likely quite rightly believe that's the only way to stop the disease once and for all. While it's absolutely not okay to be anywhere near a seniors residence, the rest is likely tolerated because it's kind of the unstated goal to the extent healthcare facilities remain un-saturated.
It's by no means a "hope and pray it goes away" or a "lets wait it out until a vaccine" -- it's a "let's get everyone not in a risk category infected as fast as possible so long as the healthcare system retains some excess capacity."
I think almost everyone you are debating in this thread would support opening things up provided a few conditions were met. The primary condition for most of us is we need adequate testing and contract tracing capacity. And the reason that this is necessary:
> healthcare facilities remain un-saturated
You can't simply look at the current burden on the health care system to guide the process. The virus has a 2-3 week lag time between when an infection cluster breaks out, and when the health care system starts to feel the impact. Without wide scale testing, we are just going to end up back in a lock down once the infection numbers start climbing again. And I can't think of a worse scenario for our economy than having to shut things down every other month because our government is to incompetent to implement a tracing program that multiple countries already have up and running.
The serological surveys are showing 50-100x infection rates of those tested. If tested healthy persons are dying at 0.5% that gets you to 0.01%-0.005% death rate in healthy persons. That number is lower than the all cause death rates of 25-34 year olds...
I see. You're correct that while the overall death rates in NYC are vastly higher than any season flu, it's mathematically possible that those are all deaths of unhealthy people.
It's a creative interpretation of the facts.
One important note: it's impossible that NYC is overestimating cases by 100x. Close to 2% of NYC has confirmed cases.
New York is at 1.3 percent positive with new cases flatlining. That fits very squarely with what you'd expect to see based on the surveys. Healthy people aren't dying by any significant number based on New York's own published data. The only other one I've seen is MA that has 97.5% deaths with comorbidity. This is a huge overreaction.
The other thing to keep in mind with death counts, especially in NY and NJ, is that all deaths of likely-infected people are being counted as COVID deaths regardless of cause of death.
This has included a couple lf suicides of folks who had respiratory illness prior to death, and one person who got in an auto accident and died of head trauma -- but he had tested COVID-positive.
Yes, most people right now who get a respiratory illness and then die probably did have COVID -- and it is possible that some folks are dying of COVID and being uncounted to offset some of the overcounting -- but with death numbers counted so loosely, it is hard to know the real story, and impossible to do simple maths using rates from one place at one time to compute rates at other times or other places.
I tend to think that NYC must be approaching saturation, and that the true new-case numbers must be falling there, but it's impossible to answer with certainty using only the numbers we have here on the internets.
> is that all deaths of likely-infected people are being counted as COVID deaths regardless of cause of death.
No, this isn't what's happening.
Doctors who are sure to the best of their knowledge and experience that the deceased had covid-19 will put covid-19 on the death certificate, and they will also say if they think it contributed to death.
That's not the same as "anyone who dies with covid-19 is being described as killed by covid-19".
I actually followed up with you on this one recently. This is in fact what's happening, and it varies by country/region/city to what extent.
In some countries they absolutely do count anyone who dies while in the possession of COVID as a COVID death, for instance Italy. " Italy’s death rate might also be higher because of how fatalities are recorded. In Italy, all those who die in hospitals with Coronavirus are included in the death counts."
“On re-evaluation by the National Institute of Health, only 12 per cent of death certificates have shown a direct causality from coronavirus, while 88% patients who have died have at least one pre-morbidity – many had two or three.” [1]
In New York they're counting speculative COVID deaths of anyone with respiratory illness even if they've never tested positive [2].
"A subtler issue is what to do when the patient has other serious medical conditions. If the person suffered from chronic lung disease, then became infected with the virus and died of pneumonia, the immediate or primary cause would be pneumonia as a result of COVID-19. The lung disease would be listed as a contributing condition, said Sally S. Aiken, president of the National Association of Medical Examiners." [3]
The CDC has guidance on this but it's fair to say its interpretation will vary from place to place. "COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death" -- that's pretty broad. [4]
Sorry, Dan, this is looking like it's not the Ebola infection you're making it out to be.
You write this long post, but then finish with the CDC guidance which agree with what I said -- that doctors have to use be able to say to the best of their knowledge and experience that the deceased had covid-19 and that it caused death, or that it contributed to death.
The problem here is that you don't know what "COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death" means.
Illegal, no, but a massive overreaction. This is why the Swedish model (of telling people to take reasonable steps like not getting too close, and staying home when sick, and away from old people) is proving so successful. They've even started to see a flattening of the curve alongside the rest of the world without lockdowns. [1, 2]
Let's not lose sight of the reason why lockdowns were advised in the first place.
Absent the data from widespread testing, the spread of the disease was consistent with two epidemiological models -- low-contagion/high-mortality and high-contagion/low-mortality. In the latter model, there's not much to be done to stop the disease from infecting the population at large, and not that much that needs to be done, since the vast majority of people will emerge unscathed. But in the former model, it matters tremendously -- doing nothing will slowly but surely result in vast loss of life but quarantines can stop the disease dead in its tracks.
Therefore the most cautious course of action, from a public health perspective, was to assume the former and proceed with a regime of extreme social distancing. Unfortunately doing so had dire economic consequences.
What if the swedes are trying to find out how to live with this virus? Like..what if there were no cure and no immunity? Unlike other governments who are betting that there will be a resolution in the near future?
What explains the difference between NY and CA if it isn't lockdowns? The difference in death per capita between the two is staggering (nearly 30x more people per capita have died in NY than CA).
The stay-at-home order can't itself have made a 30x difference, because New York only imposed it 3 days later than California. California started social distancing in general earlier and better, but I don't think anyone would dispute that social distancing in general matters. (In fact I think that's the obvious explanation, that due to the nature of NYC many people can't social distance to a useful degree.)
To be fair, the US engaged in lockdowns well after the situation had run its course in China so we had the data. It feels much more like a panic reaction than a question of which epidemiological model it best fit.
No way. China did an even more intense lockdown. If anything, the data from China indicated we should have done more, sooner. And that’s assuming you can even trust the data from China. Which you can’t.
The answer to "what is the correct action to take in the face of uncertainty" is not "whatever people say we should've done after we gain full information." That's nonsense, it assumes that next time you should just know the future.
Something that you might consider looking at when comparing countries or areas is the ratio of active cases to recovered. A higher ratio suggests that the epidemic is in an earlier stage.
Sweden has about 23 active cases for each one recovered. The US is about 10:1, Europe overall is about 2:1, so is Italy, and Spain is close to 1:1. Asia as a whole is about 1:1.
That likely has to do with availability of tests, who's getting tested, where and when. With Sweden not attempting to stop the spread, I doubt they've done wide-spread population testing. They're likely limiting it to hospitals, so the fact it's flat its really what matters until they kick off their wide-spread serological study in the works.
[edit] I misread, disregard, but preserving for posterity.
There is random sampling being done. 2,5% had an ongoing infection in Stockholm in late Mars.[1]
Now, 11/100 blood givers in Stockholm who has not been sick in the last two weeks have antibodies according to researchers.
That gives you a naive mortality rate of 0,4% per infection, assuming the spread among bloodgivers is the same as the population in general and just dividing the death count by 11% of the population. It is a high eastimate as I guess many wont get traceable antobodies and mostly asymptotic people give blood.
Active vs. recovered is a metric related to the people who are tested and confirmed. It cannot be changed by testing fewer people, only by rapidly ramping the number of tests up or down.
Part of the reason why lockdowns were resorted to in the first place is that Americans are, for various reasons, not likely to listen to official guidance about staying apart. See: all those panic articles about various people who knew they were at high risk and continued to go about their daily lives, or take flights or other forms of transporting masses of people, etc.
In fact, the general take is that Americans have been pretty good at following distancing rules with or without hard rules but some were slow to do so (and most government was in retrospect slow to strongly recommend doing so) but, to a first approximation, few will do so over an extended period.
This isn't even close to true. The Houston Livestock Show and Rodeo. Spring Break. Florida beaches. Evangelical churches.
Just because a majority of people do sensible things doesn't matter. The actions of the irresponsible and sociopathic minority still create the epidemic.
Quarantining the sick is legal but otherwise the freedom of assembly clause in the 1st amendment does not say "except when there is a pandemic or other fear". If you don't believe that should be the law fine, but since it is law and is incorporated to the states via the 14th amendment it qualifies as illegal.
Freedom of assembly, like the other 1st amendment freedoms, can be regulated by government so long, per prior SCOTUS decisions. They key is that restrictions must be content-neutral and can only pertain to "time, place, and manner," additionally surviving strict scrutiny (which, in plain English, means "there's no other way to do it"). This is bread-and-butter First Amendment court case stuff; any US government class that's at least halfway competent ought to have covered this.
The coronavirus bans on public gathering are fairly clearly permissible under SCOTUS precedent.
I would argue the second amendment doesn't explicitly say you can't own a nuclear warhead for personal amusement, but I would argue nobody's going to challenge the validity of such impositions.
"Arms" in the context used by the second amendment is AFAIK generally understood to mean "small arms" and by implication would exclude such weapons by matter of definition.
> "Arms" in the context used by the second amendment is AFAIK generally understood to mean "small arms"
No, it isn't. In the time period in which the Second Amendment was passed, privately owned warships of similar capability to the ones in national navies were common enough to play a pivotal role in two wars (the American Revolution and the War of 1812).
I'm aware of privately owned warships by privateers. It is true and a good counterargument.
However, the colloquial use at the time of the phrase "arms" almost certainly referred to small arms, but would have included knives and swords (State v. Kessler 1980[1]) which are paradoxically much more heavily regulated than firearms in most states. See also "To Bear Arms"[2], by which the modifier "to bear" suggests arms that could be carried.
Perhaps one counter argument to this point was the desire by the framers to include a conscientious objector clause obviating military service for people who had specific objections, which may support the view that "arms" was anything in use by the military of the day, but the clause was stricken from the constitution for fear that the second amendment may be interpreted by later generations as a right that could not be enjoyed by the people and was intended to encompass military service.
There is also the argument that the framers opposed the idea of a standing army, which would support your viewpoint, and there is an excellent essay[3] that makes a fair and reasonable argument to this end.
My personal opinion is that this isn't clear, but it is obvious that--at a minimum--the interpretation of "to bear arms" would suggest that the 1986 ban is unconstitutional, as is the 1934 NFA tax act that required registration. I'm not convinced "to bear arms" includes other weapons that could not be "beared," such as warships, but perhaps this will clarify my thought process for you.
> the colloquial use at the time of the phrase "arms" almost certainly referred to small arms
Perhaps it did, but whether it did or not, I don't think its use in the Constitution can be reasonably described as "colloquial". Since no specific restrictions were set out by the framers, I think they intended a broad right, not one restricted to small arms. Of course my opinion is an extreme outlier given current jurisprudence on this topic, to say the least. :-)
> it is obvious that--at a minimum--the interpretation of "to bear arms" would suggest that the 1986 ban is unconstitutional, as is the 1934 NFA tax act that required registration.
I agree.
> I'm not convinced "to bear arms" includes other weapons that could not be "beared," such as warships
Currently that question is moot in a practical sense, since nobody appears to be in the market for privately owned warships. And I will cheerfully admit that if we could get to a point where it was generally accepted that people had a right to bear small arms, and the arguments were over other categories, I think that in itself would be huge progress from where we are today. If that happened, I wouldn't spend a lot of time complaining that everyone recognized my right to keep and bear an AR-15 but I was being given grief about wanting to buy an aircraft carrier or a nuclear submarine. :-)
> I don't think its use in the Constitution can be reasonably described as "colloquial"
"Arms" was used in colloquial speech and legal use at the time which converged on approximately the same definition. It's late, but there are some resources on guncite that support this argument. They're all worth reading beyond this reason, mind you, and there are plenty of arguments there that disagree with mine.
Either way, it makes for good reading should you find a topic there you're not already familiar with, and even if you are, there are some interesting twists. I don't necessarily agree with all of them.
> Of course my opinion is an extreme outlier given current jurisprudence on this topic, to say the least.
Suffice it to say that I understand where you're coming from, but in my definition a broad application of "military small arms" is probably more accurate according to my understanding of what was intended. The problem is that I don't know, nor will we ever know for certain, since the militia (rather, the people) were to act as the nascent US' standing army. I rather wish this point were discussed in civics classes, but it's not. So, the younger generations are woefully unaware of history, much to no one's surprise...
Anyway, while I also wish we could undo some of the impingement on our rights by legal challenges made in the 20th century, I'm afraid that would be an uphill battle (as you also allude) given how panicked the general public is on loosening firearms restrictions thanks in large part to the unnecessarily excessive coverage of mass shootings that seem focused primarily on stoking fear.
But, if you allowed me, I'd probably rant all night about this subject which wouldn't do either of us much good by the sounds of it, other than reaching continued agreement and probably upsetting other readers.
> since nobody appears to be in the market for privately owned warships
...admittedly a shame!
> I think that in itself would be huge progress from where we are today.
Agreed. Somewhat surprised to have this conversation on HN of all places, TBH.
> ...that everyone recognized my right to keep and bear an AR-15
I live in an open carry state and sometimes exercise that right, particularly since I live in an area sometimes subject to wild animals. But while I would love to do the same with an AR, I suspect that would probably get me called in to the local sheriff's office. Not that they would care, but I would be disappointed if someone thought it apropos to waste their time.
It almost certainly shouldn't be covered. It states "being necessary to the security of a free State". What part of a nuclear warhead wielded by a militia contributes to the security of a free State? Nuclear weapons are rather unique in that they're essentially useless in a civil war. What are you going to do? Nuke the very land that you're fighting over?
Biological and nuclear weapons should not fall in the scope of the 2nd amendment. They're useless for guaranteeing your liberties against a tyrannical government, and the consequences for storing them improperly are severe.
> What part of a nuclear warhead wielded by a militia contributes to the security of a free State?
Depends on who's attacking and what might deter them.
> They're useless for guaranteeing your liberties against a tyrannical government
Which is only part of "the security of a free State", not all of it. The militia was also for repelling foreign invasions.
Given our current modern world, I would agree that weapons of mass destruction (nuclear, biological, and chemical) shouldn't be available to anyone who wants them; but the correct legal way to make that happen in the U.S. would be to amend the Constitution to explicitly add that exception to the Second Amendment. As the Second Amendment is currently written, it does not admit of any exceptions. As I noted in another comment upthread, privately owned warships were significant at the time the Second Amendment was passed, and the framers of the Amendment did not exempt them from the category of "arms".
The fact that legal realism--which says that the law is whatever judges say it is, even when what the judges say is patently ridiculous when compared to the actual words of the Constitution or statute--is the mainstream viewpoint in today's legal environment does not make it right.
If you read the federalist papers you'll note that the intention was literal, not to be reinterpreted for the times. That also makes the thousands of gun laws illegal. If you start reinterpreting the constitution that is how you go full banana republic, and everyone knows you never go full banana republic. You are supposed to change the law not reinterpret.
It's pretty funny to watch people try to justify "literal" readings of the text, since they often focus on interpreting just one piece of text and ignore the effects of the analysis on the rest of the text.
The First Amendment veryclearly binds only Congress in its literal reading: it begins "Congress shall make no law..." And there is no literal text in any subsequent amendment that might cause incorporation to the states--the Fourteenth Amendment only literally incorporates the Fifth Amendment, and that by literal repetition of the text.
> Underlying illnesses include Diabetes, Lung Disease, Cancer, Immunodeficiency, Heart Disease, Hypertension, Asthma, Kidney Disease, and GI/Liver Disease
That's a whole lot of people left out of your number. A quick google suggests that 1/3 of Americans have hypertension and 9% have diabetes. It seems like probably the majority of the US has at least one of these underlying illnesses.
Probably the same number of young and healthy people that die of COVID. For instance COVIDs fatality rate is lower than that of the flu for all demographics of people under the age of 35 per CDC data. It's about the same for 35-44 and really only diverges over 45. Worldwide the fatality rate under 20 rounds to zero. To date, pneumonia far outstrips COVID. [1]
This is completely inaccurate, and nowhere in the linked document is there supporting evidence to your claim that the flu has a higher fatality rate for under 35, nor the same for 35-44.
And this report relies on figures that diverge dramatically from the COVID death tolls. It lists the current total report as around 15k, when we're close to 3x.
> This is completely inaccurate, and nowhere in the linked document is there supporting evidence to your claim that the flu has a higher fatality rate for under 35, nor the same for 35-44.
Here's some more data. What I'm saying (that it's approximately the same as the flu for the young, with some early evidence pointing to it being better) isn't particularly controversial and well supported by evidence [1] (also older, April 14th). CDC as of late March "... no ICU admissions or deaths were reported among persons aged ≤19 years. Similar to reports from other countries, this finding suggests that the risk for serious disease and death from COVID-19 is higher in older age groups." [2] In Italy nobody under 30 died, and only 5 people between 30 and 39, as of April 19th -- yes in spite of a CFR of what, 13.22%? [3, 4]
More data as of April 20th from Oxford [4]. Scroll through all the data the world has to offer, I'm not wrong on this. I repeat, the facts you're arguing with me over, without citing data to support your case are not controversial. Surprising maybe, but not controvercial.
Especially when you consider (an albeit bad flu): the overall case fatality rate as of 16 July 2009 (10 weeks after the first international alert) with pandemic H1N1 influenza varied from 0.1% to 5.1% depending on the country. [4] The flu can absolutely be very bad is the take away there. And yeah this is worse than the flu, by around 3X on both ends of that range. Not bad, not great.
> It lists the current total report as around 15k, when we're close to 3x.
They actually address that critique, that it takes a few weeks for case data to be finalized and reported upwards, so the CDC data can be up to a few weeks behind. It's not divergent, it's delayed. This is a live list updated and maintained by the CDC. Make of that what you will.
> What I'm saying (that it's approximately the same as the flu for the young, with some early evidence pointing to it being better) isn't particularly controversial and well supported by evidence
There are 400 covid-19 deaths reported so far in the 18-44 age group in NYC. There are 3.5 mn people in that age group in NYC. That's over 11 per 100'000.
The mortality per 100'000 for the 18-49 group in the US estimated by the CDC was in the last nine flu seasons: 1.8, 2, 1, 1.2, 0.7, 2.5, 1.5, 0.5, 3.9.
It's a stretch to say they are approximately the same and definitely 11 is not better than 0.5-3.9.
I would call 4 vs 11 "approximately the same" for all intents and purposes when the denominator is so huge especially when COVID deaths are much more generously assigned than flu deaths, per my sources, especially [4].
Not to mention, flu deaths are attenuated by flu shots, and pre-existing immunity. It's totally plausible that there are many more COVID cases than flu cases in that age group -- and of course those flu deaths will happen year after year while COVID is a very stable virus, and if you get it once, you probably won't get it again.
Certainly not enough data to conclude it's way out of line with the flu for this age group, in this season let alone if you factor in a few seasons end on end.
Lastly, with H1N1, the numbers are quite different, too.
The COVID-19 infection fatality rate in the 18-45 group is at least 0.11% in NYC. And that's assuming that everyone single person has been infected, that no-one else is going to die from now on and that the estimates are not going to be revised upwards because of under-reporting.
The infection fatality rate for seasonal flu [edit: in the slightly older 18-49 group] is around 0.2% considering symptomatic cases, and probably there are as many asymptomatic cases which give 0.1%.
So yes, it's not impossible for the infection fatality rate to be similar. With some strong assumptions including that it's five times as contagious. So yes, it's not impossible for the acumulated lethality over five years to be similar. And if you extend the period the common flu will be much more dangerous, specially as this young people became older.
My data was older, and that difference is utterly irrelevant with a denominator so large, and there are many potential explanations. Especially since Italy, per my sources [4] in the parent post, assigns anyone who died while in possession of COVID as a COVID death. In fact they later announced up to 88% of their COVID deaths likely weren't actually COVID deaths but deaths of someone who happened to have COVID. So if we lop off 88%, well, it's hardly out of line.
88% of CFR potentially not being COVID was from [1] referencing [2]. Specifically:
"The way in which we code deaths in our country is very generous in the sense that all the people who die in hospitals with the coronavirus are deemed to be dying of the coronavirus.
"On re-evaluation by the National Institute of Health, only 12 per cent of death certificates have shown a direct causality from coronavirus, while 88 per cent of patients who have died have at least one pre-morbidity – many had two or three," he says.
"Other experts have also expressed scepticism about the available data."
"Report from the Italian National Institute of Health: analysed 355 fatalities and found only three patients (0.8%) had no prior medical conditions. See Table 1 in the paper; (99% who died had one pre-existing health condition): 49% had three or more health conditions; 26% had two other ‘pathologies’, and 25% had one." [2]
(For what it's worth in my reply I called it out as "[4] from the parent post" which is [1] here -- sorry for the confusion -- all I meant was that the Italian data skews very high, both because it's the oldest region [Lombardy] in the oldest country in Europe [Italy] -- and because they were very generous in how they ascribed cause of death).
That's from one month ago and it doesn't mean that those people would be dead equally if they had not been infected.
Actually the official number is grossly under-reporting COVID-19 deaths:
"We estimate that the number of COVID-19 deaths in Italy is 52,000 ± 2000 as of April 18 2020, more than a factor of 2 higher than the official number."
In the weeks from March 1 to April 4 there were 19824 people death in a subset of municipalities in Lombardia where data is already available, while in the last five years the number of deaths in the comparable period was in the 6767 - 7248 range. This subset normally accounts for 73% of the deaths in the region so we can estimate that there were 17000 excess deaths (27200 vs 9200-9900).
Less than 9000 COVID-19 deaths were reported in Lombardia by that time. If you think they are too generous classifying deaths as being caused by the infection, what would you say that caused the death of more than twice the usual number of people during the period?
About 1500 people in the US died yesterday of covid; year-to-date, about 1300 people died each day of the flu.
So it seems like there should be no possible way to frame it as less severe than the flu, even only comparing to when the flu is at its worst and even assuming 100% of the population has covid or whatever other extreme things you want.
This can’t possibly be true about the flu. There’s been over 100 days this year and nowhere near 130000 people in the US have died of the flu in any year in recent memory
Note that for the flu, most of the community does literally nothing. I mean, some people get vaccinated, but most literally ignore the flu as if it didn't exist - except if they get sick of course, and even then some people go to work if they can (they shouldn't, really, but they do).
If we compare this with complete shutdown of all public activity on the other end of the spectrum, we can see there are things we could be doing that are on neither ends. Maybe distancing but not shutdown, or avoiding some mass gathering but still being ok with individual meetings, or having places like restaurants operating at reduced capacity to ensure people aren't too close to each other, or ask people to wear masks when they enter stores, etc. And strict lockdown for places like nursing homes, but more relaxed for places like college, or having people of high-risk groups to work from home, but people with lower risk be allowed to work in the office. I'm not saying it's the right way to do right now, but it's a possibility, there are options.
So if this is e.g. 50 times worse than the flu, then trying any of the above may be too risky. But if it's kinda sorta like the flu, maybe somewhat worse but not 50x worse - then it may be prudent to consider measured response, given that for the flu we basically have no response at all and we're ok with that.
So, given that I was quoting global figures for the flu, on an apples-to-apples basis it seems like covid is about 10x the flu at the moment, in the US, ignoring the fact that it could be much worse without the lockdowns.
None of us know for sure, but isn't it plausible on the face of it that something 10x worse that could easily become 100x or 1000x worse warrants the lockdowns? I don't know if this is common sense, but in an alternate universe it could be.
> So, given that I was quoting global figures for the flu, on an apples-to-apples basis it seems like covid is about 10x the flu at the moment, in the US, ignoring the fact that it could be much worse without the lockdowns.
At about 10 weeks into the H1N1 influenza pandemic, we had a CFR of between 0.1 and 5.1% depending on the country as compared to 0.07 to 14.93% for COVID. H1N1 ended up with an actual IFR of 0.02% or one fifth of the low end of the CFR range.
Honestly, it's probably a maximum of three times as bad as H1N1, especially when you consider Italy later admitted of its 13.22% CFR, 88% of that number was "people who happened to die while COVID positive" instead of "likely died of COVID". [1] Huge deltas in CFR are likely attributable to mess-ups along the way and how COVID deaths are counted.
10X as bad is likely on the very highest end as Sweden hasn't locked down anyone, suggesting common sense instead, and their new case rate flattened out right alongside the rest of the world. 1000X is totally out of proportion with the data. [2]
I suppose we don't know for sure but we do have two and a half million data points which is enough to make a pretty good guess.
It's kinda starting us in the face that it's kinda like a bad flu except worse for old people, who we should lock down and look after as we build herd immunity.
That's not entirely true. The flu has a vaccine (of varying efficacy -- 10-60% depending on the year). The flu also has a large body of immunity built up across the population from having had it. It's also less contagious.
If we assume that COVID is spreading like wildfire (it is), and that we're front-loading the cases (we probably are), and that immunity will be built up as one-off instead of like having the disease mutate substantially every year like the flu, over a few years, it could easily be better.
I'm not saying it is or isn't, just that there is a path.
I think NYC almost certainly has much large prevalence than LA given hospitalization data alone.
One factor could be the nature of transmission. Regions where transmission occurs with high viral loads entering at the point of infection (for example, hospital-acquired infections) could see more severe disease progression. You would therefore expect a high hospitalization rate relative to prevalence. It's possible in other regions, including CA, hospitalization rate as a function of prevalence is lower because acquisition is largely at lower viral loads resulting in less rapid and severe disease progression. This is, of course, a hypothesis and very likely not correct. Time will tell.
Other important factors include age distribution of infection, as well as weather conditions and prevalence of pre-existing conditions.
I also suspect and absolutely think a large portion of the reason there is a decrease in hospitalizations is due to mitigation efforts decreasing the initial viral load. That is absolutely borne out of all evidence we know thus far for this virus and is also in line for what we know to be true about other viruses.
Yes agree NYC will have much higher prevalence but it has 17x the deaths - if fatality rate is similar then it would be like 50%+ of New Yorkers have contracted Covid. Just a little hard to believe.
Actually, no they don't. There are 5.6m rides per day. The average trip contains at least two rides: there, and back again. Many trips contain substantially more than two legs, and many people take more than one trip per day.
I can't tell you how many people ride the subway every day, but it's certainly less than half of the number of rides per day.
Number of rides is actually more relevant in this context than number of distinct riders. Each trip involves a new set of people next to you in the train car and a new chance of getting infected.
The number of rides is useless in estimating the percentage of the population with high risk for exposure if you don't also have the proportion of the population exposed to that risk.
There’s a decreasing number of positive test cases each day in NYC too though. At highest point was 50% and now under 30% or so. I’m not sure if those tested get the antibody test as well but it implies a larger percentage of symptomatic people are testing negative in nyc.
One way to explain the hospital surge in NYC is that a small percentage of people have had a highly fatal disease. Another plausible explanation is that a much larger number of people have had a less-fatal disease.
To date, we've had very poor estimates for fatality rates and prevalence. It wouldn't be shocking at all to find out that we're off by a factor of 10-, 20-, or even 50-fold.
Typically a sufficiently infectious disease tops out around 50-80% (don't have a source, just what people, like Angela Merkel have been saying). So this implies NYC is already off the high end of that. Hard to believe.
That's basically the
top end of the estimate of what is theoretically possible without a lockdown and infinite time. I just don't buy that it is already maxed out like that.
It's the top end estimate from the study. Even if you take the low-end estimate (~30x), there would be 4.2M cases, which is still ~50% of the New York population.
Basically, you're just saying that you don't want to believe what they're saying, therefore it's hard to believe. But there's nothing illogical or impossible about it in the slightest.
You keep making assumption. You have to concluded the lock-down did nothing, has done nothing, and hasn't prevented any infections. You can make assumption because the numbers don't align to what you assumed they'd be.
https://www.worldometers.info/coronavirus/country/us/ right now shows 12850 cases per 1M in NY, which is 1.2%. Which is pretty close to what the study shows for LA. So it's not actually suggesting NY has much more than it's already known, if we assumed overall percentage would be similar for both. It does suggest California - which currently lists 853/mln - undercounts it by more than 10x. Of course, provided that the study is true.
Also, having antibodies and being hospitalized is a very different thing. If a lot of the cases are asymptomatic or very light you won't see hospitalizations and you'd never know the person had an encounter with the particular virus until you do mass testing. And NY is quite different from LA, as far as I know, in terms of how dense population can be and how much people use mass-transportation and other high-risk scenarios, and the probability of being hospitalized may also depend on that (in other words, it may matter not only whether you had any contact with the virus, but in which circumstances and how much of the contact you had).
That doesn't prove the study is true of course, but I don't see anything from comparison with NY that makes it implausible on its face.
Yes. The consensus I've seen forming is that at least 15% of NYC has had it by now - a variety of questionably random samples have suggested numbers near that.
NY has 17x more deaths than California. That would mean if 3% of CA was infected (yes this is LA only but for ease of comparison) it would be 51%+ of NY was infected. Far more than just 15%.
LA is one of the densest parts of CA though, outside of SF/Bay Area. So if LA is 3%, you could assume the state as a whole is maybe 1-2%. Then 17x that becomes 17-34%, which seems more reasonable.
The comparison is overweighting LA county, which has 25% of California's population but 60% of its deaths. (I don't know what the math ends up being off the top of my head.)
CFR is not a great way of comparing across regions, it could be due to any number of factors. Demographics (i.e. age), co-moribidities and the amount of hospital space available.
In fact the per-country spread of CFR is huge, from a low of 0.07% in Singapore to a high of 14.93% in Belgium, a whole 3 orders of magnitude. The H1N1 influenza CFR at a similar point in the outbreak was between 0.1% and 5.1% -- but the final IFR ended up at 0.02% [1].
15.6% of pregnant woman at a given hospital were PCR confirmed over 2 weeks. I haven't seen strong arguments this wouldn't be broadly characteristic of the overall population. Worse (or better depending on how you look at it), there's significant false negatives with PCR testing.
It's credible NYC could be at 30% by April 4 and given that confirmed cases doubled since then, could be at 50%+ by now.
How off do you think this can be? Women go once a week to a medical facility (and that's to an ob-gen, not the emergency room) before birth, but also don't go out in general. You could just as well make the case this is lower than the general population.
It's for sure true that a huge number of people in NYC got it, way surpassing the confirmed cases count. The state is currently doing a serology study here and results should be published next week.
Sure I believe NYC has way higher prevalence of cases but difference in deaths would imply majority of population was infected and they’d essentially be at herd immunity now. It’s not implausible but find it hard to believe.
Case fatality rate is # of deaths / # of positive tests. It’s enormously biased by the availability of testing, and the allocation of the tests we have to the sickest people. It’s not especially likely to generalize to the population at large. We need large scale random sampling of the population for that.
Everyone is eager for these results, because there’s a good chance that the true case count is higher by anywhere from 2x to 100x. If it is, then proportionally fewer people are expected to die as the virus eventually does churn through the whole population.
It’s probably still too early to tell, but this is promising.
I think the obvious conclusion (just in logical progression if you take this premise as true) is that the percentage of NYC residents who contracted it was much higher than previously assumed.
Yes but the difference in fatalities is so large that it would imply majority of New Yorkers were infected and the city has achieved herd immunity. It’s not implausible just find it hard to believe.
It's hard to believe, I would think, because it would be such a remarkably positive development.
But just from a logical perspective it follows clearly from this data, and there's nothing fundamental about the situation that makes it particularly unlikely.
Of course it's quite possible that the LA data isn't really applicable for any number of reasons. But the idea that COVID will run through dense populations mercilessly, but quickly, and then sort of peter out is completely within the range of possible and non-shocking outcomes.
It's the natural outcome of an extremely communicable disease with a large prevalence of asymptomatic transmission and there's plenty of corroborating evidence for the idea that this is one of those.
It's also quite likely that the false-positive rate with this test is super high. Who knows.
If 50% of NYC has had the virus, they won't have herd immunity yet -- of these studies are correct, then the virus is horrendously infectious -- making 90%+ of population required for immunity.
Just speculating here but I hung out with some NYC folks last month in SoCal and they practically all smoked. It was very strange, I rarely encounter smokers living in California.
Cigarette smoking rate is widely studied; no reason to use anecdotal data here.
The tobacco smoking rate in New York is 14.1%; it's 11.3% in California. Some difference, yes, but far less difference than you see between either and some rural southern states where the numbers can be roughly double that.
No the very obvious conclusion is that covid-19 isn't very deadly. Thats what every test has concluded. Its that the death rate is closer to .3% than the 3% that the WHO quoted and much closer to the flu's death rate of .1%
Stanford reports doing a 30-person validation test with pre-COVID-19 hip surgery patients (100% negative). But this is not enough to validate these population-scale tests, especially when 98.5% specificity (rather than 99.5%) would fully invalidate the headline result.
> Stanford reports doing a 30-person validation test with pre-COVID-19 hip surgery patients (100% negative)
Hmm - I guess I should now go to the source - but from the other link: " test validation included a total of 30+371 pre-covid blood tests, and only 399 of them came back negative." https://statmodeling.stat.columbia.edu/2020/04/19/fatal-flaw... - I think I trust it more
This one's from a different company (BioMedomics) and it was a random test of people pulled straight off the streets. Happened at an entirely different geographical area (Chelsea, Massachusetts) as well. I don't want to be too optimistic but there are some signs that we are heavily under-counting the actual number of cases (at least in the US).
> Another interesting aspect of the result was 6% of men showed antibodies but only 2% of women tested.
That is interesting. So far, men are more likely to die of covid-19 (almost a 2:1 ratio) than women; if this is a repeatable result it would seem to suggest that the reason is that fewer women are becoming infected, not that women are somehow less likely to exhibit extreme symptoms when infected.
Or that women are less likely to need antibodies to clear this infection. Not everyone develops antibodies for every disease, the immune system is super complicated.
The Imperial College model was also published by press release. That will have ended up doing a whole lot more damage than any of these serosurveys, when all is said and done.
If I'm being honest, whether it's logical or not I wouldn't be surprised if this is politics and misinformation starting to creep it's way into science. In current events there's a struggle to open up the economy with Trump saying we're ready and the that governors are at fault for keeping everything closed while the are governors saying it's way too early and they need to keep everything closed. This may be an attempt to make the public think the situation is safer than it is (if more people have been sick than thought then we could potentially have herd immunity or be closer to it) which the public can then put pressure on the governors.
Who knows, but I will say it's getting harder and harder to say some thing are too tin foil hat lately.
Our initial policy was based on early data from China which seemed to imply 3% mortality, an R0 of 2.4 and a 5% hospitalization rate. In the past month, data from all sources and methods is showing this to be far less deadly but also far more contagious. Perhaps it’s time to adjust our view of the lockdown.
Uncontrolled? No, but we do know who is most vulnerable and we should be able to make more targeted and surgical measures to limit the negative impacts of social distancing.
Here is a basic question: how are the sensitivity and specificity (false negative and false positive rates, IIRC) of these tests determined?
It seems like a learning problem with noisy labels. Does anyone know what they do in practice? Especially chaotic situations like this where there aren't reference tests? Are there reference positive and negative samples?
One concern also referenced for the Stanford study was that, as mentioned in a USC popular writeup on this study linked elsewhere on this post,
> Premier Biotech, the manufacturer of the test that USC and L.A. County are using, tested blood from COVID-19-positive patients with a 90 to 95% accuracy rate.
They must be factoring this into the 2-5% infection rate number somewhere?
In the Stanford study, they used blood samples that were taken before the epidemic to determine the true/false negative rate, and blood from patients that tested PCR positive in the past to determine true/false positive rate.
Reference negative samples are relatively easy, at least in theory. Any blood samples drawn before the Covid-19 outbreak started are guaranteed to be 100% negative.
The sensitivity and specificity of these new tests is not established. It's part of why the FDA hasn't approved the tests yet. Scientists are doing their best with tests with unknown reliability, but it's a statistical mess right now.
Other more squishy factors are important, too: I don’t think Facebook posts for recruitment are a good idea. This could lead to more people taking part who have been sick or who had contact with someone who had the virus, over-representing that group of people in the sample.
This LA study has one of the same authors as the Stanford study (Neeraj Sood). It uses the same test and so it has the same issues with the false positive rate. An honest person would address those but I didn't see that in the press release.
Both the LA and Stanford studies are junk. Easily explained by false positives inherent to the test. Stuff like this is why pre-print services can have disastrous consequences.
That article is junk. Even the fact that it quotes Santa Claras deaths at 100 deaths is wrong because it was only 70 at the time. This entire blog is rife with mathematical and statistical errors.
So, in your ideal world, we would never do serosurveys for surveillance of low incidence diseases, right? Any Type 1 error is too much Type 1 error, right?
I'm sorry that literally all of the seroprevalence data hurt your priors. But the virus is widespread, not very deadly, and people are going to go back to work soon. You can stay home. You will be safe at home.
The Stanford study in Santa Clara that used Premier Biotech appears to be very misleading. Given the confidence intervals on the specificity all 50 positive cases out of the 3330 tested could just be false positives.
Especially considering the test used here was also, Premier Biotech's. Their manual [1] states, that other coronavirus strains can lead to false positives in the test:
>Positive results may be due to past or present infection with non-SARS-CoV-2 coronavirus strains, such as coronavirus HKU1, NL63, OC43, or 229E
They appear to be using an antibody test from Premier Biotech. However, I can't gain clarity on whether we know if these tests are even accurate or reliable.
"But some COVID-19 antibody tests, including those being used by public health departments in Denver and Los Angeles and provided to urgent care centers in Maryland and North Carolina, were supplied by Chinese manufacturers that are not approved by China's Center for Medical Device Evaluation, a unit of the National Medical Product Administration, or NMPA, the country's equivalent of the U.S. Food and Drug Administration, NBC News has found.
Two U.S. companies — Premier Biotech of Minneapolis and Aytu Bioscience of Colorado — have been distributing the tests from unapproved Chinese manufacturers, according to health officials, FDA filings and a spokesman for one of the Chinese manufacturers. Many of the unapproved tests appear to have been shipped to the U.S. after the FDA relaxed its guidelines for tests in mid-March and before the Chinese government banned their export just over two weeks later.
If COVID-19 antibody tests are unreliable, they can produce false results, either negative or positive, health officials said. The use of such tests has been widely discussed as a way to ensure that employees are healthy enough to go back to work and to find COVID-19 survivors who may be able to provide blood plasma to severely ill patients.
Officials at the Association of Public Health Laboratories have expressed concern about the reliability of the numerous antibody tests being sold or used across the country with little scrutiny. "
Does anyone have a link to the actual study? LA Times link doesn't include it. Other than the selection criteria for participants, it doesn't sound from the pressroom.usc.edu link that it's any better than the Santa Clara study, and perhaps worse, given the smaller number of participants. The pressroom link mentions an "accuracy" number, but not specificity and sensitivity.
That really tells you everything you need to know about this study already, doesn't it? If they put out the press release before the actual study you know what the authors' priorities are.
Just a press release, no pre-print. That being said it was being co-run by LADPH it seems. Many people involved were also involved with the Stanford study which has been very thoroughly criticized for its assumptions.
NYC has 20x the deaths / 100k people, so under the assumption that number of deaths are on comparable trajectories (due to both locations being under lockdown, maybe not as unrealistic as a few weeks ago), NYC's prevalence could be 56%-112%. This could be one explanation for why numbers are going down in NYC, or the 112% being larger than 100% could suggest that either the situations aren't comparable or that the LA study's numbers are a bit high.
Is it also possible different strains are dominant in California and NYC with different mortality characteristics? There is also the issue of weather. A recently leaked DHS report indicates sunlight radically reduces virus half-life and it's also affected by temperature and humidity. Maybe subway + weather + (possibly) strain of differing virulence can explain the discrepancy?
NYC also has the lowest car ownership in the nation, so you're basically left with walking or uber/taxi/public transportation just to get to the store/doctor/etc.
Because of demographics. The disease really, really doesn't affect young people. The younger, the better you do, to the extent that the percent of people under 20 that have died in every major area of the world rounds to 0%. Older demographics are hit hundreds of times harder than young to middle aged folks.
Italy has the oldest average population in Europe, and Lombardy has the oldest average population in Italy. 99.2% of those who died in Italy averaged 80.5 years old and had an average of 3 co-morbid conditions. [1]
Italy also appears to have suffered from tragic public health errors. Apparently they moved covid-19 infected patients around. First to other hospitals that had not yet had outbreaks and then into empty beds at long-term care facilities.
"spread" isn't "affected" -- someone with no symptoms who spreads it really isn't affected. The people who get it from them may well be.
So far the data is clear: it doesn't really do much to young people. [1] Speaking broadly, almost nobody under 20 has died, practically nobody under 30 has died, and a handful of 30-40 year olds have died. The trendline is super clear: if you're young, you're gonna be just fine. The CDC says young folks aren't even likely to wind up with serious disease or in the ICU ("no ICU admissions or deaths were reported among persons aged ≤19 years. Similar to reports from other countries, this finding suggests that the risk for serious disease and death from COVID-19 is higher in older age groups.") [2]
When I say "affected" I mean that they are not likely to develop serious symptoms, certainly severe symptoms and they're definitely not likely to develop life-threatening symptoms.
We know for a fact(well based on this report from a Chinese univesity[1])
that there are two strains of covid-19, I wonder if the less aggressive strain has been around awhile more than the more aggressive strain, it may have been around a lot longer than the outbreak was announced (think fall 2019) and may have been misidentified as the flu/bad cold. And it would explain why some people are getting extremely mild cases, they either have the less agressive strain or already had it but now have the antibodies to fight it. This is total conjecture mind you and the weather aspect is also interesting, look at Hawaii's cases and death rate(580 cases to 10 deaths, they are both very low numbers).
That study is really not very good and doesn’t tell you anything.
Sure, mutations happen all the time, however since the virus has error correction those are rare and very unlikely to have any effect on the phenotype.
Not really. Antibody tests like this one measure the total number of people who've been infected at some point in the relatively recent past, whereas the diagnostic tests New York is using only detect current active infections, so it's easy for the percentage of infections measured by the former to be higher than the latter.
When articles like this compare the number of known cases detected with the PCR diagnostic tests with the higher estimate based on antibody testing, they're comparing the sum of all cases detected since the start of PCR coronavirus testing with a point-in-time estimate of the proportion of the population that currently has antibodies for exactly this reason.
That’s true. We could have 25% active infections and 25% not currently infected but who were infected in the past. I still think the actual numbers may be be lower (but higher than the reported figures, of course).
My point was that the number of people currenlty infected is surely lower than 30%, given that only 2% of the population is already confirmed and testing a suspected case today will return a negative 70% of the time.
But as another commenter said, it's not impossible that there are many other people who had infection which is not currently active.
Using the overall test rate doesn't really help, I think.
It doesn't seem that unlikely to me (I live here). I know so many people, me included, who had it but aren't included in any official count for lack of testing, and so many others who must have had it (e.g. my cohabiting SO) who were asymptomatic.
I don’t really understand your point. The error bars would be calculated differently for a larger sample. If you had 100/100 positives with a five percent error you wouldn’t say the true amount is 95 to 105. You’ve done some sloppy statistics and claimed this invalidates the work...
NY has performed 32k tests per 1M population, CA 7k per 1M population [1]. If CA PCR-tested as much as NY, there would probably be way more official cases and your calculation would make more sense.
Death rate higher could be due to an overwhelmed health care system, so it is not linear. Once that threshold is hit, it will probably edge close to rate of severe cases among the patients.
NYC is massively overreporting deaths. Just the other day they added 3700 "new" deaths that didn't even test positive, they just suspected (ie. pulled it out of their behind) that it was cause of death.
Death numbers are overreported in general - Dr. Birx herself even claimed that if you test positive but die of something other than the virus they track it as a COVID death. It's preposterous.
Belgium is doing something similar. They're being more aggressive than any other nation in Europe at labeling deaths Covid. It has contributed to their extreme case mortality rate.
I listened to Birx[0] and perceived the same. I continue to wonder what I'm missing. If she's correct, there would be serious problems with such methods. Strange stuff.
One issue is that the study design and probably analysis is very similar to the Santa Clara study (they are using the same test kits and have shared coauthors on both studies). It's not really independent replication when both are potentially using flawed kits and analysis.
I'm not sure if we can consider this independent support: both studies use the same antibody test. Any problems with the test itself (or our understanding of problems with the tests) will be present in both studies.
Looking forward to what the science heads on my twitter feed say about it. They were positively withering in their criticism regarding the Santa Clara study.
How do you explain the low number of hospitalizations and deaths in LA vs NYC? If these numbers hold then you’d expect NYC to be close to herd immunity given the difference in fatalities.
There's a logical jump from the 2nd to 3rd paragraph that bugs me, but upon reflection, I don't know if my objection is valid:
> The initial results from the first large-scale study tracking the spread of the coronavirus in the county found that 2.8% to 5.6% of adults have antibodies to the virus in their blood, an indication of past exposure.
> That translates to roughly 221,000 to 442,000 adults who have recovered from an infection, according to the researchers conducting the study, even though the county had reported fewer than 8,000 cases at that time.
So, in the 2nd paragraph, there, a match for SARS-COV2 antibodies means an exposure. In the 3rd, it means an infection. I have always understood that exposure != infection. Yes, the former is a necessary precondition to the latter, but an exposure doesn't automatically and unconditionally become an infection.
But then it occurred to me that the body probably isn't going to be generating antibodies unless a pathogen got past the body's initial lines of defense and required a more active response from the immune system. In which case then, perhaps it would be accurate to say antibodies == infection. Are we speaking then of asymptomatic carriers or those who never experienced more than mild symptoms and might not have realized they were infected?
Am I splitting hairs here? Is this simply shorthand for "exposure that subsequently became an infection?", much in the same way I might freely use "LDAP Server", "Name Server", "Domain Controller" "Auth Server", to all refer to the same system in an on-prem Windows environment? Or, am I rightly objecting to unclear language that would lead to incorrect conclusions?
There's been a bunch of genomic analysis. Yes, the West coast strains are closer to Asian and the East coast strains are closer to European. But the differences between strains are believed to not have an effect on the disease.
How exactly? Serology has nothing to do with figuring out which version of the virus you have, ignoring the fact that there are no distinct "strains" to begin with.
89% of total fatalities in New York have at least 1 comorbidity. I wonder to what extent anyone with unknown cause of death or death due to cancer, but also had COVID-19 gets lumped into the death due to COVID-19 category.
The overall mortality rate in NYC is more than double the usual [1]. I guess that most people above a certain age have at least one comorbidity such as diabetes or high blood pressure.
One particular criticism of the Stanford study was that they recruited via Facebook adverts, and that method may be influenced in recruitment by the "desire to get a test if you had reason to believe that you, or someone near you, had the virus"
This USC study seems to have recruited differently: "Participants were recruited via a proprietary database that is representative of the county population. The database is maintained by LRW Group, a market research firm"
Via: http://www.publichealth.lacounty.gov/phcommon/public/media/m...
The flaw was that participation was voluntary, with the effects you describe. I think this is also the case with this study. I'm not sure how a non-voluntary study could be conducted, you can't really force people to participate in this.
With the NYC analysis here, everyone is missing one unknowable piece of information.
We know that the initial major outbreak in NYC was with a strain from Europe, while the initial major outbreak in California was with an Asia-origin strain.
These might be differently dangerous. If the NYC/Europe strain is just 20%-30% more fatal than the Asian one, the maths all start being believable (you start saying that 30-45% of NYC has been infected -- which, unlike 60-100+%, is quite believable.
It seems likely that some vulnerable neighborhoods in NY have effectively everyone infected, but that is probably not universally true.
It also seems quite believable that we are nearing the point where the majority of folks in NYC have been exposed. If nothing else, the subway is still seeing quite active use, and must be a breeding ground for infection -- and yet, hospitalization has started declining, meaning that new infections are slowing down -- meaning that at least among folks that still go out and about, there must just be fewer people to infect.
With subway use still being as high as it is, it seems like it must be true that, amongst the population using it, R could only fall below 1 if herd immunity (in the rider population) was beginning to form.
This also jives with the numbers we are seeing, where reported cases are down despite both increased testing and looser counting.
There's not enough data to "believe" anything, and most of the pushback is because the studies are methodologically flawed. I think that fatality rate estimates between 0.5% and 1% seem to best match some of the early data on isolated populations, but I'm far from an epidemiological expert. That said, I think people would be a lot more accepting of data contradicting those rates if it didn't have clear issues.
I feel more informed when I see an infection rate that's substantially higher than the rate of false positives from the test. So the NYC test of pregnant women at (13? 15%?) seems to be much higher than the 1% to 4% range of false positives that people seem to be estimating. But when test results are "down in the noise" I don't feel we're necessarily measuring anything.
I don’t see our squares with our Australian numbers. How have we been able to bring the case numbers right down if there are as many asymptomatic cases and as high an R0 as these studies suggest. Our overall testing of people with symptoms has an under 2% positive rate with one of the highest per capita testing rates in the world. We have a lockdown but it isn’t so extreme as to be able to stop the kind of spread these studies are suggesting.
The vast majority of our non imported cases have also been successfully contact traced, people by and large seem to be contracting it from other confirmed cases.
This recent Swedish study was done in (what seems) to be a way that's much less prone to bias than soliciting people via Facebook posts.
It was also based on tests collected from blood donor samples, so none of the specific bias of the Stanford study (which to me also seems like a huge flaw even if other sources of evidence are pointing to similar IFR numbers and undercounting)
In any case, the Swedish findings cited for Stockholm in the analysis linked to in this COVID19 subreddit post indicate a similarly low overall IFR of far below 1%.
It is worth noting that blood donors would probable tend more towards youth and general good health, which could skew its demographic profile. Also, depending on donors' motives for donating (maybe they thought it would include a COVID test) could skew results in different and unpredictable ways.
Understanding this study is very difficult because they haven't released all of the details about statistical procedures, recruitment, raw data, etc, but I would be worried that this study will have similar issues to the recently published Santa Clara study (both use the same lab test and share several authors).
I highly recommend that people read https://statmodeling.stat.columbia.edu/2020/04/19/fatal-flaw... for information about for information on how the Santa Clara study messed up. TLDR: Incorrect statistics, incorrect analysis, and incorrect sampling probably led to a vast overestimate of both prevalence and certainty in that high prevalence estimate.
In isolation I don't think this study can really shift the differing and primarily political response momentums the various states have going. Truly randomized nation-wide studies would need to be completed probably by the CDC for there to be wide-spread consensus on the results. The politics are going to move quicker than the science.
Certainly, but I don't think anyone was assuming 5% fatality rate. There were concerns about it being 1-3% fatality, but even those would be worst cases.
Worth noting that .36 is roughly 3x as deadly as the flu.
3x as deadly as the flu, in a population of people who have already been exposed to the flu (and some even taking flu vaccines constantly).
The evidence will continue to build that this virus is pretty benign compared to all of the fear mongering has been suggesting. Of course, HN downvoted me when I suggested this last time https://news.ycombinator.com/item?id=22685457
Closer to 4x and only between 10M and 40M people in the US get the flu every year. Coronavirus could be >4x more deadly and have >10x more people infected.
Lets ballpark and say it's 3x more deadly and only 3x more people get it than the flu, 40M * 3x * .0003 == 36k deaths. However, we've blown past that many deaths already, and we have nowhere near 120M infected people so far, so the math just doesn't work. It's far more than 3x more deadly than the flu, or we have far more infections than we can account for in any study.
Also keep in mind that the death statistics have largely only counted people with positive tests who died in a hospital, so it's going to be strictly under counting.
New York, Italy and Wuhan are all the evidence you need. The virus is perfectly capable of overwhelming local health care systems, at which point shut downs are the only responsible option.
In high density areas or with a substantial older population, yes.
Sweden didn't shut down and their health system isn't at the point of any of those locations. [They have a lot of people dying though - so that's not to say not shutting down is a good idea].
It's also pretty unclear from Seattle's trajectory pre-lockdown it was every going to be overwhelmed if it didn't shut down.
I don't know how you can say that when hospitals have already hit the breaking point. It's not like our health care professionals have been sitting around twiddling their thumbs.
Only a few hospitals in the very worst hotspots have hit the breaking point. I don't mean to say that we could have or should have done nothing, but the idea that there's been some widespread breakdown in medical care is just not accurate.
I'd definitely call it widespread. In many places across the world, hospitals faced bed and ICU shortages and had to go to heroic efforts to prevent them. But most of them were successful, remediating their shortages before anyone had to be denied care.
Wrong. They were more successful if they had proper supplies of masks, citizens cooperative with lockdowns, enough testing capacity, and so on. Otherwise, no, they were having triage when they normally would be able to treat everyone.
And many more hospitals were only 1-2 doubling periods away from faring as bad or worse. Doubling periods that were avoided because the surrounding regions took the actions they did.
What is the relevance of people having immunity to the flu or taking flu vaccines?
Of course there isn't any. Sure, it would matter if this were a buzzfeed listicle about "most dangerous viruses of all time". The flu might be higher in that list. But the actual question is "given this virus, which we don't have acquired immunity to, and don't have a vaccine for, how many people will it kill, and what are reasonable steps to lower that number"?
Note that the death counts we're measuring correspond to what happens when there is sufficient treatment / health care capacity. We have to assume it would be higher if hospitals were overwhelmed and patients who needed ICU beds were not given it.
You'll want to keep in mind that the death rate spikes if the medical system is overwhelmed and there are medical shortages, which has not been the case in Los Angeles. There is plenty of examples globally and in NY that this virus is capable of overwhelming metropolitan areas.
A lot of this is besides the point because what matters is hospitalization rate, and we've already reached the point where some hospital systems are stressed. Beyond that point, people die that otherwise wouldn't. So I'm not sure there's a big difference in appropriate response.
we've already reached the point where some hospital systems are stressed.
Only NYC has gotten anywhere near their capacity, and even there (as far as I know) nobody has been denied care due to insufficient resources. Many hospitals in other areas are mostly empty and even cutting shifts, e.g. https://www.insidesources.com/unexpected-consequence-of-covi...
you don't take into consideration how contagious it is, it's not only important the percentage but the absolute number and also the time (due to resources constrains)
>> There is massive difference between the appropriate response to an illness with a 5% fatality rate and an illness with a 0.36% fatality rate.
Is there? What is the appropriate response to 0.36%? The flu has a mortality rate of 0.1%, which seems comparable, except we all have some immunity from having the flu multiple times in the past. Flu shots are available for vulnerable people. So while people on average have a 1 in 1000 chance of dying if they get the flu, they also only have a 1/40 to a 1/7 chance of getting the flu at all. The probability of getting corona virus appears to be much, much higher for people exposed to it, and we have no idea what the rate of infection could get to if we let it run free and behaved normally.
Provided the people were chosen truly randomly/make up a representative sample of the population, 863 is more than enough to get an accurate estimate of the infection rate. ~1000 is a big enough sample size for the entire US, let alone just Los Angeles
I used to work on survey designs in my old job. I am having a hard time coming up with a way that they could possibly control for selection bias. We are in the middle of a pandemic, the people willing to risk leaving their homes to get tested is going to be highly skewed towards individuals that believe they have already had the infection. It's possible that they asked the participants about this, and adjusted their weights accordingly, but considering that this study shares authors with the Santa Clara study, and that study did not control for selection bias, I have my doubts.
Where's the marketing company's criteria for sample selection? How are the communicating and choosing the people? Are they all within the same locality?
And no 863 is not representative of LA. LA is made up of individual communities and some never leave those communities. Some communities are pass through for commuters. Really important where those 863 are from. All are 863 from a small area? Are some from Pasadena and other from Santa Monica. What about Covina or Redondo? LA or Beverly Hills.
Not sure what significance there is. Sure, some intermediate statistics change (e.g. what is the true percentage of bad reactions). But nothing changes the true death rate, which is what, the single largest mortality stat in America right now? Arguing over this fraction and that, isn't going to make this less deadly.
If it seems to be a highly spreadable virus, and a small percentage of the population has contracted it, then a significant reduction in the current societal restrictions will cause a spike in new infections, which leads to hospitalizations, ventilators, and deaths.
If a large percentage of the population has contracted it, and the results we have seen are due to that, then there is less risk in re-opening things.
I can't speak to this study's bearing on that, but it's one good reason to care how much of the population has already contracted the virus.
And of course the second most important thing to know I think is whether or not antibodies confer immunity. We've all been assuming herd immunity is the key, either by vaccine or the "we all get it, but slowly enough to not overrun the healthcare system" method.
Right now we're on the edge of a precipice. My read of "the people" is they will hold off a bit longer if some reasonable-sounding fact-ish based plan comes out soon. It can be a little wrong, but it has to be credible. Knowing how many people have had it, and whether or not those people are now safe, is imo the two biggest components of planning the next steps. If we can't get credible answers to those questions soon, I believe the population will just start making decisions based on hunches/beliefs/feelings/etc. Some will remain isolated, others will rebel. It won't be pretty.
No one ever said you wouldn't get this though. The whole idea from the get go was to prevent a run on the hospitals, not to prevent society from catching this.
Every passing day more data and studies are reflecting the realities of this disease not being as deadly as imagined. So perhaps, just like chickenpox, its better to just get it as a kid then to face it later as an adult.
I'm growing more sympathetic to the wild theory that this is actually no worse than the endemic coronaviruses, and the deaths we're seeing are just what happens when a bad cold hits a population that didn't build immunity to it in childhood.
---
Globally, about 3.4% of reported COVID-19 cases have died
---
That was used as a basis quite a few decisions, how to treat patients, when to close / open various locations, etc. It does say "reported" in the report, but I think often that part of ignored and many officials assumed it to be as a rate based on infected cases.
The article does mention this issue, and I wonder what the new / updated mortality rate is then based on this data.
I mean, you're tecnically correct in that the number is no longer accurate. But actually the change has been in the other direction. That link is talking about the CFR, not the IFR, and the CFR is up to 7% these days. (170k dead out of 2.5M confirmed cases).
No credible source that I saw ever suggested that the IFR was 3%, and that is not what any decisions were based on. Everyone knows that some cases will be missed. The early IFR estimates were all huge ranges like 0.1%-2%, precisely since there was so little information early on about how many cases were not being detected. (And it is surprising how little we've managed to narrow that range down).
There are other claims in that WHO link that are obviously wrong in retrospect. E.g. the claim that Covid transmits less efficiently than the flu.
Population is supposed to be representative of LA county. Tests are from Premier Biotech with some preliminary data on false negative/ false positive rates shown. I suspect assumptions related to the test itself, rather than population, may be a bigger factor for any inaccuracies. There's also the question of whether antibody response implies immunity for SARS-COV2
Another interesting aspect of the result was 6% of men showed antibodies but only 2% of women tested.
Also: --2.4% of people between the ages of 18 and 34 had antibodies to the coronavirus --5.6% of people between 35 and 54 had antibodies --4.3% of between 55 and older had antibodies
Edit: The key assumption that really, really needs to be triply checked here is the specificity of the test. They're assuming 99.5%, similar to the Stanford study I believe. If it's closer to 98% then prevalence estimate confidence intervals should include 0%, I believe. Also should be noted that the same PI on the Stanford study, Jay Bhattacharya, is also involved in this. Same criticisms likely apply here - more caution needed!
Second edit: On further reflection, I think these studies could do serious damage based on how the results are being announced by press-release - in this case, afaik there's no pre-print. In our current environment this isn't just science. Policy is being enacted and people may die because of poorly framing results. Everyone involved needs to take a step back and think through this carefully. Also - did anyone figure out why a hedge fund guy was an author on the Stanford pre-print? And why that same guy then published a Wall Street Journal op-ed about the study without identifying himself as a co-author? Shady stuff.