Population-level herd immunity against Covid-19 will not be achieved in 2020
coronavirus.jhu.edu
coronavirus.jhu.edu
- Immunity may only last a short time, in common with other Coronaviruses. Re-infection may be less severe, more severe, or not possible. We don't know.
- Infection Fatality Rate is uncertain, but could be as high as 1%.
- Long term health effects are unknown, but for example some evidence suggests permanent lung damage is possible.
Significant areas of the globe have successfully suppressed the virus, and may even be looking towards eradication. Such countries are looking to establish 'travel bubbles' or 'green zones' between them and get back to normal (within their virus free spheres).
For other areas, it may be too late for suppression, or for societal reasons suppression may never have been practicable, in which case herd immunity as an idea could be considered a palliative.
Where you get this information? I seem to recall exact opposite. Coronaviruses mutate slowly and immune responses last several years. For four seasonal coronaviruses the immunity lasts 1-2 years, longer than other viruses causing influenza.
I think this is the study they might have been referring to? And this is antibodies, not the memory cells mentioned above.
Before the hiv was understood in detail, policy responses were blunt or non-existent. Many kids including myself were taught to abstain from sex entirely, which of course is a policy which would work on paper but with which no one can comply in practice.
Here in Texas our lockdown has has little effect on the viruses rate of transmission. I believe that's because like many southern states, we could never achieve good compliance.
More realistic policies which take into account the fact that young people are at far less risk for the virus will likely win the day. We can better support and sustain quarantine if it's triaged for those at greatest risk.
Instead it Looks like policymakers will just let go of trying to control this in the US, and many older people will die unnecessarily.
What makes you say that? (Genuine question, not a challenge.) This website shows the estimated Rt going from a peak of 2.08 down to below 1, before stabilizing at 1.06:
https://www.covidactnow.org/us/tx
1.06 is still > 1 which is bad, but it's certainly a lot better than 2.
EDIT: Rt, not R0
I'm also concerned that the law currently is at variance with the reality that the parks are full here in Austin. Instead of rushing to be open, I wish the government would tailor it's message to people who need it, and come up with 'clean' food/resource pathways for people who are vulnerable to this virus rather than trying to keep kids out of the water.
Herd immunity means "it's killed as many people as it could".
If we are going to be in a situation where hospitals have ample capacity, and we figure out a way to combat the symptoms of covid-19 without loss of life, then social distancing restrictions would be lifted immediately.
Enough people are resistant so instead of becoming virus factories and spread it further they stop the infection with their immune system.
If most of the people are immune, the fire has no combustible to propagate and it stops.
The level at which this happens depends on the ease at which the virus spreads. If it spreads easily, herd immunity kicks in much later than in the reverse case.
But the degree and duration of immunity that usually follows SARS-nCoV-2 infection isn't well established yet, and coronaviruses in general are noted for not producing lasting immunity.
It depends on the costs of doing so.
You could just go for broke, let everyone catch it at once, and RIP 5-10% of your population. Even then it'll take most of a year to run its course and that year will be much, MUCH worse than just not being able to go to the pub or get a manicure.
Also the next thing that comes along (and as our human monoculture grows, that will happen more and more frequently) might be much scarier. I feel like the silver lining to our current situation is that it's forcing us to set up our society to limit contagion so that when we do run across a new pathogen which is highly contagious, has a long incubation time, and is highly deadly, maybe we'll stand a chance.
Covid is actually not a highly deadly virus and on balance we've handled it quite poorly "because economics."
There are much nastier pathogens we might have to deal with in the future.
(Well just warning them beforehand didn't work so let's use this to get the ball rolling, huh?)
Or even better, go volunteer at a hospital, I'm sure they could use the help. But leave the PPE for the people who are worried.
> These deaths could represent misclassified COVID-19 deaths, or potentially could be indirectly related to COVID-19 (e.g., deaths from other causes occurring in the context of health care shortages or overburdened health care systems).
That is not saying "those deaths were not caused by COVID-19". They explicitly mention that these could be misclassified.
https://www.washingtonpost.com/health/patients-with-heart-at...
https://boston.cbslocal.com/2020/05/04/coronavirus-covid-19-...
https://www.nydailynews.com/coronavirus/ny-coronavirus-queen...
- The herd immunity threshold depends on the reproduction number of the disease (R0). There's a basic reproduction number which is the rate they expect based on properties of the virus, and an effective reproduction number which is what ends up happening in reality. When effective R<1 the virus starts to die out.
- The basic number can't be changed but the effective number can be changed dramatically. Lockdowns and social distancing have already changed it, with varying levels of success in different locations.
- I have seen numbers thrown around that peg the herd immunity threshold for COVID-19 at 50-70%. This number is actually influenced by R, the lower R is, the lower your herd immunity threshold goes. JHU uses 70% here and that's the most pessimistic estimate I have seen.
- Wouldn't good, sustained social distancing lower effective R and thus produce a herd immunity threshold lower than 70%, perhaps much lower?
- New York is at 15-20% exposure already.
I read this article from the Oxford Centre for Evidence-Based Medicine a while back which observed that the early estimates for epidemics and pandemics are invariably worse than what ends up happening in reality, and it got me thinking. https://www.cebm.net/covid-19/global-covid-19-case-fatality-...
My personal bias is that I don't believe it is even possible, let alone wise, to keep people locked down on the time scales that are needed for controlling this virus, so they just kick the can down the road. But unprecedented levels of social distancing might be possible if we are smart enough. Every single person who works from home, wears a mask, or keeps their distance in public is helping.
Nobody who’s talking “reopening” is thinking about the drain on our society that so many dead will be.
This is not a mutually exclusive choice, but “YOLO” isn’t a good risk management theory.
Guess what? Those few million? They won't be contributing to the economy either way, even without the coronavirus - because approximately all of them are over 70.
This virus takes the old and the sick.
So yes, a lot of people talking about reopening have assessed how much economic damage that 500,000-3,000,000 deaths would cause - and it's not much. Arguably, it's negative due to reduced health care costs.
There's a lot of reasons to avoid killing off our old and infirm. But economics is generally not one of them.
https://www.google.com/amp/s/www.politico.com/amp/news/2020/...
> "While dealing with a Covid-19 pandemic, we are also on the brink of a hunger pandemic," David Beasley told the UN's security council. "There is also a real danger that more people could potentially die from the economic impact of Covid-19 than from the virus itself."
source: google "covid19 famine"
(I've generally seen this kind of not one life argument go hand in hand with the idea that Covid-19 cannot be compared to any other cause of death, for obvious reasons.)
The flu doesn't threaten healthcare capacity every season. A lot of focus is on the death statistics, but if someone recovers this does not mean they aren't a drain on hospital resources. Quite the opposite, it takes many more resources to nurse someone back to health than if they had died quickly. Corona is more contagious, leads to more hospitalizations as a proportion of those who are infected, and those hospitalizations are twice as long compared to the flu. And yes, it's also more deadly. If a hospital system collapses, then all other causes of death skyrocket. Our hospital system can handle flu without the economy shutting down, it can't handle corona. That's why we don't shut down every flu season.
"Data from China show that 20% of COVID-19 patients, though, are serious enough to get sent to the hospital. That's about 10 times more often than flu. Even though a great many people are hospitalized for the flu — the preliminary data for the 2018-19 flu season is nearly half a million — the rate of hospitalization is far lower: 1%-2% of cases, according to the CDC."
...
"Once a patient with a serious case of the coronavirus is hospitalized, the average stay is 11 days, according to a study based on January data from Wuhan — about twice as long as the five- to six-day average stay for flu.
The additional days mean additional stresses on the health care system. "To take care of intensive, really unwell people can often require two to three medical staff at one time, all in protective gear, for hours and hours," says Michael Ryan, director of the World Health Organization's Health Emergencies Program."
https://www.npr.org/sections/goatsandsoda/2020/03/20/8154082...
The flu is a recurring problem. We can't avoid it.
COVID is (hopefully) a once-in-a-lifetime occurrence (just as the 1918 flu was).
But letting Covid-19 run unchecked is more akin to raising the speed limit by 50 miles per hour. The prudent choice is not to go with it.
Yes, the flu kills people. So do lightning strikes. Airplane crashes. And anvils falling out of windows. None at anywhere near the scale of C19.
There's a balance. I don't claim to know exactly where that is, but I'm confident that "meh, granny's gonna die anyways, and I want to buy stuff and get a pedicure" is a rather ghastly approach to the current situation.
> What mindset? That preserving life is more palatable than making money? I'm ok with that. I'd rather be unemployed than dead.
You can "preserve life" by quarantining every flu epidemic. But this is an absurd argument. At some point the 'preserving life' argument breaks down.
Self-quarantine/shelter-in-place isn't about saving everybody. It's about keeping ICU units below capacity so we don't have more deaths than necessary. It's about giving scientists and doctors time to develop treatment plans.
The press in Netherlands showed the different ways of communicating a "lockdown". Though the actions were often similar, the way it was worded to the public differed quite a bit across countries. E.g. France talked about "going to war" (or something similar, cannot recall exactly). Spain and Italy both had different ways of announcing the same.
Trying to only protect/lock down people in a risk group (70+, plus overweight / lung problems) seems not to have worked. The virus spread already quickly despite a lockdown for all. It'll probably spread more quickly if that wasn't the case.
Plus bodies. So many bodies. They would have skewed older, but it's completely wrong to say this is a disease of the over-70s.
There are a lot of people of working age who barely survived infection and now have semi-permanent or possibly permanent - no one knows yet - lung damage.
The idea that we could have somehow managed this without any economic damage is pure fantasy. It has no absolutely foundation in reality.
Most critical workers are not over 70, and most do not have long-term health issues.
So yes, there was always going to be economic damage. But that doesn't meant there would have been nearly this much.
That's people with antibodies, not necessarily everyone who was exposed. There is zero information on seroconversion rates for mild and asymptomatic cases. We need to PCR-test a huge random sample, then follow up later with a serosurvey. Why are we not studying this? And why does nobody ever write a disclaimer listing their assumptions when making authoritative proclamations like this? It's going to shred their credibility into dust by the time this is all over, even if half of their assumptions turn out to be correct.
Another study suggests that 20% infected could be enough for herd immunity, if individual variation in susceptibility is similar to the estimate for SARS-CoV-1: https://www.medrxiv.org/content/10.1101/2020.04.27.20081893v...
> Even in hotspots like New York City that have been hit hardest by the pandemic, initial studies suggest that perhaps 15-21%6,7 of people have been exposed so far.
(goes on to mention how many people are dying)
> Some have entertained the idea of “controlled voluntary infection,”9 akin to the “chickenpox parties” of the 1980s. However, COVID-19 is 100 times more lethal than the chickenpox.
(goes on to completely dismiss whether that would actually lead to some level of herd immunity)
I get that the disease is more deadly. But is this one of those cases where it's better to pull hard and rip the band-aid off in one go? Nobody seems to really want to broach this topic seriously.
> To reach herd immunity for COVID-19, likely 70% or more of the population would need to be immune. Without a vaccine, over 200 million Americans would have to get infected before we reach this threshold. Put another way, even if the current pace of the COVID-19 pandemic continues in the United States – with over 25,000 confirmed cases a day – it will be well into 2021 before we reach herd immunity. If current daily death rates continue, over half a million Americans would be dead from COVID-19 by that time.
It's unclear to me whether the author is stating that herd immunity would take too long under the condition that we completely opened up the global economy or if we all continued to stay isolated.
We're definitely not going to reach 70% with the way things are.
Doing that quickly means hospitals and morgues overflowing with dead from COVID.
We HAVE talked about this from the beginning - this is the point of “flattening the curve”. If we let a million Americans die in a short time, it’ll tank the economy and overwhelm our healthcare system and ability to treat not only this disease but all others at the same time.
It’d be one thing if we were testing and tracing to try and limit. But this YOLO strategy is a recipe for complete chaos.
Even more inportantly: we don’t understand the long term impacts for those who survive. What is the 5-year mortality rate from complications? The US DOD has already decided that COVID infection disqualifies you for future military service (ostensibly because of the lung complications, but they haven’t said why formally yet).
The YOLO strategy that some folks are proposing is insane given how much we don’t know about this disease. It would be like proposing in 1982 that we get everyone infected with AIDS so we had “herd immunity”. Think about how that would have worked out for us... and then tell me that it’s a good idea to do it this time.
There may be a time when we know enough to make smart risk decisions about this disease, but less than 6 months after emergence isn’t it.
IF a few things are true:
1. The population being infected matches the demographics of the US as a whole. If the population is instead "healthy people under 60", IIRC the deaths go from 1-2M to 10-20k.
2. There is no way to infect someone deliberately in a way that reduces death rates. Variolation [0] has worked in the past, and is a possibility here. If it provides an additional factor of 10 safety benefit, you could now estimate 1-2k deaths, down from 1-2 million.
If you "rip off the band-aid" then hospitals will be overrun and more people will die because they couldn't get adequate care. The idea is to slow transmission until we get better treatments or even a vaccine developed.
Then there’s approaches like shielding that are interesting that are being discussed a lot. https://www.nature.com/articles/s41591-020-0895-3
There’s something to note about this virus infection fatality rate: we don’t seem to have an agreed upon value for it and there is quite a window. It may depend on locale, population health, age, etc. But the 1% for the diamond princess is quite high and I think that number was cited for caution.
In comparison Sweden now has 3040 dead and more than 600 new cases during the last 3 days, with no downward trend.
I don't think you can call this successful & I'm not sure how are they going to get out of this with such a sustained infection rate, if their neighbors can get things under control & minimize new infections. That could kinda turn Sweden into a plague lang that risks spreading infection in countries that handled the infection responsibly.
If you're talking about raw numbers of dead, _of course_ they have more dead, they're explicitly running this movie on fast forward. The question is number of dead in the long run.
301 dead per million in Sweden vs. 86 in Denmark, 46 in Finland, and 40 in Norway.
Based on Johns Hopkins study of confirmed cases.
So, of course Sweden's numbers are highest. Their gambit (as I understood it) was that if they infect people faster, they'll get immune faster. If they infect people faster, it obviously also means that people will die faster. It doesn't mean more will have died; Norway could just be killing people more slowly. We won't know who got it right until this is over.
(This is what I meant earlier by "raw numbers", but it's now a bad choice of words given that your figures are per capita)
Case fatality rate is defined as deaths per confirmed diagnosis. Estimates based on population testing would be of the infection fatality rate, which is a different measure. Assuming similar profile of cases diagnosed, a higher CFR could be a key indicator of greater health system strain.
> The question is number of dead in the long run.
Everyone is dead on the long run, when and how are the only things that vary and, generally speaking, sooner is less desirable.
Okay. So by case fatality rate, Sweden does look a lot worse than Norway. Your argument is that a high CFR implies a lot more dead than would have otherwise been, because it strains the health system. From a quick look at the news, it seems that they were afraid of an overload of the medical system in Stockholm 1 week ago, but it had been okay so far; not sure how it played out.
Couldn't a high CFR also just imply they triage their testing more because they have more sick people to test?
> Everyone is dead on the long run, when and how are the only things that vary and, generally speaking, sooner is less desirable.
Fair enough but that's moving the goal post. The claim as stated is "Sweden has significantly higher death rates", not "people are dying significantly sooner in Sweden".
To this point, though, I'll say that if there's a Deus ex machina that cures people next month, Sweden will have clearly lost out by letting people die sooner.
Why? The Diamond Princess has had 14 fatalities out of 712 cases which implies a fatality rate of 2%, with 8 unresolved cases still in the hospital that could push it even higher.
There’s other highly tested populations with lower CFRs and then there’s IFR estimates based on serology and indirect estimates on CFR that slew much lower than 1%
I would be careful about generalizing results from that particular study as 99.5% of the workers sampled were under the age of 70. If your sampled population does not include the most vulnerable age groups then of course your fatality rate will be lower than normal.
> There’s other highly tested populations with lower CFRs and then there’s IFR estimates based on serology and indirect estimates on CFR that slew much lower than 1%
I'm curious what you mean by "much lower"? Some of the antibody studies published recently have found prevalence rates that were not much higher than the false positive rates for the test, making confidence in their results somewhat dubious. In studies where the prevalence has been higher (e.g. NYC, Geneva), the IFR is trending towards 0.5-1%.
This meta-analysis was published yesterday which suggests an IFR of 0.75% (CI: 0.49-1.01%): https://www.medrxiv.org/content/10.1101/2020.05.03.20089854v...
What this completely ignores is that obviously only young and healthy people would attend those "corona parties" and thus the CFR of the voluntaries would rather be around 0%. And obviously those people wouldn't immediately go see grandpa but instead self-isolate for three weeks.
I hate that so many are pretending that it's like russian roulette when in reality the odds are so highly skewed.
A lot of people seem to be using it as some example of a country who haven't introduced lockdown measures, whereas in reality they are actually self-isolating and social distancing, just like everyone else.
Worth noting also that Sweden has the 6th highest case fatality rate in Europe (other Nordic countries are near the bottom of that list).
Lots of small businesses are defaulting. I'm trying to support my local restaurant by eating lunch there every day.
All in all, it's a mess.
It's very hard to make any proper comparisons of numbers before this has run its course. What if Sweden doesn't have a second wave but other countries do?
* C19 deaths per million people (Sweden vs. Others, but you can configure for your favorite jurisdiction) - https://ourworldindata.org/grapher/total-covid-deaths-per-mi...
* Architect of Sweden's no-lockdown strategy insists it will pay off - https://www.ft.com/content/a2b4c18c-a5e8-4edc-8047-ade4a82a5... (I disagree that intentionally sacrificing people like this is the moral thing to do)
As I understand it (which is poorly), they are front-loading the deaths instead of a lockdown.
I won't touch on the moral side of things.
Edit: Replaced "correct" with "accurate". Can't come up with quite the right word there.