I think I found something important.
But I will let you be the judge. And I would love to hear from any epidemiologists that read my blog.
The number of COVID cases is increasing in Oregon and Washington, while most of the other states are seeing declining or stable COVID caseloads. The increases are sufficiently worrying that Governor Inslee may reverse the opening of the State.
There is a lot of talk about the need to get herd immunity, in which enough of the population will be vaccinated or have had COVID so that the virus lacks sufficient susceptible people for spread. But no one seems to be sure about what percentage of immune individuals is needed. 60%, 80% or 90%. Some scientific papers and media accounts suggest we will never get to herd immunity. But what is the truth?
I am not an epidemiologist, but it seems to me that there should be sufficient information to get some real answers to the above questions. And so I sat down with a cup of coffee (actually several) and the data from the Centers for Disease Control and cranked through the numbers.
And I found something interesting.
For each state, I estimated what percentage of the population is immune to COVID-19 and then I compared that to the change in the number of new cases over the past few weeks.
I found a significant relationship between the two. And I may have discovered the magic herd immunity percentage.
Estimation of the Percent of the Population Immune to COVID-19
First, I started with the percentage of each state's population that has been reported to have had COVID-19, which ranges from roughly 4 (Oregon) to 14 (Rhode Island) percent.
There is an extensive literature suggesting the actual numbers of infected people are far higher than the official case numbers, with several research papers suggesting that a factor of three is reasonable. So I started by taking the CDC state case percentages and multiplying by 3.
Then I secured the vaccine data for each state, and based on the CDC numbers of single and double vaccinations, I assigned 60% of the total vaccinations as the number of people that have gotten at least a single shot. I then added this number to the 3 times case percentages noted above, to get a final fraction of immune individuals for each state, with numbers running from .56 (Oregon) to .936 (Rhode Island). This will undoubtedly be a bit of an overestimation because there will be some people that have had COVID that are vaccinated, but that would only shift my estimates of herd immunity by roughly 5-10% at most. In any case, the approach is consistent for all states.
That Change in COVID Cases for April 15-April 28th
My hypothesis is that increasing percentages of antibodies for COVID should be related to declines in COVID case numbers. Sot I next turned to the CDC COVID trends pages.
Here is the one for Oregon. Not good. Cases have been rising quite a bit (blue are daily values) and red are 7-day running means).
Models may differ based on the activities of the groups getting vaccinated. If we vaccinate mainly those who are at high risk of transmission due to work or social life then the herd immunity number could be lower. The reason we may need to get to 70%, not 60% as someone who has read Malcolm Gladwell's 'The Tipping Point' might conclude is that it's the most careful of us who are getting vaccinated.
ReplyDeleteProvocative analysis, thanks.
ReplyDeleteDid you say that you added the population vaccinated to the population previously infected? Isn't there tremendous, or at least significant, overlap between those populations? ("I then added this number to the 3 times case percentages noted above, to get a final fraction of immune individuals...") Conflating them would only lower the apparent threshold for herd immunity but I wonder if it might do anything to either promote or disrupt the state-by-state pattern you found (and I wonder if there is significant variation in the overlap of those populations from state to state that might confound things further.)
In general, my first reaction is to wonder about the variability in what it takes to reach herd immunity... i.e. it's a dynamic thing, and depends a lot on how much a population mixes, how many densely populated areas/events there are, how much travel in and out, whether people are distancing, wearing masks, what kind of local transport, weather, geography, etc., so I'd expect the magic number to vary a lot from place to place, and wouldn't want to make policy changes based on, e.g., "a week or two until 70%". Especially if (somewhat regional) variants turn out to also move the threshold a few points (or more).
I hope the epidemiologists are doing these kinds of calculations, but I agree that I haven't seen much if anything written about it publicly. I think it's dangerous to draw strong conclusions, particularly from armchair epidemiology, but I appreciate the brainstorming because sometimes the authorities actually are asleep at the switch. I want to believe that this is Global Pandemic Epidemiology 101 stuff, but I guess we know better than to make assumptions like that. :-)
Moreover, I wonder if this is the kind of thing that can be nailed down with enough confidence to be actionable.
Really good points. All of the variables you describe affect estimates of herd immunity thresholds. Epidemiologists can use models to make predictions but they require many assumptions and are accompanied by lots of uncertainty.
DeleteA few other important factors:
1) The factor of 3 that Cliff used to estimate prior infection is most certainly too high. The paper he links uses data from April 2020, when testing was extremely limited.
2) Cliff's simple analysis is confounded by other state-level factors, including policy changes and behavior (eg, school closures, masking, travel, proportion working in person) that also vary over time and affect cases. Cases go up, a state changes policy and people change behavior, and then cases flatten and fall. It isn't clear how much of the relationship between proportion protected and recent case trends is due to these other dynamics.
3) As others have pointed out below, emerging variants threaten population-level immunity. How much isn't so clear yet, but viral evolution in a partially protected population is a given. Rapid uptake of vaccination while cases are relatively low is our best chance of limiting this issue.
This seems well thought out. The only problem is that Washington uses PCR tests with a cycle threshold of 37 and 39 (they have 2 different tests). This is far too high so a huge percentage of these "cases" are actually false positives. See (Your coronavirus test is positive. Maybe it shouldn't be. NY TIMES. https://www.nytimes.com/2020/08/29/health/coronavirus-testing.html) Also https://www.publish0x.com/washington-state-covid-data-is-garbage/washington-state-doh-labs-covid-pcr-tests-are-at-the-highest-xjmwvkp?a=KQe1jX1RaJ&tid=CM In addition they don't know when 50% of these cases actually happened so any trend charts they are producing is unreliable and essentially fictitious.
ReplyDeleteIn April 2006 there was a whooping cough epidemic scare at a hospital. The public health authorities reacted. The end result was 1000 people furloughed, 172 were told they had the disease and thousands were given antibiotics and vaccines. 8 months later it was discovered that the tests were producing false positives and there was no epidemic. Whoops. The test they were relying on is the same PCR test being abused for COVID. See https://www.publish0x.com/washington-state-covid-data-is-garbage/history-lesson-faith-in-quick-test-leads-to-epidemic-that-wa-xerkgdr?a=KQe1jX1RaJ&tid=CM for a summary and https://health.maryland.gov/newsclippings/archives/2007/jan07/012207.htm for the original report.
pretty much makes everything a moot point when using an invalid test, yea? Never mind that they say the vax doesn’t help the spread but lessons the symptoms. What an upside down reality we are living in!
DeleteOregon and Washington did a good job keeping cases low, but like you point out, that success is definitely part of the problem now:
ReplyDelete“The double-edged sword of Oregon’s COVID-19 success”
https://www.opb.org/article/2021/03/17/oregon-covid-19-year-2-data-science/
The 70% looks about right, but I think finding an exact figure would be impossible because behavioral factors (such as compliance to mask wearing mandates) affect how soon herd immunity is reached.
Vaccinations are moving along much faster now than in early spring, so I expect the surge in new cases will be short lived. Full opening by June.
How would you feel if medical experts started blogging about climate and weather? I like this weather blog, but I wosh it wasn't full of politics and other things.
ReplyDeleteI had to laugh about this one. All kinds of folks who have no idea what they are talking about blog, tweet and write about climate. Take the Seattle350 advocacy group. Or Jon Talton of the Seattle Times.....read his article in today's newspaper. There is nothing wrong with no experts providing arguments, as long as they provide the details of their logic. Perhaps they found something missed by the "experts." That happens all the time....
DeleteNumbers and numbers and meteorologists are always dealing with numbers
DeleteYep, only The Annoined shall discuss the viroids. We would hate to offend poor lodokol. LOL
DeleteWhere did you find the two week change in covid cases by state? I can't find that data anywhere.
ReplyDelete@Chris May:
DeleteFrom John Hopkins University School of Medicine you can find daily cumulative test & case data in CSV or via a JSON API. From that cumulative data you can easily derive daily data and 7-day / 14-day / 30-day trends or changes. It is available by state, the District of Columbia, and the U.S. Territories.
https://coronavirus.jhu.edu/testing/jhu-covid-19-testing-data
From the NYT GitHub site you can obtain current and historical data for states as well. They offer it in the form of date/state/cases/deaths cumulative counts. Again, it is easy to derive daily counts of each variable and, from there, 14-day trends or changes.
https://github.com/nytimes/covid-19-data
The NYT's COVID site may even have 14-day charts already built and is likely available even to non-subscribers.
I've tried the CDC site after the COVID Tracking Project ceased updates in March. But I found the CDC's site a little clumsy to use and the downloadable data files I found were quite huge (sometimes in GB). But I think you can download pre-built charts from their sites.
..
I enjoy building data charts and visualizations for my own learning experience much like others like to do daily crosswords or sudoku puzzles.
https://www.litterrocks.com/states-newcasesdaily
I use the data primarily from the JHU and NYT sites. I believie Oxford University also offers data for downloading.
Cliff, interesting approach. Repeat infections are a thing and there is good reason to expect vaccine protection to wane over time. Might the number of protected individuals be even smaller when considering a group a people that got COVID early (8+ months ago) and have not been vaccinated? Many of those members likely had mild or asymptomatic cases. Additionally, with new, more infectious variants, I would expect repeat infections to increase as well.
ReplyDeletePart of this paradox is that WA and OR did relatively good jobs of preventing widespread infections early and now that is working against us in obtaining herd immunity. Working against us at a higher level are those who refuse to get vaccinated. Here in Pierce county, our case rate is climbing rapidly and there are many vaccine appointments going unfilled right now.
With other corona viruses immunity wanes over a few years. 8 months wouldn't make much of a difference but after 2 to 3 years people start getting sick again. I think as 2/3 of the population gets vaccinated and we get booster shots every 2 years we should be fine.
DeleteI agree with CLiffmasses analysis that we are close to heard immunity and because we are close to heard immunity I would like to see the government say when x % of the population is vaccinated and the number of hospitalizations drops below y all restrictions will be lifted. The government wants to totally eradicate the virus but unfortunately small amounts of the virus will always be with us, so it seems like the government has set a goal they can never reach. My co-worker won't get the vaccine because the government has messed up his life too much and he no longer trusts the government. If the government promised to remove restrictions at a certain point he would trust the government more and have an incentive to get the vaccine. I am having to rent a room in a house full of strangers even though I make 75k a year, I could afford some of the cheaper apartments but I want to save money/not live paycheck to paycheck. My new landlord doesn't want any guests until the pandemic is over trouble is there is no definition is when the pandemic is over so she might ban guests indefinitely. Its hard for a young person like me to live like a monk especially when I don't know when this will end. These restrictions have been a huge burden on the poor and its getting really hard for people like me to continue living such isolated lives/wearing mask everywhere.
DeleteCan you compute historical trajectories? For those states in the 80s and 90s what did they look like as they passed through the 60s and 70s?
ReplyDeleteThis sounds a little like pseudoscientific justification for reopening. It would be interesting to apply this math in hindsight to India which looked like it had beat this thing in January and is now exploding out of control. And then look at a place like Republic where our Republican candidate for Governor was once the town sheriff. Your magic number may be a bit more dynamic depending upon the circumstances.
ReplyDeleteThose who wanted the shot have gotten it at this point.
ReplyDeleteThe rest won't ever get it....because politics. Or religion. Or whatever. It will be a 50/50 split along the usual red/blue BS lines. If anything, we are consistent. So about 50% will be immune from the shot. The rest have to catch the actual virus.
Reading between your lines, Cliff, basically is a strategy where we ditch the masks and get on with it....allowing herd immunity to be reached by basically just allowing people to get sick. Since some would rather get actual Covid than get the shot. Well, perhaps its time to allow that to happen for them all that much easier.
As crappy as it sounds...I'm fine with that. You want to take your chances with Covid? Have at it. Your choice! Medical professionals might have other ideas, however.
Well, we need to ditch the masks outside. I did not say anything about inside. And there are plenty of people still getting the vaccine. In the end, people have a choice....get the vaccine or take a chance getting COVID. My point is that we are going to reach herd immunity in the state, one way or the other. And pretty soon.
DeleteCliff, Covid is not going to be something that magically disappears and we are right back at blissful summer of 2019. Herd immunity is a moving target that some argue is unattainable. It will end up endemic like the flu, with various strains challenging the effectiveness of vaccines at any given time. The point were we all collectively say "ENOUGH!I'll just take my chances with the damned virus. Just give me my life back!" will arrive long before any semblance of "Herd Immunity". Which, as you imply, might actually speed "Herd Immunity" along.
DeleteWhen there is broad understanding of Covid and hospitals are certain they have uniform capacity perpetually, then the restrictions will officially cease. Probably more than likely the restrictions will just be wholesale ignored after a point and their further enactment will be moot. States such as Texas will become considered safe harbor, and Texas would gladly welcome expanding their tax base at the expense of states such as California or Washington.
Ultimately, people will still get sick and they will still die, just like the flu. You have been on this Earth quite a while, and are fully cognizant of how resistant Humanity can be about doing what is actually in their own best interest to do. You can count on 30-40% doing the exact opposite of what they are told, just based on whom is telling them to do it. At least in the USA, anyway.
Getting a vaccine now or never getting one are not the only two options. Some people might prefer one of the other 300 vaccines still in development over the experimental gene therapy versions. Why can't people choose one of those? What's the deal with the insistence on people needing to take the experimental variations? (I'm not saying you are insisting this - but your post made me wonder why that's never an option in regards to this topic.) It always seems it is either take the gene therapy or you are anti-vax.
DeleteA pool of unvaccinated people increases the chance that new variants could arise that evade our current vaccinations and put us all in danger. Compulsory vaccination feels wrong, but let's not lose sight of how these "individual" decisions can affect everyone.
DeleteCliff, thank you for this analysis. I'm always leery of people who want to limit speech to specialists (yeah, I'm talking about you WAYNE). What these narrow-minded people seem to forget is the benefits of having intelligent people with diverse backgrounds look at a problem. They have the "stay in your lane" mentality, which stifles innovation and silences speech. If people like Wayne were in charge, Einstein's Theory of Relativity would never have been taken seriously because he was a lowly patent clerk. Luckily for humanity, we have many more enlightened and liberal-minded people than we have regressive elitists like Wayne.
DeleteTruth Seeking, You can now get the J&J. I volunteered for an Astrazenica vaccine trial before the Pfizer vaccine was available, and hopefully that will soon be approved. Though it is ironic that the RNA vaccines appear to be (slightly) safer! But though blood clots have been reported in those vaccinated with J&J or Astrazenica vaccines, it is very rare.
DeleteJ&J is an experimental technique as well. Different from Moderna and Pfizer, but still experimental.
DeleteAs someone who got Pfizer #2 six weeks ago, and who observes that vaccines are now available for free to anyone who wants them, I think we should go 100% back to normal by July 4, which will give ample time for anyone to get vaxxed and develop the antibodies. Anything past that, and it'll just be more social control from the "progressives," who have a deep-seated urge to tell everyone else how to run their lives.
DeleteOne thing your analysis does not address is who is getting sick, what the result of that is and what the medical system can handle. If people are still dying, should we be reopening before we can immunize via vaccine? Or do we go with natural infection and not worry about the deaths? I am not being facetious, I am not sure what the data says. Also, as I understand it, an important argument FOR restrictions is to make sure our medical system does not get overwhelmed. That seems like an important part of the equation that was left out.
ReplyDeletePlenty of statistically significant analysis has been on the subject- you're just not aware of it. The UK, US, and Israel have plenty of data that's been crunched and published within the last month in NEJM, The Lancet, Nature, and other respected publications.
ReplyDeleteEric Topol has several nice one page layperson friendly explainers on the topic in Cell, also in his twitter feed.
I'd be curious what you found for herd immunity rates in Washington by county. The phases are set by county not statewide and the herd rates, I suspect, vary quite a a bit across counties.
ReplyDeleteDon't you work somewhere that you could find well qualified epidemiologists to bounce this off of instead of posting to the internet?
ReplyDeleteI have gotten several emails from epidemiologists who said my analysis is valid.
DeleteA huge part of epidemiology is data analysis and forecasting, which Cliff is more than qualified for. There’s no magic data Jay Inslee is using that Cliff isn’t aware of.
DeleteAs a scientist, I would have thought you might have tried looking for peer reviewed literature on the topic. https://www.nature.com/articles/d41586-021-00728-2
ReplyDeleteI have read that paper and many others on the topic. The first thing I did was check out the peer reviewed literature....and was disappointed. That is why I did my analysis. Can you cite somewhere in the literature that attacked the problem the same way?
DeleteBrilliant, I agree with you. Glad you had a lot of coffee available. It makes sense and the biggest issue is that each city and state has a different way of thinking and seeing things. I am blessed to live in Washington.
ReplyDeleteOh Cliff, Do you delight in using science to confound the wise? If we reach herd immunity people will stop getting the (experimental) vaccines and the now talked about yearly booster shots. I have a question for you...... If the masks collect the covid virus (the only reason to wear them) why aren't the masks classified as hazardous medical waste and have to be disposed of as such? I see discarded masks everywhere I go.
ReplyDeleteLaundry detergent kills the virus. I wear a mask for one day, wash it and re-use it. Sunlight also kills the virus so if you left it outside all afternoon it would probably be okay.
DeleteCliff, Very insightful as usual. I respect your analysis and opinion more than the CDC and the other sources.
ReplyDeleteI'd be curious to see information about the longevity of antibodies included in these numbers. For instance, research from both Pfizer and Moderna say the body produces mRNA antibodies for approximately 6 months. How will that affect herd immunity in say, people who had Covid last year vs the population now vaccinated?
ReplyDeleteIn addition to humoral immunity, there is also cell-mediated immunity. You have got both T and B cells
DeleteIt's possible we'll be advised to get a new vaccinations each Fall, as with the flu, at least for a few years.
DeleteIt could also last for years. They'll be very conservative with their estimates.
DeleteI've been reading about the 70% figure to reach herd immunity for a while. I don't think you found anything new here. Regarding where we're at in Washington, the NY Times daily Coronavirus Tracker newsletter says the state is only at 33% of folks fully vaccinated. I think we have a long way to go.
ReplyDeleteGood analysis. I’d just point out that deaths are pretty constant and at a low level. That indicates that infections are mostly among younger people who will have a mild disease course. Rising cases is not a good metric for policy choices.
ReplyDeleteCliff, great analysis. Always enjoy how you apply basic concepts of analysis and science to topics that aren’t even in your wheel house daily. Numbers are numbers! I wonder though, can we really apply such a broad brush to a herd immunity milestone when we know that mutations of Covid -19 will detract from those who have sufficient antibodies and coverage? Two steps forward, one step back. Isn’t the the main concern with the herd approach is that it will allow too much room for development of mutations that set us far back before we reach herd on the main strain? Predominant strain in wa is already the B.1.1.7 which took over in a month since introduced (which the Vax may very well offer protection for), but perhaps not others that develop if we keep growing farm open to take its own course toward herd?
ReplyDeleteThank you sir. I think your estimates are more likely true as you dont stand to make a bunch of money either way. Happy to hear your educated estimates... well done sir.
ReplyDeleteCliff,
ReplyDeleteYour back of the envelope estimates aren’t far off from my own, and I’m an epidemiologist. There’s just one thing missing: herd immunity as a concept isn’t an equation involving constants, but rather is a balance between forces that literally evolve. Right now we might be close to herd immunity, under the assumption that vaccination yields strong protective immunity. It probably does against B.1.1.7, which is currently the dominant strain, but the P.1 strain is rapidly expanding in our region, and there’s decent scientific evidence that P.1 is more capable of either infecting a vaccinated person or re-infecting someone who previously had a bout with covid (and survived).
This is the huge unknown: we do not yet know whether additional mutations on the P.1 backbone could fully escape vaccine induced immunity. But flourishing transmission of P.1 in a community with high vaccination rates (like ours) is exactly the selective environment where such a mutation would be expected to emerge. If it hasn’t already.
So simply relying on vaccination to stop the current surge is reasonable, but only insofar as we are willing to assume that immune escape is impossible. I’m not at all certain of this assumption, so I’m strongly supportive of Inslee re-tightening restrictions until we get the current wave under control.
Dr Chris Carlson (your old ally in the math wars)
Ps here’s a very interesting preprint from a group who I trust, demonstrating that convalescent serum is less effective against P.1.
https://www.biorxiv.org/content/10.1101/2021.03.01.433466v2.full
This is what I think is Inslee's ideal plan:
ReplyDelete--Shut down everything that isn't a supermarket until November 2024. This includes restaurants (not even delivery or carryout).
--Require all citizens to wear five masks and two pairs of gloves.
--Increase social distancing to 6 meters (20 feet).
--Mandatory 6 months in jail and $5000 fine for anyone not wearing gloves and masks.
--Dusk to dawn curfew throughout the state, with violators facing 5 years in prison.
You forgot something: None of those rules should apply to "historically disadvantaged groups," because we all know that the virus is very selective. LOL
DeleteIn the NYT this Monday Morning:
ReplyDeletehttps://www.nytimes.com/2021/05/03/health/covid-herd-immunity-vaccine.html
I thought WA was estimated to be closer to 5% confirmed cases, so 3x (using your factor) would be 15% exposed to covid. Plus the 30% vaccinated. Isn't that closer to 45%?
ReplyDeleteI don't see how you say we'll be at 70% in a week.
https://www.statista.com/statistics/1109004/coronavirus-covid19-cases-rate-us-americans-by-state/
I admire your attempt to tackle a non-weather subject.
ReplyDeleteBut this analysis doesn't mean much because the case data are faulty.
COVID is the first flu in history where a "case" is defined by a positive PCR test ONLY -- PCR is notorious for false positives You could be a COVID "case" with no symptoms?
That was never before done with the flu.
So any analysis using "case: numbers is immediately tainted by poor quality data.
In addition, herd immunity ONLY means the number of infections is not increasing.
If the decline in infections is fast, that's a lot different than a slow decline, but both would be herd immunity.
The COVID death counts were also unique. Prior to 2020 flu was not a cause of death -- it was a contributory factor, if listed on the death certificate at all. In 2020 a sickly person in a nursing home with heart disease who died with flu symptoms would be listed as a COVID death, with heart disease as the contributory factor. Mere cold symptoms could be used to define COVID as the cause of death, with no PCR test at all.
The number of hospitalizations is probably more accurate.
The horrible 1918 1919 flu pandemic had two death spikes and disappeared with no vaccines and no medications to treat it. Should we expect COVID to last longer than the 1918 flu, even if there was no vaccine? All flu's mutate -- will the vaccines do as well with the mutations or will people need flu shots every year?
Outside of nursing homes the death rate is 0.2% or less -- that is not a death rate of a deadly disease. For people under age 40, COVID is no worse than ordinary influenza.
COVID is not a type of flu, it's a type of cold. It is caused by the coronavirus SARS-CoV-2. Coronaviruses have been known to exist since at least the early 1900s and are part of a group of viruses that cause the common cold. Otherwise, I agree with everything else you wrote.
DeleteHi Cliff,
ReplyDeleteI am a Public Health Nurse working in communicable disease in rural northern california. one variable that I think is important to throw out into is the reality of false negative COVID tests.
https://pubmed.ncbi.nlm.nih.gov/32422057/
I would suspect the herd immunity number might actually be a bit higher, accounting for this.
Um, in your zeal to push your panic agenda, you seem to have overlooked the far greater false positive rate. Gee, what a surprise. LOL
Delete"panic agenda" that is a gross over judgement. I actually am pretty moderate in most respects including on the COVID response, which has been the last year of my life 40+ hours per week. Our goal in public health is to protect vunerable populations, and indeed we as a society have done a terrible job of doing that, and maintaining mental sanity for our general population --- That is another issue that in my humble opinion is a downfall of our core societal value of individual freedom and neoliberal economic policy.
DeleteIf you can point me to some peer reviewed data on false positives, I would be curious to see the studies. :)
If anyone wants a reference about PCR diagnostic test in general, I found this one:
Deletehttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC7106425/
Both, false positives and false negatives are possible. But I find the section "Clinical significance of positive PCR" really interesting:
"PCR assays may detect microbial pathogens at concentrations below those of previously established gold standard reference methods. Distinguishing whether this result represents a false-positive finding and establishing the clinical significance of these findings is challenging."
PCR was originally not intended to be a diagnostic tool for viral diseases. It's an afterthought.
Without disrespect to you there could potentially be variables you don't understand that an epidemiologist might, but I do think that Inslee's logic is horribly flawed and remains reactionary a year after this whole thing started.
ReplyDeleteCliff, another thing to consider is whether prior COVID-19 infection confers the same level of immunity as the mRNA vaccines (especially if the infection was before the UK/South Africa/Brazil variants came in). The Brazil experience has shown that prior infection didn't provide much protection against reinfection and it will probably be the same way here. In other words, the multiplier may be somewhat less than three.
ReplyDeleteCliff, your immunity numbers don't seem to be accurate at all. You can see this by going to https://covid19dashboardgt.shinyapps.io/us_immunitylevel/ . Even if you use the metric of at least one vaccine dose, RI has about 75% immunity (not 93.6%) and WA state has 52%, not 63%. Note that these are likely overestimates.
ReplyDeleteAs for the magic .7 level driving case drops, note that CA has had significant drops with population immunity of only 64% (if you count partial vaccination) and 54% if you only consider fully vaccinated folks, which is probably the best approach. VA has also seen significant drops with partial immunity of only 50%.
I don't think "herd immunity" is set in stone. If 35% of people are immune, all else being equal, that should cut down on transmission dramatically. But if with 70%, what I realized the past few days is that one has to consider sub-populations. Suppose everyone over 50 is now vaccinated, for simplicity. Some social congregations are composed mostly of people under 50 - say, universities, certain churches, Amazon or Microsoft. To the extent that people are segregated by age, diseases can still surge locally within those sub-populations. So we may experience the converse of last year, when Covid spread in old folks homes. Except the younger people will just get sick, then recover, for the most part.
ReplyDeleteSo I doubt we'll have quite the surge in fatalities in a few weeks. I think we're pretty much out of the woods, so long as those over 50 have had the shots.
Also, you mentioned getting the first shot . . . Probably some adjustment should be made for the lesser effect of the first shot (and some skip the second, like my wife who is off to Japan), along with overlap with those already infected. But I think the gist of your post is on the right track.
In April 2020, Czechia (Czech Republic) was the poster child for strict mask mandates. Never mind closed borders and reduced contact, it was the magical masks that protected the people. But when restrictions--with the exception of indoor mask mandates--were relaxed for the summer, cases and deaths shot up in the fall. Expanding the mask mandate to outdoor activity did not slow the rise.
ReplyDeleteOn March 1, 2021, Czechia surpassed Belgium (another masked nation) for most per-capita cumulative deaths. Deaths also rose in masked Hungary, and on April 22 surpassed Czechia's, reaching 2,889 per 1M on May 3.
Norway, with closed borders but no mask mandate, recorded just 140 COVID deaths per 1M as of May 3, 2021. No-mandate Sweden got dissed by media (with 1,391 per 1M it's now #28 globally) while Norway (#86) got little mention.
Last fall, India took Czechia's role as mask cult darling, with a massive drop in new cases attributed to a strongly-enforced mandate (95% compliance!). But though the per-capita COVID death count was still just 161 per million on May 3, a sudden upturn currently taking place looks a lot like India could go the way of Czechia and Hungary.
Could this have anything to do with Czechs and Hungarians retreating into heated spaces in winter and Indians retreating into air-conditioned spaces in summer?
Absurd claims made for the efficacy of universal masking are insidious lies, and those who perpetuate them are not motivated by concern for the health of others.
The Indian population is younger and more rural, despite a huge urban population. Numbers per capita are nowhere near the US, still less Belgium -- though I suspect the numbers measure only a part of the spread. Going by official figures, Brazil has been the real epicenter these past weeks -- as many deaths as India, officially, despite a fraction of the population.
DeleteUSA
DeleteMedian age: 40, COVID-attributed deaths: 1,745 per 1M
Brazil
Median age: 32, COVID-attributed deaths: 1,922 per 1M
India
Median age: 28, COVID-attributed deaths: 161 per 1M
Niger
Median age: 15, COVID-attributed deaths: 0.41 per 1M
hi Cliff, good analysis as expected. I have heard 70% from the beginning some 15 months ago so now we see if we can get there.
ReplyDeleteI have wondered about the population of our state, say, and it is divided in two parts. One vaccinated and the other not. Will not over time all of the unvaccinated contract the disease until enough people either have had the disease or been vaccinated. At a certain point true 'herd immunity' is likely to take place. Years?
ReplyDeleteThe model lacks the occurrence of new variants, especially those that escape the vaccines. This may not matter now but it may matter in the future. It will be interesting in the next 12 months to compare states that have achieved herd immunity through natural exposure versus vaccination.
ReplyDeleteThe mRNA vaccines induce antibodies against the spike protein and do so very strongly. Natural exposure induces antibodies against several sites of the virus and less so against the spike protein. It is known for example that in 2009-2010 people vaccinated against the seasonal flu were more susceptible to the novel (swine) flu.
Looks like you were exactly right, Cliff! Seems like the inflection point in WA state was on 5/4! Nice job:)
ReplyDeleteHuh! What about Texas and Florida???
Delete
ReplyDeleteThe problem is that covid mutates. Give it enough leeway and variants that are impervious to the immune response to the original "natural" infection will emerge. That has long been known. Brazil was one of the first exemplars of this phenomenon. So far the immunity conferred by the mRNA vaccines has proved much more durable than and broader than that conferred by natural infection.