This report makes no attempt to control for lockdowns or other restrictions. Countries with high levels of vaccination relax those restrictions, increasing COVID infection rates.
And of course focusing on infection rates is really the wrong thing. We know antibody immunity declines over time, but protection against serious illness stays strong.
Yeah, countries absolutely have a negative feedback loop. In the UK, society is largely open and free to intermingle, yet the virus isn't spreading exponentially. In my mind, that's because the vaccine has significantly lowered the ability of the virus to spread, and society has opened up just as much as will cancel that gain out.
Not really the case. Vaccinated people get light recurrences and frequent encounters with the virus keeps the antibodies circulating almost indefinitely.
One think I noticed when getting my vaccine shot was the line going in was very socially distanced with everyone wearing masks. That changed the moment someone was vaccinated. They would come out of the centre, take off their mask, walk right next to the queue, their behavior instantly changed from getting the jab. Almost everyone did this. There is a social impact of having the vaccine that increases transmission behavior in my experience and it was almost everyone that did this.
These social impacts, from countries opening up more due to high vaccination rates are very important factors as clearly they do and have controlled the spread throughout this pandemic and if you don't control for them you can't compare transmission. Even if you compared countries measures you would still end up with this behavioral change meaning people were more likely to spread and catch once vaccinated.
That behaviour is quite culturally specific. I didn’t see any of that sort of thing here in Australia when I went to the vaccination hub. And I’ve had two people refuse social interaction recently because of vaccines. One person because she’s only had her first shot, and the other because it’s only been a couple days since getting the second dose and it takes 7-9 days for the vaccine to do it’s thing. Both highly neurotic people.
I’m sure there’s also plenty of Australians acting recklessly - but the ratios are probably very different country to country.
There's definitely an element of false security after having the vaccine. But standing near someone in a queue is a very different risk profile than simply walking past them.
It's still is an important insight though. The vaccines do not lead to herd immunity.
There isn't a viable herd immunity strategy yet, which implies very young, very old and immuno compromised individuals are still at risk.
It also means 'encouraging' people to take the vaccine won't help, except to keep those people out of the ICU, and if that's the intent it seems more 'freedomy' to me to just not treat unvaccinated covid patients when it peaks.
Because herd immunity is the argument used to defend mandatory vaccinations, and it appears to be a false argument.
Because when you run out of ICU space, there is triage. There was triage last week. There is triage next week.
Who gets care first?
My point is that many people who are unvaccinated would prefer this, than to be forced to be vaccinated. (and they are effectively forced now in many places)
There were two arguments used to justify this imposition:
(a) - the herd immunity helps protect those can't protect themselves (too young, too old, immuno compromised)
(b) - it overwelms the health care infrastructure making it so that in certain places people can't get emergency care because the ICU's are full
Now the first argument (a) seems to be not true, and the second argument (b) seems to me like 'personal responsibility'. If you want to freedom to make these decisions for yourself (such as to get vaccinated), you are only accountable for the consequences of that for yourself.
The idea that you demand to not be vaccinated, but still get first-come-first-served emergency medical care at the expense of others, sounds to me entitled, spoiled and selfish.
So, if it turned out that the vaccine caused side effects that resulted in large cases in the ICU or cancer incidents down the road, you are willing to put them at the end of the line since they decided to take the vaccine? Do you not see how many variations of this argument can be made and how nonsensical the results would be?
City dwellers have increased incidence of cancer or whatever other ailment, so their choice to live in cities precludes them from medical services. Marijuana smokers have higher incidences of x so they don’t get treatment. Motorcycle, bicycle, and scooter drivers are risky actors so their choices preclude them from care. Where does it end?
When there are enough ICU nurses available to meet the demand.
>Doesn’t seem like freedom to me.
Again, can you give me a definition. It sounds like something 'empty' an American would say. Do you mean, Liberty? Like from the French Revolution?
This discussion doesn't matter in the US anyway, because you already have economic triage, where there is just no hospital near poor people. So in the US, the poor people are just 'free to die'.
When we need to triage it absolutely makes sense to take into account how much the ailment depended on the patient's known-to-be-risky choices. Motorcyclists and drug users, sure.
Since the major vaccines have minimal known side effects, we should not now or later blame people for taking them.
It plots rate per 100k population at the time of July when fully vaccinated rate is very less. Atleast the plot should be rate per 100k vaccinated and rate per 100k unvaccinated.
> We know antibody immunity declines over time, but protection against serious illness stays strong.
The article specifically mentions that that is not the case.
> the CDC reported an increase from 0.01 to 9% and 0 to 15.1% (between January to May 2021) in the rates of hospitalizations and deaths, respectively, amongst the fully vaccinated
> This report makes no attempt to control for lockdowns or other restrictions
Haven’t we seen previous reports that basically state that lockdowns and or massive restrictions seem to have a negligible effect on spread?
> The article specifically mentions that that is not the case.
The citation they're using doesn't load for me. I'm looking at results like this:
https://www.reuters.com/business/healthcare-pharmaceuticals/...
> The analysis showed that the vaccine's effectiveness in preventing hospitalization and death remained high at 90% for at least six months, even against the highly contagious Delta variant of the coronavirus.
> Haven’t we seen previous reports that basically state that lockdowns and or massive restrictions seem to have a negligible effect on spread?
Those reports, even if they're completely valid which is questionable, only show that government-mandated lockdowns have similar effects to voluntary behavior changes in response to waves of COVID infection, which again the OP does not control for.
Another thing: Hemispheres. They specifically mention South Africa. It’s summer there, pretty much every country (at least those I looked at) in the world had lower numbers in summer with cases going up in winter.
Reporting from Africa is complete junk unfortunately. Take Africa out and you will have clear signal. Also worth mentioning: population pyramid is really a pyramid in Africa. There are over billion people in Africa but only 75 million of 60 and older.
This is deeper level of stupid. They take data, then throw in random junk and claim there is no signal in data.
When you look at each continent separately, then there is clear pattern except in Africa. It is sad but expected. There are also 3 clear and explainable outliers: Israel, Mongolia and Malaysia. Take them out and Asia has also clear signal.
Subramanian and Kumar discover that COVID cases are not correlated with vaccination rate. But everyone knows that the vaccine doesn't stop you from getting infected; rather, it reduces hospitalization and case severity.
The problem with this paper is not the result, but the authors' tone punching through, which expresses surprise at this result. This paper should have been rejected for revision on that basis.
Wasn't efficacy originally defined as preventing symptomatic covid infections?
> The mRNA-based Pfizer1, 2 and Moderna3 vaccines were shown to have 94–95% efficacy in preventing symptomatic COVID-19, calculated as 100 × (1 minus the attack rate with vaccine divided by the attack rate with placebo). It means that in a population such as the one enrolled in the trials, with a cumulated COVID-19 attack rate over a period of 3 months of about 1% without a vaccine, we would expect roughly 0·05% of vaccinated people would get diseased.
Even if you consider Delta variant, we were originally told the vaccines were very effective:
> The study said the Pfizer vaccine was 88% effective against the B.1.617.2 variant, first found in India, 2 weeks after the second dose. The Pfizer vaccine was 93% effective against the variant found in the U.K., known as B.1.1.7, two weeks after the second dose.
> Two doses of the AstraZeneca vaccine were 60% effective against the B.1.617.2 variant and 66% effective against B.1.1.7, the study said.
It's fine if preliminary results didn't hold up, but lets not gaslight the public into thinking vaccines were never about not getting infected.
i think may be a lot of people assumed initially, that the covid vaccine is gonna be like other vaccines - it prevents the virus from spreading by making it unable to take hold in your body to spread.
Isn't that what the clinical trials are for? I guess 3 months wasn't enough as places like Israel saw their infection rate go down a lot only to shoot back up recently as immunity wore off.
Again, I'm okay with new finding. What I'm not okay with is this constant gaslighting. The media tries to pretend I don't remember what was reported 6 months ago.
Sterilizing immunity is the term for that, and just as with the flu vaccine we cannot yet achieve such an outcome with Covid vaccines. Respiratory viruses seem to be particularly difficult in this respect.
And the fact the mRNA method is not technically a vaccine as such (excepting that they've changed the definition since).
The vaccine was sold as the ticket to freedom with early claims that it prevents spread and death. It does neither, might actually increase risk of more virulent strains longer term and brings a false sense of security to those not informed fully of the ongoing risks.
It will be the 'unvaccinated' - the 5-10% left - somehow still causing all spread and death because that's easier than having to continue to be cautious and responsible after 'taking one for the team' with the jab, or admit the vaccine mostly provides individual risk reduction, not societal.
Many people will still say vaccinated people get negative faster so, even if they were shedding as much virus as unvaccinated people, vaccinated people would still be less contagious.
This seems the basis for a lot of the mandates, if it stops being true then covid vaccination becomes a matter of personal health rather than public health.
Testing viral load is difficult, variance is quite high either way. This study found that during the crucial phase (first few days), the viral load is the same in the vaccinated:
There is still a reduction in the younger age groups, presumably because those have been vaccinated more recently.
> This seems the basis for a lot of the mandates, if it stops being true then covid vaccination becomes a matter of personal health rather than public health.
The mandates are based on assumptions that have been proven wrong, you are now witnessing the inertia of authorities in adjusting policy to new information. Furthermore, blaming the vaccine hesitant for reality not living up to promise is politically expedient.
The question is genuine, what are the assumptions that would support mass vaccination, instead of only people at risk of severe cases, and have been proven wrong?
> Are you implying that shedding does not correlate with being symptomatic?
No, I'm saying being vaccinated correlates with having fewer or no symptoms, which leads to different behavior.
> Your second link does not invalidate the first, and that seems to support vaccinated recover faster, hence they have less time to spread.
The crucial period is the first few days, especially before symptom onset. If you are symptomatic, you are probably at home after this period.
> I've came across the UK weekly reports, there's still much controversy on the population data used in those reports
Good point, I stand corrected. That said, data from Israel also suggests little to no reduction in infections among vaccinated.
> The question is genuine, what are the assumptions that would support mass vaccination, instead of only people at risk of severe cases, and have been proven wrong?
The wrong assumption is that vaccines "protect others" and that they will substantially reduce infections. If they actually did substantially reduce infections, we should observe that effect everywhere - we don't. Whatever reduction there is, it's likely counter-acted by the modified behavior of the people who still rely on these assumptions to be true. Either way, none of it justifies mandatory vaccination, or the scapegoating of the vaccine hesitant.
Good point on the presumably different behavior of people who might underestimate how much at/of a risk they actually are.
And I agree that the weakest the "protect others" is, the least justifiable mandates are. The story is definitely not as straightforward as it's often portrayed.
Thank you for the peaceful exchange, I've very much enjoyed it.
I don't think it's true at all that everyone knows that vaccination doesn't stop you getting infected, especially in the US - there seem to be a lot of people over there that still think herd immunity though vaccination would be a possibility if not for the evil right-wing antivaxxers screwing it up for everyone, and most of the US mainstream media has been continuing to push that narrative. There was also a big pushback against outlets reporting any evidence that vaccines wouldn't be able to achieve this anymore.
Not really - the authorities are driving vaccination in order to reduce propagation, indicating that the vaxx is supposed to materially reduce R0.
The authors are not making any claims about anything, other than demonstrating the rate of vaccination doesn't seem to affect the case rate.
The most obvious thing I see is that places with lower case rates and remote not dense populations tend to be less vaccinated in the first place. I have family in rural areas with almost zero case rates and as a result they're just less concerned and less vaxxed.
This is an important consideration for policy because it's generally not been about making sure workers have protected themselves (although that's one issue) it's really about preventing outbreaks.
“The authors are not making any claims about anything, other than demonstrating the rate of vaccination doesn't seem to affect the case rate.”
Seems like the reader can at least reasonably deduce that the vaccine does not significantly impact the transmission rate.
Mainstream thought and dogma about the vaccine certainly takes a hit based on the observations made in this article. I already see comments dismissing the findings based on personal theories about lockdown impacts and other WAGS.
Yes, they are providing evidence that maybe the vaxx doens't reduce rates - but they are not saying anything about government claims or expectations up until now.
And yes, this data would be damaging to mainstream dogma but it's only one article.
My bet is the scientific reality is somewhere in the middle: vaccines prevent transmission, but not by that much.
Public Health Policy is always going to err on the side of 'more vaxxes'.
If the data really starts to show vaxxes not as useful as we might want for R0 (because one study is just one study), they will change their tune and back off.
There are a lot of reasons this data could be odd, including the fact that spread is happening among children right now, so it could be related to school outbreaks, and, areas where there wasn't a lot of COVID previously.
In the end, their messaging will primarily be driven by net cases and uptick: if we're seeing breakouts and growth of COVID in certain areas, then you'll hear about vaccines, if case rates stay very low, we'll get back to normal.
> COVID cases are not correlated with vaccination rate. But everyone knows that the vaccine doesn't stop you from getting infected;
I don't think that's what everyone does, or should know.
In the strict technical use of the word "stop" you are correct. it is not 100% eliminated.
But this is misleading, as the risk of being infected is much reduced by vaccination, as well as the "hospitalization and case severity".
As others have noted, the behaviours of people also influences the infection rate. Lockdown without any vaccination, or no lockdown and high vaccination percentages might give similar infection rates.
They did not demonstrate nothing, perhaps only that they stupid morons.
If you take out junk data from Africa, you get clear relation for each continent. Take out 3 clear and explainable outliers: Israel, Mongolia and Malaysia and it becomes even more clear.
Here is data have fun (see source for copy paste)
Country Name Location People Fully Vaccinated % of Population Cases Per 1 Million People in the Last 7 Days
Angola Africa 2.71 31.09011888
Argentina South America 32.42 636.3661061
Australia Australia 28.37 300.2921838
Austria Europe 57.65 943.9270195
Azerbaijan Asia 26.72 2268.729292
Bahrain Asia 62.31 339.1876085
Botswana Africa 8.66 1307.336771
Brazil South America 29.68 596.5323021
Bulgaria Europe 17.12 1313.100336
Cambodia South America 50.72 150.5920474
Canada North-America 66.93 536.6461776
Cyprus Europe 58.85 1800.665537
Czechia Europe 53.56 106.7643565
Djibouti Africa 2.37 44.90135173
Egypt Africa 3.08 15.0875239
Georgia Asia 10.6 5408.851208
Greece Europe 55.43 1661.114672
Hong Kong Asia 46.02 4.766444233
Iceland Europe 76.82 1202.819199
India Asia 10.9 182.8242777
Indonesia Asia 13.19 204.0875492
Israel Asia 62.51 6224.276933
Italy Europe 61.04 610.8587852
Jamaica South America 4.88 1408.45923
Japan Asia 46.85 932.0131544
Kazakhstan Asia 28.63 1694.397008
Kenya Africa 1.47 72.96442264
Kyrgyzstan Asia 6.89 160.3718091
Latvia Europe 40.99 760.6053562
Lebanon Asia 16.15 950.4884734
Libya Africa 0.84 1215.772624
Lithuania Europe 56.32 1277.388952
Malawi Africa 2.17 25.09201977
Malaysia Asia 46.5 3788.511754
Mali Africa 0.41 4.411258991
Malta Europe 80.22 577.1876773
Mauritania Africa 0.46 210.2569365
Mauritius Asia 58.81 1577.631401
Mongolia Asia 62.99 4745.768006
Montenegro Europe 29.45 5502.737835
Namibia Africa 4.47 332.7736861
Nepal Asia 15.73 316.0918378
New Zealand Australia 25.5 71.5954806
North Macedonia Europe 25.88 2276.894799
Pakistan Asia 7.99 103.9787184
Panama South America 37.92 748.3596682
Philippines Asia 12.71 943.3400713
Portugal Europe 75.01 1088.816271
Russia Europe 25.53 753.5568248
Sao Tome and Principe South America 5.32 255.1888397
Saudi Arabia Asia 42.66 37.46390845
Senegal Africa 3.29 35.18197046
Slovenia Europe 43.67 1348.424008
Somalia Africa 0.6 27.50695315
South Africa Africa 10.23 869.8911342
South Korea Asia 31.74 199.648441
Sri Lanka Asia 38 1477.394423
Suriname South America 23.13 2316.668863
Sweden Europe 56.97 568.6918876
Taiwan Asia 3.93 2.263675619
Trinidad and Tobago South America 28.51 860.0700882
Turkey Asia 43.94 1399.061291
Ukraine Europe 8.99 299.9989279
United Arab Emirates Asia 76.03 595.8313028
United States North-America 51.91 3039.054339
Uruguay South America 72.13 177.8975494
Vietnam Asia 2.78 820.9530543
Zimbabwe Africa 10.97 64.53677208
Wow this paper is a joke. I see multiple conclusions drawn from insignificant statistical tests (if you can’t reject the null you can only say “i can’t see it”, not “its not there”). That alone should be enough to stop reading. There appears to be other elementary mistakes made as well.
Also, you can’t draw conclusions about vaccination and covid prevalence from country level data. Different countries report their statistics differently, the prevalence of covid depends on attitudes to testing which varies at the local scale let alone at the scale of countries. Do you think poor people want to get tested when they have to self isolate? A country’s wealth matters a lot for measurements of covid prevalence. As far as I can tell there’s no effort to account for this bias, and even if there was, their data is so noisy it’s all non-significant and thus useless anyway.
EDIT: This is bad science and it might be better if it wasn’t on HN.
> the CDC reported an increase from 0.01 to 9% and 0 to 15.1% (between January to May 2021) in the rates of hospitalizations and deaths, respectively, amongst the fully vaccinated [10].
Looking at the CDC slide I think they are referring to (their link is broken), 15.1% was the percentage of people in hospital who were fully vaccinated. It is fully noted on the slide: one would expect this to climb as vaccination level climbed and especially as high risk people were vaccinated first.
So they have provided no basis to care about the pandemic post vaccination.
Obviously every country wants to use its resources to the max and open as much of it's economy back up as soon as hospital occupancy and death are down so rates of infection should go up unless you reach something like 95% recent vaccination or booster.
how did this study end up being high voted on HN. I am very doubtful about mandatory vaccine for everyone, and know vaccines aren't the perfect solutions. However, this study is probably one of the worst i've ever seen. It compares countries being in totally different situations regarding the infection spread. Some are done with delta, others are just starting..
When numbers being looked at are large, they can tell us something even if we admit to them not being perfect in many ways.
At worst, we can figure out trends or, at best, get a negative correlation to learn that something is not positively correlated or even causal.
But everyone reading a statistical analysis, esp on HN, should know it for what it is: it does not account for many things (people have mentioned age, you bring up phase of the pandemic, my thought was how this doesn't account for probability of vaccinated people intermingling more which is related to someone else bringing up people in rural areas...).
The entire study, based on the data it covered, does not proclaim that vaccines are bad (even though anti-vax people will take it as such), it's just that they are at an early phase and might need more development to increase their effectiveness, and that we'll need other solutions in addition to vaccines if we want to get back to any semblance of normalcy.
I think it's of utmost importance to understand who is exactly at risk from a COVID infection (other than "old folk", which is not really saying much) so we can better prepare for dealing with them (us? :) in the future. While this study does not answer this, keeping a critical mind is important in science.
While it is sometimes true that larger sample sizes can overcome many insufficiencies in the data I don’t think that is the case here. The data is still noisy and any attempt at statistics here would result in a non-significant result. The authors appear to be drawing their conclusions based on a lack of statistical significance. This is a basic fallacy, you can’t say something is not there because your data is too crap to see it. There are statistical tests that determine if something is not different from zero, the authors did not use those tests.
I have't looked that deep into results of this particular study (hopefully peer reviews will do that: it would be unmanageable to look at everything) — but that was not the GP's complaint, but rather that conditions of each region have not been taken into account, and I am saying that this is fine if numbers are sufficiently large and statistically relevant.
You may be right that they are not, and hopefully that results in the paper being retracted.
It doesn't appear to account for age groups. Most countries have a much higher vaccination rate for older people which in general are also more likely to get sick.
It also direct contradicts data from Switzerland which clearly show a correlation between vaccination rates and infections when comparing different cantons that have higher rates compared to others.
The entire mainstream narrative and the tactics being adopted to increase vaccinations don't account for age groups either. If you look at the mainstream media, all the comparisons are based on total vaccination rate regardless of age - I know the UK government has been getting a lot of stick for "falling behind" Europe in vaccinations due to lower vaccination rates which are mostly concentrated in younger age groups. A lot of the vaccine mandates primarily target working-age people who generally aren't the highest priority in terms of age. And there's been a huge push to target under-16s for vaccination even though it's dubious that they benefit from it at all, supposedly because it'll stop them from spreading it - we'd expect that to show up in the increase in Covid-19 for sure.
Interesting (and confusing), normal person would expect that linear regression line to have opposite sign on the slope. Aren't they feeding it apples and oranges data? There must be more dimentions to this data.
This does literally nothing to identify and control for other variables that influence the case rate, uses the case rate rather than the hospitalization or death rates (important because that makes it subject to variability in detection of non-severe cases, which may be correlated with vaccine hesistancy because testing hesitancy may be.)
the thing is though, that many municipalities are using vaccination rates as their sole or primary metric for predicting case loads. and if thats the case then this study basically invalidates that metric.
How can they miss the most important variable and that is the ability to freely mingle in society and lockdown measures? It's like making a study about diabetes and a drug and missing to control for diet. The outcome of this faulty study is that vaccines don't work well and that's what we DO NOT want to project. It will become a flag of anti-vaxxers.
This analysis is of change, not of level. Say you have one country with 90% vaccination and one with 9%, and much higher incidence in the latter, but both are stable during the summer. Then the weather grows cold and people move indoors, and the number of infections double in both countries.
This report says that's so: The number of infections double in both countries. It doesn't say "they arrive at the same number" or "this is an equally big problem in both countries" or anything of the sort.
> This report says that's so: The number of infections double in both countries.
Not even that. It says the number of infections is roughly equally likely to increase in both places, but they don't say anything about the magnitude of the increase.
This is so wrong. You can clearly see the reduction in deaths in a graph when you compare UK pre-vaccination (<30% population) and post vaccination (>60% population). Idiots.
Your statement isn't contradicting the study, infections and deaths are different things and the study also mentions that in its interpretation:
>Even though vaccinations offers protection to individuals against severe hospitalization and death, the CDC reported an increase from 0.01 to 9% and 0 to 15.1% (between January to May 2021) in the rates of hospitalizations and deaths, respectively, amongst the fully vaccinated [10].
Any study that helps us understand the pandemic better is always welcomed in my book, however I agree that the study seems to be quite opinionated:
>In summary, even as efforts should be made to encourage populations to get vaccinated it should be done so with humility and respect. Stigmatizing populations can do more harm than good.
I don't see how the study proves that last point, even though I don't necessarily disagree with the statement.
And of course focusing on infection rates is really the wrong thing. We know antibody immunity declines over time, but protection against serious illness stays strong.