It is predicated on some dodgy assumptions — principally, that vaccinated individuals still transmit at the same rate as the unvaccinated, that ‘naturally immune’ individuals somehow by contrast do suppress transmission, and that the natural infection route does not select for mutants at the same (or higher) rate.
The biggest problem — and it’s a general one — is the first. I’m sick of seeing people claim that ‘the vaccine does not stop transmission’. It was never definitively demonstrated anyway, despite being widely circulated as categorical truth, and is a common misunderstanding of the scientifically cautious ‘we don’t know if …’. Well, we now have data. The vaccines have all been shown to substantially reduce transmission. If one wants to be a pedant, one could say ‘reduce’ doesn’t mean ‘stop’, but if fewer individuals shed viable virus in the vaccinated than unvaccinated group, it has clearly STOPped transmission in a greater number of individuals in that group. It’s not that it’s generated the same number of transmitting individuals but with lower levels per individual; there are fewer transmitting individuals. That matters. Those individuals are invisible to the virus. It has the same effect as reducing population density.
Much of what can be said about vaccine can also be said about ‘natural immunity’. Yet one never hears a parallel uncertainty over transmission from previously infected individuals. Well, as it happens, data shows that this too is reduced (or ‘stopped in more individuals’) for reinfection of the original strain. There’s the rub. There are worrying signs that the Brazil mutant can reinfect individuals. If reinfection is occurring, the Brazil mutant is not a ‘vaccine escape mutant’; it is a ‘natural immunity escape’. So the thing he’s worried about actually results from the strategy he advocates.
The key to reducing mutations is to reduce replications. You can create a selective pressure by doing that, but you also reduce the opportunity for mutations that can respond to it. The optimum would be to vaccinate (or infect) everyone in a day! That would knock it on the head for sure (at huge cost of life and health in the second case). Failing that, mass vaccination should proceed as fast as possible. It gives the opportunity of creating hundreds of thousands of ‘pseudo-infections’ per day without the health costs and, crucially, without the replications. It would be neither possible nor desirable to hit the same numbers ‘the natural way’. ‘Letting the disease rip’ creates a substantial excess of replications over vaccination, hence more opportunities for mutation and selection.
This is not really the way to do science: by non-peer-reviewed ‘open letter’, uncritically leapt upon by the vaccine-skeptic community with neither the knowledge base nor the appropriate scientific skepticism needed to evaluate his claims. I’m no expert, but I think his strategy would be disastrous.
Well, explain the trends in countries that have had massive vaccination campaigns using Coronavac. Can you cite a single success story?
https://graphics.reuters.com/world-coronavirus-tracker-and-maps/countries-and-territories/bahrain/
petrushka,
Yes, I can: Serrana, São Paulo.
In Bahrain, I have a nasty suspicion that there is a distinct socio-economic difference between the vaccinated and the unvaccinated, so the ‘homogeneous population’ assumption does not hold. Like I said, be wary of confounding factors. In all these cases you, petrushka, have no idea what these numbers would look like if no vaccination had occurred.
You are indulging in ignorant cherry-picking. Please stop.
The breakthrough rate for the British and American vaccines appears to be about one in ten thousand.
Reading a BBC article, I see that rates improve with very high vaccination rates.
The articles are a bit confusing, because one says 95 percent of adults were vaccinated, and the other says >75 percent. Not incompatible, but why the different numbers?
I doubt if any large nation will achieve >75 percent. Europe can’t manage 50 percent, so far. In the USA, minority communities are the big holdouts. Around thirty percent.
petrushka is quoting, selectively, from this May 5 article.
More recent data has confirmed the real-world effectiveness of the Coronavac vaccine.
The “BBC article” that confused petrushka is probably this one, which, like the AP article, notes that they observed a strong protective effect once they reached 75% vaccination rates. They have since achieved 95%.
It’s really not that confusing.
Please stop posting nonsense that may contribute to vaccine hesitancy.
There’s some seriously bad writing here:
I would point out that UK, with 60 percent having at least one dose, had a zero death day.
A vaccine with a breakthrough rate of 20 percent, or even 5 percent is not going to control a disease. Just slow it down, and produce a population dependent on boosters.
Brazil has ordered a hundred million doses of Pfizer.
India has half the vaccination rate of Brazil, and its numbers are going down. Brazil is constant.
But Brazil is going from the pity capital of the world to the first country to systematically test more than one competing vaccine. So good for them.
Newsweek turns to the dark side:
https://www.newsweek.com/exclusive-how-amateur-sleuths-broke-wuhan-lab-story-embarrassed-media-1596958
Just a little milestone you won’t read about in any newspaper.
And there’s a reason for that: it isn’t really a “milestone”. On June 4th, Florida switched from daily reporting to only providing weekly updates.
In their most recent report (June 3), they reported a one-day tally of between 1,872 and 1,878 new cases for June 2 (sources vary somewhat), and haven’t reported any new cases since then.
So it’s an artefact of the reporting frequency.
If you think that there have been zero new cases in Florida over the last 7 days, I have a bridge I want to sell you. Seriously, please stop it with the misleading tin-foil hat rubbish.
I did find that out, but it isn’t the only weirdness. Deaths in Florida haven’t been tracking the decline in cases for a while.
As for misinformation, I’ll ask you again to explain why entire nations like Chile that have strong vaccination programs, have seen little decline in cases or deaths.
I’m sure there is a seasonal component to the worldwide decline, but it really looks like some of the vaccines are dramatically more effective.
If you “did find that out”, then why on earth would you post such a stupid, misleading comment about Florida reaching a zero case “milestone”? Re Florida deaths, I may have mentioned this before, but death is a lagging indicator.
You never asked a question about Chile, merely indulged in vague “I find this interesting” musings. I did answer your question about Bahrain. Re Chile, does Google not work in Connecticut? This is the definition of cherry-picking.
Nobody has said otherwise. You, on the other hand, claimed that:
only to prove yourself wrong with the data from Serrana
As you noted previously, you are not a wizard at statistics. You are not even competent. Enough with the Q-drops.
An interesting trend among anti-vaxxers online: define the vaccine as ‘not a vaccine’ because ‘it does not stop transmission’. Which is pedantically true if one’s threshold is even a single case in a vaccinated group, but displays a maddening black-and-white obtuseness. If it does not stop it completely, it does not stop it at all? Hogwash. Likewise, condoms do not prevent babies or AIDS.
Allan Miller,
The evidence piles up daily that vaccination is safe and effective.
One of the many bizarre anti-vax tropes: that upsurges in cases are caused by vaccination. That, somehow, blood clots at a rate of 4 in a million are noticed but a direct causal link to mass hospitalisation and death, a problem with a very simple remedy, has somehow passed us by unremarked by the authorities tasked with monitoring these things 🤔. It does make me despair at the decline of critical thinking – although, more likely, the internet simply exposes us to more of the worst of it.
Apparently intramuscular injection technique used to involve drawing back the plunger to see if blood entered the syringe prior to injecting the vaccine, avoiding accidental injection directly into a blood vessel. I’m told this technique is no longer taught.
Meaning, I presume, that it does not stop the entry of the virus into your body, just kills almost all of it after that? It is clear that it has a dramatic effect on your ability to infect someone else, but I suppose they are using the word “stop” to hold it up to a standard of 100.000% prevention.
That’s the thrust of their defence when challenged – to cling to a position which is pedantically true, but thoroughly misleading. They may point to the 20% of cases in the vaccinated group in a double blind trial giving 80% efficacy. They will probably transmit. It’s a continual surprise how concepts of proportion seem so opaque to people. It’s got to be all or nothing. Although, I think it’s more dishonest than that. The scientifically cautious “we don’t know if it stops transmission” (based on original case-counting trials) has morphed into “the scientists say it doesn’t stop transmission”, repeated endlessly.
The thing is, if a massive vaccination effort by a nation has no discernible effect, it undermines confidence in vaccines in general. It has certainly undermined confidence in Coronavac.
https://www.washingtonpost.com/world/asia_pacific/china-coronavirus-singapore-data/2021/03/23/7a0582ca-8afc-11eb-a33e-da28941cb9ac_story.html
https://www.washingtonpost.com/world/2021/04/15/brazil-china-vaccine-coronavirus-coronavac/
Dare I ask what difference the source makes? Everyone can see.
The big success story for Coronavac involves a tiny closed community that was 95 percent vaccinated.
New York State is 52 percent vaccinated, and deaths are down 94 percent from January.
Chile has a higher vaccination rate, and deaths are not down at all.
UK actually has a rising case rate, but a minuscule death rate that is not rising.
Not much is being written about this, but vaccination rates are extremely high in the medically vulnerable population.
Florida, which is not a leader in vaccination, has vaccinated 83 percent of people over 60. Perhaps other states have done better.
In any case, the US had its best week since Worldometers started reporting.
petrushka,
There’s an intersection of takeup and efficacy. If a vaccine if 51% effective with 100% takeup, you still get overwhelmed health services.
The Sinovac is an inactivated whole virus (as I understand). This has potential implications for natural immunity – it is possible that the targeted antibodies of the nucleic acid vaccines actually produce a better response than the broad-surface cocktail the real pathogen produces, adjuvants aside. The reason may be binding competition between the different antibody types, though I may have misheard that.
Well, a town of 46,000 that has returned to normal, whereas neighboring communities are suffering from a nasty surge. It’s the comparison that is key.
Like I said, be wary of confounding factors.
You do not seem to understand what constitutes evidence.
Vaccinations:
Results:
Question:
Has China ever supplied documentation of its clinical trials to back up the claim of 50 percent efficacy?
Are there any peer reviewed studies of the results claimed for the Brazilian town?
Can anyone provide an evidence based argument for the failure to change the curves of cases and deaths in Chile?
Can anyone show any similar failure for the British or American vaccines?
I’ve seen reports that the mRNA vaccines provide a broader immunity than does the disease and recovery. They seem pretty effective against known variants.
One possible argument against a 50 percent vaccine is it could give a false sense of safety, resulting in increased risky behavior. A net counterproductivity.
I’m living in one of the most compliant regions in the country, but when the death toll dropped to one per day, people mostly stopped wearing masks. That has few adverse effects when efficacy is >99 percent.
There are still small crowded stores that require masks, and people comply. It’s almost like they are rational.
Yes, yes, yes, and {yes, but there’s a boring reason for that}.
But why would I bother: you do not understand what constitutes evidence.
Unlike the AZ vaccine against the delta variant, for instance…
You are starting to think like a grown-up. Keep going — you still have a long way to go.
You and I are rather lucky in that regard. Now, try to imagine this: suppose that there were pockets of people elsewhere in the same country, who 1) were not vaccinated and 2) failed to distance or wear masks. What would happen if you AGGREGATED data from these two dissimilar populations? You could have 50% “fully vaccinated” or even achieve an impressive (but hardly airtight) 65% “fully vaccinated” and simultaneously see accelerating case numbers as the virus runs rampant through the unprotected sub-population.
That’s what I reckon happened in Bahrain.
In Chile,
not quite sure how many times I can bear pointing this out.
In earlier posts I linked to comparisons of breakthrough rates.
But I appreciate the adults here who only base their opinions on peer reviewed publications, and not on “that’s what I recon.”
No, you have not. All you have posted was a comparison between an unsourced estimate of breakthrough rates for Brit and US vaccines (=1/10,000 ! Did you divide the number of breakthrough cases by the number of vaccinated individuals, perhaps? Yikes! ) and an anecdotal report of the test positivity rate in Serrana. Your equating a test positivity rate with a breakthrough rate demonstrates that you have no idea whatsoever.
But I suppose that is in line with your rather impressive track record of COVID predictions:
petrushka, April 2020
petrushka, July 2020
So why should we even bother to read what you write? Please just stop it.
I admitted I was wrong about those predictions. They were based on reports at the time that there were as many as nine unreported infections as confirmed.
I’m sure you have never reasoned from wrong assumptions.
Now reason this:
Take two populations, one of which receives a fifty percent effective vaccine, and the other a 95 percent effective vaccine.
Assume both populations are equally exposed to the virus.
What is the expected ratio of cases between the two populations.
petrushka,
Even if we were to accept your ‘equally exposed’ assumption, which is incoherent, there still is not enough information to answer your question. Which you would know if you were numerate. (For instance, think about the effect of R0…)
Your question is incoherent too — you need to specify a timeframe.
For what it’s worth, I’ll take a crack at this one. Let’s start with 100,000 subjects drawn from all demographics. At random, select two groups of 50,000 each. For the control group, use the placebo. For the other, use the 50% effective vaccine. Now, let’s say that of the placebo group, 10% (5000) get the virus, and for the vaccinated group, 2500 get the virus. That’s where the 50% effective comes from – the vaccinated group only had half as many cases.
Now, take it from the top. Same groups of 50,000 people each, one unvaccinated and the other with the 95% effective vaccine. The first group again gets 5000 cases, but the vaccinated group gets only 1/20 as many cases. That’s 250 cases, which is 95% efficacy.
Finally, we compare the 2500 from the 50% case with the 250 from the 95% case. We find that the ratio between the two populations is 10:1.
Permit me to disagree here. To get a RATIO, you don’t need to specify a time frame, you need only specify that the time frame is the same for both groups. And that if we’re doing two trials, one with a 50% and the other with a 95% vaccine, we have four groups and all four must have the same time frame to get the ratio.
In the pfizer trial, none of the subjects knew who had the placebo and who had the real McCoy, but all subjects simply went about their daily lives normally until enough cases had occurred to achieve a sufficient confidence level.
Above, I walked through the trial exactly as it was performed, and I had sufficient information to answer the question.
Oh by the way, another footnote on methodology. Before the trials started, it was necessary to specify exactly what counted as a “case”. Pfizer defined a case as EITHER two indicative symptoms, or one indicative symptom AND a positive test. Notably, this approach misses all asymptomatic cases, which are estimated to be as high as 50%. Not only that, but vaccinated people are far more likely to be asymptomatic, implying that the actual case count of the vaccinated group is probably more of an undercount than of the unvaccinated group.
It has also been noted that in the real world, these mRNA vaccines have proven (so far) to perform better than the trials predicted.
Flint,
No. Your analysis ignores a very important fact: people who are infected expose additional people to the virus.
So long as Re remains greater than one, then (1 – 1/Re) of the population will eventually get infected, but the time taken for this to happen will depend dramatically upon how much above 1 the Re is. If Re is less than 1, it’s gonna die out, perhaps rapidly, without even infecting 10% of the susceptible population. So the answer to petrushka’s question will depend on R0 and, unless you specify “total cumulative to the end of the epidemic”, on the timeframe chosen.
So, not enough information.
Vaccine trials are much simpler to model, since only a small proportion of the population is a trial participant: they can ignore the ‘knock-on’ effect and get a direct read on effectiveness.
It’s true that as cases and recoveries occur, the pool of susceptible people shrinks.
And recovered people are more or less 100 percent vaccinated.
But that’s not going to have a large effect over six months. Vaccinations are far outstripping cases.
Your R0 stuff is irrelevant. It may affect the numbers, but not the ratios.
The ten to one ratio is oversimplified, but it conforms to what is happening in Chile and Mongolia, and countries that predominantly use Sinovac/Coronavac. The effect on national death rates is minimal.
More like 99 percent effective.
Which would make them 50 times as effective as a 50 percent vaccine.
You can quibble all you want about the oversimplification, but the principle is correct. Fifty percent is not half as good as 95 or 99 percent.
A fifty percent effective vaccine is unlikely to have much effect on national numbers in a few months.
Whereas, the mRNA vaccines can bring deaths down to the noise level.
petrushka,
To chip in: R is relevant, and is affected by the vaccine itself. (R0 is the state with no immune individuals, so if you’ve created some you are no longer at R0).
Imagine the extreme situation: you’ve vaccinated 100% of one population with a vaccine that’s 100% effective, and 100% of the other with a vaccine that is completely ineffective. A placebo, effectively. There is no travel. In population 1 there is no virus and R is undefined. In population 2 R is the baseline for that virus with whatever measures are in place (R=R0) – R is different, and so is the case ratio. It starts off infinite, or rather #DIV/0. Let things run for a short enough time that infection immunity has no significant effect.
Now, in the mental model, tweak the figures. Make the vaccine in population 1 incrementally less effective, and likewise tweak population 2 to make it more effective, approaching the values you originally defined. The R values change as you do so, because they are affected by immunity, and so likewise does the ratio. It must, because it started out infinite and will end up 1:1 when all is equal. The cause of that change in ratio differential is the change in differential of individuals to infect.
Assuming finite populations, the number of susceptible people shrinks very quickly in the population having the better vaccine. This tends to increase the disparity.
This seems to be happening in the real world. Regions that reached 50 percent vaccination with British or American vaccines have rapidly declined in cases and deaths. Most are well below last summer’s lull. Some are near zero. The American Northeast is in or approaching single digit death counts.
My reading of Israel and UK numbers is that it takes two or three months after reaching 50 percent vaccinated, to bottom out. It seems likely that herd immunity is the union of recoveries and vaccinations. So the northeast had a head start.
Countries that predominantly used Chines vaccines are trying to find alternatives.
Since I started this rant, the situation in Chile has gotten worse. They have one of the highest vaccination rates in the world , and their death count has increased. Mongolia iis similar.
I haven’t found a source that documents countries that mostly use Chinese vaccines. Countries where the disease is not declining are not reporting their vaccine type.
My primary source is ourworldindata.org Their charts are accepted by many major universities.
The problem with the R0 argument is that infection remains “rare”.
It took a year for 10 percent of Americans to be infected. CDC estimates there are several actual infections for every one reported, so assume it’s 30 percent.
None of this invalidates the ratio argument. It just makes the math more complicated, and the exact ratio more problematic.
But when the simplified ratio is 10 to one or higher, the effect is quite obvious in real life. No one wants to wait while herd immunity is achieved through infection and recovery.
To match the results of 50 percent vaccination with mRNA vaccines, you would need close to 100 percent with the Chinese vaccine.
Did I read that the Brazilian village was 95 percent vaccinated?
In the absence of immunity through recovery, you can’t reach the goal of 75 percent with a 50 percent vaccine.
I realize this is oversimplified, but the death numbers are out there, and they fit my argument.
Um. If our time frame is the duration of the pandemic, you’re right. I should have made it clear that I was talking about the methodology and ratios of the actual trials performed to get Petrushka’s numbers. Those numbers are meaningful only in the context in which they are based.
If we’re talking about much larger test populations and much longer time periods, interdependent variables compound to the point where in practice, your approach verges on useless.
But those trials are where the percentages come from. Your concern with trends in levels of longer term exposure is probably too difficult to reduce to practice. Sure, over the duration of the pandemic, we can trivially say that IF the average number of people infected by those already infected is >1, then we get a chain reaction whose speed depends on the transmissibility of the virus.
And if less than1, the virus will eventually die out.
But out in the real world, we’re dealing with multiple populations with different rates of transmission, multiple virus variants with different rates of transmission, multiple regional use of measures like masks, hand-washing and distancing, major differences in inoculation based on demographs, wide differences in peoples’ innate immune systems, interference with other illnesses, and generally enough variables to confound any detailed study.
I think we can be fairly confident that Petrushka is correct that 95% efficacy is a lot better than twice 50% efficacy. To that extent, the limited Moderna and Pfizer trials are informative.
Technically speaking, we are still in a vaccine trial, since they haven’t been approved.
It’s a rather large and prolonged trial.
Which means DNA_Jock’s downstream rates are meaningful. Regardless of the efficacy of any vaccine, your chances of infection remain a function of exposure to some extent. I don’t know how to calculate what percentage of the population must be vaccinated to guarantee the virus eventually dies out, because what does “the population” actually mean? Your village, your state, your nation? You’d need to know the turnover in one population of people from another.
I don’t think “approval” is relevant here, since this administrative label has no effect on real-world results. Even WITH “approval”, all vaccines are in a state of perpetual trial. Measles was thought to have been completely exterminated in 2000, yet new outbreaks continue to occur in anti-vax communities. And the MMR vaccine is said to be 98% efficient. And almost everyone gets it. Yet there are still outbreaks.
Flint,
Yes, quite.
First the easy point: if petrushka’s “argument”[?!?] is merely that a 95% effective vaccine is superior to a 50% effective vaccine, then, err, sure… No one has ever suggested otherwise. Evah. However, he wrote a great deal of stupendously wrong rubbish, claiming that the “Chinese vaccine” (which appears to refer to Sinovac, but being specific is not his strong suit) had no effect whatsoever.
There is a certain category of people who glom onto narratives that vilify the Chinese and downplay the importance of public health measures such as mask-wearing. No idea why. Insofar as anyone believes a word they say, they are detrimental to public health.
Regarding your “out in the real world…there’s enough variables to confound any detailed study”, you make a series of accurate observations. I agree it is almost impossible to model the complex interactions of the various variants and the convoluted demography. Interesting thing is, there are pharmaceutical companies who are willing to pay large sums of money for the best forecast modeling money can buy, and there are quite a few extremely smart statistics and modeling wiz-kids who are currently working full-time to try their best to address this almost intractable minefield.
Two of them work for me.
I’m not keen on this, which is straight from Anti Vax Central. “Still in trials! Not finishing til 2023!”. One could argue that everything is an experiment, including letting the virus rip with no measures at all.
The trials in which placebos are given and no-one knows who had which are over. Subjects continue to be monitored, as is proper, and there is close observation for adverse effects. It is, I suppose, a Phase IV trial. But putting it thus carries a lot of baggage (ironically, from people who would shove chloroquine, zinc, ivermectin and vitamin D down everyone’s necks without further ado).
As soon as you get in the real world, the many biases they try so assiduously to avoid in trials immediately rear their ugly heads.
Does too. If the things the ratio depends on are dynamic, the ratio’s dynamic.