serious question

[SPUTTER]

This is just getting ridiculous. Your vaccine doesn't work but that's a good thing!

[Sousuke]

This is just getting ridiculous. Your vaccine doesn't work but that's a good thing!

[ATTACH]1025253[/ATTACH]

You probably want to read my post above. I'll add one thing. Studies have shown that 1 dose of the Oxford vaccine isn't very effective against the delta variant.

[SPUTTER]

You probably want to read my post above. I'll add one thing. Studies have shown that 1 dose of the Oxford vaccine isn't very effective against the delta variant.

And apparently the Pfizer shot isn't effective either according to the reporting coming out of Israel.

[as_rocketman]

And apparently the Pfizer shot isn't effective either according to the reporting coming out of Israel.

In this thread, we learn who doesn't understand statistics.

I already linked to evidence that Pfizer is protective against the "delta" strain, at least to the resolution of our best information in hand, and at comparable efficacy to the baseline strain.

When considering England, recall that their most common vaccine is AstraZeneca and rollout has not been perfect there, either.
ETA: It is surprisingly hard to verify the relative distribution of different vaccines in the UK. Let me know if you have a good source.

[Libertarian777]

There are many studies underway, and the statistics vary in quality. The very best controlled studies are whole-nation studies and the extended Phase III trails -- basically everyone who participated in the Phase III is still being monitored, and over time the number of exposures, infections, and hospitalizations is increasing which gives us better numbers to work with.

A few examples:

Post-Vaccination results in Israel (Nature)

Vaccination and transmission study in England (preprint)

Extended Phase III results: Survey of both Pfizer and Moderna (CDC report; feel free to go to sources at bottom if you want to see the original studies, but you'll have to synthesize results of both vaccines for yourself in that case)

We are still learning and the actual performance will jump around a bit before we come up with final numbers, but thus far all actual evidence is strongly positive.

We do have to keep ARR in context.

I analogize, if there was a cure for a disease that is 100% effective, but kills 1 in 10M (very low rate), but there was only ever 1 person who had the disease, would you use the cure on the entire world population (7.8Bn or 780 deaths)? This is the extreme end of the example

Yes SARS Cov-2 can have an R0 as high as 3, however, as all pandemics have shown it never gets to 100% of the population, so we cannot say the entire population is at risk.

Not to mention the estimates of prevalence was thought to be a lot higher (up to 6x ->https://eglobaltravelmedia.com.au/ni...andemic-cases/) early on, which means a large chunk of the population is likely already immune.

The issue with the links you've put above is its still showing relative risk reduction. I couldn't find any data in those studies that follow the placebo control group (https://www.npr.org/sections/health-...getting-immuni).

The nature article you reference has so many caveats which are quickly dismissed (note how its not held to the same double blind RCT study that other therapeutics need to be subject to).

To further this even more, VAERS has (self-)reported 6,000 deaths out of 150M vaccinated with 2 doses.
Children under 17 have had about 164 deaths in 2020 from Covid (about 1/4 that of drownings), out of a population of 70M children. (Issue obviously being we have no way to determine how many of the 70M were exposed or contracted the virus).

Just using a straight ratio as a comparison, if the death rate was similar across the population we'd be exposing children to a 73x higher risk by vaccinating them than not vaccinating them. Yes its not a perfect statistic, but it gives the bounds of the upper end of the risk.

So while the vaccines are showing efficacy, the questions are whether the risk outweighs the rewards for populations under 65 years of age. Say a 65 year old has the vaccine and it turns out 20 years later that the vaccines cause certain issues (clotting, chronic myocarditis etc), for a 65 year with an expected life expectancy of 15 years, this doesn't matter.

For a 5 year old child (with almost no risk of death from covid), there's a potential for them to have chronic issues from 25 through to 80.

Now look at India, in 1 month, from May to June, with the much feared 'Delta' variant, their case and death rates have dropped. Their vaccination rate as of today is 3.7% so that cannot be the driving factor behind the case drop.

Lastly, the infection and death rate in the USA dropped last summer. Yes rates are substantially lower now due to vaccination rates, but they're also lower because testing rates have dropped dramatically. Not to mention the number of people who have recovered from covid naturally has also grown exponentially from last summer.

My point is (as the ARR vs RRR discussion shows), vaccines aren't the be all end-all and they are not without risk (especially at population scale), and they also aren't necessary for the entire population (especially children under 17).

[five.five-six]

Assuming their vaccination rate is about 60% and the the vaccinated make up 60% of the china virus deaths, at first glance appears that the vaccine is 0% effective however the demographics of the vaccinated would tend to be the more vulnerable of the population and the unvaccinated would tend to the les vulnerable so it seems that the vax is at least marginally effective.

This is just getting ridiculous. Your vaccine doesn't work but that's a good thing!

[ATTACH]1025253[/ATTACH]

[as_rocketman]

We do have to keep ARR in context.

We should, but unfortunately there is much here that you've misunderstood. You are in good company in this respect. Let me unpack your comments first and then address this properly.

Yes SARS Cov-2 can have an R0 as high as 3, however, as all pandemics have shown it never gets to 100% of the population, so we cannot say the entire population is at risk.

That is off the low end of the scale. More adequate estimates of the basic replication constant put it a lot higher, even in modern and reasonably hygenic countries, and before we consider the possibility of more efficient variants.

Not to mention the estimates of prevalence was thought to be a lot higher (up to 6x ->https://eglobaltravelmedia.com.au/ni...andemic-cases/) early on, which means a large chunk of the population is likely already immune.

There have been specious arguments to phantom cases since before the first lockdowns. The study you cited (better link) cannot be applied in the way you suggest -- the underreporting in this country was a transient phenomenon, not a persistent undercounting mechanism.

It may be that there is a larger pool of natively immune individuals than we suspect, but this is kinda wishful thinking. Many predictions have been made with this factored in and they've all failed to match reality.

The issue with the links you've put above is its still showing relative risk reduction. I couldn't find any data in those studies that follow the placebo control group (https://www.npr.org/sections/health-...getting-immuni).

The fact that they follow relative risk reduction is a feature, not a bug. To consider "absolute risk reduction" you have to introduce a new parameter, one that you cannot directly sense and is up for interpretation. Also, the Phase III trials do indeed follow the placebo group -- you just don't see it because it's a proper RCT and does not single out that group. That's the way studies should be done whenever possible.

The nature article you reference has so many caveats which are quickly dismissed (note how its not held to the same double blind RCT study that other therapeutics need to be subject to).

Wrong. The Nature article is a whole population study. This is the other way to ensure lack of selection bias: You select everyone. Your objection is simply confused. This is a good result.

To further this even more, VAERS has (self-)reported 6,000 deaths out of 150M vaccinated with 2 doses.
Children under 17 have had about 164 deaths in 2020 from Covid (about 1/4 that of drownings), out of a population of 70M children. (Issue obviously being we have no way to determine how many of the 70M were exposed or contracted the virus).

If you sense me twitching through the screen, it's because this is practically a trope. You are among very many who do not understand VAERS. As I've been explaining for months, VAERS in the case of this vaccine is being held to a much higher reporting standard than usual, and this deliberately results in many, many false positives.

There is absolutely no support for claims that vaccination caused 6,000 deaths. A more careful accounting is important, because the number is not zero, but it is at least an order of magnitude lower. Perhaps two.

Without understanding this, you cannot possibly make a fair trade between vaccine risks and return.

Now look at India, in 1 month, from May to June, with the much feared 'Delta' variant, their case and death rates have dropped. Their vaccination rate as of today is 3.7% so that cannot be the driving factor behind the case drop.

I agree with this. I've commented on India several times.

While we're thinking of India, do not be fooled into thinking India is now on a monotonic decrease in new cases. There is a high probability of another spike, possibly even a bigger one, in the future. The reason we even call it a spike is because some guessed incorrectly that the Summer 2020 bump was the worst of it.

Lastly, the infection and death rate in the USA dropped last summer. Yes rates are substantially lower now due to vaccination rates, but they're also lower because testing rates have dropped dramatically. Not to mention the number of people who have recovered from covid naturally has also grown exponentially from last summer.

No, testing rates are not well correlated with infection rates. I also don't follow your use of "exponentially" in the above. Infection plus good recovery does lead to durable immunity in the vast majority of cases, but you don't have exponential growth in natural immunity without exponential growth in infections.

My point is (as the ARR vs RRR discussion shows), vaccines aren't the be all end-all and they are not without risk (especially at population scale), and they also aren't necessary for the entire population (especially children under 17).

So now let's finally talk about ARR vs. RRR. Many people who have never heard of these things before a week or two ago are now bandying them about without the least sense of what it means. I will blame that specific letter for popularizing this mistake, although it's not the letter's fault. Instead, this is undoubtedly caused by someone misconstruing its meaning, and then being copied uncritically over and over again. This is how we got the confusion with VAERS, for instance, and it too shows no signs of abating. Happens all the time.

This is a tempest in a teapot. The letter IN NO WAY disputes the efficacy results of Phase III trials. Those experiments are very clean. You can take to the bank the conclusion that proper mRNA vaccination will leave you with 91.3 +/- 2% chance of sterilizing immunity against wild-type SARS-CoV-2. The letter in question has no problem with this result.

What the letter is doing, in reality (subscriber, please pay attention), is proposing a different method of comparing two different efficacy results in a way designed to reduce experimental error.

The nature of the Phase III trials leaves us with a counting experiment. That experiment is very powerful in that it allows us to evaluate vaccine performance with only a relative handful of infections (order of hundreds), but when we then release the vaccine into a population without doing an RCT -- without tracking everyone individually -- the methods of sensing vaccine impact on performance in the population are different. It takes some work to compare the Relative Risk Reduction or the true vaccine efficacy against the realized efficacy in a population. To make that comparison, again, you need to sense or estimate or somehow assume an attack rate -- and that introduces bias.

What the letter proposes is to price in that bias by assuming an attack rate and using it to compute a different metric right off the bat. This is the Absolute Risk Reduction they're referring to. In no way does this calculation change our understanding of vaccine efficacy. All it does is report the same information in a different way.

This inclusion of bias can be done for the Phase III trials because, again, we have a fully monitored RCT in place. We know the attack rate in the study population, during the study period, so we don't actually introduce any new uncertainty. But there's a problem when we try to apply this number in the future: The assumption, particularly if unsophisticated readers attempt to use this measure, is that the attack rate in the wild is the same as it was for the trial.

This is a garbage assumption. The extended Phase III trials overlapped the decline in infections in early 2021. That situation may be a good fit to some future situation, and it may not.

This paper, of course, does not assume that the ARR value will be cited and used uncritically by people who don't even know the attack rate. The authors assume ARR will be corrected and used properly. That's fine, and it can be, but it won't be used that way on discussion forums.

Personally, I am not impressed by this paper. If you truly want to be free of bias, and you want to include the attack rate in the reporting, all you have to do is actually include the raw counts. In the case of Pfizer, you simply list the N (44,000) and Pvac / Pplacebo (77:850 as of 1 April 2021). Those numbers can give you the efficacy, RRR, ARR, and attack rate, whatever you want. Nothing is hidden.

The other thing the letter suggests using ARR for is to compare disparate vaccines against each other. Claimed is that there would be more noise in the efficacy value based on different approaches to placebos, testing error, and trials being held in different countries where the attack rate would be inconsistent. It's fair, I guess, but my preferred method above addresses this too.

So to your original point: Should we consider vaccination in a risk vs. reward sense? Sure.

A funny thing happened with these vaccines, though. When originally conceived, it was expected that their performance would be somewhat poor -- say, in the 70% range. Any number of individuals here berated me for even thinking a vaccine was possible. With statistics like this, the effect of vaccination is not really notable on an individual basis, but only when applied across a large swath of the population. This serves to knock down the effective replication constant, but your own individual outcome is much more random.

But as it turned out, that's not what we got. The better vaccines are in fact so effective that the choice to be vaccinated or not reduces the balance of interests question to an individual level. You have much more control over your own fate with a 95% effective vaccine than an 66% one. Thus, all the tortured arguments as everyone attempts to make this calculus with no previous exposure to medical statistics.

I have no problem with everyone making their own value calculation. All I ask is that you try to get good numbers, if indeed you will be swayed by numbers. They are not easy to find. Your post has extreme biases in it that would make a fair comparison impossible. And this letter, that many have used as an argument against vaccination, really doesn't do that at all.

[SAN compnerd]

You asked a serious question here, and you didn't get a serious answer... shocking.

The best way to answer this is to understand that it's a quantitative argument. You cannot say someone is / isn't a threat. The question is in the degree of threat.

Unvaccinated people will be at greater risk of several things, including:

• Catching the disease in the first place,
• Transmitting the disease to other unvaccinated people, and
• Transmitting the disease to vaccinated people.

Let's assume we're talking about top-tier vaccines, not any Chinese garbage. Recall that even these vaccines are not 100%, and this is true in several dimensions. Not everyone vaccinated will develop any protective response at all, but this number is thought to be < 1%. Our best estimates are that vaccinated individuals are infected at less than 10% the rate of unvaccinated, and among those, about 5% develop severe symptoms. Likewise, vaccinated individuals who are infected appear to be about 80% less likely to transmit the disease.

So once vaccinated, your overall risk is pretty low. However, your risk would be even lower if you reduce the number of exposures. In this sense, you would be at greater risk around unvaccinated people as compared to vaccinated people, all else being equal.

Does this constitute any significant risk? Judgment call. I am vaccinated and I don't have any real issue being around those who aren't, nor do I even ask. However, my heavily immuno-compromised father has a different view; he too is vaccinated but he wants to remain clear of those who aren't. It is hard to estimate what additional benefit he would gain from this behavior, but it is his choice.

Surprised you left out one of the options, vaccinated people spreading the disease to other vaccinated/unvaccinated.

Why leave that part out of your explaination?

[Libertarian777]

There have been specious arguments to phantom cases since before the first lockdowns. The study you cited (better link) cannot be applied in the way you suggest -- the underreporting in this country was a transient phenomenon, not a persistent undercounting mechanism.

It may be that there is a larger pool of natively immune individuals than we suspect, but this is kinda wishful thinking. Many predictions have been made with this factored in and they've all failed to match reality.

I haven't found an authoritive source on infections in the USA. Worldometer shows roughly 30M cases. This would imply 10% Assume, as you state, it (undercounting) was a transient phenomenon, then that's a good thing. It means the 10% infection rate is solid. And 90% of folks don't have the virus.

The fact that they follow relative risk reduction is a feature, not a bug. To consider "absolute risk reduction" you have to introduce a new parameter, one that you cannot directly sense and is up for interpretation. Also, the Phase III trials do indeed follow the placebo group -- you just don't see it because it's a proper RCT and does not single out that group. That's the way studies should be done whenever possible.

You don't seem to address the fact that folks IN the control group got vaccinated. How does this allow the trial to continue (for long term research)?

Wrong. The Nature article is a whole population study. This is the other way to ensure lack of selection bias: You select everyone. Your objection is simply confused. This is a good result.

From the Nature article:
Our study had several limitations. First, it was an observational study as opposed to a randomized clinical trial, and, therefore, causal effects are difficult to infer. Second, the comparison between the second and third lockdown might be influenced by factors such as the total number of COVID-19 cases in the beginning of each lockdown, testing policy, hospitalization policy and public compliance with the restrictions that could have changed with time. Similarly, differences among cities might be influenced by behavioral and social differences beyond availability of vaccines. However, none of these factors was likely to cause the different patterns observed in the different age groups reported here..

Not sure what selection bias has to do with anything, i never mentioned anything about that. i said the Nature article mentions confounding factors then dismisses it without expanding on it.

If you sense me twitching through the screen, it's because this is practically a trope. You are among very many who do not understand VAERS. As I've been explaining for months, VAERS in the case of this vaccine is being held to a much higher reporting standard than usual, and this deliberately results in many, many false positives.

There is absolutely no support for claims that vaccination caused 6,000 deaths. A more careful accounting is important, because the number is not zero, but it is at least an order of magnitude lower. Perhaps two.

Without understanding this, you cannot possibly make a fair trade between vaccine risks and return.

Twitch away.
Your 'held to a higher standard' applies to.... healthcare providers.

https://vaers.hhs.gov/data.html

"Healthcare providers, vaccine manufacturers, and the public can submit reports to the system. "
The reporting period tends to be very short, about 7 days (42 for GB syndrome)

https://vaers.hhs.gov/docs/VAERS_Table_of_Reportable_Events_Following_Vaccination.pdf

What percentage of VAERS is reported by the public vs healthcare providers, and what about adverse events more than 7 or 30 days beyond (2nd) vaccination?

Ok then. 6,000 people died WITH the vaccine. Much like we can say 600,000 people died WITH Covid.

(now I know you're going to respond and claim I'm saying no one died from Covid... which i'm not saying...)

I agree with this. I've commented on India several times.

While we're thinking of India, do not be fooled into thinking India is now on a monotonic decrease in new cases. There is a high probability of another spike, possibly even a bigger one, in the future. The reason we even call it a spike is because some guessed incorrectly that the Summer 2020 bump was the worst of it.

Well then is the Delta variant more lethal or not? Or is there more hype than not in the media? What happened to the South African variant we were being told was the end of civilization (ok in case you don't get it.. i'm being hyperbolic).

My point is we're being told the Delta variant is going to end us all now, but even without vaccines India has got it under control, so maybe we won't all die from it and maybe we won't need 3 shots and annual boosters and Epsilon variant shots in December in anticipation of the Gamma variant.

No, testing rates are not well correlated with infection rates. I also don't follow your use of "exponentially" in the above. Infection plus good recovery does lead to durable immunity in the vast majority of cases, but you don't have exponential growth in natural immunity without exponential growth in infections.

I don't know what to say. if you don't test how do you measure infection rates.

So now let's finally talk about ARR vs. RRR. Many people who have never heard of these things before a week or two ago are now bandying them about without the least sense of what it means. ...

This is a tempest in a teapot. The letter IN NO WAY disputes the efficacy results of Phase III trials.

And I never said it did. Did I mention the vaccine wasn't effective anywhere? No. So while you lambast others for not reading...

Those experiments are very clean. You can take to the bank the conclusion that proper mRNA vaccination will leave you with 91.3 +/- 2% chance of sterilizing immunity against wild-type SARS-CoV-2. The letter in question has no problem with this result.

BUT what about the Delta/South African/ Epsilon / Gamma / B1235 variant that we're all going to die from?
(yes this is hyperbole again).

What the letter is doing, in reality (subscriber, please pay attention), is proposing a different method of comparing two different efficacy results in a way designed to reduce experimental error.

The nature of the Phase III trials leaves us with a counting experiment. That experiment is very powerful in that it allows us to evaluate vaccine performance with only a relative handful of infections (order of hundreds), but when we then release the vaccine into a population without doing an RCT -- without tracking everyone individually -- the methods of sensing vaccine impact on performance in the population are different. It takes some work to compare the Relative Risk Reduction or the true vaccine efficacy against the realized efficacy in a population. To make that comparison, again, you need to sense or estimate or somehow assume an attack rate -- and that introduces bias.

Uhnn yes, that's what I've been saying. At population scale its different. If as we said earlier the infection rates are accurately reported, a 95% vaccine efficacy rate doesn't mean applying it to the entire population would protect 95% of people, since only 10% of the population have (or had it) it (prevalence rate).
Ie. forcing all men in the country to have mastectomies will not reduce breast cancer deaths, even if mastectomies were 100% effective against breast cancer.

And as this paper says... you have to vaccinate 217 people to prevent 1 additional infection.

What the letter proposes is to price in that bias by assuming an attack rate and using it to compute a different metric right off the bat. This is the Absolute Risk Reduction they're referring to. In no way does this calculation change our understanding of vaccine efficacy. All it does is report the same information in a different way.

But you just mentioned earlier the infection rates (even with lower testing) are not undercounted and are reliable. so we should know the attack rate.

This inclusion of bias can be done for the Phase III trials because, again, we have a fully monitored RCT in place. We know the attack rate in the study population, during the study period, so we don't actually introduce any new uncertainty. But there's a problem when we try to apply this number in the future: The assumption, particularly if unsophisticated readers attempt to use this measure, is that the attack rate in the wild is the same as it was for the trial.

This is a garbage assumption. The extended Phase III trials overlapped the decline in infections in early 2021. That situation may be a good fit to some future situation, and it may not.

This paper, of course, does not assume that the ARR value will be cited and used uncritically by people who don't even know the attack rate. The authors assume ARR will be corrected and used properly. That's fine, and it can be, but it won't be used that way on discussion forums.

Well this is always the issue, even in Bayesian statistics you have to assume a prevalence rate (even with historical data you have to assume it is constant).

Nothing i've read of what you've said so far disabuses me from what I said. The vaccine is ~94% effective. But again, the question is whether it needs to be applied to the ENTIRE population and what the ARR would be. Again, lets not force all men to get masectomies. The ARR for that is 0.

Personally, I am not impressed by this paper. If you truly want to be free of bias, and you want to include the attack rate in the reporting, all you have to do is actually include the raw counts. In the case of Pfizer, you simply list the N (44,000) and Pvac / Pplacebo (77:850 as of 1 April 2021). Those numbers can give you the efficacy, RRR, ARR, and attack rate, whatever you want. Nothing is hidden.

not sure what the 77:850 is? Probably shouldn't reference CG but these numbers show 8:162? as does the published paper (https://www.nejm.org/doi/full/10.1056/NEJMoa2034577)

[ATTACH]1025207[/ATTACH]

The other thing the letter suggests using ARR for is to compare disparate vaccines against each other. Claimed is that there would be more noise in the efficacy value based on different approaches to placebos, testing error, and trials being held in different countries where the attack rate would be inconsistent. It's fair, I guess, but my preferred method above addresses this too.

So to your original point: Should we consider vaccination in a risk vs. reward sense? Sure.

That's centrally my point... so we agree

A funny thing happened with these vaccines, though. When originally conceived, it was expected that their performance would be somewhat poor -- say, in the 70% range. Any number of individuals here berated me for even thinking a vaccine was possible. With statistics like this, the effect of vaccination is not really notable on an individual basis, but only when applied across a large swath of the population. This serves to knock down the effective replication constant, but your own individual outcome is much more random.

But as it turned out, that's not what we got. The better vaccines are in fact so effective that the choice to be vaccinated or not reduces the balance of interests question to an individual level. You have much more control over your own fate with a 95% effective vaccine than an 66% one. Thus, all the tortured arguments as everyone attempts to make this calculus with no previous exposure to medical statistics.

I have no problem with everyone making their own value calculation. All I ask is that you try to get good numbers, if indeed you will be swayed by numbers. They are not easy to find. Your post has extreme biases in it that would make a fair comparison impossible. And this letter, that many have used as an argument against vaccination, really doesn't do that at all.

Again you haven't addressed children under 17 or the long term risk factor.

Also regarding Pfizer, the primary endpoint was prevention of Covid-19 (disease), not contracting SARS-Cov-2. This is the one thing I think a lot of folks don't understand about the vaccine, it wasn't touted as preventing infection, just disease onset.

I'm not against these vaccines, but as they are EUA and don't have long term studies (like have we even had 1 woman get pregnant post vaccine and give birth yet? we're not 9 months into this yet so we have no data; and no i'm not suggesting the child will have 4 arms and 10 eyes... i'm saying we have no data).
So for EUA absolutely vaccinate >65 immediately with any available vaccine.
For 50-65... it should be recommended but on an individual basis
under 50 we can wait
under 17... not needed.

Note i'm specifically talking about mRNA vaccines.
Pfizer said their long term study is for 2 years (https://www.pfizer.com/science/coron...landmark-trial). We haven't reached 2 years.

Novavax's approach is non-replicating, JnJ/AZ are adenovirus.
I haven't yet seen any progress on an inactivated virus version, that would be the gold standard.

Lastly the major issue I have with all of this is the approach of scaring the sh|t out of the population. I still see people (likely vaccinated) driving in their car with a mask, or running outside in the sun by themselves with a mask, which indicates they have no concept of understanding risk.

[SAN compnerd]

OP, I understand you asked your question in all seriousness, but the problem here is that the people in control of this pandemic (if you can call it control) have gone to great lengths to obscure the answer to your question, media and big tech censor, once respected medical journals publish 'peer reviewed' complete crap, discussion of this very topic will get you booted from most social media, and it's clear that from the very begging we have been lied to about many aspects of this pandemic for political gain.

I feel this is all very obvious to the point of ridiculousness so I did not point it out earlier, but it's the elephant in the room here, many want to point to the 'science', but it's clear that the science has been manipulated and fraudulent in many cases.

The desire by those in charge to maintain tyrannical control via fear mongering and medical tyranny is so blatantly obvious that one should not really even need to mention it.

There are some here that refuse to acknowledge this simple and obvious truth. There are others here that are so fed up with the BS that they probably thought you were trolling with your question.

The answer to your question is intentionally being obscured because those in charge want to continue to exercise the tyrannical power they seized when the pandemic started.

Many here still refuse to admit the truth. The fact is that most of what's happened in the past 18 months was about politics and political power, not a deadly viral pandemic, that part was kabuki theater.

[as_rocketman]

I haven't found an authoritive source on infections in the USA. Worldometer shows roughly 30M cases. This would imply 10% Assume, as you state, it (undercounting) was a transient phenomenon, then that's a good thing. It means the 10% infection rate is solid. And 90% of folks don't have the virus.

Getting a firm estimate of how many infections there were is difficult. I just want to make sure we are explicit about our assumptions if we decide to throw out the numbers we do have.

You don't seem to address the fact that folks IN the control group got vaccinated. How does this allow the trial to continue (for long term research)?

Not understanding your point. The control group is the placebo group. The Phase III trial is intended to continue intense surveillance for two years, after which a different Phase IV will emerge, possibly just a continuation of the Phase III.

Not sure what selection bias has to do with anything, i never mentioned anything about that. i said the Nature article mentions confounding factors then dismisses it without expanding on it.

What you said is that it's not as tight as an RCT. The purpose of an RCT is to eliminate selection bias. You did indirectly mention this.

You are welcome to challenge the Nature paper's assertion that those confounds are minor, if you disagree with them. I think they're fine so long as we don't try to read too much into the results.

Twitch away.
Your 'held to a higher standard' applies to.... healthcare providers.

https://vaers.hhs.gov/data.html

"Healthcare providers, vaccine manufacturers, and the public can submit reports to the system. "
The reporting period tends to be very short, about 7 days (42 for GB syndrome)

https://vaers.hhs.gov/docs/VAERS_Table_of_Reportable_Events_Following_Vaccination.pdf

What percentage of VAERS is reported by the public vs healthcare providers, and what about adverse events more than 7 or 30 days beyond (2nd) vaccination?

I'm twitching. "Tends to be" is not useful here. In the case of these vaccines, virtually any severe symptom will land the sufferer in hospital, at which point their vaccination history will be discovered, and the health care provider must log the incident. That's where the overcount comes from. It's designed that way to make sure they catch all of the outlier cases, even if no connection to vaccination is suspected at all. Normally, VAERS doesn't work that way, but for COVID vaccines it does -- they changed the rules specifically to find even incredibly rare interactions. Likewise, there is no 7-30 day timeout under the current rules.

Ok then. 6,000 people died WITH the vaccine. Much like we can say 600,000 people died WITH Covid.

(now I know you're going to respond and claim I'm saying no one died from Covid... which i'm not saying...)

No, I'm not going to say that. I'm just saying you cannot use unfiltered VAERS results. All of those cases are reviewed in detail to see if a possible causal link exists, and that will give you a much lower count. That's the number you need to use in your cost/benefit equation, not 6,000 fatalities.

Well then is the Delta variant more lethal or not? Or is there more hype than not in the media? What happened to the South African variant we were being told was the end of civilization (ok in case you don't get it.. i'm being hyperbolic).

I think the variants are mostly hype. There is some evolution and they may be more virulent or even more lethal, but evidence of either is thin on the ground. I believe the papers don't understand this. I also believe certain agencies want to take credit for good news and blame bad news on factors outside their control, and this puts unwarranted focus on the "variants."

But we should try not to make decisions based on belief. I've looked for evidence of higher danger from variants since the beginning... and it's not there. CFR hasn't changed, pathology hasn't changed, there's been no correlation between variant emergence and population spread. The only signal worth reporting on is the genetic prevalance as a function of time, and this is what the newspapers use to claim new strains are "way more transmissible," etc., but that's hardly a definite conclusion.

If the new variants are more dangerous, we should see evidence of it. I haven't seen it, and I've looked. Please share if you find it.

I don't know what to say. if you don't test how do you measure infection rates.

Undistributed middle fallacy -- it's not about whether you test or not, it's whether more testing means more cases, which is what you said. Provided your test positivity rate is in the sensitive band, there should be very little correlation. In this country we never exceeded our testing capacity, except during the brief period you already explored, although places like India have done this occasionally.

And I never said it did. Did I mention the vaccine wasn't effective anywhere? No. So while you lambast others for not reading...

My point is that a fair consideration of ARR really shouldn't change your opinion. If it does, it means you've miscalculated somewhere.

BUT what about the Delta/South African/ Epsilon / Gamma / B1235 variant that we're all going to die from?
(yes this is hyperbole again).

I agree with you that this is unwarranted alarmism.

But you just mentioned earlier the infection rates (even with lower testing) are not undercounted and are reliable. so we should know the attack rate.

We do. But that's hindsight. You have to assume a future attack rate in order to drive a benefits calculation and select vaccination policy, and that opens a big political can of worms. I believe you understand this:

Well this is always the issue, even in Bayesian statistics you have to assume a prevalence rate (even with historical data you have to assume it is constant).

Right.

of the N=44K, what's the control group size?
How many actually contracted the virus?
wasn't sure is your 77:850 the hospitalization rate or infection rate?

Pfizer N = 44,000. Placebo group is 22,000. 927 contracted the virus on or before 1 April 2021, and not all of them were hospitalized. There is a different set of statistics for protective immunity, I just used sterilizing as an example.

Again you haven't addressed children under 17 or the long term risk factor.

Long term risk is hard to quantify. Here I will only argue that historically vaccine problems have shown up in the short term, and with respect to mRNA vaccines, the mechanism is not spooky and does not suggest any particular phenomenological hazards. You may disagree, of course, but we will have a hard time coming up with an objective estimate.

Regarding vaccination under 17, you're right, that wasn't part of the discussion yet. I've noted elsewhere that I do not yet see a convincing case for vaccinating children. I need more data. I may get that data, I don't know yet. The data in hand was enough for me to make my own trade, but I'm not under 17.

I'm not against these vaccines, but as they are EUA and don't have long term studies (like have we even had 1 woman get pregnant post vaccine and give birth yet? we're not 9 months into this yet so we have no data; and no i'm not suggesting the child will have 4 arms and 10 eyes... i'm saying we have no data).
So for EUA absolutely vaccinate >65 immediately with any available vaccine.
For 50-65... it should be recommended but on an individual basis
under 50 we can wait
under 17... not needed.

Again, long-term risk is hard to assess. I differ slightly with your recommendation but it is subjective. I would also suggest you reassess your conclusion in light of a better understanding with regard to VAERS, as the risk is probably a lot lower than you thought.

Note i'm specifically talking about mRNA vaccines. Novavax's approach is non-replicating, JnJ/AZ are adenovirus.
I haven't yet seen any progress on an inactivated virus version, that would be the gold standard.

Quite the opposite, in fact. Sinovac's Coronavac and Sinopharm's BBIBP-CorV, the Chinese entries, are inactivated virus types -- and they suck. Not sure why that is, but it is.

There are numerous candidates of all types but I have yet to read up on an inactivated virus candidate that is at all promising.

Lastly the major issue I have with all of this is the approach of scaring the sh|t out of the population. I still see people (likely vaccinated) driving in their car with a mask, or running outside in the sun by themselves with a mask, which indicates they have no concept of understanding risk.

I don't support this. Hopefully that comes through in my posts. From day one I've advocated putting out clear and accurate information, making recommendations, and letting everyone decide what's right for them.

[ACfixer]

Wow, some of you guys sure do a lot of writing trying to convince some other clown on the internet that you're right. Have you changed any minds?

[as_rocketman]

Wow, some of you guys sure do a lot of writing trying to convince some other clown on the internet that you're right. Have you changed any minds?

I don't worry too much about the clowns.

Sure, I've gotten a lot of positive feedback. It also helps me work out and test my own opinions. How about you?

[LBDamned]

Wow, some of you guys sure do a lot of writing trying to convince some other clown on the internet that you're right. Have you changed any minds?

does anyone even read the wall of texts?

they need to use PM... the one-upmanship of who can write the longest post is nauseating.

[Libertarian777]

I don't worry too much about the clowns.

Sure, I've gotten a lot of positive feedback. It also helps me work out and test my own opinions. How about you?

i actually appreciate this dialog.

the internet varies between
VACCINES WILL KILL US ALL (conspiracy minded folks)
and
COVID WILL KILL US ALL (mainstream media folks)

there's a middle ground which i espouse.
Vaccines are EUA. Vaccinate > 65 immediately (highest risk).
Now lets debate the rest.

And leave my kids the fk alone.

The entire discussion is around this middle part... whether RRR is relevant depending on assumptions of actual risk for the general population (< 65).

It's hard to find someone to discuss rationally with. Even my PhD friend is subject to hysteria. When I disprove his own theory with his own data he goes silent and changes subjects. The irony being I had to convince him to wear a mask in early March last year (unknown pathogen, unknown transmission vector), and this year he's arguing with me as to why we still need to wear masks (yes he's vaccinated).