Isn't vaxxed + previously immune a confound here? you can't remove previously immune from unvaxxed without also removing it from vaxxed. The vaxx efficiency could just be confounded by natural immunity
1 - In the case of "Unvaccinated" we are flipping them from the "No Immunity to "Has Immunity" column. It's literally the opposite situation. With Vaccinated with/without Infection, it is from Immunity to Immunity+(?) so the difference, while potentially real is not an actual reversal. It also probably matters which order they occurred and how long ago for both as to how much effect there is.
2 - The percent of the vaccinated that have prior infections is MUCH smaller because the percent of the population that is vaccinated is much larger. I did a back of the envelope check using Week 41 N Protein Immunity (natural only, but potentially not all infection generate N antibodies) and after first subtracting my proposed Prior Infection group from the specific N-Antibody population, I calculated the percentage of the Vaccinated population that also has N antibodies and I came up to an average of 4% (with a high of 8% and a low of 0%).
So let's be generous and say 10% of the vaccinated population ALSO has natural immunity either from infection before or after vaccination (that's another can of worms right there) - it's a much smaller portion of the overall population - the effect, though it almost certainly exists - is not going to be anywhere near as much as the effect in the Unvaccinated population where it is 85%+. I also suspect a non-trivial number of the vaccinated/naturally infected are single dose, not 2 dose vaccinees. This further dilutes the number in the relevant column (Vaccinated in my calculations is 2 Dose vaccinated).
Thanks. I think the problem is we just don't have accurate numbers on 2) and as more people get vaxxed it skews more. It's hard to see how the unvaxxed natural immunity could be 85% but vaxxed NI only 10%. That would be extremely different behaviour in what is a fairly homogeneous group (it's not like people are avoiding the vaxx in huge numbers because of prior NI)
You can see that almost all new vaccinations are coming from the non-immune section (though they do sometimes come out of the prior-infection group, it's clearly less common)
It's because the population sizes are so different. There are roughly 40m people vaccinated over 18 and under 10m unvaccinated. So even if it were exactly the same number of people, it would be 1/4 the percentage. You also have a selection bias in that those with prior-infection are less likely to also get vaccinated. (They do, but they are less likely).
But it isn't an equal number of people - because ALL excess S Protein immunity over the Single Dose count must be in the "Unvaccinated" column (I confirmed by replicating their case rates that they are calculating the Unvaccinated population this way). Also of course, the vaccinated SHOULD be catching COVID less. And they definitely were at first.
They categorize those who had the COVID shots less than 14 days ago as “unvaccinated”
The first 14 days after getting a shot accounts for at least 50% of deaths and adverse reactions. That is just cooking the books by big pharma and the CDC to make COVID shots look good. Just like Pfizer decided to conveniently not count all of deaths among the control groups which allowed them to pretend that the vaccine efficacy was over 50%. It turned out it’s closer to 23% and should have never been approved for emergency authorization use.
This is true in the US, but not the UK. They have a "One Dose < 21 days" category. Also by the time they started putting out these charts, the first doses were pretty much entirely done so there isn't any evidence of first dose waves in this data. The same couldn't be said for Dose 3 < 14 days and it is in fact missing.
I'm not saying other UK statistics might not be polluting the Unvaccinated pool in the same way as you describe, but THIS UK data doesn't.
I was thinking about this earlier today and if the apparent negative efficacy in infection rates were a result of a shrinking vulnerable population in the unvaccinated cohort wouldn't we also expect to see the death/hospitalization rates follow the same trend over time since the denominator is the same? I could be wrong, math and data analysis aren't exactly my strong suit :)
Basically almost all the cases are in the non-immune so they drive hospitalization rates. Also you really need to calculate against the number of cases for hospitalization not the total population. (because you can't be hospitalized with Covid if you don't have Covid)
I mean it has to be more positive in terms of efficacy against even catching Covid. I want to caution that these rates here aren't actually correct, they assume the prior Immunity group gets zero cases which isn't really plausible given their size. If you keep reading, you can see that I calculate a theoretical case rate among the prior infection group which is significantly lower than the Vaccinated Rate during the same period. I'm my most recent post https://justguy.substack.com/p/why-admissions-and-mortality-are I come up with a way to measure vaccine efficiency against just the the non immune population and there is a clear falloff to that signal which is time based. IE there isn't a cliff where the vaccine efficiency drops precipitously yet (which could indicate complete evasion). In your age group it's about a 2/3rds reduction in cases.
But as you say, you really have to do the risk benefit calculations and they aren't really clear. Do we believe we know all the side effects and their probabilities or not? (I tend to think no, but data gathering has been so poor on this it's really hard to know by how much). To the extent that 60 is definitely an inflection point in the outcomes graph you have to weigh the risk FROM Covid more highly. And remember that not dying is an outcome on a continuum of possible outcomes. I don't know how good our data is on what the risk factors for those various outcome endpoints are. To the extent you are properly assessing your fitness and overall health, it's a really difficult decision to make and it should be your decision not your government's.
You are welcome. I'm glad you appreciate that about my posts. One of the issues in the "substack" game, even for free ones, is that it's clear that getting audience participation is assisted by a little bit of pandering to their priors. My core thesis does not do that (though I believe the correlated implication regarding natural immunity does). But I'm really hoping to talk to truth seekers rather than position seekers.
As to your conundrum, the only way I think I could solve it personally would be to decide if you are willing to face your better defined risk to avoid the ill defined one (knowing you can't know for sure the way either will go). It sounds like that's what you are doing.
Isn't vaxxed + previously immune a confound here? you can't remove previously immune from unvaxxed without also removing it from vaxxed. The vaxx efficiency could just be confounded by natural immunity
In case you haven't seen it, I directly address this question in detail here: https://justguy.substack.com/p/prior-infection-and-the-vaccinated
This is a good question and I look at it 2 ways:
1 - In the case of "Unvaccinated" we are flipping them from the "No Immunity to "Has Immunity" column. It's literally the opposite situation. With Vaccinated with/without Infection, it is from Immunity to Immunity+(?) so the difference, while potentially real is not an actual reversal. It also probably matters which order they occurred and how long ago for both as to how much effect there is.
2 - The percent of the vaccinated that have prior infections is MUCH smaller because the percent of the population that is vaccinated is much larger. I did a back of the envelope check using Week 41 N Protein Immunity (natural only, but potentially not all infection generate N antibodies) and after first subtracting my proposed Prior Infection group from the specific N-Antibody population, I calculated the percentage of the Vaccinated population that also has N antibodies and I came up to an average of 4% (with a high of 8% and a low of 0%).
So let's be generous and say 10% of the vaccinated population ALSO has natural immunity either from infection before or after vaccination (that's another can of worms right there) - it's a much smaller portion of the overall population - the effect, though it almost certainly exists - is not going to be anywhere near as much as the effect in the Unvaccinated population where it is 85%+. I also suspect a non-trivial number of the vaccinated/naturally infected are single dose, not 2 dose vaccinees. This further dilutes the number in the relevant column (Vaccinated in my calculations is 2 Dose vaccinated).
Thanks. I think the problem is we just don't have accurate numbers on 2) and as more people get vaxxed it skews more. It's hard to see how the unvaxxed natural immunity could be 85% but vaxxed NI only 10%. That would be extremely different behaviour in what is a fairly homogeneous group (it's not like people are avoiding the vaxx in huge numbers because of prior NI)
I suggest looking at my chart in https://justguy.substack.com/p/showing-my-work
You can see that almost all new vaccinations are coming from the non-immune section (though they do sometimes come out of the prior-infection group, it's clearly less common)
It's because the population sizes are so different. There are roughly 40m people vaccinated over 18 and under 10m unvaccinated. So even if it were exactly the same number of people, it would be 1/4 the percentage. You also have a selection bias in that those with prior-infection are less likely to also get vaccinated. (They do, but they are less likely).
But it isn't an equal number of people - because ALL excess S Protein immunity over the Single Dose count must be in the "Unvaccinated" column (I confirmed by replicating their case rates that they are calculating the Unvaccinated population this way). Also of course, the vaccinated SHOULD be catching COVID less. And they definitely were at first.
They categorize those who had the COVID shots less than 14 days ago as “unvaccinated”
The first 14 days after getting a shot accounts for at least 50% of deaths and adverse reactions. That is just cooking the books by big pharma and the CDC to make COVID shots look good. Just like Pfizer decided to conveniently not count all of deaths among the control groups which allowed them to pretend that the vaccine efficacy was over 50%. It turned out it’s closer to 23% and should have never been approved for emergency authorization use.
This is true in the US, but not the UK. They have a "One Dose < 21 days" category. Also by the time they started putting out these charts, the first doses were pretty much entirely done so there isn't any evidence of first dose waves in this data. The same couldn't be said for Dose 3 < 14 days and it is in fact missing.
I'm not saying other UK statistics might not be polluting the Unvaccinated pool in the same way as you describe, but THIS UK data doesn't.
Scotland cheats https://twitter.com/freedompodcast1/status/1494000134341939204?s=21
I will check this later, but it looks like you have managed to do some good data digging here, nice one.
I was thinking about this earlier today and if the apparent negative efficacy in infection rates were a result of a shrinking vulnerable population in the unvaccinated cohort wouldn't we also expect to see the death/hospitalization rates follow the same trend over time since the denominator is the same? I could be wrong, math and data analysis aren't exactly my strong suit :)
This is a good question, I cover it here: https://justguy.substack.com/p/why-admissions-and-mortality-are
Basically almost all the cases are in the non-immune so they drive hospitalization rates. Also you really need to calculate against the number of cases for hospitalization not the total population. (because you can't be hospitalized with Covid if you don't have Covid)
flue is what the smoke escapes through from your wood burning stove.
flu is short for influenza
/nitpick.
Thanks, fixed
I mean it has to be more positive in terms of efficacy against even catching Covid. I want to caution that these rates here aren't actually correct, they assume the prior Immunity group gets zero cases which isn't really plausible given their size. If you keep reading, you can see that I calculate a theoretical case rate among the prior infection group which is significantly lower than the Vaccinated Rate during the same period. I'm my most recent post https://justguy.substack.com/p/why-admissions-and-mortality-are I come up with a way to measure vaccine efficiency against just the the non immune population and there is a clear falloff to that signal which is time based. IE there isn't a cliff where the vaccine efficiency drops precipitously yet (which could indicate complete evasion). In your age group it's about a 2/3rds reduction in cases.
But as you say, you really have to do the risk benefit calculations and they aren't really clear. Do we believe we know all the side effects and their probabilities or not? (I tend to think no, but data gathering has been so poor on this it's really hard to know by how much). To the extent that 60 is definitely an inflection point in the outcomes graph you have to weigh the risk FROM Covid more highly. And remember that not dying is an outcome on a continuum of possible outcomes. I don't know how good our data is on what the risk factors for those various outcome endpoints are. To the extent you are properly assessing your fitness and overall health, it's a really difficult decision to make and it should be your decision not your government's.
You are welcome. I'm glad you appreciate that about my posts. One of the issues in the "substack" game, even for free ones, is that it's clear that getting audience participation is assisted by a little bit of pandering to their priors. My core thesis does not do that (though I believe the correlated implication regarding natural immunity does). But I'm really hoping to talk to truth seekers rather than position seekers.
As to your conundrum, the only way I think I could solve it personally would be to decide if you are willing to face your better defined risk to avoid the ill defined one (knowing you can't know for sure the way either will go). It sounds like that's what you are doing.