I need help backing into the ARR from the latest Lancet study on vaccine effectiveness

Lancet ^ | 10/7/21 | me

[nitzy]

This is the supplemental data from the latest Lancet study on Pfizer vaccine effectiveness after 5 months.

I am trying to back into the ARR since I can't see where they mention it.

I am specifically looking at Table 6 a. and b. and not understanding what I am seeing.

It almost appears that 5 months after taking the Pfizer vaccine, you are MORE likely to test positive for Covid than if you had never received the vaccine.

I doubt this is true or at least that they would make numbers public that reflected such an outcome.

Can someone help me understand my error?

[nitzy]

For anyone unfamiliar with ARR vs RRR....You should know that RRR basically tells you "If you catch Covid, you are X.XX% likely to be in this group and you are X.XX% likely to be in that group." If they say the vaccine is 90% effective, that is what they are referring to. If you catch Covid (or have severe symptoms), there is a 90% (or 95%) chance you fall into the unvaccinated category.

The ARR is completely different. It says, "By taking the vaccine you are lowering your chances of catching Covid by X.XX%"

They try not to publish these numbers but from the original EUA it was 0.84% for Pfizer. The EUA was only looking at severe symptoms.

Yes, you read that right....According to Pfizer's own data, if you took their shot, you decreased you chances of having severe Covid symptoms by 0.84%.

[FreedomPoster]

It might be true. Go watch this when you get a chance.

Winning the War Against Therapeutic Nihilism & Trusted Treatments vs Untested Novel Therapies
https://rumble.com/vnbv86-winning-the-war-against-therapeutic-nihilism-and-trusted-treatments-vs-unte.html

Rumble — Peter McCullough, MD, MPH speaks at the 78th Annual Meeting of AAPS on October 2, 2021.

Slides: https://www.scribd.com/document/530328436/Slides-from-Peter-McCullough-MD-Oct-1-2021-Lecture

Patient Guide: https://aapsonline.org/covidpatientguide/

[Tilted Irish Kilt]

nitzy :" I am specifically looking at Table 6 a. and b. and not understanding what I am seeing.
It almost appears that 5 months after taking the Pfizer vaccine, you are MORE likely to test positive for Covid than if you had never received the vaccine. "

Aha ! You have won the prize !
Now check out the medical terminology for ADE ( Antibody-dependent enhancement ).
Still unknown are the long term effects on the immune system, since the Clinical Trial was little more than 2 1/2 months long.
Prepare to be "Kancel Cultured" by the Big Pharma Cartel !

[Myrddin]

Check out this thread on Twitter. The graphs are there showing efficacy month by month. It appears getting a booster based on the original Wuhan strain provides improved narrow immunity to the original spike, but raises even more the other antibodies that make the "patient" more susceptible to infection by the delta variant. See thread a link. ADE appears to be in play.

[Carriage Hill]

BKMRK.

[ReaganGeneration2]

You are correct in all this - I will post more links about this (after work).

For actual “informed consent”, people should know all this, not the idiotic “effectivity” relative risk.

I can give you confidence intervals too. I’ve looked at both Pfizer and Moderna Clinical trials, and a 6-month field study (maybe the same one in Lancet?). The numbers are all like the ones above. The vaccines are worthless.

Wait til you see the ARR of getting hospitalized (not just testing positive or getting a headache). It’s barely statistically valid. How infitessimal for the under-60 and healthy crowd is unknown.

These are numbers you can’t unsee.

[ReaganGeneration2]

* efficacy (not effectivity)

[BeauBo]

Check out the footnote.

The VEs (Vaccine Effects) they report, are after they control for independent variables.

The vaccine effect, though waning over time, is always positive.

If the vaccinated were more likely to be infected, the VE would be a negative number.

[nitzy]

I am trying to understand how they are getting to this number though.

For example, Appendix 6a shows 11282 cases in 1172835 person-years for unvaccinated subjects over 16 years old.

It also shows 691 cases in 13340 person-years for vaccinated subjects over 16 years old after 157 days post 2nd vaccination dose.

That works out to:
0.96% for unvaccinated over the whole period. (11282/1172835)
5.1% for vaccinated after 5 months (691/13340)

I figure maybe the difference in time factored in somehow but I’m not sure. If you divide the person-years of the unvaccinated by 12 to make them person-months it would be: 11.5%. Does that make it apples to apples?

Although, I am not sure about that because in the same chart if you look at the fully vaccinated one month after 2nd dose, it does show protection that would fall in line with the original ARR (.96 -.27 = .69 ARR) The original ARR was .84 which is pretty close to .69

I wish someone who was trained in this could look these numbers over with a critical eye and explain them rather than Pfizer explaining them to the media who lap up whatever they are told and repeat it.

[BeauBo]

“maybe the difference in time factored in somehow but I’m not sure.”

Yes, the risk of infection was very different at different times. That is one of the variables they controlled for (thank you, peer review process), to determine the vaccine effect. VE is the bottom line, not case numbers.

The five months expired after vaccination, as the Delta wave was peaking. The lack of months worth of low infection rates during the Late Spring trough, make the narrow time window for those folks (five months post-vaccination), seem relatively high.

It would be like only tracking American temperatures in March and April, to see if Global Warming were occurring. You would only capture a warming period, without the counterbalancing cooling periods.

In all time frames of this study, and across all age groups, the vaccine was protective - just to varying degrees.

[grey_whiskers]

Waning Immune Humoral Response to BNT162b2 Covid-19 Vaccine over 6 Months

New England Journal of Medicine

[ReaganGeneration2]

“That works out to:
0.96% for unvaccinated over the whole period. (11282/1172835)
5.1% for vaccinated after 5 months (691/13340)”

Doesn’t the denominator (person-years) make it normalized (”apples-to-apples”) in terms of time? Serious question.

To get ARR, why not look at the hospitalization numbers from Table 5b? - which is actually all we should care about:
173/300765 = 0.0005752 (risk of getting hospitalized with covid if vaxxed)
11282/1172835 = 0.009619426 (risk of getting hospitalized
with covid if unvaxxed)

So, the Absolute Risk Reduction provided by this vaccine is 0.904%

I welcome a math check.

[BeauBo]

“Doesn’t the denominator (person-years) make it normalized (”apples-to-apples”) in terms of time?”

Person-years accumulated during a month in which cases are low, are not the same (risk of infection) as person-years accumulated during a month in which the cases is ten times higher.

For example, May 2021 to August 2021. In May, there was almost no one who had been fully vaccinated five months prior. Big groups of people only began hitting the five month post-vaccination window when the Delta wave was raging around its peak.

They went to great lengths to make it normalized (”apples-to-apples”) - in calculating the VE. That is where all the other independent variables have been identified and isolated (through regression analysis).

[Myrddin]

The risk >> reward for the mRNA injections.

[ReaganGeneration2]

Thanks

Seems like normalization is ok if the vaxxed and unvaxxed were facing the same risk-of-infection conditions during those months?

I’m missing something. Don’t trouble yourself with explaining it to me. I gotta figure it out myself first by diving in more.

[nitzy]

That is about what the original EUA showed as well (.84%)

I have always believed that the vaccine had a marginal ability to protect people in general so that is not surprising.

However, if the data show that after 5 months the vaccine is DETRIMENTAL, that would be a huge deal.

I am just not sure if the person-years normalizes it or not. I had another person reply to me saying that some months shouldnt count the same as others. I guess that makes some sense.

[nitzy]

I agree in theory with what you are saying about certain months skewing the outcomes but I am not sure that fits here.

The range of this study was from Dec 14th 2020 to Aug 8th 2021. That is 33 weeks. There would have only been a few weeks between July 25th and Aug 8th that would have skewed bad for the vaccinated. The rest of the summer there were hardly any cases. The daily Covid numbers averaged twice as high during the 9 weeks from Dec 15 to Feb 15 and they were averaging about the same as the last two weeks of the study during the 10 weeks from Feb 15 to May 1.

To say that 2 bad weeks (in relation to the summer) is enough to skew the whole 33 week study so bad that it shows a 5% chance for the vaxxed vs a less than 1% chance for the unvaxxed, seems like a stretch.

Also, it would seem easy enough to break out a matching cohort of unvaccinated for that same post 5-month time period to compare them. I wonder why they didn’t do that? Or maybe they did and I am just not seeing it.

[BeauBo]

There are many independent variables at play.

Those people who were vaccinated were not random. They have different risk of infection (urban vs rural, supermarket cashier vs stay at home retiree, nursing home residents, ambulance EMT, etc) just as people have different risks for outcome (age , co-morbidities, etc). Essential workers whose jobs put them at higher risk of exposure (e.g. health care) were prioritized for vaccination.

Vaccination was not opened up to the general public until April, and even after that, different types of self selection biases would confound randomness of the sample.

That is the beauty of a good peer review process, and a good multiple regression analysis on a good data set - they can back out the many various factors that independently effect outcomes. Just one of those factors is the effect of the vaccine. That is what matters.

Looking at some ratio in the raw data (like cases to sample size) is confounded by many independent variables, skewing results at their own pace and direction.

When they controlled for known independent variables (apples to apples), they isolated the vaccine effect.

[ReaganGeneration2]

Too bad the clinical trials didn’t do the scientific diligence we’re asking for - extend the trials into our current timeframe. But they spoiled the sample and allowed the petrified placebos to get vaxxed.

So they must resort to this complicated VE model that can be exploited by Big Pharma bias during “peer review” (not saying it was!!! - just that it could be - I need to dive in, like I said, but with only my old-guy lapsing stats background).