83 billion COVID shots?

Post #1069

When estimating the mortality effect of an injected product, like the COVID shot, you have approximately 3 options:

1. head-to-head trial where half of participants remain uninjected

2. extrapolation based off of the best estimate of an IFR (ie, UK Tech Briefing #5)

3. extrapolation based off of the best estimate of reduced spread

The best way is number 1, but the only time that that research was done was for a few months in 2020 — and COVID was not lethal enough for there to be a direct estimate on potential lives saved. COVID was not significantly more deadly than seasonal flu.

The second-best way is number 2, acquiring the greatest understanding of the underlying lethality of the disease itself, so that you could predict what would happen if you had a product which could prevent fatalities — but, as said before, no head-to-head research ever proved any reduction in COVID fatality from taking COVID shots.

That makes option 2 your best bet, because it is as much direct evidence of potential benefit that you can acquire. Number 3 is the worst way to try to estimate the effect of a product, but it is at least included in the methods used in producing two estimates which show up in the government report on COVID, chaired by Rep. Brad Wenstrup.

The first report was a state-by-state regression of COVID shot uptake rates against COVID deaths, and it led to a model that predicted that, by 9 May 2021, there were 3.1 million COVID infections prevented by COVID shots, and 140,000 COVID deaths averted. But let’s apply option #2 above (best estimate of lethality) to those figures:

In supposedly averting 3.1 million COVID infections, the expected number of deaths prevented — using the best data set on COVID lethality (UK Tech Briefing No. 5) — you would expect for less than 5,000 COVID deaths to be prevented. That’s because there were approximately 616 COVID infections per death in the UK data.

Notice how the estimate by Gupta et al. is over 25 times higher than what the best data set predicts. It is not a disparity that can be scientifically “explained away” — because it is more than an order of magnitude of difference.

Another comparison is with the real-world data from Israel, showing how it took over 25,000 COVID shots to prevent one death. To get those 140,000 deaths prevented by 9 May 2021 would have required 3.6 billion shots. That’s 720 million COVID shots given per month, in a nation of 330 million people (more than two shots per person per month).

The other citation from the government report on COVID is from a blog for The Commonwealth Fund, and they do not reveal all of their assumptions and methods, but we can still evaluate the end-product that they put out: 120 million COVID infections prevented by 30 Nov 2022, along with 3.2 million COVID deaths prevented.

Using the real-world experience in Israel, to get the 3.2 million deaths prevented would have required 83 billion COVID shots — over 250 shots for every man, woman, and child in the USA.

---

Why the UK Tech Briefing data is the world’s best data set on pre-COVID-shot IFR:

The following reasons, brought together, are what make the UK data so good regarding published data on COVID fatality:

—ends before uptake of COVID shots (so that the effects of shots is nil)
—is prospective, tracking people (infections) forward in time
—uses lab-confirmed cases instead of suspected ones
—has more than 100 events of interest (>100 deaths)
—is in the lower third of published estimates

While actual prospective data on lab-confirmed cases followed through time before COVID shots — and with over 100 events — does exist (governments have these data), the data in the UK was actually published. Critics and detractors may cry and moan about the last point, saying it is guaranteed to produce a low number.

But to throw the criticism right back at them, there are 100 reasons for a fatality estimate to be too high (doctor didn’t treat in time, or did not use good medicine, etc.) for every reason that you could find for an estimate to be lower than reality. The background lethality of a disease will prevent estimates that are “lower than possible.”

The only way I know to get an estimate below the ground reality is to make a math error: under-estimate the actual count of deaths, over-estimate the actual count of infections. But the UK data prevents both of these errors, because it followed proven cases forward over time.

There is still the nagging issue of whether COVID caused the deaths which got counted within 28 days of lab-confirmation, but, at this point, that is true of all published data. Proving that COVID caused the death requires more work.

The evidence suggests that the government report on COVID erroneously over-estimates early benefits from COVID shots. Even worse, they do so by relying on a blog post with opaque methods and assumptions and on a regression which did not validate the veracity of state-based COVID death data.

Reference

[government report on COVID] — https://thehill.com/wp-content/uploads/sites/2/2024/12/12.04.2024-SSCP-FINAL-REPORT.pdf

[140,000 deaths averted?] — Sumedha Gupta, Jonathan Cantor, Kosali I. Simon, Ana I. Bento, Coady Wing, and Christopher M. Whaley. Vaccinations Against COVID-19 May Have Averted Up To 140,000 Deaths In The United States. Health Affairs Vol. 40, No. 9. PUBLISHED: August 18, 2021. https://doi.org/10.1377/hlthaff.2021.00619

[3.2 million deaths averted?] — Meagan C. Fitzpatrick et al., “Two Years of U.S. COVID-19 Vaccines Have Prevented Millions of Hospitalizations and Deaths,” To the Point (blog), Commonwealth Fund, Dec. 13, 2022. https://doi.org/10.26099/whsf-fp90

[Over 25,000 injections required in the real-world data in order to avert just one death] — Larkin A, Waitzkin H, Fassler E, Nayar KR. How missing evidence-based medicine indicators can inform COVID-19 vaccine distribution policies: a scoping review and calculation of indicators from data in randomised controlled trials. BMJ Open. 2022 Dec 12;12(12):e063525. doi: 10.1136/bmjopen-2022-063525. PMID: 36523237; PMCID: PMC9748517. https://pmc.ncbi.nlm.nih.gov/articles/PMC9748517/

[169 deaths by Day 28 from 117,000 early COVID infections; adjust up by dividing by 0.89] — UK Technical Briefing #5. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/959426/Variant_of_Concern_VOC_202012_01_Technical_Briefing_5.pdf

[“Time-to-death” probability for COVID; using the 95% upper bound of both the mean and the SD of the lognormal model reveals that 89% die by Day 28] — Linton NM, Kobayashi T, Yang Y, et al. Incubation Period and Other Epidemiological Characteristics of 2019 Novel Coronavirus Infections with Right Truncation: A Statistical Analysis of Publicly Available Case Data. Journal of Clinical Medicine. 2020 Feb;9(2). DOI: 10.3390/jcm9020538. PMID: 32079150; PMCID: PMC7074197. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7074197/