VAERS military death reporting rates increased 230% from 2019 to 2021

Why is the observed number of reports of death so much higher the expected number of reports for 2021?

I wanted to examine any discrepancies in reporting rates of death for military reports using pre- and post-COVID-19 era entries. I made two datasets using VAERS reports: one containing 2019 reports and one containing 2021 reports. As you can probably imagine, the former contains reports in the context of a whole bunch of vaccines, and the latter are almost all COVID-19 injectable product reports (94%).¹

To assess the military reports, I pulled out all reports from each dataset whereby the vaccine or product was administered by military personnel by selecting V_ADMINBY = “MIL” (military reports).² I am making the assumption that if a VAERS report includes a “MIL”-administered COVID-19 shot, that they are an active-duty service member.

Here’s what I got.

The 2019 dataset contains 48,444 total reports, 701 of which are military reports including 2 deaths. The 2021 dataset contains 753,047 total reports, 8,806 of which are military reports including 83 deaths. To put this into percentage-perspective, there’s a 1,454% increase in total reports, a 1,156% increase in the number of military reports and a 4,050% increase in the number of deaths within these military reports when comparing 2019 and 2021. Granted, the numbers of death reports are low, especially for the military reports filed in 2019, so this must be considered upon any interpretation of these results.

The VAERS-reported death rate ratio when when comparing 2021 to 2019 is 3.3:1 ((83/8,806)/(2/701)). This represents a 230% increase in death rate reporting with regard to military reports from the pre- and post-COVID-19 shot era as shown in the chart below.

The interesting aspect of investigating military data in VAERS is that there should technically be no under-reporting since the U.S. military are very strict and regimented. I would imagine that any emergent adverse event that occurred within temporal proximity to an administered vaccine, would be reported actively to VAERS by the military entity, since the health and safety of our service men and women are priority #1. I acknowledge, that even in highly disciplined data collection systems, under-reporting can occur due to various reasons like administrative oversight or misinterpretation of what constitutes a reportable event.

U.S. Military data

As of September 30, 2024, the U.S. military has approximately 1.32 million active-duty members. Here is a breakdown of the number of active-duty personnel by branch:

• Army: 449,816

• Navy: 332,336

• Marine Corps: 173,096

• Air Force: 315,958

• Space Force: 9,446

• Coast Guard: 40,612

Of these active-duty personnel, the number of deaths reported per recent year is estimated to be approximately 720-1,000. Let’s take the average of that: 860.³⁴

Using the actual numbers provided:

• Total number of deaths: 860

• Total number of active-duty personnel: 1,321,264

By the end of 2021, approximately 97-98% of active U.S. military personnel had been been administered with at least 1 COVID-19 shot so our denominator: the number of active-duty personnel that got a COVID-19 shot, is ~1,300,000.⁵

Thus, the background (expected) death rate for active-duty personnel is 662 deaths per 1M ((860 / 1,300,000) x 1,000,000 ≈ 662), and they all got a COVID-19 shot. We can assume that since death is considered to be a serious adverse event (SAE), that we can use the Pfizer phase III clinical trial of SAE occurrence⁶ to calculate an expected number of personnel who would die “in association with” a COVID-19 shot. Since the SAE rate was 0.7% (124/18801 - for BNT162b2 product), we can expect that of the 860 deaths that will occur, 6 will be associated with the COVID-19 shots.⁷

VAERS military report data

This is really important data because the U.S. military are very strict, and I would imagine that these reports are not filed as passively to VAERS as for regular civilians, as mentioned. I might be wrong, but it makes sense to me that this would be the case. Therefore, it should be possible to approximate the percentage of deaths due to vaccines if we do assume active reporting.

There are 83 reports of death in the VAERS military report data for 2021. If 6 active-duty service members are expected to die per year “in association with” COVID-19 shots, and 83 are reported to have died in the context of a COVID-19 shot to VAERS, then what accounts for the fact that we have 13.8 times more reports than would be expected?

It's possible that there may be some over-reporting or other biases at play, but considering what we know about the COVID-19 shots, it's also possible that the data is indicating a genuine safety signal. How likely is it that this is due to over-reporting (having not factored in under-reporting) or rather, that this is a real signal emanating from VAERS? I think it is very likely because even when we go back a little farther in time, and time frame match the pre- and post-COVID-19 VAERS data, the rate of death reports per 100,000 MIL reports in the post-COVID era (2021-2023) is 3 times higher than in the pre-COVID era (2017-2019) as shown in the Figure below.

The consistency of reporting pre-COVID-19 is excellent; not so much the case in post COVID-19 era as demonstrated below.

VAERS Data Comparison: Pre-COVID-19 vs. Post-COVID-19

Pre-COVID-19 Era (2017-2019):

• Total deaths reported: 3 (2017) + 2 (2018) + 2 (2019) = 7

• Total MIL reports: 783 (2017) + 776 (2018) + 701 (2019) = 2,260

• Rate per 100,000:

  • Deaths per 100,000 reports = (7 / 2,260) * 100,000 = 309.73

Post-COVID-19 Era (2021-2023):

• Total deaths reported: 83 (2021) + 16 (2022) + 8 (2023) = 107

• Total MIL reports: 8,806 (2021) + 1,909 (2022) + 632 (2023) = 11,347

• Rate per 100,000:

  • Deaths per 100,000 reports = (107 / 11,347) * 100,000 = 943.26

Let’s consider the facts:

1. There are 13.8 times more reports than would be expected based on the background rate calculated using the Pfizer SAE rate

2. The reported death rate in VAERS in the post-COVID-19 era is 3 times higher than the pre-COVID-19 era

Based on these facts, it seems likely that the COVID-19 shots are causing significantly higher numbers of deaths than other vaccines, and that this is not explained simply by the over-reporting “argument”, or a result of reporting bias.

The VAERS system is capturing a real signal, and it needs to be investigated.

P.S. If you have some steel-man arguments to make (or see any errors), please enter them in the comments as usual! I want to nail this.

1. In fact, 94% all reports with V_ADMINBY == “MIL” are in the COVID-19 shot context. The other reports are in the context of ANTHRAX and other vaccines.

2. https://vaers.hhs.gov/docs/VAERSDataUseGuide_November2020.pdf

3. https://sgp.fas.org/crs/natsec/IF10899.pdf

4. Brave, Grok and Venice AI searches

5. https://www.pbs.org/newshour/nation/army-98-of-soldiers-got-covid-19-vaccine-by-deadline

6. https://www.fda.gov/media/144245/download. Page 33

7. Of the 1,300,000 military personnel who were injected with at least 1 COVID-19 shot, half got Pfizer, so the new denominator is: 650,000. The number of military personnel who would be expected to suffer an SAE in the context of Pfizer shots would be 4,550 (0.7% of 650,000=4,550). Of these SAE sufferers, the expected number to die would be 6 (0.133% of 4,550=6). The 0.133% is an estimate and it depends very much on age and source data, so I went with a happy medium between VAERS and the literature.