CDC: "safety of vaccines reinforced": UNPACKED

8 Nov 2021

CDC facebook post.jpeg


On 28 Oct 2021 a brand new pre-print paper was released by the CDC response team. The CDC themselves advertise it stating "it reinforces the safety of Covid vaccines". The pre-print paper has not yet been peer-reviewed.

CDC response team paper.png



a) The paper is a pre-print. I personally have no objection to giving some weight to pre-prints. This pre-print however has several flaws and I have yet to see it be accepted by peer-review.

b) The paper uses the same VAERs data that we are told cannot be used to infer causation due to bias. Yet, the CDC uses it to infer the safety profile of vaccines. 


Analysing VAERs data has its merits in terms of detecting safety signals. It is VAERS that detected myocarditis issues. VAERS also detects a huge number of post-vaccine deaths. However, we are told there is no solid proof of injury or death and that there is no proof of association. The CDC has not looked very hard though, including in this science paper. In addition, no such investigation has been done by the FDA. There are no randomised controlled trials, certainly none that are powered enough to detect the safety signal of death seen in the VAERS data (and the original Pfizer trials were underpowered to detect deaths.)

c) The CDC summary says most adverse reports were not serious and side effects were mild and short-lived. Nobody is denying that. Most reports are just a sore arm. However, their sweeping statement downplays the seriousness of the adverse reactions that are reported, including nearly 18,000 deaths reported by year's end.

d) The CDC states that vaccines are the most effective tool to prevent severe illness. This is true only for 3-6 months after being fully vaccinated. On a population level, vaccinations increase the amount of cases on a population level.

e) In the paper's review of VAERS death reports, it says no unusual patterns for cause of death were found. Yet, to the ordinary eye, there is a clear signal of death reports compared to previous years (shown in the VAERS graph below. This graph is deaths following vaccination to the end of 2021, whereas the study ended in June).

Despite claiming no unusual patterns it does comment on the temporal reporting of deaths clustered around vaccination date. It says it may represent a bias of reporting. We don't know though because there has been no randomised controlled trial big enough to test it.


VAERS data is graphed by

f) The paper did not use the comparison of death reports to previous years as shown in the VAERS graph above. The CDC explain that the reason for that is that for the Covid vaccine, deaths following vaccination are REQUIRED to be reported whether it seems likely the death was related to the vaccine or not. Therefore, they say, a higher reporting rate of deaths is expected.

The CDC specifically says  "During the study period, all Covid vaccines were administered under emergency use authorisations (EUAs), which require vaccine providers to report serious safety events that follow vaccination to VAERS. Heightened public awareness of the Covid vaccination program, outreach and education to healthcare providers and hospitals about Covid EUA adverse event reporting requirements, and adherence to EUA reporting requirements by providers and health systems, are likely contributing to the high volume of VAERS reports following mRNA vaccination."

Dismissing this signal bears several flaws:

First and foremost, if deaths from previous years are much higher because they are not required to be reported, this should raise concerns about other vaccines' safety rather than reassuring us about the Covid vaccine.


It is quite possible that not every post Covid vaccine death is being reported and we may in fact be seeing under-reporting rather than over-reporting. It is already established that less than 1%-13% of serious drug adverse events are reported to the FDA (the number is 5% in NZ).

Most European countries, the UK and many other countries have similar patterns of deaths being reported following vaccinations. These countries have no such legal requirement to make a report.

Additionally, there is no evidence that all US authorities report all deaths after vaccination to VAERS despite this being a federally mandated procedure. To think all physicians spend at least half an hour filling in death reports for every deceased patient, for an undefined period of time after death, even if they do not suspect the death is related to the vaccine, is naive at best. Statistics alone tell us this is not happening.

In order for reported deaths to exceed mortality, in a statistically significant and meaningful way, *every single* post vaccine death would need to have been reliably collected. Then we could compare the number of deaths reported with previous years.

The all cause mortality rate in America is approximately 0.86%. The number of people receiving two doses was at least 150 million (assuming a two dose course totalling the 300 million doses reported). We would expect a reporting mortality total of approximately 1,290,000 death reports in this six month period if the rules were strictly adhered to (with every single death reported). Yet, the actual reported deaths at six months (in this study) was 4,472. Clearly, the mandated reporting was not at all being adhered to. Therefore the claim that reports were over-reported is baseless.

Given their guidelines, that all deaths are to be reported, it sounds like they expected 1,290,000 death reports to be filed. What they are saying now, is that 4,472 deaths is less than 1,290,000 expected deaths, so there is no vaccine safety issues.

Specifically, the results of the study states "Analysis of deaths reported to VAERS demonstrated lower than expected reported mortality rates compared with background mortality rates."

The paper refers back to the clinical trials that showed the vaccine had an acceptable safety profile. The (Pfizer) clinical trials had about 18,000 vaccine recipients (in the experimental group). If death was going to be detected as a rare adverse event, it would only be detected if the deaths numbers were with a 95% confidence interval, statistically significant.

If all 4,472 deaths were a genuine vaccine-related death, then out of 1,290,000 people, that is a death rate of .003% or one person in 3,333.

If the actual vaccine death rate was, say, 1:33,000, a trial with 18,000 would not detect it. It would only be about .54 of a person more than the control group. It would need a much larger sample size to be statistically significant.

Finally, the paper says this "This study has several strengths, including the large population under surveillance and comprehensive capture of national data from two complementary surveillance systems." (it also looked at v-safe but there appeared no deaths in that data). The point is, the population are under surveillance and we have death data but yet aren't taking any notice of it. Secondly, their statement proves the point that the public's response to the vaccine is under surveillance because this is indeed still an experimental vaccine, and that we are the experiment.



Post script:

*** I would be grateful if any enthusiastic or retired statistician (teacher or otherwise) is able to do the exact calculations on this. The exact question is this: Assuming an all cause mortality rate of .86%, what population sample size would be necessary in a randomised controlled trial to detect if an excess death rate of 0.003% is statistically significant? *** I bet it is more than 18,000.

Reference links

CDC fb link:


CDC paper (