Revisiting the estimate of deaths caused by covid vaccines

What proportion of reported adverse reactions were 'serious'?

In a previous article we estimated that, between Dec 2020 and Oct 2023, approximately 16,000 people in the UK died directly from covid vaccines. This estimate was partly based on the assumption that approximately 20% of the adverse reactions reported for covid vaccines could be considered “serious”. But how accurate was that 20% assumption …….

In this 2023 article we provided an estimate of the number of direct and indirect deaths that had been caused by the covid vaccines. As part of the estimate for the adverse reactions reported in the UK’s Yellow Card system, we assumed that approximately 20% of all adverse reaction reported were ‘serious’.

A colleague who is part of Robert F Kennedy Jnr’s MAHA team has asked us what the empirical basis was for the 20% estimate. In fact, that figure came from our colleague Scott McLachlan and his clinical colleagues' extensive sampling of covid vaccine adverse reaction reports described here.

It turns out that there is other evidence supporting the approximate 20% claim. For example, Grok says:

For COVID-19 vaccines, the MHRA has published summaries of Yellow Card reports. According to a recent post citing the latest Yellow Card data:

• Total adverse reactions reported: 1,598,855

• Serious ADR reports: 292,554

• Proportion of serious reports: 292,554 ÷ 1,598,855 ≈ 18.3%

Another source, specific to Scotland, reported 36,414 Yellow Card reports for COVID-19 vaccines, with 22,000 classified as serious, yielding a proportion of 22,000 ÷ 36,414 ≈ 60.4%. This higher proportion may reflect regional differences, specific reporting practices, or a focus on a subset of data, but it’s not representative of the UK-wide Yellow Card scheme.The MHRA notes that serious ADRs are defined as those that are life-threatening, require hospitalization, result in disability, or cause congenital anomalies, among other criteria. However, these reports are not confirmed as caused by the vaccine, as Yellow Card data includes suspected reactions, which may be coincidental.

ChatGPT says:

• In one cohort analysis, 61.3% of patient-reported ADRs were marked as serious by the reporter (8,955 out of 14,587 ADR reports). These reports correspond to 60.1% of the total events across all doses Reddit+15MedRxiv+15Reddit+15.

• However, when medically classified using MedDRA serious injury criteria, only 13.7% of people reporting an ADR actually had reactions considered medically serious (4,134 out of all reporters)

For the Canadian Adverse Events Following Immunization Surveillance System (CAEFISS), Grok reports:

According to the Public Health Agency of Canada, up to December 3, 2023, a total of 105,016,456 COVID-19 vaccine doses were administered in Canada. Out of the adverse event reports submitted to CAEFISS, approximately 55.9 reports per 100,000 doses administered were recorded, totaling around 58,712 reports. Of these, 11.1 reports per 100,000 doses were considered serious, which translates to roughly 11,664 serious reports. To calculate the proportion of adverse reactions considered serious:

• Total adverse event reports: ~58,712

• Serious reports: ~11,664

• Proportion: 11,664 ÷ 58,712 ≈ 0.1986 or 19.86%

So, about 19.86% of adverse reactions to COVID-19 vaccines reported in CAEFISS are considered serious. This is specific to COVID-19 vaccines, as CAEFISS data in the provided context focuses on those. Data for other vaccines is less detailed here, but the proportion of serious reports tends to be lower for non-COVID vaccines due to their longer safety track record and different reporting dynamics. Keep in mind that CAEFISS, like VAERS, is a passive surveillance system, so reports may include incomplete or unverified information, and a report doesn’t confirm the vaccine caused the event. Serious events are defined as those involving hospitalization, life-threatening conditions, or significant disability, among others.

So, for the Yellow Card and CAEFISS systems it appears that the estimate we used was pretty accurate.

There is much less agreement on the proportion of serious reports in VAERS. For example, chatGPT reports that, of the COVID‑vaccine related reports between Dec 2020 and Oct 2021:

• 12.6% involved ER visits

• 6.0% resulted in hospitalization

• 1.52% life-threatening events

• 1.47% permanent disability

• 1.27% death

That is 22.9% outcomes that are certainly classified as ‘serious’. However, ChatGPT also reports that:

One large dataset reported a lower overall rate of 4.7% serious reports, likely due to definitional nuances.

Conclusion

McLachlan et al’s approximate 20% estimate was based on a thorough clinical evaluation of a large sample of adverse reaction reports. In discussing this with him today he said:

The one thing we had over the AI methods and several of the other papers was that two sets of eyes went over every VAERS record. A third doctor (who had to remain anonymous in order to protect his/her GMC license) also reviewed a sample of the records.

The empirical evidence identified by AI tools also seems to support the 20% figure as being reasonable. Hence, we stick to our overall estimate that, between Dec 2020 and Oct 2023, approximately 16,000 people in the UK died directly from covid vaccines