‘That’ BMJ Excess Deaths Paper. An Antivax Trojan Horse

Evidence Challenge
10 min readJun 9, 2024

--

A lot has already been written about a new scientific paper, “Excess mortality across countries in the Western World since the COVID-19 pandemic: ‘Our World in Data’ estimates of January 2020 to December 2022”, published in BMJ Journals https://bmjpublichealth.bmj.com/content/2/1/e000282#ref-58

The study immediately drew the attention of The Telegraph and GB News in the UK with suggestive headlines about vaccines and excess deaths. These were followed by articles published across the world including in US, France and Australia.

Of course, anti-vaxxers have been quick to jump on these ‘mainstream media’ headlines that they would normally shun, including this tweet that has gathered 4 million views so far

But what of the paper itself? In brief, it uses Our World in Data figures to re-calculate and highlight continuous excess deaths since the start of the pandemic into 2021 and 2022 after the introduction of vaccines and mitigation measures and asks whether “non-Covid” excess deaths could be due to these interventions

Scientists and medical professionals have been busy putting out responses pointing out that the paper adds little to established scientific understanding, has significant shortcomings, and provides no evidence that vaccines are adding to excess deaths. Some commentators feel that the paper has been misunderstood and is merely calling for further investigations into excess deaths, others are more critical about the study’s design and content feeling it plays into anti-vax hands. There is also some consternation that the BMJ allowed the paper to be published given its obvious short-comings.

Here, based on my interpretation of the paper, I go further and suggest that this is not a scientific study at all and is, instead, a propaganda piece aimed at building legitimacy for well-worn anti-vax tropes through a peer-reviewed nudge nudge, wink wink article.

At its core, the anti-vax movement does two things. It minimizes disease risk while exaggerating vaccine risk. That’s it. That’s the game. And this is the central approach of the BMJ paper. It achieves this in three ways:

  1. Bias

The scene-setting Introduction has a very brief section about the risks of Covid19 amounting to a mere 44 words and 8 references. No death numbers despite official totals reaching over 7 million globally. No details of specific disease risks other than “a leading cause of death”

The authors take a different approach to vaccine risks, however:

I had to split screen prints to accommodate the size of this 346 word polemic, nearly 8 times larger than the section on Covid19 risks and 31 references. The text includes a long list of medical conditions, many taken from VAERS which, of course, does not establish cause. All these conditions are observed more frequently from Covid19 infections in the real world, although strangely not in this paper — more on omissions later

The authors even manage to devote more words (108, or 2.5 times Covid19 risks) to the risks of mitigation measures such as lockdowns and social distancing

I also conducted single/phrase word search counts on the paper. In the main body (excluding references), use of “Vaccine/Vaccination/Vaccinated” and “Measures” combined (61) was higher than “Covid-19”, “virus” and “SarsCov2” combined (51) when not used in reference to vaccines. 12 of the non-vaccine “Covid-19” instances were in relation to the pandemic period rather than the effect of the virus. This, again, points to a strong bias towards vaccine and mitigation measure risks

Another area of bias was the choice of referenced sources to minimize COVID harms and exaggerate vaccine risks

The paper examined Infection Fatality Rates citing surprisingly low numbers

All three references provided were from John Ioannidis who notoriously produced some of the lowest IFR numbers throughout the pandemic and is seen as a contrarian in some scientific circles. The paper ignored all other sources including this Imperial College London report based on 175 studies

I don’t know which is more accurate (although I’m reasonably confident which would be favourite with book-makers), the fact remains that the paper used an extreme estimate downplaying COVID risks

The paper also cited anti-vaxxers’ all-time favourite COVID-19 vaccine study, a ‘re-analysis’ of Pfizer’s trial results data on vaccine side effects from Joseph Fraiman. The paper has been roundly criticised with excellent analysis here https://sciencebasedmedicine.org/peer-review-fail-vaccine-publishes-antivax-propaganda/ and video here https://www.youtube.com/watch?v=drSAsfuMkuw. Fraiman’s co-authors include Peter Doshi who has another study referenced (92) in the BMJ’s “Excess Deaths” paper that was critical of “Big Pharma”. Doshi, for those that don’t know, is a senior editor at the BMJ. Just saying

It is also interesting to note the only paper referenced (19) that was critical of non-pharmaceutical interventions was from East Africa. Bear in mind the study under review is focused on the “Western World”

Bizarrely, the study even manages to sneak in a sentence about ‘gene therapy’ in reference to vaccines with two referenced studies

2. Omission

In order to downplay disease risks, the paper omits as much as it can get away with. The fact that there are no COVID19 death numbers while total annual figures for excess deaths are explained in great detail gives a hint to the scale of omission

The authors chronicle the introduction of vaccines and mitigation measures comparing these with annual excess death figures. Meanwhile, there is no mention of Covid waves, spikes, alpha, delta or omicron. None. The single use of the word “variant” in the main text suggests these became “milder” (again referencing 2 Ioannidis studies) without mentioning the fact they were infecting far more people leading to more deaths

In fact, Covid spikes correlate very closely with excess deaths peaks

Talking of Karlinsky & Kobak, this is exactly who Our World in Data (OWID) use to create their own excellent interactive excess mortality reports. The whole basis of this paper under review is to evaluate excess deaths based on OWID data. They are essentially recreating the wheel. A very wonky wheel

The paper also ignores any other factors other than fleetingly mentioning “other factors”. The UK, for example, faced its worst flu season in 2022/3 for over 5 years concentrated at the end of 2022, boosting excess deaths. No attempt is made to evaluate non-vaccine idiosyncratic risks

A final crucial omission is the behaviour of the general public during a pandemic. The authors focus on risks from offical mitigation measures but are silent on the changes people naturally make in the face of a deadly disease. Sweden is included in the study yet famously avoided lockdowns in 2020, its highest year of excess deaths. New Zealand, that operated a harsh Zero Covid policy in the same year, had negative excess mortality according to the paper’s own (green bar). Most delays in diagnosis and treatment may well have occured as people avoided healthcare facilities to avoid a deadly and debilitating virus that left healthcare services severely short of healthcare workers. New Zealand and others delayed this before vaccines were rolled out to reduce the virus’ worst effects

3. Ignore

The final strategy is to ignore evidence that is staring the authors squarely in the face. Let’s start with the paper’s own excess death country chart (Fig2) I highlighted above

Over the course of the pandemic, particularly during 2021 when vaccines were initially rolled out, the countries with highest excess deaths were predominantly in the Balkans region (e.g. Bulgaria) while those with the lowest excess deaths were in Scandanavia, Australasia, Western Europe (e.g. Denmark)

Levels of vaccination uptake in these regions clearly corresponded with these different levels of excess deaths — high vaccination rates had lower excess mortality during the period. The chart below looks at the top and bottom ends of the European Economic Area vaccination uptake tables

https://www.statista.com/statistics/1218676/full-covid-19-vaccination-uptake-in-europe/

The authors have actually produced powerful evidence against the role of vaccines in excess deaths yet chosen to ignore it

The paper makes much mileage out of reporting inefficiencies in some countries yet ignores the rich data available from others. The UK’s Office for National Statistics produced a detailed and robust comparison of all-cause mortality between those who had ever received a vaccine and those who had not since April 2021. The results were clear — vaccines were highly unlikely to be producing significant additional deaths

More importantly, authors ignored official vaccine death figures while making a big deal about underuse of post-mortems. Yet, based on death certificates that include post-mortem results for unusual deaths, particularly among younger people, only around 80 vaccine-related deaths have been registered in the UK. Compare this with over 223,000 Covid deaths (2787 times higher)

Or maybe higher. There is significant evidence that Covid-19 deaths have been materially undercounted globally. This is partly due to Covid-19 deaths being counted as natural causes, partly due to poor death recording, and partly due to Covid sequelae (long-term vascular damage affecting internal organs including the heart) but not registered in Covid-19 counts

https://www.ucl.ac.uk/news/2023/jan/covid-19-patients-may-retain-elevated-risk-death-18-months-after-infection

The final omission is one of the most important. The study relies on its assertion that excess deaths increased in 2021 following vaccine rollouts and remained high in 2022. The basis of evaluation is clearly implicit: 2020 = Covid + NPIs. 2021 = milder Covid + vaccines + NPIs. 2022 = even milder Covid + vaccines

But here’s the thing. COVID-19 did not bite in any country until at least very late March 2020 as this chart for daily European deaths illustrates

In fact, if the UK is anything to go by, the year started with negative excess deaths meaning that COVID-19 had an initial short-fall to make up

The high excess total deaths in 2020 were actually during 9 months at most A three-quarters year

Even worse for anti-vaxxers, the higher excess deaths during 2021’s first ‘vaccine year’ hides the fact that most people were not fully vaccinated against COVID-19 until AT LEAST May 2021 by which point excess deaths had peaked and fallen to negative

We’ve just been through the worst pandemic in over 100 years, despite everything that modern medicine and public health policies have thrown at it. That’s the cold reality.

In conclusion, this isn’t just a misjudged study. It’s not even a poor study. It isn’t a study at all. The authors have published a Trojan Horse propaganda piece, sans horse, using the well-worn anti-vax playbook of bias, omission and ignore.

And, no, they are not “just asking questions”

--

--

Evidence Challenge

Our World gets Better. Humanist. I write about health, money and climate disinformation