Excess deaths and the elephant in the room - Part 2

From InfoCheckers


As discussed in Part 1, death statistics in a number of countries world-wide suggest that Covid-19 vaccines may be responsible for significant numbers of deaths. A notable exception appeared to be Switzerland, until -perhaps- the end of 2021.

Weekly deaths in Switzerland (source:

The rise in deaths in tune with Covid-19 boosters was commented with alarm in a recent "Coronagate" blog (Google translation here). But can't we just as well blame the Delta wave for that excess mortality? And - in contrast with Germany - little to nothing can be seen from first vaccinations. Did I overestimate vaccine risk in last year's analysis?

Happily Switzerland provides age stratified data, although not fully consistent (interpolations are required for comparing data in 10 year groups with data split in over 65 and under 65 y.o.). This enabled to examine the Swiss data more thoroughly in spreadsheet format, combining excess mortality, Covid19 deaths and simulated Covid19 vaccination deaths for different age groups. File:CH-vax-deaths-analysis.ods. Links to the data are included in the spreadsheets.

The following graphs all show weekly numbers for over and under 65 y.o. just as in the picture here above; the grey dash lines indicate the expected ranges for "normal" years; note that at the end of each pandemic wave less people die than normally (mortality deficit or "early reaping" effect as can be noticed in early 2021).

Weekly deaths in Switzerland in 2021 (incl.first month of 2022)
Weekly deaths with Covid-19 in Switzerland in 2021

By subtracting the Weekly Covid deaths from the Weekly deaths we obtain:

Weekly deaths in Switzerland in 2021 that were not associated with Covid-19

Comparing with the weekly vaccinations:

Weekly Covid-19 boost vaccine doses in Switzerland in 2021 and Jan.2022


1. The peak in excess deaths of 65+ in week 49 is not explained by the Covid-19 Delta wave alone.
2. That peak follows 1-2 weeks on the peak in booster vaccinations of 65+; this confirms that splitting in age groups is necessary (see discussion in last blog post).

In the following simulations deaths occur in the same week as vaccinations [Note added: that is, the same week as second vaccinations and the same week as boosters; but see the PS]. As starting point for simulating vaccination deaths, I took my analysis of last summer. In Switzerland the great majority of people were vaccinated with Pfizer and Moderna vaccines, which I had estimated at respectively 15 and 20 deaths per million average in vaccinated people of less than 65 y.o. That turned out to fit easily with the Swiss death data (see spreadsheet "Analysis" in the ODS file) but slightly increased values fit better. With trial-and-error the following values resulted in a rather good fit between the grey dash boundaries:

Table 1. Fitted average risk to die of mRNA vaccines
Age group under 65 65+
Full vaccination 30/million 300/million
Boost dose 50/million 750/million

In view of the large differences between those age groups, based on the above numbers the simulation was refined for a little more precision (partly making use of finer age group data and partly assuming that older people tend to get vaccinated earlier - see spreadsheet "Refined analysis" in the ODS file).

Taking again the plot of non-Covid-19 weekly deaths,

Weekly deaths in Switzerland in 2021 that were not associated with Covid-19

Then by subtracting the simulated vaccination deaths, we obtain death rates in the normal range (except for the "reaping effect"):

Simulated weekly Covid.19 vaccine deaths, based on assumptions in table 1, refined
Simulation of deaths in Switzerland in 2021 without deaths due to Covid-19 and vaccines

The final graphs for the two age groups without the simulated Covid and vaccination deaths still show a slight amount of excess deaths during that period.

But what if the number of Covid deaths is in fact underestimated? Here below follow the simulations accounting for 20% more deaths due to Covid-19 than reported deaths with Covid-19, without changing the simulated vaccination deaths.

Simulation of weekly deaths in Switzerland in 2021 without deaths due to Covid-19, accounting for 20% more deaths from Covid-19
Simulation of weekly deaths in Switzerland in 2021 without deaths due to Covid-19 and vaccines, accounting for 20% more deaths from Covid-19

In conclusion, vaccinations may explain the non-Covid19 excess deaths - even for Switzerland. Of course there could be other causes of excess mortality that coincided with the boosters, in which case the estimations here are too high. Nevertheless the order of magnitude is likely correct.

But why are the excess deaths after first vaccinations so much less obvious than in Germany? Eyeballing Kuhbandtner's graphs (see last blog post), in Germany for over 65 y.o. the excess deaths seem to jump around 100 deaths per day after a jump in 50'000 doses per day, which means 1 in 500 second vaccinations - much more than simulated here for Switzerland. A possible explanation may be that in Germany also AstraZeneca's vaccine was used; last year I estimated it to be about 5 times more risky than Pfizer's vaccine. If that vaccine was largely used on elderly (please comment if you know!) then that would largely explain the difference.

This analysis matches rather well with my order-of-magnitude estimations of last summer. According to this new analysis, those earlier vaccination risk estimations for persons under 65 years of age may have been slightly underestimated. Further, estimations can now be made for the elderly and for the boosters in the context of new variants. In view of the considerably lower risk of Omicron and the reduced efficacy of those vaccines for new variants, the benefit of additional boosters becomes less clear - even for elderly people, and even accounting for higher infection risk. The following table is purely based on the simulation fit and implies therefore no claim of accuracy or reliability, but just to illustrate the combined effect:

Table 2. Estimation of average risk to die (at 20% chance of infection with an Omicron-like variant, no early treatment)
20-30 y.o. 50-60 y.o. 70-80 y.o.
COVID-19 wave mRNA boost COVID-19 wave mRNA boost COVID-19 wave mRNA boost
8/million 30/million 120/million 75/million 800/million 750/million

PS. As the "boosters" were very concentrated in time, it's not very satisfying to simulate with vaccine deaths in the same week as boosters; medically more plausible is the same and the following week. So I still did a variant simulation with half the deaths in the same week and half in the following week. The result is even slightly better:

Simulation of weekly deaths in Switzerland in 2021 without deaths due to Covid-19 and vaccines, refined

Comments, corrections and suggestions are welcome! For anti-spam, anonymous comments (without site registration) cannot include web links.



20 months ago
Score 0++

Still no better sources than Kuhbandner and a conspiracy theorist blog? Still no attempt to present any possible causality? I still have not heard of a single death caused by vaccination in Switzerland. How about this, another random article: https://www....hkeit-einher Or this one:

If you consider that coronagate blog a reliable source, I find it very difficult to take any of your analysis serious. Clutching at straws, is my impression. I wish I had time to analyse your analysis more closely ...


20 months ago
Score 0++
Thanks. My findings are consistent with those of many others, including several peer-reviewed papers - if I find them back then I will add them in a later comment and also give feedback on your links. I did not suggest that other blog to be a "reliable source" but criticized it for it totally neglects the Delta wave (incredible!); however, it points to reliable data. As I had found no reliable source to answer the questions that I ended with in Part 1, I finally provide my own analysis which is also based on the official Swiss data. And I wonder how an objective simulation can be perceived as "clutching at straws". As a matter of fact, I had collected the data when I wrote Part 1 but not yet done the analysis.

Anonymous user #1

20 months ago
Score 0++

Je n'ai pas bien compris votre " analyse " ...

En fait c'est plutot le terme d'analyse que je ne saisis pas dans son exploitation

Deja "analyser une analyse " me laisse pantois ...disons dans une posture stupide ...

je vais prendre donc le temps d'analyser votre analyse de cette analyse ( ou construction ?)

J'espère bien qu'en poursuivant cette voie ensemble avec esprit analytique ( sans se détourner des ressorts usuels de philosophie analytique ) et de la passion ardente d'analyse d'un Columbo / Sherlock / et autres Hercules de la pensée d'investigation

( au contraire d'un inspecteur Lestrade ou Mr 100% ... deduction hâtive ... c'est bien mal aimer sa maîtresse (en ) Vérité ...

(comprendra qui le voudra avec un peu d'effort ... suivant le critère : qui comprendra rira )

je disai donc pour terminer cette ebauche d'analyse d'une analyse d'analyse , ... qu'en poursuivant ensemble avec ardeur , on finira par souscrire à une analyse Freudienne / Lacanienne

( il nous faudra entre autre quelque analyste logiciel pour les besoins spéciaux de ce cas bien singulier ...)

Et surtout... Sketch Toujours !


20 months ago
Score 0++
I was not fully satisfied with the timing precision and decided to add one more refinement (more realistic) as a PS. A bit to my surprise, the result is even smoother.


20 months ago
Score 0++

As promised in my earlier reply to Lausianne here links to some other scientific articles on this topic as well as comments to his links:

My comment: peer-reviewed paper that is fully consistent with the later result that with the mRNA vaccines slightly more people died in the vaccinated groups than in the control group).

My comment: peer-reviewed paper with at first impression much higher estimations than my own - and the chief editor demanded a new review, I suppose that he also doesn't believe it. I was waiting for the result of that but maybe I'll scrutinize it myself to find the cause.

My comment: it looks high quality, by a Professor of life sciences but not (yet) reviewed (and difficult to verify?). The results are consistent with my own.

And then the two "random" ones of Lausianne:

My comment: quite interesting, looks similar to the other German analysis that I mentioned in Part 1, but with a different outcome and it's also just a preview. I mentioned in Part 1 the great importance of the time frame. Here they looked at weeks 1-48 which includes a significant Covid-19 wave when the most vulnerable people were vaccinated. Based on my analyses I find it unsurprising if states with more vaccinations also had less excess mortality. However their results do not enable to assess the risk of vaccination compared to that of Covid-19 for different age groups, and - and in week 48 the boosters had hardly started, so there's nothing about that either. Also, they didn't account for the excess deaths except for stating that "The study suggests that excess mortality can be explained, at least in part, by COVID-19 cases and that vaccination prevented infections or caused a milder course".

The other link leads me to an article of 2020, likely it was a copying mistake.

Frankly, I don't understand the idea behind a superficial look at random articles except for estimating the range of opinions - and we found huge disagreements of opinions, from totally safe to totally unsafe, often relating to subtly different questions. That's quite useless in my opinion. And when we see that data is available that has not been sufficiently well analyzed by others for our purpose then we should do it ourselves; Google cannot give a reliable answer.


20 months ago
Score 0++

A precision: when simulating with 20% more deaths due to Covid than reported "Covid deaths", that does not mean 20% undetected Covid deaths. Instead, it means that the number of undetected Covid deaths is supposed to be 20% more than the number of people who were reported to have died with Covid but who in fact did not die from Covid (people who come in and die not long after a car accident but were diagnosed "Covid-positive", are counted as "Covid deaths").

For example, if 20% of the "Covid deaths" were in fact not due to Covid, then 20% more deaths from Covid than reported "Covid deaths" corresponds to a correction of +50% (1.2/0.8=1.5) - a huge correction. Taking the number of reported "Covid deaths" at face value simply means to assume that those two errors are about equal.


14 months ago
Score 0++

In line with this and earlier discussions, according to a new peer reviewed publication:
"In the Moderna trial, the excess risk of serious Adverse Events of Special Interest (15.1 per 10,000 participants) was higher than the risk reduction for COVID-19 hospitalization relative to the placebo group (6.4 per 10,000 participants). [3] In the Pfizer trial, the excess risk of serious AESIs (10.1 per 10,000) was higher than the risk reduction for COVID-19 hospitalization relative to the placebo group (2.3 per 10,000 participants)."



11 months ago
Score 0++

Update: now a Dutch website made a model for the Netherlands. They found a good fit for 0.09% deaths per boost injection; for a fully boosted individual they arrived at an estimated total risk of about 1:250.
Their estimations for the Netherlands may sound worse than my own estimations for Switzerland. However, they did not account for age groups while I modeled vaccination risk per age group. By far the elderly were most vaccinated.
For the group of 65 y.o. and above I found a reasonably good fit at 750/million deaths per boost or 0.075% - not far away from 0.09%.

With help of Google Translate you can inspect their model here:

Add your comment
InfoCheckers welcomes all comments. If you do not want to be anonymous, register or log in. It is free.