What is the (under) reporting factor for Covid vaccine deaths?

Eyal Shahar
4 min readApr 8, 2023


To estimate the number of Covid vaccine fatalities, we must estimate the reporting “rate” (<<100%) and multiply the count of reported deaths by the inverse of that factor. For example, if only 10% is reported, the true number would be 10 times the reported number.

The computation, however, is highly sensitive to our assumptions about the reporting factor.

Under certain assumptions, a reporting factor of 10% of all Covid vaccine deaths rather than 20% doubles the number of such deaths in the US — from 60,000 to 120,000.

Recently, I argued that a substantial proportion of Covid vaccine fatalities had been reported and considered a reporting fraction of at least one-quarter:

Reports from different countries and different cultures have returned a narrow range of the fatality rate (2–7 deaths per 100,000). They cannot be wrong by a similarly large factor (x10, x20, x40). Otherwise, we would have to assume, for example, that Sweden and the US share the same huge under-reporting bias (e.g., only 10% of vaccine deaths are reported in each country.)

Here, I will try to show that an estimate of 10% is implausibly small in Sweden.

Why Sweden?

First, having analyzed and published their data since the beginning of the pandemic, in English and Hebrew (here, here, here, and here, for example), I am well-familiar with the data.

Second, I was impressed by the quality and transparency of their data.

Third, in my records, the rate of reported Covid vaccine fatalities in Sweden is similar to the rate in the US (5 per 100,000).

Parts of a recent report of adverse effects (Pfizer vaccine) in Sweden, using Google Translate

Mortality in Sweden: October 2020-September 2021

For reasons that are explained elsewhere, we should compute annual mortality statistics between October and September (“flu year”), rather than between January and December (Gregorian year).

A total of 93,642 people have died in Sweden, from all causes, between October 2020 and September 2021 (plus ~250 with missing dates). Based on linear trajectory, excess mortality during that period was around 3%, or about 2,700 deaths.

Much of the excess mortality was concentrated in two winter months –December 2020 and January 2021 — as clearly seen in the table (red rectangles added).

Vaccination and reported deaths in Sweden

Only 250,000 people were vaccinated, at least once, by the end of January 2021. By September 2021, the number was 7 million — about 90% of the total count.

According to my previous data, there were 363 reported vaccine deaths in Sweden [see table here]. (The total from recent reports, for three types of vaccines, is 427.)

Let’s assume the following:

· Only 300 deaths by September 2021

· Only 70% were vaccine-related deaths

· The reporting rate was 10%

On these assumptions, the true number should have been 2,100 excess deaths due to Covid vaccines alone.

That’s impossible to reconcile with all-cause mortality statistics. As I wrote above, the number of excess deaths in the entire flu year was about 2,700, and most of the excess was concentrated in December 2020 and January 2021, before the vaccination campaign. A 10% reporting rate of Covid vaccine fatalities is incompatible with excess mortality in Sweden between October 2020 and September 2021.

If Sweden and the US share a similar rate of reported Covid vaccine fatalities, 10% reporting is implausibly small for the US as well.

Nothing of what I wrote here change my views on Covid vaccines, their questionable effectiveness, their unacceptable rates of adverse effects and short-term fatality, and their unknown long-term morbidity and mortality consequences. Even 15,000 short-term fatalities in the US are an unacceptable figure, and the true number might have been close to 40,000.

Just imagine how the public and the media would have reacted if a new flu shot was implicated in thousands of deaths in the US.



Eyal Shahar

Professor Emeritus of Public Health (University of Arizona); MD (Tel-Aviv University, Israel); MPH, Epidemiology (University of Minnesota)