Looking at Pfizer’s vaccine alone in the table, totaling ischaemic and hemorrhagic stroke gives 203 excess cases per 10 million vaccinated people. Although the same database was used, ischaemic and hemorrhagic don’t usually occur together, so I think totaling them is reasonable.

For the two other adverse events (i.e., arterial thromboembolism and CVST) associated with Pfizer’s vaccine in the table, the studies did not compute their excess deaths. I think it’s because the risks of arterial thromboembolism and CVST were insignificant in the 28-day period, and the excess deaths were based on 28-day period too. Arterial thromboembolism and CVST were only significant at 15–21 days only for unexplained reasons. But I still decided to put their risk numbers in the table because 15–21 days is still within 28 days in my view.

That said, myocarditis, a known risk of Pfizer’s mRNA vaccine, almost never resulted in death. So I think judging the excess deaths of Pfizer’s mRNA vaccine based on ischaemic and hemorrhagic stroke is reasonable:

• Ischaemic stroke: 143 excess cases per 10 million with a 14.9% death rate gives 21.3 excess deaths per 10 million people.
• Hemorrhagic stroke: 60 excess cases per 10 million with a 23.8% death rate gives 14.3 excess deaths per 10 million people.
• Totaling them gives 35.6 excess deaths per 10 million people. For 220 million, this means 35.6 x 22 = 783.2 deaths.
• Accounting for other potential adverse events that might also result in death, I think rounding off to 1000 deaths makes sense (which is 0.005% out of 220 million).

Since Moderna’s and Pfizer’s mRNA vaccines are similar, and they form the majority of vaccinations in the U.S., I also estimate 1000 covid vaccine death in the U.S. (edit: which is still 800-times lower than covid-19 deaths, standing at over 800k in the U.S. alone.).

As you know, this is just an estimate that could be wrong. One important note is that I extrapolated the death rate from their study paper’s result section. The studies did not do a formal statistical analysis on deaths, however.

But I just use what I got. Of over the 200 papers I went through, I only found one study that analyzed vaccine mortality, which is the Xu et al. paper using VSD, finding that vaccines are actually protective against non-Covid-related deaths compared to no vaccine. But I think I’d rely on the U.K. data more than the VSD. I hope I’m clear but just let me know if anything is not.