It's a different comparison, between 2019 mortality and 2022 mortality on a whole population basis, and between 2 groups in 2022 depending on vaccination status.
Consider if there were 1000 people, and in 2019 10 would die, in 2022 12 would die if the crude death rate were to increase by 20%.
If in 2022, there are 1000 people and 800 are vaccinated, at 2019 mortality rates, 8 vaccinated and 2 unvaccinated die. If you had a hypothesis that the increased mortality is associated with the vaccinated group, and the unvaccinated had unchanged mortality, you would need to have 10 die in the vaccinated group, which is a 25% increase on the 2019 mortality rate.
Although it is a different comparison, it isn't actually very different as long as the vaccinated percentage is high, you would only get the big difference between changes to crude death rate, and relative risk of death between vaccinated/unvaccinated if the percentage of people vaccinated were lower, working under the assumption that only the vaccinated's mortality changes.
If vaccination were 50%, you would have 5 out of 500 die, and 7 of the other 500, i.e. the vaccinated group's relative mortality risk goes up by 40%, and the overall mortality risk goes up by 20%.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
branchandroot
Re: death rates
Date: 2023-01-11 12:24 am (UTC)Consider if there were 1000 people, and in 2019 10 would die, in 2022 12 would die if the crude death rate were to increase by 20%.
If in 2022, there are 1000 people and 800 are vaccinated, at 2019 mortality rates, 8 vaccinated and 2 unvaccinated die. If you had a hypothesis that the increased mortality is associated with the vaccinated group, and the unvaccinated had unchanged mortality, you would need to have 10 die in the vaccinated group, which is a 25% increase on the 2019 mortality rate.
Although it is a different comparison, it isn't actually very different as long as the vaccinated percentage is high, you would only get the big difference between changes to crude death rate, and relative risk of death between vaccinated/unvaccinated if the percentage of people vaccinated were lower, working under the assumption that only the vaccinated's mortality changes.
If vaccination were 50%, you would have 5 out of 500 die, and 7 of the other 500, i.e. the vaccinated group's relative mortality risk goes up by 40%, and the overall mortality risk goes up by 20%.
Mawkernewek