Actually, I do not get it why it is confounded by the age of the vaccinated:
you estimate the correlation between the vaxx rate (R_V) in a country (# vaxxed/ # population) and the covid mortality rate (R_M) (# covid fatalities / # population)
If the vaccines were preventing deaths it should show a negative correlation. If you use pl…
Actually, I do not get it why it is confounded by the age of the vaccinated:
you estimate the correlation between the vaxx rate (R_V) in a country (# vaxxed/ # population) and the covid mortality rate (R_M) (# covid fatalities / # population)
If the vaccines were preventing deaths it should show a negative correlation. If you use placebo, or you jab healthy young adults who almost never die of covid then there should be no correlation so the estimate would be around zero. But positive correlation of 15% is a lot! Got the confidence interval as well?
You could argue for a confounder like: the more developed nations have older and more obese populations and hence show higher covid mortality. In the same time being richer allows them to jab more people, so here you would get a confounded correlation.
I guess you could get rid of it by measuring the correlation between the time increments of the vaxx rate R_V (same country but say monthly vaxx rate increments) and the corresponding case-fatality ratio increments. You need to use the CFR because the covid seasonality would override everything anyways. That would be basically the gradient of how the CFR moves as a function of the vaxx rate.
Actually, I do not get it why it is confounded by the age of the vaccinated:
you estimate the correlation between the vaxx rate (R_V) in a country (# vaxxed/ # population) and the covid mortality rate (R_M) (# covid fatalities / # population)
If the vaccines were preventing deaths it should show a negative correlation. If you use placebo, or you jab healthy young adults who almost never die of covid then there should be no correlation so the estimate would be around zero. But positive correlation of 15% is a lot! Got the confidence interval as well?
You could argue for a confounder like: the more developed nations have older and more obese populations and hence show higher covid mortality. In the same time being richer allows them to jab more people, so here you would get a confounded correlation.
I guess you could get rid of it by measuring the correlation between the time increments of the vaxx rate R_V (same country but say monthly vaxx rate increments) and the corresponding case-fatality ratio increments. You need to use the CFR because the covid seasonality would override everything anyways. That would be basically the gradient of how the CFR moves as a function of the vaxx rate.
The same analysis with total mortality instead of covid mortality should also show the increasing 'post pandemic stress disorder ' fatalities (https://boriquagato.substack.com/p/i-see-we-have-now-reached-the-cover)