Interesting. But I would be weary of confounding variables here… E.g. those who got infected and those who got vaccinated in Jan/Feb were probably pretty differ

Interesting. But I would be weary of confounding variables here… E.g. those who got infected and those who got vaccinated in Jan/Feb were probably pretty different groups (in terms of age, social habits, health, etc). They try to control for that in the statistical analysis, but this is often hard to do well when groups are very different. EDIT: An obvious problem if you are looking at deaths is (literal) survival bias.