RE: GSK/VIR to fast-track landmark Covid drug14 Mar 2021 12:34
You are absolutely right to point out that no vaccine works 100%. As a vaccine is not 100% then one would expect some vaccinated people to die or to become seriously ill. One would also expect older people and those with underlying health conditions die at higher rates. Those are the sorts of people who live in care homes. If, say a vaccine reduces deaths by 80% then one in five of those deaths will still occur.
There are random variations in the ages of inhabitants of different care homes, and variations in the severity of their comorbidities. And vaccination might give residents, staff, and relatives a false sense of security and a change to more risky behaviour. There are random variations in death rates. So one would expect variations between areas, and care homes.
Post hoc choosing of the worst data points is misleading. It is like looking at the whole market of companies, and you at the end of the day getting to pick the ones where the SP dropped most. [this is just an analogy to illustrate bad statistical process in the presence of randomness - the randomness of markets is different (in theory Mandelbrotian because of feedback) to the normal distributions observed in the randomness of death]
There is also the possibility that the vaccines, possibly in combination with age and one or more comorbidities, are less effective, or have adverse side effects, that evidence so far has not indicated. And I do think people are a tad too complacent, and have been in a rush to focus upon the positives. People want to believe what they want to be true.