Thread by @DellAnnaLuca

VACCINE & COVID TAIL RISK

A few days ago, @NachoOliveras asked, how do the tail risks of COVID and the vaccine compare?

Here's an attempt to answer it on practical grounds.

1/7

2/ A first consideration is that one person that gets COVID can infect others, whereas one that gets the vaccine can't.

It then follows that the best practical solution is to deploy the vaccine on at least part of the population.

3/ Both the risks of getting the vaccine and in particular of getting COVID are not homogeneously distributed across the population.

An 80yo is at larger risk of dying of COVID than a 30yo; an infected restaurant worker can spread the virus more than an infected work-from-home.

4/ It then follows that, regardless of the absolute risks of both the vaccine and COVID, and given that vaccine supply is limited, it makes sense to prioritize distribution of the vaccine based on individual risks and benefits.

5/ As bad as the following sounds, if we deploy the vaccine and discover after a bit that it has side-effects, we can always stop deploying. Whereas we cannot easily stop an out-of-control COVID.

6/ Another consideration is, should the vaccine be mandatory?

I think it's not a question worth debating now. As long as the supply is limited and there are people in high-risk of becoming spreaders willing to take it, it can be made available on a voluntary basis.

7/ Regulators have been slow because they've been deliberating on vaccine's suitability for everyone.

This doesn't make sense, in a world where supply is limited and COVID is spreading.

8/ If a doctor or schoolteacher wanted to get the vaccine in September, say, I'd argue he should be let do it, once informed of the risks. He or she is at risk himself (he sees lots of patients) and is at risk of becoming a super-spreader (if he gets infected).

9/ Moreover he gifts the community data and (hopefully) protection.

10/ It doesn't make sense, in a world where individual risk (of infecting and of being infected) is not homogeneously distributed, to have one single degree of vaccine approval: for everyone or for no one.

11/ That said, would I take the vaccine now?
I'd rather wait a bit*

Does it mean I think the vaccine should not be made available?
Not at all; we waited too long.

The 2 statements above aren't incompatible.
Again, because risks are not homogeneously distributed nor static

12/ (*) Why? Because this winter I'm living an extremely secluded life and I believe that both my absolute risk-benefit and my relative one compared to the rest of the population are low (both from the point of view of risk/benefits for myself and for the population).

13/ Then, of course, over time the risks I talk about will change.

We'll soon have much more certainty about the risk profile of the vaccine. Assuming it's positive information, more and more people should consider getting vaccinated.

14/ By the way, compare the efficiencies of scale of approving the vaccine at the EU level (having to review papers only once) vs the inefficiencies (if one state went first, it would give data and protection to the others).

Again, the costs of deciding "once and for everyone."

15/ These are all tentative thoughts, shaped by what I know and what I don't.

What did I miss?