[Question] ($1000 bounty) How effective are marginal vaccine doses against the covid delta variant?

UPDATE: the bounty has now been awarded as follows:

  • $425 to Connor_Flexman

  • $300 to Josh Jacobson

  • $150 to JenniferRM/​johnswentworth

  • $75 to ChristianKl

  • $50 to SoerenMind

===

Since I got vaccinated I’ve started working in-person, going to restaurants, travelling and whatever. But the delta variant might change this risk calculation. Some of the same buzz that got covid right early on, the first time around, is now buzzing about delta potentially being very bad.

So I’m exploring ways of responding to that. This time around I’d rather stand up and fight than lock myself in a house for a year. And so this question explores one possible approach.

There’s evidence that two vaccine doses are more effective than one (though there might be much higher diminishing returns than is commonly acknowledged). I’ve also heard that past infection confers marginal immunity even in the presence of vaccination. Further, as far as I understand, the 2-dose schedule for many vaccines is more “a common thing that we happened to test in the big trials” rather than “the dose level at which marginal benefit = marginal risk”. (And giving your population more than two doses is probably much harder logistically.)

Hence: getting more than the standard 1-2 vaccine doses might be a way to protect against delta.

I live in California, and we currently seem to have a large vaccine surplus. I heard a vague rumour about a friend basically just walking into a pharmacy and getting a 3rd shot. There’s also RADVAC, which you can make yourself.

So, if there is supply… how worthwhile is it to get more vaccine doses? Should you get a 3rd, 4th, … or even more? How does getting a further dose of the same vaccine compare to getting the first dose of a different vaccine?

I’m posting a bounty of $1000 for answers that change my mind on this question (maybe increasing to many times that if this proves valuable).

The ideal thing I want would be a graph with the x-axis showing # doses, and the y-axis reduction-vs-control of the following four parameters:

  • Symptomatic infection

  • Hospitalisation

  • Death

  • Long covid

(Answers are of course impacted by the combinatorial search space of dose spacing /​ dose number /​ dose size /​ vaccine type /​ various demographics, but I won’t make any special restrictions here for now)

I will pay at least $1000 for answers that help me get clarity on this, split at my full discretion in proportion to how useful I find the answers. I’ll pay out the bounty on a rolling basis as answers come in; there is no deadline. In case this proves fruitful and there seems to be useful marginal work, it’s possible I will increase the bounty a lot (i.e. up to many thousands of dollars), or reach out to work with some answerers as contractors.