Thread by @contrarhetorics

Short answer: Nonstop, simultaneous, year-round worldwide demand & production + procurement limits is the main reason why I opted for a standard 10,000. 1/45

https://twitter.com/SirMungus/status/1318024677453975552

Long answer: 10/K a day is an abstract guesstimate. It doesn't have any particular mathematical rigor to it. But it wasn't arbitrarily arrived at. I thought long & hard before settling on it. 2/45

First factor I considered is that it's an entirely new vaccine so it has to be produced in addition to the yearly shots of rubella, flu, measles, mumps, etc. The expansion of capacity also has to maintain the existing production of vaccines. This in itself raises a worry. 3/45

Given how fiercely concerned The Experts© have been to ensure health outcomes are preserved comprehensively along all health vectors, I'm genuinely concerned about what might happen with this when/if a COVID vaccine does go into production. But I assumed it remained stable. 4/45

I then turned to look at what the actual current estimated manufacturing power is for pandemic vaccination. I used the flu shot stats as my measure since capacity there is the largest. 5/45

Note for what follows that there are 2 factors of importance in this measure: the actual seasonal flu vaccine capacity & the projected production capacity for a flu vaccine in a pandemic. 6/45

The global flu pandemic vaccine production capacity in 2019 was 6.4 billion for the full manufacturing year. It's quite incredible. Back in 2006, when the Global Action Plan to expand flu vaccine production was initiated, the pandemic production capacity was 1.5 billion. 7/45

So off that, we look remarkably set. However, what allowed the projections about pandemic vaccine production to increase so quickly was not only the expansion of real capacity but also a shift toward the production of quadrivalent flu vaccines. 8/45

A quadrivalent flu vaccine is one shot to protect against 4 different flu strains. It radically economises the outlay on separate doses & enhances efficiency in manufacturing output. But it's also why we need in C19's case to cut back the projected figure. C19 is not a flu. 9/45

The seasonal flu prod. capacity is 1.5 billion doses. In a flu pandemic, quadrivalent vaccines mean that you don't need to keep devoting capacity to 1.5B additional doses for seasonal flu while fighting the pandemic. Even if they're different strains, one vaccine covers it. 10/45

In other words, with the flu, you can produce 6.4 billion because you're incorporating in that, for that year, the 1.5B seasonal flu doses you'd otherwise be giving. With C19, the production lines will still be occupied to some degree with making those seasonal flu doses. 11/45

It should be stressed at this point that I'm operating on a best case assumption for vaccine manufacture that the same technology used in flu vax production ends up the same for a C19 vax since it has the largest world capacity. 12/45

This measure is not for worldwide capacity of *any* viral vaccine requiring different tech/production. Alternative vaccine production & technologies are liable to require more involved and radical logistical hurdles to requisite economies of scale. 13/45

See this article for a sense of the different vaccine techologies & challenges. "One big challenge in creating a lot of vaccine quickly is scaling up manufacturing, because the infrastructure needed will differ depending on the vaccine type." nature.com/articles/d4158… 14/45

So I'm being overgenerous here in translating a capacity measure that's flu predicated seamlessly to C19 production. But I decided to allow for that & also to assume a multiyear radical plunge in seasonal flu that frees some of the 1.5B capacity. 15/45

The next factor I considered is a key point @sdbaral made: that the 1st gen vaccines will not be for all. They will not be cleared for under 18s bar for safety assessment (if that). So I weighed minors as only eligible for vaccination a year on from 1st gen development. 16/45

More crucially: the immunocompromised themselves are also excluded from any first gen vax so I delayed them too. (This, incidentally, is also why "vaccinate the vulnerable first" won't work - they aren't eligible for vaccination in the first round for solid safety reasons). 17/45

(As an aside, this also hits on a core reason to suspect a decisive vax will take some time: the elderly have a significantly less robust immune response to vaxing & they're the most impacted by C19. Worse, because of the lie that this is a generally endangering virus... 18/45

... the odds of the usefulness for the aged of the vaccines underway have been crimped. See below. It is indeed too late now to go back & put a priority focus on the elderly in preclinical tests. 19/45 theconversation.com/why-vaccines-a…

Not because the evidence wasn't there from Day 1, though. It was. We can thank the even now still shamelessly ongoing panic merchantry of The Lifesavers®for this.) 20/45

So, from staggering who could be immediately be eligible, I then looked at other factors. For instance, the WHO estimate that up to 50% of vaxs are wasted globally every year, mainly due to lack of temperature control & the logistics of keeping up an unbroken cold-chain. 21/45

"At the scale of COVID-19, this spoilage rate could waste potentially a billion vaccines, which, even if valued at a non-profit cost of around $10 a vaccine, represents a staggering wasted investment." 22/45 unenvironment.org/news-and-stori…

This wouldnt have been an issue for rich countries in their supply chains before. But see my tweets on the collapse in global cargo logistics: "the rolling crisis of back-jamming in distribution" thanks to "COVIDsafe" inefficiencies & national quarantines the world over. 23/45

For the next factor, I then weighed in the question of the realities of the bureaucracies of production. I'll cite from a 2018 article by Kaitlin Schroeder on global challenges in flu vaccine supply: "Making the initial batch, filling numerous types of syringes, ... 24/45

"...packaging each product, and labeling them in multiple languages is only the beginning of the endeavor; companies must also worry about licensing their products in various countries and markets, establishing cold-chain delivery systems..." 25/45

"...& facilitating global distribution. In order to avoid producing an excess of vaccines (and suffering major financial losses), some manufacturers produce well below their capacities, aiming to meet “just below” demand. Especially when a flu epidemic worsens..." 26/45

"... in the latter months of the winter season, the delayed demand is not met with adequate supply—this may have dangerous consequences. The manufacturers have the freedom to produce as many or as few vaccines as they wish, according to their business interests; ..." 27/45

"...luckily, bulk purchase assurances from organizations like UNICEF help to stabilize the annual demand and prevent major shortages of vaccine supply." 28/45 ojs.stanford.edu/ojs/index.php/…

Relevant here is not just the multiplying factor of broken supply chain logistics for products integral to but beyond the vaccine itself (e.g. needles, packaging) or licensing legalities that'll delay upfront between production & initial distrib. to each nation... 29/45

...but also that bulk purchase *of the right vaccines* will be crucial to prevent shortages. The pressure on nations to enter into advanced purchase agreements of vaccines, out of fear of being bought out of scarce product, is high. But remembering the open question ... 30/45

... of the usefulness of 1st gen vaxs for the aged, & keeping in mind that there are 160 candidate vaxs, there's no means to be sure of whether candidates signed up for will be safe/effective. Moreover, even safe/effective vaccines are a work of compromise ... 31/45

... in what groups they work best on & which they work more poorly on. There's a perverse incentive structure for rich countries, especially, to overbuy a vax that's more effective on the young & end up underbuying a vax that's more effective on the old. Or vice versa. 32/45

So, you could have too much vaccine *and* a shortage. The problem isn't only in whether there will be a vaccine; it's also in whether there will be a bevy of them -- especially in the wake of an initial breakthrough. 33/45

Finally, the last factor was that I decided - improbably - to assume that there would not be stockpiling/hoarding by rich countries of a vaccine, that their noises about global equity in sharing a vaccine were on the level. 34/45

I set this within the frame of what they themselves deem feasible. Their internationally set production goal of the WHO's “COVAX” initiative is to deliver 2 billion doses of a COVID-19 vaccine by the end of 2021. Notably much, much lower than the 5.9B for flu vax capacity. 35/45

So 10k/day = a guesstimate as to an abstract daily standard assuming full equitable shares worldwide. Of course, that wont happen. A country like Australia has the wealth to self-prioritise. Its vaccination timescale thus shortens. But at the expense of the poorer nations. 36/45

That's probably part of why they love the idea of "safe nation" travel bubbles. It involves a logic of "reopening" in & between the rich countries while providing one of the most incredible reasons ever for drawing a ongoing quarantine with the poor ones. 37/45

So, the fabulous numbers put forth about how fast vaccination goes - 900K/day (US), "16.2 million Australians immunized in our last flu season!" - are true but need to be read next to numbers like this. In the most recent/comprehensive study we have from last decade ... 38/45

... "[T]wo-thirds of the participating countries did not have enough vaccines to cover 10% of their national populations. Worse yet, more than one-third of the 157 did not distribute enough vaccines to cover even just 1% of their populations." (citing Schroeder) 39/45

This lends some sense of the true picture *globally* that is the flipside of acceleration of delivery to rich nations. So, 10/K a day is an abstracted average I came to that weighs against all the factors above & divides by the number of countries in the world... 40/45

... as though the vaccine were being distributed equally. I will admit I took a liberty with The Lifesavers® on this as I assume that they would be aghast at the notion that Australia should accelerate its immunisations at the expense of the poorer nations. 41/45

5.5 years is a good guess at what a just distribution entails. The major caveat is less that it's too low because of what our current production shows than that I'm assuming a static picture so that 10K/day as an abstract average remains 10K/day for the whole 5.5 years. 42/45

If 2 billion doses were attained by 2021, the next year alone could well deliver twice that. Quite honestly, though, I doubt that 10K/day is too much out of the ballpark for what it would look like to have an equal Oz share in a fair admin of vaccination(s) to the globe. 43/45

If nothing else, consider that in all of this I've been operating on it being a given that the vaccine would require one dose. It's more like to require at least *two* (although perhaps not for all ages). Throw that in & all the factors compound. 44/45

But if I had to predict on the more realistic basis of the rich nations hogging supply, I'd suggest -- if they truly achieved that 2 billion by the end of 2021 -- Australia's daily vaccinations might rise to 30,000 from 2022 onward. So 5.5 years becomes 2.5 or thereabouts. 45/45