A useful rule of thumb for the vaccine is 10,000 dose-days : 1 life saved.
So if you hypothetically had a million doses of a vaccine sitting in the freezer, and you decided to keep them there for a week rather than injecting them, that's something like 700 lives.
This is sensitive to, among other things, the progress of the vaccination campaign (dose-days save exponentially more lives early when they're being administered to seniors than when they're being administered to the general population), presumed infection rates, etc.
But some calculation like this is very useful to remind us that there is a cost to waiting, and the cost for waiting on this particular topic, at this particular moment, is so wildly disproportionate to the costs to waiting we typically endure as to stagger our imagination.
This also means that we should fight like hell for every dose-day, at every point in the supply chain, at every point in the patient experience, etc.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
It is Day 100 of VaccinateCA, and now Day 1 of the Vaccinate The States. We’re a non-profit collecting and publishing the best public data set of coronavirus vaccine availability across the U.S.
We work with the federal government, some state and local governments, and partners in the community to collect, verify, and distribute information about where to get the covid-19 vaccine. We back official efforts & some of the largest publishers in the world, e.g. Google Maps.
We will increasingly act as a data broker between other community efforts, the government and provider vaccine data systems, and publishers. We are uniquely resourced and experienced in the ecosystem to verify ground truth.
Day 98 at VaccinateCA. By what I swear is coincidence, I think we'll have something fun to show off on Day 100.
And since I don't want to be a total tease, we took the embeddable map now backing e.g. Alameda County's health department and localized it into all eight languages we support.
We welcome government health officials, politicians, community organizations, and other to use it to help accelerate the vaccination campaign in your community.
The first day of US's manufacturing capacity being redirected to vaccinate a new county ex-US will save +/- 1,000X more lives than the last day of the US's manufacturing capacity being reserved for the US.
I think the ethics of the situation are reasonably complex, but the useful takeaway from this should be "If you are waiting to get the vaccine because you perceive low personal risk, please get it *as soon as possible* to free up the spot for someone with much worse risk."
(We've had a great, great deal of suffering this year because people had a model which suggested "If I don't take this shot, someone else will get it" where that decision lead to others being delayed, for example during the initial phase of vaccinating healthcare workers.)
A lot of our innovations for VaccinateCA are *extremely* low tech. One that surprises folks: “web bankers.”
There are thousands of places vaccine information can be, including county FaceBook pages, the personal Twitter feeds of elected officials, in mind of pharmacists, etc.
We use a material amount of scraping, but we’re rate limited on how many scrapers we can write, and a county dropping a PNG of their PSF announcing a clinic to their Facebook page is not easily scrapeable.
Enter the web bankers, called that to distinguish from our phone bankers.
When we identify a page/FB page/Twitter/etc which is likely to have interesting updates on it in future, we put an automated watcher on it. When it changes, that goes into a queue for the web bankers.
They go read it, manually compare against what it had last time, and type up.
Similar jobs to be done (active trading as an entertainment product); good insight on how traditional brokerages with risk appetite would think of fees.
Also fun to hear people from very different worlds allude to very foundational views that they treat as axiomatic but that their counterparties don’t understand at that level.
Most prominently, when asked about the risks of crypto, brokerage said “What do I care? Cash product.”
Meaning: because they do not offer leverage on crypto, they do not have to do the risk management they have to do on options trades, leveraged stock positions, futures, or shorting stocks.
They never take credit risk regardless of market conditions or user behavior.
There are (and I know this is surprising to many peers) a lot of people who are acutely sensitive to complexity/difficulty/mental effort required to get through things.
We understand this in B2C conversion optimization and industry spends $$$$.
Meanwhile, the process of getting a vaccine is largely not optimized from a conversion perspective.
Indeed, partially through underinvestment and partially *as an intentional rationing measure* the process is often pessimized.
Sometime after the pandemic is over, when writing about what went less than awesome, I want to write about the game theory of having N distribution systems with fungible supply, overwhelming demand, and the overwhelming incentive to *not* have the best site.