To convert an Anti-Vaxxer
For those interested, I have converted anti-vaxxers simply by showing them the FULL effect of Covid. Because if you look at just the individual risk, for anyone under 50-60, the only sensible response is "rats arse". Couple that with public officials screaming "PHEAR!!" and an angry response/dismissal of the whole rigmarole is entirely justifiable.
The eye-widening comes from looking at the whole-country impact. For some reason, no public official I've seen/heard has done more than wail "PHEARRR", has not simply laid the numbers out. And the numbers talk.
To do it: just follow the methodology laid out in Australia's federal modelling by the Doherty Institute. ("Interestingly", they hid all the following numbers.) Get the risk numbers from Knock et al (the best to date), p65 Table S9 in the Supplementary Materials: https://www.medrxiv.org/content/10.1101/2021.01.11.21249564v1.full . That's per Infection, not per Case, so you'll need to back out the Asymptotic assumption of 0.60 (p62), re-apply the age-bucketed figures from Davies et al, p14 in the PDF (https://www.nature.com/articles/s41591-020-0962-9), then if you like scale up for Alpha severity which gets you closeish to Delta (Delta is less lethal): get the scaling figures from Table S1 p23 of Doherty (https://www.doherty.edu.au/uploads/content_doc/DohertyModelling_NationalPlan_and_Addendum_20210810.pdf): +42% Hospitalisation, +61% Death.
Now you have your risk numbers.
Then get your population numbers by age-bucket and simply multiply them out, to get the total country-wide effects. This assumes that everyone will eventually get Covid. Which is the opinion of every sane epidemiologist/virologist: it's not IF, it's WHEN. Just a timing issue. For Australia, p29 of Doherty has them.
So for Australia, you'll get ~460,000 Deaths and ~2,600,000 Hospitalisations. (America's about 6.4m deaths iirc)
That's the equivalent of over 1,100 years of flu deaths.
That's enough to have the antivaxxer doing a serious doubletake, if not start talking about getting the vaccine.
Then do up a table of those Hospitalisations vs Hospital Bed Capacity. For Australia, Doherty have 55,440; UK has ~140,000 (down from 300,000 30yrs ago; NHS policy now is to shut down any hospital running at less than 85% capacity. This is great for generating headlines like "NHS nearly 90% full!" any time there's a hiccup, and for diverting money away from doctors&nurses to more deserving types like middle management). In my State of Queensland, our hospitals have been so run down that we're near 100% capacity just under normal conditions: we have ~0 Covid right now but a number of hospitals running at 105%, with Code Yellows becoming routine.
So then, Timing. How long will it take to reach that number? Doherty's projections imply 2.5yrs for Delta @ 50% vaccination if no lockdowns, masks, home isolation, etc. So run up a table showing that number over that many weeks at various time-spent-in-hospital against the number of hospital beds. A yank Covid nurse told me they're seeing 3 weeks per patient being the average, if you just want to do one number not a table.
For Australia, that implies our hospitals running at 109% capacity. Just for Covid. On top of all the usual workload.
.
(On the upside, as of a few weeks ago, Australia's oldies-first vaccination approach + rollout had dropped the Max.Deaths by 75%. Due to 99.5% of the deaths being in the Over-50yos.)
--
Other things to throw in:
* Delta looks less lethal but in Australia is Hospitalising ~20% of the UNvaccinated. So you should probably ~double the above Individual Risk numbers for Hospitalisation.
* ICUs blow out LONG before Hospitals. Average is about 8-10days in ICU iirc. Australia is currently seeing about 20% of our Hospitalised ppl need ICU, so you can whack that number in as a guesstimate and calc the result vs 8-10days duration per ICU Case.
* Point out to your antivaxxer that once we get past 100% capacity in hospitals/ICUs, for everyone else, that %Hospitalisation is death. So after a year or so of exponential growth rather than the bodgy linear you've got in your calcs above, add %Hosp+%Deaths to get actual %Death risk. So 60yos will literally be playing russian roulette every time they step outside: 1 in 6 chance of death if they bump into an infected person, even an asymptotic kid.
And in my experience, it's about this time even your rantier anti-vaxxer drops the anti.
Biting the hand that feeds IT
