[conspire] (forw) Re: [skeptic] US 1 million: this year or next?

[Rick Moen]

As long as I'm forwarding related comments, there's this.


---<begin forwards>---

Date: Thu, 12 Aug 2021 16:26:27 -0400
From: Michael Sofka 
To: Skeptic <skeptic at linuxmafia.com>
Subject: Re: [skeptic] US 1 million: this year or next?

On 8/12/21 3:16 PM, Rick Moen wrote:

> Thus, there's an automatic "detection bias". Second, as vaxed
> population has increased, especially among people like Cape Cod
> holidayers, you get the Kristiansund-blonds problem: Vaxed persons
> necessarily will then constitute a rising portion of cases.

Yep, I've been trying to explain this the past week.   Consider two
islands, both with a population of 1,000,000.  Island A has a 0%
vaccination rate, while island B has a 100% vaccination rate.  So by
definition 100% of those on island A who contract Delta Covid-19 are
un-vaccinated, while 100% of those who contract Delta Covid-19 on
island B are vaccinated.

But....

On Island A that's 800,000 people (assuming Delta has an R of 6,
closer 900,000 if it really is an R of 9, or that of chickenpox*),
while on island B it is 80,000 people.  Further, on Island A that's
8,000 to 80,000 deaths, depending on the mortality of Delta---likely
the higher number because with 80% of the population getting sick
there will be a collapse of the health care (and pretty much every
other) system.  Meanwhile, the vaccinated population of B has a very
bad super-bug season, but manages under 1,000 dead and maybe way under
1,000 if the seriously ill get good health care, and precautions
against spreading Delta, such as masks, isolation of the sick, and
quarantine of the exposed, are followed.


* The herd immunity level of an illness is 1-1/R.  For the original
Covid-19 R was calculated as between 1.2 and 2.8, with maybe 3 as the
highest value, so 1-1/3 or 67% of the population being vaccinated
should be sufficient to prevent epidemic spread.  Of course
sub-populations that are not or could not be vaccinated would still be
at risk and need protection via masks, etc. ** The dropping of mask
mandates was a fingers-crossed effort to encourage continued
vaccinations.  Delta, with an estimated R of 6-9 is a whole other
story.  At the low end, 1-1/6 or 83% of the population needs to be
vaccinated, and at the high end it is 90%.  We are nowhere close.

Those interested in the math can find an excellent derivation in The
Model Thinker, 2nd Edition by Scott Page.  The Model Thinker, BTW,
premises using multiple, different, models which have different
assumptions, strengths, and weaknesses.  The estimates of herd
immunity are based on simple diffusion epidemic models that ignore
things like contact networks, sub-groups, mutating viruses,
differential responses, etc.***

**Which largely explains why fewer older people are catching Delta, as
they have the highest vaccination rates in the country, including in
Florida.

***That was more then I intended to write.

Mike





Date: Thu, 12 Aug 2021 15:21:01 -0700
From: Rick Moen <rick at linuxmafia.com>
To: skeptic at linuxmafia.com
Subject: Re: [skeptic] US 1 million: this year or next?
Organization: If you lived here, you'd be $HOME already.

Quoting Michael Sofka:

> On Island A that's 800,000 people (assuming Delta has an R of 6,
> closer 900,000 if it really is an R of 9, or that of Chicken Pox*),

CDC's recent slideware leaked to papers estimated Delta's R0 as 8.5,
exactly the same as chickenpox and second only to typhoid (11-17) and
measles (11-18) among well-studied pathogens.  Doing the algebra for what
population percent of immunity (natural or vaccine) required to bring
R-effective below 1.0, thus declining spread:

(As you say:)

(1 - (1 / 8.5)) * 100 = 88.23%

I haven't heard 9, but if we're pessimists and assume that bad, then
(again, as you say):

(1 - (1 / 9)) * 100 = 88.9%

If a variant becomes epically bad, as bad as worst-case measles (R0 =
18), then

(1 - (1 / 18)) * 100 = 94.5%

...which is why public health agencies try to ensure herd immunity
against measles by trying to make sure at least 95% of the population
gets the MMR vaccine.

In situations where you cannot up immunity percentage, e.g., not a
super-rich country rolling in unused vaccines as the USA is, medical
countermeasures (such as masking & social distancing for respiratory
pathogens) also lower R-effective, as long as you keep doing them.

(Obviously you know all this, Michael.  Just running through the basics
for collective knowledge.)

---<end forwards>---





So, if Delta's R0 is 8.5, which would require 83.3% of the population to 
have immunity (from infection or vaccination) to reach herd immunity
where R-effective drops below 1.0 and you get decreasing spread,
then how close is San Mateo County, right now?

https://www.smchealth.org/data-dashboard/vaccination-totals-locations-data

Being goal-focussed, San Mateo County Health Dept. emphasises the 
how much of the _eligible_ population have been covered, excluding
children under 12 and the tiny fraction of people who ought not to get
the vaccine owing to current serious illness.  With apparent pride, they 
stress 92.1% of residents age 16 and up have been vaccinated, and 90.7%
of residents 12 years and up.  

The latter approximates "percentage of the eligible", and is pretty
darned good.  But the number we need is the percentage of the _whole_
population, not of the eligible.  

The county dashboard says:  539,417 total individual vaccinated, and 
total county population is 774,990.  So, percentage of the whole county
population:  69.7%

That's not _yet_ the ~88.3% we need, to kick R-eff below 1.




https://calcat.covid19.ca.gov/cacovidmodels/ claims the county's R-eff
(based on data modeling) is currently estimated at:

   1.03  Spread of COVID-19 is likely stable

The accompanying curve shows estimated R-eff having peaked on July 21st
and headed mostly down since then, with a small rise to 1.05 after that 
(coinciding with Delta) but heading mostly slowy downwards towards the
1.0 crossover point that we really want.

Again, to stress, R-eff going _downwards towards_ 1.0 is good in that it
means infection _spread_ is _slowing down_ (rate of additional infection 
is decelerating), but unfortunately is still reaching higher numbers of
infectees.  _Below_ 1.0 would mean a declining rather than growing
infection problem across the population.