Re: [Ietf108planning] Assessment criteria for decision on in-person/virtual IETF 108

[Rich Kulawiec <[e-mail]>]

Jay, Alissa, Colin, everybody:


> A requirement to prove COVID-19 immunity, vaccination or [...]

It's not clear that "COVID-19 immunity" exists or how long it will last.
It's plausible, of course, but given that we are now seeing reports
of recovered/discharged patients being re-admitted in South Korea
with a second COVID-19 infection, it's an open question.

A vaccine is likely 12 to 18 months off unless someone gets lucky and
not only manages to develop an effective one, but one whose side effects
are minor and only affect a small percentage of those to who it's administered.
There will then remain the problem of manufacturing, distributing,
and administering an enormous number of doses -- a problem made more
difficult in some countries (like the US) by anti-science cultists.

In addition, a lack of massive, reliable testing means that there is
a fundamental measurement problem in some countries (again, like the US):
we simply don't know many people are infected but asymptomatic.
Of course there are epdemiological methods/models to estimate this and
they're probably somewhere close to the truth, but even if they are,
those models don't inform individual choices.

Finally: this particular virus appears to be remarkably contagious.
Epidemiologists use the parameter R0 to denote the average number
of people infected by one infected person, and for seasonal flu it's
around 1.3.  The expected number of people infected two hops out
is 1.3^2, or about 1.7; three hops out it's 1.3^3, or about 2.2.
And so on.  Early estimates of the value of R0 for COVID-19 were
around 2.5, but emerging research suggests that it could be much
higher.  This (early release) paper from scientists at Los Alamos
via the CDC suggests a median value for R0 of 5.7.  Two hops out,
that's 32.5; three hops out, it's 185.  Here's the paper:

	Early Release - High Contagiousness and Rapid Spread of Severe Acute Respiratory Syndrome Coronavirus 2
	Volume 26, Number 7 July 2020 - Emerging Infectious Diseases journal - CDC
	https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article

Of course like so many bits of research, this has been rushed out
without full/lengthy peer review.  It could be wrong.  The authors
themselves assign a 95% CI of 3.8 to 8.9, which is a pretty wide range,
especially for a value involved in an exponential calculation.
But even if they're wrong about their estimate of R0 and we use
their lower bound of 3.8: 1.3^10 is about 13.8, while 3.8^10 is
about 630,000.

The implication here is that a single individual -- perhaps asymptomatic
and unaware that they're infected -- could be responsible for a very
large outbreak.  That individual could even be someone who repeatedly
tested negative before traveling to the conference, but picked it up
along the way.  I suggest factoring this risk into the decision
process -- although frankly I'm at a loss as to how to quantify it.

---rsk