Thursday 21 May 2009

A good risk analysis of Swine flu

A lesson on probability.

"low-probability event plus a high-loss outcome = cause for concern."

A lesson of selfishness.

"A one percent chance of one hundred million deaths is, in expected
value terms, one million deaths and that is a big deal. Probably the
United States is less vulnerable than it was in 1918, but how many
people would die in China, India and many other locales? How much
disruption to trade, travel, and the world economy would take place?
Even in the United States, our public health systems would break down
quickly and render many modern medical advances useless (e.g., when
would the Tamiflu run out?)"


http://www.huffingtonpost.com/rj-eskow/the-meaning-of-swine-flu_b_205793.html

The Meaning of Swine Flu, the Universe, and Everything
digg Share this on Facebook Huffpost - stumble reddit del.ico.us
ShareThis RSS
Read More: Ban Ki Moon, Cdc, Epidemic, h1n1, Pandemic, Swine Flu,
Swine Flu Pandemic, World Health Organization, World News

Be the First to Submit
This Story to DiggBuzz up!
Get Breaking News Alerts
never spam

*
Share
*
Print
*
Comments

Wait a second. First we couldn't get away from that swine flu story,
although we were told that it had affected less than 100 people in
Mexico and only a handful here in the US. It was getting round-the-
clock high decibel coverage on all channels. Twitter, our new
electronic central nervous system, was going ballistic. Andy Borowitz
seemed to be on to something when he wrote, "CNN Warns Swine Flu Could
Continue Through Sweeps."

By May 7 the Associated Press and others were writing stories about
the overhyped flu. It seemed resolved in a lot of people's minds: Just
another overplayed media story.

But wait. Now the World Health Organization is reporting nearly ten
thousand cases of the virus worldwide, with more than half of those
occurring in the United States. "We are not out of the woods," a CDC
director said (on CNN.) 16 schools have been shut down in New York
after an assistant principal died of H1N1. Cases are "ballooning fast"
in Japan and elsewhere, says Forbes. UN Secretary General Ban Ki Moon
alled for international solidarity, which includes ensuring that "all
have access to drugs and vaccines."

We went from apocalypse to afterthought in about two weeks, and now
we're swinging back. It's hard to know what to think. Here's a way to
reconcile some of the different and seemingly contradictory
perspectives we've been hearing, which include:

It's no big deal. The CDC estimates that 36,000 people die each year
as the result of regular flu. Very few people have died from swine
flu, but it's all people can talk about. It's an overhyped phenomenon,
a distorted perspective caused by media sensationalism.

It's an emergency. An assistant principal dies in New York. 103
students at nearby schools develop flu-like symptoms. Then a toddler
dies in Queens. Forget the statistics: Death is a singular tragedy for
the deceased and everyone who loved them. Why take chances?

It's unlikely I'll get it. 5,000 people have it in the US. Even if
that number increases a hundredfold my chances of getting it are still
only one in six hundred. And it probably won't be worse than any other
flu.

I should take serious precautions. I should carry disinfectant and
wash my hands regularly. I should stay home if I'm not feeling well,
and should look for instructions from my local health authorities.

Let's take the last two statements first. Which is correct? The
answer, based on what we know now, is: Both. It's unlikely you'll get
it, but you should take precautions anyway. You should be taking some
precautions already to avoiding getting routine illnesses. But you
should be even more careful with swine flu.

Why? To use Homeland Security symbolism, our individual threat level
is "yellow" but our collective risk is "orange." Your chance of
getting it is low, and you don't need to be too concerned about your
own health right now unless you have other complicating conditions.
But you should be concerned about all of us. As more people get it,
the chance of pandemic increases. And a pandemic is a catastrophe. So
even though your individual risk is low, our collective risk level is
high enough that you should take precautions - for humanity's sake, if
not your own.

Concepts from risk management should help. Your risk of contracting it
is low, and the outcome if you got it probably wouldn't be severe. The
chance this could become a pandemic is also low - but the outcome
would be catastrophic. So we need to treat the possibility of pandemic
very seriously.

Tyler Cowen does a good job of summarizing the way this kind of risk
analysis should be conducted:

A one percent chance of one hundred million deaths is, in expected
value terms, one million deaths and that is a big deal. Probably the
United States is less vulnerable than it was in 1918, but how many
people would die in China, India and many other locales? How much
disruption to trade, travel, and the world economy would take place?
Even in the United States, our public health systems would break down
quickly and render many modern medical advances useless (e.g., when
would the Tamiflu run out?)"

Expected value calculations are a good way to look at problems of this
kind. And the fact that the United States is less vulnerable than it
was one hundred years ago does not mean it isn't vulnerable at all.
What's more, the deaths of millions abroad should concern us for both
selfish (e.g. economic) and altruistic reasons.

This kind of risk analysis should be standard operating procedure in
the business community. That's why it's still surprising when business-
friendly politicians strip money from the budget because they think
bad events aren't likely to happen (e.g. Bobby Jindal's mockery of
volcano monitoring, or Susan Collins' removal of flu pandemic money
from the stimulus budget). Remember: A low-probability event plus a
high-loss outcome = cause for concern.

To sum up, based on what we know today: Swine flu is not that grave a
threat to us as individuals, but it's a serious danger for all of us -
collectively. Think of it as an exercise in interconnectedness, a real-
world and scientific example of that "one world" talk we're always
getting from pop stars and movie actors.

Because we are interconnected: A farmer coughs while tending livestock
in Mexico. At lunch he shakes hands with his cousin the policeman, who
writes a ticket for a visiting tourist later that day. The tourist
goes home to Belgium, where his colleague leaves the next day to visit
his cousin the day-care worker in New York. A week later a 16-month-
old child dies in Queens.

It could have happened that way. We don't know. And washing your hands
today might save thousands of lives in Bangladesh or West Africa next
month. We don't know. But we do know that, in a a very literal sense,
that "one world" business is real. It's a good reason to take health
precautions. It's a good reason to support flu management and other
risk management funds in the Federal budget. And it's a good reason to
care what happens anywhere and everywhere in this interconnected
world.

RJ Eskow blogs when he can at:

A Night Light
The Sentinel Effect: Healthcare Blog

No comments: