Smallpox Martyr Bio-terrorism
Modeling in Python
Joe Fetsch
Purpose
While smallpox has been eradicated, reserves
remain in case of an outbreak to use as a
vaccine. If Variola Major was released on a
population, many fatalities would result if
effective measures were not taken.
If this was done on a wider scale, resources
may be spread too thin to handle the situation
effectively
Explanation of Purpose
This project is intended to determine the likely
fatalities and damage caused by bio-terrorism
using a martyr attack with smallpox depending
on the time of which either a vaccine is
developed and distributed or a strict military
quarantine is implemented.
Scenario
In this scenario, two “suicide smallpox infected
terrorists” strategically target doctors’ offices
to obtain a medical certificate to explain their
absence from work. They blend in with people
suffering flu symptoms in the waiting room, but
they are already effectively spreading smallpox
amongst staff and patients. A terrorist
organization claims responsibility for the
outbreak.
Scenario, Cont'd
Citizens are too terrified to seek medical
attention, as they
are now aware that medical facilities have
been targeted. This is when the second stage
of the bio-terror
attack occurs. Smallpox is delivered to the
target city through either aerosolized delivery,
or through an
infected set of suicide terrorists passing on
smallpox through exhaled droplets.
Scenario Cont'd
Mass terror is created by the paradox that
smallpox needs to be contained and treated,
yet smallpox infection of patients and staff at
hospitals causes citizens to stay away from
medical facilities. A radio-talk-back host
ponders on-air whether by attending the doctor
to get checked for the flu,
or vaccinated for smallpox, you may acquire
smallpox before the vaccination takes effect.
Scenario Cont'd
Large segments of the community avoid
medical treatment. The strain placed on
the infrastructure of the city brings it to a halt:
planes do not arrive or leave, police at
roadblocks turn back fleeing residents, and the
“terror” caused by the bio-terror attack is
unmatched by any previously experienced
health catastrophe. The economy is bought to
a standstill and the bio-terrorists now have
political influence as they have demonstrated
their capacity to inflict terror.
Scenario Cont'd
Worse still, a rumor circulates that the
smallpox is a weaponised variant from the
former USSR, for which
there is no vaccine. Thus the containment of
infected people proves to be impossible even
though WHO
vaccines arrive quickly. There are not enough
respirator masks to go around.
Other Projects
Several government simulations involving
Smallpox taking other variables into account
have taken place, but no project has
accounted for a limited supply of vaccinations
and manpower
Several programs have accounted for
vaccination or quarantine, but few involve both
and none involve alternate supplies of these
resources
NetLogo Use

NetLogo was used to
develop a basic
understanding of the
disease modeling
system, but will not be
used to create the
smallpox model
NetLogo Virus Model
Smallpox

Child suffering from Smallpox
Much research was
done to fully understand
the Variola virus in all
forms and its effects on
a population
Project Structure

Each dot represents an
Agent with a vision,
intelligence, and a value
describing the stage in
which the disease
progresses

The infection, after
Prodrome phase, will
then progress into a
more mature phase:

Ordinary

Modified

Malignant

Hemorrhaging

Confluent
Agent Movement


The agents move in a
fashion that accounts
for the human instinct to
avoid those covered in
pustules and yet still
gather in groups
Away from infected and
towards others

This still maintains an
element of randomness
to account for ignorance
often expressed in a
human population even
in case of peril
World Structure
•
Instead of a physical •
representation,
showing the latitude
and longitude of every
agent in the world, a
social model will be
used.
•
In this model, agents
in close proximity are
likely to spend time
near each other,
drastically increasing
the risk of infection.
Agents with few
others near them are
simulations of
reclusive agents.
Visual Representation





Green agents are
healthy
Yellow agents are in
early stage where not
contagious or visible
Orange agents are in the
prodromal phase,
exhibiting flu symptoms
Red agents are infected,
contagious and visible
Blue agents are immune
Sugarscape-based model
Quarantine
•
Though this method is
not yet perfected, the
quarantine simulation
shows a world in
which a military
quarantine has isolated
everyone from each
other.
Quarantine
•
The visual
representation stops
moving, yet diseased
agents continue to
progress.
•
The line graph shows
a red line at the time at
which the quarantine
begins.
Description of the graph
In the graph above, the population of the city has gone from 5000, the initial
value, to 4052; a fatality rate of 20%. However, the population in this
situation has been quarantined after two months of the simulation, while the
rate of infection was still increasing, which would lead to many more cases
of smallpox and many more fatalities. Throughout the simulation, about
half of the agents became infected, which raises the relative fatality rate to
slightly less than 40%.
After the quarantine was implemented, the number of healthy people levels off
at 2495, as can be seen in the graph, while, at the same time, the number of
carriers no longer increases after that time, immediately after the number of
infected people becomes greater than the number of carriers in the
simulated world. From the graph, it is possible to notice the increases in the
number of carriers, infected agents, and immune agents as they progress,
defining the generation of infection. After the last generation becomes
infected, defined by the time of quarantine, all of the values drop off after
the generation progresses to the next stage of disease.
Timeline


Research Smallpox
to understand
disease in order to
better implement in
program
Using NetLogo,
obtained a basic
understanding of
the model of
infection

Using Python,
created basic
model with agents
and a 2-D vision
and 1-D movement
range to affect the
influenced
movement provided
by the infected
agents
Timeline
•
Developed a
model for
infection, hoping
to clarify my
values and prove
them accurate
with past data
•
Created a rough
draft of the model
for fatality
•
Hoping to prove
my data or create
a more accurate
representation
with the contact at
USAMRIID
Testing

Man suffering from hemorrhagic
smallpox also known as black pox
– 100% fatal
Simulation has
begun, average
fatality rate in a city
(likely to be
standard global
fatality rate) is
estimated around
20% with 60%
infected at one
point
Still needed:
•
Proof has been found to negate the
likelihood of vaccination, and the chaos
in the city would negate military
assistance for some time. No situation
such as this has taken place before, so
speculation is used to determine when
quarantine would become effective.
•
For now, my program can begin a
quarantine at any time, designated by
the user, implemented by a button.
Testing
•
Preliminary predictions are very incorrect:
–
Fatality rates are between 35 and 40%
–
All agents become infected within 6
months
–
90% of agents become infected within 4
and a half months
–
The graph on the next page shows the
results of many unhampered tests,
showing the population statuses over time