Transcript Human Mobility Network, Travel Restrictions and

Human Mobility Network, Travel
Restrictions and Global Spreading
of H1N1 pandemic
Contents
 What is Human Mobility?
 Human Mobility behaves as a channel for the disease
to spread
 GLEaM (Global Epidemic and Mobility Model)
Human Mobility
 Mobility is moving from one place to the other for
the sake of the convenience in life.
 There can be many reasons few of them are:




Weather
Business
Travel
Natural Disasters
 Every day millions of people travel across the world
covering millions of mile in a day.
 As per the research 60 million people travel via air
and covers billions of miles in one week.
 There are 3362 airports in 220 countries
 16842 connections between them, which creates a
large and very complex network.
 Model includes all the airports, 3362, in 220 countries. All the population is
divided into sub population. Each sub population is surrounding each air port.
In the figure on right shows the homogeneous mixing of infection in the same
area as well as in the different area. Areas are connected through air travels.
Background
 ¼ of the total population in Europe died because of the




black death.
14th October,1492 Columbus landed with some European
settlers and in that century more Americans died because of
disease brought by them.
European communities isolated from each other and in
1520, half of the Aztec population was succumbed to small
pox.
In 11th June,2009 World Health Organization (WHO)
raised a pandemic alert of level 6.
As of 19th July,2009 137,232 cases were registered across
142 countries.
Global Epidemic and Mobility Model (GLEaM)
 It is a structured meta-population model used for
evolution of the pandemic
 Performs a maximum likelihood analysis of the
parameters against the actual chronology of newly
infected countries.
 This method is complex as it involves Monte Carlo
generation.
 It is easy to estimate disease transmissibility as it has
the accurate and early stage data of newly affected
country.
 Human Mobility Patterns Models based on high
quality data helped to estimate disease
transmissibility.
 It also helps to know what the seasonal affects are on
the disease spreading.
Method
 Meta-population model is based on meta-population
approach.
 Whole world is divided into geographical regions
with a sub-population network.
 These subpopulations are interconnected with the
transportation and human mobility.
 Time scale separation approach for the short range
mobility between sub population defines the
effectiveness of disease.
 Model is made of three following layers:
 Population: High resolution population database provided by
the SociEconomic Data and Application center (SEDAC),
which estimates the population

Mobility Layer: Data given by IATA and Official Airline Guide,
which includes number of seats available and pairs of airlines
connected by direct flights.

Epidemic Layer: This layer defines the disease and population
dynamics.
 Stochastic procedure is used to simulate the mobility
of an individual.
 An individual can be in one of the following state
while sub population is affected by infection:






Susceptible
Latent
Infectious
Symptomatic
Asymptomatic
Permanent Recovered
In the latent (incubation) period there is no secondary
transmission occurred.
 Spreading rate of a disease is depends on the






reproduction number R0
R0 is the secondary cases produced by primary case.
β is the infectious rate for the symptomatic person.
Γββ is the infectious rate for the asymptomatic
person.
Average latency period is ε-1
Probability of entering into symptomatic
compartment is 1-pa
Probability of entering into symptomatic
compartment is pa
 Symptomatic infectious is further divided into two
cases:

An individual who can travel with probability pt.

An individual who can not travel due to illness 1-pt
 Infectious individual will recover with the rate of µ.
 Compartmental structure in each sub population. Each individual can be in :
susceptible, latent, infectious, symptomatic who cab travel, symptomatic
infectious who can not travel due to illness, asymptomatic infectious and
permanent recovered. Asymptomatic is less infectious than Symptomatic.
Latent is the incubation period when no secondary transmission occur. All
transitions are binomial or multinomial to preserve the discrete and
stochastic nature of processes.
 Global invasion of 2009H1N1 pandemic during the early stage of outbreak.
Arrows represent seeding of infection from Mexico to unaffected country.
Colors are to show different time seeding.
Travel Interventions
 Early stage of the outbreak many countries
implemented a restriction on the air traffic.
 40% of international air traffic reduction to/from
Mexico.
 After international alert 14 entities started awareness
programs.
 Stopping non essential travel to outbreak areas cause
more decrease in air traffic.
Results

Panels: A,B are the probability distribution of arrival time in UK and Germany respectively. Dotted
vertical line shows the observed arrival time and solid vertical line shows starting date of travel
restriction. Panels C,D are the cumulative travel distribution. In this any source of infection in
seeding event is considered. As per the Computational approach all the detectable and non-detectable
cases are taken into account.
 Even after giving a 90% restriction in the air travel
there was no big difference in arrival time of disease
in any country.
 Measure of arrival time with different values of travel
restrictions, which was not greater than 20 days.
 Even the greatest possible travel restriction has been
applied at the starting phase, there was no big
difference measured in arrival time of pandemic to
other countries.
QUESTIONS ???
References
 Chiara Poletto Recent Approaches in Modeling animal infectious




diseases
Christophe Fraser Pandemic Potential of a strain of Influenza A
Vittoria Coliza, Alain Barrat Modeling the worldwide Spread of
Pandemic Influenza: Baseline case and Containment Interventions
Paolo Bajardi, Chiarra Poletto Human Mobility Networks, Travel
Restrictions Global Spread of 2009 H1N1 pandemic
Roxana Lopez-Cruz Structured SI Epidemic Models with Application
to HIV Epidemic.