SCIeNTIfIC REPORTS | 7: 16725 | DOI:10.1038/s41598-017-17093-8
A Synthetic Population for
Modelling the Dynamics of
Infectious Disease Transmission in
, Kathryn Glass
, Colleen L. Lau
, Nicholas Geard
, Patricia Graves
Agent-based modelling is a useful approach for capturing heterogeneity in disease transmission. In this
study, a synthetic population was developed for American Samoa using an iterative approach based
on population census, questionnaire survey and land use data. The population will be used as the basis
for a new agent-based model, intended specically to ll the knowledge gaps about lymphatic lariasis
transmission and elimination, but also to be readily adaptable to model other infectious diseases.
The synthetic population was characterized by the statistically realistic population and household
structure, and high-resolution geographic locations of households. The population was simulated over
40 years from 2010 to 2050. The simulated population was compared to estimates and projections of
the U.S. Census Bureau. The results showed the total population would continuously decrease due
to the observed large number of emigrants. Population ageing was observed, which was consistent
with the latest two population censuses and the Bureau’s projections. The sex ratios by age groups
were analysed and indicated an increase in the proportion of males in age groups 0–14 and 15–64. The
household size followed a Gaussian distribution with an average size of around 5.0 throughout the
simulation, slightly less than the initial average size 5.6.
Agent-based models are increasingly used to investigate the processes, mechanisms and behaviours of many
complex social systems due to their ability to capture the nonlinear dynamics of social interactions. For infectious
diseases, agent-based modelling has demonstrated considerable value for informing public health policies aimed
at preparation for or response to epidemics
, including the 2014–2016 Ebola outbreak in West Africa
studies on transport simulation and disease modelling have highlighted the value of synthetic populations
especially when heterogeneities in population mixing play a crucial role in disease transmission
, or when dis-
ease incidence or risks of infection vary signicantly between subgroups, such as age groups
. e nding of
highly variant prevalence of lymphatic lariasis across gender and age groups
further highlights the importance
of demographics in transmission dynamics. For American Samoa, the age structure is distorted by the large emi-
gration to the United States and immigration from Western Samoa, leading to the incapability of present models
on long-term transmission dynamics
Synthetic populations can be categorised according to the level of detail they capture about real populations.
One important distinction is whether or not they explicitly represent the geographic locations of individuals.
Spatial models of disease transmission are an important minority. High-resolution geographic locations of human
individuals are rarely included in models for diseases transmitted by person-to-person contact
. Spatial pop-
ulation data in these studies are primarily used to map epidemics rather than project the risk of infection at the
individual level. However, high-resolution geographic locations of human individuals are more critical for infec-
tious diseases that demonstrate signicant heterogeneity at small spatial scales or show household-level cluster-
ing. Examples include vector-borne diseases, where vector abundance can vary dramatically with environmental
Research School of Population Health, Australian National University, Canberra, Australia.
School of Computing
and Information Systems, University of Melbourne, Melbourne, Australia.
Melbourne School of Population and
Global Health, University of Melbourne, Melbourne, Australia.
College of Public Health, Medical and Veterinary
Sciences, Division of Tropical Health and Medicine, James Cook University, Cairns, Australia. Correspondence and
requests for materials should be addressed to Z.X. (email: email@example.com)
Received: 8 September 2017
Accepted: 20 November 2017
Published: xx xx xxxx