Geography Reference
In-Depth Information
separation could indicate epidemic spread via long-range interactions. A higher
degree of separation between two persons means that transmission of infection
between them would take longer. Finally, the distribution of degree centrality could
identify whether super-spreaders, who have higher degree centrality values, exist
(Huang et al.
2010
); however, the diffusion of epidemic diseases is not a single
event, and the capture of the temporal process of social contacts must rely on
intensive surveys and methodological improvements in social-network analysis.
Additionally, geo-computational models, such as agent-based modeling (ABM),
have also been used to detect early warnings of critical transitions and assess
interventions against epidemics in a system that emphasizes the importance of
spatial-temporal dynamic processes (Barrett et al.
2005
). Epidemic dynamics can
be modeled as interactions between different categories of populations, including
susceptible, exposed, infectious and recovered populations. Model parameters, such
as interpersonal contact rate, infection probability, transmission rate and recovery
rate, must be derived from mathematical models or estimated from empirical
evidence, and these parameters are sensitive to predicted results (Pastor-Satorras and
Vespignani
2002
). Recently, geo-computational models have been used to simulate
the complex interactions of the host, agent and environment in models assessing
the impact of climate warming on epidemic spread (Gong et al.
2011
;Morinand
Comrie
2010
).
Although spatial-temporal clusters and diffusion processes of epidemics have
been studied extensively in recent decades (Cowled et al.
2009
; Wen et al.
2012
; Wilesmith et al.
2003
), spatial methods of tracking the possible sources of
infection have not been comprehensively developed. Current methods of tracking
sources of infection and routes of transmission are time-consuming, difficult and
intensive, involving epidemiological investigations and laboratory diagnosis with
genotyping technology. The objective of this study is to propose an innovative
methodology that considers only the spatial-temporal relationships of illness onsets,
combining exploratory spatial analysis and network topological indicators, degree
centrality and network clustering coefficient, to identify space-time clusters, track
possible sources of an epidemic and measure transmission risk of an individual.
Analyzing temporal progression of these space-time clusters of an epidemic, two
geographic diffusion patterns, contagious and relocation processes, can be further
differentiated. It will be beneficial when establishing appropriate interventions for
specific epidemic period efficiently.
The structure of this paper is as follows: in the next section we describe the
methodology for tracking epidemic diffusion in time and space. Section
15.3
presents a hypothetical example to explain how the proposed analytical procedure
works. The methodology is further applied to the dengue fever epidemic in
Kaohsiung City, Taiwan in 2009-2010 as the case study. The background of the
case study and major findings are reported in Sect.
15.4
. Final section states our
conclusions.
Search WWH ::
Custom Search