Multi-Coaffiliation Networks and Public Health Applications

Abstract

Infectious diseases are a global concern. The
challenge for public health bodies relies upon optimizing the
distribution of scarce or costly control measures to maximize
their impact on the outbreak dynamics. Risk identification has
focused on schools and child-care centers mainly because they
represent dense masses of highly immunologically naive hosts
for the pathogens. To advance the design of mitigation strategies,
epidemiology researchers have broaden their perspective
through the use of computational tools designed to provide
decision support for multiple scenarios.
To identify at-risk populations, we propose a computational
algorithm that recreates a realistic social model of the school
system of a selected study place. It is a known fact that
childhood diseases are spread through the social contacts that
occur in the classrooms while schools are in session. Through
synthetic reconstruction, the algorithm generates a synthesized
population database. The demographic simulations are created
at the level of individuals, households and schools. Then a
school to school network is built as a representation of the social
model. The algorithm outputs a new graph B′, representing the
Multi- Coaffiliation Network (MCN) with number of vertices
of order O(S), where S represents the number of schools.
The resulting weighted network includes a value associated
with each school as a possible intervention location. The riskevaluation
of the schools in the network can be derived in a
wide range of applications in both research and public policy
analysis.