SIMID | Simulation Models of Infectious Disease

Epidemics6 Conference

Nine SIMID members joined the Sixth International Conference on Infectious Disease Dynamics to share three days of intense dialogue on our ideas, data, insight, models and methods. It was great to promote our work and to meet other researchers to discuss future research opportunities.

Members and topics:

Prof. Niel Hens

We analyzed social contact dispersal with respect to age, gender, week/weekend and regular/holiday periods. We concluded that the degree distribution over distance is not well captured by the power law and adapted contact matrices across distance can improve the analysis of infection dynamics.

Dr. Marina Antillon

We evaluated routine infant vaccination with and without catch-up campaigns among older individuals in Gavi-eligible countries using a dynamic model of typhoid transmission coupled with a treatment model. Routine vaccination with TCV would be cost-effective in some settings, and additional one-time catch-up campaigns would also be economically justified at $1/dose.

Dr. Stéphanie Blaizot

Mathematical models offer the possibility to investigate the infectious disease dynamics over time and may help in informing design of studies. A systematic review was performed in order to determine to what extent mathematical models have been incorporated into the process of planning studies and hence inform study design. Despite the fact that models have been advocated to be used at the planning stage of studies or surveillance systems, they are used scarcely. With only one exception, the publications described theoretical studies, hence, not being utilised in real studies.

Dr. Pietro Colleti

Influenza activity shows a complex spatio-temporal pattern whose complete understanding is still missing. Here we study 30 years of influenza circulation in France from influenza-like-illness (ILI) cases time series. Our aim is to characterize recurrent timing pattern across seasons and assess how they change from the start of the outbreak to the time at which influenza activity reaches its peak.

Dr. Wim Delva

Calibrating individual-based models (IBMs) to empirical data through maximum likelihood estimation is typically not feasible. Instead, simulation-based Approximate Bayesian Computation (ABC) methods are routinely used to calibrate IBMs to an array of target summary statistics (a.k.a. “features”). We developed a new calibration method based on the idea that Multivariate Imputation by Chained Equations (MICE) can be used as a multivariate emulation technique that connects model inputs and outputs, and show that MICE-assisted calibration has great potential for calibrating IBMs with many unknown parameters to a large array of empirical target features.

  • Recasting agent-based model calibration as a missing data problem

Dr. Diana Hendrickx

SimpactCyan (, is a freely available simulator for IBMs, written in C++ with R and Python interfaces. Simpact is implemented in continuous time: models are updated each time an event happens, and not at fixed time intervals. This poster presents the modelling approach of Simpact and examples on how to use Simpact to address questions in epidemiology.

Dr. Lander Willem

Individual-based models are suited to combine heterogeneous within-and between-host interactions. [1] We reviewed a decade of IBM publications to present their opportunities and pitfalls to facilitate knowledge transfer within and across disciplines. [2] We modelled the stochastic nature of measles transmission and showed that individual-based social contact patterns and (clustered) immunological states provide essential information on the extent of effective herd-immunity.

Thang V. Hoang

Over the last decade, social contact data have been increasingly used to capture contact behavior of individuals, providing crucial input for more reliable modeling of infectious diseases. In this work, we report the results of a systematic review of social contact surveys relevant for close contact infections. We point out similarities and differences in collecting social contact data and issues that need to be addressed in future studies. Also, we enumerate the main findings in terms of social contact determinants.

Frederik Verelst

The impact of infectious disease transmission and policy interventions are subject to hosts’ behavior. Hence, there is an interest to incorporate human behavior in models for infectious disease transmission. [1] We systematically reviewed behavioral change models. We remain concerned that most models are purely theoretical and lack representative data. [2] We performed a discrete choice experiment in Flanders to quantify vaccination behavior.

Related Posts