Agent-based tracking of disease spread

Historically, the spread of infectious diseases is simulated with compartmental models based on differential equations that can accurately reproduce epidemic trajectories. However, a central assumption in equation models is that of a well-mixed and homogeneous population. The latter is a serious limitation, especially if the outbreak is clustered. Agent-based models can overcome these limitations by operationalising the heterogeneity in individual attributes, for example demographic attributes and pre-existing disease, and behaviours in the simulation of epidemics.

The research aim is to develop new features in MATSim (Multi-Agent Transport Simulation), a well-established framework for developing agent-based simulations. Each agent is an individual belonging to a synthetic population, which represents Switzerland’s population. The new features will enable MATSim to construct a topological graph of agents’ co-presence in vehicles and buildings over multiple days. Each agent's status is updated accordingly with a probabilistic model that considers the status of its contacts. This allows us to trace and study how the virus spreads. The results of the simulation form the basis for the explicit simulation of disease spread: EpiSim

We expect to develop a spatial, temporal, and social perspective model that enables us to accurately consider the heterogeneity of the population and their behaviours in response to the epidemic. The model can be used to explore complex and spatially-heterogeneous policies, predicting their effectiveness. The results are also relevant for preparedness against other pathogens with pandemic potential. An open-access dashboard will make our work available to all stakeholders providing forecasts of the on-going epidemic, but also the forecasts of different user-chosen containment and intervention strategies based on the simulations.

One category of forecasts that may substantially benefit from agent-based models is the projection of bed occupancy in intensive care units (ICU). The accumulation of long-stay patients can lead to patient overflow, making it difficult to anticipate how many patients can be admitted to intensive care units on a given day. In Switzerland, as of May 2020, the mean length of stay of COVID-19 patients in the intensive care units has been substantially longer compared to that across all patients in the intensive care units in 2019. Compartmental and agent-based models can help address these questions and ensure that hospitals operate within their bed capacity.

Agent-based tracking of disease spread with dynamic models of travel behaviour in a pandemic.

• MOBIS : COVID-19