Preparing for and managing pandemics using mathematical modelling

The outcome

Mathematical models of infectious disease outbreaks are helping the Australian government respond to pandemic threats.

Developed by a University of Melbourne research team, the models can be used to chart the different potential impacts of an infectious disease outbreak. For example, they can anticipate likely numbers of infections, hospitalisations and deaths.

Using modelling to assess early information about an epidemic can also predict its future course weeks in advance, providing just-in-time support for decision-makers. This information helps governments to plan the most effective response.

In Australia, the team’s models provided early evidence that COVID-19 may have a major impact. This informed the Federal Government’s response. These contributions added to 15 years of providing advice on preparing and responding to pandemic influenza.

The team’s models have been used to predict the peak of influenza season up to five weeks in advance. State governments have used this information to help plan their health resources.

Internationally, the team works with the World Health Organization on pandemic management advice for countries.

The need

Circumstances can change quickly during infectious disease outbreaks and uncertainty can be high. Governments need to be able to adapt their responses in light of rapidly emerging evidence.

Being able to anticipate the spread of an infectious disease and its impacts allows governments to identify proportionate strategies to reduce harm. These may include strengthening the health system’s capacity to respond, or applying social distancing (stay at home) measures to limit spread.

Developing the solution

A research team led by Professor Jodie McVernon, Professor and Director of Doherty Epidemiology at the Peter Doherty Institute for Infection and Immunity, and Professor James McCaw builds and refines the models. They do this by combining mathematical modelling with clinical and epidemiological studies (studies of diseases). This ongoing process ensures readiness to respond quickly when the need emerges.

Their models are based on a common type of model known as the ‘SEIR’ model. Each individual in a population is assumed to be fully susceptible (S) to infection at the outset of the outbreak, and then exposed (E) to the virus when they come into contact with an infected (I) person. After a short period, an exposed individual becomes infected and infectious (I). Once recovered (R), individuals are assumed to be fully resistant to reinfection, at least for a period of time.

The researchers refine the model by including data from actual outbreaks as they become available.