Maths has improved Australia’s response to infectious disease

Data analysis and mathematical modelling by the National Situational Assessment Consortium has shaped Australia’s response to COVID-19 since the early months of the pandemic. For Defence Science and Technology Group collaborators, the work is improving their ability to detect biological attacks.

The outcome

The National Situational Assessment Consortium is a group of experts in infectious disease dynamics, bringing skills in advanced analytics and mathematics together with extensive expertise in epidemiology and public health. The consortium advised all Australian jurisdictions on the response to COVID-19 from March 2020 until the end of 2023.

The group has produced over 200 weekly surveillance reports delivered to key national decision-making committees, including the Communicable Disease Network of Australia, the Australian Health Protection Principal Committee and the Australian Technical Advisory Group on Immunisation.

“I think the key achievement is that we established a workflow and a capability that delivered strategic analysis and advice straight to the highest level of government,” says Professor James McCaw, Professor in Mathematical Biology and senior staff member of the Infectious Disease Dynamics Unit at the University of Melbourne.

“We also co-authored Australia’s National Disease Surveillance Plan for COVID-19.”

Professor McCaw co-leads the consortium with Dr Freya Shearer. Contributors include researchers from Australian universities and research institutions, including the Defence Science and Technology Group (DSTG).

“We worked with public health and epidemiological leadership to establish a surveillance system that perfectly married up data analytics and modelling capability so that government received in a timely, routine way really clear intelligence material on what they were having to respond to,” says Professor McCaw.

For the DSTG, the collaboration has improved the team’s understanding of how to forecast infectious respiratory diseases.

“To actually have a very fluid collaboration happening between defence and a consortium of universities for advising government on a really critical issue for me has been an interesting case study to see how that works in real life in what was a high-pressure scenario,” says the DSTG’s Dr Peter Dawson.

Because of the similarities between the initial symptoms of potential biological weapon agents and respiratory diseases that regularly circulate in the community, this has fed into ongoing research to detect potential biological attacks earlier.

The consortium’s work has increased the sophistication of Australia’s monitoring and response to viral respiratory illnesses and other pathogens to support better decision making.

Learn more about Defence research

The need

Australia recorded its first case of COVID-19 on 25 January 2020. By early March, the first cases of Australian community transmission were confirmed. The World Health Organization declared the virus a pandemic on 11 March.

The Australian Commonwealth and state governments needed to make decisions on how to manage the pandemic – and fast.

Mathematical models describe how diseases spread through the population. They can be used to study how interventions may slow the spread or otherwise reduce the burden of disease on the population.

But data alone isn’t enough.

A specialist group of analysts, epidemiologists and mathematicians is needed for guiding the government response to COVID-19 and other infectious diseases.

Understanding how respiratory diseases behave in a population is also needed for detecting biological attacks.

“If you look at classical potential biological weapon pathogen, their symptoms are respiratory,” says the DSTG’s Dr Peter Dawson.

“If you want to try to detect the early signs of an attack in a particular population – say, a military population – but people are developing influenza and influenza-like illness in that same population, and the early symptoms of the attack look an awful lot like flu, you need to able to disentangle those two things.”

The research

The consortium analysed trends in COVID-19 epidemiology – patterns in the disease’s spread – over time and space, anticipating future possible trends.

“Weather forecasting is analogous in one sense. It's always being updated. Sometimes forecasts are really accurate, sometimes the models miss something. But it lets you make very local short-term decisions. And it also helps you understand the real big threats [like hurricanes or bushfires] … that you might have to manage,” says Professor McCaw.

The group’s reports to government considered:

• Real-time estimates of the rate of virus spread (effective reproduction number)
• Trends in population behaviour and the impact of that behaviour
• Immunity from infection and/or vaccination
• Health system capacity
• Case ascertainment (determining what fraction of those who have the virus are being reported as cases in health databases)
• Characteristics of new virus variants, assessing their potential to spread.

These analyses allowed the consortium to produce state-level forecasts of daily incident cases and hospital occupancy. Reports included an ‘ensemble forecast’, built from the forecasts produced by independent models, developed by independent teams.

“We built quite a complicated complimentary workflow. The team met twice a week for almost four years, producing these reports for government,” says Professor McCaw.

Developing the solution

Professor McCaw has worked with the Defence Science and Technology Group for the past 10 years. The collaboration began with the DSTG’s interest in understanding how to detect a disease outbreak or biological attack.

“We soon realised that to generalise [our tools] beyond a simulated attack and into working on real health datasets, we needed to be able to forecast disease. And we also needed to have better partnerships with health departments in order to have test data to actually run those tools,” says Dr Dawson.

Professor McCaw’s team had connections with the federal health department, which the collaboration was able to leverage.

“Certainly being able to collaborate with the University of Melbourne and health departments between 2015-2019 on forecasting influenza was an excellent opportunity for us to improve our technology,” Dr Dawson says.

“This [national situational assessment] work would not have existed if the Defence Science and Technology Group hadn't funded the flu forecasting work from 2015 onwards,” says Professor McCaw.

Partners

• Defence Science and Technology Group
• Telethon Kids Institute
• Curtin University
• Monash University
• University of New South Wales

Funding support

• Funded by the Commonwealth Department of Health Office of Health Protection

Publications

1. McCaw JM, Plank MJ, The role of the mathematical sciences in supporting the COVID-19 response in Australia and New Zealand, ANZIAM J 64: 315–337 (2023)
2. Moss R, Price DJ, Golding N, Dawson P, McVernon J, Hyndman RJ, Shearer FM, McCaw JM, Forecasting COVID-19 activity in Australia to support pandemic response: May to October 2020, Scientific Reports 13: 8763 (2023)
3. Golding N, Price DJ, Ryan G, McVernon J, McCaw JM, Shearer FM, A modelling approach to estimate the transmissibility of SARS-CoV-2 during periods of high, low and zero case incidence, eLife 12: e78089 (2023)
4. Tobin RJ, Wood JG, Jayasundara D, Sara G, Walker J, Martin GE, McCaw JM, Shearer FM, Price DJ, Real-time analysis of hospital length of stay in a mixed Omicron and Delta epidemic in New South Wales, Australia, BMC Infectious Diseases 23: 28 (2023)
5. Price DJ, Shearer FM, Meehan MT, McBryde E, Moss R, Golding N, Conway EJ, Dawson P, Cromer D, Wood JG, Abbott S, McVernon J, McCaw JM, Early analysis of the Australian COVID-19 epidemic, eLife 9:e58785 (2020)

View COVID-19 situational assessment technical reports

People

• Professor James McCaw
• Dr Freya Shearer
• Dr Rob Moss
• Dr Peter Dawson
• Professor Rob Hyndman
• Professor Nick Golding