Menu
Log in


COVID-19: modelling and public health policy, presented by Allen Cheng and Freya Shearer

13 May 2020 11:52 AM | Marie-Louise Rankin (Administrator)

The COVID-19 pandemic has sparked a lot of interest from the statistics / data science community. On the 28th of April, we invited two Melbourne based statisticians/epidemiologists who actually advise the government on the coronavirus pandemic to share their work with us.  

The first half was presented by Prof. Allen Cheng, who is both an AStat as well as an infectious diseases physician. Allen first described several models that have been used to inform the policy makers around importation risk, transmission risk, and potential load on the health care system — this underpinned the discussion around ‘flattening the curve’. Allen also discussed some ‘unhelpful models’ that are being used elsewhere, as well as other on-going analyses/investigations around the next step to combat the pandemic. A key perspective that Allen brought to us was the fact that the modelling work is just ‘one piece of the puzzle’, and that the policy makers have to weigh up many other factors into the decision making process. Allen ended his presentation by commenting on the feasibility of sampling the population for the virus — a proposal that got some attention within the statistics community.

The second half was presented by Dr. Freya Shearer, an epidemiological modeller at PRISM. Freya and her colleagues were instrumental at developing the framework that quantified the importation risk from various countries. This framework helped uncover countries that have a high likelihood of under-reporting their cases, and by extension their risk posed to Australia. This ultimately helped the policy makers to decide on international travel bans during the early stage of the pandemic.

An equally well-engaged session was the Q&A at the end, where the presenters were asked to comment on issues such as the reliability of data sources, school closure, practices in other countries, and sensitivity of some key model parameters, to list a few.

Overall, the event was extremely well attended. We had 223 attendants (9 from overseas), which is a record for a non-named lecture.

The recording and other resources (e.g. slides) can be found here.

Kohleth Chia

Powered by Wild Apricot Membership Software