Yesterday, the government finally did what many scientists, journalists and commentators had been pressuring them to do for weeks, and release some of the modelling that had underpinned their response to the COVID-19 pandemic.
The modelling, by Melbourne’s Doherty Institute, paints a stark picture of what Australia’s worst-case scenario could look like. But at the same time, experts and policymakers have repeatedly stressed that it does not offer a concrete prediction about what our fight against the pandemic might look like.
What does it tell us?
The modelling released yesterday didn’t tell us many of the things we’re desperate to know. It didn’t tell us when the lockdown would end, or whether we should close schools. Nor does it make us any the wiser about which of our beloved activities are essential or not — the modelling won’t help you avoid a fine. As both Scott Morrison and chief medical officer Brendan Murphy stressed during yesterday’s press conference, the modelling is theoretical, not predictive.
What the modelling does tell us is just how bad things could’ve gotten if we hadn’t taken strong action. Based on international case data, the team modelled four different scenarios, from one where the government had not intervened, to a scenario with quarantine, isolation and social distancing.
If we’d allowed for unmitigated spread of the virus, 90% of the country would have been infected, 35,000 intensive care beds would be needed per day. Our health system would be stretched to breaking point, and only 15% of the people who needed an ICU bed would get one, leaving doctors to make some truly painful decisions over who gets to survive, as they have had to do in Europe and the United States.
On the other hand, with quarantine, isolation and social distancing, we’d reduce transmission 33%, meaning everyone who needed an ICU bed would get one.
Why are there discrepancies?
What is harder to determine from the data is where exactly we are now. Some early expert reaction suggested that Australia is in a “relatively lucky position”, especially with the number of new cases falling each day for the past week, and with unconfirmed local transmission relatively in check.
But the nature of epidemics is what makes predictions nearly impossible.
“We can model airplanes flying because the laws of physics are predictable,” Monash University epidemiologist Allen Cheng wrote on Twitter.
“Human behaviour is not predictable. There is little data on how people are behaving at this unusual time and how that impacts on viral transmission.”
Complicating things further is the fact that other modelling has given us potential end dates, something the government’s modelling notably does not. One model suggested we could still be cooped up over Christmas. Another said our hibernation could be over by the end of winter.
Mikhail Prokopenko, a professor at the University of Sydney, developed different epidemic models, incorporating Australian data, which indicated we could reach our own pandemic peak in the next month.
“Unlike our study, the government’s model is coarse-grained, and does not include detailed predictions of what will actually occur in Australia,” Prokopenko said.
Ben Phillips, an associate professor in ecology and evolutionary biology at the University of Melbourne, says the government used international data because so much of what we know about the virus and its behaviour has come from observing what’s happening overseas. We simply don’t have enough local information.
And regardless of where the data comes from, experts say the government’s modelling is good and trustworthy.
Annette Dobson, a professor of biostatistics at the University of Queensland told Crikey the modelling released yesterday was far and away the best available, and warned that some other models that have been circulating online and in the media over recent weeks were “relatively incompetent.”
“This is the one that people should be taking notice of. Ignore all the others,” Dobson said.
Should the government have released the modelling?
It took the government weeks to reach the point of making the modelling available. Initially, Morrison was reluctant to do so. Deputy chief medical officer Paul Kelly said he would last week, before saying he wouldn’t.
One of the reservations was how the public might react — the government and their advisors seemed wary of armchair experts misinterpreting the figures and making populist pleas to change the course of our current response.
Phillips said that while that was a valid concern, he welcomed the government’s more open and transparent new approach.
“There’s certainly an issue that needs to be balanced, but in the absence of good information, there’s a greater risk that people will freak out and jump to the wrong conclusion,” Phillips said.
According to Dobson the released modelling, done by some of the country’s best experts, and carefully communicated to reduce the risk of misinterpretation, “should be such a boost in confidence.”
Dobson acknowledged that people might be disappointed with what the modelling didn’t provide — a clear end date, and a neat resolution to the present state of upheaval. But such uncertainty is likely to become the new normal.
Like a weather forecast, the modelling just tells us what might happen, not what will happen.
“In science, you really hardly ever get certainty,” Dobson says.