"We are not doing enough testing to know for certain if there is spread in our community. And so, we will only identify such spread through deaths in older people," Professor Lokuge told the National Press Club on Wednesday.

"If we wait until then to take strong action, then we risk overwhelming the health system, and as we have seen happen in Italy and is now happening in other parts of Europe and the United States.

"We risk being in a situation where many thousands, perhaps hundreds of thousands of people, may die unnecessarily."

If Australia acted early, it could buy time to establish strong health and epidemiology systems and save lives.

Professor Lokuge said Australia had done less than half the testing of Singapore, which has a population just one-fifth of Australia's. Singapore had tested about 200,000 people with a population of 5.5 million; Australia had tested 80,000 people with a population of 25 million.

Professor Lokuge, from the National Centre for Epidemiology and Population Health at the Australian National University, is a GP and epidemiologist who dealt with Ebola outbreaks in west Africa.

She said people should remove children from schools if they could, leaving schools for the children of people, including medical staff, who were unable to remove their children.

"We all have the power to stop the spread of this virus. By limiting our contact with others, by staying home, by avoiding crowds and gatherings such as sporting events and church. By working from home, by taking our kids out of school if we can do that safely.

"By not going to the gym, by not attending that party we agreed to go to a month ago. Social distancing is not something we have to practice forever, but it is something we must practise now, right now, to slow the spread of this virus."

Countries that had acted late and taken measures that were not strong enough, such as Italy, other parts of Europe and the United States, had seen a very rapid increase in cases. Australia now faced that scenario, without knowing how far the virus was spreading.

With shortages of testing kits, Australia is limiting testing to people who have been overseas in the past two weeks or who have been in contact with a case, as well as doctors with symptoms.

"If our health system is overwhelmed like it was in Italy, like it was in Wuhan, the chance of someone dying who has COVID-19 infections is four to five times higher," Professor Lokuge said.

But the director of epidemiology at Melbourne's Doherty Institute, Professor Jodie McVernon, who was also at the Press Club, said it was simplistic to compare Australia's testing with Singapore, considering the density and intensity of the population in Singapore and its place as a global, transport hub.

"No one is proposing a strategy where we let this rip," she said.

"We understand the gravity of this infection."

The government has refused to release its modelling for the number of people it is anticipating being infected. NSW Chief Health Officer Kerry Chant said authorities were working on 20 per cent infected, with five per cent of those needing intensive care and 1 per cent dying. In Australia, that translates to 5.1 million people infected, 255,000 needing intensive care and 51,000 deaths.

But asked about the latest modelling, Professor McVernon would not divulge figures.

"I will not give you a number to tell you where this is going or how bad it's going to be, because we have also just implemented a suite of interventions and we will continue to monitor how that is going," she said.

Australia was working "in the face of great uncertainty and developing preparedness scenarios based on our worst case", she said.

Asked when Australia will move to still more drastic measures, such as the shutdowns seen in Europe, Professor McVernon said there was "no magic trigger".

"What we are talking about is implementing a suite of extensive behavioural interventions well before we're at that sudden rise in cases. In many countries in Europe, there was this remarkably high death rate which basically indicated that cases were just not found. By the time this was detected, the outbreak was already out of control."

Australia was following the model of Singapore and Hong Kong and of "going in early". Most of the cases in Australia were still imported.

"The rate at which the epidemic will grow from here depends on how much is already there. And we are stepping in at an early stage," she said.

Despite increasing numbers of local cases and cases where the infection source was not known, Associate Professor Vanessa Johnston of ACT Health said there were only small pockets of localised community transmission in Australia.

Most cases were overseas sourced. Of the 414 confirmed cases by Tuesday, the source was known for 300, with some unknown and some still being investigated. Of those 300, 212 were overseas-acquired, she said.

"What this tells us is that for the majority of cases in Australia, we can identify a source of exposure for them. They are imported or they are close contacts of cases in this country. There is only a small number at this stage where we have not been able to identify a source, and that is what is concerning us now," she said.

Asked how long it took to recover once someone had the virus, Associate Professor Johnston said recovery time increased with severity.

Some people felt fully well but had not cleared the virus, so the protocol was that patients must test negative on a nasal swab twice, 24 hours apart.

"The time in which you recover does differ depending on the severity of the disease. We do know some people will express that virus for a lot longer than others," she said.

Professor Lokuge said she had worked for 25 years on controlling high-risk pandemics and epidemics all over the world, and her experience showed that they could be controlled.

Staying away from each other was the most powerful weapon Australia had to control coronavirus and protect grandparents, nurses and doctors, she said,

"There is something about this coronavirus that I'm profoundly grateful for - and that's that it won't mean the death of our children. In that respect, it's different to every other outbreak I've worked on," she said. "We can be confident, though, with the data we have, that coronavirus won't take our children from us."

We have removed our paywall from our stories about the coronavirus. This is a rapidly changing situation and we want to make sure our readers are as informed as possible. If you would like to support our journalists you can subscribe here.