COVID-19 outbreaks remain a clear and present danger despite the staged easing of restrictions, University of Newcastle virologist Nathan Bartlett says.
"It is a dangerous time. We're properly getting into winter now. Kids are back at school, they're seeing grandparents and there's more mobility," Dr Bartlett said.
"It's about hypervigilance now."
Dr Bartlett believes the government will move quickly to reinforce restrictions if outbreaks occur.
The associate professor, a Hunter Medical Research Institute scientist, said it was "now about watching things carefully, keep screening and quickly identify any potential community outbreaks".
University of Amsterdam researchers recently released results of a 35-year study into immunity from four strains of coronavirus that cause the common cold.
The study of 10 men found frequent reinfections 12 months after infection and substantial reduction in antibody levels as soon as six months post-infection.
Dr Bartlett said the duration of immunity for those who contracted COVID-19 was "a very important question moving forward".
"We need to be very cautious about speculation and conjecture when there's no empirical, scientifically robust data," he said.
"We need to see the first waves go through and resolve. We need immunity to be established and then we need to follow up months to years later to see if people become reinfected."
He said a longitudinal study, which would take years, was needed to compile the required data.
Research from the La Jolla Institute for Immunology in California suggested that people who had the common cold from other coronavirus strains may gain some protective immunity against SARS-CoV-2 - the coronavirus that causes COVID-19.
The research, published in the journal Cell, examined blood from recovered COVID-19 patients.
Dr Bartlett, who has studied various types of coronaviruses, said caution was needed in this area.
"The way a pandemic works and the way this virus is spreading would indicate the level of protective immunity in the community is essentially zero," he said.
"That would suggest that prior infection with other human coronaviruses, which circulate continuously in communities, does not confirm any appreciable level of protection [against COVID-19]. That's not to say that in fairly rare circumstances, there may be some cross-reactive antibodies."
But he said the spike proteins that SARS-CoV-2 use to attach to human receptors was "structurally quite different" to those of most other coronaviruses.