A top epidemiologist says New Zealand may have spared itself a worst-case coronavirus disaster "in a nick of time" through its dramatic shut-down.
But it will still take many weeks before we can truly see what difference the month-long measure makes against Covid-19, Otago University's Professor Michael Baker said.
Studies indicate Covid-19 has an average five-day incubation period. That meant any infections transmitted today might not turn up for at least a week.
In a month's time, however, Baker hoped to see the curve of exponential growth of cases in New Zealand – still largely made up of returned travellers – begin to flatten out.
"In general, you'd be expecting that four weeks won't be enough. But we might just be pleasantly surprised. That is the big unknown."
Because this lockdown and the pandemic itself was unprecedented for New Zealand, Baker said it was difficult to say for sure what we might expect.
"There really is nothing to draw on. We are all kind of taking part in a natural experiment where everyone stops moving around for bit," he said.
"In the past, we've seen how school closures can have a modest effect, but that is a tiny proportion of the mixing that occurs around the country."
What could be expected, he said, was a two-week lag in case rates - meaning the country's tally would be rising for some time to come.
"As the lockdown policy effectively doesn't come into place until tomorrow night, there will still be a lot of social mixing going on around the country until then," he said.
"There will still be those infected people who are infecting others in their immediate vicinity. But the opportunity for the virus continuing to progress will diminish."
The key pattern he and other epidemiologists would be keenly observing over coming weeks was the rate of infections coming from within New Zealand.
Of 155 cases of confirmed and probable cases in the country to date, just four were thought to have come from spread within the community.
"That is really the curve to watch. And if these begin to be extinguished, then that's obviously a very positive sign."
Professor Mick Roberts, a world-renowned infectious disease modeller at Massey University, agreed community transmission rates in several weeks' time would give a crucial sign of how successful the intervention had been.
Four weeks accounted for about four separate infection generations.That was the time between the infection time of an infected person and the infection time of the person they passed the virus on to.
"After four of these, you'd hope that all of the lag had been taken out of the system and things were dropping."
Provided rates would begin to drop, New Zealand could also have the added benefit of ramped up capacity for tracing, identifying and quarantining cases, Baker said.
"There is a chance for us to move back into a more normal existence, where we are gradually scaling back physical distancing measures, and, combined with our increased capability, we might find that things become pretty manageable."
Dr Ayesha Verrall, an infectious diseases doctor and epidemiologist at Otago University, said scaling up testing and contact tracing capability was critical.
"We should aim to leave the lockdown in one month with the ability to identify and trace the contacts of 1000 cases a day. We are currently struggling with 50," she said.
The country would need accessible testing, fast test turn-around times, rapid contact tracing augmented with smartphone apps and welfare support for those who struggle in isolation, she said.
"All these processes need to be fast, scaled up and integrated. China and South Korea have succeeded in this strategy of turning around large outbreaks, because they have strong public health infrastructure, following the lessons they learnt from the Sars outbreak.
"If we had better ability to find cases and isolate their contacts we would be able to manage larger number of cases without going into lockdown.
"That is how Singapore managed more than 500 cases without closing their schools."
"Building this capacity means we could look to the next 18 months with more confidence that we won't have large outbreaks or be in perpetual lockdown."
Baker credited China with showing the world why trying to tackle Covid-19 like influenza – or "flattening" a peak of cases through basic mitigation measures – wouldn't work.
While both were spread in the same way, the flu had a basic reproduction number - or the average number of people who catch the virus from an infected person - of 1.3, while the number for Covid-19 sat between two and three.
"But the huge difference here is that the incubation period for Covid-19 is about two or three times longer than the flu," Baker said.
"So you need to go to higher levels and take an approach of eradication.
"When you have a vaccine, as we do with flu, the aim is it remove susceptible hosts. With an eradication strategy, one of your only options is stopping people mixing."
Baker was aghast at reports the United States was looking at scaling back its measures to ease economic pain – something that would likely result in overflowing hospitals and a much higher number of deaths.
"You can't describe how awful that thinking is. 'Train-wreck' is one of many words you could use."
For Australia, and other countries in Europe, there was still a chance to stop the virus by effectively turning population centres into cut-off islands – an old principle called cordon sanitaire.
"It looks like, in a nick of time, we might have been able to do just this. The Government's new policy is absolutely fantastic. I just had my first good night's sleep in weeks."
How big data could help fight Covid-19
Meanwhile, a team of researchers are working on new models to give a much more detailed picture of how Covid-19 is being spread in New Zealand.
Professor Shaun Hendy, director of Te Pūnaha Matatini, a research centre specialising in complex systems and data analytics, said scientists had been using SIR models to simulate spread.
That combined data from people who were susceptible, currently infectious and recovered.
"These are based off on an aggregate population – or assuming that everyone has a chance of meeting everybody else, which is not really realistic," he said.
"So we have also got what are called individual based models, where we look at how individual people might meet and interact with other people."
"We can then use these to work out how what changes when people's behaviour changes."
Te Pūnaha Matatini researchers have already used such models to explore how flu spreads, along with how our legal system compares with other countries and how ideas move from city to city.
Hendy said one option was using anonymised cellphone data. Overseas, authorities were now using such data to observe whether people were keeping distance from each other and staying home.
"In our case, it's important to emphasise that this isn't 'big brother' stuff. We would be looking at using this data as an aggregate to tell how we are keeping apart."
Hendy said he was impressed with the Government's bold move.
"We will likely see the number of detected cases continue to rise over the next few weeks, but once they begin to fall the Government will be able to consider dropping the alert levels," he said.
"We may also see alert levels drop in some parts of the country if the government becomes confident that we are not seeing community transmission there and can restrict travel sufficiently from regions at higher levels."