New Zealand faces an "unacceptably high" risk of as many as nearly three border breaches each month - and by some measures the country is more threatened than Australia.
That's according to a new analysis by public health researchers on both sides of the Tasman, who warn that risk of new Covid-19 incursions is growing even with vaccinations and call for urgent improvements.
The researchers looked at nine border failures in New Zealand, including the August Auckland cluster that resulted in around 180 cases and three deaths, and seven in Australia - six of which led to lockdowns.
Using the estimated number of people who went through both countries' hotel-based quarantine facilities up to the end of January - along with the equivalent number of positive cases - they calculated a combined failure rate of one per 20,702 travellers.
That also worked out to one failure per 252 positive cases in quarantine and one outbreak leading to a lockdown per 47,319 travellers.
When broken down by country, New Zealand had 15.5 failures per 1000 positive cases going through quarantine, compared with two per 1000 in Australia - a seven-fold difference.
The researchers noted that, given the proportion of cases per 1000 had risen to 9.1 in New Zealand and 16.3 in Australia, that equated to 1.8 and 0.7 expected failures per month, respectively.
Moreover, if travellers came with a variant that could be around 50 per cent more infectious, those respective rates climbed to 2.8 and 1.0 - assuming a linear increase in risk.
The researchers said New Zealand's comparably riskier rate could be down to either a "lower-quality" approach - or possibly better detection, from testing more border workers over a longer period.
Still, they added that, since December, the proportion of positive cases among international arrivals was still greater in Australia than in New Zealand.
Their calculations came with some caveats, given that control approaches varied by state and country - and that estimates could be subject to "chance variations" because of the low number recorded of failures, with the potential they'd been under-estimated.
Much clearer was the likelihood that the potential number of monthly failures in both countries would increase, as a surging pandemic and new variants lifted the proportion of arriving travellers infected with the virus.
That risk could be offset with the vaccination of quarantine workers, which has just begun in New Zealand.
But the researchers still saw a need for major changes, namely capping travel from countries with high transmission, and requiring more testing, or a stint in quarantine, before travelling.
Study co-author and University of Otago epidemiologist Professor Nick Wilson said harder measures could also include daily saliva testing of all MIQ workers - and potentially also travellers - along with moving facilities to military bases outside of cities.
"I'd argue the current risk is unacceptably high, given that we know it can lead to something like the Auckland August outbreak, which cost hundreds of millions of dollars," he said.
"Even Auckland's recent lockdown would have cost tens of millions of dollars, and stressed a lot of people out. Given all of those things, more investment in improving the facilities just seems to make pure economic sense - and pure public health sense."
On the study's comparisons with Australia, he said New Zealand's testing regime might have been stronger, and thus picked more cases.
"But then there could be some genuine advantages of the Australian system. For example, they've had in place, for some time now, a requirement that no one leaves their room for the whole 14-day quarantine period," he said.
"That is obviously more rigorous than here, where we allow people out to exercise or go to smoking areas. Australia's approach might be harder on people's mental health and wellbeing, but it also might be reducing the risk."
This month, Wilson and his colleagues set out 12 such steps the Government could readily take.
Professor Michael Plank, who wasn't involved in the paper, but whose Te Punaha Matatini colleagues have been modelling the risk at the border, agreed Australia might have some lessons for its neighbour.
He pointed out that at least three of New Zealand's leakages had happened as a result of a returnee passing the virus on to another person within quarantine.
"This has happened in two different facilities - the Crowne Plaza in Christchurch and the Pullman in Auckland," he said.
"Australia has had no documented instances of this – all of Australia's leakages have occurred as a result of a quarantine worker getting infected.
"This suggests that transmission between guests in hotel quarantine might be a weakness of the New Zealand system and this could be an area for improvement."
Plank highlighted the study's finding that New Zealand's nine failures - or 10 if the latest episode was counted - was higher than Australia's seven, despite the fact 3457 cases had been through Australian quarantine, compared with 579 here.
"The lower failure rate in Australia suggests we might be able to learn from some of their systems and processes to improve our own quarantine system," he said.
"However, an important difference between the countries is that of the nine quarantine failures in New Zealand, only one - the August outbreak - has resulted in sustained community transmission, whereas in Australia this has happened on four occasions and has led to six lockdowns.
"This suggests that New Zealand is doing a relatively good job of detecting border failures before they have a chance to spread out into the community."
Plank added that, while some of the recommended measures could help stop failures from occurring, others, like saliva testing, would simply pick them up early, before they could cause outbreaks.
"Since 100 per cent perfection is unachievable, both of these are important ways to keep Covid out of the community."
Covid-19 Response Minister Chris Hipkins has been approached for comment about the new study, which has been published online ahead of peer-review.
Prime Minister Jacinda Ardern has meanwhile poured cold water over one of its key recommendations - the need for purpose-built quarantine facilities.
In the year it would take to get New Zealand's "team of five million" immunised evidence could emerge that vaccination reduced transmissibility, Ardern said.
And that could "change up" what was required at the border.
In the meantime, constructing a purpose-built facility which could hold up to 6000 people was a "significant ask", especially as some incursions were traced to returnees touching the same lift button, said Ardern.
"So not always has the facility been the problem, but the virus within it."