Coronavirus was spreading in New Zealand earlier than first thought, shows a study revealing the country's first-known Covid-19 cluster.
New Zealand's first announced case - a man who'd returned to Auckland from Iran via Bali - was reported on February 28, 2020.
But there's now evidence, discovered only through a weak positive test some seven months later, to suggest another returned traveller from Italy had already started a cluster of potentially six people.
These cases, hidden at the time, reinforced the urgency the Government faced before it eventually ordered lockdown a year ago this week.
In the paper, published in the New Zealand Medical Journal today, the researchers described an intriguing weak positive test reported to the Waikato Public Health Unit early last September.
It had come from a man in his 60s who sought out a test while suffering a sore throat, a cough and other symptoms.
Yet health officials weren't able to pin-point any potential source for acute infection of Covid-19.
The man hadn't had any contact with recent confirmed cases, or been near any of the places involved in the Auckland August cluster that was active at the time.
The patient did, however, say he thought he might have been infected months earlier.
Back on February 23, a family member, along with another close contact of the man, arrived in New Zealand from Lombardy, Italy - one of the worst-hit areas early in the pandemic.
That family member, described as the cluster's "index case", became unwell shortly after arriving home, and then sought primary care on February 25 for a "flu-like illness".
Because the person didn't meet the then-suspected case definition for Covid-19, they weren't eligible for testing, and were instead advised to self-isolate.
Over the following four days, six other household contacts became unwell with a similar illness, and all isolated themselves until the symptoms subsided.
When the "index case" returned to Italy in mid-June, the standard-given PCR test returned a negative result, but a serology test revealed the person had produced antibodies for a past Sars-CoV-2 infection.
Following the man's weak positive result in September, it was similar antibody testing that indicated he and five other household contacts had contracted the virus.
Study co-author and Waikato DHB medical officer of health Dr Richard Vipond said the findings suggested the virus was spreading silently in New Zealand from at least late February.
"It would be speculation without having any confirmatory laboratory testing, however, it seems likely that other cases went undetected in early 2020."
At the time, lab testing capacity and the case definition as it stood - requiring travel from China - was unable to allow broader testing of travellers who'd arrived from other countries.
"It wasn't until later that we learnt that there were other countries experiencing significant Covid-19 transmission early on in 2020."
Vipond and his co-authors Dr Elizabeth Becker and Chris Mansell also noted how the man's PCR test showed genetic material could still be detected many months - 201 days in this case - after infection.
Because having viral RNA in a PCR didn't always correlate to infectivity, the researchers suggested that serology should be used with repeat PCR to find out the "age" of infections revealed in positive tests.
Ultimately, the curious case supported the Government's call to lock down on this week one year ago, Vipond said, along with the benefits of following health advice for flu-like illnesses.
Otago University and ESR virologist Dr Jemma Geohegan also pointed how New Zealand wasn't testing enough over the earliest weeks, and that case definition was "very limited".
"We likely missed many cases in a similar way," she said.
"It also goes to show how prevalent it was in other parts of the world, very early on."
Te Punaha Matatini Covid-19 modeller Professor Michael Plank said genomic analysis had since told us that about 80 per cent of the first cases fizzled out without spreading far.
"However, the fact that this cluster spread to five household members makes it more likely that this would have caused ongoing transmission chains in the community during March 2020," he said.
"It's quite possible that these transmission chains were only broken after the lockdown began on March 26.
"If we hadn't locked down when we did, some of these transmission chains could easily have generated exponentially growing clusters of their own."