Researchers have reconstructed New Zealand's main Covid-19 outbreak to find that one in five adults were responsible for up to 85 per cent of the virus' spread.
The new analysis, published in the journal PLOS One, has highlighted the importance of targeting super-spreader events to combat flare-ups.
It's also suggested children under 10 infected fewer people on average and were less likely to themselves be "super-spreaders", who were defined as infecting more than five others.
New Zealand recorded nearly 1500 cases of Covid-19 between February 26 and May 22 last year, before a nationwide lockdown and several other major measures effectively eliminated the virus.
In the study, Associate Professor Alex James and her fellow Te Punaha Matatini modellers used a wealth of case data - much of it collected through contact tracing - to tease out patterns around how the virus spread in those crucial months.
They found that, before the move to alert level 4, more than half of all domestic cases resulted in at least one secondary case.
But age played a part in how many others one infected person could pass the virus on to.
The modelling showed the effective reproduction number (R) - the average number of secondary cases - was expected to have been around 0.87 for children under 10, 1.49 for people between 10 and 65, and 1.51 for those older than 65.
"Although children under 10 were equally likely to infect at least one person, adults tended to infect more people than children under 10 did," the researchers reported.
Cases among adults and older people also had a "significant" chance - 6 per cent in the 10 to 65 group and 7 per cent in the over 65 group - of being a super-spreader.
Over lockdown, the R number dropped to below one for all those age groups except the over-65s - something that may have been down to aged care facilities being over-represented in the data from the later stages of the epidemic.
Overall, the researchers pin-pointed 29 super-spreaders - of which 21 had Covid-19 symptoms before lockdown began.
Of the remaining eight that had symptoms during lockdown, six were involved in aged-care clusters.
The study also highlighted that children under 10 tended to have a lower "secondary attack rate" - a measure defining the probability of an infection being spread among a group of close or susceptible people, like a household.
That was in line with studies overseas - as was the finding that "super-spreading" events were a major contributor to transmission.
"Our results show that among adults 20 per cent of cases are responsible for between 65 per cent and 85 per cent of transmission," the researchers said.
"This suggests that interventions targeting super-spreaders or super-spreading events may be particularly effective in reducing the spread of Covid-19.
"These may include restrictions on gathering size, particularly in closed environments or crowded spaces."
Meanwhile, another paper just published in the US journal Emerging Infectious Diseases, has underlined the crucial role that real-time genomic sequencing played in the country's next major outbreak.
While health officials battled to contain the Auckland August cluster - which ultimately led to 179 infections and three deaths - scientists helped link up cases by sequencing the genomes of positive samples.
In all, they were able to generate genomes from some 81 per cent of lab-confirmed samples - or 145 of the 179 cases - and then compared them with available global genomic data.
That quickly told them that the virus behind the outbreak was part of a single cluster - and thus had stemmed from a single introduction into the community.
"Indeed, the timing and length of lockdown measures were partly informed on the basis of these data," said the authors of the study, led by Otago University and ESR virologist Dr Jemma Geoghegan and University of Auckland research fellow Dr Jordan Douglas.
"Overall, real-time viral genomics has played a pivotal role in eliminating Covid-19 from New Zealand and has since helped prevent additional regional lockdowns, leading to substantial economic savings."
Still, they said the important tool had been limited by the "biased nature" of global sampling, including the contribution of very few genome sequences from certain regions.
"We therefore advocate that potential sampling biases and gaps in available genomic data be carefully considered whenever attempting to determine the geographic origins of a specific SARS-CoV-2 outbreak," they said.
"Analyses should consider all available evidence, including that from genomic and epidemiologic sources."