Modellers trying to pick the future path of Covid-19 face a dramatically more complex landscape in 2023 – but there’s hope future waves will be “more muted” than last year’s big battles with Omicron.
Modellers trying to pick the path of Covid-19 face a dramatically more complex landscape in 2023 – but there’s hope future waves will be “more muted” than last year’s big battles with Omicron.
Ahead of our fourth year of the pandemic – and our second winter with Covid-19 – rates of deaths and hospitalisations are tracking close to their lowest levels since Omicron first took off here around a year ago.
Nonetheless, those numbers weren’t insignificant: at the most recent weekly update, there were nearly 180 people in hospital, while 18 more Kiwis – including a child younger than 10, and another person in their 20s – died with the virus.
As at the end of the week, the seven-day rolling average of reported cases stood at more than 1620, slightly up from the week before, but this didn’t yet signal our first wave of the year.
“Most regions in the South Island have seen a recent rise in cases although this has started to taper off or come back down again in many areas,” said Professor Michael Plank, of Covid-19 Modelling Aotearoa.
“This is consistent with a bump in transmission as schools and universities restart for the year.”
Reporting patterns in parts of the North Island had also been affected by the cyclone and flooding, making it tougher to interpret trends.
“It’s likely the North Island will also see a modest bump in cases, lagging behind the South Island by a few weeks.”
As for whether New Zealand could expect another winter wave, fellow modeller Dr Dion O’Neale said the coronavirus hadn’t been around long enough to show any clear seasonal trend.
“There’s a good chance that it does – we know that influenza has a seasonal component, for instance – and then we’ve also got to consider the behavioural element, with people being shut inside more.”
Plank and O’Neale expected that, if our Covid-19 landscape remained largely the same, we’d keep seeing case bumps over time, perhaps spaced a few months apart.
“There is a mixture of variants spreading at the moment, but at the moment none of them is growing quickly enough to spark a major wave,” Plank said.
“Cases will continue to rise and fall as our immunity waxes and wanes, but hopefully future waves will be much more muted than last year.”
At this point one year ago, New Zealand was in the thick of its first Omicron wave, with an under-pressure health system trying to manage more than 200,000 active cases.
As daily numbers topped 20,000, a quarter of Auckland City Hospital’s staff was put out of action.
Twelve months on, the country’s official case count stands at more than 2.2 million – although perhaps eight in 10 Kiwis have now been exposed – hundreds of thousands of us more than once.
Of those cases still being officially reported, about four in 10 were reinfections.
For modellers like Plank and O’Neale, the job has changed dramatically.
At the start of 2022, they had a largely tidy handle on the population’s state of immunity because it was nearly exclusively conferred by vaccines, which some 96 per cent of the eligible population had received by mid-summer.
Simpler, too, was the viral landscape: the country was dealing with a dwindling number of Delta cases, and an Omicron knocking at our door was spreading in just one of two forms.
“There were basically three categories of person: unvaccinated people, those who’d had one or two doses, and those who were boosted,” Plank explained.
“So, if we had an estimate of the immunity level of those three categories, we could then use that to model the spread of the virus and its impact in terms of hospitalisations.
“Of course, it’s just a lot more complicated now, because there are so many more different combinations of immunity that people have.”
The upcoming rollout of the bi-valent booster among the over-30 population will shift that picture again.
Circulating variants formed another part of the equation, with the latest ESR surveillance report showing a potpourri of Omicron sub-types driving cases.
They include BA.2 descendants CH.1.1 and BA.2.75 - making up 33 and 13 per cent of recent sequenced cases respectively – along with “recombinant” types XBB (25 per cent), XBF (14 per cent) and XBC (two per cent).
“At any point in time, there’s probably a dozen different variants out there in the community, and they each have slightly different consequences for immunity and reinfection,” O’Neale said.
“It’s almost got to the point where it’s so messy, and we can’t distinguish between all of the different combinations of variants and immunity, that we have to simplify our modelling assumptions.”
That meant boiling things down to some key parameters – namely how many people had been infected and/or boosted, and how long it had been since their last dose of the vaccine or virus.
“It’s a big simplification of reality, but it’s what we have to do in trying to model a situation like this,” Plank said.
Another confounding factor was that testing and reporting cases were likely dropping over time – meaning those figures we see reported by officials each week were growing increasingly unreliable as indicators.
Trying to fill in these blanks is a notoriously difficult exercise for modellers, who use what’s called the case ascertainment rate, or CAR, to estimate the ratio of confirmed infections to all of those missed ones.
As at late 2022, it was thought to be just 30 per cent - with a higher rate in adults but a lower one in children – meaning about two-thirds of cases are probably going undetected.
One tricky part of this is that the estimated CAR is always moving in line with our behaviour; and another is phantom infections.
Some international studies have suggested asymptomatic transmission may account for four in 10 cases, yet, without a national sero-prevalence survey built on regular blood samples in the community, it’s tough to say how many people have Covid-19 at any one time.
In the absence of a survey, modellers could turn to another “incredibly valuable” resource: wastewater surveillance.
Samples are routinely sent from dozens of screened plants across the country to ESR’s labs, where scientists can assess the concentration of the virus, and extract viral RNA from it.
When there’s enough of the virus in the sample to quantitate, scientists are able to convert that to a viral load of genome copies per day, per person – helping build a picture of infection prevalence in given catchment populations.
“It doesn’t depend on people deciding not to test or upload their test result,” Plank said.
“Ideally, we’d have a random sample of the population to use to estimate prevalence, but that costs money and time.
“The wastewater data isn’t a replacement, but it’s helping provide some information about prevalence.”
For most of 2022, reported case numbers kept in step with what trends ESR scientists were reporting from wastewater, before the two plot-lines markedly diverged over December – indicating that fewer Kiwis were testing for the virus.
Since then, however, those lines appeared to have come back together, suggesting the drop-off had been linked to the holiday season.
“So, we can try to estimate how reporting has changed over time by looking at things like wastewater, and we’ve got some work on this in progress at the moment,” Plank said.
“We can also look at hospitalisations and deaths, which are not quite as sensitive to case reporting.”
For O’Neale, one of the biggest Covid-19 questions facing us over the longer term was whether a game-changing new variant emerges to upend the pandemic – and the immunity we’ve built to date through vaccines and infection.
Another was the unequal burden of disease – including the mounting problem that is Long Covid - falling upon different parts of our society.
One alarming Ministry of Health report released last week already showed disabled people were 4.2 times more likely to be admitted to hospital for Covid-19, and 13 times as likely to die from it.
“In a situation where more actions like isolation become voluntary, and it’s up to people just to follow guidance or best practice, then those who are well-off enough to do that, will, and will benefit from those protections,” O’Neale said.
“But for those not able to isolate because they feel they have to go to work – they’re the people who’ll keep getting infected.
“So, ultimately, the more protections we remove, the more the burden falls on vulnerable groups – especially Maori and Pasifika people, and those with disabilities, and those on lower incomes.”
The PGG Wrightson livestock agent is passionate about stock and auctioneering.