New Zealand's Covid-19 death rate is rising and expected to remain high for at least several more weeks, as the virus works its way into more vulnerable communities.
But experts say a dearth of detailed data is clouding our picture of just who is becoming severely sick and dying from Covid-19.
New Zealand's total number of publicly-reported Covid-related fatalities reached 338 today, as officials announced another 22 deaths.
A Herald analysis of Covid-related deaths reported to date found about three quarters occurred in people aged over 70, while about 80 per cent had occurred since the start of the year.
About a third of deaths involved cases who weren't fully vaccinated, with roughly the same proportion involving cases having received two doses only.
However, fewer than 50 of those mortalities have so far been confirmed as being caused by Covid-19, with the data capturing any cases of someone dying within 28 days of testing positive.
The Chief Coroner's office told the Herald this week it was investigating 25 active cases where the deceased person tested positive at death, with no determinations yet made in any of them.
No data was available to offer a breakdown of what variants and subvariants were involved in the deaths.
Director general of health Dr Ashley Bloomfield said the number of deaths linked to Covid-19 appeared to be rising - and the total number of deaths per million people now was 59.
"It's important to point out that even with that increase, New Zealand still has by far the lowest cumulative death rate from Covid-19 of any OECD country and one of the lowest in the world."
Bloomfield likened a recent comparison between New Zealand and the United States to "comparing pears with beef steaks".
"If you look back and the various outbreaks that the US has had, including their Omicron outbreak most recently, the death rate so far in our Omicron outbreak, and we are now three months into it, the death rate here is a fifth of what it was at the peak of the US Omicron outbreak."
Modellers told the Herald they expected hospital and mortality rates to remain high over the medium term – perhaps one to two months – and even as case rates fell away.
As the Omicron wave subsided, Covid-19 Modelling Aotearoa's Dr Dion O'Neale anticipated a "slow transition" from hospitalisation being driven by infections in younger Kiwis to older people, which was in line with overseas experience.
"You've essentially got one trend that's going down slowly, while the fraction of cases that are in more at-risk individuals is still going up over time."
Fellow modeller Dr Emily Harvey said Covid-19 cases among young Kiwis peaked early in the wave, and their rates dropped faster.
"In older age groups – or those over 70 – rates are either flat or still on the way up," she said.
"And these are the age groups where we get a huge proportion of hospitalisations and deaths ... so case rates among these groups are the thing we need to be watching for."
Professor Michael Plank, also of Covid-19 Modelling Aotearoa, said UK data suggested the risk of death roughly doubled with an extra five to eight years of age.
"We are now seeing a similar pattern here too with deaths concentrated in older age groups and I expect this trend will continue.
"In New Zealand, cases appear to have peaked now but because deaths lag cases, I'd expect the death rate to continue at around the current rate for another couple of weeks before tapering off."
Plank added that unvaccinated people remained at much higher risk than those who are up to date with vaccinations, and were an estimated five to 10 times more likely to end up in hospital.
Hong Kong's current Covid-19 crisis offered a stark illustration of just how much of a difference vaccines were making for us.
"Like New Zealand, Hong Kong is having a major Omicron wave after previously following an elimination strategy," Plank said.
"But unlike New Zealand it has poor vaccine coverage in older age groups and as a result it has had over 7000 deaths - in a population of about 7.5 million."
Harvey said we could also expect the ethnic make-up of reported cases to change as the outbreak continued to move beyond Auckland.
"When the outbreak was first taking off, we saw the largest spread through Counties-Manukau DHB, and infection rates were highest among Pacific people," she said.
"That's linked to a known increase in infection risk that comes with being in large or over-crowded dwellings, or working in customer-facing roles, or in workforces like manufacturing, transport and distribution."
Although Māori and Pacific people typically had younger population age structures and were at higher risk of hospitalisation – something linked to higher health inequity, deprivation and pre-existing health conditions – younger age wasn't as great a protective factor as for Pakeha people.
"As the outbreak has spread out from Auckland, the demographics have shifted," she said.
Generally, however, O'Neale and Harvey said analysing the precise risk of death and hospitalisation in New Zealand by vaccination status was difficult, given a lack of publicly reported data from the Ministry of Health.
"In order to do that calculation in a fair manner, we'd need to be publishing our hospitalisations and fatalities split by vaccination and age – not by just one or the other," Harvey said.
"We can't just look at the rate of hospitalisation by people who are boosted without controlling for age, for instance, because we've got higher booster uptake in those older age groups that are the most at risk."
Another confounding factor with calculating hospitalisation rates was that the number of unvaccinated people according to the Ministry of Health's Health Service User (HSU) population estimate differed greatly from that of Census-based population estimates.
"The truth is probably somewhere in between, but without good data, we really can't say," Harvey said.
Perhaps more importantly, the modellers said we didn't know what New Zealand's overall patterns of infection looked like – modelling estimated about 1.7 million total cases as at last week.
O'Neale said, "That's why we really need an infection prevalence survey.
"It wouldn't be perfect, but it would give us much more information about what communities are most at risk of hospitalisation with this virus in New Zealand."
Otago University epidemiologist Professor Michael Baker agreed a lack of data was hampering our ability to pick apart what was driving the wave.
"At the moment, we have trouble knowing virtually every key measure of what is occurring," he said, adding that hospitalisations remained our best indicator.
Former Act president's lawyer claims sentence was too harsh, calls for home detention.