The Ministry of Health has been accused of "cooking the books" in New Zealand's Covid response by using data that undercounted Māori, inflated vaccination rates and may have led to higher rates of death and illness.
Some health statisticians say the dataset used by the ministry to measure vaccination rates undercounted Māori by as much as 20 per cent in critical age bands.
They argue that this flawed data didn't allow a proper understanding of how and where to target vaccination resources, which has contributed to gaps in Māori vaccination against Covid-19.
A Herald investigation has found there are two independent academic projects underway to assess whether Māori lives were needlessly lost to Covid-19 as a result.
In addition, Stats NZ has been reviewing the Ministry of Health's use of the Health Service Utilisation 2020 dataset to calculate vaccination rates.
It is about to report on the difference between the HSU 2020 and its own data and is expected to show vaccination rates are lower than those trumpeted by the Government.
HSU 2020 is considered flawed as it counts those who engaged with the health system during 2020. It was known Māori did so less regularly and when doing so sometimes identified as another ethnicity.
A dataset developed to compensate for those missing from HSU 2020 data showed significantly different vaccination rates for Māori than the official rate claimed by the Ministry of Health.
The dataset designed by researcher Dr Rawiri Taonui - whose expertise included indigenous human rights and racism - found Māori only reached the 90 per cent target for vaccination in Auckland and Canterbury district health boards.
In the other 18 DHBs, Māori had yet to hit the 90 per cent figure needed to provide good protection against Covid-19.
In contrast, the Ministry of Health said Māori had reached 90 per cent in nine DHBs - a count that would still leave 11 DHBs lagging below the 90 per cent cut-off that became the target to drop borders and open the country in the lead-up to Christmas.
Some of the differences in smaller and regional DHB vaccination rates are dramatic. Taonui's data shows only 82.5 per cent full vaccination coverage in Wairarapa compared to 91 per cent on the official ministry count and 81 per cent in Tairawhiti versus 87.2 per cent from the ministry.
The difference for boosters (third doses) needed to ward off Omicron is greater, even for large populations. Ministry figures say Counties Manukau has a 52 per cent booster uptake but Taonui's estimate puts this at only 35.7 per cent.
"They consciously, knowingly, cooked the books," said John Tamihere, chief executive of the Whanau Ora Commissioning Agency and Urban Māori organisation Te Whanau o Waipareira.
Tamihere said it was known there were 60,000 people in Auckland who identified as Māori yet were not in the HSU 2020 dataset for reasons that included the poverty skew.
"Māori only go to the doctor when they're crook. Going to the doctor costs money. They don't go for warrant of fitness checks."
NZ Māori Council executive director Matthew Tukaki, also chair of the National Māori Authority, said the struggle over data with the Government meant resources were not sent to the right places quickly enough.
"We lost a lot of time … we should have transferred some of the power and control to Māori early on.
"The data was flawed. The pandemic has shown us we need to do better with the collection and disbursement of data."
Tukaki said there would be ongoing problems with the use of flawed data, such as understanding the impact of Long Covid in the Māori population.
He said the question of data was one to be answered through a necessary Royal Commission of Inquiry into the management of the pandemic to learn lessons needed for future threats. "There will be another pandemic."
As part of that, Tukaki said he wanted to know if there were Māori infected, hospitalised or dead who didn't need to be.
There was a known particular risk to Māori when Covid-19 first emerged two years ago because of its contagion, demographic and respiratory characteristics.
A Waitangi Tribunal inquiry in December found the Government's handling of the pandemic in relation to Māori was a breach of Treaty of Waitangi principles. It specifically said the Crown's failure to collect data to "accurately and equitably inform the rollout of the vaccine" for Māori was a breach of the "principles of active protection and equity".
In August - as the gap between vaccination of Māori and non-Māori widened - director-general of health Dr Ashley Bloomfield said: "To further improve our vaccination reporting, we're now using what is called Health Service Utilisation or HSU data.
"Importantly, these data allow us to drill more deeply into vaccine uptake by Māori and Pacific peoples, including the numbers of first and second doses and also the rates per thousand in the population, and we can also report by DHB on each on those groups."
However, the Waitangi Tribunal finding and two High Court cases have criticised the HSU 2020 and its ability to capture the number of people who identified as Māori.
Rawiri Taonui, who built the alternative dataset, said the Māori population was undercounted by around 20 per cent in some age bands in HSU 2020. He said the dataset captured health users when it was known Māori had less access to health services than others.
He said the difference between the estimated resident population - a Stats NZ projection - and HSU 2020 was "at least 100,000 people".
The HSU 2020 Māori population was 766,569 while the Stats NZ figure was 876,600, he said. When it came to those aged 12 years or older - the age range for the most commonly used vaccination rate - the HSU 2020 figure for Māori was 571,052 while Stats NZ had 661,310.
Taonui said he built datasets to compensate for the shortfall and found significant gaps in vaccine coverage.
He said it led to a significant gap with Māori 8.5 per cent behind non-Māori. He said the booster gap was greater - those aged 18-plus were 18.7 per cent behind the general population. For Māori aged 18-34 years, just 26.6 per cent were boosted.
The greatest gap was in those aged 5-11 years which was only 33.6 per cent - 20 per cent behind the national average.
Taonui said the vaccination rate across the whole population including those aged under 5 years was key in an "Omicron hyper-transmissible environment". As of May 12, he said 64.1 per cent of the Māori population had one vaccination compared to 84.2 per cent of the total population.
"There are 314,700 unvaccinated Māori," he said.
Taonui said the flaws in the data used by the Ministry of Health meant it was less able to plan a vaccination strategy for Māori which was a factor in the Covid-19 impact experienced by Māori.
He said preliminary research results showed Māori carried the brunt of Covid-19 infection, hospitalisation and death during the Omicron outbreak, as with last year's Delta outbreak.
At 15.6 per cent of the population, Māori were more than 50 per cent of Delta cases, 38.6 per cent of Delta hospitalisations and 45 per cent of deaths.
Statistician Andrew Sporle, a lecturer at the University of Auckland, said the fact that HSU 2020 was "not robust enough" had been demonstrated in the two High Court cases brought to force the Ministry of Health to provide data on unvaccinated people to Māori health providers.
"Not only does it undercount people but it undercounts in particular age bands in the population - and undercounts those age bands that are really good at spreading respiratory viruses."
He said he had raised concern with the Ministry of Health over how it counted Māori since April 2020.
"I've been banging my head on it ever since. It's nuts. They've severely over-estimated vaccination rates. We still have populations severely at risk and we don't know what resources to put into those communities."
Sporle is a founding member of the Māori data sovereignty network, Te Mana Raraunga, the Global Indigenous Data Alliance and of AraTaiohi, which launched as the NZ Association of Adolescent Health and Development.
He said the data shortfall was one reason Māori had suffered greater adverse health effects in Covid-19 outbreaks even though it was a known issue and had been repeatedly raised with government Ministers.
Sporle said those adverse effects included whanau who had lost people, or people who needed hospital care, that otherwise did not need to die or become as sick as they did.
He said the Government needed to apply a different data approach to current and future vaccination campaigns, for Covid-19 and other current diseases or for future pandemics.
"They need to do 'different' and they have not got the capability at the moment to do 'different'. New Zealand doesn't do vaccination equity very well and we can't even count people very well."
Sporle said it was hard to understand why the Government continued using the same database when it was known - and had been shown - to serve Māori poorly.
Like Taonui, he has work underway to assess the consequences on Māori of an inequitable approach.
Lance Norman, whose role as director of health at the Whanau Ora Commissioning Agency has him involved in more than 100 Māori health organisations, said if the 100,000 missing Māori were included in the Ministry of Health dataset, the vaccination rate "would drop significantly".
Norman has worked as chief executive for the National Urban Māori Authority and Hāpai Te Hauora, New Zealand's largest Māori health agency, and sits on a number of expert advisory boards.
He said he had asked for 18 months why the Ministry of Health was using a dataset of health users when it was known Māori had unequal access to health care. "I think it's a lazy and easy dataset," he said.
He said it meant vaccination rates reported today were done on a "two-year-old number and a number that was wrong to start with".
Norman said the inequity was masked by the outcome which was still better than most comparable countries. "We've done not a bad job for the whole country and a better-than-average job for Māori."
He said it could have been better and should have been more equal - and would have been better and more equal if Government had listened.
Norman described the various phases of the vaccination strategy skewed against Māori in a way that said "you designed a system to make me lose", which made it more difficult to encourage those who were reluctant.
The flow-on effect could be seen in the number of Māori children vaccinated, which was significantly low. He said Māori households more often had multi-generational living arrangements bringing unvaccinated children in contact with vulnerable older people.
At a health select committee hearing this month, Minister for Covid-19 Chris Hipkins was quizzed over equity for Māori during the pandemic.
He replied: "Equity has been a consideration right the way through our Covid-19 response and … our Covid-19 response have been some of the most pro-equity responses a New Zealand Government has ever taken to any issue and I think that the overall outcomes we have achieved from our Covid response clearly demonstrate that."
Speaking to the Herald, Hipkins said "at the time it was adopted" the HSU 2020 dataset "was the most accurate and complete set of data that could be produced". He said it was more reliable than the Stats NZ population projection which was based on the 2018 Census.