Health officials briefly claimed Canterbury had reached a 100 per cent vaccination rate earlier this week, and are now suggesting some parts of New Zealand may have even exceeded that.
The move sparked confusion as thousands of people have attended anti-vaccine and anti-mandate protests and marches in the region in recent months. The Ministry of Health has since apologised and revised its estimate down to 99.8 per cent.
But officials are now facing criticism over the way vaccine rates are calculated, with some believing they are overestimating the success of the rollout.
The ministry Health Service User dataset (HSU) only counts people who interacted with health services over a 12-month period, and anyone who didn't is not included - so the number of people in the HSU count is smaller than the actual population of New Zealand.
Despite the obvious limitations, a media release from the Ministry of Health on Monday claimed 100 per cent of eligible people in Canterbury aged 12 and over had received at least one Covid-19 vaccination.
While the ministry backtracked soon afterwards, revising the estimate down to 99.7 per cent and then back up to 99.8 per cent, it still only allows for 985 of the approximately 482,890 eligible people in Canterbury to be unvaccinated.
READ MORE: Police and NZDF vaccine mandates 'unlawful', High Court rules
A ministry spokesperson said it was likely the initial claim of "100 per cent" was a rounding error.
"While the ministry makes every effort to ensure the accuracy and consistency of our data, we acknowledge that this most likely occurred because of human error and we apologise for any confusion this may have caused," they said.
Even so, University of Auckland senior lecturer Dr Donna Cormack, who studies the collection of data in New Zealand, believes the ministry is "over-estimating" its vaccine rates, as it uses the HSU to work out its figures.
"The HSU population is likely to miss people who have the worst access to health services," Cormack said.
"It means that we can't be as confident in the rates estimated for some groups, as we know that some groups are more likely to be missed."
Cormack says the impact is likely to be more pronounced in groups that have poor access to the health system. She also works as a researcher with the University of Otago and studies the collection and classification of ethnicity data, particularly around racism in Māori health and health inequities.
"A few per centage points at a regional or national level equates to large numbers of people potentially, so this is important to know, and important to know who has been missed," she said.
"They should be being clear and explicit about the limitations of the data that they report, including the direction of any bias."
The ministry told 1News it had been clear about its limitations, and said it had the support of a "wide range of stakeholders" including "technical experts, data users and the ministry's Directors of Māori and Pacific Health".
Rachel MacKay, group manager operations of the National Immunisation Programme, argued the smaller HSU database provided "greater accuracy" in specific population groups and was more reliable than the 2018 census.
MacKay said their data included everyone enrolled with a primary health organisation and everyone who used health services like hospitals or prescriptions during a calendar year. But she admitted they were facing some limitations, "as with any dataset".
"[There are] a small number of people potentially not included because they have had no interaction with the health system in a given year," she said.
"By definition the HSU is not a total population estimate, so does not cover everyone in New Zealand."
That meant that in some "sub-populations" more people may receive a vaccine than expected, she said.
"Some people who did not engage with health services in 2020 will receive a Covid-19 vaccine," she said.
"This means coverage rates can exceed 100 per cent."
SHARE ME