There might be thousands of Kiwis out there who've had COVID-19 without even realising it, or could be suffering ongoing effects without knowing that's the cause, researchers say. 

A study published earlier this week found eight previously unknown cases of COVID-19 in New Zealand, by looking for antibodies to the disease in blood donated by nearly 10,000 people - known as seroprevalence. The cases were found around the country, in seven different district health board regions. 

The study estimated 0.1 percent seroprevalence and an infection-to-case ratio of 2.3, " suggesting some undiagnosed infections have occurred".

But how many? New Zealand has had 2244 confirmed cases of COVID-19 to date. If there are 2.3 infections for every confirmed case, that would be about 5160 cases. Similarly, if 0.1 percent of us have antibodies, that would be about 5000 cases. 

Both figures are well above the official total of 2244 and the figure including probable cases, 2600. But in a country that has managed to completely eliminate community transmission of COVID-19, are they feasible? 

Michael Plank is a professor in applied mathematics at the University of Canterbury and part of the Te Pūnaha Matatini research centre, whose disease modelling helped inform the Government's response to COVID-19. He told Newshub there is "quite a wide range of uncertainty around" those estimates.

"For example, from eight positives out of a sample of 9806, the 95 percent confidence interval is from about 1700 infections to 8000."

Then there are the biases in the data collected. Senior research fellow at the University of Otago's Department of Public Health Amanda Kvalsvig said people who donate blood aren't typical of the wider New Zealand population. 

"Some people who are at increased risk of COVID-19 aren't eligible to donate because they have a medical condition - for example people who have diabetes and need insulin to control it... People over the age of 70 years are also less likely to be blood donors while generally more at risk of infection than others, and Pasifika might be missed in the same way."

• New Zealand Navy personnel and family members test positive in Canada
• High-profile Norwegian conspiracy theorist who said COVID-19 was a hoax dies from the virus after hosting illegal house parties
• Ministry of Health tells Auckland GP 'we don't need you' in COVID-19 vaccine rollout
• Dozens may be missing regular COVID-19 swabs

Some might even be excluded from donating blood because of a diagnosis of chronic fatigue syndrome which might actually be long COVID, where people continue to suffer symptoms long after the initial infection has gone. 

"By thinking through the biases - working out who might have been missing from the blood donor sample and why - you could develop a logical strategy by estimating the number of known cases in those missing people," Dr Kvalsvig told Newshub.

"Then you could use standard epidemiological methods to estimate how many more cases might have been found in the study if people who don't typically donate, had actually donated blood. That would give a better idea of the total number of missed cases, although there would still be uncertainty around the precise number...  I think it's too soon to jump in with an estimate, but a little more work, essentially just following the logic, could arrive at a more robust estimate than we have at the moment."

The reason it's difficult to know if COVID-19 is spreading silently in the community is that many cases appear to be asymptomatic - they don't show any signs of illness at all, but can still spread the virus. Given enough infected people however, with the unlimited movement New Zealand has largely enjoyed since the first lockdown ended, any widespread silent outbreak would inevitably rear its head in symptomatic cases. 

The fact this hasn't happened suggests most of the potential undetected cases happened in March and April last year, Dr Plank said.

"A few could be from August but it's unlikely there would be a significant number more recently than that because our rates of community transmission have been so low."

Dr Kvalsvig said it would be "interesting to know if the undiagnosed cases detected in the blood donor study can be linked to a known cluster".

"That information could be helpful for narrowing down the confidence interval a bit. Are we looking at undiagnosed cases around the edges of a known cluster, or are we seeing signs of completely unknown (albeit probably small) clusters?"

Although there's no longer any transmission risk from people infected a year ago, Dr Kvalsvig said it is still important to diagnose them if possible.

"There's concerning evidence emerging about the potential for organ damage from inflammation during the acute illness. That damage might result in further health problems down the track so it will be important for people to know that they've had this infection."