PHOTO: Timon Studler/Unsplash

How many New Zealanders haven’t caught Covid-19? – Expert Q&A

New Zealand’s second Omicron wave might just be easing up, with the latest data showing we seem to be in a downward trend. So are those who never got Covid-19 now in the minority?

The SMC asked experts about what the latest data say about rates of infection, what we still don’t know, and how this affects wider immunity to the virus.

Professor Michael Plank, Te Pūnaha Matatini and University of Canterbury, comments:

What do the data tell you about how many New Zealanders have or haven’t been infected?

“The true number of people who’ve had Covid-19 is unknown because we only know about cases who get tested and report their result. The proportion of infections that are reported is probably somewhere between 40% and 65%. It’s unlikely to be much less than 40% because at least 40% of all 20-25 year olds have already reported a case. On the other hand the reported infection rate is typically less than 65% of the infection rate in routinely tested cohorts such as border workers. This means it’s likely that at least half of New Zealanders have had Covid-19, although those that haven’t probably still represent a significant minority.”

What does this mean for our herd immunity to Covid?

“The lack of an accurate estimate for the true number of infections creates uncertainty about how much immunity there currently is in the population. This makes it harder to estimate the potential size of future waves of Covid-19 or the impact of new variants. The best way to get an accurate picture of the true number of infections would be to systematically test a representative sample of the population on a regular basis. This would provide high-quality ongoing data about the level of Covid-19 in the community and reduce our reliance on self-reported test results.”

Conflict of interest statement: “Michael Plank is funded by the Department of Prime Minister and Cabinet for mathematical modelling of Covid-19.”

Professor Mike Bunce, Principal Scientist (Genomics), Institute of Environmental Science and Research, comments:

What do the data tell you about how many New Zealanders have or haven’t been infected?

“Last week the official tally of registered cases in Aotearoa New Zealand was approximately 1.62 million people. We know that this number is under the true number of infections due to asymptomatic cases, non-reporting of positive cases, reinfections and failure/unwillingness to test (even if unwell). The Ministry of Health is planning a prevalence survey that will be key in ascertaining the true number of infections, as a person’s antibody profile can tell us if a person has been previously exposed to the virus (as opposed to the vaccine). In the UK, this approach showed that >95% of the population has been exposed to SARS-CoV-2 in the pandemic, many of these are recent infections. Looking at one cohort, primary school students, in November 2021 approximately 40% (pre Omicron) of students had been exposed. In March of 2022 this number was >80%.

“What we do know from our genomic surveillance here in Aotearoa is the approximate proportions of each of the variants (see figure and embedded table). This temporal infection history is important as the variants of SARS-CoV-2 that we are exposed to (coupled with vaccines and time since infection) are key factor in determining the traction new forms of SARS-CoV-2 will have. For example, the reason that the BA.5 wave here in New Zealand was at the lower end of the modelling predictions could be because our population was heavily exposed to the BA.2 variant of Omicron which is closer to BA.5 than original BA.1.

Image credit: ESR

What do we still not know? What data do we still need?

“The serology survey (soon to be conducted) and modelling of this data in concert with hospitalisation rates and wastewater will likely provide a more nuanced picture of the total number of Kiwis that have been exposed to Covid-19 – but working out the ‘true’ number of infected (and reinfected) people will always remain a challenge. High rates of participation in initiatives like FluTracker may also facilitate better estimates of symptomatic infections across the Motu. There are aspirations here at ESR to better characterise the (often seasonal) viruses that sweep Aotearoa using the genomic toolkit that we have developed during the pandemic. For example, PCR testing (or genomics) that can distinguish RSV, flu, Covid-19 or other viruses that give us ‘colds’ can be used to better understand the exact virus that is driving illness, and what strain of each virus is present.

What does this mean for our herd immunity to Covid?

“Determining how susceptible/resilient a population is to new wave of Covid-19 is a complex interplay of; (i) vaccine and booster rates (ii) percentage of the population with protective immunity (iii) time since infection/vaccines (iv) new variants of SARS-CoV-2 that might escape prior immunity (v) Mask use and other social factors such as staying at home while sick (vi) the efficacy of pharmaceuticals to treat Covid-19 patients and reduce viral loads and, (vii) the development of new vaccines that might reduce/prevent infections.

“Taken together, this shopping list of factors paints a picture of why it is difficult to predict how protected or vulnerable New Zealanders might be. While we would like to be more certain what the virus’s next move might be, history has taught us that diseases are adept at changing the ‘playing field’ and as we adapt our strategies we nudge the virus in different directions. So while we may well estimate a level of ‘herd immunity’ that might be protective one week, it may change the next. Predicting the trajectory of ‘new’ RNA viruses such as SARS-CoV-2 adds another layer of difficulty as they continue to rapidly explore different ways to gain a fitness advantage in a new host.”

Conflict of interest statement: “ESR receives funding from the Ministry of Health to process clinical samples for whole genome sequencing and wastewater.”

Dr Dion O’Neale, Project Lead, COVID Modelling Aotearoa; and Senior Lecturer, Physics Department, University of Auckland, comments:

What do the data tell you about how many New Zealanders have or haven’t been infected?

“Data from the Ministry of Health records just under 1.7 million confirmed cases of COVID-19 in Aotearoa New Zealand. Of these, just over 28,000 cases are recorded as probable reinfections in people who have been a confirmed case of COVID-19 twice or more. Both the number of confirmed cases and the number of recorded reinfections will be a significant under-estimate of the true number of infections. This is because not all infections end up being recorded as a case.

“The ratio of the number of infections in the community to the number of confirmed cases that we know about is called the case ascertainment rate, or CAR. This rate will typically change over time, depending on a number of factors such as how easy it is for people to access tests and how inclined they are to seek a test and to report the result. Most estimates of the CAR put it above 33% but probably below 50%. If we ignore reinfections this means that the number of infections that we have had in Aotearoa is (at least) roughly twice the number of confirmed cases — so over 3 million.

“The effect of the case ascertainment rate on estimates of the number of reinfections is a little more complicated. This is because the CAR applies for both the first infection and for later infections. If someone’s infection was not recorded as a confirmed case the first time around then – even if their second infection was recorded as a confirmed case – it will not be captured as a reinfection and we will be underestimating the number of reinfections, or possibly overestimating the number of unique people who have been infected. This can matter for disease modelling as the immunity history (the combination of past vaccinations and infections) of people affects what the landscape looks like for how susceptible the population might be for future infections, including with any new variants.

“Under-reporting of infections can happen for a number of different reasons; some people may have confirmed their infection with a rapid antigen test, but not reported the result; some people may have been symptomatic and not wanted to use a RAT to confirm their infection, or may have had a false negative on their RAT due to test timing or technique; and a sizable number of people may have not even had symptoms to alert them that they might need to test for an infection. Estimates are that as many as 40% of infections may be asymptomatic. Unless these people know that they have been in contact with a confirmed case – for example, a family member who has tested positive – or unless they are testing independently of symptoms, these people may never have had a reason to test during their infection.

“Methods like wastewater sampling (e.g. see the ESR dashboard) can help us to better understand patterns in infections and cases, based on how much RNA from SARS-CoV-2 is detected in wastewater, but such methods can be tricky to calibrate. And while wastewater methods can be useful for indicating when the case ascertainment rate in a region might be changing (i.e., if the amount of detected SARS-CoV-2 RNA is going up but the number of confirmed cases is not, then case ascertainment is probably getting worse), it is more difficult for it to tell us what the estimated number of total infections behind that signal might be.

“One of the best ways to get a good estimate of the true number of infections in the community is through what is called an infection prevalence survey. This includes methods like that taken by the ONS in the UK. An infection prevalence survey uses regular samples from sub-set of the population to calculate the prevalence of COVID-19 in the community over time and accounting for differences in location and population.

“In Aotearoa, the Ministry of Health announced in July that they would be starting such and infection prevalence survey. Once underway, this will give extremely useful information about infections in the community and how they are distributed. Infection prevalence surveys give an indication of current infections. They are often combined with something called a sero-prevalence survey. Rather than looking for indication of RNA or protein from a current COVID-19 infection, sero-prevalence surveys look for the presence of COVID-19 antibodies from past infection. While antibody markers do fade over time, seroprevalence surveys can help to give us a view back to past infections, even when people have now recovered.

“Information from this sort of survey is useful in figuring out how much protection from past infection communities are likely to have and what the susceptibility landscape looks like for the population in Aotearoa.”

Conflict of interest statement: “I, along with others from COVID Modelling Aotearoa, am funded by the Department of Prime Minister and Cabinet to provide advice on the COVID response and from a Health Research Council grant to look at equity related to COVID in Aotearoa.”