The whole of New Zealand is vulnerable to Covid-19 outbreaks – but some communities are much more than others. What does this disparity look like? And what are the inequities that drive them? Researchers Dr Matt Hobbs (University of Canterbury), Dr Steven Turnbull (University of Auckland) Dr Emily Harvey (Market Economics, Te Pūnaha Matatini) and Dr Dion O'Neale, (Te Pūnaha Matatini, University of Auckland) put some context behind 10 maps that provide the concerning picture of our uneven risk from the virus.
Just before we start: can you give us a quick explanation of "vulnerability maps"? You've also built a fascinating tool called the Aotearoa Co-Incidence Network model that draws on this kind of demographic data to track and predict Covid-19 outbreaks. How does that work?
The Aotearoa Co-incidence Network (ACN) is a model that connects up people if they are co-enrolled at the same school or co-employed at the same workplace – which we call a co-incident.
It then aggregates people together in smallish spatial units called SA2s, which include around 2000 people, although more in cities but fewer in rural areas.
Based on these co-incident links, the model counts up the number of possible interactions between people in each pair of SA2s.
This gives us a dense network where locations are linked not by people moving between them, but by people interacting with someone from that location.
Attending a shared workplace and school are obviously only some of the ways that people can interact, but they're an important one that it's possible to learn about from administrative data in the Statistics NZ Integrated Data Infrastructure (IDI).
They're also one of the ways that people interact every day, unlike some other events that might be less frequent.
This makes them a good place to start looking for interaction patterns.
People can try the model out for themselves here.
Okay, so let's look at the first map. What does it show? And how have you calculated the transmission risk here?
Each one of the co-incidence links, or potential interactions, in the ACN is a possible pathway for transmission of a contagious disease like Covid-19.
In order to find which areas have the highest risk of transmission of Covid-19, or other contagious diseases, we calculate a measure called PageRank centrality.
PageRank was the original algorithm used by Google to determine the importance of web pages in search results.
For our purposes, instead of looking at the popularity of different webpages, we use PageRank to determine "popular" areas that have many strong co-incident connections in the overall network.
For each SA2 in the network, the PageRank algorithm assigns it a number that measures both the number of connections that an area has and the importance of those connections.
For example, links that connect one region to another region that also has a high PageRank score will contribute more than a connection to a lower scoring region.
Because it measures both the number and strength of connections to a region, PageRank centrality is a good measure for estimating the transmission risk of that region.
Imagine that you dropped an infection into the network and it spread from one region to the next, following the strongest links more than weaker ones.
In this scenario, the infection would end up transmitted to the regions with high centrality scores more often than the regions with lower scores.
The figure above shows in blue the SA2s in the Auckland region that are in the top third of transmission risk scores for the country.
Since Auckland is a highly connected city, with a high number of large workplaces and schools, most of the SA2s in the city are in the top third of the transmission risk nationally.
We also see that the distribution of high transmission risk locations can be a little patchy.
This is because the transmission risk is based on where people live.
SA2s that have lots of workplaces, but few dwellings, won't rank so highly on the transmission risk, even though they might be where the interactions that lead to that transmission take place.
These vulnerability maps below also measure vulnerability according to health information. What can we tell from these two maps? Can you give some examples of health risk, and why some areas have different profiles?
Vulnerability itself can be understood in many ways and may mean different things to different people.
It is important to acknowledge that we only capture some aspects of what being vulnerable means.
We based our vulnerability data on Covid-19 risk factors identified by the Ministry of Health.
Important factors for vulnerability to Covid-19 are age – specifically people older than 65 years – and those with long-term health conditions, or living with socio-economic deprivation.
Long term health conditions include things like cancer, cardiovascular conditions, diabetes, renal conditions, and respiratory illnesses.
Health inequities, and hence vulnerability, are also often correlated with ethnicity due to experiences of systemic racism and the ongoing effects of colonisation.
We have combined the information about which regions have the highest health vulnerability with the data on which regions also have the highest transmission risk.
These are the purple areas produced by overlaying the red, representing high health vulnerability, and blue, representing transmission risk.
These areas are at the top of the list both for where Covid-19 might spread to, based on people's work or school interactions, and where it would be more likely to lead to adverse outcomes due to a high prevalence of health factors for those areas.
While most of the Auckland region has high transmission risk, the highest health - or "health-age" - vulnerability regions tend to be further away from the city centre.
In the south, these regions span from Māngere, through Papatoetoe and Ōtāhuhu, to Manukau and Manurewa. In the west they include parts of Massey, Te Atutu and Glen Eden.
Unfortunately these are also some of the areas where we've seen clusters reported in the August 2021 outbreak.
What patterns do we see emerge when we focus on just socio-economic factors alone?
Looking at socio-economic vulnerability really highlights South Auckland, especially areas around Auckland airport, as an area with increased levels of deprivation and thus being increasingly vulnerable to an outbreak.
People in these areas do tend to be younger, but the area does have increased prevalence of long-term health conditions and also higher transmission risk.
There also tends to be a higher percentage of Māori and Pacific families in this region, and evidence shows that they communities may experience adverse consequences of Covid-19 at a younger age on average.
Away from Auckland, and across the wider North Island, we see similar patterns emerge around the risk of spread based on health data.
When we zoom out to the whole of the North Island, it's difficult to see the areas of high vulnerability that are associated with some urban areas, because these cover a smaller space.
However, this bigger picture reveals that we also have high health vulnerabilities in many of our more isolated communities such as the Far North and the East Cape and Mahia Peninsula.
But to a larger extent, we can see socio-economic indicators and their influence on transmission.
What is very clear is that Auckland and Hamilton tend to have a really high risk of transmission, which is unsurprising given the size of the population and the number of connections shared between different neighbourhoods.
However, our data also show areas of high concern across places such as New Plymouth, Tauranga and Napier, especially in urban areas.
These areas have both a risk of transmission and a high prevalence of underlying health conditions that could make those communities more vulnerable to an outbreak.
We can also see the incredibly stark inequities in terms of socio-economic vulnerability, where many communities outside main urban centres, such as the Far North and the East Coast are much more vulnerable to the effects of any transmission of Covid-19.
While many of these more rural areas are protected by having lower transmission risk, we cannot underestimate the serious, devastating impact Covid-19 would have if it was left unchecked in Aotearoa.
Lastly: what are the big take-aways that you'd like readers to keep from looking at these maps? And more importantly, what do they tell us about problems that need addressing?
These vulnerability maps provide a summary of what many people across Aotearoa may already know, that there is a huge amount of inequity and that some communities and regions will experience pandemics in different ways.
We can see that there are many areas that would be considered disadvantaged in socio-economic terms, and this may also be compounded for those who already experience adverse health conditions.
What we can identify in our research are the specific neighbourhoods where an outbreak of infectious disease, such as Covid-19, could have really tragic outcomes if adequate support isn't provided.
The findings we highlight can inform equitable responses to outbreaks that will reduce the chances of these outcomes happening.
By knowing which neighbourhoods are particularly vulnerable, and also have potentially a high risk of an outbreak reaching them, we can make sure that these areas are adequately prepared and equipped to cope.
This may be done through the targeted deployment of testing and vaccination centres to better serve these vulnerable, high risk communities.
Based on current research, it can be argued that this is an area that we need to do better in, with disparities in vaccine access and availability being linked to reduced numbers of Māori being vaccinated.