Visualising population health and closing equity gaps
“I want to ensure that we have a strong Pacific voice in the analytics that we do.”
Namoe Tuipulotu, Head of Pacific Intelligence & Insights, Te Whatu Ora
Data played a critical role in informing the pandemic response in Aotearoa New Zealand.
Data analytics was used to monitor the spread of the Covid-19 virus, identify hotspots, and make decisions about public health interventions.
In the Northern region of the country, analytics was also used to target people most at-risk for complications from the virus.
Māori and Pacific regional coordination hubs had direct access to a Covid-19 dashboard, displaying clinical risk scores for local people who had tested positive.
Namoe Tuipulotu – Head of Pacific Intelligence & Insights, Te Whatu Ora, was working in the Pacific Hub at the time and says it is important to have Pacific people developing these types of dashboards if they are meant to reflect and show what’s happening for Pacific communities.
“By doing that, you get a true understanding of what the system is delivering as a whole, but also
get a better understanding of where Pacific people fit in that space,” she says.
Te Pae Tata Interim Health Plan 2022 says the interpretation and use of intelligence about communities will be led and interpreted by those communities, including Pacific people and Tāngata whaikaha – disabled people.
Tuipulotu says there is an intention to build an intelligence function within Te Whatu Ora Pacific Health Directorate.
“I want to ensure that we have a strong Pacific voice in the analytics that we do,” she says.
“It is about building relationships and partnerships across the system, but also making sure that we have key people in roles to provide the translation between what is happening in the business and how we present information to the communities that need it.”
Northern Region Health Coordination Centre (NRHCC)
As the Covid-19 pandemic spread around the world in early 2020, health districts in the Northern region of New Zealand worked rapidly to develop a regional view of hospital capacity and occupancy.
The Covid-19 Dashboard was used to support Covid-19 testing, vaccination, managed isolation and community care and is being expanded to other areas such as planned care, pulling from national datasets.
Developing a regional view
In March 2020 the Northern region was planning for an inundation of Covid-19 cases and for the hospitals to be overwhelmed, says former Head of Analytics at Waitematā, Delwyn Armstrong.
This meant the regional analytics teams, which includes Auckland, Counties Manukau,Waitematā and Northland Districts, were initially focused on getting a regional view of hospital capacity and occupancy.
Within three weeks, the team had developed a regional data store and soon after they had visibility of real time hospital capacity and occupancy across the region: something never achieved before. Due to the success of New Zealand’s lockdown, the hospital view was not as urgent as predicted, but as testing ramped up the dashboard was used to display a regional view of Covid-19 test results.
Armstrong, who is now Director Health Analytics and Insights, TeWhatu Ora – Health New Zealand, says after vaccinations began the dashboard displayed immunisation data and was also used to monitor people in quarantine or managed isolation.
When the Delta wave hit, it started to be used in relation to patient care, helping to direct ambulances carrying Covid-19 patients by identifying sites with ‘Covid-ready’ beds to care for them, she says.
As the level of infection increased, the majority of Covid patients were now in the community, and the focus turned to predicting who might need high-level care.
Visualising a clinical risk score
As the number of Covid-19 cases quickly grew, the regional dashboard was populated with more and more data.
This enabled the regional team, led by health informatics fellow at the i3 Institute for Innovation and Improvement, Cheng Kai (CK) Jin, to create a machine learning algorithm to predict the risk of hospitalisation.
Data around demographics, vaccination status, long term conditions, medications, and test results was used to create the clinical risk score, which was then automatically calculated and displayed in the operational dashboard.
This helped clinicians decide who needed to be seen and whether they met criteria for anti-viral drugs and pulse oximeters for monitoring in the community.
“The dashboard enabled data driven allocation of health services,” says Armstrong.
Data was shared with primary health organisations twice a day, so general practitioners could see all of their patients who had recently tested positive and what their risk scores were.
The Māori and the Pacific Regional Coordination Hubs had direct access to the dashboard.
Namoe Tuipulotu, Pacific
Data enabling an equity focus
Namoe Tuipulotu – Head of Pacific Intelligence & Insights, Te Whatu Ora, was working in the Pacific Regional Coordination Hub when she got involved in developing the dashboard.
“It is really important that we have Pacific people developing these types of dashboards if they are meant to reflect and show what’s happening for Pacific communities,” she explains.
Tuipulotu says the clinical risk score was particularly useful for identifying high risk patients and visualising who had been contacted or assessed within 24 and 48 hours. This information could be filtered by ethnicity and age group.
“It enabled us to see whether we were serving our Pacific people appropriately,” she says.
For example, a user could extract a list of high-risk Pacific patients who had tested positive in the last 24 hours.
“In my experience if you can’t tell the clinicians who a particular patient is they may lose trust in the data, but we were able to get down to NHI (National Health Index) level data, allowing clinical teams at the hub to sense check whether or not the data was accurate, which really helped with data quality,” Tuipulotu explains.
“It highlighted to the director of Pacific Health, who could feed up to the Executive Leadership Team, what was being done by these Pacific hubs and providers. It was really helpful to be able to drill down in different ways, because the data my manager might want would not necessarily be the same as what the doctor on call that day would want to see.
“Data is power so having this information readily available was extremely important during the pandemic,” says Tuipulotu.
“Having the right data also makes your processes so much more efficient.”
Delwyn Armstrong, Director
A scalable platform for the future
The NRHCC regional data store was originally hosted on-premise, but was shifted to the cloud as the amount of data coming in started to multiply, ultimately holding around four million Covid-19 test results.
Armstrong says this was “a real step change”.
“It was hard, but it really changed things for us in terms of scalability,” she explains.
The Covid-19 Dashboard now has around 60 different pages and Armstrong describes it as “like a big web site, providing windows to multiple different data sets in a completely integrated way”.
Data is key to the health system’s planning to enable it to ‘catch-up’ on services and care that was unable to be delivered during lockdown.
Armstrong says a range of initiatives are now underway, such as bringing together all planned care data, in order to nationally tackle wait lists.
“We’re looking to use various data sources to target the people who’ve been waiting the longest and who have the greatest need,” she says.
“Having these analytics tools means that decision makers can put in different scenarios to look at how we can shape our services.”
Tuipulotu strongly believes that “if you’ve got the right information readily available, you can make better decisions.
“Analytics is about shifting from counting widgets to starting to understand outcomes and how we can know we are truly making a difference for our patients,” she says.