Responding to the Covid-19 pandemic required all parts of the public sector to work together, and in order to collaborate successfully and to protect citizens, data had to play a key role.
In response, the Welsh Government created the Covid-19 Strategic Data Visualisation Platform, a browser based tool that drew in and shared data from Welsh public bodies and other important data resources.
Fliss Bennée, co-chair of the technical advisory group at the Welsh Government, says it was in March 2020, six months after she had joined, that the need arose for leaders in the Welsh Government, local authorities, NHS trusts and the emergency services to understand where the balance of harm would be greatest and whether the NHS would become overwhelmed.
In addition, authorities needed to know when the protective measures were working.
It needed the specialists to spend their time analysing data, rather than collating it from a range of sources, cleaning it and then distributing the information to other stakeholders.
Bring landscape together
“There were some individual dashboards and datasets in Wales, but there was nothing that brought the Welsh landscape together”, Bennée says.
“We had over-skilled people doing repetitive, time-sensitive work when these people needed to be doing the analytics. My radiation specialist was phoning wards and hand cranking data in Excel”.
She knew that the Welsh Government needed to automate the collation and presentation of data, and that decision makers across the public sector needed new tools to enable them to determine risk, support people who were self-isolating, and model and visualise the impact of the pandemic.
The Technical Advisory Cell team (TAC) had started with a spreadsheet in the early days of the pandemic, but Bennée saw the need for a secure platform that all accredited members of the public sector could access.
“We didn’t have anyone in-house that could create what we needed,” Bennée says. “Our organisation’s platform was too tightly secured, and we didn’t want to create new databases or take other people’s data, we just needed to see it in the same place to interpret it.
“The regular statistics and digital teams within the Welsh Government were flat against the wall with demand, as were digital teams in the NHS.”
The TAC worked with software engineering firm Armakuni to develop the Covid-19 Strategic Data Visualisation Platform, bringing all the data sources together then testing and deploying at speed.
It involved a change of working culture, as for some it was the first time they had worked alongside agile software developers; but soon Welsh Government officials were able to create their own scrums and began using agile methodologies for looking at scientific output as a minimum viable product. In addition, these teams then began to look at their backlogs and blockers and deal with them.
“Giving people rapid access to that data meant that the time it took to onboard them to the Welsh Government systems was taking too long. In addition, these were Microsoft systems that required authentication and licences to access,” Bennée says.
Instead, the Covid-19 Platform now provides a centralised, easy to access solution for both the Welsh Government and related public sector bodies.
The platform has provided four distinct information panels dedicated to risk, data visualisation, scientific modelling and, during the worst of the pandemic, the ability to support those self-isolating.
Application programming interfaces (APIs) are used to ensure data remains on its original host platform, such as that for a local authority or police service; and single sign-on access replaced a process of having to provide Welsh Government access to NHS and local government systems.
The risk stratification panel was developed as part of the Welsh Government emergency response and provides a national dataset that can be assessed in under 30 minutes, giving an end-user a profile of the Covid-19 risk in the population. Developed using Microsoft Azure infrastructure-as-code, it uses automated testing of the infrastructure and data pipelines. Data is collected via the QCovid algorithm as an API.
The data visualisation tool allows policy makers to understand everything they need to know about the local Covid-19 situation in two minutes, and Bennée says this has been vitally important to accelerate decision making.
When these were in place, Doctors Brendan Collins and Craiger Solomons formed a rapid action team of Welsh Government subject matter experts, led by a product owner, working with data and platform engineers from Armakuni. To ensure the data visualisation was able to respond to the changing nature of the pandemic, new visualisations were deployed three times a week.
Pipelines for accuracy
As policy leaders were making decisions with this tool, data pipelines were automated to guarantee the accuracy of the data and replace any manual and error-prone processes.
The Covid-19 modelling service, the third element to the platform, has enabled the Welsh Government to run interactive experiments, effectively allowing policy makers to run a digital twin of Wales and plot the path for dealing with the pandemic.
A delivery team of the Welsh Government and partners, along with a team of scientists, devised a set of parameters for creating and running models of possible policy responses. It is enable by Kubernetes containerisation cloud computing technology, which ensures there is no limit to the scale of resources available whilst performing the experiments.
The fourth element to the dashboard, the Self Isolation Support Scheme for direct contact with members of the public, interfaces with the NHS Test and Trace App and provides an onboarding tool for those self-isolating to get support from the Welsh Government.
It was developed in collaboration with the NHS Wales Informatics Service and NHS Digital.
A digital token service was integrated to verify when citizens need to isolate.
APIs and cloud
Together the four elements to the data platform have supported all areas of the Welsh public sector to respond in concert to the pandemic. The project succeeded as it got over the interoperability hurdles by using APIs and cloud computing platforms to provide automation and flexibility of technology service provision.
It is still in heavy use today, and Bennée says it has provided a tool for the interface between science and politics.
“Politicians need to offer certainty,” she says. “Science is built on the acceptance of uncertainty, but with data we can get closer and closer to certainty.”