AWS Public Sector Blog

One small team created a cloud-based predictive modeling solution to improve healthcare services in the UK

How do you predict and prepare for your citizens’ health and wellness needs during the COVID-19 pandemic? Healthier Lancashire and South Cumbria Integrated Care System (ICS) quickly scaled a platform on Amazon Web Services (AWS) to support the 1.8 million people in their region with Nexus Intelligence, an interactive health intelligence application that brings together data to identify and map cohorts among a region’s population and display various health measures in a dynamic way.

Nexus Intelligence contains a suite of predictive models against various measures of health needs and outcomes, which helps general practitioners identify demographics with both clinical and social vulnerabilities. Nexus Intelligence’s applications like Outbreak Mapping, Population Health Intelligence, Virtual Ward Console, and more not only supported the ICS response to the pandemic, but are expected to provide integral insights for ICS to reconfigure and re-invest in services to improve the health and well-being of the population and reduce health inequalities.

Quickly building the tools to help predict citizens’ health needs

The Digital Intelligence Unit (DIU) is part of two Fylde coast National Health Service (NHS) clinical commissioning groups (CCGs) that make up the Fylde coast Integrated Care Partnership (ICP) along with its acute hospital, mental health, ambulance trusts, local authorities, and voluntary sector in the North West of England. They are responsible for primary, community, and secondary care for their region, and form part of the wider Healthier Lancashire and South Cumbria ICS, responsible for delivering care to its population of 1.8 million.

The ICSs are new partnerships for UK Healthcare across all care providers in a region. These partnerships between organisations that meet health and care needs coordinate services and plan ways to improve population health and reduce inequalities between different social demographics in a defined region. To understand these needs, systems must share information by pulling together data from multiple organisations, in as near real time where possible, with appropriate access rights to this data based on the needs of the service. This is where Nexus Intelligence comes in.

Nexus Intelligence started as a new development within the Fylde coast ICP. It was created with the vision to develop an analytical and data science working environment that supports the generation, execution, modelling, manipulation, and visualisation of data through interactive application delivery. After starting out on-premise, the DIU decided to move the development of Nexus Intelligence, which utilised an open-source framework, into the AWS Cloud. This platform, along with its suite of applications, allowed the DIU to scale at pace and deliver several supporting applications in responding to the pandemic, including Outbreak Mapping, COVID-19 Population Health Intelligence, Virtual Ward Console to manage oximetry at home services, and support for LAMP90 testing of staff across the ICS.

Figure 1. The Nexus Intelligence interface showcasing the “Cohort selector” feature. Users can map specific cohorts of residents refined by a variety of health measures and risks. View a full size version of the image here.

These applications help public health teams and local authorities use outbreak mapping tools to geographically monitor the spread, with the ability to run outbreak detection models in as near real time as data is made available. General practitioners and their supporting multi-disciplinary teams use the interactive population health intelligence tools to identify demographics with both clinical and social vulnerabilities. Commissioned primary care services access the tools to manage their virtual wards, enabling them to quickly identify COVID-19 positive patients for clinical review and electronic referral onto the service. These applications also support care services, performing and recording weekly COVID-19 tests among clinical staff to help stop the spread of the virus.

To create the applications, the Fylde coast’s CCGs DIU moved away from a monolithic architecture to a microservice one, containerising applications to support scaling up when necessary. The team used Amazon Elastic Container Service (Amazon ECS) for containers, with 46 applications and microservices stored in AWS CodeCommit repositories with continuous improvement (CI)/continuous development (CD) pipelines. They also utilized AWS CodePipeline, AWS CodeBuild, and AWS Lambda, as well as Amazon DynamoDB CI/CD pipelines in Amazon Relational Database Service (Amazon RDS) for PostgreSQL code pipeline and code build.

Figure 2. The Nexus Intelligence architecture featuring the AWS services utilized to create the solution. View a full size version of the image here.

Amazon RDS PostgreSQL cluster is now housing almost 2 million records, including over 230,000 COVID-19 positive test results from Public Health England for the entire ICS population. Amazon DynamoDB is hosting over 680,000 COVID-19 pathology results from ICS Hospital Labs, supporting around 25,000 request per week from clinicians using the ICS Lancashire Patient Record Exchange Service. In addition to this, there are weekly LAMP90 staff testing/screening results for currently over 19,000 NHS staff, reporting over 120,000 test results to date.

This data is all secured using industry standards and following AWS best practices, and the Nexus Intelligence team periodically carries out audits and reviews using the AWS Well-Architected Framework to help the team identify any potential areas of concern, manage risks, and build a more resilient solution. This rigorous security allowed the team to adopt NHS Digital’s Internet First policy and securely deliver applications to their end users across the ICS partner organisations.

A cloud novice team built a solution in less time than using off the shelf solutions

The Fylde coast’s CCGs DIU was able to build Nexus Intelligence in the cloud with a staff of just four people. This small team, all new to the cloud, were able to use the cloud to launch innovating open-source development, enabling the delivery of several supporting applications throughout the pandemic. Their continued success in delivery has allowed the team to grow.

Working with AWS solutions architects, the DIU team found that in-house cloud development was possible and achievable even with limited resources. “Over the last 18 months we have realised the benefits of the cloud: it has enabled scaling without total redesign, quickly increasing infrastructure capacity to meet increased user demand. This along with our decision to use open-source in-house development to deliver our vision, has enabled us to react quickly in meeting the needs of the ICS throughout the pandemic,” said Stewart Bond, associate director at DIU. “We anticipate what we have achieved and continue to develop being paramount in establishing effective action for longer term impact, improving population health and reducing inequalities.”

Ben Alcock, data scientist at DIU, said, “Using AWS has enabled us to work alongside end users from across the system to develop the tools to determine the health needs of the population. This has given decision-makers the ability to run large-scale customisable machine learning models from any location, providing actionable insight to the necessary people, allowing for a deeper exploration of the health system.”

Using managed services available through the cloud also helps the team spend less time on patching and more time on innovation. “Developing systems and functionality utilizing AWS has led to a sustained period of innovation for our small, agile team. The shortened development life cycle and baked-in best practice means that we can bring insights to users of the platform at pace, without sacrificing on quality,” said Stewart Morgan, senior developer at DIU. “The stability of the services AWS provides makes sure that we spend all of our time focused on delivering improvements and new applications, instead of fixing or monitoring issues.”

What’s next?

NHS England have also shown an interest in Nexus Intelligence, and keen to explore its scope and potential for how it could be part of a collaborative toolset at a national level.

Learn more about the new opportunities for technology to revolutionize healthcare services in The Future of Healthcare Roundtable Series. In this half-day event, hear from healthcare, government, and technology leaders and experts, and discover how your organisation can meet pandemic-related challenges through cloud-based digital transformation. Register now.

Subscribe to the AWS Public Sector Blog newsletter to get the latest in AWS tools, solutions, and innovations from the public sector delivered to your inbox, or contact us.

Gemma McCarthy

Gemma McCarthy

Gemma McCarthy is a healthcare client manager in the UK public sector organisation within Amazon Web Services (AWS). She has worked with the UK National Health Service (NHS) for over five years, advising customers on their information technology decisions to improve patient outcomes, accelerate their digital transformation, and meet critical care objectives.

Stewart Bond

Stewart Bond

Stewart Bond is associate director for Digital Intelligence & Data Science at Fylde coast clinical commissioning groups and created the Digital Intelligence Unit there. With 25 years of working within the National Health Service (NHS) starting out in finance, he quickly moved into technology with a passion for innovation. In later years, he turned his attention to include intelligence, where he has been behind the creation and delivery of very successful intelligence platforms, supported by the skilled and dedicated teams he has the pleasure of leading. Nexus Intelligence is the latest platform to be delivered, created to use data to transform the way healthcare organizations prioritize and deliver care services to citizens. Nexus Intelligence is dedicated to delivering a sustainable interactive analytics platform for the Integrated Care System.