Data saves lives: building and up skilling the NHS analytics community
Sarah Culkin (Interim Head of the Analytics Unit at NHSX) and Sukhmeet Panesar (Deputy Director within NHS England and NHS Improvement’s Data, Analysis and Intelligence Service) are champions of data analytics across the health and care system. Here they set out what their teams are here to do, what they have learnt from the pandemic so far and their immediate areas of focus moving forward.
Knowledge is power. In healthcare, it is often life saving. The NHS generates a huge amount of data which can be analysed and used to drive improvements in care and how services are run as the Health Foundation reflects in its report “Untapped potential: Investing in health and care data analytics”. Clinicians can use the insights generated by skilled analysts to improve diagnosis and disease management. National and local NHS leaders can evaluate innovations and new models of care to find out if expected changes and benefits were realised. Local NHS leaders can make complex decisions about allocating limited resources and setting priorities for care. Patients, service users and the public will be able to better use and understand health care data.
Ultimately, data analysis results in improved patient outcomes and experience, as well as optimal use of NHS resources. This has never been more evident than during the pandemic where the NHS COVID-19 Data Store has enabled better decision making at a national, regional and local level. The use of a single integrated data, analysis and modelling platform (Foundry) has enabled NHS analysts to develop various tools to support local systems to respond to the pandemic including short-term forecasts, supply management capability (for critical equipment such as oxygen, generators and PPE), an integrated planning tool and an early warning system to support regional and local teams to anticipate pressures and make best use of resources.
And yet, in general, the NHS is failing to make the most of its data because there are not enough people with the right analytical skills to make sense of the information being collected.
The NHSX Analytics Unit works in partnership with NHS England and NHS Improvement’s Data, Analysis and Intelligence Service, to provide leadership to analysts working in the system and raise data analysis up the health and care system agenda.
Building a community in the fight against COVID-19
We know that analysts work best in communities where they can learn from each other. But at the moment individuals are often isolated within teams and don’t always have access to the right tools and support.
COVID-19 is most probably the single biggest challenge the NHS has faced since its inception. In response to the challenge, we established one of the largest online communities of data professionals and analysts (14,000) to come together to share tools and methods, encourage collaboration and reduce duplication beyond traditional organisational and geographic boundaries. It has since grown with organisations such as HDR UK, techUK, AI Tech North UK, Interworks, Tableau, Tech4CV19, AphA, Logan and Tod joining and bringing their expertise to our community-driven movement, which is now leading data-driven, evidence-based decision-making across health and social care.
The challenges that lie ahead
Access to state of the art tools
Communities need to share languages and common tools to work and grow together. R and Python are both free, open source, state of the art programming languages. Unlike HTML and JavaScript, they can be used for other types of programming and software development besides web development. They’re considered cutting edge in their field and are free to use. Because of this they are used in data science and analysis environments by large parts of academia and industry, including Facebook, Google, Spotify and Netflix.
Yet analysts in health and care organisations often find getting access to analytical tools such as R and Python a challenge, even when they’re open and free to use. Why? Whilst on our home devices we are free to install what we like, with good reason, admin privileges required to install these tools are restricted in most of our organisations. Instead, IT teams decide appropriate tools on behalf of their organisations and act as gatekeepers to installation. A lack of understanding around open source tools compared to software that is purchased results in many analysts missing out.
At all levels of the health and care system, open access to analytical tools and methods allows the system to extract more value from data, move beyond the more repetitive manual tasks and support automation. We need to create a culture of ‘build it once, share the methodology and learn with others’. Our teams aim to do that, hopefully prompting others to the same, inside as well as outside of health and care.
Recruitment, training and support
Recruitment, training and career opportunities are other key issues. This article in the Journal of the Royal Society of Medicine highlights how analysts are often recruited into very junior roles, given little to no guidance on the skills they need in order to progress and have few inspiring leaders to look up to. Internal training and support networks help reduce barriers of entry into the R and Python community. Colleagues can gain a great deal from linking to professional support networks such as the NHS-R community and the Association of Professional Healthcare Analysts (AphA) or accessing other suitable training opportunities and online online resources. We have embarked upon understanding and building the learning and development landscape for data professionals, analysts and indeed consumers of the outputs through the members of the Developing Data and Analysis-as-a-profession Board.
Staff need to be given protected time to learn whilst using these tools, which includes what we would term ‘time to fail whilst learning’. This is good practice, as we all know we can learn a great deal more from our mistakes than our successes.
The support of senior leaders is also integral in promoting and rewarding the uptake and continued use of these tools which bring cutting edge data science into their organisations.
Open and reusable code
As set out in the NHS Digital Service Manual, public services are built with public money. So unless there's a good reason not to, the code they're based on should be made available for other people to reuse and build on. We should make all new source code open and reusable and publish it under appropriate licences (such as MIT, OGLv3 or GPLv3, alongside suitable open datasets or dummy data). Open sharing of technical skills and domain knowledge through sites like Cross Validated and StackOverFlow, and sharing code and methodology through platforms like GitHub, will build high quality analytics throughout the system.
How will we help?
We are here to lead by example, as well as to push the case across the system to modernise and strengthen the use of data and analytics. In the short term, we are busy focussing on three areas:
Making sure that NHSX and NHS England and NHS Improvement really are leading by example, ensuring business intelligence and analysis runs throughout the policy lifecycle and that we demonstrate working in an interoperable and transparent way.
Developing and spreading innovation of new techniques through collaboration and uptake of new data science tools.
Building a stronger analytical workforce, with access to the right tools and support including:
Agreed frameworks, guidelines and policies to operate within
Communities of practice, be they local, regional, national or virtual, where innovation, sharing and working in the open is the norm
Learning and development opportunities
To find out more, visit the Analytics Unit pages of the NHSX website, join our online community ‘AnalystX’ on FutureNHS and if you want to ask us a question or have a great idea, email us at analytics-unit@nhsx.nhs.uk.