The need for a data strategy
All healthcare organisations I speak to and work with want to get to predictive analytics (i.e. use advanced analytics such as ML, be proactive and do forecasting etc) or even prescriptive analytics (where you identify what you should do based on predictions).
However, many are still in descriptive analytics stage, looking back at things that have happened in the past (maybe weeks ago) and often presented in non-interactive methods such PowerPoints or PDFs.
In my experience, getting from descriptive analytics, through what we call the ‘Trusted insights’ phase where users develop confidence in the data and insights, and start engaging with them meaningfully, through towards predictive analytics and beyond is a journey.
And it’s not an easy journey. Which has to be done in stages, built on strong foundations around data architecture and governance.
No matter where you are in the journey, having a robust data strategy helps take stock of where you are, everything that’s going on nationally and regionally that is relevant, and what you need to do to get to your vision.
Developing a good data strategy is not easy
Writing a data strategy is easy. Writing a good data strategy, that gains buy in from Executives, operational folk and clinicians, and helps move you along the data transformation journey is not that easy.
Having worked on data strategies in different settings and organisations, I find there are 3 key pillars of a successful data strategy:
- Engagement – the human element is key. Although you can probably write most of the content sat in a dark room, you shouldn’t. You need to talk to the key stakeholder groups, understand their problems. Get their ideas and co-develop the data strategy with them.
- Clear technology blueprint and roadmap – there is an increasing amount of national initiatives, new technologies and tools. There needs to be clear blueprint on how all of this works together and comes together to deliver on your vision. You also need a clear roadmap on how to get to the vision
- Data culture and literacy – the most clearly structured and thought plan doesn’t help by itself, the rubber really hits the road with the people – if stakeholders don’t believe in the change needed and value it will deliver it’s not going to go anywhere. They need to be empowered to be able to engage with the tools. And the data and analytics teams need to have a different mindset and the right skills to support the transformation in a new NHS data landscape and collaborative working environment (both technical skills and what I call core skills such as curiosity, communication, problem solving and storytelling with data.
How to develop an effective data strategy
The methodology we use for developing a data strategy has 8 key elements which are aligned to the 3 pillars above.
For example, within the engagement pillar, we start off by fully understanding the strategic objectives of the organisation and engaging with the key user groups to understand challenges and ambitions for the future.
Within the ‘Clear technology blueprint and roadmap’ pillar, we explore the policies and controls to provide governance and accountability needed, and how to then collect, store and share safety. All of this needs to underpinned by robust and modern data infrastructures including data platforms, which in turn sets the scene for meaningful insights that can really drive decision making.
Within the ‘Data culture and literacy’ pillar, we need to understand what types of roles and skills are required to deliver the strategy, and a prioritised roadmap on what need to be done and how it should be communicated.
The real impact
The implementation of the data strategy is when it becomes real.
Our prioritised roadmaps identify initiatives across three stages – Foundational, then Transformational and Leading-edge.
On the Foundational stage, we work with client to get the basics right – for example, data governance, data quality and modern data platforms. This sets the stage for more ambitious and impactful initiatives. On the Leading-edge stage, we are working with clients to help them use data for population health and predictive analytics. Examples include solutions that identify patients at risk of deterioration or that helps predict future service demand and to model capacity scenarios.
Reflections from HETT 2023
This year as sponsors for HETT, I presented on: ‘Data-driven transformation’ with clients from Essex Partnership University NHS Trust and Royal Free London NHS Foundation Trust, where we focussed on the components that build a successful data strategy.
Our clients were able to bring the following elements to life through their personal and real-world experiences:
- What their organisation was looking to get out of their data strategy
- What benefits they gained from engaging a wide group of stakeholders during development of their data strategy
- What’s helping them to implement their data strategy
- How the data strategy implementation is helping them with their real operational issues
During the audience Q&A we were asked: ‘Do we really need a data strategy?’. The answer was ‘Yes’. Why? Because a well constructed and co-developed data strategy gives a clear vision of the future and what needs to be done when to get there from where you are currently. The technology, processes and healthcare demand is always changing – therefore the staretgy needs to be ‘live’ and continually evolve. In practice, you implement a strategy and have regular evaluation milestones, at which point the implementation roadmap may need to be adjusted.