More data than ever is now collected from citizens and deposited in digital repositories and in the NHS we have some of the biggest silos of data in the world with longitudinal data going back more than 20 years in both digital and non-digital records.
Many years ago, the NHS ran on spreadsheets and old technology: data was unwieldy and cumbersome to manage, and decisions were based on imperfect, out of date, or partial information.
Fortunately the digital evolution is well underway in the NHS and data flows more freely across the system and is much more widely available.
However, the important question is: Are we making better decisions on the basis of this information and are we confident enough to act on what we know?
It is very easy in this day and age to suffer from information overload. It’s harder to make sense of what is in front of us without the help of digital technology and algorithms to spot trends, patterns and, most importantly, skilled teams of analysts and data scientists to help visualise and interpret multiple streams of data.
The two main issues seem to be about access to data and using the data in a way that allows us to make critical decisions with enough confidence about the outcomes we can expect to achieve.
So what should we do to make this happen and avoid some common bear traps?
There are two key lessons here:
Know when to stop
Join up what you have got
1. ‘What if…?’
Harnessing these advanced analytical techniques to help make decisions about the design of complex future care models can be very powerful but they reveal a hidden complication.
Getting a useful answer to the ‘what if’ question is not as easy as it first seems even with the application of sophisticated predictive modelling techniques. Often, the answer to one question will uncover other questions. This is not a new phenomenon, however it can lead to the condition known as ‘analysis paralysis’, where the original objective of making a decision becomes lost in a flourishing ecosystem of ‘ what if’ questions and everyone starts to drown in a plethora of answers that can seem as confusing as the original question.
In order to take the decision, you need a multidisciplinary team that works in conjunction with your analysts. Your team should be led by an ‘Intelligence Partner’ who can combine an in-depth understanding of the analysis with the leadership skills to bring the knowledge of the multidisciplinary team and create a coherent, evidence-based narrative to make recommendations for system leaders.
Most importantly of all, an Intelligence Partner will know when there is no more value in asking ‘What if….’ and will be able bring the process to a conclusion.
Intelligence Partners are a rare breed and you need to invest in your workforce to help create them. Just as you prepare to have a digital-ready workforce, you should plan for an information-ready one as well.
2. ‘If only we could join up the data’
Determinants of health and wellbeing are well understood and it is clear that these extend well beyond health data into environmental and social factors.
The most exciting advances in artificial intelligence indicate that we will make better decisions if we can take more of these factors into account. However, so many times we hear that integrating the data needed to do this is impossible.
It’s not, but it will take time and a level of investment you should plan for. Here are three things you can do to make it happen:
Get your system-level Information Governance sorted before you do anything else. Agree an Integrated Care System-level Authority to set data sharing agreements and ideally make system-level decisions about data sharing. Make sure they can see the big picture and can evaluate your integrated data and interoperability programmes within that context. Create a group that includes more voices than just the Information Governance experts, and can make credible recommendations to the constituent organisations.
To get to the point of being able to have a population health analytics capability across a system, you need the right foundations in place. Key to that, you need an Information Strategy. This will help you ensure you can create a roadmap that will build the right foundations and focus on a phased approach with benefits delivered at each stage.
This should include core dimensions such as governance, people, data and technology. A common misconception is ‘if we have a Digital Strategy, we have an Information Strategy’ but the former can often only incorporate the technology elements of what is required and not key elements of workforce capability and data strategy.
A lot of products on the market promise some dazzling capabilities in providing population health analysis. This may be the case but they will not work without the correct data, strategic vision and skilled workforce.
In short, these things will not be fully realised quickly and you should be prepared to invest the time and resources to get these elements implemented correctly.
If you have been involved in delivering a major Digital Programme, you will be aware of what is required to make it happen. It should be no different with your integrated data platforms, regardless of whether you provide them in-house, outsource or have a blended model. If you are considering buying-in specialist platforms or tools, then you need to understand the investment you and your system will need to achieve the desired results.So, in summary, my advice is be pragmatic. Invest in an Information Strategy and use it to build the right foundations to take you forward. Invest in people to have the right skills and capabilities in your workforce to understand analytics and lead the decision-making process, using a multidisciplinary team approach to build consensus based on evidence.