NHS Data – do we need a map or atlas?

Die ganze Welt in einem Kleberblat – Heinrich Bünting 1581

Building on their recent report on preparing the NHS for the “AI era“, the Tony Blair Institute have announced an effort to map the the national data architecture of the NHS. This builds on the work published last year in The Lancet Digital Health which described national data flows using publicly available data.

Personally, I love Ursula Franklin’s wonderful description that a good map is

“…meant to be useful, to assist the traveller and bridge the gap between the known and the as yet unknown”.

If this work from the TBI shows us better ways to use information to improve outcomes, reduce the burdens on staff and make our services more sustainable it will be a triumph. But still…

• Maps are not neutral. Physical maps can reflect political and social issues more then they describe topography. The mappae mundi pictured above better presented religious concepts than geographical realities. Arguments between the Gall-Peters and Mercator projections even made it into an episode of The West Wing. Maps have implicit meanings that we need to look for and consider.
• Maps trigger cognitive biases. What is the correct level of abstraction that makes a data map accessible, prevents readers gaining a false sense of expertise and allows the map to be created and maintained? Is there one?
• Maps are ephemeral. The NHS doesn’t stand still. Systems, pathways, process and people are ever changing. Describing connections that exist is not the same as understanding or being able to influence / govern them. How should a data map be designed to tell us how to better cultivate what has grown, not just where to plant more systems (Margunn Aanestad sets this out much more eloquently!)

So then what does an atlas offer that a map does not? Firstly, an atlas is less bound by a singular scale and can show one region in granular detail with another in a much more zoomed out manner – not all data and data flows are created equal and understanding some in greater detail is sensible. Secondly, an atlas can layer in details making direct linkages between disconnected information. This adds richness, context and meaning. Creating and reading in this way allows us to look at information and “re-edit and piece together again without thinking we are summarising or exhausting it”. This is what drives new insights, perspectives and ideas.

This all builds on arguments better made made by Kate Crawford that for complex, critical issues, we need creative ways to better understand the social, political environmental and technological aspects that drive them. Healthcare and data is possibly as complex as it gets in all of these domains. This data map might be great – an atlas could be better.