How we made the data visual at Southampton University Hospital
There is no shortage of data in modern healthcare, but the means to quickly and easily gather it and then understand it is often lacking. This is not the case at University Hospital in Southampton, where a small team adapted a 1990s data visualization concept from American academics into a working system they describe as transformative. Project manager and consultant surgeon David Rew explains more.
In the late 2000s, we realized that the NHS ‘impending digital transformation agenda would create major challenges in integrating, visualizing and understanding information as paper records were taken away from our trust. We sought a new format for visualizing clinical data that would make the most of our records, which had been formally structured and standardized since 1995.
We learned about the conceptual work of Professor Ben Shneiderman and his colleagues at the Human Computer Interaction Lab at the University of Maryland. The power of their framework model for an interactive health record, codenamed Lifelines, which we also adopted, had not been recognized or taken up at that time.
Agile and iterative
I proposed Lifelines as a concept to explore with the IT team at Southampton University Hospital (UHS) and UHS IT systems consultant Alan Hales. Alan and I then worked in a very agile and iterative fashion to adapt Lifelines’ original concept from a curiosity-based discovery to a practical system.
The evolving product, UHS Lifelines, is now available to more than 5,000 users in our electronic patient record (EHR) and for 2.5 million individual health records. It is subject to continuous development as new data sources are incorporated.
UHS Lifelines is based on a stack of synchronized timelines, on which documents, results and reports are plotted as interactive icons, based on their historical subject, their timestamps (also known as metadata) on a dynamic two-dimensional chart. The X axis displays time from left to right, and the vertical axis can display any number of data timelines for which a subject taxonomy exists.
It is a conceptually very simple but operationally radical approach to real-time visualization and interrogation of a complex clinical record. It allows large amounts of clinical information to be displayed on a single computer screen. It allows users to harness the high-speed powers of the visual brain to “read the story” from the mass of information contained in iconographic models, and to select and access key documents in seconds.
UHS Lifelines also adheres to Professor Shneiderman’s mantra for visualizing large data sets. It gives an overview of the entire documentary file; and it allows the user to zoom in on key documents and filter out unnecessary screen clutter, while directly accessing details on demand. It enables faster, better and safer clinical decisions with less user fatigue.
Intuitive and fast
The methodology is very intuitive in practical use with minimal training. It is very fast compared to traditional electronic document search strategies using tabs, multiple windows and lists. It cuts the time spent in ambulatory clinics searching for digital record sets from minutes to seconds, while dramatically reducing the risk of missing critical information and patterns in the record set.
UHS Lifelines also brings clinical immediacy to historical electronic documents of all ages. For example, this prompted the restoration and uploading into the system of 500,000 UHS histopathology reports dating back to 1990.
We believe the UHS Lifelines example provides an elegant framework for a standard, safe and controllable EPR for all UK citizens. It can integrate and display all key data from primary, secondary, social, community and mental health care in one interactive framework.
Control of access to information
The interposition of icons between the viewer and the source information also makes it possible to control access to sensitive information, while recognizing their existence. With it, first responders could have controlled and selective access to critical personal data in an emergency.
The UHS Lifelines model could enable citizens to access their own comprehensive health record in a controlled and structured manner, whether from mobile or desktop apps. Iconography also provides a simple and understandable framework for the application of machine analysis to predict health patterns, trends and future NHS resource requirements.
UHS Lifelines was built at a negligible cost for UHS or for the NHS. It illustrates the power of user-centric digital design and local agile principles in the NHS to address the complex global challenges of healthcare data management.
David Rew is a Consulting General Surgeon at University Hospital Southampton and Concept Manager for the UHS Lifelines Project as part of the Trust’s Global Digital Example Program. He is a member of the Engineering and Physical Sciences Research Council (EPSRC) Strategic Advisory Team for Health Technologies and Visiting Professor of Information Technology at the University of Southampton Solent, whose students have contributed significantly to the development of the UHS Lifelines system.
He and his colleagues welcome inquiries from anyone interested in learning more about the job.