The 3 C's of health data

...coherent, coordinated & connected...

In 2006, Tim Berners-Lee, inventor of the World Wide Web said, “Data is a precious thing and will last longer than the systems themselves.” Twelve years later, we collect mountains of health data in our clinical systems – in an infinite variety of formats, from a myriad of locations and clinical contexts, and with inconsistent meaning.

Every new health system - every project or clinical application - establishes yet another data silo, constructed in such splendid isolation so that it has no correlation or relationship to any other. Only later do we collectively scratch our heads and puzzle over how to connect them; to make the data more available and usable.

Safe, incremental innovations in messages and documents have demonstrated some progress in the interoperability of small parts of a health record, but will not scale to a whole health record nor be sustainable as message numbers grow.

Semantic interoperability - "the ability of two or more systems or elements to exchange information and to use the information that has been exchanged" - remains our nemesis despite three decades of experience in collecting digital health information. Eternally optimists, we expect semantic interoperability to miraculously materialise.

This is not rocket science, people! We have no strategy for convergence, yet the intrinsic value of health data will only be realised if it is coherent, coordinated and connected.

It’s time for a common-sense approach that has potentially radical implications. We need to focus on the atomic data, because it is the data itself that should be the health record, not the clinical system.

The data is the health record

The radical innovation here is that the spotlight is focused on the data as basis for a lifelong health record. The role of any clinical software is reduced to a technical tool, effectively a specialised system that provides to access the health record data, and which can be substituted at any time if an improved tool becomes available. With this approach we ensure data integrity, potentially from cradle to grave, without the risks associated with data migrations and transformations.

As a natural consequence of that pivot in mindset, data silos are banished and semantic interoperability becomes orders of magnitudes easier.

In order for this to succeed, the data needs to be:

  • open;
  • standardised;
  • strategically designed and coordinated to be coherent across the breadth of healthcare; 
  • clinically agreed as 'fit for purpose’;
  • freely available;
  • independent of any single vendor, project or clinical application; and
  • a foundation for working more intelligently with health data in ways not currently possible.

It is already happening...

There is an international community actively working toward this. Mostly based in Europe, a community of clinicians, vendors, health informaticians and organisations have been working collaboratively to describe the data that clinicians need, in a way that technicians can implement. The approach is based on an independent, open standard known as openEHR, open Electronic Health Records. Data created based on the shared library of openEHR data patterns, known as archetypes, can be exchanged and reused by others who also use the same archetypes as the basis for their data. It means that systems whose business model previously relied on the ‘specialness’ of their unique database structure and ‘locked in’ data, can now connect and share data seamlessly with other openEHR-enabled systems. Vendors now compete on value-add ‘smarts’ such as enhanced functionality, user interfaces and workflow.

Momentum is building, along with increasing evidence about successful implementations.

Our current fragmented approach to health data design severely limits the potential impact offered by digital health. There is no reason to delay further...

The future is already here — it’s just not very evenly distributed.
— William Gibson