The Decentralized Web Podcast: 005 Eric Prud’hommeaux, W3C

In episode 5 of The Decentralized Web Podcast, host Justin Bingham is joined by Eric Prud’hommeaux, Data Sanitation Engineer at W3C and Data Architect at Janeiro Digital. W3C is the main international standards organisation for the world wide web. During the podcast, Justin and Eric discuss the evolution of FHIR and the decentralized web, and the future of interoperability in health systems.

The early advent of RDF

At W3C, under Tim Berners-Lee, an early vision for RDF and linked data saw information being decentralized to be controlled by the people. Today RDF and linked data is interfaced with constantly and drives the quality of information return. People write their own structured data to be harvested, bringing value to search results. The ability to combine two RDF documents allows data to be interoperable between disparate applications, enabling system descriptions and machine interpretations resulting in useful outputs.

Healthcare

The publication of the draft NHS policy Data Saves Lives aims to reshape health and social care using data and technology to promote patient-centred care. For the first time, the NHS wants patients to control their own data by creating interoperability across disparate legacy systems. This approach is at the heart of decentralization and uses technology and standards pioneered by Prud’hommeaux.

Digital solutions in healthcare need to capture vast sets of relationships and deploy data in structures that people can read. W3C used their healthcare and life sciences team to model and develop an RDF query language that led to the development of SPARQL. Later the RDF representation of FHIR – an interoperable standard for healthcare data – was developed. Most healthcare systems have FHIR compatibility or support and it’s essential to the future of healthcare interoperability.

The evolution of healthcare data technology

Prud’hommeaux started out with the RDF-ification of CCDA (Consolidated Clinical Data Architecture) then progressed to working with Lloyd McKenzie at HL7 developing the OWL representation of the underlying data model for CCDA. The RIM (Reference Information Model) captured the semantics of the clinical data but wasn’t popular with users. A second version wasn’t semantic and had better uptake; version 3 again wasn’t popular, being difficult for developers to use. Finally came the FHIR project that simplified the capture of data, moved away from semantics, and was engaging for the user. It’s now possible to have decentralized data that’s mobile, understood by different machines and can be validated so it’s interoperable. Having standards like FHIR has been key. The work in healthcare led to a need for a serious validation language and so ShEx was developed to describe and validate the FHIR RDF representation.

The future

We want to learn from caring for a patient, to enable us to also care for others. To do this effectively, inputted patient data needs to include the rich context surrounding each interaction. This will directly benefit the patient as the improved information is shared across services and between healthcare professionals.

To find out more about linked data and decentralization in healthcare, listen to the full podcast episode.

Guest: Eric Prud’hommeaux | LinkedIn