Healthcare Data Standardisation and Interoperability

Healthcare data standards support the implementation of NHS and international standards, to achieve the Trust’s vision of delivering top quality patient care, excellent education and world class research.

With more and more integrated health care delivery needs, data can no longer be restricted to one system or one organisation. Systems must be connected to ensure that clinicians have immediate access to relevant and appropriate patient data from care providers and settings. Clinical Data Standards are one of the building blocks of the digital architecture for healthcare. The Clinical Data Standards initiative provides a foundation that will enable data aggregation, cross study analysis and more importantly, direct patient care.

The generalised flow of patient information within the NHS and its relationship across the different standards are shown in the diagram below. The pyramid demonstrates how the accurate representation and recording of patient information, within any Electronic Health Record (EHR) system, equally serves as the determinant for non-clinical activities.


Description automatically generated

We apply various national and international standards within Epic.

Some key standards are:

  • Terminology standards: SNOMED CT, LOINC, dm+d
  • Classification standards: ICD-10, OPCS-4
  • Interoperability standard: FHIR
  • Other relevant standards: UCUM (for Units of Measure)

UCLH is the first Epic site in the world to implement native SNOMED-CT recording for diagnosis recording in Epic. All future Epic sites in the UK will follow suit, and this will affect up to 20 million people by 2025. To make this process easier, Epic has developed a specific utility that is now part of their core software system - we contributed heavily to the development of this utility.

Our presentation in the SNOMED-CT International Expo 2021 was in the top 5 most-attended sessions across 100s of sessions.

We are a major player for the national interoperability projects:

  • HIE (Health Information Exchanges)
  • OneLondon

Two national Natural Language Processing (NLP) projects (MiADE and Cogstack) use SNOMED CT as the underlying data standard.

We supported NHS England on the UK data standards strategy. Some completed work includes:

  • Designed National Congenital Anomaly and Rare Disease (NCARD) Registry Service Diagnosis refset in SNOMED CT
  • Designed National Surgical Devices and Implants procedure refset for the treatment of stress urinary incontinence (SUI) and pelvic organ prolapse (POP)
  • Shared Unit of Measure standardisation (to UCUM) local implementation experience
  • Designed national Neurology Outpatient Diagnoses refset in SNOMED CT [for national Getting It Right First Time (GIRFT) program]

We led the national call for the use of the OMOP-CDM for data sharing. The successful examples are:

  • 4 new Health Informatics Collaborative (HIC) themes led by UCLH to use OMOP-CDM, e.g. Multiple Myeloma, Hearing Health, Critical Care, Transfusion dependent anaemia
  • Partnered with IgniteDATA and AstraZeneca as the first in Europe to use FHIR standard to demonstrate EHR2EDC (Electronic Health Record to Electronic Data Capture) technology for oncology trial. More clinical trials with other pharmaceutical companies to follow.
  • Too many standards! Standards should be created jointly but they were not. They differ between industries and countries. The standards sometimes are not even the same depending on which EHR systems the organisations are using. 
  • Selling the benefits. For majority of the clinicians and managers, data standardisation is still a novel concept. What clinical or operational benefits has it led to? For system vendors, does this mean more sales?
  • Build a culture of data re-use amongst UCLH clinicians and Epic analysts, where there is an expectation that data generated will be shared and used outside UCLH.