1. Background

Welcome to SNOMED CT for Data Analysts! We hope you enjoy following this learning and learning about how to leverage the capabilities of SNOMED CT to achieve your clinical data analytics goals!

Why is SNOMED CT Useful for Clinical Data Analytics?

SNOMED CT is fundamentally designed around the meaning of clinical ideas, which we call ‘concepts’. This means that clinical data can be queried and analyzed based on its clinical meaning and the relationships between these meanings. Furthermore, because SNOMED CT provides concepts at different levels of detail, clinical data can be recorded at a granular level of detail (if required), and then aggregated or queried at various higher levels of clinical abstraction.

SNOMED CT also provides multilingual synonyms to refer to these clinical concepts. The concept centric approach ensures that equivalent queries (which use terms from different languages) all refer to the same clinical meanings.

Another important feature of SNOMED CT, which distinguishes it from statistical classifications (such as ICD) is that it uses a “polyhierarchy”. This means that each code can be grouped under more than one category, based on different aspects of its meaning. This enables SNOMED codes to be queried from different perspectives and aggregated in different ways.

One of the most significant reasons why SNOMED CT is so powerful for data analysis is that it uses formal Description Logic as the foundation for its clinical concept definitions. This means that we can leverage formal machine reasoning techniques to enable new relationships to be inferred from the clinical data.

Other reasons why SNOMED CT is useful for analyzing clinical data include:

  • SNOMED CT is the most comprehensive clinical terminology in the world, and as such it provides a broad clinical coverage across a range of clinical disciplines, specialties and domains
  • SNOMED CT is regularly updated which allows new clinical knowledge to be included and continual improvements to be made
  • SNOMED CT provides robust versioning and retains a full history of all changes, to help manage queries over longitudinal health records that may span many years or decades
  • SNOMED CT is an international terminology, so queries, subsets, rules and maps can be shared and reused between countries
  • SNOMED CT has localization mechanisms that allow queries to be applied to data from different countries, dialects, regions and applications

What Types of Clinical Data Analytics does SNOMED CT Support?

SNOMED CT is designed to support the accurate and precise capture of clinical information at the point of care, and the subsequent meaning-based retrieval of this data. This allows SNOMED CT to support a wide range of analytics tasks, including: 

  • Point of care reporting
  • Patient summaries
  • Clinical decision support
  • Trend analysis
  • Pharmacovigilance
  • Clinical service and quality audits
  • Predictive medicine
  • Clinical research
  • Semantic search