Overview of ICD-11 architecture and structure

The International Classification of Diseases (ICD) has progressed from a short list of causes of death to become the predominant classification of human diseases, syndromes, and conditions around the world. The World Health Organization has now explored how the ICD could be revised to leverage the advances in computer science, ontology, and knowledge representation that had accelerated in the twentieth and early twenty-first centuries.

Many teams of clinical specialists and domain leaders worked to fundamentally revise the science and knowledge base of ICD-11. Development of the ICD-11 architecturally was a fundamental revision. The architecture for ICD-11 proposed in 2007 included three layers: a semantic network of biomedical concepts (Foundation), a traditional tabulation of hierarchical codes that would derive from that network (Linearization), and a formal ontology that would anchor the meaning of terms in the semantic network. Additionally, each entry in the semantic network would have an associated information model of required and optional content (Content Model).

This paper describes the innovative architecture developed for ICD-11.

ICD11 is a revolutionary transformation of a century long medical classification that retains is historical rendering and interface while expanding the opportunity for multiple linearization and underpinning its content with a formally constructed semantic network. The new artifact can enable modern data science and analyses with content encoded with ICD11.