Rigor in electronic health record knowledge representation: Lessons learned from a SNOMED CT clinical content...

Area of Science:

• Clinical Informatics
• Health Data Standards
• Medical Terminology

Background:

• Rigor in clinical knowledge representation is crucial for electronic health record (EHR) data interoperability and reuse.
• Clinicians must understand clinical standards for knowledge representation within EHRs.

Purpose of the Study:

• Educate clinicians and students on knowledge representation principles.
• Evaluate the success of manual lookups for assigning Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) concept identifiers.
• Assess the use of formally mapped concepts from the Omaha System interface terminology.

Main Methods:

• Clinicians performed 21 manual lookups of Omaha System terms in SNOMED CT browsers.
• Success was defined as an exact match with the corresponding code from the January 2013 SNOMED CT-Omaha System cross-map.

Main Results:

• Manual lookups achieved a 57.1% success rate (12 out of 21 attempts).
• Identified errors included semantic gaps, differences in granularity, synonymy, and partial term matching.

Conclusions:

• Achieving rigorous clinical knowledge representation is a global challenge.
• Terminology cross-maps show potential for improving SNOMED CT encoding rigor.
• Further research is needed to assess outcomes of using cross-maps for SNOMED CT encoding.