Related Experiment Videos
Culling a clinical terminology: a systematic approach to identifying problematic content
Proceedings. AMIA Symposium
|February 5, 2002
Summary
The integration of SNOMED RT and NHS Clinical Terms into SNOMED CT identified problematic concept descriptions requiring review. This systematic approach categorizes fourteen types of terminology issues for improved clinical data quality.
Area of Science:
- Medical Informatics
- Clinical Terminology Management
- Health Data Standards
Background:
- The College of American Pathologists and the UK's National Health Service (NHS) are merging SNOMED RT and Clinical Terms Version 3.
- This initiative aims to create a unified clinical terminology: SNOMED Clinical Terms (SNOMED CT).
- Mapping between the existing terminologies is a crucial step in this merger process.
Purpose of the Study:
- To present data on problematic concept descriptions identified during the early development phases of SNOMED CT.
- To categorize and describe types of terminology content requiring further review.
- To establish a systematic method for identifying and addressing problematic terminology.
Main Methods:
- Analysis of concept descriptions during the mapping process between SNOMED RT and Clinical Terms Version 3.
- Identification of descriptions flagged for further review by the SNOMED editorial process.
- Categorization of identified problematic descriptions into distinct types.
Main Results:
- Numerous concept descriptions were identified as potentially problematic during the SNOMED CT mapping.
- Fourteen distinct types of problematic terminology content were described based on this analysis.
- These problematic descriptions require review for incorporation into SNOMED CT or retirement.
Conclusions:
- A systematic approach to identifying problematic terminology content is feasible and necessary.
- Addressing these issues is vital for the successful development and integrity of SNOMED CT.
- The findings support ongoing refinement of clinical terminologies for improved healthcare data.