Aligning an interface terminology to the Logical Observation Identifiers Names and Codes (LOINC^®).

Area of Science:

• Medical Informatics
• Health Information Systems
• Clinical Terminology Standards

Background:

• Interface terminologies in healthcare, such as TLAB, often lack standardization.
• Integrating diverse health information systems requires common data standards.
• Logical Observation Identifiers Names and Codes (LOINC) is a crucial standard for biological data exchange.

Purpose of the Study:

• To align the Bordeaux University Hospital interface terminology (TLAB) with LOINC.
• To enhance the integrated use of biological results across health information systems.
• To evaluate the feasibility and performance of mapping TLAB to LOINC.

Main Methods:

• Developed an innovative approach to decompose and recompose LOINC concepts.
• Anchored TLAB entities to LOINC attributes and aligned them to LOINC concepts.
• Applied an instance-based filtering process using laboratory results from a hospital data-warehouse.

Main Results:

• Achieved high alignment rates: 99.8% of TLAB entities related to LOINC analytes and 61.0% to LOINC systems.
• Successfully mapped 55.4% of TLAB entities used in the hospital data-warehouse to LOINC concepts.
• Manual evaluation of 1-1 mappings yielded a precision of 0.59, considering TLAB label quality.

Conclusions:

• Alignment of TLAB and LOINC demonstrated reasonable performance despite TLAB label limitations.
• Interoperability can be improved by enhancing the formality of LOINC structure.
• Querying LOINC attributes, not just concepts, offers potential for better data integration.