Semantic enrichment of Pomeranian health study data using LOINC and WHO-FIC terminology mapping principles

Objective To semantically enrich the laboratory data dictionary of the Study of Health in Pomerania (SHIP), a population-based cohort study, with LOINC to achieve better compliance with the FAIR principles for data stewardship. Materials and Methods We employed a workflow that maps codes from the SH...

Full description

Saved in:
Bibliographic Details
Published in:JAMIA open Vol. 8; no. 2; p. ooaf010
Main Authors: Inau, Esther Thea, Radke, Dörte, Bird, Linda, Westphal, Susanne, Ittermann, Till, Schäfer, Christian, Nauck, Matthias, Zeleke, Atinkut Alamirrew, Schmidt, Carsten Oliver, Waltemath, Dagmar
Format: Journal Article
Language:English
Published: United States Oxford University Press 01.04.2025
Subjects:
Computational linguistics
Data dictionaries
Laboratories
Language processing
Natural language interfaces
Research and Applications
Semantics
Terms and phrases
South Korea
Germany
LOINC
WHO-FIC
FAIR
Study of Health in Pomerania
data dictionary
semantic enrichment
ISSN:2574-2531, 2574-2531
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Objective To semantically enrich the laboratory data dictionary of the Study of Health in Pomerania (SHIP), a population-based cohort study, with LOINC to achieve better compliance with the FAIR principles for data stewardship. Materials and Methods We employed a workflow that maps codes from the SHIP-START-4 laboratory data dictionary to LOINC codes following the terminology mapping principles and best practices recommended by the World Health Organization Family of International Classifications (WHO-FIC) Network. Results We were able to annotate 71 out of 72 (98.6%) of the source codes in the SHIP-START-4 laboratory data dictionary with LOINC codes. 32 source codes were mapped to a single LOINC code (cardinality 1:1) and 39 resulted in a complex mapping. All of the successful mappings are equivalent (=) matches. Discussion We increased the FAIRness of the SHIP laboratory data dictionary by semantically enriching laboratory items with links to an accessible, established, and machine-readable language for knowledge representation (LOINC). Our mapping improves semantic data retrieval and integration. However, not all clinically and significantly relevant data are included in the LOINC code. Therefore, these missing aspects have to be considered in data interpretation as well. Conclusion Semantically enriching the SHIP-START-4 laboratory data dictionary has contributed to its improved data interoperability and reuse. We recommend that data owners and standardization experts collaboratively perform annotations before data collection starts instead of doing this retrospectively. These experiences may inform the development of standard operating procedures for annotating data dictionaries developed for other population-based cohort studies. Lay Summary The Logical Observation Identifiers Names and Codes (LOINC) is an internationally recognized coding system that supports the unique identification and coding of medical laboratory tests. This study aimed to semantically enrich the SHIP-START-4 laboratory data dictionary by linking it to LOINC thus making it more findable, accessible, interoperable, and reusable (FAIR). The mapping principles and best practices recommended by the World Health Organization Family of International Classifications (WHO-FIC) Network were implemented in the semantic enrichment. Seventy-one out of 72 codes from the SHIP-START-4 laboratory data dictionary were successfully mapped to LOINC codes. This semantic enrichment enhances data retrieval, integration, and reuse. It is important to remember that some clinically relevant data is not covered by LOINC and must be considered separately. This work illustrates the importance of early collaboration in data annotation for better interoperability in future cohort studies. It also illustrates that it is better to perform annotations on data dictionaries before data collection starts instead of afterwards. The experiences gained while conducting this work may be relevant for other researchers who seek to develop a standard method for annotating medical data dictionaries for population studies.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2574-2531
2574-2531
DOI:10.1093/jamiaopen/ooaf010