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

Autor: Monsen KA; a School of Nursing, University of Minnesota , Minneapolis , MN , USA ., Finn RS; a School of Nursing, University of Minnesota , Minneapolis , MN , USA ., Fleming TE; b Gillette Children's Specialty Healthcare , St. Paul , Minneapolis , MN , USA , and., Garner EJ; a School of Nursing, University of Minnesota , Minneapolis , MN , USA ., LaValla AJ; a School of Nursing, University of Minnesota , Minneapolis , MN , USA ., Riemer JG; c Consultant , Riverside , CA , USA.
Jazyk: angličtina
Zdroj: Informatics for health & social care [Inform Health Soc Care] 2016; Vol. 41 (2), pp. 97-111. Date of Electronic Publication: 2014 Oct 17.
DOI: 10.3109/17538157.2014.965302
Abstrakt: Unlabelled: Rigor in clinical knowledge representation is necessary foundation for meaningful interoperability, exchange and reuse of electronic health record (EHR) data. It is critical for clinicians to understand principles and implications of using clinical standards for knowledge representation within EHRs.
Purpose: To educate clinicians and students about knowledge representation and to evaluate their success of applying the manual lookups method for assigning Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) concept identifiers using formally mapped concepts from the Omaha System interface terminology.
Methods: Clinicians who were students in a doctoral nursing program conducted 21 lookups for Omaha System terms in publicly available SNOMED CT browsers. Lookups were deemed successful if results matched exactly with the corresponding code from the January 2013 SNOMED CT-Omaha System terminology cross-map.
Results: Of the 21 manual lookups attempted, 12 (57.1%) were successful. Errors were due to semantic gaps differences in granularity and synonymy or partial term matching.
Conclusions: Achieving rigor in clinical knowledge representation across settings, vendors and health systems is a globally recognized challenge. Cross-maps have potential to improve rigor in SNOMED CT encoding of clinical data. Further research is needed to evaluate outcomes of using of terminology cross-maps to encode clinical terms with SNOMED CT concept identifiers based on interface terminologies.
Databáze: MEDLINE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje