ABSTRACT
OBJECTIVES: LOINC(R)(Logical Observations Identifiers, Names, Codes) is being used as the global standard for sharing laboratory test information and standardization. However, difficulties have been encountered in transferring local code to LOINC. Use in existing laboratory information systems(LIS) is possible with maximized local codes and LOINC mapping. Since the existing mapping tool has parts that do not match domestic medical environments, it is difficult to use without modification or supplementation. To this end, we have developed algorithms for LOINC mapping and have evaluated their usefulness. METHODS: We used 2,376 M-codes transformed from Pusan National University Hospital's 1,150 local codes, and codes from various laboratory test domains(Diagnostic Hematology, Clinical Chemistry, Seroimmunology, Molecular and Cytogenetics, Microbiology, Transfusion Medicine). In materializing the automatic mapping algorithms, spread sheet programs(Excel, Microsoft) and existing mapping tools(RELMA, Regenstrief) were used. The accuracy of the mapped codes was verified by a specialist of the Laboratory Medicine Department. RESULTS: Of the 2,376 M-codes, mapping on LOINC was found to be possible for 78.7%(1,871) while LOINC corresponding with the local codes could not be found for 21.3%(505). Of the mapped codes, 90.8%(1,699) were mapped accurately automatically, while the rest were mapped manually. CONCLUSIONS: The LOINC mapping algorithm that was developed in this study was useful for mapping various forms of local code with LOINC.