Standardization of Terminology in Laboratory Medicine II

In 2004, the National Health Information Standard Committee (NHISC) was established under the Korean Ministry of Health and Welfare (MOHW) to standardize national health information, and 14 practical committees composed of field specialists were organized The laboratory test terminology practical committee was in charge of standardizing laboratory medicine terminology, in vitro radioimmunoassay, and pathology.

In Korea, the work regarding the terminology used in the medical field started in 1977 with the publication of the 1st edition of medical terminology dictionary "English-Korean, Korean-English medical terminology", which was revised in 2001 in its fourth edition (1), but research into standardizing terminology regarding information exchange has been rare. Yoon et al. tried to standardize and to code laboratory medicine terminology in the 1990's, but implementation of this terminology into local laboratory systems was unsuccessful.

Whereas digitalization and computerization has progressed rapidly in Korean medicine, data have not been easily interchangeable because medical information systems were developed independently at individual institutes. To fix this problem, medical terminology should be standardized because it is essential in data transmission between health institutes and would maximize the benefits of information technology by decreasing, unnecessary test duplicates, enabling interchangeable electronic medical record (EMR)/electronic health record (EHR), and utilizing the individual health care Smart Card.

Among efforts to standardize medical terminology, one of the most representative systems in laboratory terminology is the Logical Observation Identifier Names and Codes (LOINC) system developed by the Regenstrief Institute for Health Care at Indiana University in 1994 (2). It provides a standard for data exchange among the laboratories in America, is continuously revised to reflect changes in laboratory tests, and includes most of the laboratory observations.

Based on the first year of this study, from December 2004 to May 2005 (3), we decided to standardize test name, specimen, units, result description terms, etc and to adopt the six axial structure of LOINC that can be used in national public health network. During the second year, a standard database of concept names for all laboratory terms was developed, including the terms used by tertiary health care institutes and reference laboratories in Korea.

Entire laboratory terms among LOINC database (4) were considered to be adopted. Test names that either had been deleted or were not currently used in Korea were excluded. It was discussed with the staff in Regenstief whether LOINC could be used as one of the key terminology sets to be adapted, and we were told that there was no specific obligation in using them.

Standard expression was determined by an "English-Korean, Korean-English medical dictionary" (1) published by the Korean Medical Association and "the rules for foreign language expression" were legislated by the national institute of the Korean language (5). The names of microorganisms and antigens for allergy tests were expressed in English. Abbreviations frequently used in routine work were selectively adopted after inquiry to the committee or a section of specialists in the relevant field. A program for mapping between LOINC and electronic data interchange (EDI) codes, and a program for translation from English into Korean were used. To collect the opinions of experts, a public hearing and a workshop for the clinical pathologists were held on 28 September 2005 and 14 April 2006, respectively, and opinions from these meetings were reflected in this work.

We named the Korean standard laboratory terminology system the Korean Logical Observation Identifier Names and Codes (K-LOINC). Korean concept names for laboratory tests were expressed as combinations of six-axis concept names: components, property, time aspect, system (specimen), scale type, and method type. The naming principle followed the LOINC user manual (6).

The basic concept name table consisted of K-LOINC code, K-LOINC component, K-LOINC property, K-LOINC time aspect, K-LOINC system, K-LOINC scale, K-LOINC short name, LOINC code, LOINC component, LOINC property, LOINC time aspect, LOINC system, LOINC scale, LOINC short name, EDI code, EDI test name, UMLS unique concept identifier (CUI) code, and UMLS name for 28,858 test observations (Table 1, 2). Among the 30,740 LOINC codes, 89.6% were adopted and 1,307 were added. A mapping table for the EDI codes and UMLS was also prepared. Synonym tables for concept names, words, and terms using the six axes were prepared to help find concept names with common terms used in the fields. The final product from the 2nd year of this study is accessible on-line at the homepage of the National Health Information Standard Committee (7).

Table 1
Final products of the standardization in laboratory terminology

• Click for larger image

Table 2
K-LOINC and LOINC by class

• Click for larger image

During the 1st year's work, we examined a suitable structure in Korea and found that four axes, test name, specimen, method, and scale type were enough to express all laboratory observations. However, the six-axis concept names of LOINC, consisting of components, property, time aspect, system (specimen), scale type, and method type, was adopted in consideration of international data exchange in the future.

In the LOINC system, the method field was specified only when there was a significant difference in the test results or reference intervals. To reflect the special situation of the presence of an independent department in the analysis and reimbursement for radioimmunoassays (RIA) in Korea, codes for in vitro radioimmunoassay were duplicated and the method was specified as RIA.

Because terminology for microbiology in the LOINC system was quite different from those currently use in Korea and EDI codes, several inquiries and consultations were conducted with the relevant experts of the Korean Society of Laboratory Medicine. Finally, this study decided to adopt the LOINC system per se and to make tries to apply into local laboratories by combining concept names. Still, mapping with EDI codes was almost impossible.

It was necessary to match K-LOINC (LOINC codes adopted in Korea) with EDI to facilitate implementation of a standard code at the local hospital and institutes. Several problems, however, arose because the structure and granularity for concepts between the two systems were significantly different. Because EDI codes were developed to reimburse laboratory tests fees, it used the same codes for running tests at the same cost, regardless of methods or specimens, which leads to n:1 mapping of the standard codes vs. EDI. On the contrary, there were EDI codes, including several tests as a set, which led to 1:n mapping. Most standard codes for describing results were not matched to EDI codes because of the absence of concepts in the EDI system.

LOINC for molecular biology tests did not completely cover the present tests. Because this field is growing rapidly, it was suggested that a standing committee be organized to create a consistent naming system for the upcoming tests, which need new test codes and to refine present test names. The National Health Information Standard Committee (NHISC) uses UMLS codes to connect standard health information terminologies in each division. To map K-LOINC to UMLS, mapping tables of the LOINC database and UMLS were used.

To facilitate the implementation of standard terminology in local laboratories, it is necessary to have instruments and/or reagent companies provide accurate standard codes for the instruments and reagents they produce, as well as provide mapping tables with Korean EDI codes for reimbursement.

Terminology for laboratory tests is one of the well standardized areas with a report for coding clinical laboratory information in Japan as well as LOINC and UMLS in America.(8) The LOINC system, however, initiated by the civil institute, and UMLS, provided by the government, were not fully implemented in laboratories and there has been no report of trying these systems in Australia or Europe. In this respect, nation-wide standardization of health information terminology would be a model when it is successfully implemented and would greatly contribute to the global competitiveness of Korea. Here, we report the Korean standard laboratory terminology database for test names, result description terms, and result units covering most of the laboratory tests in Korea.