LOINC (Logical Observation Identifier Names and Codes)爲美國Regenstrief Institute與病理學會(the College of American Pathologists)資訊委員會所共同制定,其主要目標就是提供統一通用的名稱與編碼用來辨識檢驗(laboratory test)和臨床觀察(clinical observation)結果報告,作爲電子病歷訊息交換的標準。本文應用「健保碼與LOINC」(NHI-LOINC)檢驗對應資料庫,建立LOINC資料庫與醫院檢驗資訊資料庫的對應關係與轉換模式,從而開發一套電腦輔助比對系統,協助醫院從檢驗資訊系統中將院內檢驗資訊轉換成國際使用之LOINC碼。本系統利用恩主公醫院檢驗資訊系統兩個月的線上作業資料來實際測試轉換能力與效益。恩主公醫院檢驗資訊系統兩個月的線上作業資料總共有172,167筆,自動完成比對轉換(完全不需專家輔助選擇)的有123,504筆,佔71.73%;需要經過專家輔助完成比對轉換的有48,663筆,佔28.27%。本研究結果顯示,在實際運作時,只有不到30%的檢驗項目需要專家介入,因此,本系統可以有效率的協助檢驗院內碼對應至LOINC的工作,將有助於推動LOINC成爲台灣檢驗資訊交換標準。
LOINC (Logical Observation Identifier Names and Codes) has been developed by the Regenstrief Institute of the United States and the College of American Pathologists Information Committee. The main goal is to provide unified names and codes used for laboratory tests and clinical observation reports to facilitate the exchange of electronic medical records. In this paper, we used the NHI-LOINC mapping knowledge base to establish a mapping from the laboratory test codes of a hospital to LOINC, and developed an auto-transformation system to demonstrate the feasibility of such mapping. In order to assess the performance of the automated mapping system, we collected two-month laboratory tests performed by En Chu Kong hospital. There were 172,167 laboratory tests in total. 71.7% (123,504) of them were automatically mapped to the corresponding LOINC without any assistance from laboratory experts, but the rest (28.3%) needed some involvement of the experts to complete the mapping. Our study shows that in practice we can design such system to perform the mapping from the laboratory test codes of a hospital to the corresponding LOINC efficiently, and this may result in speeding up the adoption of LOINC in Taiwan.