為有助於提升國內疾病分類編碼品質，本研究採文獻探討的研究方式，分析自1992年至2005年共29篇編碼正確性評估相關文獻，從中歸納出編碼品質定義、評估方法、編碼流程、及編碼錯誤原因。本研究發現多數編碼正確性研究是在探討ICD代碼的再現性或變異性，而不是病歷資訊的正確。此外較複雜病例、及醫師病歷書寫品質不良與編碼錯誤有密切地關聯。疾病分類人員之編碼錯誤原因有主要診斷選擇錯誤、錯誤編碼、主次診斷倒置、不當切割、及編碼高報等。最後參照研究結果對編碼品質研究人員，醫師及保險相關機關提出建議。 To help promote the quality of ICD codes in Taiwan, this paper analyzed twenty-nine papers published in English from 1992 to 2005 about evaluation of ICD code accuracy. From these the author generalized the definition of quality, methods for evaluating code accuracy, coding process, and potential error sources. This paper found many studies examined the reproducibility or the variability between the original and the reviewer’s code rather than the accuracy of medical record information. More complex cases and poor quality documentation were more prone to coding errors. Other researches revealed the reasons of errors were coder errors such as misspecfication, miscoding, resequencing, unbundling, and upcoding. Based on the results mentioned above, suggestions for the research, hospital and insurer were provided.
