目的:本研究旨在探討某區域醫院疾病分類編碼品質。方法:擷取92年度某區域醫院住院醫療資料檔之膽囊及/或膽道結石病例共507個案例為研究樣本,首先由二位疾病分類專家就病歷重新再編碼,比對原始疾病分類編碼與專家第二次編碼之一致性,再隨機抽取50本病歷由第三位疾病分類專家進行再測信度測試。資料分析方法有描述性統計、成對T檢定、卡方檢定、Pearson相關分析、及羅吉斯迴歸等。結果:(1)整體編碼錯誤率為13.53%,診斷錯誤率為12.86%,而處置錯誤率為14.38%;整體編碼錯誤個數以忽略可編之次要處置(29.38%)佔居多,其次以診斷編錯代碼(20.94%),再其次此診斷可忽略不編碼(15.00%);(2)診斷編碼個數、處置編碼個數及住院天數分別與錯誤個數都呈正相關(p<0.0001);(3)開刀房處置與非開刀房處置編碼錯誤有顯著性關聯(p<0.05),在忽略編碼也有顯著性關聯(p<0.05);(4)改變DRGs分派之編碼錯誤原因,編錯代碼為55.56%;病歷書寫不完整為11.11%;(5)在控制病例特性診斷個數、處置個數、及住院天數等變項之下,處置編錯代碼、主要處置選擇錯誤、診斷病歷書寫不完整、診斷編錯代碼、忽略可編之次要診斷等編碼錯誤原因改變DRGs分派有統計上顯著影響(p<0.05)。結論:住院天數愈長病情愈複雜,則編碼錯誤個數愈多。DRGs分派改變之主要原因為處置編錯代碼,次要的為主次處置倒置。值得一提的是診斷病歷書寫不完整卻比主要診斷選擇錯誤更重要。因此疾病分類人員對於複雜度較高之病歷更應謹慎判讀,此外病歷書寫品質之改善不容忽視,方能確保最低的編碼變異。 Objectives: The aim of the study was to evaluate the accuracy of coding in a regional hospital. Furthermore, identifying that errors of coding was associated with the changes of DRGs assignment. Methods: The subjects were a sample of 507 admissions with cholelithiasis and/ or choledocholithisis from January 1st 2003 to December 31st 2005 in a regional hospital inpatient claim data. A retrospective audit was carried out, involving comparison of the original coding in the medical records with blindly recoding by two senior certified coders. The main outcome measures were the levels of disagreement between original and auditing coding. The reasons for DRGs discrepancies after DRGs grouping were also determined. The descriptive statistics, pair t-test, chi square test, pearson’s correlation and logistic regression analysis were performed with SPSS 12.0. Results: (1) Coding errors occurred in 13.53 per cent of all codes, 12.86 per cent of diagnoses and 14.38 percent of procedures. The three common coding errors were other procedure undercoding (29.38%), diagnosis miscoding (20.94%), and diagnosis unnecessary coding (15.0%). (2) The number of coding errors were positively related to LOS, number of diagnosis and procedures codes (p<0.0001). (3) Significant associations were found between operating room (OR) and non OR procedure for coding error and undercoding. (4) Miscoding and incomplete medical record documentation accounted for the change in DRGs assignment 55.56% and 11.11%, respectively. (5) After controlling for patient characteristics, predictors of change in DRGs assignment were miscoding of procedures, resequencing of procedure, incomplete diagnosis documentation, miscoding of diagnosis and undercoding of secondary diagnosis (p<0.05). Conclusions: Coding errors were more prevalent in complex cases with multiple codes. The predictors of DRGs discrepancy suggest that coders need to work cautiously for coding in more complex cases and the need to improve medical record documentation to ensue that DRGs variation is kept to a minimum.