目的:以已申報的回溯性住院資料,模擬改善疾病分類編碼品質後,探討對住院醫療給付的影響。方法:以分層隨機抽樣法選取外科住院樣本病歷,再交由6位疾病分類人員編碼審核,最後實際編碼審核257本。結果:所有257本病歷的CMI值為1.27,經試算系統推估之最佳CMI值為1.95,兩者的差值為0.68,推估後可能增加的醫療給付佔總醫療給付的比率為33.8%。經審核後,共計修正36本,修正前的CMI值為1.24,而修正後則成為1.50。修正個案推估後可能增加的醫療給付佔總醫療給付比率為1.8%,與33.8%相比較,其佔率比為5.3%。結論:透過Tw-DRGs試算系統最佳化推估之醫療給付差額中,大約有5.3%是實際可爭取到的部分。再以實際醫療給付來推估,平均每個外科出院個案可額外爭取到1,158點的醫療給付。建議醫院宜在疾病分類編碼品質方面加強投資,包括疾病分類人員之教育訓練、Tw-DRGs最佳化資訊系統、改善病歷書寫品質等,相信可爭取到更合理的醫療給付。
There were 257 cases of medical records by stratified sampling depending on the proportion of year and office number from 2000 to 2006 in our hospital. The CMI value of all 257 sample cases was 1.27. After verification by certified coding specialists, these results revealed 36 revised cases, 1.24 of the CMI value, 0.26 of the difference of the CMI value, and 1.8% of payment difference proportion. Through the DRG optimization estimates system, 5.3% of medical benefit payment differences could be achieved in reality. Estimating with 103 actual medical benefit payments, the extra average Relative Value Unit (RVU) per case was 1,158. The hospital could enforce more ICD-coding quality for striving to achieve more rational medical benefit payments, including Certified Coding Specialists, optimization information system and the quality of medical records.