目標：模擬若住院費用由論量計酬制度改採疾病診斷關係群（diagnosis related group；DRGs）制度，對不同特質醫院財務表現之影響。方法：以健保局1998年至2002年身份證字號加密的全民健保「住院醫療費用清單明細檔」，利用該局提供之第二代醫療資訊服務系統，轉出DRGs代碼；進而模擬平均每住院人次收益差為平均之各年度各醫院每住院人次在DRGs支付制度費用減去論量計酬制度下之費用（以下簡稱收益差），以廣義估計方程式模式（generalized estimating equation；GEE），探討年度效應、醫院特質（權屬層級別、平均編碼個數、平均住院日數、病例組合指標）、區域特質（縣市別醫療市場競爭程度、人口密度、每萬人口醫師數）與醫院收益差之關聯性。結果：若住院申報費用由論量計酬改採DRGs支付住院費用時，對地區醫院較有利，其各年整體收益差呈現正值、區域以上醫院則呈負值。就權屬別而言（除2002年未有評鑑等級醫院外），各層級醫院的平均收益差皆為公立高於私立；年度效應達統計上顯著差異，從1999年開始，醫院受益有逐年增加（受益差分別為-406元、-268元、-125元、46元）；而醫院特質中平均編碼個數每增加一個、收益差增加1640元，平均住院日數每增加1天、收益差減少1460元，病例組合指標值每上升1單位、收益差增加5055元。結論：模擬過去五年住院費用改採DRGs制度，呈現各年度的收益差有逐年改善，這可能與相關單位宣示實施DRGs致產生年度效應有關，而不同特質的醫院其收益差會有不同；建議健保局在全面實施DRGs前應注意DRGs支付制度對不同特質醫院所造成之財務衝擊。

　　　　　Objectives: To simulate the financial impact of different kinds of hospitals after the payment system changed to DRGs from fee-for-service.Methods: Data of inpatients from 1998 to 2002, who had their ID scrambed, were grouped into DRGs by the 2nd version software of medical information service system released from BNHI. The difference in the average revenue of per admission per hospital per year between the DRG and fee-for-service payment system was simulated in this study. The association factors with the difference in revenue, such as the year effect, hospital characteristics (ownership-accreditation, average of coding numbers, average length of stay and case-mix index) and regional characteristics (medical competitive level of county, population density and number of physicians) were investigated by the generalized estimating equation model.Results: If payment system was changed from fee-for-service to DRGs, the difference in revenue was positive for district hospitals every year, but negative for regional hospitals and medical centers. The difference in revenue for all of the public hospitals was more than private hospitals (except the hospitals without accreditation in 2002). The year effect for the difference in revenue was also statistically significant. The difference in revenue went from negative to positive compared to 1998 in the past 4 years (NT$ -406 in 1999, NT$ -268 in 2000, NT$ -125 in 2001 and NT$ 46 in 2002). The difference in revenue was increased NT$ 1640 per coding number added and decreased NT$ 1460 per day of length of stay increased. The difference in revenue was decreased NT$ 5055 per case-mix index increased.Conclusion: When the payment system changed from fee-for-service to DRGs, the difference in revenue became better year by year in this simulated study. This year effect was likely related to the announcement of DRG implemented by BNHI. The financial impact was different for different kinds of hospitals.