醫療費用的上漲已是全世界實施全民健康保險國家共同的壓力，依政府之規劃全 民健康保險長遠目標將採用診斷關係群 (DRGs) 作為醫院服務項目支付基準。在前瞻性付費 制度下，許多專家學者提出了控制醫療成本及維護醫療品質的管理方法與對策，長庚醫院將 各專家學者研究發表實際有效之管理方法，深入研究、探討，去蕪存菁、整合串連起來，建 立了前瞻性付費制度下之醫療管理模式。本研究之重點為針對此一管理式在效率及效益的發 揮、醫療服務品質的確保、資訊作業在簡化工作及異常管理之功效等方面做一實證研究，而 以長庚紀念醫院臺北林口、基隆、高雄三院區 79 年 7 月至 82 年 6 月之DRG371及DRG373 之住院病患資料共30,076例作實證研究。本研究以Unpaired t-test, ANOVA,Stepwise Logistic Regression ,Cross Validation 等研究方法統計分析在住院醫療管理模式運作 下之PPS制與FFS制是否在住院日、住院醫療費用及醫療品質等方面有所差異，以實證此管 理模式之義。研究結果顯示在住院醫療管理模式之運作下，醫療品質並未因給付制度係PPS 或FFS而有差異，且在醫療費用上確可發揮控制功能，顯示管理的機能可因嚴謹的制度訂定 及整合而發揮至極致，實證住院醫療管理模式確是一可提高效率、有具體效益、能維護醫 療品質且可行之管理模式，並且可供全民健康保險逐步實施DRGs制度之政策推動及執行上 參考。(中華衛誌1997；16(2)：149-159)

　　　　　Huge increase in medical expenses is common among countries with national health insurance. Hence, the National Health Insurance Bureau of Taiwan Has Adopted several polices such as referral, co-payment, case payment and DRGs as the long term payment goal,in order to protect its financial status against abuse of the newly introduced national health insurance system. To face this insurance scheme, new management models have to be developed by hospitals to minimize cost, but at the same time maintaining the quality of health care Standardization of procedures, concurrent and retrospective review, establishment of supporting services, as well as continuing education for physicians have all proven to be effective under the prospective payment system (PPs). To suit its organizational structure, Chang Gung Memorial Hospital (CGMH)has implemented an integrated management system, combining all the above mentioned concepts. The-oretical foundations of the model include contingency theory as well as management by participation. In addition, data were all processed by compters for abnormal management and to increase efficiency. The aim of this research therefore was to evaluate empirically the effectiveness of the system. Medical imformation of 30,076 in-patients admitted to various branches of CGMH from July 1990 to June 1993 are available. Only DRG 371 and DGR 373 were included in the analysis as two payment methods (PPS & FFS) can be found in the same period of time. Statistical techniques such as unpaired t-test,ANOVA, stepwise logistic regression and cross validation were used to test if there were any differences in length of stay (LOS), fees, and quality of care between the two methods. Results indicated that under this integrated management system,PPS was significantly better in cost control,but there was no differ-ence in LOS or quality of care. This is an implication that our model is both effect and efficient,yet is practical and can easily be implemented. Our experience may therefore serve as a reference for other hospitals when designing their own management models.(Chin J Public Health.(Taipei)： 1997； 16(2)：149-159)