醫學中心營運標竿分析與管理決策路徑模式之建構

Abstract

醫學中心是台灣醫療體系中最高等級的醫院，也具有區域醫療的領導地位，更享有較高的健保給付。醫學中心無論在規模、設備、人員投入上或是醫療服務的提供上都比一般醫院掌握了數量上的優勢。因此醫學中心若能有效的營運，對於醫院的管理及健保的發展上均有良好的示範作用。 本研究係探討醫學中心民國94年~98年的營運效率表現，資料來源為衛生署統計之「醫療機構現況」及「醫院醫療服務量」兩大資料庫。在橫斷面的資料分析上，應用了CRS ( CCR )、VRS ( BCC )，Super-Efficiency，Context Dependent等模式作分析；而在縱斷面資料分析方面，則是利用了Malmquist Productivity Index ( MPI )、Metafrointer 等模式分析；其中在共同邊界模式分析中，並運用的Bilateral模式來比較公私立醫學中心的技術效率的差異及以無母數的Rank-Sum Test來檢定其差異情形。最後再利用各模式的效率分析結果，建構了標竿學習路徑圖、公私立醫學中心技術落差比趨勢路徑圖及醫院競爭優劣路徑圖等。 在資料包絡分析法的基本模式（CRS、VRS）中發現約佔43.9%的醫院是具相對有效率的表現，技術效率達有效率表現者更佔57.1%，規模效率達有效者為43.9 %，而且只有一家呈現規模報酬遞減（現有規模過大）的情形。在9家具有效率表現的醫院中，利用Super-Efficiency模式再找出營運成效最佳的標竿醫院，結果發現營運效率表現最好的是一家北部私立的財團法人醫院。在跨年度的時間序列分析上，不論短、中、長期的分析比較中發現醫學中心技術變革大都呈現進步的情形且生產力指數成長進步的家數均佔大多數，這表示醫學中心的生產技術有不斷提升的狀況。在共同邊界模式中發現公立醫學中心的平均技術落差比私立醫學中心為低，是為生產技術的落後；而由Bilateral模式及Rank-Sum Test檢定中得知公立醫學中心的技術效率較私立醫學中心為差並達顯著水準。而由情境相依模式中， 可以知道醫學中心的營運效率可分成四個不同階層，其中Level 1有9家醫院，Level 2有6家醫院，Level 3有4家醫院，Level 4有2家醫院；結果發現A醫院是各階層吸引力最大的醫院，其餘各階層各有不同吸引力與進步力改善情形。最後，本研究利用各模式的效率分析結果建構了三個主要的管理決策路徑模式：標竿學習路徑、技術落差比的趨勢路徑、競爭優劣路徑等，並分析各管理決策路徑模式中醫院所處位階與改善路徑。 本研究以橫斷面及緃斷面的效率分析模式分析醫學中心的營運效率表現並建構管理決策路徑模式，由效率分析與標竿學習方向希望醫學中心能以較少的資源達到最大的產出，並進而提升經營績效與競爭力，對於醫學中心的發展與國家醫療資源的運用可以更有幫助。

Medical centers are the highest level of hospitals in the medical system of Taiwan, and they also play the leading role in regional medical system. Moreover, they are provided with higher national health insurance coverage. The scale, equipment, manpower input and provision of medical service of medical centers are superior to those of general hospitals. Therefore, if medical centers can be effectively operated, they can become a great example for management of hospitals and development of national health insurance. This study intended to investigate the operational efficiency of medical centers from 2005 to 2009. The data sources were two major databases, Medical Care Institution’s Status, and Hospital’s Utilization, from the Department of Health. In terms of cross-sectional data analysis, models such as CRS (Constant Return Scale), VRS (Variable Return Scale), Super-Efficiency, and Context Dependent, were used to perform analyses. In terms of longitudinal data analysis, models such as Malmquist Productivity Index (MPI) and Metafrointer, were used to perform analyses. In Metafrontier, Bilateral model was used to compare the difference in technical efficiency between public and private medical centers. Moreover, nonparametric Rank-Sum Test was used to test the difference. The path diagrams of benchmarking, trend of technology gap ratio of public and private medical centers, and competitive advantages/disadvantages of hospitals were constructed based on the efficiency analysis results of various models. The basic models (CRS and VRS) of data envelopment analysis found that approximately 43.9% of the hospitals exhibited relatively efficient performance, 57.1% of them exhibited technical efficiency, 43.9% of them exhibited significant scale efficiency, and only one hospital exhibited decreasing returns to scale (the existing scale was too large). Among 9 hospitals with efficient performance, Super-Efficiency model was used to find out the benchmark hospital with the best operational efficiency. The result showed that the operational efficiency of a private medical center in the northern Taiwan was the best. The multi-year time series analysis found that the technological reformation of most of the medical centers improved and so did their productivity index in short-term, mid-term, and long-term analyses, suggesting that the production technology of medical centers were constantly improved. Metafrontier found that average technology gap of public medical centers was lower than that of private medical centers, suggesting that the production technology of public medical centers fell behind. The Bilateral model and the Rank-Sum Test showed that the technical efficiency of public medical centers was significantly poorer than that of private ones. The Context Dependent Model showed that medical centers could be divided into four different levels according to their operational efficiency. There were 9 Level 1 hospitals, 6 Level 2 hospitals, 4 Level 3 hospitals, and 2 Level 4 hospitals. The result showed that the attractiveness of Hospital A was the strongest and other levels of hospitals were characterized by different attractiveness and improvement in progress. This study constructed three managerial decision-making path models based on the results of efficiency analysis on various models: benchmarking path, trend of technology gap ratio path, competitive advantage/disadvantage path, and analyzed the level of hospitals in various managerial decision-making path models and the paths to be improved. This study used cross-sectional and longitudinal efficiency analysis models to analyze the operational efficiency of medical centers and constructed managerial decision-making path models. It was hoped that medical centers can yield the maximum output by using fewer resources, and their operational performance and competitiveness can be further improved from the perspectives of efficiency analysis and benchmarking, which may be more beneficial to the development of medical centers and national investment in medical care resources.