本研究建構含服務品質的Malmquist生產力指數,評估醫院長期生產力變動情形。生產力變動指數可分解為品質、效率與技術變動指數。考量四項投入、四項產出、一項服務品質,分析1994至1998年40家醫學中心與區域醫院生產力變動情形。結果顯示,各醫院長期生產力略呈退步現象,主因技術退步。平約而言,生產力退步2.7%,技術退步3.8%,效率進步2.0%,純效率進步1.0%。上述各變動幅度,經bootstrap信賴區間檢定約達統計顯著,然而服務品質並未顯著改變。無母數檢定發現,納入服務品質顯著影響對整體生產力、效率與技術變動的評估。 以迴歸分析法,探討影響醫院單期效率與跨期生產力、品質、效率與技術變動原因。在影響醫院單期效率因素方面,公立、平均住院日與疾病嚴重度較高的醫院效率略差,病床使用率與市場集中度較高的醫院效率較佳。總病床數較多的醫院效率較差,其程度隨病床數增加而減緩。在影響醫院跨期生產力、品質、效率與技術變動因素方面,全民健保實施、病床使用率與平均住院日顯著影響生產力、品質與效率的跨期變動;而公立醫院與總病床數顯著影響品質與效率變動;市場集中度亦影響品質變動;疾病嚴重度是減少技術變動主因。研究結果並歸納出醫療管理之政策意涵。
The objective of this study is to evaluate efficiency and productivity changes of hospitals. Data sources are the statistics for public and private medical institutions and for hospital medical services in Taiwan. The sample consists of panel data for inputs and outputs of 40 medical centers and regional hospitals from 1994 to 1998. An attribute-incorporated Malmquist productivity index is constructed and decomposed into attribute, efficiency, and technical change indexes. Bootstrap confidence intervals for the geometric means of these indexes are constructed for statistical inference of the average changes. Regressions are also performed to assess factors affecting efficiency and productivity changes of hospitals. Results indicate that most hospitals experienced a significant productivity slowdown due to technology and attribute regresses but efficiency significantly improved. The inception of the NHI program has had positive effects on hospitals' productivity and attribute changes but negative effects on efficiency changes. The findings support the necessity to incorporate attributes of medical services into the Malmquist productivity evaluation of hospitals. Additional studies may consider using better attributes to measure a hospital/s quality of services.