非参数DEA方法在陕西医院可变要素投入上的相对效率之研究

Nonparametric DEA and Relative Efficiency of Shaanxi Hospitals in Use of Variable Factors

目的将引入投资约束的DEA模型应用于陕西医院在1997到2000年间的绩效分析。方法医院在可变投入要素使用上的总效率被分解为技术效率与资源配置效率分别估算。讨论半参数对称修剪最小二乘回归分析(STLS)在DEA结果检验问题上比经典的Tobit分析更为适切。结果若所有医院均以相对最优方式运作,可以节省6.24亿元,占总预算的7.71%。有264个决策单位大致可以认为规模大小是理想的,其余的规模或者过大,或者过小。结论改进的DEA模型可修正传统模型对的低估。陕西医院总体效率较高,但规模有待调优。

Objective The performance of Shaanxi hospitals from 1997 to 2000 is assessed by using a DEA model augmented with investment constraints. Methods The overall efficiency of hospitals in use of variable factors is decomposed into technical and allocative efficiency scores. It is also discussed that the semi-parametric STLS regression is more relevant than the classical Tobit analysis in testing DEA estimate of efficiency. Results About 0.624 billion Yuan, i. e., 7.71% of the budget, could be saved if all hospitals had run at their optimal level. 264 DMUs are found of almost ideal size while the others are either too small or too large. Conclusion The augmented DEA model can improve underestimation of traditional model for hospitals＇ effciency. In general Shaanxi hospitals were quite efficient but the sizes of most hospitals are to be optimized.