【摘要】 目的比较随机前沿分析方法在评价2003～2007年四川省县级及以上公立医院成本效率和技术效率过程中的应用,为该方法在我国医院效率评价方面的应用提供建议。方法采用随机前沿分析评价医院的成本效率和技术效率;似然比检验、秩相关分析、主成分分析等方法相继用于研究函数的最适模型和指标体系。结果随机前沿生产函数应变量为工作量所得技术效率为0.413,而应变量为主成分分析所得综合指标时技术效率上升为0.794。随机前沿成本函数应变量为总可变成本所得成本效率值为0.824,应变量为总成本所得技术效率值为0.540,二者相关系数为0.398;解释变量不含价格指标所得技术效率值为0.413,包含价格指标所得技术效率值为0.472。结论采用随机前沿生产函数评价医院的技术效率时,若应变量只选用工作量所得效率值并不理想;采用随机前沿成本函数评价医院的成本效率时,解释变量选用总可变成本所得效率值更好,在解释变量中加入价格指标后所得效率值要优于不含价格指标的测算方法。更多还原

【Abstract】 Objective:To investigate the cost efficiency(CE) and technical efficiency(TE) of all the public hospitals the county level and above by stochastic frontier analysis(SFA) from 2003 to 2007,so as to [WT6B1]provide reference to the proposed application of the methods for evaluation of hospital efficiency in China.Methods:CE and TE of hospital were evaluated by SFA.Likelihood ratio test,rank correlation analysis and principal component analysis were used to investigate the optimal model of function and the index system.Results:TE was 0.413 when chosen the workload as dependent variable of stochastic frontier production function,while that increased to 0.794 if changed dependent variable to a composite index derived from component analysis-based.TE was 0.413 when the explanatory variable did not include price indices,while that was 0.472 when including the price indices.CE was 0.824 when chosen the total variable cost as dependent variable of stochastic frontier cost function,while that was 0.794 if changed dependent variable to total cost.The correlation coefficient of the two CE values was 0.398.Conclusion:It is not suggested to use workload as dependent variable when evaluating TE of hospital by stochastic frontier production function.It is suggested to choose total variable cost as explanatory variable when evaluating CE of hospital by stochastic frontier cost function.Adding price indices in the explanatory variable is recommended.更多还原