主成分分析和数据包络分析的联合应用

目的 联合应用主成分分析和数据包络分析以综合评价某医院各临床科室的相对效率,提出改善建议.方法 根据医院情况,选择财务、工作量和病人满意度等各方面有代表性的12个指标,计算出16个临床科室的主成分得分和相对效率值,将2者分别排序并综合于二维象限,直观地评价和判断各科室相对效率.并得出非有效科室的松弛量,指出需调整的投入产出量,以达到相对有效.结果 二维象限显示各科室2种方法间排序一致,差异无统计学意义（P〉0.05）.位于第一象限的普外科、神经内科、肿瘤内科相对效率高（最优解θ=1）、综合情况较好 （PCA得分〉1）;而第四象限的小儿外科、眼科等相对效率高（最优解θ=1）,但主成分得分低（PCA得分〈-1.5）,规模偏小综合实力较差,应考虑予以发展;第二、三象限的感染病科、血液内科和口腔科则效率较差,为达到科室相对有效,在产出不变时,实际占用总床日数、职工人数、工资和福利、固定资产、材料及一次性消耗等投入应根据松弛量进行调整.结论 主成分分析和数据包络分析能互相补充,既体现评价单元的综合差异,又可比较相对效率,联合应用可以更全面地评价各决策单元.使用二维象限能更直观的判断较多评价单元的相对效率和综合情况,可以考虑在医院管理决策中予以推广.

Combined application of principal component analysis and data envelopment analysis

Objective To evaluate clinical departments＇ relative efficiency by principal component analysis （PCA） and data envelopment analysis （DEA） and provide improvements for the inefficient clinical departments. Methods To select 12 representative indicators including finance, workload, and patient satisfaction from 16 clinical departments to calculate scores by PCA and values of relative efficiency by DEA. To sort them separately and to integrate the results in dimensional quadrant and appraise each clinical department＇s operation directly. Results The difference of the sorting by the two methods was not significant （P〉0.05）. Some departments in the first quadrant such as surgery, neurology and oncology were effective（ DEA= 1 ）and outstanding （PCA scores〉 1 ）, while those in the fourth quadrant should be developed which were effective but less input-output. The others were less efficient and input should be adjusted according to slack when outputs were fixed. Conclusion PCA and DEA can complement each other and evaluate decision making unit more comprehensively. So they should be recommended to management decision-making in the hospital.