摘要
目的:探索均值填补方法在卫生项目评价指标值缺失中的应用并评价其填补效果。方法:通过秩和比评价法和Spearman相关系数比较评价均值填补法对实例卫生项目评价指标缺失值处理后填补结果的合理性。结果:3种方法的秩和比评价结果差异均有统计学意义(P<0.05),但评价指标缺失值数据的评价排序与2012年对应区卫生绩效排序结果差异较大,项目平均进度替代法的参与区评价排序与对应区卫生绩效排序结果一致(γs=0.761,P<0.05)。结论:卫生项目中指标值缺失会对秩和比评价的结果产生偏性,均值填补是一种简单有效的处理秩和比分析方法中指标值缺失的方法。
Objective： To explore the application of mean imputation method in dealing with missing indicator in comprehensive evaluation of health project and evaluate the effect. Methods ： The rationality of mean imputation method in dealing with missing indicator in comprehensive evaluation of health project was evaluated by rank sum ratio （RSR） evaluation method and Spearman correlation coefficient comparison. Results ： There was no statistically significant difference in RSR evaluation result among three methods （ P 〈 0. 05 ）, but the difference between missing value rank of evaluation index and health performance rank in the corresponding area was large, the rank of project mean imputation method agreed with health performance rank in the corresponding area （ys = 0. 761, P 〈 0. 05 ） . Conclusion： Missing indicator values of health project will produce bias in comprehensive evaluation of health project. Mean imputation method dealing with missing indicator is objective, simple and effective in RSR method for comprehensive evaluation of health project.
出处
《中国妇幼保健》
CAS
北大核心
2014年第24期3871-3874,共4页
Maternal and Child Health Care of China
基金
广东省广州市科技计划重大科技专项〔2011Y2-00025〕
广东省广州市母婴安康行动计划〔穗卫(2012)29号〕
关键词
卫生项目
秩和比
均值填补
Health project
Rank sum ratio
Mean imputation
