论文部分内容阅读
目的提出大型纵向监测健康体检数据统计分析策略。方法从数据预处理、马尔科夫链蒙特卡罗填补、广义估计方程纵向分析等几个方面介绍该策略,并以白细胞对代谢综合征的影响研究为例说明其应用性。结果将粗糙的原始数据进行预处理后,填补了大量不完整数据;利用广义估计方程进行纵向统计分析,克服因忽略组内相关造成的问题,使时间因素得到控制和体现。结论大型纵向监测健康体检数据统计分析策略可有效地利用查体记录的各类信息,极大地提高了统计分析质量。
Objective To propose a statistical analysis method of large-scale longitudinal monitoring health examination data. Methods This method is introduced from the aspects of data preprocessing, Markov chain Monte Carlo filling, longitudinal analysis of generalized estimation equations and so on. The application of leucocyte to metabolic syndrome as an example is illustrated. The results of the rough raw data pretreatment, to fill a large number of incomplete data; the use of generalized estimation equations for longitudinal statistical analysis, to overcome the neglect of group related problems caused by the time factor to control and reflect. Conclusion The statistical analysis method of large-scale longitudinal monitoring health examination data can effectively utilize all kinds of information recorded by physical examination and greatly improve the quality of statistical analysis.