摘要
目的为解决多中心间的异质性问题。方法系统介绍Bayes线性混合效应模型原理与方法。结果通过WinBUGS软件和SAS软件实例分析,认为二者固定效应结果相同,WinBUGS软件Bayes分析混合效应结果可提供更多效应变异的信息。结论Bayes线性混合效应模型分析结果解释更合理,推论范围更广。
Objective To solve the heterogeneity among centers in multi-center clinical trials. Methods To explore Bayes basic inference and several method in analysis of mixed model. Results Appling the WinBUGS and SAS to model for the example of phase Ⅱ multl-centre clinical trials, told us Bayes approach makes the same explanation in solution for fixed effects as SAS but more information in solution of mixed effects. Conclusion Bayes analysis of linear mixed effects model could be used to more variety of states, it is a very important method worth to generalize.
出处
《中国卫生统计》
CSCD
北大核心
2005年第4期200-203,共4页
Chinese Journal of Health Statistics
