摘要
目的探讨无金标准情况下诊断试验的评价方法及患病率的估计。方法从Beyesian参数估计原理出发，用Gibbs抽样法完成参数后验密度的估计。结果将有金标准的资料作为无金标准处理，所得参数估计值前后吻合，效果良好，并可得到参数的后验密度分布图。结论上述方法可有效地估计无金标准情况下诊断试验的评价参数。
Objective To find out an estimating method for prevalence and the parameters of diagnostic tests in the absence of gold standard Metkods Bayesian estimation thcory is used. Marginal posterior densties of all parameters are estimated using Gibbs sampler. Results We illustrate the methods using data sets that are in the presence of gold standard. The results of diagnostic test assessment are homogeneously either presence or absence of gold standard. We can also obtain the distribution of these parameters. Conclusion The method can be used to assess diagnostic test and estimate prevalence effectively in the absence of gold standard.
出处
《中国卫生统计》
CSCD
北大核心
1999年第4期203-205,共3页
Chinese Journal of Health Statistics
关键词
诊断试验
无金标准
BAYESIAN
参数估计
Diagnostic test Absence of gold standard Bayesian estimation Gibbs sampler
