中国卫生统计
中國衛生統計
중국위생통계
CHINESE JOURNAL OF HEALTH STATISTICS
2010年
1期
11-12,16
,共3页
陈炳为%郑水龙%刘沛%陈启光
陳炳為%鄭水龍%劉沛%陳啟光
진병위%정수룡%류패%진계광
潜在分类模型%无金标准%诊断试验%灵敏度%特异度
潛在分類模型%無金標準%診斷試驗%靈敏度%特異度
잠재분류모형%무금표준%진단시험%령민도%특이도
Latent class method%Diagnostic test%Absence of gold standard%Sensitivity%Specificity
目的 探讨无金标准诊断试验下灵敏度与特异度的估计方法.方法 以三个诊断试验为例,利用潜在分类模型估计无金标准诊断试验的灵敏度、特异度及其95%可信区间.并将潜在分类模型估计的结果与Bayesian方法估计结果进行比较.结果 三个诊断试验的灵敏度与特异度的95%可信区间均存在一定的重叠,表明两种方法估计结果较为相似.结论 诊断试验的灵敏度与特异度在无金标准时是无法直接计算的,可利用潜在分类模型与Bayesian方法估算诊断试验的灵敏度与特异度.
目的 探討無金標準診斷試驗下靈敏度與特異度的估計方法.方法 以三箇診斷試驗為例,利用潛在分類模型估計無金標準診斷試驗的靈敏度、特異度及其95%可信區間.併將潛在分類模型估計的結果與Bayesian方法估計結果進行比較.結果 三箇診斷試驗的靈敏度與特異度的95%可信區間均存在一定的重疊,錶明兩種方法估計結果較為相似.結論 診斷試驗的靈敏度與特異度在無金標準時是無法直接計算的,可利用潛在分類模型與Bayesian方法估算診斷試驗的靈敏度與特異度.
목적 탐토무금표준진단시험하령민도여특이도적고계방법.방법 이삼개진단시험위례,이용잠재분류모형고계무금표준진단시험적령민도、특이도급기95%가신구간.병장잠재분류모형고계적결과여Bayesian방법고계결과진행비교.결과 삼개진단시험적령민도여특이도적95%가신구간균존재일정적중첩,표명량충방법고계결과교위상사.결론 진단시험적령민도여특이도재무금표준시시무법직접계산적,가이용잠재분류모형여Bayesian방법고산진단시험적령민도여특이도.
Objective To explore the methods to evaluate sensitivity and specificity of diagnostic test without gold standard.Methods Taking three diagnostic test without gold standard for example,Sensitivity,specificity and their 95% confidence interval were computed by latent class model(LCM).And results were compared with those evaluated by Bayesian model.Results The 95% confidence interval of sensitivity and specificity by LCM and those of Bayesian models had overlapped,their parameter were similar.Conclusion Without gold standard,sensitivity and specificity would not to be computed directly.But,latent class model and Bayesian model could be used to estimate sensitivity and specificity and their 95% confidence interval in absence of gold standard.