参数法与非参数法计算 AUC 评价诊断试验准确性的样本量对比
摘要
目的:比较不同参数下参数法与非参数计算 A U C 时所需样本量,为评价诊断试验准确性估算样本量提供参考。方法:据Obuchowski NA 等人提出的公式估计初始值,Monte Carlo 模拟调整参数法、非参数法计算 AUC 时样本量。结果:参数法计算 AUC 时,所需样本量较非参数法少,预期q > 0 . 9 0 0 时,所需样本量少于 5 0。结论:评价双正态模型时,可据参数法计算 A U C 进行样本量估计。参考
Krzanowski WJ and Hand DJ.ROC curves for continuous data.Boca Raton,
FL:CRC Press,2009.
Zhou XH,Obuchowski NA and McClish DK.Statistical methods in diagnostic medicine.Vol.712,New York:John Wiley&Sons,2011.
Feng Dai,Cortese Giuliana,Baumgartner Richard.A comparison of
confidencecredible interval methods for the area under the ROC curve for continuous diagnostic tests with small sample size[J].Statistical Methods in Medical
Research,2017,26(6):1-23.
宋花玲,贺佳,黄品贤,李素云.ROC曲线下面积估计的参数法与非参数
法的应用研究[J].第二军医大学学报,2006,27(07):726-728.
Obuchowski NA,McClish DK.Sample size determination for diagnostic
accuracy studies involving binormal ROC curve indices[J].Stat.Med.,1997,16(13):
-1542.
谷红梅,李康.诊断试验ROC参数估计双正态样本量估计方法探讨[J].
中国卫生统计,2006,23(1):2-4.
余松林.主编.医学统计学.北京:人民卫生出版社,2002:164-178.
李康.主编.医学统计学[M].7版.北京:人民卫生出版社,2018:72.
##submission.copyrightStatement##
##submission.license.cc.by-nc4.footer##