Harel Ofer, Zhou Xiao-Hua
Department of Statistics, University of Connecticut, 215 Glenbrook Road Unit 4120 Storrs, CT 06269-4120, USA.
Stat Med. 2006 Nov 30;25(22):3769-86. doi: 10.1002/sim.2494.
In the case in which all subjects are screened using a common test and only a subset of these subjects are tested using a golden standard test, it is well documented that there is a risk for bias, called verification bias. When the test has only two levels (e.g. positive and negative) and we are trying to estimate the sensitivity and specificity of the test, we are actually constructing a confidence interval for a binomial proportion. Since it is well documented that this estimation is not trivial even with complete data, we adopt multiple imputation framework for verification bias problem. We propose several imputation procedures for this problem and compare different methods of estimation. We show that our imputation methods are better than the existing methods with regard to nominal coverage and confidence interval length.
在所有受试者都使用一种普通测试进行筛查，而只有这些受试者中的一个子集使用金标准测试进行检测的情况下，有充分记录表明存在一种偏差风险，称为验证偏差。当测试只有两个水平（例如阳性和阴性），并且我们试图估计测试的敏感性和特异性时，我们实际上是在构建一个二项比例的置信区间。由于有充分记录表明，即使有完整的数据，这种估计也并非易事，因此我们针对验证偏差问题采用多重插补框架。我们针对这个问题提出了几种插补程序，并比较了不同的估计方法。我们表明，在名义覆盖率和置信区间长度方面，我们的插补方法优于现有方法。
