Maximum likelihood, multiple imputation and regression calibration for measurement error adjustment.

Maximum likelihood, multiple imputation and regression calibration for measurement error adjustment.
复制标题

DOI:
10.1002/sim.3458
复制
发表时间:
2008-12-30
影响因子:
2
通讯作者:
Natarajan, Loki
Natarajan, Loki
中科院分区:
医学3区
文献类型:
--
作者:
Messer, Karen;Natarajan, Loki

文献摘要

参考文献

被引文献

相似文献

在疾病相关性的流行病学研究中,对于大多数样本,通常只有暴露的替代指标。可进行验证子研究,以估计替代测量值与真实暴露水平之间的关系。在这篇文章中,我们讨论了这样一个主要研究/验证研究设计的三种估计方法:(i)最大似然(ML),(ii)多重插补(MI)和(iii)回归校准(RC)。对于logistic回归,我们展示了每种方法如何依赖于不同的数值近似的可能性,我们适应标准软件来计算多重插补和最大似然估计。我们使用模拟来比较现实和极端设置的估计器的性能,以及内部和外部验证设计。我们的研究结果表明,大的测量误差或足够大的样本量,ML表现一样好或优于MI和RC。然而,对于较小的测量误差和小样本量,ML或RC可能具有优势。有趣的是,在大多数情况下,RC与ML的相对优势是由相对方差而不是估计量的偏差决定的。提供了SAS中所有三种方法的软件代码。
In epidemiologic studies of exposure-disease association, often only a surrogate measure of exposure is available for the majority of the sample. A validation sub-study may be conducted to estimate the relation between the surrogate measure and true exposure levels. In this article, we discuss three methods of estimation for such a main study / validation study design: (i) maximum likelihood (ML), (ii) multiple imputation (MI) and (iii) regression calibration (RC). For logistic regression, we show how each method depends on a different numerical approximation to the likelihood, and we adapt standard software to compute both multiple imputation and maximum likelihood estimates. We use simulation to compare the performance of the estimators for both realistic and extreme settings, and for both internal and external validation designs. Our results indicate that with large measurement error or large enough sample sizes, ML performs as well or better than MI and RC. However, for smaller measurement error and small sample sizes, either ML or RC may have the advantage. Interestingly, in most cases the relative advantage of RC versus ML was determined by the relative variance rather than bias of the estimators. Software code for all three methods in SAS is provided.
DOI: 10.1093/ije/dyl097
发表时间: 2006-08-01
影响因子: 7.7
作者:
Cole, Stephen R.;Chu, Haitao;Greenland, Sander
通讯作者: Greenland, Sander
DOI: 10.1093/aje/kwj082
发表时间: 2006-04-15
影响因子: 5
作者:
Natarajan, L;Flatt, SW;Pierce, JP
通讯作者: Pierce, JP
DOI: 10.1093/jnci/88.23.1738
发表时间: 1996-12-04
期刊: JOURNAL OF THE NATIONAL CANCER INSTITUTE
影响因子: --
作者:
Prentice, RL
通讯作者: Prentice, RL
DOI: 10.1093/aje/kwg091
发表时间: 2003-07-01
影响因子: 5
作者:
Kipnis, V;Subar, AF;Carroll, RJ
通讯作者: Carroll, RJ
DOI: 10.2307/2004418
发表时间: 1969-01-01
影响因子: 2
作者:
GOLUB, GH;WELSCH, JH
通讯作者: WELSCH, JH