Bias Correction Methods for Misclassified Covariates in the Cox Model: Comparison of Five Correction Methods by Simulation and Data Analysis

Bang, Heejung; Chiu, Ya Lin; Kaufman, Jay S.; Patel, Mehul D.; Heiss, Gerardo; Rose, Kathryn M.

doi:10.1080/15598608.2013.772830

Cited by 18 publications

(21 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More advanced designs and models have been proposed and may be considered in future [16], [22], [23], [24]. Second, statistically speaking, this problem is a special case with measurement error or misclassification in outcomes [25]. However, we are cautious to recommend statistical correction with hard-to-satisfy-or-verify assumptions partly because of non-statistical reason; more data collection about blinding may do more harm than good [20], [21].…”

Section: Discussionmentioning

confidence: 99%

Random guess and wishful thinking are the best blinding scenarios

Bang

2016

Contemporary Clinical Trials Communications

Self Cite

View full text Add to dashboard Cite

Blinding is a methodologic safeguard of treatment evaluation, yet severely understudied empirically. Mathieu et al.’s theoretical analysis (2014) provided an important message that blinding cannot eliminate potential for bias associated with belief about allocation in randomized controlled trial; just like the intent-to-treat principle does not guarantee unbiased estimation under noncompliance, the blinded randomized trial as a golden standard may produce bias. They showed possible biases but did not assess how large the bias could be in different scenarios. In this paper, we examined their findings, and numerically assessed and compared the bias in treatment effect parameters by simulation under frequently encountered blinding scenarios, aiming to identify the most ideal blinding scenarios in practice. We conclude that Random Guess and Wishful Thinking (e.g., participants tend to believe they received treatment) are the most ideal blinding scenarios, incurring minimal bias. We also find some evidence that imperfect or partial blinding can be better than no blinding.

show abstract

Section: Discussionmentioning

confidence: 99%

Random guess and wishful thinking are the best blinding scenarios

Bang

2016

Contemporary Clinical Trials Communications

Self Cite

View full text Add to dashboard Cite

show abstract

“…Other discussed methods include probabilistic bias analysis [1, 50], Bayesian bias analysis [1, 73, 74], regression calibration [76, 77], modified maximum likelihood [78–80], multiple imputation [77, 81], and propensity score calibration [82–84] which all also require that appropriate assumptions are met. The review also articulates the research settings when each of these approaches is most appropriate.…”

Section: Methods and Assumptions Necessary To Reduce Bias Due To Missmentioning

confidence: 99%

Are All Biases Missing Data Problems?

2015

View full text Add to dashboard Cite

Estimating causal effects is a frequent goal of epidemiologic studies. Traditionally, there have been three established systematic threats to consistent estimation of causal effects. These three threats are bias due to confounders, selection, and measurement error. Confounding, selection, and measurement bias have typically been characterized as distinct types of biases. However, each of these biases can also be characterized as missing data problems that can be addressed with missing data solutions. Here we describe how the aforementioned systematic threats arise from missing data as well as review methods and their related assumptions for reducing each bias type. We also link the assumptions made by the reviewed methods to the missing completely at random (MCAR) and missing at random (MAR) assumptions made in the missing data framework that allow for valid inferences to be made based on the observed, incomplete data.

show abstract

“…The performance of MI has been compared with other methods, including RC and MR in settings of linear and logistic regression 19,23,28 and Cox regression. 22,29 These authors found that the optimal method depends on the size of the validation subset and degree of measurement error. Shepherd et al 28 noted MI worked well in the setting of correlated covariate and outcome measurement error in the linear model.…”

Section: Ta B L Ementioning

confidence: 99%

STRATOS guidance document on measurement error and misclassification of variables in observational epidemiology: Part 2—More complex methods of adjustment and advanced topics

Shaw

Gustafson

Carroll

et al. 2020

Statistics in Medicine

View full text Add to dashboard Cite

We continue our review of issues related to measurement error and misclassification in epidemiology. We further describe methods of adjusting for biased estimation caused by measurement error in continuous covariates, covering likelihood methods, Bayesian methods, moment reconstruction, moment-adjusted imputation, and multiple imputation. We then describe which methods can also be used with misclassification of categorical covariates. Methods of adjusting estimation of distributions of continuous variables for measurement error are then reviewed. Illustrative examples are provided throughout these sections. We provide lists of available software for implementing these methods and also provide the code for implementing our examples in the Supporting Information. Next, we present several advanced topics, including data subject to both classical and Berkson error, modeling continuous exposures with measurement error, and categorical exposures with misclassification in the same model, variable selection when some of the variables are measured with error, adjusting analyses or design for error in an outcome variable, and categorizing continuous variables measured with error. Finally, we provide some advice for the often met situations where variables are known to be measured with substantial error, but there is only an external reference standard or partial (or no) information about the type or magnitude of the error.

show abstract

Bias Correction Methods for Misclassified Covariates in the Cox Model: Comparison of Five Correction Methods by Simulation and Data Analysis

Cited by 18 publications

References 50 publications

Random guess and wishful thinking are the best blinding scenarios

Random guess and wishful thinking are the best blinding scenarios

Are All Biases Missing Data Problems?

STRATOS guidance document on measurement error and misclassification of variables in observational epidemiology: Part 2—More complex methods of adjustment and advanced topics

Contact Info

Product

Resources

About