STRATOS guidance document on measurement error and misclassification of variables in observational epidemiology: Part 1—Basic theory and simple methods of adjustment

Abstract

Measurement error and misclassification of variables frequently occur in epidemiology and involve variables important to public health. Their presence can impact strongly on results of statistical analyses involving such variables. However, investigators commonly fail to pay attention to biases resulting from such mismeasurement. We provide, in two parts, an overview of the types of error that occur, their impacts on analytic results, and statistical methods to mitigate the biases that they cause. In this first part, we review different types of measurement error and misclassification, emphasizing the classical, linear, and Berkson models, and on the concepts of nondifferential and differential error.We describe the impacts of these types of error in covariates and in outcome variables on various analyses, including estimation and testing in regression models and estimating distributions. We outline types of ancillary studies required to provide information about such errors and discuss the implications of covariate measurement error for study design. Methods for ascertaining sample size requirements are outlined, both for ancillary studies designed to provide information about measurement error and for main studies where the exposure of interest is measured with error. We describe two of the simpler methods, regression calibration and simulation extrapolation (SIMEX), that adjust for bias in regression coefficients caused by measurement error in continuous covariates, and illustrate their use through examples drawn from the Observing Protein and Energy (OPEN) dietary validation study. Finally, we review software available for implementing these methods. The second part of the article deals with more advanced topics.

Association Between Residential Greenness, Cardiometabolic Disorders, and Cardiovascular Disease Among Adults in China

“…In addition, the cross-sectional design may have overestimated the mediation effects. 62 Second, we assessed greenness exposure based on communities and not individuals, which might have produced measurement error (ie, Berkson error 63 ). Although this error did not bias our effect estimates, it could have reduced statistical power.…”

Comparison of Bayesian and frequentist methods for prevalence estimation under misclassification

“…The model, expressed in the BUGS language, is basically a description of the process that provides parameters for data generation: Importantly, this guarantees that all probability mass for each of these variables is restricted to the domain [0, 1] thus eliminating the problem of prevalence estimates that are negative or larger than unity, and the same conveniently holds for limits of credible intervals. For the prevalence we use the uniform Beta (1, 1) as a non-informative prior distribution 2 . In contrast, prior information on the sensitivity and specificity of the diagnostic test is available from their respective validation studies.…”

“…Frequentist and Bayesian methods for bias adjustment of epidemiological risk estimates have *Correspondence: matthias.flor@bfr.bund.de 1 German Federal Institute for Risk Assessment, Max-Dohrn-Str. [8][9][10]10589 Berlin, Germany Full list of author information is available at the end of the article been reviewed in Keogh et al [2] and Shaw et al [3]. Estimation of prevalence is always based on the application of a diagnostic test to classify samples with respect to the binary trait under investigation.…”

Comparison of Bayesian and frequentist methods for prevalence estimation under misclassification

“…For the prevalence we use the uniform Beta (1, 1) as a non-informative prior distribution [2] . In contrast, prior information on the sensitivity and specificity of the diagnostic test is available from their respective validation studies.…”

“…A 95% credible interval (CrI) denotes a range of prevalence estimates that together account for 95% of the probability mass of the distribution. The 95% highest density interval (HDI) is the shortest 95% CrI, such that any value outside the HDI is considered less plausible [2] More accurately, the Beta (1, 1) prior is weakly informative, as it implicitly considers all possible values to be equally likely. However, for the sake of simplicity we will use the term non-informative throughout this article.…”