Bias correction methods for misclassified covariates in the Cox model: Comparison of five correction methods by simulation and data analysis

Measurement error/misclassification is commonplace in research when variable(s) cannot be measured accurately. A number of statistical methods have been developed to tackle this problem in a variety of settings and contexts. However, relatively few methods are available to handle misclassified categorical exposure variable(s) in the Cox proportional hazards regression model. In this article, we aim to review and compare different methods to handle this problem - naive methods, regression calibration, pooled estimation, multiple imputation, corrected score estimation, and MC-SIMEX - by simulation. These methods are also applied to a life course study with recalled data and historical records. In practice, the issue of measurement error/ misclassification should be accounted for in design and analysis, whenever possible. Also, in the analysis, it could be more ideal to implement more than one correction method for estimation and inference, with proper understanding of underlying assumptions.