A toolkit for clinical statisticians to fix problems based on biomarker measurements subject to instrumental limitations: From repeated measurement techniques to a hybrid pooled–unpooled design

The aim of this chapter is to review and examine different methods in order to display correct and efficient statistical techniques based on complete/incomplete data subject to different sorts of measurement error (ME) problems. Instrument inaccuracies, biological variations, and/or errors in questionnaire-based selfreport data can lead to significant MEs in various clinical experiments. Ignoring MEs can cause bias or inconsistency of statistical inferences. The biostatistical literature well addresses two categories of MEs: errors related to additive models and errors caused by the limit of detection (LOD). Several statistical approaches have been developed to analyze data affected by MEs, including the parametric/nonparametric likelihood methodologies, Bayesian methods, the single and multiple imputation techniques, and the repeated measurement design of experiment. We present a novel hybrid pooled–unpooled design as one of the strategies to provide correct statistical inferences when data is subject to MEs. This hybrid design and the classical techniques are compared to show the advantages and disadvantages of the considered methods.