Patterns of bias due to differential misclassification by case–control status in a case–control study

Case-control study is still one of the most commonly used study designs in epidemiological research. Misclassification of case-control status remains a significant issue because it will bias the results of a case-control study. There exist two types of misclassification, differential versus nondifferential. It is commonly accepted that nondifferential misclassification will bias the results of the study towards the null hypothesis. Conversely, no reports have assessed the impact and direction of differential misclassification on odds ratio (OR) estimate. The goal of the present study is to demonstrate by statistical derivation that patterns exist on the bias induced by differential misclassification.