Example of Differential Misclassification Bias

Pregnancy and Congenital Malformations

Lead Author(s): Jeff Martin, MD

Differential Misclassification of Exposure Variables
Schematic of Differential Misclassification
Differential Sensitivity and Specificity of Exposure Measurement
Result: Overestimation

Differential Misclassification of Exposure Variables

One of the classic examples of differential misclassification of exposure are studies of exposures during pregnancy and congenital malformations.

These are commonly done as case-control studies because of the rarity of the outcomes.

Schematic of Differential Misclassification

Differential Sensitivity and Specificity of Exposure Measurement

In these studies first, we may have differential recall of various exposures during pregnancy,

with cases less apt to forget things and controls more apt to forget things,
i.e. differential sensitivity of the exposure measurement.

Second, we may have differential specificity.

The cases may tend to recall things that truly did not happen.

Result: Overestimation

The result of all of this is that there is overestimation of the association between the exposure and the outcome.