| dc.description.abstract | Background. Biases are factors that introduce lack of internal validity i.e., factors that lead to an incorrect assessment of the correlation between an exposure and its effect on the population. Exposure misclassification bias is when data collection is imperfect within a categorical exposure variable. Job-exposure matrixes (JEM) are tools used in exposure assessment that assign exposures to job-titles and can introduce this bias, as exposure here is often categorised in two or more groups. Methods can be applied to adjust for this bias, called ‘quantitative bias analyses’. These methods are not widely used, however, there is good literature present to help consider whether it is useful for your research and if so, how to apply it.
Objective. The objective of this study was to review studies that applied bias analysis for exposure misclassification which was introduced by using job-exposure matrix (JEM) as their exposure assessment method.
Methods. A literature review about the application of QBA on exposure misclassification bias introduced by using JEMs in occupational epidemiology was performed.
Results. Of the total of 5 papers, three eligible papers were found in the query, and one in the afore mentioned papers’ supplementary materials references. Two of these papers performed sensitivity analyses, while the other two applied actual methods of QBA namely, Probabilistic Bias Analysis (PBA) and a Bayesian approach.
Conclusion. While methods of QBA have been successfully applied to some papers using JEMs as exposure assessment, it is not generally used or even considered in occupational epidemiology. Even though the consideration and potential application of QBA can be very important in this field as policy recommendations and decision making in e.g., work safety are made based on, among others, occupational epidemiological studies. | |
