Quantitative bias analysis for mismeasured variables in health research: a review of software tools
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Mismeasurement (measurement error or misclassification) can cause bias or loss of power. However, sensitivity analyses (e.g. using quantitative bias analysis, QBA) are rarely used.
Methods
We reviewed software tools for QBA for mismeasurement in health research identified by searching Web of Science, the CRAN archive, and the IDEAS/RePEc software components database. Tools were included if they were purpose-built, had documentation and were applicable to epidemiological research.
Results
16 freely available software tools for QBA were identified, accessible via R and online web tools. The tools handle various types of mismeasurement, including classical measurement error and binary misclassification. Only one software tool handles misclassification of categorical variables, and few tackle non-classical measurement error.
Conclusions
Efforts should be made to create tools that can assess multiple mismeasurement scenarios simultaneously, to increase the clarity of documentation for existing tools, and provide tutorials for their usage.
