lodi is a package that implements censored likelihood multiple imputation (CLMI) for single pollutant models with exposure biomarkers below their respective detection limits. Additionally, implementations for standard methods such as single imputation with a constant and complete-case analysis are provided, although we do not recommend these methods for datasets with a relatively high percent below their respective detection limits (say >25%).

You can learn more about how to use CLMI by working through the example provided in browseVignettes("lodi").


lodi requires rlang >= 0.3.0 to be installed, so you may want to update rlang before installing.


Development version

To get a bug fix, or use a feature from the development version, you can install lodi from GitHub.

# install.packages("devtools")
devtools::install_github("umich-cphds/lodi", build_opts = c())

Getting help

If you encounter a clear bug, please file a minimal reproducible example on github. For questions, please email Jonathan Boss at bossjona@umich.edu.


Boss J, Mukherjee B, Ferguson KK, et al. Estimating outcome-exposure associations when exposure biomarker detection limits vary across batches. Epidemiology. 2019;30(5):746-755. 10.1097/EDE.0000000000001052