CRAN Version Github Version Travis CI

Limit of Detection Multiple Imputation

lodi is a R package that implements censored likelihood multiple imputation (CLMI) for single pollutant models with exposure biomarkers below their respective detection limits. lodi also contains implementations for standard methods such as single imputation with a constant and complete-case analysis, although those methods are primarily designed for comparison with clmi.


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

The package can be installed from CRAN


Or from Github

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

The Github version may contain bug fixes not yet present on CRAN, so if you are experiencing issues, you may want to try the Github version of the package.


Once lodi is installed, you can load up R and type


to learn how to use the method.


If you encounter a bug, please open an issue on the Issues tab on Github or send us an email.


For questions or feedback, please email Jonathan Boss at or Alexander Rix


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