Targeted Minimum Loss-Based Estimation (TMLE) for Survival Analysis with Competing Risks

Authors: David Benkeser and Nima Hejazi


Description

survtmle is an R package designed to use targeted minimum loss-based estimation (TMLE) to compute covariate-adjusted marginal cumulative incidence estimates in right-censored survival settings with and without competing risks. The estimates can leverage ensemble machine learning via the SuperLearner package.


Installation

For standard use, we recommend installing the package from CRAN via

install.packages("survtmle")

You can install a stable release of survtmle from GitHub via devtools with:

devtools::install_github("benkeser/survtmle")

Issues

If you encounter any bugs or have any specific feature requests, please file an issue.


Example

This minimal example shows how to use survtmle to obtain cumulative incidence estimates with a very simple, simulated data set.


Contributions

Contributions are very welcome. Interested contributors can consult our contribution guidelines prior to submitting a pull request.


Citation

After using the survtmle R package, please cite both of the following:

    @manual{benkeser2017survtmle,
      author = {Benkeser, David C and Hejazi, Nima S},
      title = {{survtmle}: Targeted Minimum Loss-Based Estimation for
               Survival Analysis in {R}},
      year  = {2017},
      howpublished = {\url{https://github.com/benkeser/survtmle}},
      url = {http://dx.doi.org/10.5281/zenodo.835868},
      doi = {10.5281/zenodo.835868}
    }

    @article{benkeser2017improved,
      author = {Benkeser, David C and Carone, Marco and Gilbert, Peter B},
      title = {Improved estimation of the cumulative incidence of rare
               outcomes},
      journal = {Statistics in Medicine},
      publisher = {Wiley-Blackwell},
      year  = {2017},
      doi = {10.1002/sim.7337}
    }