Bayesian Methods for Identifying the Most Harmful Medication Errors



This package implements two distinct but related statistical approaches to the problem of identifying the combinations of medication error characteristics that are more likely to result in harm: 1) Bayesian hierarchical models with optimal Bayesian ranking on the log odds of harm, and 2) an empirical Bayes model that estimates the ratio of the observed count of harm to the count that would be expected if error characteristics and harm were independent. In addition, for the Bayesian hierarchical model, the package provides functions to assess the sensitivity of results to different specifications of the random effects distributions.


Package: mederrRank
Type: Package
Version: 0.0.6
Date: 2011-05-03
License: GPL 2 or greater
LazyLoad: yes

The package is loaded with the usual library(mederrRank) command. The most important functions are bhm.mcmc, bhm.resample and mixnegbinom.em.


Sergio Venturini,

Jessica A. Myers


Myers, J. A., Venturini, S., Dominici, F. and Morlock, L. (2011), "Random Effects Models for Identifying the Most Harmful Medication Errors in a Large, Voluntary Reporting Database". Technical Report.

See Also

bayes.rank, bhm.mcmc, bhm.resample, mixnegbinom.em.

Want to suggest features or report bugs for Use the GitHub issue tracker.