Science Inventory

Categorical regression dose-response modeling


Davis, Allen AND Jeff Gift. Categorical regression dose-response modeling. Society of Toxicology Annual Meeting, Baltimore, MD, March 10 - 14, 2019.


This will be a training session for toxicologists/biologists/risk assessors interested in learning how to use EPA's CatReg software.


The goal of this training is to provide participants with training on the use of the U.S. EPA’s Categorical Regression soft¬ware (CatReg) and its application to risk assessment. Categorical regression fits mathematical models to toxicity data that have been assigned ordinal severity categories (i.e., minimal, mild, or marked effects) and that can be associated with up to two explanatory variables corresponding to exposure conditions, usually concentration and duration. CatReg calculates the probabilities of observing the different severity categories over the continuum of the explanatory variables describing exposure conditions. The categorization of observed responses allows the expression of dichotomous, continuous, and descriptive data in terms of response severity and supports the analysis of data from single studies or across multiple studies. CatReg can also estimate the lower confidence limit on the dose associated with a given severity probability and exposure duration for the purpose of identifying points of departures in human health risk assessments. Additionally, the meta-analytical capability of CatReg allows for the filtering or stratification of data in order to determine statistically significant differences in response between sexes, strains, and/or species. A recently developed graphical user interface for CatReg greatly increases the efficiency with which users can perform categorical regression analyses and will be the primary focus of this training. Disclaimer: The views expressed in this abstract are those of the authors and do not necessarily reflect the views or policies of the U.S. EPA.



Record Details:

Product Published Date: 05/11/2018
Record Last Revised: 05/11/2018
OMB Category: Other
Record ID: 340687