An Integrated Benchmark Dose for Acute and Short-Term Exposure using EPA's CatReg Program
Kenneth Brown
KBinc, Chapel Hill
NC kbinc@mindspring.com

Integrated Risk Analysis (IRA) is a term used for joint analysis from multiple studies, either statistically combining results from individual studies or pooling data after suitable evaluation of data quality and testing for statistical heterogeneity. It is intended to imply a level of qualitative and statistical rigor not always associated with use of the terms "meta-analysis" and "harmonization". The objective of IRA is to utilize as much data as reasonable, without compromising relevant qualitative or statistical considerations. EPA's CatReg program was originally designed for categorical regression, but it also provides the statistical tools for IRA when the objective is dose-duration response assessment leading to estimates of toxicity markers, such as the benchmark dose (usually the ED10), or other effective doses levels ("concentration" may be substituted for "dose"). An advantage of the integrated benchmark dose is utilization of data from multiple studies (when available), and estimation of what might be called a benchmark dose-duration, i.e., benchmark doses for different durations of dosing. Such estimates also lead to time-scaling, which refers to the relative affect of dose level and duration on response, an important consideration in acute and short-term exposures. The concept is illustrated with experiments of rats and mice exposed to hydrogen sulfide.