Y-007. Old Data Provides New Dose-Response Curves for Bacillus anthracis

B. Thran1, R. Lee1, S. Taft 2, T. Nichols2, D. McKean2;
1USACHPPM, Aberdeen Proving Ground, MD, 2US EPA NHSRC, Cincinnati, OH.

Statistical analysis of dose-response data for Bacillus anthracis causing inhalation anthrax is not new. However, analysis of the data set housed in the Pathogen Information Catalog (PI CAT) has not been previously performed. The PI CAT contains a large amount of dose-response data for inhalation anthrax that was generated in the 1950’s and 60’s. A strength of the PI CAT is the details associated with the data, because the data was extracted from original research reports, not peer-reviewed journal articles. A careful analysis of the PI CAT culled the data into a working data set which consists of 11,495 guinea pig exposures, from 20 studies, representing 1047 doses ranging from 20 - 2,627,907 spores. Four descriptive study variables, preparation (Prep), chamber size (CS; liters), relative humidity (RH) and temperature (T), were present for most of the studies and, therefore, were examined for possible survival dose differences. Doses corresponding to 99%, 95%, 90% and 50% survival were estimated. A common probit slope analysis was used to estimate the survival dose response curve and relative median potency between levels of study variables. Significant differences with respect to relative median potency were observed for RH (low / normal), CS (50,755 / 115K / 1M) and Prep (dry / wet). Further analyses were conducted for dry and wet preparations separately. With respect to survival doses, differences were observed for RH (low / normal / high) for the dry preparation and CS (115K / 1M) for the wet preparation.