You are here:
RSCABS: An R package for performing the Rao-Scott Adjusted Cochran-Armitage trend test By Slices
Citation:
Swintek, J. RSCABS: An R package for performing the Rao-Scott Adjusted Cochran-Armitage trend test By Slices. U.S. EPA Office of Research and Development, Washington, DC, 2016.
Impact/Purpose:
RSCABS[3] (Rao-Scott adjusted Cochran-Armitage trend test By Slices) is a modification to the Rao-Scott[5] adjusted Cochran-Armitage trend test[1, 2] that allows for testing at each individual severity score often seen in histopathological data. The test was originally developed and implemented in SAS? by John Green(1) as part of the Medaka Extended One Generation Reproduction Test[4] (MEOGRT) with the purpose of testing the effects of endocrine disruptors on histopathological endpoints. The RSCABS package allows for easy use of the RSCABS anal- ysis with the capability of using both command line and graphical user interface (GUI) driven operations. The RSCABS analysis is specificity designed to analyze histopathological results from stan- dard toxicology experiments, for example the MEOGRT. These experiments typically have a number of organisms (e.g. medaka) contained in various amounts in different holding apparatuses (e.g. fish tanks). Multiples of each holding apparatuses are exposed to a either an experimental control (plain water) or one of several different concentrations of a chemical. At the end of the experiment several histopathological endpoints are evaluated on every organism. A severity score is assigned to every histopathological endpoint, which is typically an integer ranging from 0 (no effect) to 5 (an incredibly severe effect). A greater severity score indicates a more severe effect but the differences in severity scores are not consistent. For example the difference in severity of a score of 2 to a score of 1 is not the same as the change in severity moving from a score of 5 to of 4. Thus, even though severity scores have an order to them, they must be treated as categorical variables.
Description:
RSCABS[3] (Rao-Scott adjusted Cochran-Armitage trend test By Slices) is a modification to the Rao-Scott[5] adjusted Cochran-Armitage trend test[1, 2] that allows for testing at each individual severity score often seen in histopathological data. The test was originally developed and implemented in SAS? by John Green(1) as part of the Medaka Extended One Generation Reproduction Test[4] (MEOGRT) with the purpose of testing the effects of endocrine disruptors on histopathological endpoints. The RSCABS package allows for easy use of the RSCABS anal- ysis with the capability of using both command line and graphical user interface (GUI) driven operations. The RSCABS analysis is specificity designed to analyze histopathological results from stan- dard toxicology experiments, for example the MEOGRT. These experiments typically have a number of organisms (e.g. medaka) contained in various amounts in different holding apparatuses (e.g. fish tanks). Multiples of each holding apparatuses are exposed to a either an experimental control (plain water) or one of several different concentrations of a chemical. At the end of the experiment several histopathological endpoints are evaluated on every organism. A severity score is assigned to every histopathological endpoint, which is typically an integer ranging from 0 (no effect) to 5 (an incredibly severe effect). A greater severity score indicates a more severe effect but the differences in severity scores are not consistent. For example the difference in severity of a score of 2 to a score of 1 is not the same as the change in severity moving from a score of 5 to of 4. Thus, even though severity scores have an order to them, they must be treated as categorical variables.
