5.6. Summary

The growing complexity of guideline studies, and the increasing number of measurements required therein, present major challenges to nonclinical scientists and statisticians. Central to defining hazards and subsequently assessing risk, is a clear understanding of the study design and data characteristics. This chapter briefly defines statistical aspects of toxicology studies with examples implemented using SAS. We have described randomization schemes for two commonly used designs in toxicology, namely, the parallel design and Latin square design. We have presented the statistical power evaluation of the heart-rate corrected QT intervals from a large animal toxicology study. We have also discussed the statistical analysis of treatment-related effects on body weights from general toxicology studies. From creating randomization schemes, calculating the power of a test, performing statistical analysis and finally to reporting results in nicely formatted tables, SAS has been an indispensable tool for nonclinical statisticians in the pharmaceutical industry.