Statistical Evaluation Methods in Toxicology
What is specific to the statistics in toxicology, and why not just use textbook statistics? The reason is the aim of regulatory toxicology: “be confident in negative results.” By toxicological studies, one would like to prove the harmlessness of new drugs. By means of the so-called proof-of-safety approach, the false-negative error rate (consumer’s risk) is directly controlled. Unfortunately, in most of the statistical textbooks and publications, the alternative proof of the efficacy of new drugs with the direct control of the false-positive error rate is used, denoted in toxicology as proof of hazard. Therefore, in this chapter, the basics of the falsification principle are presented simplistic. The commonly used proof-of-hazard approach is discussed hereinafter, focusing on testing a dose-related trend. Finally, the proof-of-safety methods for selected study types are explained by means of examples.
KeywordsEqual Sample Size Williams Test Contrast Coefficient Regulatory Toxicology Observe Effect Level
- Bretz F, Hothorn LA (2003) Statistical analysis of monotone or non-monotone dose-response data from in vitro toxicological assays. ATLA (Alternat. Lab. Animals) 31(1):81–96Google Scholar
- Chow S-C, Liu J-P (eds) (1998) Design and analysis of animal studies in pharmaceutical development. Marcel Dekker, New YorkGoogle Scholar
- Fairweather WR, Bhattacharya A et al (1998) Biostatistical methodology in carcinogenicity studies. Drug Inf J 32:401–421Google Scholar
- Hothorn LA (2003) Statistics of interlaboratory in vitro toxicological studies. ATLA-Altern Lab Anim 31(Suppl 1):43–63Google Scholar
- US-EPA (2013) Section 13, Test Method, Daphnid, Ceriodaphnia Dubia, survival and reproduction test. http://water.epa.gov/scitech/methods/cwa/wet/upload/2007_07_10_methods_wet_disk3_ctf13.pdf. Accessed 3 March 2013