Improving the application of quantitative methods in validation work

Abstract

The acceptance of quantitative methods in in vitro studies has seen a steady increase over the last 5 years. ECVAM's Biostatistics Task Force report in the mid-1990s (Holzhütter, H.-G., Archer, G., Dami, N., Lovall, D.P., Saltelli, A., Sjöström, M., 1996. Recommendations for the application of biostatistical methods during the development and validation of alternative toxicological methods. ECVAM Biostatistics Task Force Report 1. ATLA 24, 511–530) laid out clear guidelines for biologists in this field. Nevertheless, the application of more formal approaches to experimental design and statistics is still far from complete in many studies. Furthermore, the regulatory paradigm often does not encourage the use of standard statistical techniques; for example, lowest detectable concentration levels in assays are defined in international guidelines as a simple multiple of standard deviation, rather than tests which are accepted in the statistical field as more reliable and more powerful. ECVAM is currently defining a new approach to facilitate the improved acceptance of these methods into the development and validation process. This will cover a review of appropriate quantitative approaches that can be applied in the validation of alternative methods, as well as an educational element for the community as a whole. This paper will discuss some issues of concern, and outline the milestones planned to help facilitate the acceptance of quantitative methods.