Animal testing
A more important contribution to safety assessment was probably my 10-year crusade to get rid of animal-based genotoxicity assays for assessment of medical devices, usually an in vivo micronucleus assay. This was well before the ground swell against animal testing, and might be seen as prescient, but the truth was that medical devices themselves or device extracts could not reach the OECD guideline dosimetry requirements for a valid assay. Thus, there were hundreds of invalid assays, usually using 90 animals per test. Of course they were always negative. I recall starting even earlier to eliminate the carcinogenicity tests, first their occurrence and then the option for performing them under any circumstances. As practiced, they were even farther removed from the requirements for any sort of scientifically reasonable carcinogenicity assessments, based not only on dosimetry but also on the numbers of animals used. I remember reviewing one test on a sum of 5 animals. The use of animals for medical devices was an incredible waste with no benefit.