Yes, obviously. But what about new chemicals where testing has been done, without significant observed damage? Is that considered safe after enough observations, or would a single fatality skew your risk-minimizing behavior?