Going from principles to rules in research ethics.

In research ethics there is a canon regarding what ethical rules ought to be followed by investigators vis-à-vis their treatment of subjects and a canon regarding what fundamental ethical principles apply to the endeavor. What I aim to demonstrate here is that several of the rules find no support in the principles. This leaves anyone who would insist that we not abandon those rules in the difficult position of needing to establish that we are nevertheless justified in believing in the validity of the rules. I conclude by arguing that this is not likely to be accomplished. The rules I call into question are the rules requiring: - that studies be designed in a scientifically valid way - that risks to subjects be minimized - that subjects be afforded post-trial access to experimental interventions - that inducements paid to subjects not be counted as a benefit to them - that inducements paid to subjects not be 'undue' - that subjects must remain free to withdraw from the study at any time for any reason without penalty. Both canons, the canon on principles and the canon on rules, are found in the overlap among ethical pronouncements that are themselves canonical: the Nuremberg Code, the Declaration of Helsinki, the Belmont Report, CIOMS's International Ethical Guidelines for Biomedical Research Involving Human Subjects, and NBAC's 2001 report, Ethical Issues in International Research: Clinical Trials in Developing Countries.