In a clinical trial, there are hazards to the rights of the participants and to the safety of the participants. There are also hazards to the trial itself--to its completion and to its reliability. The risks emerging from these hazards are complex, and the framing of these risks within the constraints of the ethical principles, I believe, is demonstrated in an institutional review board's (IRB's) determination of what is "reasonable" with respect to costs and benefits given the unique characteristics of each investigator-trial-participant grouping.
The Belmont Report of the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research outlined the ethical principles--respect for persons, beneficence, and justice--which serve as the moral foundation for protecting human research participants ( see http://ohsr.od.nih.gov/guidelines/belmont.html ). In addition, Benjamin Freedman later introduced the principle of equipoise as an ethical principle of IRBs (see http://content.nejm.org/cgi/content/abstract/317/3/141 ).
"Respect for persons" demands attention to informed consent, maximization of choice, protection of privacy and confidentiality, and the protection of vulnerable populations. "Beneficence" involves examining the experimental design, evaluating the risks and benefits of the research, minimizing the risks of research participation, and investigating the qualifications of the investigators. "Justice" requires attention to participant recruitment, the equitable distribution of the costs and benefits of research, and the inclusion and exclusion criteria for selecting participants. And, in "clinical equipoise," the IRB decides whether or not an arm of a clinical trial is justified based on whether there is genuine uncertainty among the expert medical community about the comparative therapeutic value to each arm of the clinical trial.