Institutional Review Boards have been recently audited in the US, and a great deal of concerns and criticisms appeared in the medical literature and the lay press. Globally, these comments questioned the efficiency of Institutional Review Boards in fulfilling their basic mission, the protection of patients subjected to biomedical research. Vulnerable populations, such as psychiatric and incompetent patients, were particularly identified. In intensive care units, conducting research in critically ill patients unable to give informed consent themselves remains a hot issue. Several possibilities have been proposed and assessed in the recent past: pre-emptive (global) consent, proxy consent from a surrogate, deferred consent, and, more recently, waiver of consent for emergency research. Obviously, the inability to deal with this major obstacle (i.e. obtaining consent from incompetent patients) would lead to the cessation of any research in intensive care units, which is hardly acceptable and certainly unethical. Recent published studies have reported over-mortality in treatment groups, posing new questions: who should be informed of those failures, and when? What is the role of Institutional Review Boards in preventing such an issue? What type of indemnity should be proposed to the family of deceased patients during a trial?