From Ruth Faden, Tom Beauchamp, and Nancy Kass in the NEJM:
One major question is whether informed consent should always be required for randomized comparative-effectiveness studies, particularly studies conducted in a learning health care system. Our answer to this question is no. It will often be unethical to go forward with CER [Comparative Effectiveness Research] in which patients are randomly assigned to different interventions without their written, prospective, informed consent. However, in a mature learning health care system with ethically robust oversight policies and practices, some randomized CER studies may justifiably proceed with a streamlined consent process and others may not require patient consent at all.
They argue that creating health care systems that continually improve themselves through applied research will require relaxed informed consent procedures for low risk studies. This is important work from scholars who have thought deeply about medical and research ethics.