Should Zelen pre-randomised consent designs be used in some neonatal trials?

My aim is to suggest that there is a case for using a randomised consent design in some neonatal trials. As an example I use the trials of extracorporeal membrane oxygenation (ECMO) in neonates suffering pulmonary hypertension. In some trials the process of obtaining consent has the potential to harm the subject, for example, by disappointing those who end in the control group and by creating additional anxiety at times of acute illness. An example of such were the trials of extracorporeal membrane oxygenation (ECMO) in neonates suffering pulmonary hypertension. Pre-randomised consent could avoid or lessen these harms. However, a number of ethical objections are made to these research designs. They involve denial of information, using people, denial of choice, and 'overselling' of allocated treatment. Furthermore, they are the wrong response; better communication might be the answer, for example. I argue that these objections are not completely persuasive. However, they are enough to suggest caution in the use of such designs.