We weigh the presumed benefts of routinely searching all research scans for incidental fndings against its substantial risks, including false-positive and false-negative fndings, and the possibility of triggering unnecessary, costly evaluations and perhaps harmful treatments. We argue that routinely searching for IFs may not maximize benefts and minimize risks to participants.

How far does a researcher’s responsibility extend when an incidental finding is identified? Balancing pertinent ethical principles such as beneficence, respect for persons, and duty to rescue is not always straightforward, particularly in neuroimaging research where empirical data that might help guide decision making are lacking. We conducted a systematic survey of perceptions and preferences of 396 investigators, research participants, and Institutional Review Board members at our institution. Using the partial entrustment model as described by Richardson, we argue that our (...) data supports universal reading by a neuroradiologist of all research MRI scans for incidental findings and providing full disclosure to all participants. (shrink)