"I think the most famous existential crisis in Oncology is when a patient asks to be put on some new treatment still in trials, after it's been deemed his mortality is inevitable. There is a reason for studies to be randomized, double-blinded, etc., and it's to advance the science. If he or she were in such a study, I would be unwavering in following the rules. If he or she were not in the study, I would offer to write a letter to the institution and it's Investigational Review Board for a special provision for this patient, with the understanding that his or her response to therapy could not be part of the study. If they denied the request, there's nothing more I could do in this respect."

Evidence-based medicine is Gaussian in its distribution, and the further away from the mean the patients wander, the more art you will need to put into your practice. Although there may be a point at which you may choose an 'outlier' of therapy on that bell curve and say, 'What harm can it do?' (especially for a terminal patient), you still have a responsibility to the science. This sentiment is part of the professionalism the interviewer wants to see.

"I practice according to the guidelines of my specialty, and when these blur in complicated cases, I feel I am free to adjust on-the-fly, as long as I can document a solid rationale to my decisions and an appreciation of the risks vs benefits."