Lies, Damn’d Lies, and Statistics: Is it possible to communicate cancer risk accurately?
15 October 2015 - 18:30-20:00
Lecture: Steve John, Lecturer, History and Philosophy of Science, University of Cambridge
Comment: Deborah Ruddy, Consultant in Clinical Genetics, Guy’s and St. Thomas’ Hospital
Henriette Raphael House, Function Room, Guy’s Campus
(See this event also on Facebook)
The fact that different ways of presenting the same risk information can have very different effects on patients’ choices raises an ethical challenge: how, if at all, should medical professionals use such “perlocutionary knowledge”? In this paper, I suggest that reflection on the nature of risk estimates poses an even harder problem: which reference classes should physicians use to calculate patients’ risks in the first place? I then sketch one answer to this challenge, based on an account of rational gambling. Unfortunately, this answer seems to conflict with much current medical practice. In the final part of the paper, I try to justify current practice. I conclude with some remarks on the implications of my argument both for clinical ethics and for the philosophy of science.