Behaving in a way congruent with one’s expected role is the motive for “role-based decisions”  (March, 1994).  A woman might think that a loving granddaughter is expected to do everything to keep her grandfather alive as long as possible.  Therefore she should choose aggressive care, even though its probability of success is remote and it might cause substantial discomfort to the patient.  Note that normative theories require that the option with the highest expected utility should be selected.  Merely multiply the probability of each outcome by its utility.  That product—expected utility—should be the basis for making a rational choice.  By following this procedure one will select the option which maximizes one’s best interests.  “Role-based decisions” do not require the calculation of expected utility.  Thus if the woman is trying to adhere to the role of a loving granddaughter, providing the woman with statistics showing that the decision may not maximize expected utility may be completely irrelevant.

Lynn et al., (2000) speculated why providing patients, their physicians, and their surrogate decision makers with accurate prognostic information did not change their collective decision making behavior. Accurate prognostic information is one of the key components of expected utility, so such information should have been extremely beneficial.  However this assumption was predicated on the belief that patients would want to maximize the patient’s expected utility.  If patients or their decision-making surrogates wanted to adhere to certain roles, then maximizing expected utility would not be the goal in these instances. 

Page 16 of 21
Previous Section

Next Section