Discount Functions and the Measurement of Patients' Values

Abstract

This paper discusses the measurement of patients' values for future outcomes and examines some problems clinicians confront when making management decisions that attempt to comply with those values. It presents a model that implies that a patient's preference varies with the passage of time, and that during certain periods of time a patient's values may not be representative of his or her long-term preference. An examination of the attitudes of 18 pregnant women toward avoiding pain and avoiding anesthesia bore out the predictions of this model. The women preferred to avoid using anesthesia during childbirth when asked one month before labor and during early labor; however, during active labor their preferences suddenly shifted toward avoiding pain (p less than 0.05). Their preference shifted again toward avoiding the use of anesthesia (p less than 0.05) when evaluated at one month postpartum. The women's preferences one month before labor were the best predictors of their postpartum preferences. The women's preferences during active labor and transition phase of labor were unrelated to their postpartum preferences. These results suggest that, depending upon when a patient's values are measured, (1) they may not be representative of his or her long-term preference, and (2) their use in decision analyses may not maximize the patient's long-term satisfaction so much as they maximize the probability that the decision will be based on the patient's impulsive response. These results also imply that (3) patients may make inherently less reliable value assessments of abstract outcomes they have never experienced, and (4) the distinction between current and long-term values can create a problem when determining whether people behave rationally.