A central but unmet challenge in health care delivery is to increase the probability that the care patients receive near the end of their life is consistent with their goals, values, and preferences. Providing patient-centered care at the end of life is challenging. In their final days, nearly a third of older Americans need critical decisions made regarding the use or nonuse of life-sustaining interventions, but the patients themselves cannot participate in those decisions. Although this observation highlights the promise of advance directives (ADs), to date ADs have not delivered on this promise. This article provides a new framework, based in behavioral economic theory, that may explain the current failures of ADs and point to potential solutions. Specifically, it discusses how five well-described cognitive biases that pervade human decision making (affective forecasting errors, optimism bias, present-biased preferences, focusing effects, and default options) may account for deficiencies in the uptake, efficacy, and patient-centeredness of ADs. The text suggests potential solutions in need of evaluation, discusses metrics for assessing these interventions' benefits, and considers alternatives to the approaches advocated.
Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.