THE ASSESSMENT OF VALUES IN MEDICAL DECISION MAKING.

INTRODUCTION

This article addresses the importance of and challenges for the assessment of values for medical decision making. The emphasis on patient autonomy in today's health care system including mandates to educate patients about their rights to participate in advance care planning entails a responsibility to support patients in making "rational decisions that reflect their personal values" (Bruhn 1991:200). Edelstein [this issue] has addressed the challenges for assessment of a patient's capacity to make a rational decision for him- or herself. Similarly, we are faced with the challenge of how best to aid patients, families, and health care providers to assess, clarify, and/or communicate the values that underlie health care decisions or the advance directives aimed to guide future care decisions. Here, values refer to the beliefs--and the relative importance of different beliefs--that guide a person's behaviors across various situations (Schwartz and Bilsky 1987). Although the assessment of values is relevant for understanding everyday and long-term care decisions, this article focuses on the assessment of values relevant for acute or end-of-life medical treatment decisions.

The assessment of health care values in clinical settings is important for several reasons. The most general argument rests on the premise that individuals differ in their "judgments about the value of life in different states of health" (Llewellyn-Thomas et al. 1988:395). In facing complex medical decisions, individual patients often need support in clarifying their values as they weigh risks and benefits of treatment alternatives (e.g., O'Connor et al. 1999). Cultural, religious, and cohort influences as well as particular life experiences, health status, and emotional factors (Caralis et al. 1993; Garrett et al. 1993) may influence a person's treatment preferences. For example, there are significant cultural and cohort differences in beliefs regarding the extent to which the patient should be aware of his/her prognosis, the role of the individual versus family as the central decision-making unit, the extent of trust in the health care system, whether miracles can occur, and wishes for life-prolonging treatments (e.g., Eleazer et al. 1996; Hauser et al. 1997; High 1988; Kapp 1991; Klessig 1992). Individual differences in treatment preferences are not, however, readily predicted by demographic or psychosocial variables (e.g., Danis et al. 1988; Lo, McLeod, and Saika 1986; Mold et al. 1994). Within-group differences are important; for example, a study of hospitalized patients over the age of 80 found wide variation in health values (Tsevat et al. 1998), suggesting the importance of individualized assessment.

A second argument for values assessment is its potentially important role in advance care planning. The current emphasis on advance directives (living wills, health care proxies) as tools to promote patient self-determination during times of incapacity has not necessarily lead to improved communication about end-of-life care. Living-will documents, in which an individual may specify treatment preferences, may be confusing to patients and have limited reliability and validity (Hoffman, Zimmerman, and Tompkins 1996; Wetle 1994). The widely publicized SUPPORT study (Teno, Lynn, et al. 1997; Teno, Licks, et al. 1997) documented that even if patients had completed advance directives, treating physicians were often not aware of them, general instructions were difficult to translate to specific circumstances, and the care provided was frequently inconsistent with written preferences. Many authors have argued that it may be less helpful to ask people what types of treatments they would want (e.g., respirator, tube feeding, surgery; e.g., Emanuel and Emanuel 1989) than to ask people on what basis they would make certain decisions, or why? (Ditto et al. 1996; Forrow 1994; Gibson 1990; Sehgal et al. 1992). Knowing an individual's underlying values and concerns, through a process of values assessment, may be helpful for guiding treatment planning across a range of (potentially unforeseen) circumstances.

Health care proxies, appointed to make decisions on behalf of patients in the event of incapacitation, ideally make decisions based on the standard of "substituted judgment," that is, making a decision reflecting the values and preferences of the patient. However, many studies have questioned the ability of surrogates to match, beyond chance, the treatment preferences stated by patients (Hare, Pratt, and Nelson 1992; Ouslander, Tymchuk, and Rahbar 1989; Seckler et al. 1991; Suhl et al. 1994; Tomlinson et al. 1990; Uhlmann, Pearlman, and Cain 1988; Zweibel and Cassel 1989). Several studies do suggest that prior discussion between patient and proxy may help proxies to better predict patience preferences (Suhl et al. 1994; Sulmasy, Haller, and Terry 1994). Facilitation of such discussions between patients and proxies as well as care providers would be one use for a values assessment tool. A number of "values histories" interviews are now being published and distributed as tools to help families in health care planning (e.g., AARP 1996; Doukas and McCullough 1991; Gibson 1990).

Finally, values assessment is critical for the resolution of ethical dilemmas and, more generally, in negotiating decisions that involve more than one party. Most health care decisions (especially those with more immediate and serious implications) occur in an interpersonal context and thus entail the negotiation of values of different players (patients, families or friends, various health care providers). Even the nature of the physician-patient relationship is based upon the balance of input between the patient's and the physician's values (Emanuel and Emanuel 1992; Pellissier and Venta 1996; Quill and Brody 1996). Ethical dilemmas arise when conflicts of values arise between parties (Doolittle and Herrick 1992). Further, questions about a patient's competency frequently arise when the patient's wishes differ from those of the family or health care team; evaluations of decision-making competency aim to determine whether a patient has impaired capacity for reasoning or whether there is a legitimate conflict in values between the parties. In these situations, understanding the patient's and others' values, the consistency of the patient's expressed values with his/her personal history, and the stability of stated values over time are all important aspects of an assessment to resolve ethical dilemmas. As argued by Moye [this issue], values assessment is a critical component of a competency evaluation. We need to be open to the possibility that elders with cognitive limitations affecting decision-making abilities may still be able to express deeply held values and beliefs.

This article aims to review: (1) methods used to date for health care values assessment; (2) challenges for the measurement of health care values, with illustrations of "lessons learned" through the process of developing and piloting a Health Care Values Survey; and (3) directions for future research. Although formidable challenges remain for objective assessment of a construct as complicated as values pertaining to serious health care decisions, this review will suggest that it is a task well worth pursuing.