| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Rachel A. Pruchno, PhD, Director of Research, Center for Aging, University of Medicine & Dentistry of New Jersey, School of Osteopathic Medicine, 42 E. Laurel Rd., Suite 2300, Stratford, NJ 08084. E-mail: pruchnra{at}umdnj.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: End of life • Autonomy • Decision making • End-stage renal disease • Substituted judgments • Proxy
Nevertheless, most people feel that they can rely on close family members to make health care decisions for them (Hayley, Cassel, Snyder, & Rudberg, 1996; High, 1988, 1994), and most health care providers turn to family members when critical decisions have to be made and the individuals involved are unable to make them (High & Rowles, 1995; Keating, Moss, Sorkin, & Paris, 1994). The President's Commission for the Study of Ethics Problems in Medicine and Biomedical and Behavioral Research (1983) has endorsed substituted judgment as a means for promoting patient autonomy. Substituted judgment calls for the surrogate to make decisions in a manner that approximates the patient's wishes. It assumes that the surrogate understands patient preferences and correctly represents the wishes of the patient. In this article we examine the extent to which spouses of patients with end-stage renal disease use substituted judgment as they make hypothetical decisions about the patients' likelihood of continuing dialysis, and we raise questions about whether there may be alternative principles guiding their decision-making strategies.
Although the substituted judgment standard may be ethically sound, it is unclear whether people actually use this standard in practice. Several studies have shown that, when faced with hypothetical decisions about life-sustaining medical care, family members and physicians do not consistently predict a patient's preferences at levels of accuracy beyond those expected by chance alone (Druley et al., 1993; Fried, Bradley, & Towle, 2003; Gerety, Chiodo, Kanten, Tuley, & Cornell, 1993; Hare, Pratt, & Nelson, 1992; Suhl, Simons, Reedy, & Garrick, 1994; Sulmasy et al., 1998; Uhlmann, Pearlman, & Cain, 1988). Rather, family proxies consistently tend to overestimate the frequency with which patients would like to receive treatment (Uhlmann et al.). Moreover, Ditto and colleagues (2001) reported that, even with structured interventions to encourage the use of instructional or values-based advance directives and discussion between patients and family members about treatment wishes, proxy decision makers remain inaccurate in providing decisions that are similar to those of the patient.
One explanation for why family members may not be using substituted judgment when making health care decisions is that they have difficulty separating their own values and beliefs from those of others. Citing a principle variously termed assumed similarity (Cronbach, 1955), attributive or social projection (Holmes, 1968), and false consensus effect (Marks & Miller, 1987), Fagerlin, Ditto, Danks, Houts, and Smucker (2001) suggested that people tend to overestimate the extent to which their own opinions and choices are shared by others and, as a result, attribute their own preferences to others. In their study of undergraduate students and their parents who engaged in a guided discussion about the hypothetical life-sustaining-treatment preferences that parents would make, Fagerlin and colleagues found that students' predictions of parents' preferences were strongly and consistently related to the students' own preferences for parents' medical treatment, and not to their parents' actual preferences, suggesting that projection, rather than accurate substituted judgment, guided their predictions. In another study, Fagerlin, Ditto, Hawkins, Schneider, and Smucker (2002) replicated these results with older adults and their chosen surrogates, where issues of life-sustaining treatment were presumably more realistic and salient. Surrogates' accuracy, however, was only somewhat better than student accuracy, suggesting that projection prevails even for actual surrogates and their predictions of an older population of patients. Moreover, this pattern was observed even when discussion of treatment preferences happened minutes before the prediction task, suggesting that projection occurred despite other available information (Ditto et al., 2001; Krueger & Clement, 1994).
Understanding the extent to which proxies use substituted judgment and projection is particularly important for persons with end-stage renal disease who are on hemodialysis. Hemodialysis is a common form of life-maintaining treatment, with approximately 300,000 patients in the United States, and about 95,000 new patients each year (U.S. Renal Data System, 2002). Most dialysis patients (more than 90% in 1999) have or develop significant comorbid conditions, including diabetes, congestive heart failure, ischemic heart disease, peripheral vascular disease, and cerebrovascular disease (Merkus et al., 1999; Shidler, Peterson, & Kimmel, 1998; Valderrabano, Jofre, & Lopez-Gomez, 2001). Although the actual number of people discontinuing dialysis prior to death is unclear, recent data indicate that between 17% and 22% of people with end-stage renal disease who die discontinue dialysis just prior to their death (Cohen, Germain, & Poppel, 2003; Kliger & Finkelstein, 2003).
The prevalence of comorbid conditions, the option of discontinuing dialysis, and the high mortality rate in this population (23% annually; U.S. Renal Data System, 2002) raise the salience of surrogate decision making for this population. Moreover, the tendency for dialysis patients to lose decision-making capacity near the end of life as a result of dementia or other organic brain syndromes (Neu & Kjellstrand, 1986) makes it critical to understand the extent to which surrogate decision makers use substituted judgment or projection to guide their decision-making strategies.
The analyses that follow contrast preferences for continued dialysis treatment expressed by patients with end-stage renal disease with those expressed by their spouses who were presented with a series of hypothetical situations. They address the question of whether substituted judgment or projection affect surrogate decisions, as they seek to explain the mechanisms underlying surrogate decision making. Finally, the analyses examine the influence of communication between spouses, the extent to which the patient feels the spouse understands his or her preferences, and the degree to which the patient believes that the spouse will act in ways that are consistent with the patient's preferences on the accuracy of substituted judgments.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Couples were eligible for inclusion in the study if patients (a) had end-stage renal disease; (b) were on hemodialysis for at least 6 months; (c) were at least 55 years old; and (d) were married and cohabitating with their spouse for at least 5 years. Study eligibility required both patients and their spouses to be English speaking and to be free of cognitive, hearing, and speech impairments that would preclude their ability to answer questions on the phone. These characteristics were based on the screener's evaluation of the patient and spouse during the screening session. During the screening process, the recruiters determined that 28 patients or spouses had insufficient cognitive ability to participate and 16 patients or spouses had hearing impairments that would interfere with their ability to participate in the study. We did not enroll these couples in the study. In addition, one patient and one spouse each had more than four wrong answers to the Short Portable Mental Status Questionnaire (which was administered at the beginning of the interview), and staff withdrew these two couples from the study.
Respondents were interviewed in person (n = 37) if they lived within 1 hr of Boston, MA, and over the phone (n = 254) if they lived farther from Boston. Preliminary analyses (available on request from R. Pruchno) revealed that there were no significant differences in means or patterns of response as a function of mode of interview. Patients and spouses were interviewed within a mean of 5.6 days (SD = 8.1) of one another. As a way to maintain confidentiality, patients and spouses within each couple were never interviewed by the same interviewer.
Recruitment efforts yielded a sample of 315 couples. As the analyses that follow address the spouses' role as health care proxy, 24 couples in which the patient did not indicate that his or her spouse would be the health care proxy were deleted, leaving a sample of 291 couples. The average age for patients and spouses was 69.9 (SD = 8.05) and 67.8 (SD = 8.98) years, respectively. The majority (73.9%) of patients were male. Patients included 85.2% Caucasians, 10.7% African Americans, 1% Native Americans, and 3% who indicated other or mixed races. Race distribution of spouses was nearly identical. The mean length of patients' treatment for end-stage renal disease was 6.75 years (SD = 6.2). Patients had been on hemodialysis for a mean of 4.68 years (SD = 3.3); 4.1% were on hemodialysis for less than 1 year, 12.0% for 12 months to 2 years, 21.7% for 25 months to 3 years, 26.3% for 37 months to 5 years, and 35.9% for more than 5 years. Patients and spouses had been married for an average of 41.5 years (SD = 13.1).
Measures
To assess patient and spouse preferences for continuation of dialysis, we had respondents presented with a series of hypothetical situations and asked about the likelihood of the patient continuing on dialysis given each condition. The hypothetical situations inquired about were conditions likely to characterize people with end-stage renal disease. They included those in which the patient developed mild and severe stroke, mild and severe dementia, permanent coma, and terminal illness, both with and without pain.
To ensure a common understanding of each condition, interviewers read descriptions developed by Singer and colleagues (1995) to the respondents prior to asking questions about the condition. For example, for the severe stroke scenario, respondents were told "A stroke means that the patient would have damage to the brain causing permanent physical disability such as paralysis. She or he would have severe paralysis on one side of her or his body, be unable to walk, and would need to be in a chair or bed. She or he would not have meaningful conversations, be unable to carry out routine daily activities, need a feeding tube for nourishment, and would be able to live at home with someone caring for her or him day and night; otherwise she or he would probably need to be cared for in a chronic care hospital." Respondents were asked to assume, for the purposes of the research study, that these problems would stay the same for the rest of the patient's life; they would not get better.
After hearing each description, patients were asked the following question: "If you had [condition], how likely would you be to want to continue your dialysis treatments? Would you say very likely (5), somewhat likely (4), not sure (3), somewhat unlikely (2), or very unlikely (1)?" Responses to this question are referred to as patient preference. For spouses, each description was followed by two questions. The first was, "If the patient had [condition], how likely would you be to want his or her dialysis treatments continued? Would you say very likely (5), somewhat likely (4), not sure (3), somewhat unlikely (2), or very unlikely (1)?" Responses to this question are referred to as spouse's preference. The second question posed to the spouse was, "If patient had [condition], how likely do you believe he or she would be to want his or her dialysis treatments continued? Would you say very likely (5), somewhat likely (4), not sure (3), somewhat unlikely (2), or very unlikely (1)?" Responses to this question are referred to as substituted judgment. To examine the most clinically significant differences in patient preference, spouse preference, and substituted judgment, we collapsed responses for each variable to the following three levels: (0) unlikely (previously responses 1 and 2), (0.5) not sure (previously response 3), and (1) likely (previously responses 4 and 5). A similar procedure was used by Fagerlin and colleagues (2001). This scoring system also allows one to easily discerning the percentage of people who prefer to continue dialysis.
We had patients asked the following question to determine whether they had discussed their preferences for medical treatment with their spouses (communication): "Have you talked with your spouse about your wishes regarding medical treatment to maintain your life if you become seriously ill and cannot speak for yourself?" Responses were either yes (1) or no (2). We assessed the extent to which patients believed that their spouse understood their wishes about medical treatment by use of the following question: "How well do you believe your spouse understands your wishes about medical treatment if you got very sick? Would you say poor understanding (1), fair understanding (2), good understanding (3), very good understanding (4), or excellent understanding (5)?" We assessed the extent to which patients felt confident in their spouses' intentions to follow their preferences by use of the following item: "How much confidence do you have that your spouse will make medical treatment decisions according to your wishes, if the time comes that you are not able to speak for yourself?" (Responses included 1, not much confidence; 2, a fair amount of confidence; and 3, a lot of confidence).
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
= 0.01–0.14). Relationships between patient preferences and substituted judgment were modest ( = 0.05–0.20). However, relationships between spousal preferences and spouses' substituted judgment were large and significant ( = 0.40–0.66). In every scenario, the relationships between spousal preference and spouses' substituted judgment were larger than those between substituted judgment and patient preference.
|
|
Prior Communication
Two hundred thirty-four (80.4%) patients indicated that they had talked with their spouse about their treatment preferences (communication). Hierarchical regression analyses estimated spouses' substituted judgments from patients' and spouses' preferences, communication, and product terms representing the interaction of communication with patients' and spouses' preferences. These models examined the extent to which prior communication of preferences affects spousal preference and substituted judgment. Analyses revealed that the interaction coefficients were insignificant in six of the seven models, ps >.27, suggesting that substituted judgments were not improved by prior communication. The one significant interaction, for the mild dementia scenario, suggested weaker projection effects when patients indicated prior communication:
|
|
Patients' Felt Understanding and Confidence in Spouse
As we expected, patients tended to believe that their spouses had an excellent understanding of their wishes (M = 4.49; theoretical range = 1–5), and they had confidence that their spouse would make medical decisions according to their wishes (M = 2.90; theoretical range = 1–3). To test whether patients' felt understanding and confidence affected the accuracy of spouses' substituted judgment, we regressed spouses' substituted judgments onto patients' and spouses' preferences, felt understanding, confidence, and product terms representing interactions of patients' and spouses' preferences with felt understanding and confidence. In six of the seven scenarios, the interaction coefficients were insignificant; p >.13. For the severe stroke scenario, patients' confidence moderated the effect of patients' preferences, ß = –.13, p <.02, and the effect of spouses' preferences, ß =.13, p <.05, on substituted judgment, but these effects were in a counterintuitive direction (decreased accuracy and increased projection effects at higher levels of confidence). Thus, although most patients felt that their spouses understood their wishes and had confidence in their spouses' intention on following their wishes, these beliefs did not improve spouses' substituted judgments.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This study's finding that spouses are more in favor of dialysis continuation than are patients, although consistent with other studies (Fried et al., 2003; Hare et al., 1992; Seckler, Meier, Mulvihill, & Cammer Paris, 1991), has important clinical implications. It suggests that if the desires of spouses are followed, patients would receive more life-sustaining care, and less palliative care, than patients themselves would like. Therefore, an argument can be made that these inaccurate substitute judgments both threaten respect for patient autonomy and raise questions about the appropriate use of financial and health care resources (Teno, Fisher, Hamel, Coppola, & Dawson, 2002).
These findings, along with those of other studies, help to confirm the inability of surrogate decision makers to accurately represent the wishes of the people whose perspectives they are presumed to represent. Even those patients with much at stake who had talked with their spouses about their wishes regarding medical treatment at the end of life, those who believed that their spouses understood these wishes, and those who had confidence that their spouse would make medical treatments according to their wishes had spouses whose substituted judgments were in error.
Results from this study suggest the need for better and more structured communication between patients and their surrogate decision makers. Holley and colleagues (1999) have suggested that surrogates may be inaccurate because many dialysis patients do not adequately discuss the various conditions under which dialysis should be discontinued. Decisions about terminating or continuing dialysis are extremely complex, including a host of biological and psychosocial issues. Developing a structured, ongoing dialogue between patients and spouses that is sensitive to these issues and recognizes the complexity of the decision represents an important next step in improving the ability of surrogates to make decisions that preserve patient autonomy. Encouraging patients and surrogates to discuss treatment preferences at different points in the illness is especially important, as it is likely that these preferences are fluid. Moreover, encouraging patients and surrogates to engage in discussions about the principles underlying patients' decisions (e.g., what are the goals of care? when is death preferable to illness?) may be more helpful to improving surrogate decision making than knowing the manifest decisions. Furthermore, an intervention that would help surrogate decision makers distinguish their own desires from those of the patient would be critical.
A word of caution must be inserted, however, because data regarding communication between patient and spouse were gathered only from the patients. Communication is a two-way street; hence a comprehensive understanding of communication requires reports from multiple sources. Future research in this area would be advised to inquire about communication from both the perspective of the patient and the spouse, enabling measures of spouse concurrence to be examined as well as individual reports.
Health care providers can play an important role in facilitating communication and understanding these preferences. Cohen and colleagues (2003) present helpful guidelines that medical staff may use in doing so, involving explicit discussions between the patient, surrogate, and physicians. These discussions may help physicians and surrogates gain a sense of the patient's preferences and values regarding life-sustaining treatment, which can then be utilized if the patient loses decision-making capacity and a decision must be made (Cavalieri, 2001; Gerety et al., 1993; Hare et al., 1992). Currently, many physicians do not adequately discuss advance planning with their patients (Cavalieri, Latif, Ciesielski, Ciervo, & Forman, 2002; Doukas, Gorenflo, & Coughlin, 1991; Gamble, McDonald, & Lichstein, 1991; Stefanacci, Cavalieri, Flynn, & Pomerantz, 1994), and many patients feel more comfortable discussing treatment preferences with family members than with physicians (Hines et al., 1999). These barriers must be overcome before clear and informed communication about end-of-life options can take place.
Several limitations are important to note. First, the study design is cross-sectional and thus does not address issues of causation. Future longitudinal research should examine whether the accuracy of substituted judgments changes over time (see Danis, Garrett, Harris, & Patrick, 1994; Emanuel, Emanuel, Stoeckle, Hummel, & Barry, 1994; High & Rowles, 1995). Second, the interview was structured such that assessment of spouses' surrogate judgments always occurred immediately following assessment of their own preferences. Therefore, the substituted judgments may have been due to the salience of spouses' own preferences when making their predictions, although Fagerlin and colleagues (2001) found evidence of projection when asking surrogates to make predictions before stating their preferences. Third, it is not clear whether actual behavior would mimic responses to the hypothetical scenarios. Many social psychological studies suggest that people are generally poor at predicting their future preferences and behavior (Osberg & Shrauger, 1986). Fourth, although patients participating in this study were cognitively intact, it is unknown whether results would generalize to those situations in which cognition is not intact. Finally, results are limited to an understanding of the conditions under which patients and their spouses would discontinue dialysis. They provide no information about preferences for other life-sustaining treatments, such as feeding tubes or ventilators. In fact, research by Singer and colleagues (1995) indicates that decision making varies as a function of treatment.
These findings raise serious questions about whether autonomy should continue to be the principal foundation guiding the way in which decisions are made at the end of life. This ethical debate has been an evolving one (Blustein, 1993; Callahan, 2002; Cavalieri, 2001; Cohen et al., 2003; Hardwig, 1990; High, 1994; Meier & Morrison, 2002; President's Commission, 1983). Drought and Koenig (2002) contend that the contribution of bioethics to clinical care at the end of life deserves critical scrutiny. Moreover, it is likely that embracing ethical principles other than autonomy as we struggle to make decisions at the end of life may obviate the concern about the accuracy of substituted judgments. There is some evidence, for example, that patients may actually prefer that others make end-of-life decisions for them. Research by High, Hines and colleagues (1999), and Terry and associates (1999) finds that many competent patients often defer to the wishes of their family members in making their treatment decisions and may in fact withhold completing an advance directive in order to give family members greater latitude (High). Once such individuals are incompetent, then, their self-determination (President's Commission) would not be compromised by spouses' making decisions on the basis of considerations other than patient autonomy.
In addition, many treatment decisions are not based on perceived preferences of the incapacitated patient, but instead adhere to the principal of beneficence (Cavalieri, 2001), in which decisions are based on consensual perceptions of the patient's best interest (President's Commission, 1983) shared by surrogates, other family members, and physicians (Cohen et al., 2003). Indeed, many dialysis patients want their doctors to use their judgment of what's best for them, rather than an advance directive, in deciding on life-sustaining treatment should they become mentally incompetent (Sehgal et al., 1992). These barriers notwithstanding, the issue of accuracy in substituted judgments becomes irrelevant to medical decision making when other standards are implemented.
Future research should be designed to remedy some of the limitations of this study and also to test whether ethical principles other than autonomy underlie decision making at the end of life. More information is needed to discern the factors underlying surrogates' tendency to exaggerate patients' desires for life-sustaining treatment. Withdrawal of life-prolonging treatment may represent to spouses the premature relinquishment of hope, directly conflicting with perceived obligations to honor the sanctity of life. Or perhaps those who have witnessed patients' abilities to continually adapt to a downward health trajectory have a unique understanding of the limits to their tolerances for adversity. Research aimed at understanding the ways in which the moral principles other than patient autonomy affect decision making at the end of life will be critical as our society struggles to determine whose preferences matter and when.
|
Footnotes |
|---|
1 Initiatives on Aging, Boston College, MA.
2 Associate Provost and Professor, Boston University Medical Center, MA.E. Lemay, Jr. is currently at the Department of Psychology, Carnegie Mellon University, Pittsburgh, PA; L. Field is currently with the Human Research Quality Improvement Program, Partners Healthcare System, in Boston; and N. Levinsky is now deceased.
Decision Editor: Linda S. Noelker, PhD
Received for publication February 10, 2005. Accepted for publication June 10, 2005.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. A. Pruchno, M. J. Rovine, F. Cartwright, and M. Wilson-Genderson Stability and Change in Patient Preferences and Spouse Substituted Judgments Regarding Dialysis Continuation J. Gerontol. B. Psychol. Sci. Soc. Sci., March 1, 2008; 63(2): S81 - S91. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
