Carolyn Pevey. Handbook of Death and Dying. Editor: Clifton D Bryant. Volume 2. Thousand Oaks, CA: Sage Reference, 2003.
Living wills and durable powers of attorney may be understood as nips and tucks in the cultural lag that our rapidly advancing health care technology has produced. They promote individual autonomy (McLaughlin and Braun 1998; Samuels 1996; Pope 1999), they may save scarce medical resources from being spent unnecessarily (Schneiderman et al. 1992; Stephens and Grady 1992; Broadwell et al. 1993; Fins 1997), and they may protect health care providers from lawsuits for wrongful death or unnecessary treatment (Schniederman et al. 1992, Broadwell et al. 1993; support Investigators 1995; Stephens and Grady 1992; Glick 1992). This chapter will review the history of advance directives in the United States, the controversies surrounding advance directives, and data regarding advance directive completion rates and their correlates.
What Are Living Wills?
Although regular wills typically provide information about what to do with a person’s material possessions after death, living wills provide information about what a person wants to have done while still living, but incapacitated, and unable to make his or her own health care wishes known. Discussions of living wills are often framed in the context of the “right-to-die,” but they may also direct health care providers to render, rather than only to withhold, treatment. Although there is not a uniform standard formulation for all 50 states, living wills address whether the person would want to have specific treatments such as cardiopulmonary resuscitation (CPR), artificial feeding, artificial respiration, intravenous fluids, or medications (including antibiotics) in cases of extreme debilitation and incapacitation without hope of a cure. Living wills may include important information for the physician who is deciding how to instruct health care personnel to respond should the patient experience cardiac or respiratory arrest. For example, the living will may be the deciding factor in whether or not a physician completes a “Do Not Resuscitate” (DNR) order for an individual patient, which would bind staff to stand by and allow a death to occur in the event of a heart attack.
Because it may be that the person signing the living will does not mention a specific treatment preference for the particular situation that arises, today’s living wills usually include a provision for a surrogate decision maker, more formally known as the person designated to exercise durable power of attorney for health care.
Durable Power of Attorney for Health Care
Durable power of attorney for health care (DPAHC) is the permission that one person gives another to make his or her health care decisions, in the event the first individual becomes incapacitated. It is “durable” because it is permanent until revoked (Moore and Wooten 1998), and the person named to exercise it is known as a surrogate or proxy. The living will and DPAHC are usually both implied by the term advance directive or advance medical directive, because they are health care instructions given in advance of a situation of incapacitation, although living wills without DPAHC likely still exist.
Role of the Surrogate
The surrogate has three levels of interpretation by which to make decisions. First, the proxies may make “directive” decisions in which the patient’s wishes are easily ascertained by a reading of the advance directive (Alpers and Lo 1999; Collopy 1999). Because it is impossible for people signing living wills to make provisions for every possible circumstance, however, there are often cases in which the proxy must use “substituted judgment.” In these situations, the surrogate must make a decision based on what the surrogate imagines the incapacitated patient would choose, were the patient able to do so (Alpers and Lo 1999), and this decision is based on the surrogate’s knowledge of the patient’s beliefs and values (Collopy 1999). In the absence of either a direct written preference or adequate knowledge of the patient’s beliefs and values as they pertain to a particular health care question, the surrogate must decide on the treatment that is in the patient’s “best interests” (Alpers and Lo 1999; Collopy 1999). Surrogate laws may be especially important for people who do not have a traditional family structure (Jansen and Ross 2000), because physicians will not otherwise easily know to whom questions or information should be directed. Fields and Scout (2001) suggest that the significant others of lesbians and homosexuals in particular may be ignored, and they might find advance care planning laws especially helpful. There have been some problems associated with the use of surrogate decision making, and these will be visited below. Understanding the issues, however, requires an understanding of their history.
Although the first U.S. living will was created in 1967 by a member of the Euthanasia Society of America (Webb 1997), early attempts to pass living will legislation were unpopular (Glick 1992). Because living wills may direct health care providers to withhold or withdraw treatment after which the patient will likely die, they were promoted by pro-euthanasia groups but opposed by Catholic theologians until 1984 (Hays and Glick 1997). The first attempt to pass a living will law in Florida in 1968 failed, as did the Natural Death Act legislation attempted in California in 1974 (Glick 1992). It was not until 1976 that California passed the first U.S. state legislation making living wills legally binding (Hays and Glick 1997).
By unlikely coincidence, 1976 was also the year that Karen Quinlan’s case was decided by a New Jersey court. Although Quinlan (Webb 1997) was not the first person to have her life artificially sustained against the wishes of family members, and there have been other more dramatic cases (such as that of Samuel Linares; see Nelson and Cranford 1989; Lantos, Miles, and Cassel 1989), her case was followed by the popular media with great interest (Webb 1997). This media attention may have been because of the mistaken but widespread belief that Quinlan had fallen into an irreversible coma as a result of drug and alcohol abuse (Webb 1997). For young people who were considering experimenting with mind-altering substances, Webb (1997) argues that Quinlan’s story was a potentially potent cautionary tale.
After realizing that Quinlan’s condition would not improve, her family asked the physician to turn off the machine that was breathing for her (Webb 1997). Although he initially agreed, the hospital attorney later advised against it, and a court battle ensued. Because Quinlan was 21 when she became incapacitated, she was without a legal guardian. In a highly publicized trial, her father was granted legal guardianship, and the court ruled that it would not be homicide to remove her respirator. In 1976, her father was given the right to refuse treatment for her, and the respirator was removed. Quinlan died 9 years later. The many witnesses who testified to having heard her express the wish that she not be maintained on life support were probably instrumental in the court’s final decision. This testimony was important because it helped establish that removing the respirator was the judgment Quinlan would have made had she been able to express herself (Webb 1997).
The 1983 case of Nancy Cruzan in Missouri, however, is the most cited regarding the advance directive legislation because it seems to have been the impetus for the Patient Self Determination Act (PSDA) of 1991 (Glick 1992). In the Cruzan case, the U.S. Supreme Court in 1990 ruled that before a treatment is refused on behalf of an incapacitated patient, there must be “clear and convincing evidence” that the family member or physician is acting as the patient would have chosen (Webb 1997:162). A year later, the PSDA was passed with the intent of enhancing patient autonomy (Silverman et al. 1995). It requires that all health care providers receiving Medicare or Medicaid funds inform patients about advance directives (Wolf et al. 1991). This information must be given in writing and include a description of the patients’ legal rights concerning end-of-life care and decision making in that state, and health care providers must educate not only their staff but also the community at large; documentation of advance directives must be placed in patient charts as well (Wolf et al. 1991). The clear-and-convincing standard coupled with the PSDA is likely driving the interest of U.S. health care providers in attaining higher completion rates of advance directives.
Completion Rates and their Correlates
Scholars estimate that between 5% and 25% of American adults have completed advance directives (Pope 1999; VandeCreek, Frankowski, and Johnson 1995). For health care providers who are required by law to promote the use of advance directives and who may feel dependent on them for protection against lawsuits (Glick 1992), these are not promising figures.
Ideally, a person completes a living will after a thoughtful discussion with his or her primary physician with whom the patient has an established and meaningful relationship of mutual respect and trust. Many researchers believe that doctor-patient discussion is a necessary prerequisite for any type of informed consent that living wills imply (Samuels 1996; The SUPPORT Investigators 1995; Lo and Steinbrook 1991; Elpern, Yellen, and Burton 1993; Singer et al. 1998; Caralis, Davis, Wright, and Marcial 1993; Pope 1999). Mansell et al. (1999) show that patient-physician communication does seem to increase completion rates. Those patients who had discussions with their doctors were twice as likely as those who hadn’t to complete advance directives in Mansell et al.’s (1999) study. In addition, such discussions will increase the likelihood that the patient’s doctor will know about the advance directive on admission and therefore be more likely to recognize and implement it (Morrison et al. 1995).
Although the support investigators (1995) found that providing physicians with information on patient prognoses and end-of-life preferences did not produce higher completion rates, others have found that completion rates do increase when physician reminders are placed on charts (Walker, Mandell, and Tsevat 1999). In this study, researchers put notes on selected patient charts just before a clinic visit, telling the doctor that the person’s lab values suggested a decline in condition and that advance directives were important; the notes asked the physicians to discuss and document advance directives during the coming visit (Walker et al. 1999). Although the bulk of the research does seem to support the idea that physician-patient discussion increases completion rates, some people do not learn about advance directives in the context of a relationship with a physician. Caralis et al. (1993), for example, found that the people they studied were more likely to have learned about advance directives from the media (19%) than from the physician (10%). Only 14% of their respondents had a living will, even though 83% them had spoken with their doctors about their medical conditions.
Adolescents may be comfortable discussing advance directives, but in one study (McAliley et al. 2000), most of them thought living wills were important only for people over age 70 to have. And indeed, age was the only variable predicting advance directive completion among the health care workers studied by Orlander in 1999, in which the oldest quartile in the sample was three times more likely to complete than was the youngest quartile. VandeCreek et al. (1995) found that age 60 and up was correlated with 5% to 15% completion rates, but that age 80 and up was correlated with 46% completion rates. Among the exclusively male patients studied by Mansell et al. (1999), age, severity of illness, and previous illness of spouse predicted completion of advance directives. In this study, marital status also predicted completion, with widowers more likely to have completed advance directives but married men one-third as likely to have completed one.
Education of the patient, staff, and community is required by the PSDA, and education may be an important factor in advance directive completion rates. Rubin et al. (1994) showed that exposure to an educational pamphlet and DPAHC form resulted in an 18.5% completion rate versus a rate of 0.4% in a control group while Caralis et al. (1993) found that people with relatively less education were less likely to have completed advance directives. In addition, Silverman et al. (1995) found that those scoring lower on education quizzes about advance directives were less likely to have completed them and that education was an independent predictor for a history of discussions of preparation of advance directives. Lack of education alone does not account for low completion rates, however; staff members at a teaching hospital, a presumably well-educated subset, had lower completion rates than the general population (Orlander 1999). In addition, Caralis et al. (1993:199) found that although lower educational levels were associated with less knowledge of advance directives, those with less education were more likely to have formally named a proxy than were those with more education. Hallenbeck and Goldstein (1999) report that those people with more education who are also white and affluent are most likely to have completed advance directives.
Silverman et al. (1995) report that African Americans scored lower than European Americans on an education quiz and suggest that their lower rates of completion may be a function of lower relative rates of education. Other researchers, however, find a definite correlation between race and ethnicity and attitudes about end-of-life issues, which could affect completion rates. Whites, for example, are more likely to believe in euthanasia than are African Americans (Caralis et al. 1993), and African Americans and Hispanics are less likely to complete advance directives than whites or Asians (Rubin et al. 1994). Caralis et al. (1993) found that African Americans and Hispanics are more likely to want treatment if terminally ill and report that African Americans in particular are more likely to feel that they will get less care if they sign advance directives. Hispanics and African Americans were more likely than non-Hispanic whites to want dialysis, antibiotics, artificial nutrition, and surgery (Caralis et al. 1993). Although a majority of all respondents disagreed with the idea of doctors making all the decisions alone, Hispanics were more likely to say they would accept the doctor’s opinion than were the other groups (Caralis et al. 1993:158). Hispanics find the impact of illness on the family very important, and more so than for non-Hispanic white or African Americans, but they still want the doctor to rule if there is no living will and the family is in disagreement (Caralis et al. 1993).
Some cultures may be more group or family oriented and less interested in individual autonomy than mainstream U.S. culture, and these differences may influence relative rates of advance directive completion. In a study by Blackhall et al. (1995), 53% of Korean Americans and 35% of Mexican Americans said it would be best to keep a terminal diagnosis from the patient, but only 13% of European Americans and 12% of African Americans agreed. Mexican Americans, Korean Americans, and Japanese Americans are less likely than non-Hispanic whites and African Americans to believe that the dying patient should be informed of his or her prognosis (Blackhall et al. 1995; Kalish and Reynolds 1976). Kalish and Reynolds (1976) reported that in their study, 51% of African Americans did not believe that people should be informed of a terminal condition, but 71% would wantto be told of their own terminal prognosis. Similarly, Caralis et al. (1993) found that although most Hispanic respondents wanted to be informed if they had a terminal prognosis, they might also pressure a physician to keep such news from a loved one.
Japanese doctors are more likely than Japanese American doctors to aggressively treat the terminally ill and less likely to tell the patient a terminal prognosis (Asai, Fukuhara, and Lo 1995). This may not be a pressing issue in Japan, where families usually defer to the physician and wrongful death lawsuits are rare, but in the United States, a subculture in which collective family interests take priority over those of the individual could make advance directive planning between the patient and physician more problematic (Asai et al. 1995).
In some ethnic groups, the emphasis on preservation of harmony may be a value at variance with the mainstream, which may result in fewer discussions among patients and providers. Although doctors are respected authority figures, they are expected to make decisions that benefit the collective rather than the individual needs in many Asian groups (McLaughlin and Braun 1998). Among some groups, families want to be involved in major decision making, but at the same time, they may be so interested in the maintenance of harmony that they are unable to say “no” to a health care provider (McLaughlin and Braun 1998). This could make it difficult to clearly grasp the wishes of a proxy.
Some religions may find standard living wills problematic for a variety of reasons as well. Under Islamic law, for example, standard living wills are inappropriate because they assume that life should not involve suffering and contain instructions to withhold treatments, food, and water, which is unlawful for the Muslim ill (Ebrahim 2000). Ebrahim (2000) suggests that, in keeping with Islamic law, Muslims name a surrogate rather than completing a standard living will or, instead, that they can write alternate living wills requesting only that no overzealous treatment take place. Practitioners of Shinto, as well, may find discussion of living wills difficult because in Shinto, merely talking about death may cause contamination (McLaughlin and Braun 1998).
Although these cultural, ethnic, and religious barriers to efficient communication regarding end-of-life care are interesting and a worthy area of study, some researchers also caution that health care providers must be on guard for assuming that such group data apply to individuals (Kalish and Reynolds 1976; Hallenbeck and Goldstein 1999). Caregivers are advised to ask open-ended questions to learn the assumptions and expectations of the people they serve and to pay attention to body language and lack of response for cues that there may be value differences (McLaughlin et al. and Braun 1998). In addition, Blackhall et al. (1995) advise physicians to ask patients in advance whether they would want news of a terminal prognosis.
Low completion rates are not the only problem with advance directives; there also may be problems with adherence to them, as well as problems with their underlying assumptions.
It may be that advance directives do not travel with the patient from one health care setting to another or from home to the health care facility (Morrison et al. 1995). In one study, only 26% of patients with advance directives had them recognized by the health care facility to which they had been transferred, and 14% of those were not honored (Morrison et al. 1995). Documentation was showing improvement, however; 13% in 1991, 25% in 1992, and 31% percent in 1993 of the advance directives in this study were recognized by the health care facility. But in 74% of the admissions in which the patient had an advance directive and was unable to make decisions, the advance directive was not recognized, and the person who acted as proxy was not the person named in the advance directive (Morrison et al. 1995:480) When advance directives are recognized, however, they are usually followed (Morrison et al. 1995:482).
Another problem that may arise is that emergency medical service (EMS) personnel do not honor advance directives because their standard of care is to “resuscitate aggressively unless a doctor on the scene directs otherwise” (Sabatino 1999:298). This means that if someone is expected to die but has opted out of hospital care and decided to die at home, the person may be at risk for aggressive and intrusive “life-saving” treatments if the EMS system is mistakenly activated. EMS organizations have responded to this situation by encouraging state laws to support DNR orders for the home, and as of 1999, 42 states had DNR orders that when signed by physicians, EMS may legally honor in out-of-hospital settings.
Doctors themselves are not always comfortable with living will implementation, especially when it entails withdrawal of treatment (Gordon and Levitt 1996), and do not always do what the patient wants (Madson 1993).
Ely et al. (1992) received 439 questionnaires completed by physicians that gave the case history of an 89-year-old man who had had a severe stroke but whose prognosis was still uncertain. All questionnaires stated that the patient had previously completed a living will. Half of the physicians were told that the living will specifically requested that tube feedings be withheld, and half were told that the living will was not specific. The physicians were then asked whether or not they would recommend inserting a feeding tube. Although most physicians who received information about a specific living will were opposed to tube feedings, when they were asked whether they would change their recommendation if the family members pushed for the tube feeding, 66% of those previously recommending no feeding tube changed their opinions to favor the tube placement, contrary to stated patient wishes. Of those physicians who initially favored tube feedings, 42% changed their recommendations if the family opposed them. In other words, even in the presence of a living will, the doctors reported that they would recommend treatments favored by the family over treatments favored by the patients (Ely et al. 1992).
Some of those surveyed reported that legal issues were their primary guiding force, and others were uncomfortable with the timing of the decision or the certainty of the patient’s mental state. It is clear, however, that the wishes of the family are often more important than the wishes of the patient. Yet in spite of this preference for family wishes, health care professionals sometimes view surrogate decision makers with suspicion (Collopy 1999). The source of this suspicion probably originated in concerns about conflicts of interest. For example, the person who is chosen as a trusted proxy may also be a beneficiary of the death of the one who is terminally ill. On the other hand, because aggressive and prolonged treatments require hospital facilities, the hospital and staff generally can be considered beneficiaries of protreatment decisions as well or, in cases where the patient is uninsured or underinsured, beneficiaries of hastened deaths.
Surrogates Don’t Always Obey
Hospitals, unlike most individuals, have experts to advise them in the legal ramifications of treatment decisions, and hospitals are subject to intense scrutiny by accrediting organizations. But laws concerning surrogates are intended to regulate private and unsupervised individuals (Collopy 1999). In some states, DPAHCs are valid only if they are written on a specific form (Lee and Berry 1991), and health care providers are not obligated to obey the person exercising durable power of attorney for health care in all states (Moore and Wootton 1998). Indeed, state laws are more likely to allow for proceedings against bad proxies, but not against health care providers who ignore proxies; the primary concern of state law regarding proxies is that patients be protected “from proxies who overreach their authority” (Collopy 1999: 40). The existence of one paper arguing that a proxy needs full disclosure of a patient’s condition before making a choice suggests that the distrust care providers feel toward proxies is widespread (Jansen and Ross 2000).
In addition, proxies may also have their own interests that are not necessarily financial. For example, Lee and Berry (1991) discuss the case report of a man who was unable to make his own decisions but who had previously informed his physician that he would not want CPR if he ever experienced a heart attack; also, his wife had promised him that she would never allow him to be hooked up to machines in a hopeless situation. When the man became seriously ill without hope of recovery, the wife agreed to have the doctor sign a DNR but later canceled it. She then reinstated it and then rescinded it again. Evidently, the alternate proxy, who had been like a son to her, and on whom she depended for many of her needs, had been influencing the wife to cancel the DNR because he didn’t think it was moral to withhold CPR. The wife had felt compelled to do as he wished rather than as her husband had wished (Lee and Berry 1991). But if proxies are nervous or unsure of patient wishes in specific cases, health care providers may become suspicious that the proxies don’t really speak for the patient, which can cause increasing distance between the health care providers and the already isolated proxy (Collopy 1999).
Surrogates are obligated to use substituted judgment if no specific instructions were left (Alpers and Lo 1999; Collopy 1999). In other words, they must ask themselves what the patient would choose for himself or herself in the given situation were he or she able to do so. The proxies make this decision based on personal knowledge of the patient’s values and wishes. A deeper level of criticism of surrogate decision making addresses the “best-interests” standard (Collopy 1999). This is the perspective proxies must use when there is no directive covering the situation and when proxies do not know what the patient would choose. But this standard may conflict with the actual purpose of the durable power of attorney for health care. The New York State Task Force on Life and the Law (Collopy 1999) claims that the best-interests standard must be objective, but Collopy (1999) argues that people may choose their surrogate decision maker in hopes that the proxy will not be objective:
From the patients’ perspective, a principal motive for naming a proxy would seem to be the desire to protect oneself, when incapacitated, from treatment decisions made by nameless others. Yet, here, the (moral) consensus requires that proxies, when they have no explicit instructions or grounds for substituted judgment, disallow their own personal values and follow a best interests standard—in short, that they hand patients over to the judgment of “most people,” to nameless others. (P. 42)
Several scholars claim that because end-of-life issues are too important to relegate to signatures and checklists (Caralis et al. 1993; Lo and Steinbrook 1991; Silverman et al. 1995; Pope 1999), research on advance directives focuses too heavily on completion rates and attempts at improving them rather than on the quality of the documents in expressing patient wishes. Lo and Steinbrook (1991) argue that the formal rationality fostered by the focus on form completion is the wrong emphasis and that, instead, we should apply substantive rationality to questions of end-of-life preferences. First and foremost, they argue, the patient must be truly informed (Lo and Steinbrook 1991). Pope (1999) argues that because most people do not know enough about medicine to properly complete them, the PSDA is actually promoting underinformed and uninformed advance directives (Pope 1999). Caralis et al. (1993) claim that the emphasis on form over substance is a “Mirandizing” of patient rights and includes a subtle coercion to sign the papers. Stone (1998) also notes that coercion may be involved in health care providers’ emphasis on getting documents signed. Silverman et al. (1999) contends that the mere signing of a document without an appropriate dialogue between the patient and the physician results in patient decisions that are not carefully investigated. It is clear that meaningful dialogue with one’s physician is considered by many to be the most important step in formulating living wills.
More serious than the bureaucratic ritualism that may surround the completion of living wills, however, is the suggestion that people who sign living wills when already in health care facilities are at risk of having the documents automatically interpreted by underinformed or confused staff members as requests for DNR orders (Stone 1998). These DNR orders, Stone (1998) argues, may go into effect even if the person who initially approved the living will is not dying so that the person does not get any treatment for what may be treatable conditions. Because living wills and durable powers of attorney are not synonymous with requests for physician-assisted suicide, these concerns expressed around the possible overemphasis on formal rationality are more than stylistic quibbles.
Autonomy Versus Interdependence
Although advance care planning focuses on the completion of legal forms and ideally happens in the context of doctor-patient communication, it is generally assumed to be based on the ethical principle of autonomy.
In interviews of 48 dialysis patients, however, Singer et al. (1998) found that advance care planning is also based on personal relationships and relieving burdens on others. They argue that from the perspective of the patient, rather than the health care provider, the focus of advance care planning is on the social process, and that it occurs in the context of discussions with loved ones. They found that 69% of those surveyed involved their loved ones in their discussions of end-of-life issues, and the choice of who to involve in these discussions was determined by emotional closeness or intimacy. In fact, many felt that the topic was too personal to discuss with their physicians. Family and community considerations were also very important among the 36 elderly Israelis interviewed by Leichtentritt and Rettig (1999:352), extending even to considerations for the impact their end-of-life decisions may have on future generations.
Studies have shown that many people prefer that others make serious health care decisions for them if they are incapacitated (Leichtentritt and Rettig 1999; Singer et al. 1998; High 1988). Particularly, the most importance is often placed on having the family help them decide what to do (High 1988), and the elderly favor group decision making.
Although at least one study shows that family opinion is the most important factor that physicians consider when making decisions for incapacitated patients (Ely et al. 1992), advance directives may actually serve to undermine families by making it possible for the physician to ignore them (Lo and Steinbrook 1991). Because families deliver 75% of long-term care, ignoring their wishes may be unwise and counterproductive (Hardwig 1993).
The importance of autonomy in understanding end-of-life issues may be misguided, not only because many people do not wish to have complete autonomy but also because the reality is that, in varying degrees, we are deeply embedded in social networks. According to Durkheim (1965), the most basic categories of science, such as time and space, come to us from the social milieu. But today many social scientists agree that even seemingly autonomous individuals have had their personalities and much of their interior life shaped, at least in part, by interaction with others. If our sociality is an inescapable part of who we are, then the efforts to insist on autonomy may be counterproductive. In addition, there is some disagreement about whether the supposedly autonomous individual who plans his or her future medical care is actually the same person who later falls ill.
Berghmans (1998), for example, argues that advance directives that give permission for the signer to be used in nontherapeutic dementia research are probably unethical. Berghmans argues that dramatic changes in personal identity come with dementia, and although a person before the onset of dementia may be willing to submit to nontherapeutic research later, it is unlikely possible to know whether that same person with dementia would agree, as that may not actually be the same person. Indeed, even in persons without dementia, there is some evidence that opinions change. A 5-year study in Rhode Island, for example, showed that 9.7% of people for whom EMS was activated revoked their DNR orders (Sabatino 1999). On the other hand, Kuczewski (1994) contends that the self is so socially embedded that not even a loss of an internal sense of continuity of self, as in dementia, for example, can change the basic identity of a person.
Many of the critiques presented here may be resolved. The increasing rates of advance care document recognition in the Morrison et al. study (1995), for example, is evidence of a growing awareness of the importance of living wills. On the other hand, the question of whether the person who signed a living will was truly informed is unlikely to abate because formal legislation requiring substantive rationality is irrational in essence. Although the controversies discussed in this chapter will likely continue, it is doubtful that they will slow the trend for planning one’s health care in advance.