Text Size: A A A
Physician-Assisted Death in the United States: Are the Existing 'Last Resorts' Enough?

*This essay is being provided in full free of charge. To see others, or to subscribe to the Hastings Center Report, register at the bottom of this page.

Although there has been relatively little activity in the last ten years with regard to legal access to physician-assisted death, this fall a citizens’ initiative in the state of Washington is proposing an Oregon-style law that would allow legal access to potentially lethal medication for terminally ill patients, subject to defined safeguards. As the rhetoric inevitably heats up, this seems like a good time to review areas of progress in palliative and end-of-life care and to consider whether laws like the one on the table in Washington are either needed or desirable.

Several things are clear: (1) Palliative care and hospice have improved in terms of access and delivery, and they remain the standards of care for addressing the suffering of seriously ill patients. (2) Despite state-of-the-art palliative measures, there will remain a relatively small number of patients whose suffering is insufficiently relieved. (3) Several “last resort” options, including aggressive pain management, forgoing life-sustaining therapies, voluntarily stopping eating and drinking, and sedation to unconsciousness to relieve otherwise intractable suffering, could address many of these cases.1 The question remains as to whether physician-assisted death—that is, providing terminally ill patients with a potentially lethal prescription that they could ingest on their own to relieve otherwise intractable suffering by directly hastening death—should be one of these last resort options.

My own answer to this last question is a cautious “yes”: open access to physician-assisted death, subject to the safeguards of excellent palliative care and access to other last resort options, gives patients an important additional option, and the benefits of legalization outweigh the risks.

Progress in the Last Decade

Perhaps the most dramatic sign of progress has been the coming of age of the palliative care movement, which allows fully informed decision-making and the provision of treatments to maximize quality of life for all seriously ill patients alongside any and all disease-directed treatment that patients want to continue. Almost all major medical centers now have inpatient palliative care consultation services, and similar services are spreading into community hospitals. Consultation possibilities are also spreading into the outpatient and home settings, although the gaps between need and availability are much wider in these contexts. The American Board of Medical Specialties has recently given palliative care the status of being a board-certified subspecialty, and fellowship programs are sprouting up across the country.

There remain serious challenges. There are not enough skilled palliative care clinicians to meet the growing needs, and reimbursement for palliative care services—which rely heavily on counseling and coordination of care rather than expensive procedures—remains problematic. Similar gaps exist in providing basic palliative care education for all clinicians who care for seriously ill patients, and in generating an evidence base for the field. Nonetheless, palliative care seems to have passed the “tipping point” as a field; most patients and families can find the treatments that they need regardless of their stage of disease.

When patients become terminally ill, access to palliative care is facilitated by the proliferation of hospice programs. Hospice remains our premiere program to provide palliative care for terminally ill patients who are willing to forgo further treatment of their underlying disease, as it provides, pays for, and coordinates comprehensive quality-of-life-oriented treatments for terminally ill patients. Hospice has expanded considerably in the last ten years, primarily in two domains: the inclusion of terminally ill patients with diseases other than cancer—congestive heart failure, dementia, and chronic lung disease, for example—and the ability to supplement the palliative aspects of care for terminally ill patients who reside in skilled nursing homes.


Despite this progress, the majority of patients who die in the United States are never transitioned to hospice, mostly because of a requirement that once they are in hospice they will forgo disease-directed therapy. Some larger hospices are experimenting with loosening these restrictions through “bridge” programs that let patients continue active treatments that are important to them and have some potential for helping while at the same time receiving the full benefits of hospice. Since hospices are paid on a per diem basis (on average, about $135 per day), only the very large, affluent hospices can afford to offer expensive disease-directed therapy at the same time that they are providing and paying for comprehensive palliative care, but there is much more flexibility and willingness to experiment with these areas now than ten years ago.

Significant progress has also been made in the articulation and provision of last resort options for patients whose suffering becomes unacceptable to them despite state-of-the-art palliative care. Ten years ago, the problem of intractable suffering was often not acknowledged or was blamed on the clinician (“They don’t have adequate expertise”) or the patients (“They want too much control over their fate”). Now, it is much more widely acknowledged that even with the best possible palliative care, there will always be a small percentage of cases where suffering sometimes becomes unacceptably severe, and that clinicians are obligated to treat these circumstances as a palliative care emergency that requires consultation and committed efforts to respond in the most helpful, least harmful way.

Other Last Resort Options

When unacceptable suffering persists for terminally ill patients despite state-of-the-art palliative care, four options have emerged as last resort possibilities. The first two are widely accepted and relatively uncontroversial:

Right to intensive pain and symptom management. Pain can almost always be sufficiently relieved without any significant risk of hastening death, but there will be a few cases where pain accelerates severely toward the very end of life and pain medicines must be proportionately increased, bringing an increasing risk of hastening death. Patients and families, having given their informed consent, should be able to count on their doctors’ willingness to take these risks as part of their commitment to relieve suffering, and medical ethics and law stand firmly behind this commitment.

Consider a case like this: a patient with advanced bladder cancer is nearing death and must make daily tradeoffs between adequate pain relief and sedation. She reaches a point where she “just wants to be out of pain.” After extensive discussion with patient and family, her dose of opioids is gradually increased until her pain is sufficiently relieved, but after that point she eats and drinks very little and sleeps most of the time. She dies a week and a half later. Although death may have been unintentionally hastened by a small amount, in this process the obligation to relieve her severe suffering warranted taking that risk.

Right to forgo life-sustaining therapy. Medicine has made great strides prolonging life, and a small but significant part of this progress has been the availability of life-sustaining technologies. These technologies no longer just mean ventilators and feeding tubes; they now include radical technologies such as ventricular-assist devices. But while the array of medical choices faced by patients and families has grown more complex, ethics and law remain clear that patients have a right both to forgo such treatments and to stop them once started. Here’s a typical case that turns on this kind of decision: an elderly man fractures his spine falling off a ladder. He initially accepts intubation and surgery in hope that he will recover neurological function, but four weeks later, he has not. He cannot imagine living this way, and after extensive discussions with his medical team and his family, he is mildly sedated and removed from the ventilator. He dies within twenty-four hours.


Of course, some patients suffer unacceptably from symptoms other than pain, and their lives (and deaths) are not directly dependent on life-sustaining technologies. A portion of these patients look for options to escape suffering and even hasten death by means other than physician-assisted death. Two additional possibilities have emerged over the past decade. While clearly more morally complex and less settled than the first two options, these options seem to generate less legal and ethical controversy than physician-assisted death:

Voluntarily stopping eating and drinking. VSED is different from the natural loss of interest in food and drink that occurs when a person is actively dying. It is an informed decision to stop food and fluids while one is still physically capable in an effort to hasten death and escape suffering. The decision is entirely within the patient’s control, but it requires considerable resolve, since the patient generally continues to live for one to two weeks after stopping. Although VSED is not directly “physician-assisted,” it needs to be “physician-supported”: the patient must be assessed at the outset to ensure that he or she is competent and has adequate palliative care, and the physician must help the patient and family to address any unforeseen complications as the process unfolds.

VSED is only an option when the patient retains mental capacity. Suppose, for example, that a patient with amyotrophic lateral sclerosis has become tired of living in the circumstances imposed by his illness. He does not want a feeding tube or a mechanical ventilator, and he feels that he is dying too slowly. He asks about physician-assisted death. After careful assessment of his capacity and motivation, his physician raises the possibility of VSED. Since eating and drinking are already a daily struggle, he and his family welcome this alternative. His family can then come together for his last two weeks, during which he will be supported by home hospice. In the final phase of this process, he may be mildly delirious because of the metabolic changes of dehydration, but usually these and other symptoms can be well managed as part of his hospice support.

Sedation to unconsciousness. This type of sedation is distinct from the proportionate use of sedatives to treat severe anxiety or delirium, which is a relatively common part of terminal care. Instead, it is an explicit decision to render the patient unconscious so that he or she can escape otherwise severe and intractable suffering. Other potentially life-prolonging treatments such as food and fluids are usually discontinued at the same time. There is general agreement on its use as a last resort to relieve severe, intractable suffering in imminently dying patients, but as one gets further away from these circumstances, agreement about its permissibility falls off sharply.2 

Consider an elderly holocaust survivor dying of end-stage congestive heart failure who was transferred to an acute hospice/palliative care unit to manage his terminal shortness of breath. The opioids used to treat his dyspnea made him delirious, and he began having flashbacks from his time in the concentration camps. Usual sedatives did not control his agitation, and family and staff agreed that this suffering was unacceptable, so he was sedated to unconsciousness with phenobarbitol. The dose was adjusted until he appeared to be resting comfortably, but he was completely unresponsive. Artificial hydration and nutrition were not provided, with the family’s consent. His level of sedation was maintained until he died seventy-two hours later.


The Empirical Data

By far the best data about physician-assisted death in the United States come from Oregon, where the practice is reported to the health department and where annual summaries have been prepared every year since legalization. The practice has been remarkably stable over the ten intervening years, accounting for approximately one out of every thousand deaths per year.3 This appears to be a very small number given the amount of controversy surrounding the practice, but one in fifty patients talk to their doctor about it, and one in six talk with family members, suggesting that the availability of such an escape may be much more important to many patients than its actual use.4 The Oregon statute requires that patients be informed of “feasible alternatives,” including hospice and palliative care and other last resort alternatives, and some choose alternatives other than physician-assisted death.

There are also data from Oregon to suggest that the legalization of physician-assisted death enhances rather than undermines other aspects of palliative and end-of-life care. Oregon has one of the highest rates of hospice referral in the nation, and the vast majority of patients who choose physician-assisted death are simultaneously enrolled in hospice. Oregon also has relatively high rates of opioid prescription per capita, and physicians as well as other medical professionals have very high rates of attending training courses in both palliative care and end-of-life medical decision-making. There is also a statewide form, “Physician Orders for Life-Sustaining Treatment” (POLST), for recording a patient’s wishes about cardiopulmonary resuscitation and other potentially life-sustaining therapies. The form has become a model for other states.5 Overall, Oregon appears to be among the leaders in comparison to other states in virtually all aspects of palliative and end-of-life care, including allowing open access to physician-assisted death, subject to safeguards.

The secret practice of physician-assisted death in the rest of the nation is very difficult to study. To admit to participation, a physician has to admit to a crime, and along with any family present, runs the risk of prosecution. On the other hand, there appears to be very little interest in prosecuting such cases providing they are not discovered or flaunted, leading to a “don’t ask, don’t tell” policy that is unpredictable and potentially dangerous. Under this policy, there is no opportunity to get second opinions from experts in palliative care, no documentation, and considerable potential for idiosyncratic responses from clinicians.

In the mid-1990s, I was part of a team that conducted an empirical study of the secret practice, using research techniques that protected anonymity. We found that physician-assisted death and euthanasia accounted for approximately 1 to 2 percent of deaths. Although this appears to be ten to twenty times higher than the rates reported in Oregon, the reporting techniques were so different that the rates are not directly comparable. On the other hand, we know from Oregon that conversations with doctors about these issues are common, and it appears to be much better and safer to have the conversations out in the open rather than in secret.6


Very little is known about the frequency of other last resort practices in the United States. Data from the Netherlands, where all end-of-life practices are regularly studied, suggest that forgoing life-sustaining therapy and prescribing “opioids in large doses” are each reported to account for approximately 20 percent of deaths, which fits with my clinical experience in the United States.7 There are no reliable data about the frequency of voluntarily stopping eating and drinking in the United States, although the practice is thought to be rare. In our two hospice programs in Rochester, New York, where VSED is permitted and supported as a last resort, it accounts for less than 1 percent of deaths. Sedation to unconsciousness to treat otherwise intractable physical symptoms appears to be used very variably in the United States, apparently depending more on the values and practice patterns of the practitioners than of the patients. Reports vary from no deaths to half of all deaths, depending in part on definitions but also on practice patterns.8 At our hospital, where sedation to unconsciousness for treatment of intractable symptoms is subject to guidelines and restrictions that include a mandatory palliative care consult, it accounts for less than 1 percent of deaths.

The Clinical Context

All last resort options, including physician-assisted death, make sense only if excellent palliative care is already being provided. Mandatory palliative care consultation should therefore be a standard safeguard for any and all of these practices. Over the next ten years, medical institutions and professional groups should ensure that all clinicians who care for seriously ill patients are competent in the basics of palliative care and that specialty-based palliative care consultation is available for the more challenging cases.

There is also a need to develop explicit, predictable strategies to respond to difficult clinical situations where patients experience severe suffering despite state-of-the-art palliative care. Many of these patients will benefit from a discussion and exploration of last resort options that may or may not include physician-assisted death. There are two main clinical situations in which this might come up:

Patients who are worried about future suffering and wonder what options would be available to them. This conversation begins with an exploration about hopes, fears, and prior experiences of family and friends. Such patients frequently want to know what options they could have in the future if their suffering becomes unacceptable to them. In response to these inquiries, the clinician should talk to the patient about how he or she approaches such situations and what last resort options could be provided if needed. Many patients are reassured by learning about options other than physician-assisted death and by the willingness of the clinician to explore this domain and to commit to working with them and addressing their suffering throughout the illness until death. They can then be free to spend their remaining time and energy on other important personal and family matters.

Patients who eventually experience suffering that is unacceptable to them. This is a much smaller population than those who are worried about the future, but these patients’ needs can be more challenging. The starting point is always to explore the patient’s suffering in its totality, including why it is now experienced as unacceptable. Part of this assessment is to ensure that standard palliative care is being skillfully applied, and that the request does not emanate from anxiety or depression that might be otherwise addressed. A second opinion by a specialist in palliative care should be obtained. If there are no good alternatives, then the last resort options that are legally available should be explored in the approximate order presented in this paper. Usually, but not always, options other than physician-assisted death will adequately address the patient’s clinical situation and be acceptable to the patient. In the event that no other possibilities are workable or acceptable, physician-assisted death would need to be considered in light of the legal environment (the approach in Oregon will be very different than the rest of the country) and the values of patient, family, and clinician.


Will Physician-Assisted Death Be Necessary?

Some patients will prefer access to physician-assisted death even if the other last resort options are predictably available. Patients who request and eventually act using physician-assisted death in Oregon have a strong interest in controlling their fate, and physician-assisted death puts more choice directly in their hands. However, all the last resort options, including physician-assisted death, are imperfect. Although each addresses some situations particularly well, there are other situations where they would not be as helpful.

For example, voluntarily stopping eating and drinking has the advantage of putting the decision in the patient’s hands, but it requires tremendous discipline not to drink if one is thirsty and capable of drinking, and the duration of the process is too long if symptoms are severe and immediate. On the other hand, medical sedation to unconsciousness may be very frightening to those who value consciousness and being in charge, and there is no way to verify that the sedated patient is not still suffering but unable to report it. Finally, physician-assisted death requires that the patient be physically capable of self-administration and able to swallow a concentrated amount of lethal medication. In addition to these practical issues, any of these options may be morally troubling for patient, family, physician, or staff.

Adding physician-assisted death to the list of last resort options has both risks and benefits. One benefit is that it adds another important possibility for terminally ill patients who experience unacceptable suffering. We should be as responsive as possible to these patients without violating fundamental values, but it is clear that the patient’s values in this context count the most, followed by the family and then the clinician (if the course of action requires the physician’s participation). Most patients will be reassured by the possibility of an escape, and the vast majority will never need to activate that possibility. But some patients will need a way out, and arbitrarily withholding one important option from patients whose options are so limited seems unfair.

The main risk of including physician-assisted death with other last resort options is that it seems to be very polarizing in the United States, where there is wide agreement about palliative care and hospice being the standards of care and also a surprising level of acceptance of the other last resort possibilities. Even the Supreme Court, in its 1997 decision, made it clear that “obtaining medication . . . to alleviate suffering even to the point of causing unconsciousness and hastening death” is legally acceptable.9 Opponents of physician-assisted death may work to further restrict access or even prohibit access to other last resort alternatives as they become more well known and predictably available.

On the Horizon

It seems highly likely that palliative care and hospice will continue to expand in the United States and other Western countries. There are very few places in the health care system where we can simultaneously save money and improve quality, but palliative care and hospice have the potential to do both. On the quality side, improvements in pain and symptom management, more informed medical decision-making, and enhanced patient and family support are core elements of palliative care and hospice. The cost savings will come not from restricting access to expensive treatments and technologies, but from a better informed consent process for patients and families concerning medical treatments with marginal benefit and coordinated care for patients with very complex treatment. Hospice continues to be the gold standard for end-of-life care, but its challenge is to design programs that would allow patients to simultaneously continue some disease-directed therapies in order to serve a wider range of dying patients.10 If these programs can be proven cost-effective, then perhaps the hospice benefit can be expanded, and hospice and palliative care could be better integrated into traditional medical care. Palliative care needs to be part of the standard of care for all seriously ill patients, whether or not they choose to continue disease-directed therapy in any form.


The last resort options other than physician-assisted death must become more standardized, available, and accountable. There is currently too much variation. Some patients may be denied access to them because clinicians or institutions are reluctant to use them, while at the same time, others are given last resort options when more standard palliative measures would have been more appropriate. Better policies and procedures should begin at the national level, with local programs following suit. Fortunately, medicine seems to be moving in this direction. Witness a recently published American Medical Association guideline on sedation to unconsciousness that is consistent with position statements by the American Academy of Hospice and Palliative Medicine and others.11

Similar guidelines and policies are needed for voluntarily stopping eating and drinking. Because these options are intended to be rarely used, all institutions should review their own practices against the position statements of leadership organizations. For example, my institution has a guideline on sedation to unconsciousness for treatment of refractory suffering, and we review every single case in which it is utilized. The state of California is considering a law entitled the California Right to Know End-of-Life Options Act, which requires that patients be given information about both hospice and last resort options, should they request it. The goal of all these initiatives is to improve predictable access and accountability both for hospice/palliative care and for legally permitted last resort options.

In the domain of physician-assisted death, the most pressing change on the horizon is the Oregon-style citizens’ initiative in the state of Washington. Residents of the state are already relatively well educated on the subject, having been through an unsuccessful initiative in 1991, and they are more aware of the Oregon experience, being adjacent geographically and relatively similar demographically. Citizens’ initiatives begin by gathering signatures of support from a large number of residents. If the legally determined threshold is achieved, the initiative is then placed on the ballot during the next election cycle and subject to the popular vote. Not all states sanction citizens’ initiatives, but in my opinion, they are more likely to be successful in legalizing physician-assisted death than legislative approaches, given the relatively high level of public support and the tendency for the issue to get polarized in legislative situations. Nonetheless, legislative processes to legalize physician-assisted death were recently attempted in California, and before that there were near misses in Hawaii and New Hampshire.

Although some proponents of physician-assisted death will find the incremental, state-by-state approach to legalization frustrating, it has some value. For one thing, it gives us time to study the intended and unintended effects of legalization before national implementation. We can simultaneously evaluate the impact of better and more widespread access to hospice and palliative care, and of more predictable and accountable availability of other last resort options. In addition, it keeps the movement a grass-roots one—particularly when it happens through citizens’ initiatives. That may not make the practice of physician-assisted death less controversial, but it can perhaps make the process through which physician-assisted death is legalized less controversial.

 

1. T.E. Quill, B. Lo, and D.W. Brock, “Palliative Options of Last Resort: A Comparison of Voluntary Stopping Eating and Drinking, Terminal Sedation, Physician-Assisted Suicide, and Voluntary Active Euthanasia,” Journal of the American Medical Association 278 (1997): 1099-2004.

2. See F.A. Curlin et al., “Religion, Conscience and Controversial Medical Practices,” New England Journal of Medicine 356 (2007): 593-600.

3. See “Death with Dignity Act, Records and Reports Data on the Act,” http://www.oregon.gov/DHS/ph/pas/.

4. S.W. Tolle et al., “Characteristics and Proportion of Dying Oregonians Who Personally Consider Physician-Assisted Suicide,” Journal of Clinical Ethics 15 (2004): 111-18.

5. “POLST: Physician Orders for Life-Sustaining Treatment Paradigm,” at http://www.ohsu.edu/ethics/polst/.

6. Tolle et al., “Characteristics and Proportion of Dying Oregonians Who Personally Consider Physician-Assisted Suicide.”

7. P.J. van der Maas et al., “Euthanasia, Physician-Assisted Suicide, and Other Medical Practices involving the End of Life in the Netherlands, 1990–1995,” New England Journal of Medicine 335 (1996): 1699-1705; B.D. Onwuteaka-Philipsen et al., “Euthanasia and Other End-of-Life Decisions in the Netherlands in 1990, 1995, and 2001,” Lancet 263 (2003): 395-99.

8. L.A. Jansen and D.P. Sulmasy, “Sedation, Alimentation, Hydration, and Equivocation: Careful Conversation about Care at the End of Life,” Annals of Internal Medicine 136 (2002): 845-49.

9. Washington v. Glucksberg, 117 S.Ct 2258 (1997); Vacco v. Quill, 117 S.Ct. 2293 (1997).

10. A.A. Wright and I.T. Katz, “Letting Go of the Rope—Aggressive Treatment, Hospice Care and Open Access,” New England Journal of Medicine 357 (2007): 324-27.

11. K.B. O’Reilly, “AMA Meeting: AMA OKs Palliative Sedation for Terminally Ill,” American Medical News, July 7, 2008, at http://www.ama-assn.org/amednews/2008/07/07/prsi0707.htm.

*This essay is being provided in full free of charge. To see others, or to subscribe to the Hastings Center Report, register at the bottom of this page.

Although there has been relatively little activity in the last ten years with regard to legal access to physician-assisted death, this fall a citizens’ initiative in the state of Washington is proposing an Oregon-style law that would allow legal access to potentially lethal medication for terminally ill patients, subject to defined safeguards. As the rhetoric inevitably heats up, this seems like a good time to review areas of progress in palliative and end-of-life care and to consider whether laws like the one on the table in Washington are either needed or desirable.

Several things are clear: (1) Palliative care and hospice have improved in terms of access and delivery, and they remain the standards of care for addressing the suffering of seriously ill patients. (2) Despite state-of-the-art palliative measures, there will remain a relatively small number of patients whose suffering is insufficiently relieved. (3) Several “last resort” options, including aggressive pain management, forgoing life-sustaining therapies, voluntarily stopping eating and drinking, and sedation to unconsciousness to relieve otherwise intractable suffering, could address many of these cases.1 The question remains as to whether physician-assisted death—that is, providing terminally ill patients with a potentially lethal prescription that they could ingest on their own to relieve otherwise intractable suffering by directly hastening death—should be one of these last resort options.

My own answer to this last question is a cautious “yes”: open access to physician-assisted death, subject to the safeguards of excellent palliative care and access to other last resort options, gives patients an important additional option, and the benefits of legalization outweigh the risks.

Progress in the Last Decade

Perhaps the most dramatic sign of progress has been the coming of age of the palliative care movement, which allows fully informed decision-making and the provision of treatments to maximize quality of life for all seriously ill patients alongside any and all disease-directed treatment that patients want to continue. Almost all major medical centers now have inpatient palliative care consultation services, and similar services are spreading into community hospitals. Consultation possibilities are also spreading into the outpatient and home settings, although the gaps between need and availability are much wider in these contexts. The American Board of Medical Specialties has recently given palliative care the status of being a board-certified subspecialty, and fellowship programs are sprouting up across the country.

There remain serious challenges. There are not enough skilled palliative care clinicians to meet the growing needs, and reimbursement for palliative care services—which rely heavily on counseling and coordination of care rather than expensive procedures—remains problematic. Similar gaps exist in providing basic palliative care education for all clinicians who care for seriously ill patients, and in generating an evidence base for the field. Nonetheless, palliative care seems to have passed the “tipping point” as a field; most patients and families can find the treatments that they need regardless of their stage of disease.

When patients become terminally ill, access to palliative care is facilitated by the proliferation of hospice programs. Hospice remains our premiere program to provide palliative care for terminally ill patients who are willing to forgo further treatment of their underlying disease, as it provides, pays for, and coordinates comprehensive quality-of-life-oriented treatments for terminally ill patients. Hospice has expanded considerably in the last ten years, primarily in two domains: the inclusion of terminally ill patients with diseases other than cancer—congestive heart failure, dementia, and chronic lung disease, for example—and the ability to supplement the palliative aspects of care for terminally ill patients who reside in skilled nursing homes.


Despite this progress, the majority of patients who die in the United States are never transitioned to hospice, mostly because of a requirement that once they are in hospice they will forgo disease-directed therapy. Some larger hospices are experimenting with loosening these restrictions through “bridge” programs that let patients continue active treatments that are important to them and have some potential for helping while at the same time receiving the full benefits of hospice. Since hospices are paid on a per diem basis (on average, about $135 per day), only the very large, affluent hospices can afford to offer expensive disease-directed therapy at the same time that they are providing and paying for comprehensive palliative care, but there is much more flexibility and willingness to experiment with these areas now than ten years ago.

Significant progress has also been made in the articulation and provision of last resort options for patients whose suffering becomes unacceptable to them despite state-of-the-art palliative care. Ten years ago, the problem of intractable suffering was often not acknowledged or was blamed on the clinician (“They don’t have adequate expertise”) or the patients (“They want too much control over their fate”). Now, it is much more widely acknowledged that even with the best possible palliative care, there will always be a small percentage of cases where suffering sometimes becomes unacceptably severe, and that clinicians are obligated to treat these circumstances as a palliative care emergency that requires consultation and committed efforts to respond in the most helpful, least harmful way.

Other Last Resort Options

When unacceptable suffering persists for terminally ill patients despite state-of-the-art palliative care, four options have emerged as last resort possibilities. The first two are widely accepted and relatively uncontroversial:

Right to intensive pain and symptom management. Pain can almost always be sufficiently relieved without any significant risk of hastening death, but there will be a few cases where pain accelerates severely toward the very end of life and pain medicines must be proportionately increased, bringing an increasing risk of hastening death. Patients and families, having given their informed consent, should be able to count on their doctors’ willingness to take these risks as part of their commitment to relieve suffering, and medical ethics and law stand firmly behind this commitment.

Consider a case like this: a patient with advanced bladder cancer is nearing death and must make daily tradeoffs between adequate pain relief and sedation. She reaches a point where she “just wants to be out of pain.” After extensive discussion with patient and family, her dose of opioids is gradually increased until her pain is sufficiently relieved, but after that point she eats and drinks very little and sleeps most of the time. She dies a week and a half later. Although death may have been unintentionally hastened by a small amount, in this process the obligation to relieve her severe suffering warranted taking that risk.

Right to forgo life-sustaining therapy. Medicine has made great strides prolonging life, and a small but significant part of this progress has been the availability of life-sustaining technologies. These technologies no longer just mean ventilators and feeding tubes; they now include radical technologies such as ventricular-assist devices. But while the array of medical choices faced by patients and families has grown more complex, ethics and law remain clear that patients have a right both to forgo such treatments and to stop them once started. Here’s a typical case that turns on this kind of decision: an elderly man fractures his spine falling off a ladder. He initially accepts intubation and surgery in hope that he will recover neurological function, but four weeks later, he has not. He cannot imagine living this way, and after extensive discussions with his medical team and his family, he is mildly sedated and removed from the ventilator. He dies within twenty-four hours.


Of course, some patients suffer unacceptably from symptoms other than pain, and their lives (and deaths) are not directly dependent on life-sustaining technologies. A portion of these patients look for options to escape suffering and even hasten death by means other than physician-assisted death. Two additional possibilities have emerged over the past decade. While clearly more morally complex and less settled than the first two options, these options seem to generate less legal and ethical controversy than physician-assisted death:

Voluntarily stopping eating and drinking. VSED is different from the natural loss of interest in food and drink that occurs when a person is actively dying. It is an informed decision to stop food and fluids while one is still physically capable in an effort to hasten death and escape suffering. The decision is entirely within the patient’s control, but it requires considerable resolve, since the patient generally continues to live for one to two weeks after stopping. Although VSED is not directly “physician-assisted,” it needs to be “physician-supported”: the patient must be assessed at the outset to ensure that he or she is competent and has adequate palliative care, and the physician must help the patient and family to address any unforeseen complications as the process unfolds.

VSED is only an option when the patient retains mental capacity. Suppose, for example, that a patient with amyotrophic lateral sclerosis has become tired of living in the circumstances imposed by his illness. He does not want a feeding tube or a mechanical ventilator, and he feels that he is dying too slowly. He asks about physician-assisted death. After careful assessment of his capacity and motivation, his physician raises the possibility of VSED. Since eating and drinking are already a daily struggle, he and his family welcome this alternative. His family can then come together for his last two weeks, during which he will be supported by home hospice. In the final phase of this process, he may be mildly delirious because of the metabolic changes of dehydration, but usually these and other symptoms can be well managed as part of his hospice support.

Sedation to unconsciousness. This type of sedation is distinct from the proportionate use of sedatives to treat severe anxiety or delirium, which is a relatively common part of terminal care. Instead, it is an explicit decision to render the patient unconscious so that he or she can escape otherwise severe and intractable suffering. Other potentially life-prolonging treatments such as food and fluids are usually discontinued at the same time. There is general agreement on its use as a last resort to relieve severe, intractable suffering in imminently dying patients, but as one gets further away from these circumstances, agreement about its permissibility falls off sharply.2 

Consider an elderly holocaust survivor dying of end-stage congestive heart failure who was transferred to an acute hospice/palliative care unit to manage his terminal shortness of breath. The opioids used to treat his dyspnea made him delirious, and he began having flashbacks from his time in the concentration camps. Usual sedatives did not control his agitation, and family and staff agreed that this suffering was unacceptable, so he was sedated to unconsciousness with phenobarbitol. The dose was adjusted until he appeared to be resting comfortably, but he was completely unresponsive. Artificial hydration and nutrition were not provided, with the family’s consent. His level of sedation was maintained until he died seventy-two hours later.


The Empirical Data

By far the best data about physician-assisted death in the United States come from Oregon, where the practice is reported to the health department and where annual summaries have been prepared every year since legalization. The practice has been remarkably stable over the ten intervening years, accounting for approximately one out of every thousand deaths per year.3 This appears to be a very small number given the amount of controversy surrounding the practice, but one in fifty patients talk to their doctor about it, and one in six talk with family members, suggesting that the availability of such an escape may be much more important to many patients than its actual use.4 The Oregon statute requires that patients be informed of “feasible alternatives,” including hospice and palliative care and other last resort alternatives, and some choose alternatives other than physician-assisted death.

There are also data from Oregon to suggest that the legalization of physician-assisted death enhances rather than undermines other aspects of palliative and end-of-life care. Oregon has one of the highest rates of hospice referral in the nation, and the vast majority of patients who choose physician-assisted death are simultaneously enrolled in hospice. Oregon also has relatively high rates of opioid prescription per capita, and physicians as well as other medical professionals have very high rates of attending training courses in both palliative care and end-of-life medical decision-making. There is also a statewide form, “Physician Orders for Life-Sustaining Treatment” (POLST), for recording a patient’s wishes about cardiopulmonary resuscitation and other potentially life-sustaining therapies. The form has become a model for other states.5 Overall, Oregon appears to be among the leaders in comparison to other states in virtually all aspects of palliative and end-of-life care, including allowing open access to physician-assisted death, subject to safeguards.

The secret practice of physician-assisted death in the rest of the nation is very difficult to study. To admit to participation, a physician has to admit to a crime, and along with any family present, runs the risk of prosecution. On the other hand, there appears to be very little interest in prosecuting such cases providing they are not discovered or flaunted, leading to a “don’t ask, don’t tell” policy that is unpredictable and potentially dangerous. Under this policy, there is no opportunity to get second opinions from experts in palliative care, no documentation, and considerable potential for idiosyncratic responses from clinicians.

In the mid-1990s, I was part of a team that conducted an empirical study of the secret practice, using research techniques that protected anonymity. We found that physician-assisted death and euthanasia accounted for approximately 1 to 2 percent of deaths. Although this appears to be ten to twenty times higher than the rates reported in Oregon, the reporting techniques were so different that the rates are not directly comparable. On the other hand, we know from Oregon that conversations with doctors about these issues are common, and it appears to be much better and safer to have the conversations out in the open rather than in secret.6


Very little is known about the frequency of other last resort practices in the United States. Data from the Netherlands, where all end-of-life practices are regularly studied, suggest that forgoing life-sustaining therapy and prescribing “opioids in large doses” are each reported to account for approximately 20 percent of deaths, which fits with my clinical experience in the United States.7 There are no reliable data about the frequency of voluntarily stopping eating and drinking in the United States, although the practice is thought to be rare. In our two hospice programs in Rochester, New York, where VSED is permitted and supported as a last resort, it accounts for less than 1 percent of deaths. Sedation to unconsciousness to treat otherwise intractable physical symptoms appears to be used very variably in the United States, apparently depending more on the values and practice patterns of the practitioners than of the patients. Reports vary from no deaths to half of all deaths, depending in part on definitions but also on practice patterns.8 At our hospital, where sedation to unconsciousness for treatment of intractable symptoms is subject to guidelines and restrictions that include a mandatory palliative care consult, it accounts for less than 1 percent of deaths.

The Clinical Context

All last resort options, including physician-assisted death, make sense only if excellent palliative care is already being provided. Mandatory palliative care consultation should therefore be a standard safeguard for any and all of these practices. Over the next ten years, medical institutions and professional groups should ensure that all clinicians who care for seriously ill patients are competent in the basics of palliative care and that specialty-based palliative care consultation is available for the more challenging cases.

There is also a need to develop explicit, predictable strategies to respond to difficult clinical situations where patients experience severe suffering despite state-of-the-art palliative care. Many of these patients will benefit from a discussion and exploration of last resort options that may or may not include physician-assisted death. There are two main clinical situations in which this might come up:

Patients who are worried about future suffering and wonder what options would be available to them. This conversation begins with an exploration about hopes, fears, and prior experiences of family and friends. Such patients frequently want to know what options they could have in the future if their suffering becomes unacceptable to them. In response to these inquiries, the clinician should talk to the patient about how he or she approaches such situations and what last resort options could be provided if needed. Many patients are reassured by learning about options other than physician-assisted death and by the willingness of the clinician to explore this domain and to commit to working with them and addressing their suffering throughout the illness until death. They can then be free to spend their remaining time and energy on other important personal and family matters.

Patients who eventually experience suffering that is unacceptable to them. This is a much smaller population than those who are worried about the future, but these patients’ needs can be more challenging. The starting point is always to explore the patient’s suffering in its totality, including why it is now experienced as unacceptable. Part of this assessment is to ensure that standard palliative care is being skillfully applied, and that the request does not emanate from anxiety or depression that might be otherwise addressed. A second opinion by a specialist in palliative care should be obtained. If there are no good alternatives, then the last resort options that are legally available should be explored in the approximate order presented in this paper. Usually, but not always, options other than physician-assisted death will adequately address the patient’s clinical situation and be acceptable to the patient. In the event that no other possibilities are workable or acceptable, physician-assisted death would need to be considered in light of the legal environment (the approach in Oregon will be very different than the rest of the country) and the values of patient, family, and clinician.


Will Physician-Assisted Death Be Necessary?

Some patients will prefer access to physician-assisted death even if the other last resort options are predictably available. Patients who request and eventually act using physician-assisted death in Oregon have a strong interest in controlling their fate, and physician-assisted death puts more choice directly in their hands. However, all the last resort options, including physician-assisted death, are imperfect. Although each addresses some situations particularly well, there are other situations where they would not be as helpful.

For example, voluntarily stopping eating and drinking has the advantage of putting the decision in the patient’s hands, but it requires tremendous discipline not to drink if one is thirsty and capable of drinking, and the duration of the process is too long if symptoms are severe and immediate. On the other hand, medical sedation to unconsciousness may be very frightening to those who value consciousness and being in charge, and there is no way to verify that the sedated patient is not still suffering but unable to report it. Finally, physician-assisted death requires that the patient be physically capable of self-administration and able to swallow a concentrated amount of lethal medication. In addition to these practical issues, any of these options may be morally troubling for patient, family, physician, or staff.

Adding physician-assisted death to the list of last resort options has both risks and benefits. One benefit is that it adds another important possibility for terminally ill patients who experience unacceptable suffering. We should be as responsive as possible to these patients without violating fundamental values, but it is clear that the patient’s values in this context count the most, followed by the family and then the clinician (if the course of action requires the physician’s participation). Most patients will be reassured by the possibility of an escape, and the vast majority will never need to activate that possibility. But some patients will need a way out, and arbitrarily withholding one important option from patients whose options are so limited seems unfair.

The main risk of including physician-assisted death with other last resort options is that it seems to be very polarizing in the United States, where there is wide agreement about palliative care and hospice being the standards of care and also a surprising level of acceptance of the other last resort possibilities. Even the Supreme Court, in its 1997 decision, made it clear that “obtaining medication . . . to alleviate suffering even to the point of causing unconsciousness and hastening death” is legally acceptable.9 Opponents of physician-assisted death may work to further restrict access or even prohibit access to other last resort alternatives as they become more well known and predictably available.

On the Horizon

It seems highly likely that palliative care and hospice will continue to expand in the United States and other Western countries. There are very few places in the health care system where we can simultaneously save money and improve quality, but palliative care and hospice have the potential to do both. On the quality side, improvements in pain and symptom management, more informed medical decision-making, and enhanced patient and family support are core elements of palliative care and hospice. The cost savings will come not from restricting access to expensive treatments and technologies, but from a better informed consent process for patients and families concerning medical treatments with marginal benefit and coordinated care for patients with very complex treatment. Hospice continues to be the gold standard for end-of-life care, but its challenge is to design programs that would allow patients to simultaneously continue some disease-directed therapies in order to serve a wider range of dying patients.10 If these programs can be proven cost-effective, then perhaps the hospice benefit can be expanded, and hospice and palliative care could be better integrated into traditional medical care. Palliative care needs to be part of the standard of care for all seriously ill patients, whether or not they choose to continue disease-directed therapy in any form.


The last resort options other than physician-assisted death must become more standardized, available, and accountable. There is currently too much variation. Some patients may be denied access to them because clinicians or institutions are reluctant to use them, while at the same time, others are given last resort options when more standard palliative measures would have been more appropriate. Better policies and procedures should begin at the national level, with local programs following suit. Fortunately, medicine seems to be moving in this direction. Witness a recently published American Medical Association guideline on sedation to unconsciousness that is consistent with position statements by the American Academy of Hospice and Palliative Medicine and others.11

Similar guidelines and policies are needed for voluntarily stopping eating and drinking. Because these options are intended to be rarely used, all institutions should review their own practices against the position statements of leadership organizations. For example, my institution has a guideline on sedation to unconsciousness for treatment of refractory suffering, and we review every single case in which it is utilized. The state of California is considering a law entitled the California Right to Know End-of-Life Options Act, which requires that patients be given information about both hospice and last resort options, should they request it. The goal of all these initiatives is to improve predictable access and accountability both for hospice/palliative care and for legally permitted last resort options.

In the domain of physician-assisted death, the most pressing change on the horizon is the Oregon-style citizens’ initiative in the state of Washington. Residents of the state are already relatively well educated on the subject, having been through an unsuccessful initiative in 1991, and they are more aware of the Oregon experience, being adjacent geographically and relatively similar demographically. Citizens’ initiatives begin by gathering signatures of support from a large number of residents. If the legally determined threshold is achieved, the initiative is then placed on the ballot during the next election cycle and subject to the popular vote. Not all states sanction citizens’ initiatives, but in my opinion, they are more likely to be successful in legalizing physician-assisted death than legislative approaches, given the relatively high level of public support and the tendency for the issue to get polarized in legislative situations. Nonetheless, legislative processes to legalize physician-assisted death were recently attempted in California, and before that there were near misses in Hawaii and New Hampshire.

Although some proponents of physician-assisted death will find the incremental, state-by-state approach to legalization frustrating, it has some value. For one thing, it gives us time to study the intended and unintended effects of legalization before national implementation. We can simultaneously evaluate the impact of better and more widespread access to hospice and palliative care, and of more predictable and accountable availability of other last resort options. In addition, it keeps the movement a grass-roots one—particularly when it happens through citizens’ initiatives. That may not make the practice of physician-assisted death less controversial, but it can perhaps make the process through which physician-assisted death is legalized less controversial.

 

1. T.E. Quill, B. Lo, and D.W. Brock, “Palliative Options of Last Resort: A Comparison of Voluntary Stopping Eating and Drinking, Terminal Sedation, Physician-Assisted Suicide, and Voluntary Active Euthanasia,” Journal of the American Medical Association 278 (1997): 1099-2004.

2. See F.A. Curlin et al., “Religion, Conscience and Controversial Medical Practices,” New England Journal of Medicine 356 (2007): 593-600.

3. See “Death with Dignity Act, Records and Reports Data on the Act,” http://www.oregon.gov/DHS/ph/pas/.

4. S.W. Tolle et al., “Characteristics and Proportion of Dying Oregonians Who Personally Consider Physician-Assisted Suicide,” Journal of Clinical Ethics 15 (2004): 111-18.

5. “POLST: Physician Orders for Life-Sustaining Treatment Paradigm,” at http://www.ohsu.edu/ethics/polst/.

6. Tolle et al., “Characteristics and Proportion of Dying Oregonians Who Personally Consider Physician-Assisted Suicide.”

7. P.J. van der Maas et al., “Euthanasia, Physician-Assisted Suicide, and Other Medical Practices involving the End of Life in the Netherlands, 1990–1995,” New England Journal of Medicine 335 (1996): 1699-1705; B.D. Onwuteaka-Philipsen et al., “Euthanasia and Other End-of-Life Decisions in the Netherlands in 1990, 1995, and 2001,” Lancet 263 (2003): 395-99.

8. L.A. Jansen and D.P. Sulmasy, “Sedation, Alimentation, Hydration, and Equivocation: Careful Conversation about Care at the End of Life,” Annals of Internal Medicine 136 (2002): 845-49.

9. Washington v. Glucksberg, 117 S.Ct 2258 (1997); Vacco v. Quill, 117 S.Ct. 2293 (1997).

10. A.A. Wright and I.T. Katz, “Letting Go of the Rope—Aggressive Treatment, Hospice Care and Open Access,” New England Journal of Medicine 357 (2007): 324-27.

11. K.B. O’Reilly, “AMA Meeting: AMA OKs Palliative Sedation for Terminally Ill,” American Medical News, July 7, 2008, at http://www.ama-assn.org/amednews/2008/07/07/prsi0707.htm.

Timothy E. Quill, "Physician-Assisted Death in the United States: Are the Existing 'Last Resorts' Enough?" Hastings Center Report 38, no. 5 (2008): 
Support The Hastings Center

Make a Gift

Other ways to give

  Full Access Ad

IRB ad

Support