7

Reflections on the Workshop and Evidentiary Gaps

In the workshop’s final session, moderators of previous sessions were asked to summarize the themes that emerged from the 2 days of presentations and discussions. This chapter includes the themes as summarized by workshop session moderators as well as evidentiary gaps and potential directions for future discussion of physician-assisted death that were suggested by individual workshop participants. Summary statements made by session leaders do not imply consensus among workshop participants.

SESSION ONE REFLECTIONS: EVIDENCE AND TERMS OF DISCUSSION

Linda Ganzini
Professor of Psychiatry and Medicine
Oregon Health & Science University

Linda Ganzini, who chaired the workshop’s first session on evidence and terms of discussion, began by saying that she thought Daniel Sulmasy of Georgetown University had raised an interesting question, which concerned how one’s beliefs affect the research questions one asks. Ganzini also recounted how Sulmasy pointed to the importance of keeping ethical ideas in mind when asking scientific questions about physician-assisted death. She said that one of the larger themes she heard was that even if it was possible to be assured about the motivations, competence, and mental status of every patient who asks for physician-assisted death, the

potential for acceptance of this practice may increase distrust among those in marginalized groups.

The challenge of assessing decision-making capacity, competence, and voluntariness were reccurring themes during the workshop, Ganzini said, adding that there are significant data and methods in the literature regarding determining decision-making capacity that could be applied in the context of physician-assisted death. This includes evaluating whether mental health professional are applying similar types of standards to bear on their evaluations of patients requesting physician-assisted death. Reflecting on Anthony Back’s presentation, Ganzini commented that many of the patients who have been the first to avail themselves of these laws seem to celebrate the occasions of their deaths, and she asked whether these individuals might be changing society’s ideas about death and how to die. She concluded her summary by reiterating the importance of research so as to avoid drawing conclusions based on information in the blogosphere.

SESSION TWO REFLECTIONS:
PROVIDER EXPERIENCES AND APPROACHES

David Magnus
Thomas A. Raffin Professor of Medicine and Biomedical Ethics,
Stanford University
Co-Chair, Stanford Hospital Ethics Committee

Neil Wenger
Professor of Medicine, University of California, Los Angeles (UCLA)
Director, UCLA Healthcare Ethics Center
Chair, UCLA Medical Center Ethics Committee

David Magnus and Neil Wenger served as co-chairs of the session on provider experiences and approaches. Wenger observed that it might be possible to learn from what is essentially a natural experiment with nearly identical interventions being variably implemented in somewhat divergent populations. “How do we evaluate a paradigm shift that is an anathema to some and considered essential to others?” he asked. “This is a question that we should take to heart.” He proposed that there could be a case-controlled trial comparing patients with terminal conditions who want aid-in-dying in three states that permit the practice and three that do not or in health care organizations that opt in and those that opt out. The difficulty, he acknowledged, would be in finding patients who want it, not necessarily patients who follow through. Such a study could also

evaluate the end-of-life care received by patients and its effect on families and clinicians.

For Wenger, physician-aided dying raises several questions that, he said, can create skepticism about current practices around which there formerly was no skepticism, such as: How certain are we about capacity and undue influence in withholding or withdrawing treatments that are common in hospitals today? Are there conflicts of interest regarding very aggressive treatment decisions that are decided between oncologists and patients or between surgeons and patients every day? What is the approach toward persons with disabilities? And, reflecting on the presentation by Joanne Lynn of the Altarum Institute, are we so comfortable focusing on the medical model of care because it is simply too painful to think about the social determinants of health and medical care?

The request for physician-assisted death is very powerful, Wenger said, and it stimulates a cascade of communication and intervention that appears to be missing in routine care but that should not be. He said that the resources devoted to preparing institutions to deal with requests for physician-assisted death far outweigh the number of patients affected, though these efforts may enable other necessary discussions about advance care planning and palliative care. Wenger also noted the importance of paying attention to the slippery slope discussed by several speakers.

Magnus commented that he had hoped this workshop would focus on the micro- and meso-level issues without as much focus on the macro-level moral and ethical issues, but this workshop has shown that to be impossible because all of these levels of discussion are intertwined in multiple ways. Though there are many questions, detailed below, that require research and data to answer, Magnus said that the data available are reassuring in some ways. However, he added, the fact that there is so much unknown about the practice of physician-assisted death in the United States is concerning, particularly regarding the ancillary effects on patient care. There is a potential, he said, that the availability of physician-assisted death and its use by a very small number of people will lead to a significant increase and improvement in high-quality end-of-life care in general. Alternatively, physician-assisted death may have negative ancillary effects on end-of-life care. Or, there may be no effects at all. It may turn out that physician-assisted death will only be relevant for a tiny number of people, in which case it will turn out to be a relatively unimportant topic. The bottom line is that research is needed to monitor these vastly different outcomes, Magnus said.

SESSION THREE REFLECTIONS:
PHYSICIAN-ASSISTED DEATH IN THE BROADER CONTEXT

Joanne Lynn
Director, Center for Elder Care and Advanced Illness
Altarum Institute

Richard Payne
Esther T. Colliflower Professor of Medicine and Divinity (Emeritus),
Duke University
John B. Francis Chair, Center for Practical Bioethics

James Tulsky
Chair, Department of Psychosocial Oncology and Palliative Care,
Dana-Farber Cancer Institute
Professor of Medicine, Harvard Medical School

James Tulsky remarked that the real focus of Session 3 ended up being how providers engage with and adapt to physician-assisted death when it becomes legal. He found this discussion useful, given that he lives in Massachusetts, where he expects physician-assisted death to become legal in 2020.

Tulsky’s first of three take-home points from the session was that requests for physician-assisted death must stimulate deep conversation between clinicians and patients and enhance quality end-of-life care. The second was that context is important in that the implementation of a physician-assisted death program in a particular community cannot be separated from the ethnic and social make-up of that community. The third key point was that this issue is complicated for physicians and other health care providers. Moral distress, he said, is going to be present, and it is important to acknowledge and manage. Moral distress cannot be ignored, regardless of whether physician-assisted death is legal in a given jurisdiction.

It was clear from the presentations, Tulsky said, that creating an institutional policy requires massive stakeholder engagement and that patients must opt in rather than be offered these services by providers. It was also clear, Tulsky said, that in response to these new laws, health care systems have developed varied policies that focus as much on improving quality of end-of-life care as they do on allowing adherence to the laws. These policies all encourage the use of physician orders for life-sustaining treatment (POLST) forms and other measures to ensure that well-developed alternatives to physician-assisted death exist. Furthermore, each of these policies manages the requirement for secondary consultation differently, but generally uses this provision to improve the palliative care of

patients considering assisted dying. Several speakers made the point that systems that opt in need to provide some level of patient navigation and provisions for those providers who opt out. Tulsky said that it appears to be helpful to have a list of willing prescribers, but the question remains whether it is desirable or undesirable to have one or two “go-to” providers. Tulsky’s final takeaway, he said, was that having willing health care providers present at the time of death can provide great comfort to patients and their families.

Richard Payne began his summary by noting his concern about the implications and impact that the widescale adoption of physician-assisted suicide will have on the profession of medicine, particularly in the context of the fragmented, disorderly way medicine is practiced in the United States. In particular, he said, he believes that such adoption may cause various issues for vulnerable, medically underserved minorities who have few long-term relationships with a health care provider. Reflecting on the different perspectives within various communities—in particular, African American communities and religious groups—Payne said he was surprised to learn that Archbishop Desmond Tutu recently voiced support for assisted death, reversing his lifelong position on the topic, in part due to observing the prolonged death of Nelson Mandela.¹ Payne stressed that no minority community is a monolith and said that there is a need for ethnographic data that crosses racial-ethnic strata as well as socioeconomic strata within a racial and ethnic group.

Payne said that voluntariness is the primary concern when considering vulnerable populations and physician-assisted death. He asked how slippery the slope will become and how voluntariness can remain protected as the processes around physician-assisted death are scaled up. Payne also challenged the workshop participants to have a subtler and more expansive view of the possible threats to the integrity of the voluntariness concepts. For instance, he said that the social determinants of health, living, and dying are critical and must be explored by hearing from the people most directly affected by the impact of key social factors on decisions about health care, including end-of-life care.

Lynn expressed concern that widespread adoption of physician-assisted death laws will trigger a very subtle but broad community-based discrimination against the elderly and people with disabilities, given the budgetary challenges regarding how the nation cares for those who are frail and elderly and who have disabilities. In her opinion, she said, the challenges facing older Americans should be a rallying cry for action that

___________________

¹ For more information, see https://www.npr.org/sections/parallels/2017/01/04/507294833/at-85-desmond-tutu-calls-for-the-right-to-an-assisted-death (accessed April 25, 2018).

goes far beyond concerns about physician-assisted death. Once the nation straightens out how it cares for older Americans and those with disabilities, she said, then physician-assisted death will become an intriguing question to address.

SESSION FOUR REFLECTIONS: DATA COLLECTION IN THE UNITED STATES AND OTHER COUNTRIES

Nancy Berlinger
Research Scholar
The Hastings Center

Nancy Berlinger, who served as the chair of the session on data collection in the United States and other countries, remarked that Jennifer Gibson in her presentation said that data on access, quality, equity, and societal impacts would all be important indicators of the effect of physician-assisted death on the communities where it is legal. Gibson said that the perspectives of the public and patients are not always well captured by the reported data but are an important part of the story along with the perspectives of providers. One important question raised in the session was who has accountability for the collection and use of data on physician-assisted death. There was much discussion, Berlinger said, about what data the government in any country should collect, as well as which data public health officials are authorized to collect and analyze under the scope of legislation permitting physician-assisted death, and which data are beyond this scope but may be collected and analyzed by others. She remarked that communication about physician aid-in-dying was mentioned several times as something that would be hard to collect data on but which is fundamental to understanding the practice, including how patients may use the existence of a legal provision to initiate discussions about end-of-life care, whether or not they actually request a prescription.

Filling in gaps in understanding about how the physician-assisted death process works can compete directly with patient-centered goals that are seen as central to the process, she said. Examples would be trying to get patient accounts of their motivation for requesting physician-assisted death or getting in-depth patient and family experiences of the practice. Berlinger explained that patients and families undergoing the process of considering physician-assisted death or completing the process should not necessarily be expected to participate in research as it could create too heavy of a burden on what is essentially a private act.

Matthew Wynia raised a concern about how the medical community should develop and adhere to standards on physician-assisted death, including in jurisdictions where it is not yet or will never be legal, and how data collection should reflect those standards. Wynia proposed the idea of a national registry for information on physician-assisted death.

DATA COLLECTION

Scott Kim, who acknowledged that he believes there is hardly a problem for which the answer is not “more data,” said he found himself struggling with the call for more data in this space. He said that the data presented during the workshop did not address the question of whether physician-assisted death is a good thing, and that is typically the standard applied in health care. He noted that Wenger’s suggestion for a case-controlled study would provide part of the answer, but that a better option might be to use the approach Oregon undertook when it randomized expanded access to Medicaid. Kim explained that the data collected on Medicaid use in Oregon through this approach has proven incredibly useful, and it is conceivable that one could take a similar approach in researching physician-assisted death.

Kim then raised a question for the workshop participants to consider: “If we are not going to make the social commitment to have that standard of data about physician-assisted death—and arguably, we ought not to—then would it be better for us to stop saying that this is a data-driven entity entirely, stop even calling for data, and just agree that this is something that is on that kind of uncomfortable saddle between the political and the metaphysical?”

Tulsky replied that he empathized with Kim’s question because what has become clear to him over the course of the workshop was that no data will convince those who are firmly opposed to physician-assisted death to endorse the practice, nor will it convince the people who are firm advocates to oppose it. However, he said, he believes that data can provide insights into how to manage physician-assisted death, assuming it is going to be legal, so that it is done in a safe, efficacious, and effective manner with the best outcomes. Tulsky said that he did not want to make physician-assisted death a “data-free zone” but rather challenged workshop participants to be honest about what data can and cannot do as well as what the goals are of collecting those data. Magnus said that getting quality data about best practices in this area is important, and Anthony Back noted the importance of using research and data to effectively create a process of public and professional engagement on this issue.

RESEARCH AND FURTHER DISCUSSIONS

Throughout the workshop, many participants spoke about gaps in the evidence concerning physician-assisted death and end-of-life care in general, as well as areas for further discussion regarding ethical and moral considerations surrounding physician-assisted death. In terms of empirical research, Lynn raised the question of who will fund research in this area. She noted that the mission of the National Institutes of Health concerns cures and the prevention of disease, but does not focus on researching how people will live. In her experience, a number of federal agencies and philanthropies are also uninterested in sponsoring research in this area. A research agenda has never been developed but nonetheless is needed, not least of all because end-of-life decisions are in most people’s future, Lynn said.

Unanswered questions or potential opportunities for additional research mentioned by individual workshop participants during the workshop include the lists below. Some of the questions and issues listed below may have been raised by more than the one or more individuals attributed to each statement.

Patient Care and Patient Choices

• What are the ancillary impacts of physician-assisted death laws on the care received by people who are not accessing the option and can the impact of physician-assisted death on clinical care beyond those who received the intervention be measured? (David Magnus, Neil Wenger)
• How has the legalization of physician-assisted death affected end-of-life care and palliative care for the people who are not going to take advantage of the law? (Barbara Koenig)
• If physician-assisted death leads to an activation of resources—time with providers, enhanced communication—with a positive impact for patients, can we better understand why this happens and export these lessons to contexts beyond physician-assisted death? (David Magnus)
• What is the consent process for palliative sedation if the drugs used for physician-assisted death do not work as planned and who is the responsible physician after the patient takes a lethal medication? (Timothy Quill)
• Can we better understand how physician-assisted death is similar to or different from suicide in other contexts? (Linda Ganzini)
• What are the financial, emotional, and other pressures being faced by older individuals using long-term services and supports, and

• how do they affect considerations of physician-assisted death? (Joanne Lynn)

• Why has patient demand for medical aid-in-dying increased in recent years, and how should that increase inform the delivery of patient-directed care? (Kim Callinan)
• Do any patients complete physician-assisted death because their symptoms are not being managed? (Barbara Hansen)
• What are the rates of cognitive impairment and dementia in patients in home and community-based long-term care settings? (Cheryl Phillips)
• Are the legal safeguards regarding mental health screening failing to screen out some people with impaired judgment who should not be getting a prescription for lethal medication? (Thaddeus Pope)
• What is the frequency of complications arising during the procedure? (Matthew Wynia)

Implementation

• What harms occur because of the fracturing of continuity of care as a result of physicians or institutions opting out of physician-assisted death? (David Magnus)
• How do the prices of drugs used in physician-assisted death affect how people of different socioeconomic status make decisions regarding physician-assisted death? (Matthew Wynia)
• How is “terminal illness”—meaning a 6-month prognosis—currently being determined in jurisdictions where physician-assisted death is legal? (Scott Kim)
• How can we better understand the decisions physicians are making in terms of a patient’s mental health and competence in the process of requesting physician-assisted death? (Tom Strouse)
• How strong a presumption of capacity is used, and what thresholds are applied in assessments of decision-making capacity? Are evaluators using a checklist to determine if a patient makes certain statements, or is there an in-depth clinical interview probing the person’s understanding? Who provides the second opinion on these determinations? (Scott Kim)
• How often are patients referred to low-threshold physicians (e.g., physicians more likely to say yes rather than no to a request for physician-assisted death) and is this desirable or undesirable? (Scott Kim)
• What are the appropriate standards for evaluating capacity in the context of physician-assisted death, given that the current tools for evaluating capacity for other types of medical decision making are

• based on the patient’s ultimate goal being health, as opposed to death? (Tom Strouse)

• What is the effect of the waiting period and what constitutes an official request from a patient for physician-assisted death? What is the appropriate length of time to ensure that such a monumental decision consistently reflects the patient’s wishes? (Barbara Koenig, Peter Reagan)
• What is the appropriate balance between legal safeguards and access to medical aid-in-dying? Which legal safeguards and regulatory requirements are necessary, and which ones create unnecessary delays and stigma? (Kim Callinan)
• What are best practices in provisions for safe disposal of unused drugs? (Matthew Wynia)
• What are the current practices in long-term care settings regarding physician-assisted death in locations where it is legal? (Joanne Lynn)

Access and Utilization in Disabled and Vulnerable Populations

• What are the views about physician-assisted death in the disabled community, and what is the use of physician-assisted death by disabled patients? Are there some disabled patients unable to access physician-assisted death who otherwise would choose that option? Are disabled patients getting access to physician-assisted death who otherwise would not be eligible? Is there evidence that physician-assisted death is having ancillary negative impacts on care for disabled patients? (David Magnus)
• What is the impact of physician-assisted death on vulnerable populations, particularly African American and other underserved minority communities and how do these communities feel about the practice of physician-assisted death? What will ethnographic research that crosses racial and ethnic strata as well as socioeconomic strata with a racial or ethnic group show about the acceptance of this practice and how legalization affects trust or mistrust in the health care system? (Jeffrey Berger, Barbara Koenig, Richard Payne)
• Is the lower use of physician-assisted death among low socioeconomic status groups reflective of unequal access, challenges in navigating the system, lack of information, or less of a preference for aid-in-dying among these groups? (Mara Buchbinder, Barbara Koenig)

• Does physician-assisted death pose a threat to voluntariness as it is scaled, particularly for vulnerable populations? What kinds of social determinants of health, living, and dying pose threats to the integrity of the voluntariness concept? (Richard Payne)
• What is the effect of physician-assisted suicide on individuals with psychiatric disorders? Does publicity about the practice trigger an increase in suicides? (Dan Sulmasy)
• What are the major challenges in accessing physician-assisted death in states where it is legal? (Kim Callinan, Omega Silva)

Impact on Families, Clinicians, and Health Care Systems

• Is there a difference in the grief process for survivors of a person who completed physician-assisted death compared to survivors of a person who died a “natural” death, who voluntarily stopped eating and drinking, or who committed suicide using a more violent means? (Barbara Hansen)
• What is the psychiatric effect of assisted suicide on families? (Dan Sulmasy)
• What are the psychological effects on clinicians of participating in assisted deaths? (Anthony Back)
• Does opting in contribute to more or less “burnout” among clinicians? (Barbara Hansen)
• What is the effect on institutions that opt in or opt out and to what extent do participating institutions believe they have an obligation to identify a provider if a patient’s physician chooses not to participate, and how is this obligation carried out? (Barbara Hansen)
• What is the impact of not passing medical aid-in-dying laws? Does a lack of medical aid-in-dying laws create a more dangerous underground practice? (Kim Callinan)
• What mechanisms exist to support providers in carrying out physician-assisted death, and which practices and policies work best? (David Magnus)
• What is the impact on patients and families if their hospice program has a policy that does not allow their staff to be present in the home when the patient takes the medication? (Barbara Hansen)

Ethical and Social Norms

• What would serious, deliberative polling reveal about the impact of physician-assisted death on social and cultural norms? What can be done to implement genuine democratic public engagement on

• physician-assisted death? (Barbara Koenig, Dan Sulmasy, Matthew Wynia)

• How are the various proposed expansions in laws outside of the United States likely to affect laws and regulations in this country? For example, what will be the positive and negative implications of expanding physician-assisted death to minors, of broadening or eliminating the terminal illness requirement, of modifying the self-administration requirement, or allowing physician-assisted death to be included in advance directives in other countries? (David Magnus)
• How do the conflicts of ethics and values for physicians, family members, and others involved in physician-assisted death affect those who need long-term services and supports? (Joanne Lynn)
• What are the potential and realized conflicts of ethics and values for providers of long-term care? (Joanne Lynn)
• How do societal processes, including how publics learn from each other directly or indirectly through mass and social media, affect the acceptance of physician-assisted death? How do those processes affect grieving process? (Nancy Berlinger)
• Can ethical frameworks be used to consider the impact that physician-assisted death may have in terms of a “gift exchange” between the patient and their loved ones, allowing them to confront and address death and dying? (Anthony Back)
• Is the public’s interest in legalizing physician aid-in-dying part of a broader set of issues involving lack of trust in the health care system? (Barbara Koenig)
• How has long-term use of physician-assisted death in Oregon changed the nature of perceptions about this practice, and to what extent is it having an impact on moral dimensions of society? If there is a change in perceptions about this practice, is it occurring because of broader changes in society or because passing a law does produce profound changes in the ethos of a culture? (David Magnus, Neil Wenger)
• Separate from how providers adjust to physician-assisted death legalization, what will it take to develop a better understanding of where physician-assisted death sits in relation to a provider’s own practice in terms of deciding how they will react to legalization? (Nancy Berlinger)
• Is there a psychological slippery slope in terms of how practitioners and patients begin to see physician-assisted death as part of normal practice and whether there is pressure to participate? (Daniel Sulmasy)

Data Collection

• Additional research is needed on the validity of the legal reporting mechanisms in states where physician-assisted death is legal so as to better understand whether those data are true, as well as data on incidences of abuse. (Daniel Sulmasy)
• In states where physician-assisted death is legal, additional data are needed in order to assess how the laws are working. For instance, data are needed on the two types of slippery slope: (1) expansions of physician-assisted death within an accepted category of practice (e.g., “terminally ill”) and (2) expansions in the categories of persons who can receive physician-assisted death (e.g., children, non-terminally ill, advance requests). (Scott Kim)

Broader Research on End of Life

• What is the range and general practice of end-of-life alternatives to physician-assisted death—proportional palliative sedation, palliative sedation to unconsciousness, voluntarily stopping eating and drinking? How are these alternatives being used and what are the policies governing their use? (David Magnus, David Orentlicher)
• Can predictive models of when death is likely to occur help to operationalize the 6-month prognosis in a more precise manner? (Joanne Lynn)
• The desire to control how one dies is often dismissed as an illegitimate reason for physician-assisted death but is often a reflection of a patient’s lifelong desire to overcome times of his or her life when controlled by others. Additional examination is needed of the range of end-of-life interventions and an individual’s ability to leave this world in control. (Linda Ganzini)

This page intentionally left blank.