Canadians who meet eligibility requirements can opt to self-administer or have a clinician administer these medications; the vast majority of people choosing MAiD have had their medications delivered by physicians or nurse practitioners. Canada is the first country to permit nurse practitioners to assess for medically assisted dying eligibility and to provide it.
Our most recent research involved interviews with 59 nurse practitioners or registered nurses across Canada who accompanied patients and families along the journey of medically assisted dying or who had chosen to conscientiously object. Nurses worked across the spectrum of care in acute, residential and home-care settings.
Neither of these perspectives do justice to the complexities of MAiD as it is enacted. Without an understanding of those complexities, it is difficult for patients and families to make good decisions.
Nurses accounts of MAiD
Nurses told us that medically assisted dying is about so much more than the act itself. Medically assisted dying is a conversational journey with patients that lasts weeks or even months.
These discussions patients have over time with skilled and compassionate health-care professionals help to determine whether this is what they really want, or whether there are other options that might relieve their suffering.
Conversations between patients and their families are essential to negotiating a common understanding and moving forward together.
Indeed, evidence has suggested that these conversations, when experienced as meaningful by patients, may help to alleviate the suffering that leads to the request for a medically assisted death. This is particularly true if the suffering has arisen from the sense of isolation.
If and when patients decide to proceed with MAiD, then conversations are required to ensure that all of the organizational details (what, where, when, how) are patient-centred choices and that those who are involved know the part they are to play. After the act of medically assisted dying, it is compassionate conversations that support families in navigating an uncharted bereavement process.
So yes, medically assisted dying is about supporting autonomy, but it is also about understanding that autonomy exists within, and is shaped by, our constellation of relationships. We need to be talking more about the essential nature of what it means to have a good death.
Complex reasons to choose death
MAiD is often spoken of as the definitive intervention that ensures control over the alleviation of suffering. But, we have learned that MAiD can also be chosen as the antidote to a system that fails in compassion or equitable palliative care access.
We heard a story of one man who had overstayed the time allowed on a palliative care unit. His doctor was a conscientious objector to medically assisted dying so each time health professionals planned to transfer him to residential care, the man asked for a medically assisted death. In doing so his stay in palliative care was assured.
So, while medically assisted dying does promise control over people’s suffering, it can also be used as a form of resistance to a challenging system or depleted support.
We need to plan ways to ensure that inequitable access or lack of caregiving networks do not become the default reasons for requesting a medically assisted death.
Nurses emphasized how important it is to have preparatory conversations repeatedly. Organizing an assisted death is labour-intensive for all involved; it requires thoughtful and detailed planning within the care system and among families and support networks.
Often the first time that patients and families hear a detailed explanation of the process is when the nurse or the physician first assesses eligibility. Nurses said it is not uncommon for patients to experience uncertainty, to vacillate in their decision around an assisted death, or to experience fear at the moment of death.
It is tough to talk about your uncertainty when so many have invested time and energy into planning your death. At the time of assisted death, nurses and physicians go to extraordinary lengths to ensure a “good death” by normalizing the process, fulfilling patient wishes and providing exemplary clinical care.
Despite all of this, the death is often deeply impactful because it is so different than the death we have known where people gradually fade away. Persons receiving medically assisted death are fully there one minute, and gone the next.
Within minutes they go from talking, to unconscious, to a grey pallour that signifies death, and this “greying” affects even seasoned health-care providers. The death can provoke an array of overwhelming emotions in health-care providers and families alike, both positive and negative.
With the changing landscape of medically assisted dying in Canada, the need for reflective conversations becomes ever more urgent. We need to better understand how medically assisted dying changes the nature of death to which we have become accustomed and how those changes impact all those involved.
Death may be the great equalizer but the availability of good end-of-life care is rarely equitable. Now, a new mobile palliative care program designed to address that inequity is providing care and dignity to people with life-limiting illnesses who are homeless and living in poverty in Victoria.
The Palliative Outreach Resource Team (PORT) is a collaboration of the University of Victoria, Island Health, Victoria Cool Aid and Victoria Hospice. PORT acts as a bridge between people with serious illness and their caregivers, palliative care, and other health and social support systems.
The program is built upon lessons learned from a three-year study led by UVic palliative care researcher Kelli Stajduhar, lead investigator of the Equity in Palliative Approaches to Care program with the Institute on Aging & Lifelong Health and the School of Nursing. The study followed 25 people living homeless or barely housed while struggling with life-threatening medical conditions. The 2018 report Too little, too late: How we fail vulnerable Canadians as they die and what to do about it, found that homeless and barely housed people have to navigate many systems—health care, housing, social care—and that as their health declines, their ability to access these systems also declines. The big takeaway: despite a terminal diagnosis of cancer, heart failure or lung disease, those who were able to access palliative care actually experienced an improvement in quality of life.
For PORT’s first year, the clinical team will be funded by Island Health and Saint Elizabeth Health Community Enterprise, a social enterprise with a commitment to end-of-life care for marginalized communities. Mirroring similar models in Toronto and Calgary, people can self-refer or be referred by their caregivers to a palliative care nurse and a physician who provide whole person care, manage the pain and symptoms related to life-limiting illness, support chosen family and caregivers, and provide grief and bereavement support. Chosen family and caregivers in this population include “street family” and shelter, housing, harm reduction, and peer and support workers from inner-city community organizations who are doing the bulk of end-of-life care for people living in poverty.
The Vancouver Foundation is funding the UVic-led evaluation of the program, as well as the development of initiatives to increase access to and quality of palliative care in the inner city. The PORT team, which began service in July, has supported three deaths and is currently supporting seven people who are dying.
“For almost a decade, providers in our community have cobbled together resources to meet the needs of our clients who are living with unmet palliative needs,” says Grey Showler, director of health and support services at Cool Aid. “We are thrilled to see PORT come to life.”
“Over the next year, we will be implementing this model of palliative care in collaboration with organizations and people who have expertise in care and support for homeless and vulnerably housed people at end-of-life including street families,” says Jill Gerke, director of the palliative and end-of-life care program with Island Health. “We are using research and promising practices to inform the development of this model adapted to our community that bridges existing support and services.”
“Palliative care isn’t a ‘thing’ or a ‘place’ but an approach that focuses on whole-person care for the person, their family and community. This approach necessitates a community response where everyone sees their responsibility and their part in care for dying people,” says Stajduhar.
Ivette Jeffries-Logan and Omisade Burney-Scott are friends for life – and collaborators in death. Three years ago when a mutual friend realized she wouldn’t survive pancreatic cancer, the two central North Carolina women were within the circle of friends she summoned.
Over the course of about three months, the women stayed at Cynthia Brown’s side, as the community activist and one-time Durham City Council member went about the process of dying.
They rubbed her head, kept a watchful eye on her pain, and helped her decipher doctorspeak. And when her spirits appeared to lag, they’d tell her jokes and sing at her bedside.
This, Jeffries-Logan says, was a good death: “If I can help someone at the end of life heal and be clear, I will. There are some things we are required to do alone, but we are not isolated. We are community people. What happens to my nation happens to me. What happens to me happens to my nation.”
Jeffries-Logan and Burney-Scott are death doulas; their form of caregiving is both old and new. The ancient Greek word “doula,” meaning “woman servant” or “slave,” was repurposed in the 1960s to describe birth workers who offer encouragement, back rubs, and other assistance during childbirth.
These days, end-of-life doulas, sometimes called death midwives, are an emerging profession in the growing death positivity movement, which urges a paradigm shift for thinking and talking about death as natural and not inherently traumatic.
They provide nonmedical support to help ease the final transition for the terminally ill. But it’s not merely about that culminating moment, “The End.” They help the dying and their loved ones navigate death with all its “before and afters” – including sickness, acceptance, finding resources for all the legal housekeeping, funeral planning, and bereavement.
For Burney-Scott and Jeffries-Logan, it’s the highest calling.
Sisters in ritual, they performed sacraments of soothing and release drawn from their West African and Indigenous spiritual traditions. Burney-Scott is African American and was initiated in the West African Ife religious practice, and Jeffries-Logan is a member of the Occaneechi Band of the Saponi Nation, a tribe rooted in the North Carolina Piedmont region.
Being a death doula “is not fun. But it’s an honor,” says Burney-Scott, a healer and longtime advocate who most recently worked as a reproductive justice organizer in North Carolina.
She stumbled into the practice when her mother’s dear friend, a hospice nurse, showed Burney-Scott what to do at her mother’s passing.
“I didn’t want to do it,” she says. “The thing I feared most, from when I was a little girl and even when my mom was healthy, was losing my mother. She was that mom that all my friends would talk to, the mom who could let you know [you] were the most special person in the world even when she was yelling at you to do your laundry.”
Near the end, her mother made her retrieve a manila envelope containing her will, insurance information, deeds – the bureaucracy of death. But without ever using the word “doula,” her friend guided Burney-Scott in ushering out of this world the woman who had brought her into it.
“Aunt Cora” encouraged Burney-Scott to whisper her love in her mother’s ear, to hold her hand, play music, and to be present in “an organic practice.” One day, when her mother struggled to breathe, Cora assured Burney-Scott that she didn’t need to fetch doctors – that nothing was wrong.
“She’s leaving,” Cora told her, a simple statement that’s also a tenet of end-of-life care: Death can’t be controlled, but you can prepare for some aspects of it.
Because there is no universal or official training, no licensing and no regulation, there is no official estimate of how many death doulas operate in this country.
But death and dying are constant. And beyond the eulogies and coffins, there’s a clear and growing need for death-related services. The number of Medicare-approved home- and hospital-based hospices, for example, rose from barely 30 to slightly more than 3,400 between 1984 and 2009. A decade later, more than 4,500 exist, according to the Centers for Medicare & Medicaid Services.
Groups such as the International End-of-Life Doula Association and others train and certify doulas, providing hands-on experience, like a practicum. Still, many death doulas enter the field as Burney-Scott did, pressed into duty by a family member’s passing.
Few can make it into a full-time, paying job. Others have a background in the clergy or are people of faith, are volunteers involved in work with the sick and shut-in, or are shamans or healers.
Still others start end-of-life doulaing because they are nurses, midwives, or health care professionals who, through experience, have come to know that end of life is more than just what happens to your body.
Merilynne Rush, a nurse and home-birth midwife, co-founded Lifespan Doulas, an organization that trains and certifies end-of-life doulas. In three years, she says, the group has trained 200 people. She sees the need to educate and vet death doulas even while she thinks that community-trained doulas are valuable and necessary.
“There are so many people who are called in their communities [to do this] that no one should tell them they can’t,” Rush says. “I’d never be able to go into every community. That’s one reason for never having any kind of regulation that imposes a state-sanctioned structure that says you are in or out.
“At the same time, when you are working within a medical organization, they need to know you are OK and there are some standards,” she adds. “Training should never be mandatory, but optional.”
A diversity consultant who focuses on Native communities and trauma, Jeffries-Logan distrusts what she believes is a move toward professionalization.
Her death doula work is grounded in Indigenous customs, and communicating with the ancestors does not happen through curricula. Heeding a call from her ancestors, she did a traveling ceremony, designed to pave a deceased person’s road to the afterlife, for an infant relative who died before he turned a year old.
As part of a common tribal custom, she won’t speak the name of the deceased aloud for a year; to do so could keep the spirit tied to its temporal life – now a thing of the past – and distract it from the arduous journey to the ancestors.
Neither she nor Burney-Scott takes money for what they do. Rather, they extend their services to family and friends based on existing connections and an understanding that death is cultural and clinical. “It’s not like I was going to roll up and do this with just anyone. I don’t do shallow-ass relationships,” Jeffries-Logan says.
She questions what happens when the training moves out of informal community pedagogy and into a classroom.
“Who’s the certifying body? Who has the funds to pay for services?” she asks. She thinks of formalizing death doula work in the same vein as yoga, an Indian spiritual system that has been co-opted from communities of color and networks of caring to be dominated by White instructors who teach a fraction – the poses, the breathing – of the whole for pay.
Both women know that communities of color lag in accessing end-of-life care – whether due to cultural beliefs, experience and well-founded fear of racism in medical settings, lack of insurance or financial resources, or misconceptions about what’s available.
For example, Black people represented 8% of those receiving Medicare-funded hospice benefits in 2017, compared to 82% for White people.
In many Southern Black communities, people won’t talk about death, Burney-Scott offers. “There is truth in our mouth. You can manifest things with your word. Don’t talk about death [lest] you invite it in.”
That goes for other communities, as well. A 2010 study comparing Latino immigrant to White cancer caregivers found that the Latinos were surprised and even disturbed by transparent talk about death in hospice pamphlets and consultations.
Furthermore, Rush says that generally when death is imminent, “most people are overwhelmed and don’t know where to turn. They don’t even know that they can get hospice earlier. And even then, they may have a nurse come in for a few hours or an aide, but they aren’t there all the time. People have to rely on their community and network.”
And that’s just what Cynthia Brown did once she accepted that she wasn’t going to beat cancer, calling on the women her family members sometimes referred to as “Cynthia’s girls.”
“She invited us into the process from the very beginning. We swung into action on the logistical things: running errands, taking her to appointments, making meals,” Burney-Scott says.
“And then she said, ‘I want to cut my hair.’ She had 12 braids left. Each one of us cut two braids. Then, she called and said, ‘Hey, will you come over and help me write my memorial?”
She summoned Jeffries-Logan and another friend to help her assemble and bless her ancestors’ altar. With trademark precision and humor, she even planned who would cook at her funeral repast or meal: not her many loving White friends; she didn’t trust their chops in the kitchen.
Her death doulas and friends, in turn, called on each other, their own histories of loss, and their ancestors to help guide Brown through her own departure.
And when the end came, the friends all rolled to the hospital one last time. Burney-Scott donned her trademark white head wrap and packed a bag with crystals and Florida water, a citrusy blend believed to have calming properties.
Jeffries-Logan carried tobacco as an offering; red cedar to represent blood and life force; water from the Eno River, which courses through her tribal nation’s territory; and a ceremonial turtle rattle, used by tribes in special ceremonies.
“Cynthia fed me, I laid up on her couch, we carpooled to anti-racism trainings around the state,” Jeffries-Logan says, her eyes moist and a catch in her voice. “And when we did a ritual for my mother [who died from Alzheimer’s disease] in the ocean, Cynthia told me, since she had lost her parents at a young age and had to be like a mother to her younger siblings, she knew what it was like to be a motherless child. I was going to do whatever I could for her.”
She didn’t want her beloved sister-friend “scratching and clawing to stay here.” So she stroked the soles of Brown’s feet – which got cooler and cooler as death approached – not to bring back sensation, but to help untether her from this earth.
When Brown took her last breath, Burney-Scott’s and Jeffries-Logan’s hands were among those resting on her body. It was a fitting end: a social death for a community advocate who told her friends, “You continue to fight the good fight, and you have to promise me that you won’t leave anyone behind.”
One afternoon in the summer of 2018, Bob Gramling dropped by the small suite that serves as his lab in the basement of the University of Vermont’s medical school. There, in a grey lounge chair, an undergrad research assistant named Brigitte Durieux was doing her summer job, earphones plugged into a laptop. Everything normal, thought Bob.
Then he saw her tears.
Bob doesn’t baulk at tears. As a palliative care doctor, he has been at thousands of bedsides and had thousands of conversations, often wrenchingly difficult ones, about dying. But in 2007, when his father was dying of Alzheimer’s, Bob was struck by his own sensitivity to every word choice of the doctors and nurses, even though he was medically trained.
“If we [doctors] are feeling that vulnerable, and we theoretically have access to all the information we would want, it was a reminder to me of how vulnerable people without those types of resources are,” he says.
He began to do research into how dying patients, family members and doctors talk in these moments about the end of treatment, pain management and imminent death. Six years later, he received over $1 million from the American Cancer Society to undertake what became the most extensive study of palliative care conversations in the US.
The resulting database contains over 12,000 minutes and 1.2 million words of conversation involving 231 patients. This is the basis of the Vermont Conversation Lab, which Bob created to analyse this data and find features of those conversations that make patients and family members feel heard and understood.
Brigitte’s job in the lab that summer was simple: listen to moments of silence and categorise them. The idea was that they could indicate emotionally charged connections between doctor and patient. Once the silences were coded, they would be used to train a machine-learning algorithm to detect them automatically – and, with them, moments of emotional connection.
You might ask what place algorithms could possibly have in this sensitive realm. The reality is that healthcare communication needs help, especially in palliative care, where practitioners seek to bring patients to their deaths as meaningfully and painlessly as possible.
In 2014, the US Institute of Medicine made improving doctor-patient communication a priority in its landmark study, ‘Dying in America’. An analogous publication in the UK, Ambitions for Palliative and End of Life Care, emphasised the need for patients, family and caregivers to have “the opportunity for honest, sensitive and well-informed conversations about dying, death and bereavement”. It reiterated that doctors need to make those conversations possible.
Most of the resulting communications training seems to offer scripts and templates to help doctors deliver bad news and make decisions with patients. But this is not enough. In this context, doctors really need to understand conversations more broadly. They need to appreciate everyone’s role in a conversation. They need to learn the ability to listen and be silent. They need to confidently recover from conversational missteps.
“Oncologists are in general very uncomfortable with this kind of thing. They want to focus on treatment, and they talk eloquently about different protocols and clinical trials,” says Wen-Ying Sylvia Chou, a programme director in the Behavioral Research Program at the US National Cancer Institute. She oversees funding on patient-doctor communication at the end of life. “But sitting in the place of being a listener is not something that clinicians are trained for or necessarily comfortable doing.”
Enter Bob Gramling. Hospitals track infection rates, bed occupancy and many other measures. Why not good conversations, too?
Amiable and serene, wearing a bracelet of Buddhist meditation beads, Bob sees a big role for artificial intelligence (AI) that can detect and measure the features of clinical interactions that matter to patients, then report those measurements to numbers-oriented healthcare systems.
Once such technology is widely available, he says, “we can incentivise our hospitals to build systems to improve those interactions and reward doctors for doing it”.
“How are you?” asks the nurse practitioner, who’s just come into the patient’s room.
“Fine,” the patient says. She’s a 55-year-old white woman with stage 4 breast cancer. Neither she nor the nurse practitioner know that she’ll be dead in five days.
“No, you’re not,” the nurse practitioner retorts.
“Oh, a loaded question,” the patient laughs.
“It’s been a long – well? No,” says her spouse.
“No,” says the patient. “It’s a polite question, it’s a polite answer.”
This is a snippet of a conversation in Bob’s database that he played to his brother David, a linguistics professor at the University of Arizona. David recognised the dynamics of this specific moment. The people in that room hadn’t been talking about care or disease, but they had been doing something important in the conversation that would affect the quality of the care.
When the Gramlings’ father died, David flew home from a literature studies fellowship in Berlin. But years earlier, he’d been intimately involved as a caregiver, witnessing a “smörgåsbord of insane, irrational communication failures” with lawyers, nurses, nutritionists and priests.
For a year after their father’s death, the brothers were swallowed by family matters. As they emerged, they began talking about palliative care communication and linguistic research in healthcare settings, and began to collaborate professionally.
The most recent result is a book, Palliative Care Conversations, published in early 2019. It aims to show physicians how conversations work, such as how clinicians and patients often understand words and phrases differently. David looked at the conversations at a granular level, using the tools of a linguistic subfield called conversation analysis. He spent years listening to audio recordings of the conversations, noting moments worth closer analysis.
Meanwhile, Bob provided clinical details about medical culture. In the last few years, he has also hung out with jazz musicians, who are master communicators when they’re improvising, and visited the Stanford Literary Lab to see how digital tools can be applied to massive literary corpuses to understand patterns too diffuse for human readers to catch.
As the Gramlings note in the book, the above back-and-forth between patient, spouse and nurse practitioner is remarkable for a first exchange between strangers. They explain that’s because “the clinician is willing to risk conventional rapport-building pathways by contradicting the family member’s self-reported state of mind”. In other words, the physician has opened the door to a looser mode of relating – and it works.
Another conversation doesn’t go as well. It’s a “pragmatic failure”, as David would say.
“When I came in,” says the nurse practitioner, “I saw you were watching Scrubs.”
“Scrubs?” the patient says. He’s a 63-year-old black man with stage 4 kidney cancer, who will live for 135 more days.
“Have you ever seen Scrubs?” asks the nurse practitioner, who is white.
“Yeah,” the patient says. “No, I wasn’t watching Scrubs.”
As the exchange unfurls, it’s clear the patient and clinician won’t connect. The clinician then seems to want to force their way to the task at hand, and forget the small talk where rapport could be built.
“When you study communication in healthcare, you’ll see a lot of monologues from doctors,” Bob says. “I don’t mean that in an insulting way – it could be really good information.” In palliative care, he explains, conversations are different: “It might be just because it’s the nature of palliative care. It’s what we do and what our value is… there is a lot of turn-taking.” That’s another term he learned from his brother. It refers to the back-and-forth of conversation.
“This is not a clean, rational, logical experience that fits on an 8-and-a-half-by-11 piece of paper, it’s a human-engaged relational endeavour,” he adds. “If we’re going to develop metrics for that, we’d better be looking at both the beauty and the science from many angles.”
Research on end-of-life communicating and decision-making typically looks at what doctors or nurses say. It rarely takes into account the deeper linguistic and cognitive factors that influence patients’ abilities to communicate in the first place.
One study, by speech-language pathologists in the late 1990s, showed just how large these language challenges can be. They gave a battery of language comprehension and memory tests to 12 hospice patients: 11 of them couldn’t recall words, had difficulty understanding things and pronouncing words, and had difficulty remembering what was said to them. These symptoms get in the way of normal activities, like having conversations.
Even something as crucial as how well older patients can hear gets overlooked. In a 2016 survey of 510 hospice and palliative care providers across the US, 87% of them said they did not screen for hearing loss, even though 91% of them agreed that patients’ hearing loss impedes conversation and negatively affects the quality of the care they receive. Only 61% said they felt confident nonetheless that they could deal with patients with hearing problems.
The Gramlings pay a remarkable amount of attention to another factor: the pain, shortness of breath, fatigue and medications that can keep patients from communicating normally.
In his research, David has addressed what he calls “language in extremis”: what happens when people’s ideas about language and communication buckle under the strain of circumstances, as in multilingual experiences in Nazi concentration camps, or interpreting in border patrol detention facilities.
End-of-life medical conversations also often involve language in extremis. As cancer brings a person’s life near to its end, they may have lost some of their lifelong communicative powers to the disease or its treatments. They may have less ability to speak subtly and indirectly, which is important for politeness. Shallow breathing shortens utterances, and drugs may block word-finding. All of this reinforces an asymmetry in communication that doctors don’t always grasp.
A physician might encourage a patient to speak openly, and indicate their willingness to listen, but in practical terms, “That gesture doesn’t quite work,” David says, and doctors need to understand why.
At the same time, people still hew to lifelong social conventions about being a user of their language. They might be dying, but “They don’t back away from their interactional responsibilities,” David says. They honour turn-taking; they don’t interrupt. They tell jokes, they use family language, and they create mini-rituals of inclusion and exclusion, often to deal with the communication asymmetries.
“If I were picturing the developmental arc,” says David, “it wouldn’t be coasting down into death. It would be all the way and sometimes heightened. The kind of complex literacy you need to use in a hospital setting in a serious illness, and managing all your oncological terms – it’s almost like the competencies themselves get expanded in this end of life.”
In his lab, Bob is examining even more fleeting aspects of conversations, such as pauses. It’s an interesting choice, because pauses might be considered as a sign that a speaker has lost their way or that an interaction is breaking down. On the other hand, pauses are easy to locate in the acoustic signals of recorded conversations. And they might indicate where someone is listening or about to say something important, so they might be a good thing.
Bob’s team used machine learning to identify pauses of 2 seconds or longer in spoken conversations, then human coders like Brigitte Durieux tried to categorise them, looking for ones that were more than just silence.
Because they didn’t have access to what the doctors or patients were actually thinking, they looked for the presence of emotional words and other sounds like sighs or crying on either side of the pause. Did a question about the quality of life, treatment hopes, prognosis or dying precede the pause? If so, the pause may have been because the doctor invited the patient to consider something.
The team found that during some of these pauses, some connection, shift or transformation was occurring. These “connectional silences” were rare. Out of a set of 1,000 clips with pauses, a mere 32 were connectional in nature. They were brief, as well, most lasting less than four seconds. But there’s still power in them.
The dynamics of a conversation change dramatically after such a connectional silence. Suddenly, a patient will be talking more than they did earlier. They’ll be directing the conversation, not the doctor. It’s as if the mutual agreement to pause for two seconds spilled into an agreement to shift roles.
“No, for some reason I guess I just in my head was gonna be on [chemotherapy] for the rest of my life and everything was gonna be hunky dory and…” a patient begins.
A 2.9-second connectional silence follows. The doctor inhales audibly, to signal they will respond, which makes the patient pick back up.
“You know. I knew early on, I mean you told me early on it’s not like and then this will be the rest of my life. Something, you know, might go down.”
The doctor responds. “Something. That can be a very hard thing to think about. That here we found something that’s helping but you can’t stay on it for the rest of your life.”
In other moments, the silence comes after a doctor has said something empathetic.
“It’s rare of me to tell somebody point-blank you’ve got to stop. However, I will say you have my permission to set limits,” the doctor says.
“Okay,” says the patient, then falls silent for nearly seven seconds.
His wife chuckles. “He can’t stand the thought of it. I can tell by his laugh,” then she laughs.
“I know he can’t stand the thought of it,” the doctor says.
“No, that’s okay,” the patient says. “I’ll get used to it.”
Or in another instance, the doctor tells a patient’s spouse, “what you feel is really hard. It’s really hard.” There’s a 2.8-second silence.
“I just wish he had a better quality of life.”
“I know, I know,” says the doctor.
Even though these connectional silences don’t happen often, Bob thinks they’re good linguistic markers of connection exactly because doctors don’t commonly use them. When someone good at monologuing and interrupting falls silent, it may mean they’re allowing something else to happen.
Bob surmises, “More often than not, the conversations that have a lot of space in them are probably going to lead to people feeling more heard and understood.”
Judy had a question. Having come to the hospital at the University of Vermont to recover from the flu, this elegant, 83-year-old woman was lying in her bed. Two doctors had come to her room bearing news. It was cancer, not the flu, and it had spread from her liver. She could undertake a course of chemo, or she could have her pain managed as she died.
She turned to her daughter, Kate, sitting beside her. “What should I do?” she asked.
When the doctors had requested this meeting, Kate had dropped everything to be there. It seemed unusually serious. Now she knew why. She wondered why she hadn’t seen the signs of her mother’s cancer. Judy’s skin had started to look yellow, she recalled. But instead of recommending a check-up, she bought her mother some pinker make-up.
In this pivotal conversation, the doctors presented the options but also wanted to know what was important to Judy. They knitted the science together with thoughtfulness and compassion. Kate was struck by their slow, almost languid approach to delivering the news.
Slowly it dawned on her that this was a conversation about her mother’s death. Neither of them had prepared for this. Not now, not so soon.
“It had the nature of a conversation with a clergyperson rather than a doctor,” she remembers. Pastoral kept coming to mind.
At the end of the conversation, one of the doctors gave her his card. It was Bob Gramling. Kate has since seen the bright blue spectrographs showing gaps in conversation – where the pauses occur. She thinks these are important moments as well.
“Where there’s silence, where there are gaps, that’s where the caring shows up,” she says. “I think it’s incredible work to point out to doctors there’s a lot going on in the silences.”
Bob and David have only scratched the surface of how these conversations work. So far they have only studied English speakers, for example; pauses work differently in other cultures, so they need data on those moments, too. And because their data comes from people with cancer, there’s a concern that the analysis may be skewed.
With cancer, says Wen-Ying Sylvia Chou of the National Cancer Institute, most patients have time: “They continue to be themselves and continue to be part of the conversation and any ongoing discussion.” With other diseases, though, there could be more risk that the person would “lose cognitive function or physical function”. In those cases, she says, conversations “would look very different”.
Healthcare’s use of natural language processing – technologies that treat language as data – is expanding, and the chances are good that research like that of the Gramlings will expand to cover conversations with people who have other serious illnesses.
Bob isn’t the only researcher exploring the use of artificial intelligence in palliative care. In 2017, James Tulsky, a palliative care physician at Dana-Farber Cancer Institute in Boston and a Harvard professor who studies health communication, stressed that “mass-scale, high-quality automated coding will be required” to give feedback that helps clinicians improve their expressions of empathy.
Tulsky turned to Panayiotis Georgiou, a computer engineer at the University of Southern California, to develop automated detection of emotional connections between doctors and patients. In 2017, a team headed by Georgiou showed that certain acoustic features of the speech of couples in counselling could be used to predict marital outcomes. What if algorithms could do the same for palliative care conversations?
“The technology in theory exists out there to do all this,” Tulsky says. “It’s just a matter of doing enough research, running enough iterative trials, training up the machines to actually get these algorithms trained well enough so you could apply them to more random talk.”
I ask Judy’s daughter Kate what she thinks of using artificial intelligence to enrich human connections. “I wouldn’t worry about the technology,” she says. “The more technology, the more sacred the conversation becomes.”
What does she mean? Anything that enables humans to use their voices more effectively with each other is a good thing, she explains: “It’s because of the increasing technology that the interaction becomes more wonderful.”
What is a conversation? It’s a setting where humans interact, often for a purpose but sometimes for none at all. People have to learn how to have conversations but when they become expert in their culture’s conventions, conversing becomes so automatic it feels natural.
Modern healthcare has hijacked conversation and made it a tool by which physicians can achieve their ends.
According to David, “The contemporary hospital still understands ‘conversation’ as ‘making a pre-determined X happen through conversation’.” This is a barrier in serious illness and end-of-life care, where the conversations need to be venues for figuring out what the X might be.
At the end of a patient’s life, there may not be effective medical treatments, just things to discuss and plans to make. This may need a more natural conversation than a medical one, a conversation in which none of the participants may know what the outcome will be.
After all, these conversations aren’t just for doctors; they’re for patients, too. And family members, nursing aides, housekeeping staff. “There are a lot of human beings who have a vested interest in this other human,” Bob says.
There are critics who don’t think artificial intelligence and machine learning have a role to play in palliative care. Bob’s view is that shying away from analysing this kind of conversation in this way means that essential opportunities for improving it will be missed.
“It is helpful, as a discipline that has historically thought of communication as just the art of medicine, to actually think that, no, this is a science,” he says. And understanding that science could help us re-engineer the healthcare system to support more meaningful conversations.
He’s aware of the delicacy in institutionalising and commodifying a human interaction, though. “As a physician,” he says, “I was afraid of being a researcher that was going to oversimplify this kind of sacred experience into something that’s measurable and convenient and essentially meaningless.”
That’s where Brigitte Durieux struggled with her feelings as she listened to thousands of audio clips of pauses. In some conversations, people were laughing, but she was struck by the loneliness in others. She had begun to recognise patients’ voices and wondered what had happened to them.
“Nobody is perfect, but there are times when one realises there’s something that could be said to make this feel less like a loss,” she says. Sometimes she whispered under her breath something the doctors could have offered instead.
After Bob found Brigitte crying, he wrote an ethics proposal to the hospital so that he could introduce a new procedure into his lab. He borrowed an idea from the hospital’s palliative care unit, where staff gather every week to say the names of people who have died, then ring a singing bowl.
Now, at the start of every Vermont Conversation Lab meeting, a researcher reads the name of one of the patients from the database and rings the bowl. So far, they have gone through the list of names twice.
The ceremony helps, says Brigitte, because it reduces the guilt of turning a sensitive moment in someone’s life into a piece of data.
“What it does ultimately,” she says, “is recognise the humanity of things.”
When Maine passes a law allowing physician-assisted dying (PAD), it will be joining nine other jurisdictions in the United States.
By October, one in five Americans (22%) will have a law that allows terminally ill patients, most of whom have cancer, to choose an end to their life with medical help from a doctor.
However, the practicalities of actually doing so are formidable, and patients who choose this option find there are many obstacles in the way.
First is finding a doctor who will participate. Many doctors have moral objections to PAD, refuse to participate, and will not refer patients.
This sounds familiar to Charles Blanke, MD, professor of medicine at the Knight Cancer Institute at Oregon Health and Science University in Portland, who has been participating in PAD since it was legalized there in 1997.
Blanke says patients have told him that after being turned down by their physician, they also were not given a referral; instead, they were told by their doctor that “they don’t know anyone, and good luck finding someone.”
I believe this is patient abandonment.
Dr Charles Blanke
“I believe it is patient abandonment,” Blanke told Medscape Medical News. “For some patients, it takes them months to find me, so it’s no wonder many are too ill by then to proceed.”
In general, eligible patients say that PAD was not offered to them, Blanke said, but he argues that “it is legal and should be put on the table.”
He emphasized that physicians should never be pressured to participate in PAD, but they should refer patients. “We need to make it more patient friendly and more accessible.”
For years, Oregon was the only state that allowed the practice.
In recent years, however, other states have passed similar laws — Washington in 2008, Montana in 2009, Vermont in 2013, California in 2015, Colorado in 2016, Washington, D.C. in 2017, Hawaii in 2018, and New Jersey just a few weeks ago.
Lack of Training
That some doctors do not want to participate in PAD is understandable; many have moral objections to the whole idea, citing the Hippocratic oath to ‘do no harm.’
But there are signs of a shift toward more acceptance.
For instance, a 2018 Medscape ethics report found that 58% of doctors who responded to the survey said physician-assisted death should be available to the terminally ill, similar to 57% in 2016, and up from 54% in 2014 and 46% in 2010.
However, doctors who are willing to participate find it difficult to do so.
“The law makes no provision for medical training, there is no formal system, and I believe that is one of the major barriers and a shortcoming of the law in every state where it is legal,” said Lonny Shavelson, MD, a California physician based in the San Francisco area who specializes in aid in dying. He founded Bay Area End of Life Options in 2016.
“I agree that sometimes there is a moral objection, and there is sometimes institutional resistance, but most commonly it is lack of training,” he said.
Doctors, as a rule, like to do things they’ve been trained in.
Dr Lonny Shavelson
“Doctors, as a rule, like to do things they’ve been trained in and don’t like to do things they haven’t been trained in,” he added.
He noted that his practice has received more than 800 requests for medical aid in dying from different patients throughout California.
“Every patient who comes to us does so because they can’t find another doctor,” he said. “Everyone thinks it’s because of moral objections or that the patients live in rural communities, but it’s not the case for most of the patients.”
Shavelson told Medscape Medical News that he always calls the patient’s doctor, and most of them are not morally opposed to participating in PAD. “But what they tell me is that they’ve never been trained and that they don’t know anything about it. They don’t know what medications to use, or anything about the paperwork or protocol,” he said.
Barriers To Access
“The great news is that we have 22 years of data in Oregon, and the law is protecting patients,” says Kim Callinan, CEO of Compassion & Choices, the largest national advocacy group for aid in dying.
“But we also have robust data showing that the law is not meeting its intentions and that we have erected too many barriers for many to access it,” she told Medscape Medical News.
Callinan believes that improvements are needed to allow the original intention of the law to take place. “We want to keep the right safeguards in place,” she said. “But we are seeing such small numbers of people using it, and in many cases it’s because they can’t get access.”
Recent reports confirm that the number of patients who have chosen PAD — and who have completed the process — remains small.
For example, data from Oregon show that from 1997–2018, prescriptions have been written for 2217 people, and 1459 patients have died from ingesting the drugs.
In California during a single year (2017), 577 individuals received prescriptions and 374 people died after ingesting the medication.
Shavelson feels the actual demand for PAD is not reflected in the current statistics, and the numbers would probably be much higher if there was more access to physicians.
He argues that a more accurate survey would be to identify how many patients have requested PAD but could not find a physician to help them, he said. Shavelson believes that number would be significantly higher than what has been documented.
In some cases, it is not the physician making the decision but the healthcare system.
A recent survey of 270 California hospitals, conducted 18 months after implementation of the state’s End of Life Option Act, found that 61% of hospitals had a policy forbidding physicians to participate (JAMA Intern Med. 2019;179:985-987).
“We found that of the 164 hospitals in California that opted out, 56% allowed physicians to refer patients to another provider and 29% of hospitals did not provide any guidance on this question,” said lead author Cindy Cain, PhD, assistant professor in the Department of Sociology at the University of Alabama at Birmingham.
“I support the idea that a health system can opt out,” says Peg Sandeen, PhD, MSW, executive director of the Death with Dignity National Center, a nonpartisan, nonprofit organization. “As much as I don’t like it, and think physicians should be free to practice, the health system has that right to do so,” she said.
However, not referring patients is an entirely different issue. “The outright act of refusing to refer a patient puts the physician into an ethical quandary,” she said. “Referral is part of how medicine is practiced, but it is up to the individual physician to make that determination.”
Waiting Times Present Another Barrier
The whole PAD process requires two oral requests with a waiting time of at least 15 days between them, and also a written request using the statutory form included in the state’s aid-in-dying law.
There are slight variations among states (eg, Washington, DC also requires two witnesses). Many states also require a second waiting period, in which the physician must wait 48 hours from the time of receiving the written request to write the prescription.
Callinan believes that the waiting periods, as well as the need for two doctors to confirm eligibility, are redundant in some cases. “The eligibility is that a patient has 6 months or less to live, and 2 doctors have to certify that,” she said.
“But if someone is already enrolled in hospice, as many are, it has already been determined that they meet the 6-month criteria and that the decision has been made to forgo treatment. In this case, they should only need one doctor to authorize it,” she argues.
A new law in Oregon may cut some of the waiting time, as it allows physicians to make exceptions to the waiting periods if the patient is likely to die before completing them.
“Oregon law has not evolved since it was written 20 years ago,” said Blanke. “This new bill will eliminate the waiting period for those who are imminently terminal. It won’t affect very many people, but it will help a few get quicker access.”
Shavelson praised the new Oregon law. “I think the 15-day waiting period is obscene because it’s not 15 days,” he said, explaining that it may be more like 3 or 4 months, as patients have to find a doctor and then may have to wait weeks for an appointment.
“The idea was that it was supposed to be a period of contemplation, but many patients have been contemplating since they got their diagnosis,” Shavelson pointed out. “They didn’t start thinking about it when they first made their request — they have been thinking about this for a long time.”
Patients in this waiting period may be dying or losing the mental and/or physical ability required for self-administration of the drugs, he explained. In his own clinical practice, about 30% of patients die during the 15-day waiting period, he estimates.
This is a similar proportion to that found in recent study from Kaiser Permanente Southern California, where one third of patients became too sick or died before the process was completed (JAMA Intern Med. 2018;178:417-421).
Accessing and Taking the Drugs
Even for patients who do manage to get through the bureaucracy, there are challenges in the practical steps of actually obtaining the drugs. A physician can only write the prescription and it is up to the patient to procure the drugs.
When states began to first legalize PAD, the drugs of choice were oral pentobarbital and secobarbital. However, as of 2015, both of these drugs have been largely unavailable, as previously reported by Medscape Medical News.
Through trial and error, a group of physicians eventually developed a drug regimen (DDMP2), which contains diazepam 1 g, digoxin 50 mg, morphine 15 g, and propranolol 2 g. It is more complicated than the barbiturates but has been found effective.
Shavelson explained that an updated version known as D-DMA (no propranolol and amitriptyline 8 g added), which is both faster and more reliable than all other protocols, is in the process of replacing DDMP2.
Both formulations are compounded by a pharmacist and available as a powder, which then must be mixed with 4 oz of apple juice and taken as a liquid/suspension.
Shavelson noted that physicians may not know where a patient can fill the prescription.
“It’s not something that can be filled at the local CVS or Walgreens,” he said. “A regular pharmacy doesn’t have the ingredients on hand, and for the DDMP2 combination, it has to be compounded.”
In California, two pharmacists currently fill about two thirds of the prescriptions. “Pharmacists need training as well,” Shavelson contends. “They are an integral part of this process.”
Even the last step in the whole process, the actual ingestion of the drugs, can be difficult for some patients.;
State law requires that the lethal dose be self-ingested via the digestive tract (orally or through an nasogastric (NG) or gastrostomy tube). The restriction that the drugs must be self-administered was to help ensure no one could harm a patient against his or her will.
However, many terminally ill patients are so sick they can’t physically mix the solutions, pick up and take the medicine, or swallow the drugs. Blanke estimates that around 10% of the patients he has evaluated have swallowing issues, and they fear that they will be unable to swallow the medications when they are ready to die.
To get around these practical difficulties, a proposed bill in Oregon sought to allow patients to self-administer intravenous drugs.
“There are many people who cannot swallow or administer through an NG tube, so just pushing the button on a pump syringe would allow them to take the medication,” said Blanke. “The IV could be put in right before they used it.”
Putting in an IV is easier than an NG tube, he explained, and much less invasive than a gastrostomy tube. “There’s really no difference between them, as far as putting medication in,” Blanke said. Both require some intervention and hold the same risk that someone else can administer the drugs.
Although the bill passed through the Oregon House of Representatives, it stalled in the Senate and has not moved forward. Some opponents of the bill feared that it would move Oregon closer to allowing euthanasia, while others cited the high cost of pump syringes.
Blanke believes that much of the opposition was really directed at the concept of assisted dying. “The arguments were with Death with Dignity,” he said. “Not the idea of making changes in the law or the use of an IV.”
The practical difficulties of PAD in the United States contrast with a much simpler process in Canada. Since 2016, Canada has legalized medical assistance in dying, which allows for both physician-assisted euthanasia and self-ingestion of a lethal dose. Patients have overwhelmingly selected physician-assisted euthanasia, where the lethal dose is administered intravenously by a clinician. According to Health Canada, of the nearly 7000 Canadians who have chosen to end their lives since the law went into effect, only six people have opted to self-administer drugs.
Physician Education and Training Needed
The biggest barrier — and the most imperative need — is physician education and training in PAD, argues Shavelson.
“Traditionally, teaching happens at large institutions, medical schools, universities, academics — but they won’t touch this,” said Shavelson. “They don’t want their reputation so-called ‘sullied,’ and are frightened that their reputation will take a hit. I don’t think that’s true, and I think people would feel that it’s a good thing to have medical centers more involved in this.”
Academia has fallen down on their responsibility, he contends. “This is a legal medical procedure and there is not one medical institution in my state [California] that is doing formal training on this. It’s not part of any conferences or any continuing medical education.”
As an example, the University of California, San Francisco, forbids palliative care residents and fellows from participating in aid-in-dying practices. The end result is that there are palliative care fellows coming out of training who have no experience in this area.
“Their patients will be asking about it, since palliative care doctors get asked about it more than any other specialty except for oncology,” said Shavelson. “So we will have palliative care and hospice doctors who have no training in it, and that’s absurd. This is part of the realm of what they are going to have to deal with in their practice, and institutions have forbidden it.”
However, next year the first conference for clinicians on medical aid in dying will be held in Berkeley, California, and will really delve into the nuts and bolts of practicalities, Shavelson explained. “The topic has come up at conferences, and there have been other gatherings to discuss it, but the focus has been on policy and ethics.”
Some hospitals and hospices have policies that forbid nurses to be part of the process or even to discuss end-of-life options.
By Emilie Le Beau Lucchesi
When Ben Wald, 75, was dying of cancer in 2012, he wanted to use Oregon’s Death with Dignity Act to receive a prescription for a lethal medication that would end his life. His hospice nurse, Linda, was part of the discussion and provided both information and support, said his wife, Pam Wald, of Kings Valley, Ore.
His colon cancer had spread to his lungs, and his weight dropped from 180 to 118 pounds. He struggled to speak or eat.
When he was ready to end his life, the couple wanted Linda with them, but the hospice organization she worked for did not allow it, Mrs. Wald said. The organization allowed other hospice workers, such as social workers and massage therapists, to be present, but not the doctors or nurses it employed.
Without a nurse present, Mrs. Wald was going to be alone with her husband when he died. She wanted someone knowledgeable to support her through the process. She reached out to Compassion & Choices, a national advocacy group for aid in dying. The group paired her with two volunteers, one of whom was a retired intensive care nurse.
“You watch your husband die and you hear that change in breathing,” Mrs. Wald said. “Jane, the I.C.U. nurse, she said, ‘The breathing changes. You’re doing fine, Pam. Keep holding his hands.’”
But many may still run into the problem the Walds had, because some hospitals and hospices see medical aid in dying as conflicting with their traditional mission of protecting life and avoiding harm to patients. Those that are faith-based typically follow church policy against medical aid in dying.
Some have policies that forbid nurses even to discuss end of life options. Others hold a “neutral” stance on aid in dying, but bar doctors or nurses from being in the room while a patient self-ingests the medication and begins the dying process.
“We want to be clear: Nurses absolutely do not have to be present or provide that comfort if they feel they have a moral or religious objection. Our code of ethics states they have the right to object,” Ms. Stokes said.
But for those who wish to support their patients, the new statement defines key words such as “participation” and “presence.” These definitions are meant to encourage organizations to be clearer in terms of a nurse’s ability to answer questions during the decision-making process or offer support in the final moments.
Currently, Ms. Stokes said many company policies are vague and difficult to interpret. Even terms like “to witness” or “to be present” may be open to interpretation. Ms. Stokes said the association has received inquiries from nurses wondering if covering their eyes qualified as not witnessing.
Of the policies analyzed, 78 percent prevented nurses or other staff members from being present during or after the prescription was taken. The authors described the policies as “relatively silent” about the rationale for their decisions, but some referred to medical aid in dying as being “outside the scope of hospice practice.” Others did not want to be seen as “taking sides.”
The hospices that allowed staff members to be present made note of the core hospice value of not abandoning patients.
The study found that although the policies tend to be vague, there is a clear distinction between the role of the physician and that of the hospice. Physicians who write the prescription might not be employed by the hospice and therefore not subject to the organization’s particular policies. Policies note that physicians have a responsibility to respond to any complications that might occur after the prescription is ingested.
Each jurisdiction that permits medical aid in dying publishes annual reports on who took the medication, and why, where and whether medical practitioners were present.
In California’s report for 2018, only 54.3 percent of aid in dying patients were reported to have a medical care provider present at the time of ingestion. In Oregon’s 2017 report, only 33 percent of patients did. Many of these medical providers did not remain at the bedside, and 70 percent of patients in Oregon did not have a provider present at the time of death.
Keith Seckel, a registered nurse in Corvallis, Ore., believes it can be helpful to have a medical practitioner present. He has taken care of about a dozen patients who utilized their state’s aid in dying law. He was with them and their families as they took the lethal medicine and died. Mr. Seckel said a nurse is helpful in managing a patient’s discomfort or pain before taking the medication.
Many patients at the end of life experience anxiety, constipation, nausea, pain or shortness of breath. A patient who is short of breath, for example, might get anxious about swallowing the medicine for fear of choking. A nurse can provide reassurance, which Mr. Seckel said takes the pressure off the patient and family members to “get it right.”
He said that having a nurse in the room can also ease the stress for family members, particularly when their loved one makes unfamiliar sounds or unexpected movements.
Mr. Seckel said the timeline varies for each patient. The patient usually takes an anti-nausea medication anywhere from 15 to 60 minutes in advance. In all jurisdictions, the patient must administer the medicines themselves. Nurses and physicians are prohibited from assisting.
Mr. Seckel said some patients then take an anti-anxiety medication before the fatal dose. Within minutes, patients typically report feeling drowsy.
“I might offer to the patient, ‘If you can feel it hitting you, if there is something you want to say, say it now,’” Mr. Seckel said.
The patient then slips into unconsciousness. Mr. Seckel said he watches for signs of discomfort or pain. Some family members ask him for updates as their loved one’s breathing begins to slow or color drains from their skin. Others, Mr. Seckel said, are too connected to the moment to ask questions but want to review the experience with him later.
Because the laws clearly state that a patient must be able to take the medicine without assistance, Mr. Seckel said patients often have questions about their disease progression and how much time they likely have until they can no longer take the lethal medicine on their own. Often, the role of the nurse is to give patients information so they can determine a timeline for themselves.
He said there have also been times when he has been called to the bedside after the patient passed. He said it’s not uncommon for family members to want confirmation that their loved one is truly gone. “I’ve had more than one person say, ‘I’m glad you were there, we wouldn’t have known what to do,’” Mr. Seckel said.
A palliative care physician struggles with the complex realities of dying at home, and the unintended consequences of making it a societal priority.
By Richard Leiter, M.D.
“If time were short, where would you want to be?”
As a palliative care physician, I regularly ask my patients, or their family members, where they want to die. The specific language I use depends on what they know, what they want to know and how they process information, but the basic premise is the same. Having asked this of hundreds of patients, I have come to expect most will tell me that they want to be at home.
But recently I have struggled with the complex realities of dying at home, and the unintended consequences of our making it a societal priority.
It is emotionally and intellectually compelling that patients should die in their own homes, surrounded by loved ones in a comfortable, familiar environment. For patients dying of end-stage disease, be it cancer, heart disease or something else, even the best hospitals are unlikely to be able to “fix” the underlying problem. We worry that people will go through expensive and potentially painful tests and interventions that have little chance of changing the ultimate outcome. And the opportunity costs are high; time waiting for a scan or procedure could be spent getting financial affairs in order or saying goodbye.
While there are still those who subscribe to the idea that excellent health care demands doing everything possible to prolong a life, many doctors and patients now prefer a less intensive approach when time is short. Rates of hospice enrollment have increased and the home has re-emerged as a place to die, not only preferred by patients and families but also heavily recommended by clinicians, especially in my field.
The system is imperfect, though. Unless a family has the significant resources necessary to hire aides or nurses, informal caregivers become responsible for nearly everything — from feeding to bathing to toileting. These tasks often get harder as the dying person weakens. In my experience, most family members want to care for their loved ones at home, but many are unaware of caregiving’s physical and emotional toll. And the length of time a patient spends in hospice care is difficult to predict, sometimes requiring caregivers to take significant time away from work or other family members.
Complicating matters, I frequently detect ambivalence in patients who tell me they want to die at home. Some are comforted by the reliability of the nursing care and easier accessibility of IV medications in the hospital. For others, dying at home may not be their top priority. Parents may want to protect their young children’s physical space from death. Similarly, one patient’s wife told me through tears that their adult son had died suddenly in their home a few years earlier; she could not bear the thought of watching her husband die in the same place.
We should not be surprised, then, that some patients who do enroll in hospice end up back in the hospital. And yet we in palliative care often view these cases as failures. We wonder what the critical gap was that led the family to call 911 or come to the emergency department. Was the patient’s pain uncontrolled? Were medications unavailable? Did the family panic? Something must have gone wrong.
I wonder, though, if we’ve adopted the wrong approach. As a doctor who regularly asks my patients where they “want” to die, I often worry about what this will look like if they choose home. I am concerned about the unacknowledged caregiving burden for families and friends. In addition, many people with advanced disease experience escalating symptoms, like pain or shortness of breath, that even the best hospices have difficulty managing in the home. In these situations, I am caught between the passionate rhetoric of my field, the spoken and unspoken wishes of my patients, and my clinical judgment. The patient in front of me always takes precedence, but my cognitive dissonance is difficult to escape.
To be sure, dying in the hospital has its own trade-offs. Though we can make more, and faster, medication adjustments, severe symptoms can be difficult to treat regardless of the setting. And as much as we try, it’s nearly impossible to alter the health care system’s usual rhythms. Overflowing hospitals often lack the flexibility to give dying patients the privacy of a single room. We cannot guarantee that they will not be woken up by the squeal of a malfunctioning IV pump or the chaos of clinicians scrambling to help another patient. Family and friends may live hundreds of miles away, preventing them from being with their loved ones at critical moments. While an inpatient hospice facility, which represents a third option, can provide hospital-level care in more of a homelike environment, Medicare and other insurance providers have set a high threshold for the few available beds. Most patients are only eligible if they are in the last few days of life or have severe, uncontrolled symptoms that would otherwise require hospitalization.
This dilemma entered my personal life earlier this year. The caregiver for my 96-year old grandmother found her slumped over and unresponsive in her wheelchair in her apartment, where she lived alone, but with the support of aides around the clock. She did not regain consciousness, and the paramedics arrived to take her to the hospital. When my uncle called to tell me what was going on, I was unsure of how to respond. My grandmother’s health and cognition had been declining over the past few months, but her quality of life was still good. In that moment, though, my clinical intuition was that she was dying. As a palliative care physician, wasn’t it now my job to protect my grandmother from spending what could be her final hours in a hospital? On the other hand, without seeing her how could I be sure that whatever was happening could not be fixed? With uncertainty and emotion clouding my judgment, I froze.
The paramedic took the phone and gently explained that he wanted to ensure my grandmother had all the care she needed, whatever the outcome. Taking her to the hospital was the right decision. The doctor in the emergency department empathically told us he thought my grandmother was dying and recommended we focus on ensuring that the short time she had left was as comfortable as possible. The nurses quietly checked on her throughout the night, looking for any signs of distress. My grandmother died the next morning — in the hospital and at peace.
The quality and consistency of end-of-life care are not where they need to be. To ensure that all people receive the same compassionate care that my grandmother did, we need to focus not only on where, but also on how they die. When we view all deaths in the hospital as failures, we risk neglecting a critical opportunity to improve the dying experience for many of our society’s sickest and most vulnerable. Clinicians across medicine should elicit and, whenever possible, honor their patients’ preferences for where they want to die. At the same time, we need to acknowledge our own uncertainties and be honest — with ourselves and our patients — about the difficult trade-offs these choices entail.