End-of-Life Care for Homeless Patients: Title and subTitle Break“She Says She Is There to Help Me in Any Situation”

Mr K is a 66-year-old African American man treated in the palliative care clinic of a university hospital for metastatic renal cell carcinoma. Mr K has been chronically homeless for most of the past 50 years. He sells drugs to support his daily cocaine and heroin use. He was in relatively good health until February 2002, when he presented to a nearby hospital with flank pain. The diagnosis based on biopsy results was renal cell carcinoma. The lesion was localized to one kidney. For unclear reasons, resection was made contingent on cessation of drug use. After this examination, Mr K was lost to follow-up for almost a year, when he then presented to the emergency department with abdominal pain and heroin withdrawal. When questioned about the delay in treatment, he said that thinking about the diagnosis made him anxious. He was treated for severe hypertension with labetalol, and further evaluation showed bone metastases. He was evaluated for chemotherapy, but given the lack of anticipated benefit, Mr K opted to forgo it. In April of 2003, Mr K was first referred to the palliative care clinic. His physician, Dr E, focused on the challenges of controlling his pain while Mr K continued to use heroin and broached planning for the end of his life because Mr K lacked close friends and family and had no home. Although Mr K declined to enter a methadone program and has no stable place to stay, he keeps his appointments and has become more comfortable with clinic staff. He speaks frankly about his drug use, and as a consequence, he and his physicians created a safe and effective treatment program for pain and other symptoms. About his prognosis, he observes, “I did not expect to make it to age 21, never mind 66.” He feels that he has had a full life, and his goal is to live as well as he can in the time that remains.

Mr K and his palliative care physician, Dr E, were interviewed by a Perspectives editor in October 2004, and Dr E was reinterviewed in August 2006.

DR E: Mr K has been a very heavy drug user. He continues to [use] heroin and was living with a lady friend . . . but their relationship has come apart now. He is . . . probably going to be relocating to a shelter.

MR K: That's a hassle, trying to keep up with my medicine because I got to kind of keep it bagged up, and my bag gets teared up. Sometime it just made me want to give up . . . ’cause it is so strenuous. Sometime, you can't take your medicine and you know you have to take it. Then you can be dealing with death more, . . . but thank God that so far I be able to hold on and take it.

Homelessness annually affects an estimated 2.3 million to 3.5 million individuals living in the United States. Approximately 10% to 20% of these will become chronically homeless, defined as remaining homeless for more than 1 year.^{1 - 2} Homeless people at the end of life face numerous challenges, including lacking a safe place to store medications and inadequate transportation. Mr K, like other homeless patients, has difficulty prioritizing health care over more compelling needs.³ Like Mr K, African Americans and men are overrepresented among the chronically homeless.⁴ The chronically homeless have little to no social support and sparse contact with their families.⁴ But Mr K was additionally burdened: in the face of terminal illness, he dealt with addiction, and had few choices about where to receive end-of-life care.

In this article, we discuss the challenges of providing health care for homeless people and offer specific management recommendations. These include information on shelter options, improving medication adherence, approaches to pain management for patients who concurrently use illicit drugs, and legal issues surrounding use of opioid analgesics among patients at high risk for substance abuse. We discuss particular issues attendant to goals of end-of-life care and advance directive planning. Although published literature regarding end-of-life care for the homeless population is limited, we cite the best available evidence. Where no data exist, we share our experience and suggest practical steps clinicians may use.

With poor access to ambulatory care, homeless people have rates of hospital use 3 to 4 times greater than those expected.^{5 - 7} Homeless people comprise an estimated 26% of patients seen at one urban public hospital⁸ and have lengths of stay 36% longer than poor nonhomeless patients admitted with the same problems, leading to an excess cost of $2414 per admission.^{9 - 10} More than 50% of homeless persons in the United States lack health insurance. The remaining half are generally covered under a public program such as Medicaid or Medicare or Veterans Affairs insurance.⁵ An estimated 62% of homeless people have a lifetime history of alcohol problems, 58% have drug problems, and 57% have mental health issues.¹ Among the chronically homeless, approximately one third report that their health is fair or poor^{11 - 12} and premature mortality is 3 to 4 higher times than expected.^{13 - 16} In the United States, the majority of homeless people, as with housed individuals, die in hospitals.¹⁴

DR E: Because he frequently doesn't have a mailing address or telephone, we get his appointment written down on paper before he leaves. . . . He’ll voluntarily walk in if he is feeling unwell. He . . . knows where our social worker and nurse [are], and if he has any problems, he walks right up and finds them.

The daily need for a homeless patient to secure housing, food, and transportation complicates the delivery of health care. Incorporating an assessment of daily needs into clinical care is thus critical ( Article ). In health care settings, social workers, discharge planners, or case managers have expertise regarding local resources and are central to a multidisciplinary approach to caring for homeless patients. By acknowledging the realities of the patient's life and introducing end-of-life planning over repeated visits, the clinician and the team have an opportunity to build trust. Of course, building trust is hampered by the prevalence of mental illness, substance abuse,^{17 - 18} and chronic illness in this population.^{2 ,19 - 20} Nevertheless, as illustrated by Mr K and Dr E, these problems are surmountable.

Conduct a complete needs assessment of competing priorities, including the following:

Shelter

Food

Communication

Transportation

Money

At diagnosis, Mr K's original physician offered him potentially curative surgery, contingent upon his agreement to stop using illicit drugs. The rationale to deny surgery is not stated but in all likelihood was both medically and ethically inappropriate. From the clinical standpoint, while heroin and cocaine use can complicate anesthesia, neither is an absolute contraindication.²¹ During hospitalization, heroin withdrawal can be prevented with opiate replacement therapy (methadone, levomethadyl acetate, or buprenorphine).^{22 - 25} One potential explanation for the physician's response was that he viewed drug use in moralistic terms.²⁶ This reaction may have been a factor in Mr K's not seeking follow-up until his pain became intractable and curative options were no longer feasible. The missed opportunity is particularly poignant because Mr K has worked in diligent partnership with his physician, suggesting that he might have been a very good surgical candidate.

DR E: We have talked a lot with Mr K about housing options, and he is in close contact with the social worker. He probably could get public housing if he wanted it, but he says he does not. He has various friends he goes to stay with for periods of time and then moves on to somewhere else.

The McKinney-Vento Homeless Assistance Act of 1987, which provided the groundwork for federal support of homeless programs, defined a homeless person as one who lacks a fixed, regular, and adequate nighttime residence or as an individual who has a primary nighttime residence that is a supervised shelter designed to provide temporary living accommodations (including welfare hotels, congregate shelters, and transitional housing for the mentally ill), an institution that provides a temporary residence for individuals intended to be institutionalized, or a place not designed for, or ordinarily used as, a regular sleeping accommodation for human beings.²⁷ Thus, the definition recognizes that, like Mr K, homeless people do not necessarily spend every night out of doors or in emergency shelters.

Options for housing for homeless people vary widely and include emergency shelters, transitional housing, single room occupancy hotels, permanent supportive housing, and subsidized public housing. For homeless people with health problems, medical respite may be an option.

Emergency shelters are congregate living arrangements for homeless people; they may be administered by local governments, faith-based organizations, or nonprofit organizations. Procedures, rules, and supportive services vary greatly. Some facilities require clients to reapply for shelter on a nightly basis, others guarantee space for a set period of time, and some require clients to vacate during the day and to be sober. Transitional housing is designed to transition people from homelessness to permanent housing; stay is generally limited to 24 months.²⁸ Single room occupancy hotels provide low-cost housing to indigent adults. Unlike apartments, many accept rent on a weekly or monthly basis. Single room occupancy hotels generally have a small room with shared restroom and cooking facilities.²⁹

Supportive housing is subsidized housing that provides either on-site or closely linked supportive services, including medical care, mental health care, substance abuse treatment, and case management to chronically homeless persons.³⁰ The goals of supportive housing are to “achieve residential stability, increase skill levels or income and obtain greater self-determination.”³¹ Many supportive housing providers now embrace the concept of “housing first,” a harm reduction strategy that does not demand abstinence from illicit substances before obtaining housing.³² Supportive housing can be successful in maintaining housing in difficult-to-house populations, including those with mental illness and substance abuse problems,^{33 - 36} and can reduce inpatient psychiatric hospitalizations.

Federally funded public housing was established by the government to provide “decent and safe rental housing for eligible low-income families, the elderly and persons with disabilities.”³⁷ Local housing agencies may also give housing vouchers that permit eligible persons to find low-cost housing on their own. With a voucher, an individual or family pays 30% of their income for the housing and the voucher program subsidizes the difference. Demand exceeds supply and waiting times can be long.³⁸ The long waits make public housing or vouchers unrealistic for homeless patients, like Mr K, with life-limiting illnesses.

Medical respite describes short-term shelter with basic medical services for homeless persons who are too ill to function on the street but who don't need hospitalization. Respite care models range widely, from homeless emergency shelters with part-time nursing care to extended-stay facilities with physicians available. Respite availability decreases subsequent hospitalization days.³⁹ Although most respite providers recognize the need for end-of-life care, only a few programs address this issue.⁴⁰ Unfortunately, most programs do not have sufficient staffing to attend to the needs of patients nearing death (David Buchanan, MD, written communication, February 7, 2006).

DR E: [Mr K] applied for Medicaid very early in the course of disease and qualified without difficulty because of his extreme poverty and terminal illness. [The] managed Medicaid program provides all medications free through our hospital. We try to keep medications to a minimum because he has trouble taking the ones that he needs . . . primarily for hypertension and renal failure.

MR K: Sometimes I don't get a chance to take my medicine properly either because sometimes I don't be where I can take some water or eat. . . . 

Homeless people have difficulty obtaining prescription medications and adhering to treatment regimens.^{5 ,41 - 42} Difficulty adhering has been attributed to many factors, including financial barriers to obtaining medications,^{3 ,43 - 44} mental health or substance use problems,^{44 - 46} difficulty storing or refrigerating medications, and loss or theft of belongings.^{41 ,44} Programs to improve adherence have been implemented particularly for infectious diseases.^{44 - 45} Expert recommendations include using once-daily medications, avoiding medications that require refrigeration, minimizing medications that require food and water or that have adverse effects that require frequent use of restrooms, and dispensing small quantities of medication with frequent refills.⁴⁷

DR E: [Mr K] has pain primarily related to his large abdominal mass . . . [O]ne of the biggest struggles . . . is to . . . have good pain control and have him be safe. . . .He very effectively self-medicates with the heroin. [Early in our management], we negotiated that we would give him some acetaminophen-hydrocodone for occasional use but he would continue to use heroin for pain management. He is unusual in that he is very frank about his drug use. . . . I said, “[W]hat I am worried about is that if I give you these pain medications, . . . you are going to be at high risk for an overdose.” He said, “Yes, I agree with that.” I said, “What I would recommend is that if you find what you are doing is not working, then let me know and we will add something else. . . . I don't want to do anything that is going to hurt you.”

A critical component of the management of cancer pain in this patient population is trust, beginning with the physician's acceptance of the patient's self-report of pain and distress. Performing a comprehensive pain assessment and substance abuse assessment is important, as is assessing for comorbid psychiatric disorders,^{48 - 51} because treating depression and anxiety may improve pain management and reduce the risk of relapse.

However, in the absence of data, such approaches are based solely on clinical experience. An interdisciplinary team (eg, pain management or palliative care specialist, addiction medicine specialist, oncologist, nurse, social worker, mental health care provider) is often needed to develop a pain management plan.⁵² Clinicians should not expect a complete remission of the patient's substance abuse. Conflict with staff may decrease if there is an understanding that the goals of care are to provide optimal pain and symptom management within a structure that includes both social and emotional support and setting limits to contain the harm done by occasional relapses.⁵³

DR E: He told me [last spring] he is not using heroin any more because he does not have the energy “to hustle for it.” His pain is well controlled on morphine sulfate sustained release with Roxanol for breakthrough. I chose this because the half-life is not too long and the risk of diversion is relatively low compared to oxycodone or Dilaudid—he didn't like the fentanyl patch. The fentanyl patch is a problem for anyone who spends a lot of time outdoors in the heat. They fall off very easily and are not well absorbed through dry or dirty skin. He may also have been suspicious of it because he could not feel it working. I give him only sustained-release preparations of morphine, and he shows up once a month, on time (not early) to get his refills. Diversion is necessarily a concern and is virtually impossible to rule out in this case. However, I have seen him in the ER on an occasion when he was discharged from another hospital without his morphine SR [sustained release], and he was in terrible pain. The fact that his pain is generally controlled encourages me to believe that he is taking it as prescribed, at least most of the time.

An effective pain management plan requires a complete assessment of the pain (ie, description, intensity, location, aggravating and relieving factors, and impact on function). Patients who use illicit drugs may be tolerant to the therapeutic effects of opioid analgesics,⁵³ but there are no systematic investigations of this phenomenon. Clinicians may initiate a conservative dose to start, but if the patient has ceased using illicit opioids, clinicians must monitor patients frequently to facilitate rapid-dose titration and to prevent withdrawal. To treat poorly controlled pain in patients who have been receiving regularly scheduled opioids for at least 5 days, a 30% to 50% increase in dose is appropriate as frequently as every 5 half-lives of the medication.^{52 ,54} Dosages must be reduced for patients with impaired hepatic or renal function.⁵²

Dose-escalation requests by patients with substance-use problems may be viewed by clinicians as an aberrant drug-related behavior. Aberrant behaviors are those that increase the possibility of addiction, such as repeatedly running out of medications (which may suggest diversion—selling or trading the controlled substances for illicit drugs), or obtaining opioids from nonmedical sources.^{55 - 56} Some aberrant behaviors may be indicative of pseudoaddiction—that is, behaviors necessitated by an individual's seeking adequate pain relief.⁵⁷ As with Mr K and his clinicians, the fact or perception of addiction should be openly discussed. Clinicians should give clear guidelines to patients (including addicted patients) with life-limiting illness about responsible analgesic behaviors (ie, keeping all appointments, taking analgesics as prescribed).⁵³

As in the program that Dr E developed for Mr K, once the pain is controlled (and analgesic dose has been appropriately titrated), chronic cancer pain is best managed with the use of long-acting opioid analgesics (eg, controlled-release opioids, methadone, transdermal fentanyl) for the persistent component of the pain and use of short-acting opioids for breakthrough pain. Long-acting opioids, in controlled-release or transdermal formulations, are recommended because of their longer duration of action and slower onset of analgesia, which may reduce aberrant drug-taking behaviors.⁵⁸ Dr E notes that she chose long-acting morphine sulfate because it has a lower risk of diversion than either oxycodone or hydromorphone, but her anecdotal observation is controversial.^{59 - 60} Despite Mr K's preference not to use transdermal fentanyl, some patients may prefer it because drinking water is not required, although its higher cost may limit its utility. Also recommended are providing patients with a written pain agreement and list of instructions about the parameters of responsible analgesic use; frequent visits (perhaps weekly) to assess effectiveness and illicit substance use; prescribing small quantities, as Dr E did, to reduce the temptation to sell or trade medications; and renewing analgesic prescriptions contingent on attendance at clinic appointments.

Pain agreements are widely used, although their efficacy has not been evaluated.⁶¹ These written documents are signed by both the clinician and the patient and outline the conditions of continued provision of analgesics. Components usually include indication for use, goals of treatment, name of prescribing clinician, dose and frequency of prescribed analgesic, number dispensed, and frequency of refills, requirements for clinic visits, conditions for lost or stolen medication, explanation of plan of care for severe pain, use of urine toxicology tests (including consequences of positive test results), specification of prohibited behaviors, and consequence of violating agreements.^{62 - 64} Toxicology testing could be done to evaluate whether the patient has been taking the prescribed pain medication as well as any illicit substances, although it can be difficult to distinguish metabolites of heroin from certain opioid analgesics. Thus, testing may be most useful to detect other illicit drugs, such as cocaine or methamphetamines. Dr E said she knew that “the dogma is that you should make a contract . . . that precludes the use of street drugs, but he admitted . . . that he was incapable of complying. There really wasn't much use in getting tox screens on him, because they would be positive.”

Dr E consciously elected to prioritize pain control, chose medications that she believed had low diversionary value, and dispensed smaller monthly amounts with no early refills.

DR E: I could not be 100% confident he was not selling at least some of what he was getting . . . the habit may have been fueled by his access to prescription drugs. [U]ltimately he quit using [heroin] on his own because it was requiring too much effort.

Suggesting that a patient continue to use illicit drugs as a form of pain control is ethically and potentially legally problematic. Although it is important to acknowledge the reality of a patient's drug use, suggesting that the illicit drug itself be a part of the pain management strategy may suggest tacit approval of the illegal activity. Likewise, a physician who continues to prescribe opioid analgesics when he or she knows that medication is being diverted could be subject to legal consequences.

Both regulatory and legislative agencies recognize that the use of opioid analgesics to treat pain is a legitimate medical practice even if the patient has a past or current history of substance abuse problems.⁶⁵ The goal of regulations is to ensure the availability of controlled substances for pain management while preventing their diversion.⁵⁸ Physicians report that the fear of investigation, sanctioning, and even prosecution for prescription of opioid analgesics is a barrier to appropriate treatment of pain.^{65 - 67} Although the likelihood of sanction is low when opioid analgesics are used legitimately for the treatment of pain, physicians' fears may be increased by the media attention afforded the cases in which sanctions were imposed.⁶⁵ In 2001, of 923 839 physicians registered with the Drug Enforcement Agency, 831 (0.09%) were investigated by the agency for potential involvement in criminal diversion of opioid analgesics; 739 (0.08%) of these investigations resulted in regulatory action; 92 (0.01%) physicians were arrested.⁵⁸

The Model Policy for the Use of Controlled Substances for the Treatment of Pain,⁶⁷ adopted by the house of delegates of the Federation of State Medical Boards of the United States in 2004, was designed to provide state medical boards with information regarding the appropriate management of pain in compliance with state and federal laws and regulations. It defines inappropriate treatment as including “non-treatment, undertreatment, overtreatment, and the continued use of ineffective treatments.”⁶⁷ Twenty-four state medical boards have adopted all or part of the guidelines for their own policies.⁶⁸ The Model Policy identifies the components of effective management: a complete medical history, including information on the nature of pain, current and prior treatment, coexisting disease, effect on physical and psychological function, history of substance use and medical indications for use of controlled substance; written treatment plan, including objectives, further evaluations planned, and other modalities used; informed consent and agreement for the treatment, including risks and benefits of use of controlled substances, designation of one clinician to prescribe, and written agreement if the patient has a history of substance use; and periodic review of progress toward goals, including information on improved or diminished function. They also recommend consultation with substance use and psychiatric experts, as needed.⁶⁷

MR K: [The doctors] asked me . . . did I want to be on a machine. . . .Well, at first I was planning for everything, but then I told them that I don't think that I want to put up with that if I don't have no sense come back. So I just decided that I’d kind of pray on it and . . . left it up to my doctor, too. If you don't see where I can come back to my mind, just go ahead and take me off. I will be prepared and I just feel like I am going to turn it over to the Lord.

The few studies of end-of-life wishes in the homeless population have shown both similarities to and differences from those of the general population.^{69 - 71} Although there is no reason to believe that homeless people differ from the domiciled in terms of a wish to receive adequate symptom control and to be treated with dignity at the end of their lives, living on the margins of mainstream society may elevate some concerns and change their goals of care.

Because many homeless people have had negative experiences with the mainstream health care system (including having therapy denied, as did Mr K), homeless people may fear that care will be inappropriately withheld at the end of life.⁷⁰ A study of homeless persons' beliefs about resuscitation found this population was more likely than a cohort of domiciled persons with chronic obstructive pulmonary disease to wish for cardiopulmonary resuscitation and aggressive attempts at curative care, with 88.6% of homeless persons requesting cardiopulmonary resuscitation compared with 61.6% of domiciled respondents.⁷¹

Some studies have found that homeless persons expect their deaths to be sudden and violent.^{69 - 70} Because of this, they may be less willing to discuss end-of life-planning. Clinicians can explicitly acknowledge past negative experiences and take concrete action to engender trust, reassuring patients that their wishes will be honored.

DR E: [Mr K] really doesn't like to be cooped up. To be honest, I can't tell you whether we have had an explicit discussion about it [end-of-life care] or not. My impression is that he wants to keep going as long as he can and then when he is done, he is done. He would not be happy living in a nursing home and he really would not be happy being stuck in the hospital. My biggest fear is that he will die out on the street, but his biggest fear seems to be loss of independence.

MR K: Well, she told me that if I wanted to . . . that she could do everything in her power to get me there at a nursing home. But I told her that I will let her know. . . . I hear so many different things about nursing homes, how they treat patients, . . . and I am not exactly at that stage of mind.

Dying at home is not a realistic goal for people who lack both the literal shelter of a home, as well as the companionship and social connections often associated with stable housing.⁶⁹ Chronically homeless individuals may chafe at institutional care. Having spent years living on the street, they may not easily accommodate to the rules and restrictions found in skilled nursing facilities nor feel comfortable receiving care in facilities that cater to geriatric populations.

Homeless persons with substance abuse problems may have difficulty living in a setting where there are specific prohibitions against illicit drugs. Whereas Mr K can obtain and use heroin on the street, it would be nearly impossible to do so in a nursing home. Homeless persons cannot generally use institutional hospices because they lack resources to pay for them. Thus, in all likelihood, homeless people are often left with the option of presenting to the clinic or emergency department for admission to the hospital when their pain becomes too severe or they become too weak to function.

The creation of palliative care services in hospitals that serve large numbers of homeless people would provide a setting for end-of-life care. The proportion of hospitals that offer dedicated palliative care programs has grown significantly in the past decade, from 15% in 2000 to 25% in 2003.⁷² Academic medical centers, Veterans Affairs hospitals, and Catholic hospitals are more likely to have palliative care services than nonprofit or public hospitals.^{72 - 73} Because homeless patients are more likely to receive care in public hospitals, they are less likely to die in settings that offer palliative care services.

MR K: Really, I don't have nobody to discuss it [my illness] with. I told a couple of people that I knew in the street, my friends . . . But they really don't know what to do. So I just kind of left it up to my doctor. Whichever way I want to go, she is willing to help me . . . I feel pretty good with her.

Homeless persons are less likely than the general population to have a family member or close friend to identify as a surrogate; in one study of end-of-life decision making for homeless people, only 60.7% of homeless adults were in contact with their family.⁷¹ With the increased likelihood that a surrogate decision maker will not be available, physicians are more likely to be called on to make decisions on behalf of homeless patients than for members of the general population.⁷¹ A report by the American Bar Association on the “unbefriended elderly” (adults without decisional capacity, advanced health care directives, or surrogate decision makers) found that most states do not address how decisions should be made; 8 states have enacted statutory authorization to give default surrogates the authority to make medical decisions without judicial involvement. In these states, authority is sometimes given to attending physicians, either alone or in consultation with other physicians or an ethics committee. Other state laws recommend the use of either trained volunteers or court-appointed legal guardians.⁷⁴

Physicians making decisions for unbefriended adults appears to be relatively common; one third of critical care physicians acknowledged having discontinued mechanical ventilation for unbefriended adults.⁷⁵ A study of decision making for the critically ill found that 16% of intensive care unit patients in a metropolitan hospital were unbefriended. For 37% of these patients, attending physicians considered making decisions to limit life-sustaining treatment on behalf of the patient. In 56% of these cases, attending physicians made decisions alone, in 33% of cases, they consulted with another attending physician, and in 11% of cases, they sought input either from the hospital ethics committee or the courts.⁷⁶

Clinicians caring for homeless people should encourage them to discuss and document their wishes for the end of life. As with all patients, these conversations should begin before a homeless person is terminally ill. Physicians should plan for multiple conversations to ensure adequate understanding and to build trust. Some recommend using the scenario of a hypothetical medical crisis to make the patient feel more comfortable.⁴⁰ Clinicians should elicit the patient's concerns about dying, asking specifically about experiences that the patient has had with friends or family. They should probe for fears about dying alone and on the street. It is appropriate for the clinician to allay the patients’ fear that their death will pass unnoticed by acknowledging that they will be missed and remembered.

Patients should be reassured that the information discussed will be used to carry out their wishes.⁴⁰ Clinicians should document patients' requests and implement strategies to increase the likelihood that the documentation is available when it is needed, including knowing who patients wish to be contacted in case of emergency and recommending that homeless patients give their emergency contacts a copy of their advanced directive. Clinicians should also ask patients if there is a safe place that a copy could be held. In addition, clinicians should recommend that patients carry advance directive documents on their person.

As demonstrated by Dr E and Mr K, it is possible for clinicians and homeless patients to establish trusting and meaningful relationships. New paradigms, such as “housing first” and ensuring that curative and palliative care are not contingent upon cessation of street drug use, are necessary to improve care at the end of life for chronically homeless individuals. Clinicians who take the time to establish trusting relationships with these most vulnerable of their patients facing the end of life may find great personal as well as professional satisfaction from helping a fellow human being face death with dignity.