Meeting Palliative Care Needs in Post–Acute Care Settings: Title and subTitle Break“To Help Them Live Until They Die”

Mr D was an 84-year-old man with metastatic melanoma who was hospitalized for a rapid decline in functional status and diminished oral intake. On admission, he was intermittently able to follow commands and had a left facial droop. He was dehydrated and received intravenous fluids. Other causes of functional decline such as infection, hypothyroidism, or hypercalcemia were excluded. Central nervous system imaging confirmed multiple enlarging metastases.

Diagnosed with melanoma 3 years earlier, the patient had declined biopsy and potentially curative surgery. Metastases to his adrenal glands and brain were confirmed during the year prior to admission, and he received palliation with radiation and corticosteroids.

Mr D was admitted from home where he lived with his wife of 58 years. They had no children, and the patient's only other close family were 2 physically frail sisters. His wife was his durable power of attorney for health care. However, she had dementia, and her capacity to continue fulfilling this role was uncertain. Mr D's physician estimated his life expectancy to be “days to weeks,” depending on nutritional intake. When the patient resumed eating and drinking small amounts, the physician recommended discharge to a nursing home. The family expressed a desire for Mr D to stay in the hospital until his death, which was untenable for financial reasons. The family's other major priority was location—they wanted to find a facility close to Mr D's home so that they could visit daily.

A Perspectives editor interviewed Dr T, Mr D's palliative care physician in the hospital, and Ms R, the director of nursing at the nursing home to which Mr D was discharged.

DR T: The prognosis was probably weeks. Medicare would pay for an acute hospital stay for an acute hospital need, but they wouldn't pay enough for an acute hospital stay for hospice-level care, and therefore the family would end up with a huge hospital bill. Mr D was unable to care for himself at home. He had become progressively weaker, slurring his speech, and had difficulty expressing himself. He was really unable to get around and do activities of daily living. His wife is also diminished in her capabilities because she has multi-infarct dementia. It was pretty clear that his 2 sisters, who were even older than he, weren't going to be able to care for him. So, it was basically a situation where he was too ill to care for himself, and he had nobody at home to care for him.

For many people in the United States, the most desirable option for end-of-life care is to die at home with hospice.¹ However, even the most loving families may lack the physical stamina, economic resources, or emotional resilience to provide terminal care. Mr D's caregivers—his sisters and wife—were unable to fulfill this role.

How can high-quality palliative care be ensured when acute care is no longer necessary and home care is not feasible? Reimbursement rules favor short hospital stays, but transfers between care settings are fraught with risks of medical errors and increased patient and family distress.^{2 - 3} Transfer from hospital to post–acute care requires that new clinicians assume responsibility at a critical time for patient and family. Hospital physicians who recommend transfer should ensure that (1) the patient cannot benefit from additional acute care, (2) treatments can minimize pain and other discomfort during transfer, and (3) the patient's life expectancy is sufficiently long to realize benefits from the new setting. When these conditions are satisfied, hospital-based practitioners can facilitate transfer to post–acute care.

We review herein options for post–acute palliative care settings. We outline ways that hospital-based physicians can make wise recommendations from among these options and write orders to ensure continuity of care. Finally, we discuss research on interventions that may expand and improve palliative care services in post–acute care settings.

When a patient receiving palliative care leaves the hospital but cannot go home, post–acute care options usually include inpatient hospice, nursing home with or without hospice, or residential care with hospice. Choosing among these options requires consideration of (1) patient and family preferences, (2) patients' needs for professional nursing, (3) insurance and ability to pay, and (4) type and quality of local services. Preferences are paramount but tempered by payment options and the type and quality of local services. To ensure optimal decisions, physicians need to collaborate with a knowledgeable social worker or discharge planning nurse.

Settings vary by their provision of professional nursing services, support for activities of daily living, and access to hospice and palliative care. Level of care ranges from inpatient hospice units to assisted living or residential care facilities with a visiting hospice team. Hospital-based physicians need to understand the basic parameters of inpatient hospice, nursing home care, and assisted living or residential care. They can then use key questions to identify local facilities that are best prepared to treat patients requiring palliative care. We have developed questions based on clinical experience, provided in the Table.

Inpatient hospice provides the most comprehensive palliative care. Medicare- and Medicaid-certified hospice beds may be available in free-standing hospice units or in hospitals or nursing homes. Inpatient hospice facilities are certified to provide either short periods of intensive symptom management or long-stay supportive care.

Most post–acute care is provided in nursing homes. Terminal care in US nursing homes is increasing, from 18% of deaths in 1986 to 23% of deaths in 2001.^{4 - 5} Nursing homes may offer onsite palliative care services but more commonly work in collaboration with hospice. Hospices contract with nursing homes to develop a coordinated plan of care and add specialized nursing care for symptoms, nursing aide time, spiritual care, and bereavement support. Most daily care is provided by nursing home staff.

Although collaborative care may be challenging, in 2000 76% of nursing homes in the continental United States contracted with hospices. Hospice was used by 5.6% of nursing home residents who died from 1995-1997.⁶ More recent studies indicate that hospice services are provided in 16% to 26% of deaths in nursing homes and assisted living, with significant variation by state.^{7 - 9} Several studies have demonstrated that the addition of hospice to usual nursing home care improves the quality of end-of-life care.^{10 - 12} Family caregivers perceive improved quality of care for symptoms, and half believe hospice reduces the need for hospitalization. Residents who died with hospice received more pain treatment and decreased use of restraints, intravenous fluids, tube feeding, and hospitalization.^{10 - 12}

Nursing home hospice is covered by Medicare and also by most state Medicaid plans. Hospital-based physicians may find it difficult to discharge a Medicare patient directly to hospice in a nursing home. When the patient's only insurance is Medicare, at discharge a skilled nursing benefit is also available to cover rehabilitative services for as long as the patient responds to this treatment. The patient must elect one or the other, and the skilled nursing benefit is economically advantageous to both patient and nursing home due to more complete coverage.¹³ Coverage is more complete when the patient's primary or secondary insurer is Medicaid.

Patients may also receive hospice in residential care or assisted living facilities. These settings provide a sheltered living environment with housekeeping services, augmented by limited professional nursing and certified nursing assistant services. Care is typically paid for out-of-pocket although the addition of visiting nurse or hospice services will be covered as in a home setting. The only systematic study of end-of-life care in this setting found that hospice is used more often in assisted living than in nursing homes, and families are equally satisfied with the quality of end-of-life care.⁸

DR T: It wasn't clear [to me] who the physician would be. And it wasn't clear. . . that in this setting there would be sufficient numbers of either nurses or CNAs [certified nursing assistants] to help the patient. I certainly didn't think that they had the other members of the multidisciplinary team. . . . They're really not focused on end-of-life care. They're focused on rehabilitation or custodial care.

I visited Mr D in the nursing home twice after his transfer. He did quite well for about a month and a half before he died. He was quite comfortable and well cared for. He denied pain and had no seizures. He said he was getting lots of attention. He initially was in a semiprivate room with a nice view and music, but near the end it was planned to move him to 1 of their 2 private “hospice” rooms. His wife and sisters were able to visit frequently but not to stay overnight.

Dr T echoes concerns expressed by patients, families, and health care professionals. Distrust of this setting is common; 29% of seriously ill adults would rather die than live permanently in a nursing home.¹⁴ Compared with other settings for end-of-life care, family caregivers are least satisfied with conventional nursing home care; in a national study, recently bereaved family reported excellent nursing home care only 41.6% of the time, compared with higher ratings for all other sites of terminal care.^{15 - 16} Persons who die in nursing homes may experience unrelieved pain and emotional distress and have unmet needs for personal care and communication about treatment decision making.^{17 - 22} These findings are consistent despite varied methods and populations.

Major barriers to high-quality care include staff turnover, limited training, insufficient staffing, regulatory pressures, and reimbursement rules.^{23 - 24} Nursing home staff gain skills through experience but rarely have palliative care education.²⁵ Federal regulations require onsite visits by surveyors who evaluate quality of care. Some survey domains—weight loss, functional decline, and use of antipsychotic medication—are common aspects of palliative care. Caregivers may alter treatment out of fear of citation by regulatory bodies, but clear documentation of palliative goals of care can decrease inappropriate citations.²⁶

In contrast, post–acute care facilities may offer improved comfort over that offered by hospitals. The best-quality facilities are comfortable settings with low rates of staff turnover.²⁷ The decor is more homelike and may include personal items such as photographs and favorite chairs. Some facilities provide private rooms with extra space for family and friends, and many accommodate visits by pets. Longer lengths of stay facilitate close therapeutic relationships between nursing staff and dying patients.^{28 - 29}

DR T: We basically wrote “hospice plan of care” orders. When the patient was ready for transfer, he didn't have a lot of symptoms. So, we were able to write some “as-if” orders in case certain symptoms appeared. The good news was that I spoke to the family about the fact that I would be able to go out and see him there. That, I thought, would help smooth the transition.

Transfer can result in loss of continuity of care, increasing patient suffering and family distress.^{30 - 31} Nursing homes must provide for a physician admission assessment within 30 days of transfer.³² Dr T anticipated a potential gap in medical care and wrote orders to cover a variety of possible end-of-life symptoms.

Inpatient palliative care teams rarely visit their patients after discharge. However, they can improve the quality of care by anticipating patients' future needs. Discharge orders may be written to anticipate escalating symptoms. Movement during transfer may escalate pain or dyspnea, and delirium may worsen as the patient enters a new environment.³³ The palliative care team can write “on-call” orders for medication when the patient is moved. Sliding scale orders can provide for different dosages when a symptom is assessed as mild, moderate, or severe. Protocols may be included to anticipate and guide management of acute complications such as gastrointestinal bleeding or seizure and to avoid readmission to the hospital. Although published protocols are not available, clinicians may write individualized protocols to anticipate and guide management of acute complications.

Transfer records should include clear documentation of goals of care and treatment decisions. Records should include copies of living wills and health care power of attorney documents and information about life-sustaining treatment decisions. This information facilitates the work of the accepting physician and provides a rationale to forgo efforts at rehabilitation. All but 5 US states have “out-of-facility” forms to document do-not-resuscitate orders, and the hospital team should prepare these documents prior to transfer. In addition, they should anticipate and discuss decisions such as artificial feeding or rehospitalization and provide orders consistent with these discussions.

MS R: We saw that there was a problem with end-of-life care here. In this setting, where there's so much rush, and so much turnover, we set aside 11 beds for a hospice subunit. We did a lot of training with housekeepers, janitors, unlicensed assistant personnel, as well as the nurses and managers. We have a private room for those people who are—I hate this term—actively dying, and [for whom] more family members tend to congregate.

When we are managing a hospice patient, or multiple hospice patients, I will often add an additional unlicensed person. The thing that I’m most concerned about is presence. In addition to physical care, I think it's important that touch happens and that there's another spirit, another person in the room. It's not just bringing someone in to help them die; it's to help them live until they die.

To ensure excellence in palliative care, nursing homes must provide support for advance care planning, expert symptom management, care for psychological and spiritual suffering, and bereavement services for family and staff caregivers. Over the past several years, investigators have tested interventions to improve care in this setting. Research provides some evidence for innovations in advance care planning, provider training, and changes to health care systems that make hospice or palliative care easier to access ( Article ). These interventions are not widely available but provide models for future research and practice in post–acute palliative care.

Successful innovations in advance care planning take into account characteristics of nursing homes: the relative absence of physicians, the need to address decisions beyond resuscitation, and the high rate of death. Gillick et al³⁴ piloted a goal-oriented approach to advance directives with 38 nursing home residents and surrogates. After education, residents or their surrogates were approached by a physician, nurse, or social worker to rank the goals of maximizing comfort, maintaining function, or prolonging life. Seventy-eight percent of residents and surrogates were able to complete the ranking and to agree to a treatment plan extrapolated from their goals. Given 5 choices, 21% chose the most intensive treatment plan and 6% chose comfort care only. Investigators did not study effects on patient outcomes.³⁴

In Oregon, health care professionals created a “physician orders for life-sustaining treatment” (POLST) form for statewide use. The POLST includes information from advance care planning discussions and covers resuscitation, antibiotic use, artificial feeding, and overall goals of medical treatment. In postimplementation surveys 71% of all Oregon nursing homes reported using the POLST and 88% of residents with a POLST form had orders to limit life-sustaining treatments.³⁵ Finally, 2 randomized controlled trials provide evidence for training nursing home social workers or nurses as advance care planning facilitators. Intervention patients had greater documentation of treatment preferences, care more concordant with preferences, and decreased hospitalization.^{36 - 37} Taken together, these studies suggest that systematic approaches to advance care planning can increase documentation of preferences and influence how patients are treated. Future research is needed to determine how generalizable these approaches are and whether they improve satisfaction with care.

Palliative care training for nursing home staff increases their knowledge and skills. In 1989, Linn et al³⁸ published the results of a randomized trial in which staff training about care of terminally ill patients resulted in greater resident satisfaction at intervention sites. Ersek et al^{39 - 40} designed a comprehensive curriculum entitled “Palliative Care Educational Resource Team” (PERT) for nursing home staff. Evaluation before and after the intervention of 169 trainees demonstrated significant increases in knowledge, self-assessed skills, and supervisor-assessed skills in providing end-of-life care. In an educational intervention for physicians, Keay et al⁴² combined a half-day conference with feedback of chart audits on quality of care. After training, in follow-up chart audits for 203 deaths, physicians were more likely to address pain, dyspnea, overall comfort, and other palliative care concerns. Education improves clinicians' knowledge of and attention to palliative care domains but may not be sufficient to improve practices; future studies should extend these results to measurement of patient and family outcomes.

Quality improvement interventions combine education of health care professionals with feedback of performance data and technical assistance to facilitate practice change. This method has been used successfully in 3 intervention studies to improve pain management practices in nursing homes.^{43 - 45} In a controlled quality improvement study, Hanson et al⁴⁶ worked collaboratively with local hospice organizations to deliver a structured palliative care curriculum and provide feedback of palliative care performance measures to nursing home staff and physicians. Outcomes were measured using chart audits conducted at baseline and at 6-month follow-up. Measures of quality of care before and after training showed that intervention sites increased the percentage of residents enrolled in hospice (4.0% vs 6.8%; P<.01), residents with pain assessments (18% vs 60%; P<.001), residents in pain with nonmedication treatments administered (15% vs 35%; P<.001), and residents with detailed advance care planning discussions (4% vs 17%; P<.01). Quality of care measures were unchanged in control sites.

Providing information to residents and family caregivers about hospice can increase hospice enrollment. In a controlled trial, Cassarett et al⁴⁷ randomized nursing home residents to a routine screen for hospice eligibility. Residents' physicians were asked to agree to an informational visit from a hospice nurse after screening. The intervention increased early hospice enrollment. Families of residents who died expressed more satisfaction if they had received the intervention although at the time of death there was no difference in hospice enrollment.

Nursing home palliative care units provide dedicated staff and a modified environment to support dying residents and their families, but few data on process and outcomes of care have been published.⁵² Kovach et al^{48 - 49} reported on implementation of 5 “hospice households” in 3 nursing homes. Households provided comfort care plans, increased staffing, provided consistent staff for patients, and added family support. Staff satisfaction increased and observational measures of resident discomfort decreased after implementation. Volicer et al⁵⁰ compared outcomes of residents with advanced Alzheimer disease who were admitted to a dementia special care unit with similarly ill residents admitted to traditional nursing home care. Special care unit staff had additional training in dementia care and focused on residents' comfort. Family caregivers were included in advance care planning discussions, guided by a framework of 5 levels of intensity of medical intervention. Overall mortality risk did not differ during 2-year follow-up, but regression analysis showed an increase in mortality risk for residents in the special care unit who had less severe dementia at baseline. Measures of observed discomfort were lower on the special care unit, as were total health care costs.⁵¹

Patients may be unable to receive palliative care at home with hospice services. For these patients, post–acute care options include inpatient hospice or nursing home or residential care with hospice. Nursing homes provide most post–acute palliative care outside private homes, and hospice use in this setting is increasing. Despite attitudinal, regulatory, and payment barriers to hospice and palliative care, the best of these facilities can provide more personalized care than hospitals, while relieving burden on family caregivers.

To ensure high-quality palliative care for patients after discharge, hospital-based physicians should learn to identify the most promising post–acute care settings in their community. When benefits of transfer outweigh risks, they should maximize quality and continuity of care by communicating treatment goals, life-sustaining treatment orders, and anticipatory orders for symptom management.

Physicians and nurses in long-term care will serve increasing numbers of dying patients. Training in generalist disciplines, geriatrics, and palliative care should prepare these health care professionals to treat dying patients in long-term care settings. Early research on innovative approaches to advance care planning, clinician training, and new ways to deliver hospice and palliative care can serve as a guide for larger and more rigorous studies and for dissemination of improved palliative care services.