The Gerontologist
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 QUICK SEARCH:   [advanced]


     


This Article
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Berger, A.
Right arrow Articles by Bolle, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Berger, A.
Right arrow Articles by Bolle, J.
The Gerontologist 42:49-53 (2002)
© 2002 The Gerontological Society of America

A Commentary

Social and Cultural Determinants of End-of-Life Care for Elderly Persons

Ann Berger, RN, MSN, MDa, Donna Pereira, RN, MAa, Karen Baker, RN, MSNa, Ann O'Mara, RN, PhD,NCIa and Jacques Bolle, RN, DNSca

a Warren Grant Magnuson Clinical Center, National Institutes of Health, Bethesda, MD

Correspondence: Ann Berger, RN, MSN, MD, Warren Grant Magnuson Clinical Center, Chief Pain and Palliative Care Service, National Institutes of Health, Building 10, Room 2N236, 10 Center Drive, Bethesda, MD 20892. E-mail: aberger{at}cc.nih.gov.

The limited data about the epidemiology concerning the settings where elderly persons die include a variety of social, cultural, and economic factors. Unfortunately, even less is known about the interplay among these factors. Approximately two thirds of all cancer patients prefer, but often do not succeed in, dying at home (Gallo, Baker, and Bradley 2001Citation; McCusker 1983Citation; Tang and McCorkle 2001Citation). The purpose of this article is to provide a survey of current literature reviewing the social and cultural determinants as they relate to the settings where elderly individuals in the United States die. Given the paucity of data in this subject area, the review will include research conducted over the past 20 years. The problem with this approach is the profound economic and technological advances that have occurred in the health care delivery system and that in turn have affected how and where we spend our last days of life. We will use the model of Mor and Hiris 1983Citation that outlines possible factors for choices of settings at the end of life: (a) sociodemographic characteristics of the patient, (b) characteristics of available support networks, (c) measures of functional characteristics and degree of dependence on support network, and (d) health system and institutional factors.


    Review of the Literature: State of Science
 TOP
 Review of the Literature:...
 Research Priorities
 Conclusions
 References
 
Sociodemographic Characteristics
Among the sociodemographic characteristics most commonly identified by researchers, those of sex, age, marital status, ethnicity, socioeconomic status, and educational level have been explored most often. In general, the physiological variables, sex and age, do not consistently correlate with a place of death in the United States. For example, Tang and McCorkle 2001Citation review of the literature revealed that European women were less likely than men to die at home; whereas, quite the opposite was found in the United States where women were more likely than men to die at home. In their review of studies exploring age and site of death, Tang and McCorkle found no significant correlation between age and home or hospital death.

Studies of the relationship between marital status or educational status and site of death have also not been confirmatory. Several studies demonstrated that single, divorced, and separated patients were more likely than married patients to die in the hospital (Mor and Hiris 1983Citation; Tang and McCorkle 2001Citation). At least three studies showed no significant difference (Tang and McCorkle 2001Citation). Although one study revealed no correlation between educational level and place of death, two studies concluded that patients with higher levels of education had a tendency to die at home (Mor and Hiris 1983Citation; Tang and McCorkle 2001Citation).

The impact of ethnicity, cultural affiliations, and socioeconomic status on site of death has shown more consistent correlations than sex and gender. African Americans are more likely to die in a hospital than in a nursing home, at least for those who cannot be cared for at home (Merrill and Mor 1993Citation). Individuals of higher socioeconomic status are more likely to die at home, compared with those of lower socioeconomic status who die in hospitals in both Europe (Tang and McCorkle 2001Citation) and the United States (McCusker 1983Citation; Mor and Hiris 1983Citation). However, the exception to this trend was found in a study of individuals with the lowest socioeconomic status who died at home (McCusker 1983Citation).

The tone and content of the conversations that health care providers have with patients who are terminally ill of different cultures may be closely linked to where the patient dies. There are clearly different ways of communicating about death and dying to individuals of different cultures, which often does not fit into the Western model of telling the patient everything. Cultures that have been noted not to speak with the patient about death and dying include Indians, Filipinos, Chinese, Italians, and Anglo-Celtic Australians (McGrath, Vun, and McLeod 2001Citation). The end result can be a misunderstanding of how and where terminally ill patients want to spend the last few days or weeks of their lives.

Support Networks
Social networks, in the form of having a full-time caregiver present and the physical and emotional abilities of the caregiver, can influence where terminally ill patients die. A number of studies have explored these relationships and, in general, the findings have shown that patients with caregivers or good family support are more likely to die at home than in the hospital (Tang and McCorkle 2001Citation). In addition, if the patient is the primary caregiver or not the spouse or a child, death in the hospital is more likely (Mor and Hiris 1983Citation; Tang and McCorkle 2001Citation).

Typically, the ability of the caregiver to allow the patient to die at home is determined by many factors. The literature illustrates that the younger the age of the caregiver, the more likely the patient would die at home, whereas if the primary caregiver is dependent on outside employment, the patient will die in an institution (Mor and Hiris 1983Citation; Tang and McCorkle 2001Citation). To understand the needs of caregivers, one needs to assess their physical, psychological, social, and spiritual well-being—all realms of quality of life. In a study of family caregivers of older cancer patients receiving care at home, the greatest problems for the caregiver were fatigue, sleep changes, pain, and overall physical condition (Ferrell, Juarez, Borneman, and ter Veer 1999Citation). Many caregivers were facing their own declining health. Other studies have identified similar problems that cut across the physical, psychological, and spiritual domains of the caregiver. Both family function and structure are affected when the caregivers are inadequately supported (Borneman and Ferrell 2001Citation).

Caregiving activities can interfere with caregiver employment and lead to a subsequent loss of income, which can in turn influence the settings where patients die. Losing one's job and, in many cases, health insurance benefits to remain at home and care for a terminally ill loved one is, for the vast majority in this country, simply not an option. The economic burdens of both income loss and informal caregiver costs are substantial and longstanding. To date, the strength of the relationship between income loss and the settings where patients die has not been studied. One area that has been explored is the difference in caregiver expenditures between terminally ill patients and survivors. Given, Given, and Stommel 1994Citation compared 3-month caregiver expenditures of terminally ill and survivor breast cancer patients. On average, expenditures were twice the amount for decedents compared with survivors.

One particular aspect of psychosocial abilities of the caregiver is the notion of hope. Of particular note is how caregivers perceive or define hope throughout the dying process of their loved ones. In a study by Yates and Stetz 1999Citation, family caregivers who were interviewed had two stages of hoping. At first, the hope started as a hope for cure, many times continuing until the end. As the disease progressed, hope sometimes shifted to hope for relief from suffering. There are no known studies of how this influences the setting of death; however, when hope for cure remains until the end, clinically it is obvious that the patient will need to die in the hospital because everything, including heroics, may need to be done.

Functional Characteristics and Degree of Dependence on Support Network
In a similar vein to caregiver attributes, researchers have studied terminally ill patient variables that may impact on site of death. Among the more frequently studied are functional status, diagnosis, length of survival since diagnosis, caregiving needs and referral to home care, terminal symptoms, pain management, causes of rehospitalization, institutionalized patients and hospitalization, and patient wishes. The consistency and strength of the impact of these variables on site of death have not been thoroughly established. Because of these inconsistencies and paucity of data, considerably more research is still needed in determining how these variables impact on place of death. A study by Mor and Hiris 1983Citation revealed that patients who are admitted into the hospital are more functionally dependent; however, functional status is not directly related to place of death. The reason for this is unclear.

There are many conflicting data on diagnosis and its relationship in terms of place of death. The only diagnoses that have repeatedly been shown to predict hospital death are hematological malignancies (McCusker 1983Citation).

Several studies revealed that patients with a shorter survival time were more likely to die in the hospital (McCusker 1983Citation; Polissar, Severson, and Brown 1987Citation); however, Mor and Hiris 1983Citation found length of survival was not related to place of death.

There is little literature on the need for specialized clinical care needs, such as equipment needs and place of death. Mor and Hiris 1983Citation suggested that the need for specialized skilled care was not related to place of death.

Only a few studies have investigated the difference between symptoms of patients who die at home and those who die in an institution. According to Mor and Hiris 1983Citation, mildly disoriented patients died at home, whereas those patients who were severely disoriented or oriented may have died in an institution.

Izquierdo-Porrera, Trelis-Navarro, and Gomez-Batiste 2001Citation cited that for elderly cancer patients predicted hospital death for men was related to digestive comorbidities, and predictive factors for women were functional dependence and lack of social support.

Research has revealed that terminal cancer patients in hospitals had more severe pain than those in the home group (Tang and McCorkle 2001Citation). The SUPPORT study also showed that people in the hospital had moderate-to-severe pain at the time of death (SUPPORT Principal Investigators 1995Citation). Many studies have shown that a major issue for family caregivers is management of the patient's pain, especially if it is not under good control (Borneman and Ferrell 2001Citation; Ferrell and Borneman 1999Citation; Ferrell, Grant, Borneman, Juarez, and ter Veer 1999Citation; Schachter and Coyle 1998Citation).

A study done by McCusker 1983Citation revealed that patients were hospitalized according to medical or social factors. Medical indications included surgical procedures, administration of treatments, or management of pain and other symptoms. Social indications included absence of a caregiver, need for respite of the caregiver, or financial reasons related to the patient's insurance (McCusker 1983Citation).

For institutionalized elderly persons, the incidence of an acute condition appears to precipitate hospitalization, whereas among community dwellers, the presence or absence of a social support network and race are the only factors predicting hospital death. The oldest of the old nursing home residents are least likely to be discharged to the hospital to die. Those who enter a nursing home in the last year of life almost never die at home. Physical disability, lack of social contacts, incontinence, and a diagnosis of dementia were significantly associated with a long, terminal, institutional stay (Brock, Foley, and Salive 1996Citation).

Encouraging and empowering patients to make their wishes known to families and health care providers is an important component to achieving the goal of dying in a preferred setting. However, simply knowing a patient's desire to die in a particular setting may not be sufficient to making the desire a reality. For the past two decades, researchers have explored barriers to implement patients' wishes and to identify ways in which both the health care team and family caregivers can help terminally ill patients die in their preferred setting. In a study done with 100 elderly patients under the care of a home health care team, the key factor in the place of death was the patient's expressed wish. The failure to provide death at home for 19 of the patients was because there was a prolonged terminal phase that led to emotional and physical exhaustion in the patient or family (Junker-Groth and McCusker 1983Citation). A recent study conducted by Leff, Kaffenbarger, and Remsburg 2000Citation demonstrated that making plans to die in a particular place for 80 patients followed in a physician-led house call program was common and implemented successfully. The most frequently identified setting was the home, and 91% of the patients were able to achieve their goal. Having a do not resuscitate order also contributed to the patient making specific plans on site of impending death. Clearly, more research is needed to explore the processes that patients, their families, and the health team undergo as plans are made for a comfortable and preferred place of death.

In addition to understanding terminally ill patient variables and caregiver variables, the health system environment in which the death will ultimately occur must also be studied. In the United States, the more commonly studied health system factors include hospices, hospitals, health insurance coverage, home care, and long-term care facilities. Several trends are emerging from the ongoing research in this area. For example, the number of hospital beds and specialists in a given area can play an important role in determining where patients die.

Mor and Hiris 1983Citation demonstrated that patients who were in a hospice that had access to hospital beds were three times more likely to die in the hospital than hospices that had no easy access to hospital beds. Areas with more hospital beds per capita had higher rates of in-hospital deaths and lower hospice use, but areas with higher health care reimbursement and more physician availability had higher rates of hospice use (Virnig, Kind, McBean, and Fisher 2000Citation).

In a study done with 28,828 hospice and nonhospice cancer patients, hospice participation was found to be the variable most strongly associated with death at home. Admission to hospice appears to override the tendency for certain subgroups, such as elderly persons, to die in an institutional setting (Moinpour and Polissar 1989Citation).

In general, insurance coverage is less comprehensive for patients dying at home than for the patients dying in hospitals (Mor and Hiris 1983Citation). There is some suggestion that increasing the hospice Medicare benefit to include provision of respite care might afford patients the ability to die at home (McCusker 1985Citation).

Many people suggest that providing nursing or physician care at home may facilitate improved end-of-life care for older persons; however, there are no documented studies that looked at the impact of 24-hr available supportive care on the place of death (Merrill and Mor 1993Citation; Tang and McCorkle 2001Citation).

The death of a resident in a long-term facility is not an isolated incident, perhaps affecting only the immediate family and a few close friends. To the contrary, other residents and staff are impacted, in a number of ways, by this event. A small study conducted in England and applicable to long-term residents in the United States bears this out. Katz, Siddell, and Kormaromy (2001) investigated how death and dying are managed in long-term-care facilities in three geographical areas of England. Managers and staff of these facilities believed that the best way to support other residents when someone was terminally ill was to limit their knowledge of the situation. When a resident died, managers recognized that family members needed an array of psychosocial supports. In addition, managers acknowledged that staff required support during the dying period and beyond. In both instances, however, budgetary and staff constraints precluded management's ability to provide the resources.

Berger, Termini, Hirshfield, and Iesle (1998–1999) conducted a research project in the state of New Jersey in conjunction with the Office of the Ombudsman for the Institutionalized Elderly. The project was funded by the Healthcare Foundation of New Jersey. The project goals were to improve pain and symptom management and quality of end-of-life care in long-term care settings by: (a) teaching the principles of palliative care, (b) empowering the long-term care staff to develop a Palliative Care Consultation Team that would manage pain, and (c) plan the end of life within their facility and to encourage referrals to hospice when appropriate. The consultation team was to develop sensitivities toward end-of-life issues leading to institutional changes, as well as changes in attitudes and beliefs regarding death and dying.

A pretest evaluated the team members' (a physician, nurse, spiritual minister, and social worker) knowledge of pain and symptom management. The mean score was 14.02 out of 30 (47%). Following the 2-day training, which included bedside rounds, focus groups were done to evaluate how the staff felt the training had helped them and what barriers existed that made it difficult to practice pain management and palliative care. The focus groups revealed that the staff felt empowered to serve residents' needs in the arena of pain and palliative care. They also felt that there was a greater acceptance of discussion of the death and dying process, before the event occurrence. Posttest scores were not measured after team member training.

The barriers to good end-of-life care that were identified included staff turnover, specifically in the long-term care facility, leading to loss of the key team members; lack of administrative structure, leading to lack of appropriate utilization of staff; lack of support from other health care members in their facilities; lack of availability of physicians; and lack of knowledge and myths about the use of opioids for elderly patients by other staff members in the institution, as well as the residents' resistance to utilize recommended medications because of fear or belief that pain is to be expected. Even though there are no studies to investigate how all of these barriers interplay to determine the setting in which institutionalized elderly persons die, it may be that inability to manage end-of-life care in long-term care facilities may affect where people die.


    Research Priorities
 TOP
 Review of the Literature:...
 Research Priorities
 Conclusions
 References
 

More Light!—Goethe's last words (1832; as cited in Field and Cassel 1997Citation).

These last words are not necessarily a plea for more research on the end of life, but they are fitting nonetheless.

What do we need to know to design an end-of-life care system that meets the needs of older populations? Universally, there is a need to evaluate the outcomes of older people to define clearer concepts in the last phase of life. Before moving forward, we need to ask: What are the reasons for inadequate applications of existing knowledge for provision of accessible, effective and affordable care? How has professional education about pain and palliative care changed the attitudes and knowledge of health care providers, and how is this translating to advice and place of setting of death for the patient?

How do patients' preferences and knowledge about end-of-life care options interact with physicians', health care providers', and public attitudes and knowledge about end-of-life care? In our diverse population, how do we respect cultural differences, such as withholding disclosure of terminal care status, in a litigious society? Diversity is not just ethnicity. Recent research has focused on pockets of ethnic groups, such as elderly African American women, but the 2000 census indicated people are less willing to box themselves into select ethnic categories. Dying experiences have not been explored in several cohorts, such as the aging gay and lesbian community and newly arrived immigrants lacking residency or citizenship status. The media have such a powerful influence on society's attitudes toward and knowledge of end-of-life decision making: How can market research support the design of programs to meet pluralistic end-of-life needs?

How does hope for cure come into play in both the patients' and families' decisions on the use of technology and wanting "everything done"? What does "everything done" mean to patients and families? How do all of the family caregivers' physical, psychosocial, financial, and spiritual needs determine where their loved one will die? What does quality of life mean to patients and families? Suffering is often enmeshed with physical symptoms requiring integrative research in a multivariate format. How do we evaluate variables in the clinical pathways of dying, as in duration, patterns, and comorbidity, such as advanced dementia?

What are different models that can be developed to incorporate pain and palliative care management in all long-term care facilities? What are family members' preferences as to place of death, and how does this change over the course of the dying process? What triggers a change or transition of patient care settings (e.g., home care, hospice, hospital, nursing home, or combinations)? What are the effects on the patient and family with differing care deliveries and processes?

What are the organizational and economic obstacles that impede reliably excellent care at the end of life? Administrative boundaries often separate out medical and social services and tend to focus on isolated documents, such as advance directives. Economic benefits and burdens affect patients, families, care settings, and insurance companies.


    Conclusions
 TOP
 Review of the Literature:...
 Research Priorities
 Conclusions
 References
 
It is essential, in light of the paucity of literature, to have prospective studies to test hypotheses that look at the interplay of all of these factors in terms of both preferences as to place of death and actual place of death. As the recent report about transforming death in America prepared for Last Acts stated, "The conclusions were clear: Far too many people needlessly suffered physically, psychologically, spiritually, and socially at the end of life. Their preferences regarding care were too frequently ignored, and their families often were left in emotional despair and financial ruin" (Metzger and Kaplin 2001Citation).

Have we truly asked the right questions?

The Forum

Book Reviews

Practice Concepts

Received for publication October 4, 2001. Accepted for publication May 29, 2002.


    References
 TOP
 Review of the Literature:...
 Research Priorities
 Conclusions
 References
 




This article has been cited by other articles:


Home page
Journal of Applied GerontologyHome page
D. Gu, G. Liu, D. A. Vlosky, and Z. Yi
Factors Associated With Place of Death Among the Chinese Oldest Old
Journal of Applied Gerontology, February 1, 2007; 26(1): 34 - 57.
[Abstract] [PDF]


Home page
Palliat MedHome page
G Visser, M Klinkenberg, M I B. van Groenou, D L Willems, C P M Knipscheer, and D J H Deeg
The end of life: informal care for dying older people and its relationship to place of death
Palliative Medicine, July 1, 2004; 18(5): 468 - 477.
[Abstract] [PDF]


This Article
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Berger, A.
Right arrow Articles by Bolle, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Berger, A.
Right arrow Articles by Bolle, J.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
All GSA journals Journals of Gerontology Series A: Biological Sciences and Medical Sciences Journals of Gerontology Series B: Psychological Sciences and Social Sciences