The median age at death in the United States is now 78 years and has been associated with a steady and linear decline in age-adjusted death rates since 1940. In 1900, life expectancy at birth was <50 years; a girl born today may expect to live to age 81 and a boy to age 76. Those reaching 65 years can expect to live another 18 years on average and those reaching age 80 can expect to live an additional 8 years. These unprecedented increases in life expectancy (equivalent to that occurring between the Stone Age and 1900) are due primarily to decreases in maternal and infant mortality, resulting from improved sanitation, nutrition, and effective control of infectious diseases. As a result, there has been an enormous growth in the number and health of the elderly. By the year 2030, 20% of the United States’ population will be over age 65, as compared with <5% at the turn of the 20th century (7).

Although death at the turn of the 20th century was largely attributable to acute infectious diseases or accidents, the leading causes of death today are chronic illness such as heart disease, cancer, stroke, and dementia. With advances in the treatment of atherosclerotic vascular disease and cancer, many patients with these diseases now survive for years. Many diseases that were rapidly fatal in the past have now become chronic illnesses.

In parallel, deaths that occurred at home in the early part of the 20th century now occur primarily in institutions— 68% of all deaths occur in hospitals or nursing homes (8). The reasons for this shift in location of death are complex, but appear to be related to health system and reimbursement structures that promote hospital-based care and provide relatively little support for home care and custodial care services despite the significant care burdens and functional dependency that accompany life-threatening chronic disease in the elderly. The older the patient, the higher the likelihood of death in a nursing home or hospital, with an estimated 76% of persons over 85 years experiencing an institutional death and a similar number spending at least some time in an institution in the year prior to death (8). These statistics, however, hide the fact that most of an older person’s last months and years are still spent at home in the care of family members, with hospitalization and/or nursing home placement occurring only near the very end of life. National statistics also obscure the variability in the experience of living with serious illness. For example, the need for institutionalization or paid formal caregivers in the last months of life is much higher among the poor and women. Similarly, persons suffering from cognitive impairment and dementia are much more likely to spend their last days in a nursing home compared with cognitively intact, elderly persons dying from nondementing illnesses.

The needs of older adults living with serious illness are not well matched by current models of care. Specifically, multiple studies demonstrate that the personal and practical care needs of patients who are seriously ill and their families are not adequately addressed by routine office visits or hospital and nursing home stays and that this failure results in substantial burdens—medical, psychological, and financial—on patients and their caregivers (9). Neither paid personal care services at home nor nursing home costs for the functionally dependent elderly are covered by Medicare, but instead are paid for approximately equally from out-of-pocket and Medicaid budgetary sources that were originally developed to provide care for the indigent. In the context of chronic progressive disease, the burden of coordinating an array of social and medical services falls on primary physicians and more often, individual families.

In response to the needs of seriously ill older adults and their families, palliative care teams have become increasingly prevalent in United States hospitals and provide comprehensive interdisciplinary care for seriously ill patients and families in collaboration and consultation with primary physicians. Over 80% of hospitals with over 300 beds now report a palliative care team and over two-third of all hospitals report a palliative care team—a steady 140% increase in prevalence since 2000 (10). Hospital palliative care teams have been shown to significantly reduce symptom burden, enhance patient and family satisfaction, and lower costs (Morrison nejm, health affairs, NEJM). Other programs, although less well developed, focused on reducing functional decline and delirium and improving transitions from hospitals also show promising early results (11).

In the ambulatory care setting, programs are less well developed than in hospitals. Hospice services, under the Medicare benefit, are available in most US communities and provide palliative care, primarily at home, for patients with a life expectancy of 6 months or less who are willing to forgo insurance coverage for life-prolonging treatments. Overall, about 40% older adults access their hospice benefit prior to death and median length of stay on hospice is on the order of 3 weeks (12). Reasons for the low rate of utilization of the Medicare Hospice Benefit vary by community but include the inhibiting requirements that patients acknowledge that they are dying in order to access the services, that physicians certify a prognosis of 6 months or less, and that very few hours (usually 4 or less) of personal care home attendants are covered under the benefit. In addition, the fiscal structure of the Medicare Hospice Benefit lends itself well to the
predictable trajectory of late-stage cancers, but not so well to the unpredictable chronic course of other common causes of death in the elderly such as congestive heart failure, chronic lung disease, stroke, and dementing illnesses.

Other programs that coordinate care for patients who have complex illnesses, outside of hospice, are becoming increasingly available in many communities—primarily for younger adults or for individuals enrolled in Medicare Advantage (i.e., Medicare managed care plans). These programs typically focus on intensive telephonic case management and have been shown in early studies to improve care for those with serious illness (13). The quality, cost, and extent of the services provided are highly variable.

Finally, comprehensive multidisciplinary home care programs that serve frail older adults have been developed in several specialized settings. The Program of All-Inclusive Care for the Elderly (PACE) is a capitated Medicare and Medicaid benefit for frail older adults that offers comprehensive medical and social services at day health centers, in homes, and at inpatient facilities. Patients enrolled in PACE have higher rates of advance directive completion and lower rates of nursing home admission, hospitalization, and hospital deaths than do patients who do not use the services (ref Morrison NEJM). Similar programs of team-coordinated home-based care exist within the Veterans Administration (VA) (VA home-based primary care and VA palliative care programs). All VA hospitals are required to have both a home-based primary care program for homebound veterans and a palliative care team. Furthermore, under recent VA regulations, all veterans are allowed access to hospice at the same time as they are receiving disease-directed or curative treatments.

In nursing homes, the site of care for many of the most seriously ill and cognitively impaired older adults, incentives promoting palliative care standard are lacking. Indeed, nursing home quality metrics focus on improvement of function and maintenance of weight and nutritional status. Evidence of the decline that accompanies the dying process is typically regarded as a measure of substandard care (14). Therefore, a death in a nursing home is often viewed as evidence particularly by state regulators of poor care rather than an expected outcome for a frail, chronically ill, older person. The financial and regulatory incentives and quality measures that currently exist in long-term care promote tube feeding over spoon feeding and transfer to hospital or emergency department in the setting of acute illness or impending death (Table 62.1). They fail to either assess or reward appropriate attention to palliative measures, including relief of symptoms, spiritual care, and promotion of continuity with concomitant avoidance of brink-of-death emergency room and hospital transfers (15). Although provision of hospice services has been shown to improve quality in nursing homes (16), penetration of hospice services into most nursing homes remains low and increasing federal scrutiny on long-lengths of stay of nursing home residents on the hospice benefit has led to concerns about enrolling this patient population.

Although death occurs far more commonly in older adults than in any other age group, remarkably little is known about the course of serious illness in the oldest old, that is, those over age 75. Most research on the experience of living with serious illness has been done in younger populations, and most studies examining pain and symptom management have focused on younger populations. Studies in older adults have focused primarily on patients’ preferences for care rather than on the actual care received. Indeed, the largest study to date of the experience of living with a serious illnesses in the United States (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments [SUPPORT]) studied the hospital experience of patients with a median age of 66 (17). The median age of death in the United States is 78 years, and many of the oldest old die in nursing homes or at home rather than in hospital. Data from Medicare and state Medicaid registries suggest that expensive and high technology interventions are less frequently applied to the oldest patients, independent of functional status and projected life expectancy. Whereas these discrepancies may reflect patient preferences and indicate appropriate utilization of resources and patient preferences, it is more likely that they represent a form of implicit rationing of resources based on age. The implication is disturbing, considering that half of the highest cost, Medicare enrollees survive at least 1 year (18).

Aside from pain and other sources of physical distress (see Section “Symptom Management: The Challenge of Pain”), the key characteristic that distinguishes the experience of serious illness in the elderly from that experienced by younger groups is the nearly universal occurrence of long periods of functional dependency and need for family caregivers in the last months to years of life. In SUPPORT, the median age of participants was 66 years and 55% of patients had persistent and serious family caregiving needs during the course of their terminal illness (19), and in another study of 988 terminally ill patients, 35% of families had substantial care needs (20). This percentage rises exponentially with increasing age. Although paid care supplements provide the sole source of care in 15% to 20% of patients (transportation, homemaker services, personal care, and more skilled nursing care), the remaining 80% to 85% of patients receive most of their care from unpaid family members (20). Furthermore, most family caregiving is provided by women (spouses and adult daughters and daughters-in-law), placing significant strains on the physical, emotional, and socioeconomic status of the caregivers. Those ill and dependent patients without family caregivers, or those whose caregivers can no longer provide nor afford needed services, are placed in nursing homes. In the United States, this typically occurs after patients exhaust all of their financial savings in order to become eligible for Medicaid. At present, 20% of over the age 85 population reside in a skilled nursing facility, and this number is expected to increase dramatically in the next 50 years (21). Present estimates suggest that the current number of skilled nursing facility beds in the United States will be woefully inadequate for the needs of our aging population.