End-of-life (EOL) care has major financial implications for every stakeholder across the health economy, from patients and their families, to payers, to the healthcare system writ large. A disproportionate share of total healthcare spending is concentrated in the last 30 to 60 days of life, driven primarily by inpatient care, emergency department utilization and late-stage intervention (e.g., ventilators, high-cost medicines, infusion therapy). For patients and families, the EOL journey is often characterized by fragmented care, unwanted hospitalizations and diminished time at home. Patients are more likely to die in their preferred location when palliative care is involved, and caregivers report higher satisfaction with this outcome, particularly when hospital stays are avoided.1 Understanding mortality trends and better aligning care with patient preferences could increase the value of EOL care.
When asked, almost 90% of people report a desire to die at home or in a non-institutional setting. In reality, the standard experience in the U.S. healthcare system often involves hospitalization, preceded by repeated emergency department visits and inpatient stays.2,3 Not only are intensive care settings associated with less favorable experiences for decedents and families, but they are also the most costly sites of care.
Research has consistently shown that a significant portion of Medicare spending – up to 25% – occurs in the final year of life, with inpatient services accounting for the majority of this spending.4 Researchers estimate the average cost for the last month of a person’s life in a hospital can amount to $32,379, often leaving a substantial financial burden to surviving families.5 Intensive care unit (ICU) stays, late-stage procedures and transport between care settings introduce costs and increase the potential for adverse health events (e.g., sepsis, infection, surgical complications). In contrast, earlier hospice enrollment and palliative care are associated with lower emergency department and acute care utilization, fewer in-hospital deaths and improved family-reported outcomes.6 For anyone focused on managing the total cost of care – patients, CMS, policymakers, etc. – managing EOL care and place of death is an obvious opportunity for cost avoidance and patient satisfaction.
Medicare covers hospice services for beneficiaries with a life expectancy of six months or less, providing EOL care including nursing services, medical equipment and medications related to the terminal illness. Enrollment in hospice requires patients to forgo curative treatment for that illness, a tradeoff that may deter some individuals from accessing hospice earlier in their disease trajectory. Separately, while hospital-based care is covered, Medicare does not cover long-term custodial care in nursing homes, and while hospice services may be delivered in facility settings, beneficiaries are generally responsible for room and board costs unless they qualify for Medicaid.
Addressing the cost of EOL care is increasingly urgent as the U.S. population continues to age. By 2030, one in five Americans is projected to be 65 or older, and the number of deaths is projected to exceed 3.6M by 2037, approximately 1M more than in 2015.7,8 Longer life expectancy, combined with a rising burden of chronic disease, has resulted in a larger population of adults with complex medical needs, leading to increasing complexity in EOL care considerations.
This analysis examines place of death in the U.S. between 2018 and 2024, assessing variation by age, race, geography and clinical cause to offer insight into the differential value in various EOL care options.
Leveraging mortality data from the CDC WONDER database, this analysis examines deaths occurring between 2018 and 2024. Each death was categorized by place of death, including home, hospice facility, outpatient or emergency department, inpatient hospital, long-term care facility and other or unknown locations. Changes over time were assessed based on the proportion of deaths occurring in each location, nationally and by age group, race/ethnicity and state. Selected chronic and acute conditions were used to examine trends in place of death by cause.
Between 2018 and 2024, the share of deaths occurring at home increased from 31.4% to 34.3%, representing the largest shift across care settings (Figure 1). Deaths occurring in acute-care hospital departments (inpatient, outpatient, emergency) – accounted for the most deaths each year, declining slightly from 35.1% of deaths in 2018 to 34.6% in 2024. Over the same period, deaths in hospice facilities declined slightly from 7.8% to 6.9%, while deaths in nursing home or long-term care settings fell from 19.0% to 17.1%. Deaths categorized as unknown or other increased nominally by 0.4 percentage points.
Between 2018 and 2024, the share of deaths occurring at home increased across all age groups (Figure 2). Among adults ages 65 and older, deaths at home increased from 30.5% to 34.2%. For adults ages 45-64, the proportion increased from 36.0% to 37.1%, while for adults ages 15-44, the share remained essentially unchanged at 31.3% in 2018 and 31.4% in 2024. Among adults ages 65 and older, deaths occurring in nursing home or long-term care settings declined from 24.1% to 21.1% and hospice deaths decreased from 8.6% to 7.7%. For adults ages 45-64, hospice deaths fell from 7.0% to 5.5% and nursing home deaths declined from 6.0% to 5.3%, a change of 0.7 percentage points. The share of deaths occurring in inpatient and outpatient or emergency department settings remained relatively stable across age groups, with changes of less than one percentage point between 2018 and 2024.
Between 2018 and 2024, the share of deaths occurring at home increased across all racial groups (Figure 3). Among white decedents, home deaths accounted for 31.9% of deaths in 2018 and 34.8% in 2024. Among Asian decedents, the share increased from 31.0% to 34.8% and among Black decedents from 27.9% to 30.9%. Smaller increases were observed among American Indian or Alaska Native decedents (30.8% to 32.6%), Native Hawaiian or Other Pacific Islander decedents (33.7% to 34.8%) and individuals reporting more than one race (32.7% to 34.5%). The share of deaths occurring in inpatient settings was similar in 2018 and 2024 across most racial groups, yet the proportion varied between racial groups. For example, inpatient deaths accounted for 27.5% in both years among white decedents and 35.7% in 2018 compared to 34.9% in 2024 among Black decedents. Nursing home or long-term care settings accounted for a smaller share of deaths in 2024 than in 2018 for several groups. Among white decedents, nursing home deaths declined from 20.3% to 18.3%, and among Black decedents from 12.5% to 10.9%. Hospice deaths represented a smaller share of deaths in 2024 than in 2018 for all racial groups except American Indian or Alaska Native decedents, including declines among Black decedents (6.7% to 5.4%) and white decedents (8.1% to 7.3%). While outpatient or emergency department deaths accounted for a comparable share of deaths across racial groups in both years, white decedents were comparatively less likely to die in that setting.
In 2024, the proportion of deaths occurring at home ranged from 25.0% in South Dakota to 44.3% in Utah with a median of 34.9% (Figure 4). Between 2018 and 2024, the percentage point change in deaths occurring at home varied widely across states, ranging from a decrease of 8.0 percentage points in Utah to an increase of 6.5 percentage points in Iowa. Most states saw increases between one and four percentage points, with the largest increases in Wyoming at 5.1 percentage points, Montana at 4.9 percentage points and Connecticut at 4.9 percentage points.
In 2024, the proportion of deaths occurring in medical facilities ranged from 24.3% in Utah to 44.2% in New York with a median of 33.4% (Figure 5). Between 2018 and 2024, the percentage-point change in deaths occurring in medical facilities varied widely across states, ranging from a decrease of 5.5 percentage points in Nevada to an increase of 2.4 percentage points in Massachusetts. Most states saw decreases between one and three percentage points, with notable declines in Mississippi at 3.9 percentage points, Tennessee at 3.1 percentage points and Washington, D.C. at 3.0 percentage points.
The location of death varies by clinical condition. In 2024, patients with malignant neoplasms were more likely than those with diabetes or ischemic heart disease to die in hospice, with 11.9% of cancer deaths occurring there compared with 3.6% for diabetes and 3.7% for ischemic heart disease, which includes myocardial infarction (Figure 6). Inpatient deaths are also highest for malignant neoplasm deaths (25.4%), compared with 20.4% for diabetes and 20.8% for ischemic heart disease. By contrast, deaths from ischemic heart disease and diabetes occur more frequently at home, representing 41.6% of heart disease deaths and 45.3% of diabetes deaths, and in long-term care, with 15.9% and 15.0% respectively. Outpatient and emergency department deaths are relatively uncommon for cancer patients at 1.9% but occur more frequently for heart disease at 11.8% and diabetes at 10.9%.
This analysis finds that the share of Americans dying at home has increased slightly, while deaths in nursing homes and hospice facilities have declined. While deaths in medical facilities have nominally declined, these most expensive settings remain the default, while lower-cost alternatives remain underutilized, either due to entrenched cultural norms or availability of palliative care.
Given ongoing health policy discussions prioritizing choice and healthcare consumerism, health economy stakeholders seemingly have a moral obligation to enable Americans to die at home, the preferred choice of almost 90% Americans. In turn, the fact that only 35% Americans die at home raises numerous questions about underlying incentives in the health economy.
While some institutional deaths are unavoidable (e.g., trauma patients in the emergency department, sudden heart attacks or thrombosis), many of the 34% of deaths occurring in hospitals could reasonably occur in less acute settings. Many institutional deaths result from the uniquely American characteristic of “heroic medicine,” fostered in large part by fee-for-service reimbursement. Alternatively, Americans would benefit from a purposeful approach to EOL care policy emphasizing dying at home. Designing and implementing such a policy would logically include a combination of Wennberg’s patient-centered care standards, living wills and advance directives and patient education about do-not-resuscitate orders.
Researchers consistently have found that dying in a hospital is twice as expensive as dying in hospice, with minimal or negative added value (i.e., improved quality of life or health outcomes).9 Accordingly, policymakers are obligated to understand – and then eliminate – the barriers and perverse incentives that continue to prevent EOL care from migrating to lower cost care and more patient-focused settings like home or hospice. The known cost savings of dying at home, in hospice or even in long-term care runs counter to current Medicare coverage policy. While hospice services are covered, beneficiaries must forgo curative treatment, and facility-based hospice often requires out-of-pocket payment for room and board unless Medicaid eligibility is met. At the same time, Medicare’s lack of coverage for long-term custodial care continues to limit access to nursing homes for patients with prolonged functional decline. Medicare’s hospice policies have not been adjusted in over 40 years, further adding to the necessity of revising benefits to include coverage of palliative treatments while in hospice, given the known cost-effectiveness relative to dying in a hospital. Medicare patients would be more incentivized to access hospice care if palliative, while not curative therapies, were covered. However, demand for hospice aids is outpacing supply and characteristically low wages deter employment growth.10,11
Where people die is one of many factors that will further strain the already faltering health economy. The U.S. population is aging rapidly, with deaths soon outpacing births, and more individuals are living longer with complex chronic conditions. At the same time, healthy life expectancy lags total life expectancy by a decade or more, meaning that additional years of survival often come with substantial morbidity and healthcare needs. In this context, reliance on inpatient and emergency-based EOL care drives spending upward without commensurate gains in quality or value.
Given the scale of the challenge, along with the accelerating demographic pressures, EOL care represents one of the clearest opportunities to bend the cost curve while simultaneously improving “value for money” in the healthcare system. At its core, approaching EOL care with an emphasis on the dignity and quality of life requires a continuum of transparent communication and dialogue about choices. Every American is entitled to know the odds of “success,” rather than focusing on the hope of what the “outcome” might be.