The Medicare Hospice Benefit: Policymakers Respond to the Challenges of Growth

March 2011


The Medicare Payment and Advisory Commission (MedPAC) reported that in 2009 nearly 1.1 million Medicare beneficiaries received hospice services at a cost totaling 12 billion dollars. During the period from 2000 to 2009, MedPAC also found that the supply of hospices grew 50 percent.1  The growth in the number of hospices has raised questions regarding program design and its effects on costs and quality. As both the use of the benefit and program spending have expanded, MedPAC and other governmental agencies—such as the Government Accountability Office (GAO)—have recommended payment reforms to better reflect the patterns of care that occur in hospice.

The goal of payment reform is to improve both program value and quality of care for beneficiaries. Currently, hospice providers are reimbursed by Medicare at a constant daily rate for each day that the beneficiary is enrolled in hospice. There are four base levels of reimbursement for different categories of care: routine home care, continuous home care, inpatient respite care, and general inpatient care. More than 95 percent of hospice care days fall into the category of routine home care. This method of payment has not been changed since the introduction of the hospice benefit in 1983.2  Since that time, more information has become available about the patterns of care and service utilization in hospice. The new information has led policymakers to consider changing payment structures to better reflect the different levels of intensity of care during a hospice stay.

The GAO first examined hospice payment in a 2004 report.3  They used data from 2000 and 2001 on the mix of services provided—such as home health, nursing, and social services—and compared this data to utilization patterns in 1983. The GAO found that there was a change in the mix of services provided and that this change reflected a shift in the delivery of care and the resources necessary to provide care. In its report, the GAO recommendation to Congress and to CMS was to gather more comprehensive, patient-specific data on hospice use and to determine if payment methodologies needed adjustments.

After studying the issue in recent years, MedPAC has formulated specific recommendations for hospice payment reforms. These recommendations followed from two key findings from its 2008 report4—first, that the constant payment rate does not accurately reflect patterns of care, and secondly, that the current methodology incentivizes a longer length of stay (LOS). The average LOS in hospice has increased significantly in the last decade from 54 days in 2000 to 86 days in 2009. Additionally, during the same time period, the longest LOS (those in the 90th percentile) grew from 141 days to 237 days.5 While MedPAC found that this longer LOS can be partly attributed to more patients with non-cancer diagnoses choosing hospice benefits, the trend also raises questions as to whether some providers are engaging in patient selection under a current payment structure that makes long stays more profitable than short stays. In order to realign incentives and promote greater accountability in the provision of the hospice benefit, MedPAC has recommended that hospice payment should change from a constant daily rate to a “u-shaped” payment curve that better represents patterns of care. This would result in higher payments per day at the beginning and end of episodes of care—when resource use tends to be the greatest—and lower payments per day in the interim period as the length of the episode increases.

The work from GAO and MedPAC on the hospice benefit has set the stage for restructuring the benefit to promote better value and quality of care. While there are gaps in the knowledge base—particularly around the issue of quality of care—there is a body of research that can provide insights and add to payment reform discussions. With increased focus on cost containment and quality improvement, other research in this area can help to inform Medicare policy.

Selected Research on Hospice Costs and Setting of Care

Previous research has focused on the effect of hospice on Medicare expenditures. In a HCFO-funded study, Donald H. Taylor, Jr., Ph.D., that hospice use reduced expenditures by an average of $2,309 per hospice user. He also found that the maximum reduction in Medicare expenditures per user was approximately $7,000, and occurred when the beneficiary had a primary diagnosis of cancer and used hospice for the last 58 to 103 days of life.6  In addition, Taylor examined whether the presence of an informal or formal care network in a beneficiary’s home leads to different hospice utilization patterns near death.  He and his colleagues found that beneficiaries receiving formal home care had a much higher chance of enrolling in hospice prior to death.7  He concluded that future research should examine whether the absence of formal caregivers impeded hospice access for beneficiaries.

Other research has focused on the setting of care at the end of life. In previous HCFO-funded work, Korbin Liu, Sc.D., studied nursing home use by  dual eligibles in their last year of life. He found that use of hospice made beneficiaries more likely to die in a nursing home than a hospital. He and his colleagues recommended improving access to hospice care as one avenue to avoid hospital deaths.8  More recently, Susan Miller, Ph.D., of Brown University, examined the growth of hospice care in nursing homes specifically and the implications for beneficiaries. She and her colleagues found that between 1999 and 2006 rates of nursing home hospice use more than doubled, and with this growth there was also a doubling of the mean length of stay and an increase in the proportion of nursing home decedents with non-cancer diagnoses.9  She concluded that efforts to curb Medicare hospice expenditures should consider potentially negative access effects for nursing home residents with non-cancer diagnoses.

Future Directions: Quality of Care 

In its March 2010 report, MedPAC noted a lack of sufficient evidence to assess quality of care in hospice. Current and forthcoming research may help to fill the knowledge gaps in this important area. A recent report from the Robert Wood Johnson Foundation, “Improving Care at the End of Life: How the Robert Wood Johnson Foundation and its Grantees Built the Field”10  highlights the contributions of the foundation to all facets of the field of palliative care. These contributions include past work, conducted under the foundation’s Targeted End-of-Life Projects Initiative, which focused on tool development to assess quality of care in hospice. Led by Jean Teno, M.D., M.S., of Brown University, this project resulted in a web-based toolkit and family survey to help health care facilities improve end-of-life care.11 This work led to the adoption of the family survey by the National Hospice and Palliative Care Organization, which represents 80 percent of hospices in the United States.

The Centers for Medicare and Medicaid Services (CMS) has also supported initiatives to improve the quality of hospice care. In 2006, CMS began the PEACE Project in conjunction with the Carolinas Center for Medical Excellence, Medicare’s quality improvement organization for North and South Carolina. The purpose of the project was to identify quality measures on end-of-life care and analyze the instruments available to gather data on those measures. Recent work on the development of these measures was published by a team from the Carolinas Center for Medical Excellence. Their work led to specifications and data collection tools for 34 PEACE quality measures that were highly rated by experts in hospice and palliative care.12

Current HCFO-funded research will also help to inform the knowledge base on quality of care at the end of life. In a recently awarded grant, Donald H. Taylor, Jr., Ph.D., will assess the relationship between hospice use, quality, and costs in the Medicare program.  The purpose of this research is to focus concurrently on costs and quality when assessing the value of palliative care and to help inform payment policies to promote sustainable financing mechanisms for this care. While quality of care in hospice has been an understudied issue, these efforts will provide more evidence on quality and inform efforts to reform the hospice benefit. 


As the number of Medicare beneficiaries increases the trend of growth in hospice care is likely to continue. More expansion in benefit use will lead to greater need for evidence-based recommendations to guide CMS in any reform efforts. Research can help to inform payment policies that ensure continued program viability and access to care. Details on studies led by HCFO grantees can be found at

Related HCFO Grants

Title: The Treatment of Dying Medicare Managed Care Patients: The Role of Social and Economic Factors
Grantee Institution: Health Research and Educational Trust 
Principal Investigator: Jon R. Gabel
Grant Period: November 1, 2002-June 30, 2005

What is the cost and utilization of services during the last two years of life for Medicare managed care patients and fee-for-service Medicare patients? The researchers analyzed data from provider, beneficiary, plan benefit, prescription drug, clinic and office encounters, and laboratory and x-ray services database files for the managed care and fee-for-service populations. Patients who disenrolled from United HealthCare's Medicare managed care and returned to fee-for-service were also studied. Major causes of death such as cancer, chronic obstructive pulmonary disease [COPD], chronic heart failure, stroke, and dementia were analyzed for utilization and costs. The researchers addressed the following questions: 1) What are the costs and use of services associated with end-of-life care for major causes of death? 2) For the major causes of death being studied, how do cost and treatment patterns in end-of-life care vary according to area resources and financial arrangements? 3) What are differences in the site of death of Medicare managed care patients for these three major causes of death? 4) How does continuity of care vary among managed care settings? 5) What are the economic and other factors that determine continuity of care? 6) What is the cost of prescription drug coverage for end-of-life patients? 7) What is the appropriate of prescribing and dispensing for the five study conditions? 8) What economic and socio-demographic factors explain differences across areas in end-of-life prescribing? The objective of the project was to guide administrators, legislators, and providers as they make decisions about end-of-life care.

Title: Factors Affected End-of-Life Care for Beneficiaries Who are Dually Eligible for Medicare and Medicaid
Grantee Institution: The Urban Institute
Principal Investigator: Korbin Liu, Sc.D.
Grant Period: October 1, 2002-June 30, 2005

What are the factors affecting end-of-life care for the dually eligible? Using the “Multi-State Dual Eligible Data Files” developed by Mathematica under contract with CMS, researchers at the Urban Institute examined (1) services provided by the Medicare and Medicaid programs, as well as variations among states; (2) the composition of and proportion of expenditures on end-of-life care (e.g., hospital, physician, prescription drug, long-term care); (3) variation in use and expenditures over time between health and long-term care services; and (4) variation by race and age in utilization, expenditures and source of financing end-of-life care. The objective of this study was to inform policymakers and providers about the patterns of use of care at the end-of-life by dual eligibles in order that they might re-examine and improve current policies addressing such care.

Title: Does Hospice Save Medicare Money?
Grantee Institution: Duke University
Principal Investigator: Donald H. Taylor, Jr., Ph.D. 
Grant Period: October 1, 2003-September 30, 2006

Does hospice save money for the Medicare program and does hospice have any effect on out-of-pocket expenses incurred by families of terminally ill Medicare beneficiaries? The study focused on the Medicare hospice benefit. The researchers conducted a literature review and empirical analyses to answer the above questions. They hypothesized that “hospice can provide effective palliative treatment to terminally ill patients while saving money for the Medicare program, by reducing expensive curative care that must be foregone to obtain hospice benefits.” The objective of this study was to inform policymakers about key issues that affect the ability of hospice to save money, including length of stay, underlying disease, and the propensity of hospice and non-hospice patients to use health care services.

Title: Identifying the Use, Cost, and Quality Tradeoff in the Medicare Hospice Benefit
Grantee Institution: Duke University
Principal Investigator: Donald H. Taylor, Jr., Ph.D. 
Grant Period: March 1, 2011-February 28, 2013

The researchers will assess the relationship between use, quality and cost of palliative care across the settings in which this care is delivered, including community-base palliative care, hospital-based palliative care, community-based hospice, and inpatient hospice. Research questions include: (1) What is the relationship between palliative care use, quality, and cost in North Carolina palliative care programs; (2) What is the relationship between hospice use, quality, and cost in the Medicare program; and (3) Do newly developed quality metrics yield a similar pattern of use, quality, and cost in prospective North Carolina palliative care data?  The objective of the proposal is to focus on quality and cost concurrently in assessing the value of palliative care and help inform policy around the development of sustainable financing mechanisms for this care.  

1. Medicare Payment Advisory Commission. (2011). Report to the Congress: Medicare Payment Policy. Washington, DC: MedPAC.  
2. Medicare Payment Advisory Commission, 2011.                                                                    
3. Government Accountability Office. (2004). Medicare Hospice Care: Modifications to Payment Policy May Be Warranted (GAO-05-42). Washington, DC: Government Printing Office. 
4. Medicare Payment Advisory Commission. (2008). Report to the Congress: Medicare Payment Policy. Washington, DC: MedPAC.  
5. Medicare Payment Advisory Commission, 2011.
6. Taylor Jr., D.H. et al. “What Length of Hospice Use Maximizes Reduction in Medical Expenditures Near Death in the US Medicare Program,” Social Science & Medicine, Vol. 65, No. 7, June 2007, pp. 1466-78.
7. Van Houtven, C.H. et al. “Is a Home-Care Network Necessary to Access the Medicare Hospice Benefit?,” Journal of Palliative Care, Vol. 12, No. 8, August 2009, pp. 687-94.
8. Liu, K. et al. “Nursing Home Use by Dual-Eligible Beneficiaries in the Last Year of Life,” Inquiry, Vol. 44, No. 1, March 2007, pp. 88-103. 
9. Miller, S.C. et al. “The Growth of Hospice Care in U.S. Nursing Homes,” Journal of the American Geriatric Society, Vol. 58, No. 8, August 2010, pp. 1481-88.                                   
10. Patrizi, P. et al. “Improving Care at the End of  Life: How the Robert Wood Johnson Foundation and its Grantees Built the Field,” RWJF Retrospective Series, March 2011. 
11. Teno, J.M. “Web-based Toolkit Features Family Surveys and Resource Guides to Help Health Care Facilities Improve End-of-Life Care,” The Robert Wood Johnson Foundation, Grant Results, July 2006. See also:  
12. Schneck, A.P. et al. “The PEACE Project: Identification of Quality Measures for Hospice and Palliative Care,” Journal of Palliative Medicine, Vol. 13, No. 12, December 2010, pp. 1451-59.