Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner

AHCPR Research on Long-Term Care

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

As the lead Federal agency charged with supporting and conducting health services research, the Agency for Health Care Policy and Research (AHCPR) has undertaken and funded important studies in each of these areas of long-term care research. The bibliographies that follow provide citations for selected findings as well as a complete listing of long-term care research conducted and funded by AHCPR since 1990.


Use, Cost, and Financing
Access and Quality of Care
Organization and Delivery of Care
Consumer and Caregiver Behavior
Special Populations
Data Development and Methodological Studies
Published Papers, 1990-96
   Use, Cost, and Financing
   Access and Quality of Care
   Organization and Delivery of Care
   Consumer and Caregiver Behavior
   Special Populations
   Data Development and Methodological Studies
Funded Grants, 1990-96


Long-Term care is an increasingly important and rapidly changing component of today's health care delivery system. Four out of every ten people turning age 65 will use a nursing home at some point in their lives, and many will need home care and other related services as well. As the population ages, the need for these services will continue to grow, particularly for women. Long-Term care services are essential to many younger populations as well—children with disabilities, people with mental health problems, people with Alzheimer's disease, people with acquired immunodeficiency syndrome (AIDS), and others.

The increasing need for these services is creating significant budget concerns for Federal and State Governments, as well as straining family finances. Combined Medicare and Medicaid outlays have been growing dramatically. About 40 percent of long-term care costs are paid by the Federal/State Medicaid program. Although the Medicare program accounts for only a small share of total expenditures, its share has been growing. Despite rising Government expenditures, out-of-pocket payments continue to be a large source of financing for long-term care. As a result, for many individuals who have chronic care needs, long-term care remains a catastrophic cost.

These financial pressures, combined with similar pressures related to acute care, are fueling unprecedented changes in the health care marketplace. Long-Term care providers are diversifying and combining forces to maintain and expand market share as the influence of managed care spills into the long-term care market. Pressures to reduce costs have pushed sicker and more disabled persons into lower levels of care. These market and delivery system changes, in turn, are underscoring questions about the appropriateness, cost, and quality of services delivered in each of these settings and are prompting increased interest and concern on the part of consumers, providers, and Federal and State Governments charged with regulating and paying for these services. In recent years, for example, Congress has considered prospective payment systems for nursing home and home health agency payments under the Medicare program, as well as proposals to diminish the Federal role in the Medicaid program. Many States are exploring managed care options for the elderly and disabled under Medicaid.

As they consider how to respond to these cost pressures and market changes, purchasers, providers, consumers, and policymakers will need answers to fundamental questions in six broad areas:

  • Use, cost, and financing.
  • Access and quality of care.
  • Organization and delivery of care.
  • Consumer and care giver behavior.
  • Special populations.
  • Data development and methodology.

The following sections highlight selected findings.

Return to Contents

Use, Cost, and Financing

The growing cost of long-term care has stimulated policymakers to develop new approaches to control public expenditures and also has altered market strategies for providing that care. Accurate estimates of use and cost and identification of factors that influence care decisions provide the basis for designing new public policies and new market strategies to meet demand with fewer resources.

AHCPR's research has documented the high use of care by, and expenditures for, elderly and long-term care populations in general and the significant overlap of long-term and acute care:

  • In 1987, per capita health care expenditures (in 1992 dollars) for the elderly living in the community were over $6,000 compared with $1,700 for the nonelderly. Per capita expenditures for nursing home residents were about $29,000 (Cohen, Carlson, and Potter, 1995).
  • Nursing home residents are also important users of hospitals. About 9 percent of all Medicare hospital admissions are transfers from nursing homes. About 28 percent had one or more hospital visits during their nursing home stay. (Murtaugh and Freiman, 1995; Freiman and Murtaugh, 1995).

Nursing home costs represent about 70 percent of long-term care expenditures, as well as a large part of AHCPR long-term care research. Studies such as the following document the likelihood of nursing home use, reliance on Government funds, and characteristics of users:

  • About 43 percent of persons turning age 65 will use a nursing home before they die. About 20 percent of users will spend 5 or more years there (Kemper and Murtaugh, 1991). Of those turning 65, 17 percent can expect to use a nursing home and receive Medicaid reimbursement (Spillman and Kemper, 1995).
  • More than 70 percent of nursing and personal care home residents are women and two-thirds of them are widowed or divorced. About 40 percent are demented and about 59 percent require assistance with four or more Activities of Daily Living (ADLs) (Lair and Lefkowitz, 1990).
  • About 10 percent of residents in nursing and personal care homes are under age 65 (Lair, 1992) and 11 percent do not need help with ADLs (Lair and Lefkowitz, 1990).

Other studies estimate the rate of nursing home use by age and functional status, and, for users, expenses per day and source of payment (Feinleib, Cunningham, and Short, 1994; Short, Feinleib, and Cunningham, 1994). Examples of additional studies include a model of hospital discharges to nursing homes (Roberge, Grant No. HS07953), predictors of long stays in nursing homes (Muller, Grant No. HS06672), and a comparison of nursing home admissions in urban and rural areas (Netzer, Grant No. HS000088). Differences in long-term care use and expenditures among African-American, Hispanic, and white populations have also been studied (Pourat, Grant No. HS08034; Proctor, Grant No. HS06406; Wallace, Grant No. HS07672).

AHCPR studies have also contributed to our knowledge of home care use and expenditures, as these examples illustrate:

  • Nearly 6 million persons used home health services in 1987 (Altman and Walden, 1993).
  • The most frequently used home care services are home health aides and homemakers, providers that frequently are not covered by either public or private insurance (Altman and Walden, 1993).
  • Women were twice as likely to use home care as men. Those persons aged 85 and older, widows, those living alone, and those having difficulties with basic daily activities were the most likely to have a home care visit (Altman and Walden, 1993; Short and Leon, 1990).

Generally, home care is provided to a less functionally dependent population, compared with nursing home care, but policies have been designed to try to encourage home care use in lieu of nursing homes, as AHCPR studies have shown:

  • Married persons are half as likely as unmarried persons to be admitted to a nursing home; having at least one daughter or sibling reduces those chances by about one-fifth (Freedman, 1996).
  • Generous home care programs increase the likelihood that unmarried persons will live independently rather than live in shared housing or enter a nursing or personal care home (Pezzin, Kemper, and Reschovsky, 1996). Evidence was not found that these interventions displaced informal care (Kemper and Pezzin, 1996).

Although there is little evidence that subsidized home care reduces aggregate long-term care costs, one study suggests that if home care services were more tightly targeted to appropriate individuals, the potential for program-level cost savings might be significantly improved (Greene, Lovely, Miller, and Ondrich, 1995).

Studies of the financing of long-term care have analyzed the impact of waiting periods, coinsurance, and deductibles on coverage and the percentage of lifetime cost that is covered (Kemper, Spillman, and Murtaugh, 1991; Short, Kemper, Cornelius, and Walden, 1992). The impact on public costs has also been studied (Short and Kemper, 1994). Another study found that between 12 and 23 percent of persons would be ineligible for private long-term care insurance because of underwriting criteria if everyone applied at age 65 (Murtaugh, Kemper, and Spillman, 1995).

Return to Contents

Access and Quality of Care

Concern for controlling the cost of long-term care has increased interest in ways to reduce home care costs by targeting home care programs to those who need the most care. AHCPR studies have contributed to this debate with estimates of the number of persons with long-term care needs and the numbers that are eligible under different criteria. Studies have also contributed to questions of appropriateness of nursing home placements:

  • About 20 percent of the elderly living in the community (5.4 million persons) in 1987 had difficulty with at least one instrumental or basic ADL (Leon and Lair, 1990). About 4 million of these persons need human help with these activities (Stone and Murtaugh, 1990). About one-fourth of those needing human help are cognitively impaired, and about one-half receive help with just ADLs (Spector, 1991).
  • A conservative estimate is that 15 percent of nursing home residents could be cared for at lower levels of care (Spector, Reschovsky, and Cohen, 1996).
  • About one million persons aged 65 and older living in the community would be eligible under criteria of needing active or standby help with three or more ADLs. The number would increase as much as 70 percent if eligibility were included for cognitively impaired persons, depending how that was defined (Spector, 1991). Other criteria are also potentially important, such as the need for complex medical treatments (Kemper, 1992).

One study illustrated the tradeoffs of changing targeting criteria and the difficulties of establishing criteria using simple eligibility measures such as ADLs and cognitive impairment (Spector and Kemper, 1994). Another study found less access to nursing home care for Medicaid-covered persons as opposed to private payers (Reschovsky, 1996).

AHCPR research has also made major contributions to the understanding of what factors affect the quality of long-term care services. The Agency sponsored development of guidelines for prevention and treatment of pressure sores, detection and prevention of depression, prevention of incontinence, and assessment of Alzheimer's disease—all common problems faced by long-term care patients. These guidelines have been adapted by the American Medical Directors Association for use in long-term care facilities. Evaluations of the implementation of guidelines have also been funded, including a study of the implementation of the AHCPR urinary incontinence guideline (Watson, Grant No. HS08491).

In addition to sponsoring development of clinical guidelines, AHCPR has researched clinical factors associated with important clinical problems facing long-term care patients. Studies include risk of pressure sores (Spector, 1994); behavior problems (Spector and Jackson, 1994); basic functioning in daily activities (Spector, 1995; Spector and Takada, 1991); and sensory impairments (Laforge, Spector, and Sternberg, 1992).

A number of additional studies have focused on long-term care quality. A study currently underway is looking at the difference in outcomes for home health care patients in urban and rural areas and examining differences in both the quantity and quality of care (Shaughnessy, Grant No. HS08031). One study developed a nursing home quality information system (Harrington, Grant No. HS07574), while another study has tested a facility-level outcome quality assurance system (Caro, Grant No. HS07585). Other research includes an analysis of consumer preferences for restraints versus other protocols for preventing falls (Mion, Grant No. HS06923) and a study of psychoactive drug regulations (Maloney, Grant No. HS07954).

Return to Contents

Organization and Delivery of Care

The changing dynamics of the health care system have affected how long-term care is organized and delivered, and the major questions focus on how these changes have affected quality, access, and cost. The Agency has been engaged in a number of studies in which the organization of care and the structure of long-term care industries and markets have been the focus. The results of some of these studies follow:

  • Cost-based reimbursement increases the number of registered nurses, but reduces the number of practical nurses compared with flat-rate reimbursement. Higher reimbursement levels encourage the employment of practical nurses (Cohen and Spector, 1996).
  • Higher professional nursing staff levels and lower turnover of staff improve nursing home outcomes (Cohen and Spector, 1996; Spector and Takada, 1991).
  • Nursing facilities with few private pay residents are associated with negative outcomes (Spector and Takada, 1991).
  • For-profit nursing homes have lower operating costs (Cohen and Dubay, 1990) but higher hospitalization rates than nonprofits (Freiman and Murtaugh, 1993).
  • Case management can reduce cost per participant and maintain quality if the agency has power to authorize public payments, maintains an average expenditure cap, and provides the case management services itself (Kemper, 1990).

Another study developed a theoretical framework for explaining the existence of nonprofit nursing home ownership in a market for long-term care characterized by less information for consumers than providers and the underprovision of quality (Hirth, Grant No. HS06934). This study received the Kenneth Arrow Health Economics Award in 1994.

Additional AHCPR-funded research includes the changing structure of the home care industry, clients, and outcomes (Peters, Grant No. HS08892); a comparison of unlicensed and licensed home care providers (Estes, Grant No. HS06860); State variation in nursing home and home health markets (Harrington, Grant No. HS06174); the dynamics of the hospice industry (Hamilton, Grant No. HS06619); a comparison of social HMOs and risk-based Medicare HMOs (Dowd, Grant No. HS07171); and the supply of special care units for Alzheimer's patients in nursing homes (Leon, Potter, Cunningham, 1990; 1991).

Return to Contents

Consumer and Caregiver Behavior

Studies of caregiving behavior provide a basis for understanding the dynamics of home care demand and supply and help guide long-term care policies. Because long-term care is generally a family decision, research encompasses not only the disabled person but also the family and other informal caregivers. AHCPR has conducted a number of studies on consumer and family responses to long-term care needs. Findings from these efforts indicate that:

  • About 95 percent of noninstitutionalized elderly persons in need of long-term care rely on family members and friends for help with activities of daily living. About 70 percent of the combined community and institutional population receiving long-term care rely entirely on private resources (Spillman and Kemper, 1992).
  • Publicly-provided formal home care results in only small reductions in the overall amount of care provided by informal caregivers (Pezzin, Kemper, and Reschovky, 1996) while enabling elderly persons with disabilities to live more independently and to reduce their probability of institutionalization (Kemper and Pezzin, 1996).
  • The effectiveness of alternative government long-term care policies, such as caregiver allowances and other cash incentives designed to promote family care, depends critically on whether transferred resources are targeted to the elderly person or the caregiver (Pezzin and Schone, 1997).

Other studies have reached the following conclusions:

  • Caregivers of elders with greater care needs are more likely to take unpaid leave, reduce work hours, or rearrange their work schedules to assume elder care responsibilities. Being female, white, and in fair-to-poor health also increased the likelihood of the caregiver requiring accommodation in his or her work (Stone and Short, 1990).
  • Major determinants of caregiver attrition are the degree of dependence of the frail elderly care recipient and the need for help on demand (Boaz and Muller, 1991).

Other research has attempted to quantify the relationships between employment, caregiving hours, and financial assistance from the adult children (Boaz and Muller, 1992).

Return to Contents

Special Populations

Long-Term care services are required by several different populations, and the needs of these populations vary. In addition to the elderly, many of the long-term care users are younger persons with physical disabilities; persons with developmental disabilities (DD); and persons with chronic diseases such as diabetes, emphysema, and AIDS. AHCPR studies of special populations indicate the following:

  • About 95 percent of the facilities for the mentally retarded are community based, but they represent only about 60 percent of the beds because of large State institutions (Beauregard and Potter, 1992).
  • About half of persons with DD in institutions are admitted after they are past the age of 40. About two-thirds of persons with DD aged 65 or older are in nursing homes or other long-term care facilities that are not specially designated for DD (Altman, 1995a).
  • The most ADL-disabled persons tend to reside in State institutions, and the least disabled tend to reside in small private or public facilities (Cunningham and Mueller, 1990).
  • The type of services needed are very similar across age groups, although the reasons for these needs differ (Altman, 1995a).

AHCPR has funded a substantial amount of research that examines service use, costs, and Medicaid coverage of HIV-infected patients. Those studies have determined the following:

  • Home care use is the greatest among persons with AIDS, compared with persons at less advanced stages, such as HIV+ status (Buchanan, 1995).
  • HIV-infected persons covered under Medicaid are more likely to receive paraprofessional help than those covered by private insurance (Fleishman, 1997).
  • The course of functional loss is variable and episodic, which suggests the need for flexible and responsive systems for authorizing and managing in-home services for persons with HIV (Crystal and Sambamoorthi 1996).

Other HIV-related research include a study of access to home care and use (Hanley, Grant No. HS06404), resource use in chronic care facilities (Blustein, Grant No. HS00034), and models to predict service intensity and visit-related costs (Payne, Grant No. HS06843). One other study looked at the effects of case-managed home care on the health of AIDS caregivers (Reynolds, Grant No. HS06971).

Return to Contents

Data Development and Methodological Studies

AHCPR has also been an important contributor to the infrastructure of long-term care research through its survey activities. Over the years, AHCPR surveys have been a major source of data for policy and basic research in long-term care:

  • The 1987 National Medical Expenditure Survey (NMES) included data for persons living in the community, persons institutionalized in nursing and personal care homes, and persons in intermediate care facilities for the mentally retarded.
  • The 1996 Medical Expenditure Panel Survey (MEPS) provides population-based information on the consumers of health care, the services used (including home care), and expenditures made regardless of insurer. The nursing home component of MEPS provides information on a nationally representative sample of nursing homes and their residents.
  • The HIV Cost and Services Utilization Study (HCSUS) is a national sample of persons with HIV infection who are in treatment. Long-Term care issues that can be addressed include utilization of home health care and nursing facility care, case management services, and patterns and correlates of health-related quality of life, including functional status. The AIDS Cost and Services Utilization Study (ACSUS), the predecessor of HCSUS, was also funded by the Agency.

The development of national estimates of long-term care use and expenditures from the institutional component of NMES and other large data sets involves many methodological issues, some of which in themselves have contributed to the statistical and measurement literature. Research staff have contributed to methodological studies in diverse ways using statistical methods to develop nationally representative estimates of measurement issues. In addition, AHCPR has funded projects to improve research methodologies for studying long-term care populations.

Examples of statistical and measurement studies include the following:

  • Problems of nonresponse in the development of year-long profiles of nursing home residents and their movement in and out of other settings (Cohen and Potter, 1990).
  • Imputation and weights for national estimates (Cohen and Potter, 1993; Cohen, Potter, and Flyer, 1994).
  • Episode-based estimates of nursing home length of stay, which include short-term hospitalizations (Short, Cunningham, and Mueller, 1990).
  • Development of a nationally representative sample of decedents who used nursing homes during their lives (Carlson, Kemper, and Murtaugh, 1995).

Other papers have focused on the measurement of disability and cognitive impairment with national data (Altman, 1994; Spector, 1990), and the measurement of ADL change (Mathiowetz and Lair, 1994). Research has also been funded to improve methods to study duration in nursing homes with multiple outcomes using hierarchical duration models (Morris, Grant No. HS07306; Morris, Norton, and Zhou, 1994).

Return to Contents
Proceed to Next Section


The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care