Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner
Future Directions for Community-Based Long-Term Care Health Services Research

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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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.

Expert Meeting Summary, June 20-21, 2000

In June 2000, long-term care experts met to help guide the long-term care research agenda for the Agency for Healthcare Research and Quality (AHRQ). The meeting was cosponsored by AHRQ's Center for Organization and Delivery Studies and Center for Cost and Financing Studies.

Select for PDF File (251 KB). PDF Help.

By William D. Spector, Thomas J. Shaffer, R. Tamara Hodlewsky, Jan J. De La Mare, and Jeffrey A. Rhodes


Day One
Long-Term Care Research Priorities
   Quality Dimension Categories
      Organization and Delivery
      Financial and Market Incentives
      Consumer Issues
   Immediate Research Priorities
      Workforce Issues
      Family Decisionmaking
   Special Concerns Regarding Children and Adults With Disabilities
Quality Measurement
   Determining the Scope of Quality Concerns in Community-Based Long-Term Care
   Defining a Quality Indicator
   Other Conceptual Issues
   Measuring Outcomes in Home Health Care: OASIS
   Recommendations About Specific Quality Measures
      Global Measures
      Service-Specific Measures
Day Two
Data Presentations
   Medical Expenditure Panel Survey (MEPS)
      Community-Based Variables
      Home Health Care
   Medicare Current Beneficiary Survey (MCBS)
      Community-Based Variables
      Home Health Care
   Health and Retirement Study (HRS)
      HRS Questionnaire Topics
   Health Dynamics Among the Oldest Old (AHEAD)
      AHEAD Questionnaire Topics
      Other Variables
   National Long-Term Care Survey (NLTCS)
      Community Questionnaire
      Informal Caregiver Survey
   National Home and Hospice Care Survey (NHHCS)
      Agency Variables
   Patient Variables (Current and Discharged)
Improving Data
Identifying and Prioritizing Data Gaps
Other Data Issues
   Measuring Unmet Need in ADLs and IADLs
   State-Level Data
   Capturing Transitions
   Special Issues Related to CSHCN
   Measuring Home Care Quality and Staffing
1. Meeting Agenda
2. List of Meeting Participants
3. Participant Information and Pre-meeting Questionnaire
1. Summary of Quality Dimensions
2. Priority Long-Term Care Quality Concerns


On June 20-21, 2000, the Agency for Healthcare Research and Quality (AHRQ) sponsored "Future Directions for Community-Based Long-Term Care Health Services Research." This meeting of long-term care experts was convened by AHRQ's Center for Organization and Delivery Studies (CODS) and Center for Cost and Financing Studies (CCFS) to help guide the Agency in developing its long-term care agenda.

This was the second of two AHRQ-sponsored meetings on long-term care health services research. (Go to Appendix 1 for the meeting agenda.)

The first meeting, "Future Directions for Residential Long-Term Care Health Services Research, was held in October 1999.1 That meeting focused on residential long-term care, which included both assisted living and nursing home care. The scope of the second meeting was community-based long-term care, including both formal and informal care.

Community-based long-term care services were defined as services that help maintain community living, maximize independence, enable social integration, and provide supports for instrumental activities of daily living (IADLs) and activities of daily living (ADLs). IADLs include activities such as shopping, help with taking medicines, and help with light housework; ADLs include help with activities such as bathing, dressing, and eating.

The meeting focused on community-based services including service-enhanced housing settings, such as group homes and personal care homes. Assisted living was discussed as part of the first meeting on residential care.

Specifically, the goals of the meeting were to:

  • Identify the highest research priorities for community-based long-term care.
  • Identify the most important quality indicators and identify measurement needs to help set priorities for the future development of quality measures.
  • Assess the capability of current survey and administrative data to answer priority research questions.
  • Get advice on data strategies to identify and reduce data gaps.

The 2-day meeting was based on experts' responses to a series of questions asked of the participants prior to the meeting. (Go to Appendix 2 for a list of the meeting participants.) The questions focused on research priorities, quality indicators, and adequacy of existing national databases.

1 Spector WD, Potter DEB, De La Mare J. Future Directions for Residential Long-Term Care Health Services Research. Expert Meeting Summary, October 14-15, 1999. Rockville, MD: Agency for Healthcare Research and Quality; 2000. AHRQ Pub. No. 01-0007.

Return to Contents

Day One

Long-Term Care Research Priorities

Before the meeting, participants were asked to list their top five research and policy questions concerning cost, quality, and outcomes of community-based long-term care (Select to access Appendix 3). Respondents were asked to focus on the age group best corresponding to their primary field of expertise: children with special health care needs (CSHCN), adults with disabilities, or the elderly.

Although we were interested in high priority topics, their combined responses covered an expansive range of issues, which we assigned into six categories:

In some cases topics overlapped these categories, but assignment was made to only one category based on the major focus of the topic. (Go to Table 1 for a summary of these dimensions.)

Quality Dimension Categories

Organization and Delivery. Participants identified a number of priorities related to organization and delivery issues. The major concerns were the impacts of organizational factors and program design on cost and outcomes. Some respondents wanted more evaluation of models of care that integrate long-term care and acute care, while others stressed the need to evaluate new models (e.g., care being provided in existing housing arrangements or neighborhood-based designs).

There was also interest in the impact of other aspects of community care such as consumer-directed care and case management. The appropriate roles of clinicians in providing care were also a concern (e.g., physicians vs. nurses, nurses vs. aides).

Equity/Access. Respondents believed that issues related to access and equity were important. Included in the equity/access category were research on racial and ethnic disparities in access to community care, the need to develop models of care sensitive to cultural differences, and the need to better understand how financing influences access. When focusing on care for children, respondents emphasized the need to assure that long-term care services were integrated with educational services.

Financial and Market Incentives. Many respondents emphasized the need to study the overall impact of Medicare home health prospective payment. Others were concerned about how the aging of the population would affect financing of long-term care. An additional concern was to better understand the changing market place for long-term care, especially the implications of the growth of managed care, the decrease in Medicare home health spending, and the expanding assisted living market.

Consumer Issues. There were three main themes promoted by respondents concerning consumer issues.

  • First, they recommended evaluation of models of care that are sensitive to consumer preferences and expand consumer choices.
  • Second, they were concerned about family caregiving and understanding the costs and burdens of that care. There was interest in a number of topics including the decisionmaking process that determines the mix of formal and informal care, prevention of caregiver health problems, and the prevention of elder abuse. For children and adults with disabilities, there was concern that we have insufficient data to currently describe basic patterns of care and caregiver burdens.
  • Third, there was interest in improving ways to provide information to families such as quality report cards and approaches to promoting self-directed care.

Quality. Many respondents included research on quality as a high priority. For some, this meant monitoring certain aspects of quality, such as quality of life, unmet needs, and safety of the home environment. Also, there was concern about how regulatory approaches and the mix of informal and formal care affect the overall quality of care provided. There was also interest in improving care guidelines for long-term care. Because patients frequently move from one level of care to another, there was interest in decisionmaking that results in care transitions and the appropriateness of care settings.

Finally there was great interest in being able to understand what were the best staffing patterns for long-term care, including the amount of staff and the skill mix. Because of the apparent shortage of staff throughout the long-term care spectrum, there was interest in improving ways to hire and retain staff.

A final area of concern was medication management.

Methodology. The final category of priorities was methodology. The main methodological priority for participants was the improvement of measures. Two areas for improved measures were consumer satisfaction and processes of care. Specialized areas mentioned were satisfaction measures for the cognitively impaired and measures to assess the burden of informal care. For children, there was interest in better measuring functional limitations. Finally, there was interest expressed in improving projections of long-term care needs.

Immediate Research Priorities

The combined list of research questions was narrowed to a shorter list of immediate priorities during the course of the discussion. Three priority topics emerged as especially high priorities: outcome studies, workforce issues, and family decisionmaking. Basic information about needs and utilization for children and adults with disabilities was also a very high priority.

Outcomes. Participants suggested that one of the highest priorities would be a focus on outcomes in community-based care. Once important patient risk factors were identified, the influences of care provision, health system, and market factors could be studied. Participants noted, however, the need for more research on risk-adjusted outcome measures. Areas for which measures remain lacking are social outcomes, access to care, and family caregiving.

Participants stressed the need for evaluations of services to include cost. A focus on the amount of care needed—or dose-response—was emphasized as very important. Some participants also noted the importance of impact of new technologies in long-term care.

Participants noted that the need for research on evaluation of outcomes applies to each of the three age groups of long-term care users (CSHCN, adults with disabilities, elderly), but outcomes may need to be defined differently for each population. For example, standard measures of functional and cognitive status for the elderly are inadequate for children, whose long-term care goals must be integrated with developmental and educational goals.

Long-Term care outcomes for adults with disabilities must include employment goals as well as other social goals. It was also noted that outcomes may be viewed differently by recipients of care than by the families of recipients.

Workforce Issues. Participants suggested that the recruitment and retention of qualified staff appears to be in crisis across the spectrum of long-term care. Participants emphasized the need for research on the development, preparation, and maintenance of the workforce; approaches to monitoring staff quality; and the effectiveness of recruitment and retention strategies. Research should include the development of new approaches to meeting consumer preferences in staffing and the study of existing approaches such as self-directed care. Approaches to increasing the supply of informal caregivers and assuring the quality of their care should also be studied.

Family Decisionmaking. Participants expressed the importance of more research on how families make decisions about meeting long-term care needs. This was viewed as important for all three age groups:

  • CSHCN.
  • Adults with disabilities.
  • The elderly.

It is important to determine where families receive information and what people know about options for care. Research should include an analysis of the incentives provided by different State programs and reimbursement systems and their effect on choice of care setting. Experts also expressed strong interest in transition decisions, the appropriateness of transfers from one setting to another, and the quality of care provided during transition periods.

Special Concerns Regarding Children and Adults With Disabilities. Unlike long-term care for the elderly population, for which there are a variety of data sources, respondents declared that very little is known about long-term care for children and adults with disabilities. Experts lamented the dearth of data even at a very basic descriptive level, including long-term care needs and utilization by age and by disability. Trajectories and projections of need/use would also be helpful. For these groups, participants believed there was a need for basic descriptive data that could give direction to subsequent research on the appropriateness of care in various settings and the effect of those settings on outcomes.

Participants raised a number of issues that specifically concerned children or adults with disabilities, including the following:

  • Little information exists about the effectiveness of therapies and home health care, especially for children.
  • Transitions in care have not been studied for children; it is especially important to study transitions from home to school and hospital to home.
  • Information about parental caregiving for children is insufficient. Participants stressed that informal care should not be confused with the normal care provided as part of parenting and that the two are difficult to separate in practice. This is similar to separating out normal tasks provided in the role of spouse from caregiving for a disabled spouse, but it may be more difficult.
  • AHRQ's Medical Expenditure Panel Survey (MEPS) was seen as an ideal survey to add questions about one's experience with the long-term care system and to incorporate subjective judgments about the quality of care from both the children's and their families' viewpoints.
  • MEPS is the only national data set that has longitudinal information about children's health services and has good health status data.
  • MEPS was seen as important in describing caregiving arrangements and the relationship between informal and formal care, and caregiver burden. It remains the only source of caregiving information for these populations.
  • Health status measures, in general, need improving. More developmental work is needed to adapt ADL and IADL measures for children. Measuring outcomes for children is difficult because of the particular diversity of conditions they may have.

Quality Measurement

In the afternoon of day 1, the participants discussed quality issues related to community-based long-term care. The discussion had three goals:

  • To determine how quality concerns in community-based, long-term care differed from residential long-term care.
  • To determine if quality concerns differed for the three long-term care subpopulations.
  • To get advice about specific long-term care quality measures that could be achieved with existing data, especially measures that would be appropriate for an annual national quality report that has been mandated by Congress. (The purpose of that report is to describe the changing state of quality in the health care system over time.)

Determining the Scope of Quality Concerns in Community-Based Long-Term Care

Prior to the meeting, the participants were shown a table that specified the scope of quality concerns associated with residential care that was an outgrowth of the October 1999 meeting on residential care. They were asked to add to this table so that it would reflect concerns of both institutional and community-based long-term care. They were asked also to think about the three populations of long-term care users and to focus on the population in which they were most expert. Table 2 presents the revised and expanded table of quality dimensions.

The most obvious addition to the table was the inclusion of informal caregiving concerns. Table 2 responses are categorized into three dimensions of care: caregiving, quality of (formal) care, and quality of life. Caregiving bridges both quality of care and quality of life. With respect to caregiving, there are concerns about the technical skills and interpersonal skills of caregivers, but there is also a concern that the quality of life of the caregiver can be detrimentally affected by caregiving responsibilities. These dimensions were further classified as process and structure or outcome (Table 2).

When discussing informal caregiving, respondents were concerned with the adequacy of informal caregiver training, caregiver burden and stress, negative impacts on the physical health of caregivers, and the economic impact of caregiving—especially impact on work.

In the areas of quality of care and quality of life, respondents added a number of measures that reflected the fact that a community-based care population is less disabled than a residential care population and includes more children and working-aged adults.

Additions included: IADLs, concerns about social integration into the community, ability to engage in productive activities, mental and social development for children, prevention of institutionalization, prevention of accidents during transportation, achieving a high level of autonomy and control, and assuring care is provided in the least restrictive environment.

In addition, respondents added other items that extended the scope of quality concerns (e.g., fears about physical or financial harm, unmet needs, adequacy of care plans, aspects of the home environment, and elapsed time to receive services). In some cases these were concerns that are more important for a community population; and in other cases, respondents suggested aspects of long-term care that were not stressed in the prior meeting but are applicable to all long-term care populations.

Attendees were asked to discuss any other items that should be included in the table as well as other related concerns. The following are some issues that were raised:

  • The importance of quality-of-life measures such as control, autonomy, choice, respect, and dignity. Participants believed these were difficult to measure.
  • Importance of assessing the housing and community environments.
  • The difficulty of measuring satisfaction with care for cognitively impaired persons. Participants believed this information could be provided, although imperfectly, by family members. Research about what can be measured by self-report from cognitively impaired persons is an area for future research.
  • Concern that different clinical disciplines have difficulty communicating with each other while providing care, which sometimes results in clinical errors.

Defining a Quality Indicator

Participants discussed the meaning of the term "quality indicator" and described at least two common uses of the term. One definition is often referred to as a "red flag." A red flag is a value of a measure that triggers concerns that indicates quality problems first, but it is not a direct indicator of a quality problem. A red flag usually leads to reviews of processes of care to determine if care problems exist.

Quality indicators being designed by the Centers for Medicare & Medicaid Services (CMS) for use in the home health survey and certification process and for nursing homes are examples of red flag indicators.

Knowledge of red flags for a particular facility may affect how surveyors spend their time in a facility when reviewing potential quality problems. These indicators also can be used by facilities to monitor potential quality problems throughout the year.

Risk-adjusted outcome measures of quality are also examples of red flag indicators. For most outcomes there is no specific value that would necessarily suggest poor quality. Typically the expected value of the outcome would depend on the casemix of the patients being treated.

One approach is to account for the mix of patients and compare mean outcomes for a risk group with what would be expected if care were adequate (or average). Since it is difficult to perfectly risk-adjust, this approach may not adequately differentiate poor or good quality. Therefore, risk-adjusted outcome measures are typically recommended as part of a quality assurance process rather than indicators of quality per se.

A second definition is a measure that in itself indicates unacceptable care. An example is a measure based on a care guideline if there is strong evidence that the guideline improves outcomes.

An illustration would be turning patients every 2 hours, which is a standard for good care to prevent pressure ulcers. Any immobile resident who is not turned at least every 2 hours would be viewed as not receiving adequate care.

Another example may be outcomes that should not happen if care meets current practice—for example, stage-four pressure ulcers or residents dying of malnutrition.

Even in these cases it may be difficult to set the standard at no incidents because there may be unusual circumstances that would make some small prevalence possible without the existence of a quality problem.

Other Conceptual Issues

When discussing quality, the participants spent extensive amounts of time on conceptual issues. It was agreed that in order to think about specific outcome measures, it was necessary to attribute the outcome to specific services. Participants also stressed the importance of controlling for the health risks of the populations when making comparisons of service-outcome relationships. Some pointed out the importance of accounting for consumer preferences in some way.

The scope of quality concerns depends on the type of services being considered. In adult day care, quality-of-life issues may be very important; examples include comfort of furniture, noisiness of common areas, usability of the library, and ability to choose from a variety of organized activities. In addition, clinical concerns include staff competency and the ability to meet the needs of persons with dementia. For home care, the measures should be more clinically focused and include indicators such as the responsiveness of caregivers, staff turnover, technical competence of the clinical care, and the quality of caregiver-patient interactions.

Quality concerns also may vary by type of long-term care population. For adults with disabilities, some services would have the goal of social integration into the community or attainment of employment. Children may be receiving services to improve mental or physical development to reach age-appropriate developmental goals.

Goals may also depend on the level of disability. An elderly person who is on a ventilator and receiving skilled care from a home health agency may have extensive clinical and functional needs. In contrast, a person with mild cognitive impairment and few physical disabilities who needs help with medicines and transportation may have few clinical issues but may want to be involved in care management decisions.

A final point discussed was the potential for outcomes to conflict and the willingness of people to accept tradeoffs of goals. For example, some patients may be willing to accept less safety to increase physical activity, autonomy, or control. They may prefer a caregiver whom they know personally and trust but who has less training than a stranger from a certified agency. Therefore, the need to incorporate preferences in quality measurement was seen as important. However, few specifics on how this should be accomplished were provided.

Measuring Outcomes in Home Health Care: OASIS

The Outcome and Assessment Information Set (OASIS) refers to a collection of outcomes and associated risk factors developed for home health care. It is currently being used to assess home health care by CMS. Because it is a potentially important source of quality outcome measures, it was discussed extensively throughout the meeting. This section highlights the discussion.

In OASIS, outcome measures are at the patient level and reflect changes in health over two points in time (every 60 days, admission to discharge). Measures are dichotomous, defined as either stabilization or improvement vs. other. Negative outcomes also include hospital readmissions and emergency room use. There are 40 outcomes covering standard domains: functional, physical, cognitive, and emotional.

It is now part of the assessment process for home health agencies; 19 items are also used in the Medicare prospective payment system. Data are transmitted for all Medicare and Medicaid residents who receive skilled care. The goal is to compare risk-adjusted outcomes to a reference group or to performance in the prior year. The system will be used similarly to the Minimum Data Set system to flag potential indicators of poor quality and then to follow up with more intensive review to assess ways to improve processes of care.

The discussion of OASIS focused on its perceived limitations and strengths:

  • It was viewed as an important model for outcome measurement; but since it is limited to home health and is very clinically oriented, it may not be applicable to other home care services.
  • There was concern about using OASIS data for research because it may not be possible to follow all home health "users" over time. For non-Medicaid, non-Medicare persons, or those not receiving skilled care, identifiers are masked, making it impossible to follow persons or link to other data.
  • Its strengths are the uniformity of the data across home health agencies, its integration into the agency assessment process, its use for quality measurement and payment, and the fact that risk-adjusted measures have been developed.

The University of Colorado group is currently developing an OASIS-like set of measures for personal care services in New York State.

Recommendations About Specific Quality Measures

Participants were somewhat reluctant to recommend specific quality indicators. To some extent, this is a reflection of the dearth of peer-reviewed literature on quality indicators. They focused more on what should not be used than what should. The discussion remained at a fairly conceptual level most of the time. Some examples of possible measures were suggested, but without consensus. In some cases, the limitations of the measures were discussed more than the strengths. The discussion focused primarily on indicators that would in themselves indicate a quality problem.

Participants discussed two types of measures: those that evaluate the long-term care system globally and those that might be used to evaluate specific services or settings. A few highlights from the discussion are summarized below.

Global Measures.

  • Participants suggested considering measures such as the percent of persons who are institutionalized. They considered viewing it as a possible measure because of the belief that persons prefer to be in the least restricted setting; therefore, remaining in the community is a goal in itself. Following this logic, evidence of a reduction in the rate of nursing home use among those with long-term care needs may be seen at a system level as an improvement in the quality of care provided.
  • A second measure was unmet functional need. The argument is that one of the major goals of the long-term care system is to meet functional needs. In all but rare cases, unmet need would suggest a failure of the underlying care system. Unmet need in IADLs and ADLs can be constructed from some data, such as the National Long-Term Care Survey. Questions were raised about the psychometric properties of these measures, and it was noted that more development work is necessary. An alternative to measuring unmet need in ADLs and IADLs is measuring greater than expected declines in ADLs and IADLs. Expected declines would need to be adjusted for important health risk factors.
  • Participants suggested another global measure—the percent of informal caregivers expressing high levels of stress or burden. They did not recommend preferred measures of stress or burden.
  • Participants discussed the limitations of global measures. Global measures, by their nature, are not closely linked to specific services, resulting in potential attribution problems. Changes in these measures may not reflect changes in the quality of care in the system; rather, they may reflect other factors such as changes in health behavior or changes in the level of disability of the population. For example, if the long-term care population continues to become more disabled, the system quality may unfairly appear to be declining because we would likely observe increases in the percentage of persons in nursing homes and more unmet needs among the long-term care population. One approach to minimizing this effect is to adjust expected outcomes for changes in the disability level of the population.

Return to Contents

Service-Specific Measures.

  • Participants did not recommend any specific measures of satisfaction, although there was a discussion about the need to narrowly define the services received. It was reported that the Administration on Aging is developing satisfaction measures to be used by State Agencies on Aging. Instruments include measures designed by Scott Geron at Boston University. CMS and AHRQ, as part of the nursing home Consumer Assessment of Health Plans (CAHPS®) project, are attempting to develop satisfaction measures; but this project is limited to nursing homes.
  • Beyond OASIS, participants did not establish consensus on specific outcome, process, or structural measures that go beyond the detail of Table 2. They emphasized several times that because OASIS is designed for home health care, it is very clinical and does not capture many of the social and quality-of-life dimensions of home care and other community-based care.
  • Participants emphasized that more investment in needed in the development of measures of quality for community-based care. It was felt that AHRQ could play an important role in stimulating research in this area.

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