Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
7 Issues Requiring Further Study There are five sets of issues that need study before federal policy positions on these issues can be developed and prescribed: (1) the scope and design of information systems needed to regulate nursing homes effectively and to facilitate development of sound policies for long-term care; (2) policies governing the methods and amounts of payments to nursing homes for care of residents eligible for support under the Medicaid program; (3) policies affecting the supply of nursing home beds in the context of the growing demand for all types of long-term-care services; (4) regulatory policies concerning (a) the training and qualifications of all staff in nursing homes and (b) minimum staffing patterns needed to provide adequate care to mixes of residents with varying needs; and (5) policies governing construction of new nursing homes, specifically, the proportion of single rooms that should be required. 190
ISSUES REQUIRING FURTHER STUDY /191 INFORMATION SYSTEMS Data About Residents The introduction of the requirement (recommended in Chapter 3) for standard assessment data on every resident will produce a vast body of data about the characteristics of nursing home residents and how they change while in nursing homes. These data have potentially major significance for three purposes: (1) for improving nursing home management, (2) for improving the effect- iveness of regulation, and (3) for obtaining essential information with which to develop more effective and efficient nursing home regulatory policies, and for facilitating development of more appropriate long-term · · . care policies. It is a large undertaking to install a national standard resident assessment system in 15,000 nursing homes that has the capability of allowing needed information to be retrieved readily. It involves, among other things, determining the standard data to be collected and designing and testing techniques for collecting it reliably, developing instruction manuals, and training thousands of people to conduct the assessment routinely and with reasonable integrity and reliability. It also involves developing case-mix groupings based on definitions related primarily to assessment scores, and developing auditing procedures and the standards to be used by state auditors to determine whether the error rates they find are acceptable. With good planning, adequate resources, and strong, competent leadership, this set of tasks could be accomplished in 2 or 3 years. Complex technical and policy decisions are involved in designing a sound system for gaining access to these data by computer. The decisions will require careful study and will take time. Introducing a manual resident assessment system should not be delayed until this study is completed. A great many nursing homes now have their own computers. Some--perhaps many--are likely to enter resident assessment data into their own computer files so that they can use it for their own management purposes."
192 / NURSING [IO3fE CARE But a great deal of work would have to be done to determine how best to use these data to meet both state and federal government (as well as resident care) needs. Some questions occur, such as whether all data on all residents should be acquired by the state regularly or whether the data should be sampled, and, if so, how, and how often. These questions can be answered largely by determining the priority uses to which the data are to be put--for example, for developing case-mix-controlled outcome standards for quality assurance purposes, for use in Medicaid payment decisions, for developing staffing and other resource algorithms tied to case mix, or for utilization review. There also are questions of cost, technical feasibility, privacy, authorized access to and uses of the data, and a number of other technically and legally complex and politically sensitive matters. These questions can and should be resolved. The rapid advances and decreasing costs of computer technology make a computerized system for handling resident assessment data feasible from technical and economic standpoints. A study should be commissioned by the Department of Health and Human Services to design the system. Responsibility for conducting the study should be assigned to a group of technically competent and broadly knowledgeable people who are sensitive to the concerns and needs of all interested parties--the residents, the nursing home operators, state governments, and the federal government. Such a study will have implications for the future role and contents of the National Nursing Home Survey conducted by the National Center for Health Statistics. This survey has been the most important source of information on nursing home residents and care resources. However, its utility has been limited by its small sample size, long intervals between surveys (almost 10 years since the last completed survey), the modest amount of data on the health and functional status of residents, and absence of longitudinal data. The recommended study could lead to a new strategy that would resolve these problems. In this process, consideration needs to be given to relevant recommendations of the National Committee on Vital and Health Statistics for a minimum data set on long-term care.
ISSUES REQUIRING FURTHER STUDY /193 Other Data About Nursing Homes As mentioned in Chapter 1, the committee was impressed by the paucity of information about nursing homes and their operations, as well as about regulatory activities available on both national and state levels. With more than half of all nursing home revenues coming from public funds, and with growing demand for nursing homes and other types of long-term-care services, the need for more infor- mation seems clear. But moving from that general conclu- sion to specific decisions on what information should be collected, how frequently, how it should be done, how it should be aggregated, analyzed, and made publicly available, and who should be responsible, is quite another matter. A study by a technically competent and broadly knowledgeable group--possibly the same group that is responsible for studying the resident assessment data system--should be asked to study the requirements and make recommendations on how they should be handled. Recommendation 7-1: The Secretary of HAS should order a study to design a system for acquiring and using resident assessment data to meet the legitimate and continuing needs of state and federal government agencies. The Secretary also should order a study to determine the needs for other data about nursing homes that would facilitate regulation and policy development. This study should recor''nzend specific ways to collect, analyze, and publish or otherwise make such data publicly available. MEDICAID PAYMENT POLICIES The Medicaid program was originally designed to pay for health care services for those on welfare and selected others whose incomes were low and who were"medically needy" because they had no health insurance. Medicaid was--and is--perceived at the state and federal levels as a component of the welfare system. As is true of the other components of the welfare system, the states are responsible for administering it under broad federal guidelines. This means that each state determines who
194 / NURSING HO3lE CARE shall be eligible for Medicaid, what services it will pay for, and how (and how much) it will pay the health professionals and the institutions who provide authorized services to eligible patients. Medicaid funds are appropriated annually (or biennially) by state legisla- tures as are the matching federal funds. The federal contribution to state Medicaid budgets ranges from 50 to 78 percent. In most states, Medicaid is the second largest budget item after education and in recent years has been the fastest growing.2 About 50 percent of nursing home revenues come from Medicaid. The funds pay for some or all of the costs of about two-thirds of the residents.3 In 1984, Medicaid expenditures for nursing home care totaled about $14 billion.4 Medicaid payment policies--both the methods used to calculate how much to pay, and the actual rates of payment--provide strong incentives to nursing home operators. (Eighty percent of the beds are operated on a for-profit basis.) Nursing home operators adjust their operations so that the revenues they receive cover all of their costs (including capital costs) plus a profit. Nursing homes can control costs by controlling admissions (choosing a mix of residents whose needs for care can be paid for by the revenues they bring in), and by controlling such variable operating expenses as staffing, food, laundry, housekeeping, and plant maintenance. Because Medicaid rates are as much as 30 percent lower than private rates for comparable residents in some states, there is a clear incentive to try to attract and keep as many private-pay residents as possible. At least six goals have been suggested (or implied) as appropriate for state Medicaid payment policy. It should 1. control public expenditures for Medicaid: 2. ensure adequate provider participation and access to care by those eligil~le--or likely to become eligible--for Medicaid, irrespective of degree of disability; 3. encourage appropriate and high-quality care; 4. deliver service efficiently (provide the maximum appropriate service per dollar); 5. be administratively simple to implement; and 6. minimize the potential for fraud and abuse.
ISSUES REQUIRING FURTHER STUDY /195 The committee commissioned a review of the research literature to ascertain what is known about the effects of different Medicaid payment policies both on access and quality of care in nursing homes.5 The findings suggest that the relationship between quality and payment policy is highly variable and somewhat unpredictable. Some facilities provide excellent care at the same payment rate, and with the same resident mix, as other facilities in the same geographic area that provide substandard care. Some rates or payment levels may be insufficient to provide desirable quality of care and quality of life, but the distribution of the payment into cost line items within a facility may have a greater impact on quality than the amount of the total payment. Furthermore, such aspects of facility performance as the quality, motivation, and efficiency of the care-giving staff, and managerial skill, are not price-sensitive. These performance characteristics vary greatly across facilities. Since the relationships among costs, charges, and quality of care are very complex, simply paying more is no guarantee of improved quality. In some cases, paying less (up to a point) need not lower quality. No studies, however, have adequately investigated the complex relationships among costs, charges, reimbursement, and quality. The fact that the current literature does not show strong relationships among cost, charges, payment policy, and quality does not imply that there are no relationships. Available ways to measure are neither valid nor reliable. Most studies have measured structure or process because they were the only known proxies for quality. It would be desirable to know how the quality of life (as perceived by the resident) and the quality of the care (as determined by case-mix-controlled outcomes compared to national norms) are related to costs. In sum, there is now no evidence to establish the superiority of any Medicaid payment policy with respect to its effects on quality of care. There has been little systematic evaluation of the impact of different systems, in part because many have been implemented recently. Ideally, if a particular approach to Medicaid reimbursement policy proves to be more successful than
196 / NURSING HO3lE CARE others in having the desired effects on nursing home behavior, it should be adopted by all states. Thus, it is an important area for further federal policy development. But much additional research is needed and it merits continued high-priority attention by the HCFA. Research and demonstrations to test innovative payment systems should be encouraged. They should include, for example, all-payer rate setting, to assess its effects on quality of care and access for heavy care and Medicaid residents. The research opportunities in these questions will be enhanced considerably once the standardized resident assessment data system is in place. It will then become possible to control for case mix and apply outcome measures of quality. DEMAND FOR AND SUPPLY OF NURSING HOME BEDS Demand In most states6 there is evidence of excess demand for nursing home beds: occupancy rates are well over 90 percent and there are waiting lists at many facilities. The demographic trends suggest that the demand for the kinds of long-term care services now being provided mainly by nursing homes is certain to increase. The number of persons over 65 is projected to increase from 25.7 million in 1980 to 36.3 million in 2000, a 41.2 percent increase. For the over-85 group, the projected increase is 108 percent during this period, from 2.6 to 5.4 million.7 The rapid growth of the population aged 85 and over is likely to have a significant impact on the size and structure of the nursing home population.8 If current age and sex-specific institutionalization rates hold, the proportion of the residents in nursing homes who are age 85 and over can be expected to rise from 31 percent in 1980 to 43 percent in 2000. This increase in the mean age of the nursing home population implies a greater proportion of heavy-care residents. Two additional factors may affect demand for nursing home beds. One is the rate and direction of change in health status at advanced ages. That is, in addition to
ISSUES REQUIRING FURTHER STUDY /197 the effects of health status on survival, changes in health status may directly affect the risk of institutionalization for the elderly. For example, with increasing numbers of persons over 85 there may be an increase in the prevalence of chronic diseases if the increases in survival are concentrated among chronically morbid and impaired people. This would increase the demand for intensive nursing home care. (Alternately, it may be that the future survivors will be proportionately healthier than those in the same age group today, and that projecting current rates will overestimate future morbidity rates. But there is no evidence to suggest this is happening.) Another factor affecting future demand will be increases in the availability of alternative long-term care services, perhaps stimulated by the emergence of alter- native financing patterns, for example, the growth of private, long-term-care insurance. Irrespective of the nature and extent of such developments, the need for nursing home care will not diminish during the next 15 years. There is no evidence that either the population's health status (physical, functional, and mental) will so improve that nursing home care requirements will decrease, or that other long-term- care services could be substituted for nursing home care for the majority of individuals now found in nursing homes. The population in nursing homes is likely to be more aged and more disabled, and some form of mental disability (particularly Alzheimer's disease) is likely to be more common. But the current pattern of a mix of residents with different treatment and service needs (that is, a fairly heterogeneous population) is likely to continue. Bed Supply In most states there are more people seeking admission to nursing homes than there are beds available. This excess demand results from three interrelated factors: (1) There are many individuals now living in the community WhO are just as disabled as nursing home residents, and some of them would enter a nursing home if a bed were
198 / NURSING HOME CARE available; (2) to control the costs of the Medicaid program, some states have sharply constrained or completely stopped construction of new nursing homes or expansions of existing nursing homes; (3) in some states Medicaid reimbursement rates are so low that at current rates, the marginal cost of treating some (heavy-care) residents may exceed the reimbursement rate. The supplier's market for nursing home beds that exists in most states allows nursing home operators to select among applicants for beds. Business logic suggests that they will try to optimize net income by favoring private-pay over Medicaid-eligible and, generally, the easier-to-care-for resident over those who are more difficult (and, therefore, more costly) to care for properly. However, the latter judgment is affected by payment policy: if payment for heavier-care residents is more than sufficient to offset higher costs so that it is more profitable to admit them than lighter-care residents, and if adequate staffing to care for such residents is feasible, heavier-care residents may be given preference for admission. But there are still likely to be some types of residents who will be hard to place. A recent study was done on patients in Massachusetts hospitals who were "backed-up", that is, they were in hospital beds for "administrative necessity."8 Although they no longer needed hospital care, they could not be discharged to their homes because they needed nursing home care. The hospitals could not find nursing homes willing to admit them, so they were allowed to remain in more costly hospital beds pending availability of nursing home beds. The study found that the backup population consisted of two groups of patients: one group spent a short time in the queue before being admitted to nursing homes; the second group spent a long time in the queue. The second group of patients was sicker and often had severe mental problems. The findings of this study support the judgment that, other things being equal, nursing homes will select the easier-to-care-for patients from the queue. Another important factor about which there is currently insufficient information is the influence of hospital prospective payment systems and other cost-containment
ISSUES REQUIRING FURTHER STUDY /199 measures on demand for, and availability of, nursing home beds. A special effort is needed to study these relation- ships. Regardless of the payment system, however, the supplier's market also places heavier demands on the regulatory system to assure quality because the market pressure to maintain high quality to attract residents does not exist. And the business logic that tempts nursing home administrators to cut operating costs to minimum levels or even below is hard for many to resist. The question of what constitutes the best policy with respect to bed supply has no simple answer. The variation among states in the number of nursing home beds per thousand is very large: the national average is about 1 bed per 20 persons aged 65 or over. Minnesota has more than twice the national average; Florida has about half the national average. Thus, there is a fourfold vari- ation. But bed occupancy rates--and excess demand--appear to exist in most states.6 The uncharted policy areas that are related to bed supply are (1) alternative ways of financing long-term care--particularly the possibility of private insurance arrangements for financing long-term-care services that are not primarily health-related and are not limited to payment for services provided in nursing homes; and (2) the development and greater availability in most communities of a much larger number of alternative long-term-care arrangements such as home health care, homemaker services, congregate housing (including domiciliary care), meals-on-wheels, special transportation facilities, adult day-care, and respite care. If good alternatives to nursing homes were readily available ant! could be paid for from third-party insurance, a fraction--the exact number is not now known--of residents could be cared for more appropriately in alternative service arrangements or facilities. If this were to occur, then a certain number of nursing home beds now occupied by residents requiring lesser levels of care would become available to help cope with the growing numbers of older heavy-care residents who must be in nursing homes. The development of alternative long-term
200 / NURSING HOME CARE care arrangements on a large scale is not likely to occur until alternative financing arrangements become available. Exploration of some alternative care arrangements--on a small scale--is proceeding. The HCFA has sponsored some innovative long-term-care demonstration projects during the past 10 years. More recently, under statutory authority contained in the 1981 budget legislation, states have been granted waivers to permit them to use Medicaid funds to finance services in community-based, long-term projects designed to prevent unnecessary institutional- ization for individuals who otherwise could receive Medicaid support only in nursing homes.9 Systematic evaluation of these programs has just begun. In sum, the policy issue concerning supply of nursing home beds is related to the broader policy issue of developing a more appropriate array of long-term-care services. This, in turn, hinges on the development of more appropriate private and public financing arrangements and policies. A systematic study of these issues is necessary to design sound public policies to facilitate development of both the new financing mechanisms and of the array of long-term-care services needed. This study should be viewed as a matter of high priority by both the Congress and the executive branch of the federal government. STAFFING OF NURSING HOMES Once a data base derived from systematic, periodic resident assessments becomes available, two kinds of staffing studies will become possible that have not yet been done satisfactorily. The first will be studies to develop an algorithm for relating minimum nursing staff requirements to case mix. Perhaps something analogous to the "management minutes" concepti° could be devel- oped. (Management minutes is an empirically derived algorithm used to estimate the daily nursing time requirements for a resident based on his/her assessment scores and service needs.) This is a complex study that will require considerable sophistication in study design and execution to produce valid and reliable results. If
ISSUES REQUIRING FURTHER STUDY /201 completed successfully, it could provide the basis for a regulatory tool of considerable power. The second kind of study deals with the issue of staff qualifications and training. At present, only profes- sional judgment is used to define the requirements. But with outcome measures that can be derived from resident assessment data, studies to compare the effectiveness of different staffing patterns, types of staff, and training requirements will be possible. The HCFA should support well-designed studies within this area. If convincing evidence becomes available that some approaches to staffing and training are distinctly superior (in quality of care/life and cost) to others, the HCFA will be in a position to incorporate the desirable approaches into its regulatory standards. SINGLE- VERSUS MULTIPLE-OCCUPANCY ROOMS Most nursing homes have been constructed with either all or most of their rooms designed for double occupancy. Because beds in most places are in short supply, residents seldom can choose either private rooms or their roommates. The nursing home population is heterogeneous, so this is a thorny issue. It is clear that quality of life for an undemented resident can be seriously affected by the functional, mental, and behavioral status and service needs of a roommate. Moreover, the issues of privacy and of choice--for example, whether or not to watch TV or listen to music, and which programs--loom very high in the quality-of-life values of most residents. Most mentally alert residents probably would prefer private rooms if they could! have one. The question is: Should the HCFA require that all new construction, or additions to existing nursing homes, be required to have a specified fraction of private rooms? If so, what should that fraction be? Not enough is known to answer this question with confidence. The effects on construction costs or on operating costs of requiring a specific proportion of single rooms are not known. Moreover, not enough is now known either about the preferences of residents for private rooms and of the desirability of
202 / NURSING HOME CARE having certain residents--particularly some who are demented--share a room, and whether four-bed rooms might be better than two-bed rooms. In the next 10 to 20 years there may be a substantial amount of new construction or major remodeling of nursing homes. The committee believes that the HCFA should commission a study of this issue to determine the proper balance between single- and multiple-occupancy rooms that should be required in newly constructed nursing homes and in additions to--or major remodelings of--existing homes. Recommendation 7-2: The NCFA should commission a study of the costs anc! benefits of single-occupancy rooms compared to multiple-occupancy rooms in nursing hones. The study should be designed to obtain data about the effects of single rooms on the quality of life of various types of nursing home resid ents. The stud y should be completed within 2 years after it has been authorized. It shorted contain recon~n~endations for the desired proportions of single- and n'~cltiple-occupancy rooms in nursing hones. It should recon~n~end required proportions in future new construction and ma jor reflood cling of existing build ings.