Click for next page ( 127


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 126
6 Quality of Care National data systems have been a largely underdeveloped source of information for measuring quality of care among the elderly. There are compelling reasons for identifying needs for this type of informa- tion and the extent to which such systems could meet them. The Office of Technology Assessment (1985a:77) points out that "Medi- care's prospective payment system (PPS) has intensified concern with the complex relationship between cost and quality of medical care" and that "assessing PPS impacts on quality of care is critical." Among the reasons given is that, if PPS succeeds as a cost contain- ment strategy, "its effect on the quality of care will be a deciding factor in the program's continual survival." Another report (Insti- tute of Medicine, 1986) focuses on the importance of information systems for assessing and improving the quality of care of nursing home residents and recommends (p. 43) that the secretary of the Department of Health and Human Services "undertake a study to design a system of acquiring and using resident assessment data." These reports reflect the emergence of quality of care as a matter of central interest for policies directed at controlling costs or improv- ing services for specific subgroups of the elderly. The larger context consists of an agenda of policies concerned with the application of prospective payment systems to sectors other than hospital care, modifications in Medicare and Medicaid benefits, and availability of staff, facilities, anct alternative sources of care, particularly for the Tong-term care patient. Policies that have been adopted, or are being contemplated, may 126

OCR for page 126
QUALITY OF CARE . . 127 have both short-term and long-term effects on quality of care, but the character of these effects is not clear. The potential for differential effects exists because of the diversity of health care systems; the increasing competition among providers; and the heterogeneity of the elderly population. The elderly vary in their health and mental health care needs, social and economic status, and living arrangements- circumstances subject to rapid change as age advances (Shanas and Maddox, 1985~. The utility of information on quality of care for national policy purposes is directly related to the ability to examine relationships involving these variables. Selected policy questions relating to the quality of care, listed in order from issues that require research or evaluations to questions that depend on or are related to measurement of quality of care include: Should quality assurance mechanisms for health care services of the elderly be strengthened because of the adverse effects of cost containment? Should Professional Review Organizations (PROs) under Medicare be strengthened to ensure quality of care for Medi- care beneficiaries? . To what extent should government and private insurance programs regulate quality of care under their reimbursement mechanisms? How has the change in Medicare's payment system to DRGs affected the quality of care for elderly patients? Has the emergence of competitive health maintenance or- ganizations (HMOs) provided economic incentives for their elderly enrollees or affected the quality of care provided to them? Who should be responsible for monitoring quality of care in HMOs? How can we develop, refine, and implement improved and standardized measurements of quality of care for older peo- ple? Who should be responsible? The discussion that follows presents the framework within which quality of care is most often examined, identifies several measures of quality of care suitable for national data systems, and considers steps needed to ensure availability of such measures. A FRAMEWORK FOR QUALITY OF CARE ASSESSMENT It is now possible to talk about a generation of effort to progress

OCR for page 126
128 AGING POPULATION IN THE TWENTY-FIRST CENTURY from broad definitions of quality of care to the identification of crite- ria that can be applied to measure quality. Although specific details of quality of care assessment have varied, the framework that is most widely accepted consists of three components-structure, process, and outcome (Donabedian, 1980~. Pr~rnary emphasis in explicating these concepts has been directed at the diagnosis and management of illness and is focused on transactions between providers and patients to achieve an appropriate balance of health benefits and risks. This is clearly an essential aspect of quality of care and is ex- emplified by approaches taken by Professional Review Organizations in developing programs for quality assurance of inpatient care for Medicare beneficiaries (Mohr, 1985~. Other aspects of quality of care that require greater prominence and that are readily accommodated within the categories, structure, process, and outcome, are oriented toward prevention and a consideration of how a system of care is functioning, not only with respect to those who appear for care at the time they do so, but also in terms of the need that exists in a community or general population. Quality is affected by both the technical performance of providers and the extent to which a system reaches those in need, for example, hypertensives under treatment may have high levels of control, but on a population basis there may be a high prevalence of uncontrolled hypertension. The usual formulation of what is included under each category is given below, expanded to cover some of these considerations. Structure refers to relatively stable parts of the health care sys- tem, such as numbers, types, and qualifications of professional per- sonnel, physical facilities, and medical technologies; it also includes organization ant] distribution of different types of services, proximity to the population being served, and other factors affecting access to care. Process refers to what is done to and for the patient or a de- fined population and how it is done; this includes the application of preventive, diagnostic, and therapeutic procedures, nursing and home care, counseling, referral practices, and coordination of care, follow-up, and monitoring procedures. Outcome refers to changes in health status or maintenance of desirable levels of health influenced by health services; this includes one or more dimensions of physical and psychosocial functioning, dis- ability, prevention of disease or complication of disease, satisfaction with health care, and mortality.

OCR for page 126
QUALITY OF CARE 129 A comprehensive assessment of quality of care would be con- cerned with all three components and their interrelationship, with the ultimate measure of quality determined by the outcome of care (Kessner et al., 1973~. In practice, there are often severe constraints in linking particular structural factors and processes of care to spe- cific measures of health status. Standards for many elements of structure are based on professional judgments and are used for such purposes as accreditation, licensing, Medicare certifications, and es- tablishing qualifications for staff in, for example, hospitals, group practices, and nursing homes. However, standards do not always represent a consensus; they are subject to change as new knowledge is acquired and new technology developed, and different structural patterns may achieve similar levels of quality. Furthermore, the fact that standards are met does not provide assurance that a favorable effect on health status will follow. In short, structural factors such as the availability of highly qualified primary care providers, special- ists and support services, or regionalized emergency medical service systems may increase the probability of receiving high-quality care, but they are not sufficient conditions for determining quality. Process of care measures provide more generally accepted indi- cators of quality of care, reflecting as they do what care is delivered and how. The measures are derived mainly from professional norms of practice or research in which some link to outcome is unplicitly or explicitly expected, whether this is improved functioning, delay in dependency, or relief from pain or depression. The degree of cer- tainty about a link may be weak in some cases because of lack of information about the natural history of disease, the relative effec- tiveness of alternative treatment modalities, or the role of behavioral and biological characteristics of patients in determining the course of illness. The selection of process measures for assessing quality of care seeks to minimize these limitations. Measures are available or being developed that range from adherence to algorithms for the treatment of specific conditions (Greenfield et al., 1981) to commu- nication and counseling (Inui and Carter, 1985), detection of need for health services (Aday and Andersen, 1975; Shapiro et al., 1985), matching need with appropriate sources of care in acute episodes of illness and in Tong-term care, for which coordination of services is important (Mathematica Policy Research, Inc., 1986~. The pessimism that long pervaded the outlook for having out- come measures to assess quality of care has been replaced by a more optimistic view. Although there are many uncertainties about how

OCR for page 126
130 AGING POPULATION IN THE TWENTY-FIRST CENTURY health status is influenced by personal and social factors, knowledge has cumulated regarding the effectiveness of specific health care in- terventions. Advances have been made in the development, testing, and application of health status measures to supplement indices of mortality that are useful in examining both technical and health care system effects on the health of a population (Bergner, 1985; Brook and Lohr, 1985~. This holds true for Al age groups, but it is par- ticularly relevant for the elderly, whose neecis for health services are high and large proportions of whom could be affected positively or negatively. We specify outcome measures under the three functional goals of health care below: Caring involves reassurance and relief from anxiety, satisfaction, and communication between provider and patient, which may affect whether regimens prescribed are followed and response to symptoms and care-seeking are appropriate. Curing an] maintenance involve a reduction in symptoms and dependency, avoidance of complications and iatrogenic effects of treatment and reversing the effects when they occur, improvement or stabilization in functional capacity, and reduction in avoidable mortality. Many specific conditions are candidates for observation, such as hypertension, angina, hip fractures, major depression, glau- coma, cataracts, respiratory diseases, and hearing, visual, and dental defects. Functional status measures detailed in Chapter 3 represent a spectrum of physical, emotional, and social indicators of health sta- tus. They reflect the results of an interaction between population and environmental conditions and processes of care determined by how the delivering system and individual practitioners function. This type of interaction affects aspects of health care such as matching need with demand as affected by availability and.access to care, timeliness of medical intervention, planning of long-term care and post-hospital care, coordination of multiple sources of care, and con- tinuity of care. Prevention measures are designed to prevent, delay, or reverse the progression of disease, and reduce functional lirrutation, handi- cap, or dependency due to health conditions. A preventive measure may be interpreted as a surrogate or intermediate outcome measure when its effect on a health condition or health status has been estab- lished, when it reaches the population that is at risk and can benefit from it, and when there is appropriate follow-up.

OCR for page 126
QUALITY OF CARE 131 The potential and limitations of what can be gained from specific measures are discussed in Chapter 5. A full treatment of prevention would consider degrees of certainty about efficacy, cost-effectiveness of particular prevention measures, state of readiness among providers to apply them, and method and frequency of the interventions. How- ever, the application of prevention procedures that are generally viewed as desirable would serve as an indicator of quality of care. These inclucle examination and follow-up of asymptomatic individ- uals for such conditions as hypertension, visual defects, impaired hearing, hypercholesteremia and counseling on smoking cessation, exercise, rest, diet, and rehabilitation aimed at restoring function consistent with the maximum achievable level. THE ROLE OF NATIONAL INFO1IMATION SYSTEMS National data systems have the capacity to produce information needed to derive quality of care measures to determine how quality varies within the population and how it may be influenced by sources of care and policies that affect services received. Many of the specific items required are included in surveys that are already operational or about to start; others are dependent on the extension of existing items or the introduction of new ones. Additional efforts will be needed to develop and test new mea- sures of outcome and processes of care, as well as methods for making reliable and valid observations on the elderly who are well and those who are infirm or cognitively impaired. In this regard, particular attention needs to be given to methodology and research pertaining to cognitive aspects of survey design, i.e., aspects of survey research interviews pertaining to cognitive processes such as understanding language, remembering and forgetting, perception, judgment, and inferring causes. One direction for such activity is the specification of better questions as well as better ways of expressing questions with respect to cognitive considerations (Fienberg et al., 1985; I,essler and Sirken, 1985~. Another is the use of flagging procedures by interview- ers to indicate their degree of belief in the correctness of reported responses. Such information could then play a helpful role in the evaluation of the sensitivity of estimates to possibly varying levels of data accuracy. Methodology for expressing the results of such efforts in terms of bias or accuracy also needs better development. A significant expansion in the measurement of quality of care can be achieved before results of new research are available through

OCR for page 126
132 AGING POPULATION IN THE TWENTY-FIRST CENTURY the use of currently available or readily adaptable measures, linkage of multiple data sets, and longitudinal information systems. Imple- mentation of the recommendations in earlier sections of this report on health status and health promotion, as well as those in later sec- tions on utilization and Tong-term care, would achieve this purpose. The additional requirement is a coherent strategy for placing the measures in a quality of care framework. In developing this strategy, it is important to bear in mind that national information systems should have the capacity to measure quality directly. They should also provide broad indicators of quality of care that can signal that other sources of information are needed (Rutstein et al., 1976~. For example, an alert was signaled in 1986 when the Health Care Financing Administration (HCFA) released hospital mortality rates derived from the Medicare hospital bills. The result was to focus attention on how patients are managed in the hospital. Another effect should be to intensify the development of severity of illness measures that are feasible for inclusion in data systems on hospital discharges (Horn and Horn, 1986~. Indicators of quality of care would become more readily available with the adoption of the Long-Term Health Care Minimum Data Set (U.S. Department of Health and Human Services, 1980b), and this would provide a basis for probing more fully into processes of care in institutional and noninstitutional settings. Also relevant for the development of quality of care measures are the recommendations of the Institute of Medicine's Committee on Improving Nursing Home Care for a study to determine the scope and design of an information system useful for regulating nursing homes effectively and considering policies for long-term care (Institute of Medicine, 1986~. Quality of care represents a major interest; the results will also be valuable for other data systems. A concerted effort is needed to develop quality of care measures on the elderly suitable for national information systems. A task force should be established to serve as the locus for this effort. The task force could be the responsibility of the National Committee on Vital and Health Statistics, the Health Information Policy Council in the Department of Health and Human Services, or the Forum on Aging- Related Statistics, the coordinating council of principal government agencies concerned with the provision and use of information on ag- ing. Special expertise and experience will be required, and the task force activity will need to be paralleled by funding of methodological studies through research grants and contracts. The task force should

OCR for page 126
QUALITY OF CARE 133 be concerned with assessments of structure, process, and outcome of care for subgroups of the population, with a special emphasis on outcomes of care influenced by delivery system and provider char- acteristics. Consideration should be given to measures and methods that have the capability of identifying rapidly possible effects of health policy decisions. Items contained in several of the national health surveys can be used as building blocks for developing quality of care measures. Candidates are found in the NHIS and its supplement on aging and prevention, the NNHS, the NHANES, and the NAMCS, all of which are NCHS surveys; in the Long-Term Care Survey (HCFA and the Office of the Assistant Secretary for Health, DHHS), and in the forthcoming NMES (NCHSR and HCFA). These items should be reexamined for modification, amplification, and replacement based on results of analysis. Recommendation 6.1: The pane! recommends that agen- cies having cognizance of national data systems, whether based on adrn~nistrative records or survey data: (1) jointly review these data systems to identify items that can be used in developing quality of care measures and identify missing items that are needed to produce such measures and (2) make provision for collecting the data needed to fill gaps and for producing quality of care measures. Priority should also be given to development of new items on func- tional status (as recommended in Chapter 3) and their adoption once they have been tested. The Long-Term Health Care Minimum Data Set (discussed in Chapter 7), should be incorporated into relevant data systems. Appendix D lists the items contained in this minimum data set. Indicators of quality of care can also be developed by linking data obtained in surveys to other files, such as the National Death Index, or by reinterviews to obtain information on a completed episode of illness. Recommendation 6.2: The pane] recommends that multi- pie data sources be linked to provide indicators of quality of care during episodes of illness and that continuous, Ton- gitudinal data be developed to track the flow of individuals through episodes of illness to recovery, dependency, or death, including their patterns of use of acute ant] Tong-term care facilities and community and home care services.

OCR for page 126
134 AGING POPULATION IN THE TWENTY-FIRST CENTURY Continued support should be ensured for steps already taken, such as the conduct of the National Mortality Followback Survey; provision for searches of the National Death Index over long periods for deaths among people identified in sample surveys of the NCHS; and follow- up of individuals in the NHANES, the NHIS, and the Long-Term Care Survey; HCFA's Medicare Automatic Data Retrieval System should be completed and continued annually; quality of care assess- ments of hospital prospective payment systems need to be concerned with pre- and post-hospital care; and urgent attention should be given to the introduction of severity of illness measures. The pane! recognizes that most elderly are capable of responding for themselves when interviewed, but it is uncertain whether issues affecting the quality of information provided by them are similar or different from those at younger age groups e Interviewers encounter additional data quality problems in surveying the elderly because, with advancing age and for many in nursing homes, the number of cognitively impaired people increases and the interviewer is depen- dent on proxy respondents. The validity of data from proxies may be affected by many factors not well understood at this time. Re- search on these issues will need to address not only the magnitude and characteristics of the problems involved, but also the methods and sources for improving the quality of the data. Recommendation 6.3: The panel recommends that the Na- tional Center for Health Statistics develop methods for in- creasing the validity and reliability of health data (health status, needs, and care received) collected in surveys of the elderly, whether from elderly respondents themselves, from their caregivers or next-of-kin, or from both.