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 45
Medicare: A Strategy for Quality Assurance - Volume I 2 Concepts of Assessing, Assuring, and Improving Quality In this chapter we describe quality assurance concepts and models as a context in which to understand where our proposed quality assurance program for Medicare fits in the long tradition of quality assurance in this country. We attempt to answer these questions: What do quality assessment, quality assurance, and quality improvement mean? What are the roles of structure, process, and outcome in these concepts? What are the key properties of a quality assurance program? The final section of this chapter examines two quality assurance conceptual models, that is, the traditional structure-process-outcome model and the continuous quality improvement model. During this study the health care industry became increasingly interested in the potential application of the continuous improvement model for health care. The committee took advantage of numerous opportunities to explore its concepts and practices, and it thoroughly debated how this relatively new model might be incorporated into a strategy for quality assurance in Medicare. QUALITY ASSESSMENT, QUALITY ASSURANCE, AND QUALITY IMPROVEMENT Definitions The concepts of quality assessment and assurance in the health sector are not new; the literature documents efforts over the last 80 years or so to place them into operational frameworks.1 Quality assessment is the measurement of the technical and interpersonal aspects of health care and the outcomes of that care. Assessment is expressly a measurement activity;
OCR for page 46
Medicare: A Strategy for Quality Assurance - Volume I although it is the first step in quality assurance, it does not imply a solution to problems that may be uncovered. Classically, quality assurance encompasses a full cycle of activities and systems for maintaining the quality of patient care. One definition has it as “a formal and systematic exercise in identifying problems in medical care delivery, designing activities to overcome the problems, and carrying out follow-up monitoring to ensure that no new problems have been introduced and that corrective steps have been effective” (Lohr and Brook, 1984, p. 585). Generally, that cycle involves a set of steps proceeding from identification and verification of quality-related problems and their causes to the implementation of solutions to the problems with the specific intent that the solution be long-lasting or preventive; these activities are followed by a timely review to determine if the problem has been solved and no new ones generated in the process. If the last two conditions are met (one problem solved and no new ones generated), attention turns away from that aspect of patient care to other areas or topics. Quality improvement is a set of techniques for continuous study and improvement of the processes of delivering health care services and products to meet the needs and expectations of the customers of those services and products. It has three basic elements: customer knowledge, a focus on processes of health care delivery, and statistical approaches that aim to reduce variations in those processes. In understanding the place of continuous quality improvement it is helpful to think in terms of a bell-shaped curve that distributes numbers of providers or volume of care against quality. The leading tail is the province of research, the lagging tail is the focus of regulation; and the middle is the focus of continuous improvement. The design of a quality assurance system (for Medicare or for any other health care program) should attend to all three parts of the distribution. Purposes of Quality Assurance Quality assurance neither promises nor guarantees error-free health care. Its ultimate goal is to build confidence and faith in the quality of the health care being rendered. Achieving error-free health care at all times is impossible; trying to do so ultimately discourages quality assurance efforts. The ultimate goal, however, is achievable and thus encourages continuous effort. An effective quality assurance program is not an end in itself; rather, it is a means of maintaining and improving health care (O’Leary, 1988). Quality assurance programs vary widely in their purposes, targets, and methods. Four major purposes can be identified.2 First, in some circumstances the main goal is to identify providers whose delivery of care is so far below an acceptable level that immediate actions are needed to ensure that they no longer deliver care or that responsible third-party payers, such
OCR for page 47
Medicare: A Strategy for Quality Assurance - Volume I as Medicare, no longer reimburse for any care delivered. Second, quality assurance programs, on identifying providers whose delivery practices are determined to be unacceptable, may concentrate on working with those providers to correct the problems and bring care up to an acceptable level. Both these examples reflect an orientation to “outliers,”3 although the remedies (and hence the underlying philosophies) differ. A third purpose focuses on improving the average level of quality of care delivered by a community of providers. Improving average performance, sometimes referred to as “shifting the curve,” by moving a large number of providers forward on the quality scale usually occurs only gradually. It implies, among other things, a considerable educational effort. From a statistical point of view however, one can shift the curve by removing outliers from the professional community when their abnormal practices are truly extreme and constitute a significant percentage of the community’s practices. Implicit in this process of shifting the curve is the understanding that acceptable levels of quality are protected from erosion. Fourth, quality assurance may also motivate and assist providers to achieve high levels of quality. Programs may identify excellent providers who serve as models or mentors. They may explicitly recognize and reward exemplary performance or underwrite incentives for practitioners and organizations to reach and surpass desirable levels of quality. This approach supplements direct efforts at improving average practice by highlighting and rewarding superior performance. Internal and External Programs This report often distinguishes “internal” and “external” quality assurance. Internal quality assurance programs are those implemented by organizations or systems, for example by hospitals, HMOs, home health agencies, or similar groups of practitioners. Internal programs have (or could be brought to have) several key features. They can be integrated into ongoing patient care and adapted to the local environment and the degree of sophistication and interest of the practitioners. They can emphasize professionalism and the desire on the part of most practitioners to do better. They can minimize the adversarial “we-they” attitude that can and does provoke mistrust of the outside reviewer (Weiss, 1972). To the extent that internal programs involve their entire staffs in quality-of-care matters, they can reinforce the concept of the “virtuous” organization. Theoretically, they can serve all four purposes of a quality assurance program outlined above, although they may find it difficult to identify or, more importantly, remove the true outlier provider. Further, they are not likely to identify systemic problems that exist within the organization. Internal programs face several problems. Conflict over authority, lack of
OCR for page 48
Medicare: A Strategy for Quality Assurance - Volume I commitment or expertise, and concern about financial repercussions for individuals or of financial stress in the organization may result in inaction despite well-known and well-documented problems. When individual organizations develop idiosyncratic methods, collect data without an external reference for comparison, or fail to follow through when internal data suggest problems, valuable resources and opportunities for improvement are wasted. The time-honored principle of peer review may be difficult to implement in a small organization. External quality assurance programs typically serve a broader social purpose and clientele. Examples include the accrediting activities of the Joint Commission on Accreditation of Healthcare Organizations and the National League for Nursing, the Medicare Utilization and Quality Control Peer Review Organization (PRO) program, and state medical licensing or disciplinary boards. Through mechanisms such as the threat of exposure, imposition of financial sanctions, or withdrawal of accreditation or licensure status, they may be able to deal with outliers (and especially to remove them). They may also be helpful by assessing the quality of the internal quality improvement process and, where appropriate, offering technical assistance. Often, however, external programs will be less well suited to improving the average performance of providers, and most are not in a position to identify exemplary providers or to offer assistance in reaching higher levels of quality care. Some external programs may have uniform data collection and reporting methods that allow comparisons across settings and institutions; information can be shared with organizations entitled to their own data. External programs can force attention to problems that would not otherwise be addressed, pressure providers to correct problems, and make it harder for incompetent practitioners to move from one facility or organization to another. For external programs, successful review is feasible only if standards enable accurate assessment of the variations in individual organizations. External review must protect those under review when the reviewers belong to antagonistic or competitive groups. Review of care rendered by physicians in prepaid group practices by those in fee-for-service practice is an example of potential conflict among practitioners who by all other accounts may be considered peers. Both internal and external quality assurance programs are necessary for a comprehensive approach to quality assurance; neither presents a sufficient response to quality problems. The distinctions may seem somewhat arbitrary because, at one level, everything outside the patient-practitioner pair is external to the interaction. Denoting a hospital or HMO program as internal and the Medicare or a state’s program as external, or agreeing on
OCR for page 49
Medicare: A Strategy for Quality Assurance - Volume I TABLE 2.1 Desirable Attributes of a Quality Assurance Program • Addresses overuse, underuse, and poor technical and interpersonal quality • Intrudes minimally into the patient-provider relationship • Is acceptable to professionals and providers • Fosters improvement throughout the health care organization and system • Deals with outlier practice and performance • Uses both positive and negative incentives for change and improvement in performance. • Provides practitioners and providers with timely information to improve performance • Has face validity for the public and for professionals (i.e., is understandable and relevant to patient and clinical decision making) • Is scientifically rigorous • Positive impact on patient outcomes can be demonstrated or inferred • Can address both individual and population-based outcomes • Documents improvement in quality and progress toward excellence • Is easily implemented and administered • Is affordable and is cost-effective • Includes patients and the public who can legitimately be considered a peer, is related more to almost incidental traditional professional boundaries of accountability than to intrinsic ones. One challenge in devising a strategy for quality assurance is to combine the strengths of both internal and external approaches yet avoid replication and counterproductive effects, such as the poisoning of the atmosphere for professional involvement in internal programs. CRITERIA FOR JUDGING AN EFFECTIVE QUALITY ASSURANCE PROGRAM What are the attributes of a successful quality assurance effort that would be acceptable to those with a stake in the process (i.e., patients, providers, payers, and policymakers)? This section outlines our view of the criteria that a successful quality assurance program, either internal or external, should strive to meet (see Table 2.1). Some of these criteria may appear contradictory (such as minimal intrusion into the patient-provider relationship and ability to deal effectively with outlier providers), but the mark of a good program is an appropriate balance between such elements. We use these attributes later in this chapter to evaluate two conceptual models of quality assurance. A successful quality assurance program has the following 15 attributes):
OCR for page 50
Medicare: A Strategy for Quality Assurance - Volume I It is able to address a full range of quality problems—poor technical quality, overuse, and underuse. Reviews of the literature and discussions with providers of care during the study site visits led us to understand that problems of all three kinds occur to varying degrees in different settings of care (e.g., underuse of home health services and overuse of hospital or outpatient surgical procedures) (Chapter 7). Hence, an effective quality assurance program should have a range of methods such that it can prevent, identify, and correct problems of underuse, overuse, and poor technical and interpersonal quality in all patient care settings and under various reimbursement mechanisms. Additionally, an effective quality assurance program should be flexible enough to prompt appropriate responses to new problems as settings, reimbursement mechanisms, and clinical practice change over time. It intrudes minimally into the patient-provider relationship. Because the core of health care is the patient-provider relationship, no quality assurance program should jeopardize the relationship of trust or the ability of the practitioner to use his or her best judgment to guide the care of the patient. Neither should it diminish the autonomy of patients in seeking and obtaining care that conforms to their preferences. It is acceptable to professionals and providers. The essence of quality assurance is improving the care provided by individuals to individuals, through better decision making, enhanced skills, more adequate support systems, and similar elements affecting health care. To accomplish these goals, a successful program must be accepted by the professionals and organizations in which it is embedded or to which it is directed. The program must support, and be seen as supporting, the goal of the well-motivated health professional to provide compassionate and competent care. This implies that judgments about care and recommendations about change in practice are made by peers. It fosters improvement throughout the health care organization and system. A quality assurance program is incomplete without a focus on improving the processes through which patient care is delivered, and these processes involve individuals throughout the health care organization, practice, or institution. On the broader external level, quality assurance should strive to improve the health care delivery system as a whole. It is able to identify and ameliorate outlier practice. A quality assurance program must be able to identify not only problematic patterns of care in the aggregate, but also the individual outlier practitioner, institution, facility, or agency. It must have the tools to intervene in that provider’s practice, when necessary, to prevent actual or future harm to patients. It can invoke positive and negative incentives for change and improve-
OCR for page 51
Medicare: A Strategy for Quality Assurance - Volume I ment in performance. To support the goal of providing compassionate and competent care, incentives and rewards for high quality are generally preferable to penalties for poor quality, but both should be available to the quality assurance program. In any case, no quality assurance system can rely solely on coercion through sanctions applied at a time and place remote from the site of care. It provides well-motivated people with timely information to improve their practice. Two essential functions of a quality assurance program are the correction of identified problems and the improvement of care generally. If a quality assurance program is to fulfill these functions, it must be able to provide practitioners with timely data at a level of aggregation or disaggregation clinically relevant to their practice. This implies that the data should be based if at all possible on rates, so that comparisons can be made to standards of practice. When problems in individual care are detected, interventions must be sufficiently timely to prevent further harm to that patient and to others who might be at risk. It has face validity to public and professionals. A quality assurance program must be understandable and reasonable to the public and to the health professionals who are subject to assessment; it must reflect their quality objectives and respond to their quality complaints. The program’s methods should have clinical relevance to practitioners for the kind of care they provide, and it should include appropriate adjustments for nonpractitioner-related variables. For the public it should illuminate decision making relevant to those aspects of health care under patient or regulatory control. Its individual elements meet requirements for reliability, validity, and generalizability. A successful quality assurance effort should demonstrate scientific rigor of its methods, beyond simple face validity to practitioners or providers. In so doing, it should minimize the need for separate and unique programs among various external organizations. It improves patient outcomes. The focus of compassionate and competent health care is patient well-being and outcomes. Ultimately, therefore, a quality assurance program should affect patient outcomes in ways that can be measured and evaluated over time. When outcomes cannot be measured or evaluated directly and when the process of care is the appropriate aspect of care to be assessed, there should be a demonstrated link between those health care processes and expected patient outcomes such that the impact on patient outcomes of improving the process of care can be inferred. It can address both individual patient and population-based outcomes. The outcomes of care for both individual patients and populations (e.g., collections of patients enrolled in HMOs or the Medicare beneficiary
OCR for page 52
Medicare: A Strategy for Quality Assurance - Volume I population as a whole) are important targets of quality assurance. This orientation helps ensure that underuse of services, especially as it is reflected in poor access to care in the first place, receives due attention as part of a quality assurance effort. It documents improvement in quality and progress toward excellence. A quality assurance program must be able to track and evaluate the effect of its efforts. Documenting improvement implies that quality-related information will be analyzed with appropriate statistical tools over time and that such information will be shared among appropriate staff and organizations. It is easily implemented and administered. A successful quality assurance strategy must find a middle ground between (on the one hand) an excessively simplistic system that has little specificity, relevance, or involvement of professionals but is relatively easy to mount and administer and (on the other) an excessively complex, costly, labor-intensive system that may itself detract from the ability of providers to render adequate patient care. Thus, although we do not intend by this criterion to discount the need for considerable investment in a Medicare quality assurance program, we do want to emphasize the need for ease of implementation and administration. It is affordable and cost-effective. Determining whether a quality assurance program is acceptable in cost and is cost-effective is very difficult; in the public sector the history of evaluations of the Professional Standards Review Organizations makes that clear. Nevertheless, the use of public monies to assure quality requires accountability. Part of the challenge of the Medicare quality assurance strategy will be to develop mechanisms to evaluate the costs and cost-effectiveness of that program; this requires articulating criteria for acceptability and effectiveness in this domain and applying those methods and criteria objectively. It includes patients and the public. An effective program must have a mechanism to listen to and respond to complaints and suggestions and a mechanism for change to act as a safety valve. Considering the size and complexity of the Medicare program, this is especially important for the external quality assurance efforts. QUALITY ASSURANCE CONCEPTUAL MODELS Designing a strategy for quality assurance for the Medicare program requires us to describe the program we expect to be in place and the steps we believe are necessary to implement that program. This in turn requires a clear conceptual framework and a set of program goals. This section summarizes important conceptual models relating to quality of health care and introduces concepts and terms that we will rely on throughout this report.
OCR for page 53
Medicare: A Strategy for Quality Assurance - Volume I The Traditional Structure-Process-Outcome Model Avedis Donabedian has articulated what continues to serve as the unifying conceptual framework for quality measurement and assurance. His widely accepted model of structure, process, and outcome has guided two decades of research and program development (Donabedian, 1966, 1980, 1982, 1984, 1988a, 1988b, 1988c). We do not depart significantly from this conceptualization. Structure Structural measures, the characteristics of the resources in the health care delivery system, apply to individual practitioners, to groups of practitioners, and to organizations and agencies. They are essentially measures of the presumed capacity of the practitioner or provider to deliver quality health care, not of the care itself. Deficiencies in structural measures are not evidence of poor care (and certainly not of poor outcomes); they may, but do not necessarily, point to crucial areas requiring improvement or reform. For health care professionals, these variables include demographic factors (e.g., age) and professional characteristics (e.g., specialty, licensure and certification, practice setting and style). For facilities and institutions, they include size, location, ownership and governance, and licensure and accreditation status. They can also include many physical attributes (e.g., special units and computer capabilities) and a large set of organizational factors (e.g., staff-to-staff or staff-to-patient ratios; employee morale and turnover). Quality assurance programs acquire information on structural measures in several ways. The simplest is probably through mechanisms of licensure, certification, or accreditation that are maintained by states, professional associations, or third-party payers. Provider surveys can also provide relevant information. Some observers question the relationship of structural measures to either process or outcome variables because of inadequate measures and little empirical evidence of direct connections; they tend to downplay the importance of structural characteristics of health care organizations as markers of quality. Nevertheless, over the last decade various elements of the quality assurance field have more explicitly emphasized accountability, governance, and lines of authority, especially for hospitals. Regulatory agencies such as states, voluntary associations such as the Joint Commission, and institutional quality assurance officials are much more likely to repose the ultimate responsibility for quality care in boards of trustees or directors and equivalent organizational executives. Consequently, structural measures that reflect organizational patterns, lines of authority, and communication
OCR for page 54
Medicare: A Strategy for Quality Assurance - Volume I within health care delivery systems command attention in quality assurance programs. Process The process of care embodies what is done to and for the patient. Process measurement is the most common approach to quality assessment and assurance today. The relative merits of process measurement versus outcome measurement have been debated vigorously over the years; the consensus is that both are necessary but that neither is sufficient to successful quality assurance (Table 2.2). Process measurement can be directed at individual practitioners, teams of practitioners, or entire systems of care. It can include aspects of whether and how patients seek and obtain care. Process measures seek information to identify problems that occur during the delivery of care. Elements of care delivery are evaluated against criteria that reflect professional standards of good quality care and, increasingly, patient-oriented measures of satisfaction. Data about processes can be obtained in numerous ways. These include patient reports of care rendered, direct observation of care, review of medical records (or abstracts of records) and similar documents, and analysis of insurance claims or other utilization data. The presumed advantages of process-of-care evaluation are several. It has great appeal to practitioners because it is directly related to what they do. It is easy to explain and to interpret the approach and its findings. Reliable, valid criteria and methods are available. In some cases, review of care against process criteria can be nearly “real time,” meaning that corrective actions can be very timely. Data can be analyzed by individual providers or aggregated in various ways that support comparisons of practice patterns across communities or health delivery systems. For certain settings, process measurement may be the preferred strategy (compared to outcome measurement), such as for ambulatory office practice (Palmer, 1988). Finally, process measurement can point directly to specific areas needing performance improvement, which is a fundamental aim of quality assurance. Process-of-care assessment is not without its limitations. Resource costs can be high, which explains the considerable attractiveness of using administrative or insurance claims data sets. For some settings, such as home health care, the data sources are poor; for others, such as long-term-care facilities (Kane and Kane, 1988), process measures may be less informative than some alternative outcome measures because of the repetitive nature of much that is done in such settings. Process measures may focus on issues
OCR for page 55
Medicare: A Strategy for Quality Assurance - Volume I TABLE 2.2 Strengths of Process and Outcome Measurement on Selected Dimensions Dimensions Type of Measurementa Process Outcome Relevance to goal of health care − ++ Appeal to practitioners and institutions ++ + Appeal to patients and public + ++ Can detect problems of overuse ++ – Can detect problems of underuse + + Can detect poor technical or interpersonal quality ++ − Real time review and timely intervention possible ++ − Points directly to specific areas needing performance improvement ++ − Reflects important trends over time + ++ Particularly useful for certain settings physician office care ++ − hospital care ++ ++ post-hospital care (e.g., home health) + + Minimizes intrusiveness for providers − ++ Minimizes intrusiveness for patients ++ − Reliable and valid assessment methods + + Reliable and valid evaluation criteria + − Relevant informaton recorded in administrative (billing or utilization) data + − − Consensus on best practices ++ to − − NA Consensus on best outcomes NA ++ Can account for biological variability − + Can account for patient preferences − − + Can account for patient behavior − − − Can assign accountability for performance when care is from multiple providers over time − − − Costs of measurement + to − + to − − aCode: ++, strong in this dimension; +, adequate in this dimension; −, fair in this dimension; − −, poor in this dimension. NA means not applicable.
OCR for page 58
Medicare: A Strategy for Quality Assurance - Volume I limited by the fact that outcomes are not routinely or uniformly recorded, especially when those data are not needed for reimbursement purposes. Finally, outcome measurement can incur high resource costs in time, manpower, and dollars, especially to overcome the problems caused by incomplete documentation of what transpired. Within the traditional model, more effort has been directed toward quality assessment than quality assurance. Formal methods exist for completing the full cycle, using the information from the assessment to assure quality. No single assurance system, however, parallels the structure, process, and outcome framework. Continuous Quality Improvement Within the last five years or so, a model of quality improvement that had its start in the manufacturing field has begun to be applied in the health field. It goes by various names, including “quality improvement process,” “total quality management,” “organization-wide quality improvement” (called Total Quality Control in Japan), and similar phrases; “continuous improvement” may be the most universally recognized term (Batalden and Buchanan, 1989; Berwick, 1989). The philosophic and technical basis for this model evolved from a set of management and statistical control methods pioneered decades ago by U.S. statisticians and engineers (but implemented chiefly by post-World War II Japanese industrialists) for application in industry, primarily manufacturing (Deming, 1986; Walton, 1986; Garvin, 1986, 1988; Juran et al., 1988). These concepts, as translated for health care delivery, are described below. Four Core Assumptions First, people involved in delivering health care work mainly in organizations; therefore, quality improvement uses the energy and lines of accountability of an organization for improvement. To do so, top leadership must be committed to quality improvement. Second, health care workers—administrators, physicians and other professionals, paraprofessionals, and support staff—wish to perform to the best of their capacity. Third, when workers cannot attain their best performance, wasteful, needlessly complex, and undependable systems or organizational methods of work are often to blame. Fourth, the interaction of individuals and the organizations and systems within which they practice can always improve. Eight Key Constructs First, the emphasis is on external customers or recipients of care. All that is done is done for the benefit of the patient.
OCR for page 59
Medicare: A Strategy for Quality Assurance - Volume I Second, all that precedes the benefit to the patient—for example, facilities, equipment, providers, support staff, and organizational policies—must be involved in a relentless, systematic, and cooperative effort to improve care. The continuous improvement model calls on all involved parties to participate in quality monitoring and to shift the quality curve upward rather than just eliminate the outliers. Continuous improvement, by its very name, assumes that there is no permanent threshold to good performance; in health care, it implies that health professionals and the settings in which they practice should never be content with present performance. Third, as illustrated in Figure 2.1, activities are cyclic, involving continuous “planning, doing, checking, and acting” (PDCA). One expert has described the PDCA cycle as the “demoralization of the scientific method,” making it possible for all to participate in the application of these methods to daily work (Paul Batalden, citing George Box, personal communication, 1989). In health care, this implies attention to processes of care that are responsive to patients’ needs. It recognizes that the only way to improve outcomes is to work on what produced them. Fourth, it views the work of individuals and departments within health care organizations as interconnected. One supplies “work products” to others; one also receives from others. As such, people and departments serve as their own internal “suppliers” and “customers,” and their interconnected activities are intended to benefit the external customer. Patients may be the key external customers but, for example, subspecialty physicians may be FIGURE 2.1 A Simplified Diagram of the Continuous Improvement Model for Both Internal and External Customers
OCR for page 60
Medicare: A Strategy for Quality Assurance - Volume I customers of primary care physicians, and pharmacy staff may be suppliers for nursing staff. In short, “I am a supplier for people who depend on what I do, and I am a customer when I depend on what others in the organization do” (Donald Berwick, personal communication, 1988). Fifth, it places considerable emphasis on systems or “processes” as the way care gets delivered (Paul Batalden, personal communication, 1989). (See the flow diagram in Figure 2.2.) Suppliers provide inputs that are transformed by a series of actions into outputs. These are received by customers who have needs, expectations, and values by which they judge outcomes, that is, ascribe benefit to that output.6 Organizations such as hospitals or physician offices are viewed as large networks of interrelated processes with thousands of internal customer-supplier relationships. Constant improvement of every production process by everyone involved is the central focus. In this view, every health worker has two roles: first, doing his or her job and, second, improving the job. Sixth, improvement occurs by integrating the voices of customers and of processes of care into the cyclical redesign of service and care (Figure 2.1). Both the customer and those serving the customer must contribute to information gathering—the former through surveys, complaint procedures, and similar channels (often external to the organization) and the latter through continual internal monitoring of procedures, resource use, and patient care. Seventh, active, visible commitment of the highest leadership of the organization is necessary. Because this view holds that the major impediment to improving quality is found in the way people and organizations work together, engagement in quality improvement must permeate an organization, starting at the top. Eighth, the continuous improvement approach uses a set of practical FIGURE 2.2 Flow Diagram of the Process-Outcome Relationship in the Continuous Improvement Model
OCR for page 61
Medicare: A Strategy for Quality Assurance - Volume I techniques that facilitate learning and action. These tools (e.g., flow charts, fishbone diagrams, and run charts) have been adapted from the decades of organization-wide efforts at quality control in industry and are intended for use by people at all levels of the organization. All can be used to identify and analyze the various processes of health care delivery and to monitor the effectiveness of quality improvement interventions in organizations. Applications Outside the health field, the industries represented by companies that have implemented various quality control programs using this model vary widely: automobile manufacturers, public utilities, communications, and consumer products. In 1987, Congress established the Malcolm Baldrige National Quality Award (P.L. 100–107), which is patterned after the prestigious Deming prize awarded in Japan since 1951.7 Its aim is to encourage quality accomplishments and excellence in U.S. manufacturing and service companies and small businesses by awarding a prize to companies that pass a rigorous examination of their quality measurement program and demonstrable accomplishments.8 As the continuous improvement model is increasingly being diffused in the nonhealth sectors of the U.S. economy, its appeal to the health care community is spreading.9 For instance, the Joint Commission’s Agenda for Change (Joint Commission, 1988; Jurkiewicz, 1988) has adopted a new set of “Principles of Organization and Management Effectiveness” that strongly emphasizes total organizational commitment to the continuous improvement of the quality of patient care.10 Several hospitals and hospital chains have already implemented various forms of quality improvement activities. For instance, West Paces Ferry, a hospital owned by the Hospital Corporation of America, has embarked on a quality improvement plan (National Demonstration Project, 1989); the hospital has focused specifically on “improving the mean rather than only policing some lower margin of acceptable performance” (Chip Caldwell, personal communication, 1989).11 Quality Assurance Concepts from an International Perspective The commitee sought information on international aspects of quality assurance for the purpose of learning whether concepts and methods in other countries might be helpful to the United States. In preparing this brief discussion we draw on a paper prepared for the study (Reerink, 1989). After analyzing the efforts of several countries Reerink concludes that the United States is the front runner in the field of quality assurance. In a great majority of countries, the Netherlands being a notable exception, quality assurance is viewed with disdain or mistrust. By comparison, in the
OCR for page 62
Medicare: A Strategy for Quality Assurance - Volume I United States quality assurance in health care is studied and implemented with considerable intensity. Some countries have followed the U.S. example and developed (or are in the process of developing) their quality assurance programs on the traditional structure-process-outcome model. In a few countries (e.g., the Netherlands and Malaysia), performance indicators and health accounting are dominant elements of the quality assurance programs. Reerink reports that the most important facets of quality assurance in countries he studied are the dominant role taken by health professions in establishing quality assurance systems and the widely held perception of the necessity of health professional leadership for successful quality assurance programs. TRADITIONAL AND CONTINUOUS IMPROVEMENT MODELS COMPARED Several aspects of the continuous improvement model resemble those of contemporary systems of quality assurance or performance monitoring described a decade or more ago for the health care field. These systems include the bi-cycle concepts of Brown and Uhl (1970) and the health accounting approach of Williamson12 (1978, 1988), both of which have cycles quite analogous to the planning-doing-checking-acting (PDCA) approach. Moreover, both approaches incorporate notions of structure (e.g., organizational factors and high-level accountability), process (e.g., patient care activities), and outcomes (e.g., patient well-being or satisfaction). For instance, both the traditional and the continuous improvement models of quality assurance stress the importance of outcomes (or achievable benefit); Williamson’s health accounting approach, for instance, starts with achievable benefit not being achieved and works back to the process of care. The main distinction is that the latter more explicitly involves patient (i.e., customer) values as a critical element of outcomes. Both approaches also acknowledge the importance of information that links processes to outcome. Thus, in many ways the continuous improvement approach is consistent with traditional notions of quality assurance. The two concepts depart from each other mainly in five ways. The continuous improvement model, first of all, emphasizes continual efforts to improve performance and value even when high performance standards appear to be met. In the latter case, traditional quality assurance activities would cease or shift attention elsewhere. Second, continuous improvement stresses the evaluation of simple and complex systems from the perspectives of the customers. This has the effect, among other things, of directing attention to the way people and departments in organizations work together and, thus, to sources of variation and multidisciplinary quality-of-care issues that traditional quality assurance approaches might not detect or target for change.
OCR for page 63
Medicare: A Strategy for Quality Assurance - Volume I Third, it emphasizes understanding the views of patients and other customers about the care process and their outcomes. This tends to draw more attention to patient satisfaction than has been heretofore the case. Fourth, the continuous improvement model is designed to improve the overall (or average) performance of individuals and the organization more than to remove outliers. That is, poor practitioners or institutions are not the target. Finally, although both approaches would place the accountability for quality high in the organization’s leadership, the continuous improvement model explicitly vests ultimate responsibility for quality and quality improvement at the very top of the management structure while still emphasizing the personal responsibility of all members to contribute to quality improvement. In contrast to traditional quality assurance, however, the utility of the continuous improvement model in dealing with clinical problems encountered in ordinary medical practice is yet to be learned. For instance, its applicability to problems of poor physician decision making in choosing diagnostic or therapeutic modalities is unproven, and its ability to deal with issues of overuse or underuse remains to be shown. Most of the health applications to date have targeted organizational processes and customer or patient satisfaction. Whether health care institutions and facilities can successfully implement the continuous improvement approach with a focus on meaningful medical, nursing, and other professional quality-of-care issues will have to be tested rigorously over the next few years. Earlier in this chapter we identified 15 attributes of quality assurance programs. How do the two models reviewed in this chapter—the traditional and the continuous improvement models—stand when judged against these criteria? Both approaches doubtless have the ability to address overuse, underuse, and poor technical and interpersonal quality, to minimize intrusion into the patient-provider relationship, to deal with outlier practice and performance, and to provide practitioners and providers with timely information to improve performance, although they would do so with varying degrees of success. For health care, neither the traditional nor the continuous improvement approach can be proven, at the moment, to be especially rigorous scientifically, to have a consistently positive impact on patient outcomes, to document improvements in quality and progress toward excellence, to be easy to implement or administer, or to be affordable or cost-effective. The traditional approaches appear valid and are acceptable to professionals and providers, at least for internal programs. They are more suitable for non-organization-based practitioners (such as independent, fee-for-service practitioners) and for population-based outcomes of care. The continuous improvement approach more explicitly includes patients and the public, fosters improvement throughout the health care organization and system,
OCR for page 64
Medicare: A Strategy for Quality Assurance - Volume I attempts to use the scientific method, and applies both positive and negative incentives for change and improvement in performance. SUMMARY This review and comparison of the traditional and continuous improvement models for quality assurance sets forth concepts and terms used throughout the report. The two models are also judged against 15 attributes the committee identified as desirable for a quality assurance program. The lesson is that no single approach or conceptual framework is likely either to suit our purposes or to meet the criteria we have identified for an effective quality assurance program. The study committee had widely divergent views on the benefits of the continuous improvement model given the limited knowledge about its application to clinical problems and settings of care outside the hospital; the lack of consensus on the committee about a “best” approach to quality assurance and quality improvement is a microcosm of the larger debate in the health care community on concepts and models of quality assurance. The continuous improvement model has already had a considerable impact on the quality assurance field. The committee’s debate underscored the need for flexibility and innovation in the Medicare quality assurance program over the coming decade. Caution was taken to avoid locking the Medicare program into, on the one hand, an older, familiar conceptual model or, on the other, a very appealing yet unproven new system. Rather, we draw on all the elements of these approaches to fashion a program for quality assurance in Medicare (Chapter 12) that we believe will foster improvement in the quality of health care as it was defined in Chapter 1. NOTES 1. The quality-of-care literature is enormous. The few historical overviews of quality assurance programs include: Egdahl, 1973; Williamson, 1977, 1988; Williams and Brook, 1978; Lohr and Brook, 1984; and Donabedian, 1989. 2. The mechanics of quality assurance may be tailored to specific categories of quality problems or to given settings, such as efforts designed to address problems of underuse of outpatient or inpatient care in a health maintenance organization (HMO) or those intended to improve nursing aide care in a home health agency. Chapters 7, 8, and 9 discuss these points at greater length; examples of methods are found in Volume II, Chapter 6. 3. The term outliers typically refers to clinicians or institutions that render seriously substandard or unorthodox care. A rough rule of thumb might be that it refers to the worst 1 percent or 5 percent of providers on a quality measure, although it can be defined in strictly statistical terms.
OCR for page 65
Medicare: A Strategy for Quality Assurance - Volume I 4. Outcome measurements are not new. One of the earliest documented efforts to monitor outcomes is that conducted by the surgeon Ernest A. Codman in the early 1900s. Codman’s approach, known as the “end result system,” included a retrospective review of the outcomes of each of his surgery patients one year following the operation (Brook et al., 1976; Nash and Goldfield, 1989). 5. The literature on the assessment of health status and quality of life is very large. For recent overviews of this field, see Bergner, 1985; Bergner and Rothman, 1987; Katz, 1987; Lohr and Ware, 1987; Lohr, 1989; Greenfield, 1989; Kane and Kane, 1989; Mosteller and Falotico-Taylor, 1989; and Sabatino, 1989. Avery et al., 1976 and Brook et al., 1976 provide landmark overviews of quality assessment using outcome measures. 6. These uses of the terms process and outcomes are not the same as in the familiar triad of structure, process, and outcome, but they are not in conflict. For example, as depicted in Figure 2.2, this use of outcome includes the value set of the customer or recipient of the output of the process. Process includes inputs (which in turn may include structural elements), actions (which may be conditioned by structural factors), and technical outputs. 7. In 1989, Florida Power & Light Co. became the first non-Japanese company to win Japan’s Deming award. 8. Several firms participated in the 1988 and 1989 competition rounds. In 1988 awards were given to Westinghouse Electric Corporation’s Commercial Nuclear Fuel Division; Motorola, Inc.; and Globe Metallurgical, Inc. The 1989 awards went to Milliken & Company and Xerox Corporation’s Business Products and Systems. 9. Several firms participated in an innovative project begun in 1987—the National Demonstration Project on Industrial Quality Control and Health Care Quality—designed to permit high-ranking quality assurance professionals to collaborate on ways to improve the state of health care quality assurance (National Demonstration Project, 1989). The current phase of this project involves a series of courses on quality improvement and studies of clinical and administrative problems tackled through continuous improvement techniques (Donald Berwick, personal communication). 10. The new principles will be elaborated in standards for accreditation of health care providers. They are intended to encourage organizations to weave continuous improvement attitudes and methods throughout the organization, specifically in strategic planning, allocation of resources, role expectations, reward structures, performance evaluations, and the role of the organization in the community. To elaborate these principles, the Joint Commission also appointed a Task Force charged to develop principles for continuous improvement; its work will continue into 1990. 11. One technique the West Paces Ferry program stresses is “benchmarking,” which Camp (1989a, 1989b) defines as a process “for uncovering…best industry practices….” Camp argues for seeking best practices wherever they might exist (e.g., levels of performance attainable in another setting or industry). 12. The system supported by CBO, the National Organization for Quality Assurance in Hospitals in the Netherlands, incorporates many elements common to Williamson’s health accounting approach (Reerink, 1989).
OCR for page 66
Medicare: A Strategy for Quality Assurance - Volume I REFERENCES Avery, A.D., Lelah, T., Solomon, N.E., et al. Quality of Medical Care Assessment Using Outcome Measures: Eight Disease-Specific Applications. R-2021/2-HEW. Santa Monica, Calif.: The RAND Corporation, August 1976. Batalden, P.B. and Buchanan, E.D. Industrial Models of Quality Improvement. Pp. 133–159 in Providing Quality Care: The Challenge to Clinicians. Goldfield, N. and Nash, D.B., eds. Philadelphia, Pa.: American College of Physicians, 1989. Bergner, M. Measurement of Health Status. Medical Care 23:696–704, 1985. Bergner, M. and Rothman, M.L. Health Status Measures: An Overview and Guide for Selection. Annual Review of Public Health 8:191–210, 1987. Berwick, D.M. Sounding Board. Continuous Improvement as an Ideal in Health Care. New England Journal of Medicine 320:53–56, 1989. Brook, R.H., Avery, A.D., Greenfield, S., et al. Quality of Medical Care Assessment Using Outcome Measures: An Overview of the Method. R-2021/1-HEW. Santa Monica, Calif.: The RAND Corporation, August 1976. Brown, C.R., Jr. and Uhl, H.S.M. Mandatory Medical Education: Sense or Nonsense? Journal of the American Medical Association 213:1660–1668, 1970. Camp, R.C.Benchmarking: The Search for Industry Best Practices That Lead to Superior Performance. Part III. Why Benchmark? Quality Progress 22:76–82, March 1989a. Camp, R.C. Benchmarking: The Search for Industry Best Practices That Lead to Superior Performance. Part IV. What to Benchmark. Quality Progress 22:62–69, April 1989b. Deming, W.E. Out of the Crisis. Cambridge, Mass.: Massachusetts Institute of Technology Press, 1986. Donabedian, A. Evaluating the Quality of Medical Care. Milbank Memorial Fund Quarterly 44:166–203, July (part 2) 1966. Donabedian, A. Explorations in Quality Assessment and Monitoring. Volumes I-III. Ann Arbor, Mich.: Health Administration Press, 1980, 1982, 1984. Donabedian, A. Quality Assessment and Assurance: Unity of Purpose, Diversity of Means. Inquiry 25:173–192, 1988a. Donabedian, A. The Quality of Care. How Can it be Assessed? Journal of the American Medical Association 260:1743–1748, 1988b. Donabedian, A. Monitoring: The Eyes and Ears of Healthcare. Health Progress, 69:38–43, November 1988c. Donabedian, A. Reflections on the Effectiveness of Quality Assurance. Paper prepared for the Institute of Medicine Study to Design a Strategy for Quality Review and Assurance in Medicare, 1989. Egdahl, R.H. Foundations for Medical Care. New England Journal of Medicine 288:491–498, 1973. Elinson, J., quoting himself in Donabedian, A., et al. Advances in Health Assessment Conference Discussion Panel. Journal of Chronic Diseases 40(Supplement 1):183S-191S, 1987. Ellwood, P. Outcomes Management. A Technology of Patient Experience. New England Journal of Medicine 318:1549–1556, 1988.
OCR for page 67
Medicare: A Strategy for Quality Assurance - Volume I Garvin, D.A. A Note on Quality: The Views of Deming, Juran, and Crosby. Harvard Business School Note Publication No. 9–687–011. Cambridge, Mass.: Harvard College, 1986. Garvin, D.A. Managing Quality. The Strategic and Competitive Edge. New York, N.Y.: The Free Press, A Division of Macmillan, 1988. Greenfield, S. The State of Outcome Research: Are We on Target? New England Journal of Medicine 320:1142–1143, 1989. Joint Commission (Joint Commission on Accreditation of Healthcare Organizations). Agenda for Change. Update 1:1, 6, 1987. Joint Commission. Agenda for Change. Update 2:1, 5, 1988. Juran, J.M., Gyrna, F.M., Jr., and Bingham, R.S., Jr. Quality Control Handbook. Fourth edition. Manchester, Mo.: McGraw-Hill, 1988. Jurkiewicz, M.J. Spectrum 1988. The Joint Commission and the Agenda for Change. American College of Surgeons Bulletin 73:20–25, 1988. Kane, R.A. and Kane, R.L. Long-Term Care: Variations on a Quality Assurance Theme. Inquiry 25:132–146, 1988. Kane, R.A. and Kane, R.L. Reflections on Quality Control. Generations 13:63–68, Winter 1989. Katz, S., guest ed. The Portugal Conference: Measuring Quality of Life and Functional Status in Clinical and Epidemiological Research. Journal of Chronic Diseases 40:459–650, 1987. Lohr, K.N. Outcome Measurement: Concepts and Questions. Inquiry 25:37–50, 1988. Lohr, K.N., guest ed. Advances in Health Status Assessment: Conference Proceedings. Medical Care (Supplement) 27:S1-S294, March 1989. Lohr, K.N. and Brook, R.H. Quality Assurance in Medicine. American Behavioral Scientist 27:583–607, 1984. Lohr, K.N. and Ware, J.E., Jr., guest eds. Proceedings of the Advances in Health Assessment Conference. Journal of Chronic Diseases 40:1S-193S, (Supplement 1) 1987. Lohr, K.N., Yordy, K.D., and Thier, S.O. Current Issues in Quality of Care. Health Affairs 7:5–18, Spring 1988. Mosteller, F. and Falotico-Taylor, J., eds. Quality of Life and Technology Assessment. Monograph of the Council on Health Care Technology. Washington, D.C.: National Academy Press, 1989. Nash, D.B. and Goldfield, N. Information Needs of Purchasers. Pp. 5–24 in Providing Quality Care: The Challenge to Clinicians. Goldfield, N. and Nash, D.B., eds. Philadelphia, Pa.: American College of Physicians, 1989. National Demonstration Project on Quality Improvement in Health Care. Quality Improvement in Health Care. A Newsletter. Boston, Mass.: Harvard Community Health Plan. Premiere Issue, 1989. O’Leary, D. Quality Assessment: Moving from Theory to Practice. Journal of the American Medical Society 260:1760, 1988. Palmer, R.H. The Challenges and Prospects for Quality Assessment and Assurance in Ambulatory Care. Inquiry 25:119–131, 1988.
OCR for page 68
Medicare: A Strategy for Quality Assurance - Volume I Reerink, E. Study on International Aspects of Quality Assurance. Paper prepared for the Institute of Medicine Study to Design a Strategy for Quality Review and Assurance in Medicare, 1989. Roberts, J.S. Quality Assurance in Hospitals: From Process to Outcomes. Pp. 48–62 in Quality of Care and Technology Assessment. Lohr, K.N. and Rettig, R.A., eds. Report of a Forum of the Council on Health Care Technology. Washington, D.C.: National Academy Press, 1988. Sabatino, C. Homecare Quality. Generations 13:12–16, Winter 1989. Walton, M. The Deming Management Method. New York, N.Y.: Dodd, Mead & Company, 1986. Weiss, C.H. Evaluation Research. Englewood Cliffs, N.J.: Prentice Hall, 1972. Williams, K.N. and Brook, R.H. Quality Measurement and Assurance: A Literature Review. Health & Medical Care Services Review 3:1, 3–15, May/June 1978. Williamson, J.W. Improving Medical Practice and Health Care: A Bibliographic Guide to Information Management in Quality Assurance and Continuing Education. Cambridge, Mass.: Ballinger, 1977. Williamson, J.W. Assessing and Improving Outcomes in Health Care: The Theory and Practice of Health Accounting. Cambridge, Mass.: Ballinger, 1978. Williamson, J.W. Future Policy Directions for Quality Assurance: Lessons from the Health Accounting Experience. Inquiry 25:67–77, 1988.
Representative terms from entire chapter: