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
Guidelines for Clinical Practice: From Development to Use 2 Developing Clinical Practice Guidelines Medicine ... is mobile, and many of us get breathless not so much by trying to keep up with medical progress as by trying to avoid being run over by it. Roger I. Lee, 1958 The rapid pace of development in biomedical science and technology can, on the one hand, make guidelines a useful aid to busy practitioners and, on the other hand, subject guidelines to rapid obsolescence. This reality underscores the importance of the review and updating process noted in Chapter 1. More generally, the way clinical practice guidelines are developed can strongly affect their potential for effective use by practitioners, patients, and others. Thus, planning for successful implementation must begin with development and continue through cycles of revision and updating. One theme underlying the attributes presented in the preceding chapter is that those who devise guidelines should anticipate what will make the guidelines practical and credible. Most obvious is that guidelines need to be specific, comprehensive, and flexible enough to be useful in everyday clinical practice. In addition, the logic, language, and symbols used in the guidelines should be unambiguous to intended users and easy to follow. Because practitioners and other potential users will want to know who developed the guidelines and how they did it, the development process and the participants in the process should be documented. Developers of guidelines should describe the strength of the evidence and the relative importance of the projected benefits and risks; they should also, in general, indicate how compelling is the case for particular interventions. A multidisciplinary development process that includes all key groups will encourage acceptance of guidelines by the members of these groups—including patients and their surrogates. Inclusion in this sense need not be
OCR for page 46
Guidelines for Clinical Practice: From Development to Use limited to representation in the group drafting the guidelines; it can extend to participation in hearings, reviews, pretests, and similar activities. Such an inclusive approach, although potentially requiring more time and funds, can help developers of guidelines better understand the situations in which guidelines may be applied. The sense of this committee, therefore, is that planning for implementation and later evaluation of guidelines must take place during the development phase. What does or does not happen at the development stage may materially affect the success or failure of a set of guidelines, independent of the quality of the implementing efforts that follow. The examination of guidelines implementation in Chapters 3, 4, 5, and 6 underscores this point and leads to further discussion of guidelines development in Chapter 7. The succeeding sections of this chapter first describe the major types of organizations involved in guidelines development in both the public and private sectors and then discuss how the enterprise is evolving. The chapter closes with a brief commentary on methods and costs of guideline development. For this overview, the study committee relied heavily on the published literature, its site visits, the public hearing, focus groups, and other study activities, as well as on its own expertise. PLURALISM AND DIVERSITY IN GUIDELINES DEVELOPMENT Systematic efforts to develop clinical practice guidelines have grown dramatically in recent years. Professional societies, public agencies, health care institutions, and researchers have become appreciably more active and visible in the guidelines arena (Woolf, 1990b; Kosterlitz, 1991); the field also has at least one regular newsletter, Report on Medical Guidelines and Outcomes Research (Robinson, 1991). According to the American Medical Association (AMA), 8 physicians organizations reported active involvement in developing guidelines before 1980; now, more than 50 organizations can report such activity (AMA, 1991a). 1 The creation in 1989 of the Agency for Health Care Policy and Research (AHCPR), and its Forum for Quality and Effectiveness in Health Care, provided new focus and visibility for public-sector activities. Insurers, health maintenance organizations (HMOs), and other private organizations have also become more active. This pluralism of sponsorship reflects the breadth of interest in guidelines, the special concerns of different sponsors, and the varying outlooks on what topics warrant guideline development and by what methods (Audet 1 The AMA publishes a Directory of Practice Parameters (AMA, 1991a); as of late 1991, it listed and cross-referenced 1,319 practice parameters developed by 45 U.S. organizations and provided information on how to obtain the actual parameters. The AMA also publishes quarterly updates of this directory, indicating newly completed and withdrawn guidelines.
OCR for page 47
Guidelines for Clinical Practice: From Development to Use et al., 1990). A recent General Accounting Office (GAO, 1991b) survey of medical specialty societies succinctly described this diversity: ''No two medical specialty societies with whom we spoke have produced similar guidelines for similar reasons in a similar fashion" (p. 11). This observation applies broadly to the whole field of guidelines development. The following overview of types of activities and trends illustrates some of the diversity of this undertaking without attempting to characterize the quality of specific activities in terms of the attributes identified in Chapter 1. The discussion, however, tends to slight less well-supported and less publicized efforts at guidelines development, which may be quite numerous and idiosyncratic to particular institutions; it also does not cover individuals working independently to develop guidelines, medical texts, and similar materials. As a consequence, this chapter risks implying that most efforts at guidelines development have the characteristics of the better-organized work described here. Such programs still appear to be atypical, although many organizations are attempting to improve their procedures and methods as discussed further in Chapter 7.2 This chapter's discussion of the work of specific organizations does not imply endorsement. PROFESSIONAL SOCIETIES AND RELATED ENTITIES The last century has been marked by the development of professional medicine and by broad deference to its judgments, a deference that has been challenged in recent years. Although guidelines for clinical practice emerged initially as one sign of professional responsibility, the current acceleration of activity is a function both of the challenges to the profession and of the profession's responses. Medicine is characterized by a wide array of professional organizations. Some are general in focus (most notably, the AMA), but most center on the medical specialties and subspecialties with which most physicians strongly identify. These "academies," "colleges," or "societies" often sponsor specialty-specific, peer-reviewed clinical journals that serve as a major source of information for physicians; a number of these organizations have added the development of practice guidelines to their agendas. Associations for health care professionals other than physicians—for example, the American Nurses Association (ANA) and the American Dental Association (ADA)— are also increasingly involved in guidelines efforts. Overall, professional societies are generally seen as having taken the lead in organized efforts to develop practice guidelines, thereby serving as an important, perhaps key, 2 Readers interested in technical descriptions of current methods might examine the following: the AHCPR Forum manual developed to assist guidelines panels (Woolf, 1990a), the American College of Physicians procedure manual (ACP, 1986), and the introductory manual commissioned by the Council of Medical Specialty Societies (Eddy, 1991c).
OCR for page 48
Guidelines for Clinical Practice: From Development to Use source of authoritative communication to practitioners about what constitutes appropriate care. As illustrated below, the involvement of individual professional organizations varies both in scope and purpose. How Are Professional Societies Involved? Relatively formal specialty society activity in the area of guideline development appears to date back at least 50 years to the American Academy of Pediatrics (AAP) monograph on control of infectious diseases, first published in 1938 and now in its 22nd edition (AAP, 1991). Other organized efforts reaching back more than a decade include those of the American College of Obstetricians and Gynecologists (ACOG; beginning in 1959), the American Society of Anesthesiology (ASA; in 1968), the American College of Physicians (ACP; in 1976), and the American College of Radiology (ACR; in the 1970s).3 Some efforts related to guidelines are those of single-specialty societies; others are collaborative efforts. The ACP through its Clinical Efficacy Assessment Project has developed a particularly broad program that cuts across all facets of medical care and many medical specialties. Topics run from expensive and inexpensive diagnostic tests (e.g., magnetic resonance imaging and complete blood counts) to indications for surgery (e.g., carotid endarterectomy) to cardiac rehabilitation. Narrower in range and more typical is the work of the American Academy of Ophthalmology (AAO) to develop "preferred practice patterns" for comprehensive eye examinations and various disease specific topics (Sommer et al., 1990). In a broad-ranging look at a major medical problem, the American College of Cardiology (ACC) recently published a supplement to the Journal of the American College of Cardiology (ACC, 1989) based on a symposium covering quality-and cost-conscious cardiovascular care and the role of decision modeling. In 1980, the ACC and the American Heart Association (AHA) together started a Task Force on Assessment of Diagnostic and Therapeutic Cardiovascular Procedures.4 Work by this group on coronary artery bypass graft 3 Space does not permit a complete listing of all the topics on which these specialty societies have developed guidelines in the past decade or so. As noted earlier, the AMA is now tracking such efforts as are some commercial firms. The Medical Technology Assessment Directory (IOM, 1988) provides a profile and complete list of assessment documents and guidelines produced by many groups. See also the February 1990 issue of the Quality Review Bulletin for detailed descriptions of selected specialty society programs. 4 Among the ACC/AHA Task Force reports related to acute myocardial infarction (AMI) are guidelines on the following topics: ambulatory electrocardiography, percutaneous transluminal coronary angioplasty (see the excerpt in Appendix A), clinical use of cardiac radionuclide imaging, exercise testing, and permanent cardiac pacemaker implantation (IOM, 1990a). Some recent guidelines publications have discussed early management of patients with AMI (Gunnar et al., 1990).
OCR for page 49
Guidelines for Clinical Practice: From Development to Use surgery (Kirklin et al., 1991) is an instructive example of what can be accomplished by cooperative actions of specialty societies. To cite another example, the American College of Nuclear Physicians (ACNP) and the Society of Nuclear Medicine (SNM) may begin jointly to develop guidelines for nuclear medicine practitioners. The AMA has spearheaded two broad, multiorganizational activities limited largely to physician groups. The first is a collaboration between the AMA and 14 national medical specialty societies known as the Specialty Society Partnership. The second is the Practice Parameters Forum comprising "all national medical specialty and state medical societies interested in participating . . . [in an activity] created for the purpose of facilitating the development, evaluation, and implementation of practice parameters" (AMA, 1990c, p. 16); the Forum numbers nearly 50 volunteer specialty and state medical societies. (The AMA prefers the term practice parameters to practice guidelines.) In addition to playing a coordinating role, the AMA issues recommendations on specific clinical problems and technologies through its Diagnostic and Therapeutic Technology Assessment (DATTA) program and its Council on Scientific Affairs. The Journal of the American Medical Association has published a widely read and frequently cited series of articles on guidelines (Eddy, 1990a-j,l; 1991d,e). The organization also has issued a set of attributes to guide the development and assessment of practice guidelines (AMA, 1990a). In the recent history of guideline development, the 1987 annual meeting of the Council of Medical Specialty Societies (CMSS) was a significant point (CMSS, 1987; Gschwend, 1990). Representatives of several societies that had taken a leading role in guideline development up to that time (for example, ACOG, ACR, ASA) debated the importance and risks of standard setting with representatives of many specialty groups that had not yet become involved. In summarizing the meeting, the president of the CMSS concluded that it is "the duty of specialty societies and physicians to lead in shaping quality guidelines" (CMSS, 1987, p. 71). The CMSS is also involved in coordination and training activities (Woolf, 1990b). A 1988 conference sponsored by the congressional Physician Payment Review Commission (PPRC) provided further national focus by linking specialty society and other efforts to develop practice guidelines to a policy perspective that emphasized guidelines as one vehicle for rationalizing and controlling expenditures. AMA leadership in several areas has already been cited and also includes publication of a major paper (1990b) on legal implications of practice guidelines. That paper provided important reassurance to some groups that they would not be imprudently inviting legal problems by developing guidelines—as long as they followed objective procedures, focused on scientific and medical considerations, and exercised due care in
OCR for page 50
Guidelines for Clinical Practice: From Development to Use formulating and updating their recommendations. (For commentaries on the AMA paper, see Miike, 1989; Brennan, 1990; Hall, 1990; and Havighurst, 1990b.) The creation of the Society for Medical Decision Making in 1980 was a further important milestone for guidelines development, insofar as the society serves as a locus for experts in guideline and algorithm development and clinical decision making to exchange information and share experiences. Some experts, in fact, regard the advanced medical decision analyses of this group as having provided the best advice about medical technologies and interventions in recent years. The field has published its own specialty journals (such as The Journal of Medical Decision Making) since the early 1980s. An array of related professional initiatives, including certain kinds of textbooks and other publications, also deserve recognition. For example, the Manual of Medical Therapeutics (known colloquially as "The Washington Manual") from Washington University in St. Louis is a widely used pocket reference, now in its 29th edition. It is aimed at and used chiefly by house officers and others for assistance in how to perform diagnostic and other procedures, select correct therapeutic dosages for medications, and conduct other patient care activities. In that sense, it is more of a "how to do it" manual (for instance, how to perform a lumbar puncture) than a ''whether to do it" guideline document (that is, whether it is appropriate to perform the lumbar puncture). Nonetheless, its popularity reflects intense interest, across the medical education and practice spectrum, in easy-to-use references and guidelines-like materials. Several nonphysician professional associations are also engaged in guidelines development and related activities. Since the late 1960s, the American Nurses Association has developed standards of nursing practice and universal practice guidelines in consultation with specialty nursing organizations (ANA, 1990).5 The American Dental Association recently developed its first practice guideline on the general-initial dental examination; the group is not planning a guideline for 1992 but may return to the effort in 1993. Why Are Professional Societies Involved? It is not surprising that professional societies are involved in guidelines development. In the GAO survey cited earlier (1991b), the 27 responding 5 The American Nurses Association defines a standard as an "authoritative statement enunciated and promulgated by the profession by which the quality of practice, service, or education can be judged" (ANA, 1990, p. 4). It defines universal practice guidelines as "a process of client care management for nursing diagnoses with recommended interventions to accomplish desired client outcomes for a specific cluster of phenomena within a nursing specialty. These guidelines are established by research and/or professional consensus by practitioners in the specialty" (ANA, 1990, p. 5).
OCR for page 51
Guidelines for Clinical Practice: From Development to Use societies noted two primary reasons for involvement: (1) improving quality of care and (2) defending against outside forces. The second objective may cover efforts to reduce malpractice and its associated costs, to encourage greater uniformity in health insurance coverage and utilization review criteria, and to counter conflicting guidelines developed by other specialty societies. Traditionally, the purposes and missions of specialty societies and professional organizations, whether or not they are explicitly stated, include fostering the provision of appropriate clinical care based on professional standards. In this vein, the American College of Preventive Medicine (ACPM) sees its role as "leadership in research, professional education, development of public policy, and enhancement of standards for preventive medicine" (ACPM, 1989, p. 56). This is surely not an atypical view, and it is entirely consistent with involvement in guidelines development. The AMA argues that physician organizations need to be involved in guideline development "to ensure that practice parameters are properly developed and that quality improvement, rather than cost containment, serves as the foundation for their development" (Kelly and Swartwout, 1990, p. 54). Risk management—that is, the effort to lower or curb the number of events that might lead to malpractice litigation—is another possible goal of guidelines development. The American Society of Anesthesiology and the Risk Management Foundation at Harvard University have been particularly prominent in this arena (Holzer, 1990). Building on the work of anesthesiologists from the Harvard University-affiliated hospitals, the ASA recently developed "stricter standards intended for risk management activities, focusing on clinical practices that give rise to malpractice claims" (Pierce, 1990, p. 61). The lowering of malpractice insurance rates for physicians who agree to follow these anesthesia standards has attracted much interest among other specialty societies, although anesthesiologists themselves have noted that not all types of clinical practice lend themselves to this approach to risk management.6 As discussed in Chapter 5, policy makers have also become interested in guidelines as one element in malpractice law reform. In addition to professional societies, malpractice underwriters and self-insured physicians groups have themselves developed guidelines for practice, an effort that is sometimes characterized as a form of risk management (Holzer, 1990; Pierce, 1990).7 For instance, the Doctors' Company in Santa Monica, California, produces for its member physicians a series called "Risk 6 The anesthesia guideline that had the greatest influence in malpractice terms was a simple recommendation to use continuous oxygen saturation monitoring. The development of transcutaneous oxygen monitors played a significant role in the success of this guideline, which antedated the enormous increase in attention to guidelines at the end of the 1980s. 7 For a broader discussion of risk management in this context, see Morlock and colleagues (1989) and Kapp (1990), as well as Chapter 5 of this volume.
OCR for page 52
Guidelines for Clinical Practice: From Development to Use Management Guidelines," which provides recommendations on diagnostic workups, treatment, clinical management, and practice management. One guideline on the evaluation of patients with chest pain, for example, makes recommendations about the taking of an appropriate medical history, the focus of the physical examination, and the proper use and interpretation of electrocardiograms. Finally, some professional societies have tried to integrate concerns about quality with concerns about appropriate payment for physician services. For example, an explicit goal of the ACP Clinical Efficacy Assessment Project is to produce sound definitions of good medical practice that can contribute to a rational system of payment for medical care (White and Ball, 1990). Although in the past some parties have criticized the ACP for its explicit cooperation with third-party payers, the ACP maintains that "with the involvement of the professional societies, guidelines for payment have a far better chance to reflect appropriate medical practice" (White and Ball, 1990, p. 52). This broad policy perspective is also reflected in the college's recent decision to create a Center for Applied Research that will expand the ACP's involvement in both guidelines and outcomes research. (Appendix A includes an excerpt on the use of erythrocyte sedimentation rates from the organization's compendium Common Diagnostic Tests.) PUBLIC AGENCIES Government support for practice guidelines can serve two sets of aims: the broad goal of promoting the public health and welfare and the narrower ones of improving the quality and controlling the cost of government-funded health care programs. The activities described in the next section fall into both categories. For the most part, guidelines development has been the domain of federal rather than state government, as described below.8 The federal government has been involved in guidelines development in at least three distinct ways: 8 States do get involved, however. The state of California, for example, requires patients who may undergo blood transfusion to receive a standardized explanation of risks, benefits, and options approved by the state Department of Health Services (an excerpt is included in Appendix A). "Guidelines" relating to AIDS prevention, tracing, reporting, and the like have been promulgated by virtually every state in the union. Many if not most state guidelines appear to start with guidelines published by federal agencies. To the extent that experimentation with health care reform takes place at the state rather than the federal level, activities in the states may become more prominent in coming years. A 1990 conference sponsored by the California Public Employees Retirement System and the Oregon Medicaid initiative described in Chapter 6 reflect state interest in defining basic health benefits using guidelines and similar materials as one basis for decision making.
OCR for page 53
Guidelines for Clinical Practice: From Development to Use directly convening and managing groups to develop practice guidelines (or similar statements of good clinical practice); funding the development of guidelines by other groups; and funding and conducting basic and applied research to strengthen the clinical knowledge base and the methodologic tools that support better guideline development. U.S. Public Health Service The major part of federal activity now occurs in the several agencies of the U.S. Public Health Service (PHS)—especially AHCPR and the National Institutes of Health (NIH). In addition, the Office of the Assistant Secretary for Health, through its program on health promotion and disease prevention, has had a major role in guidelines relating to screening and prevention. Other PHS agencies with related activities include the Food and Drug Administration (FDA) and the Centers for Disease Control (CDC). Some guideline-related activities for the Medicare program are also carried out under the auspices of the Health Care Financing Administration (HCFA). Food and Drug Administration Perhaps the earliest government entry into an activity similar to guideline development came with the assignment to FDA of responsibilities for the assessment of drugs (1938) and medical devices (1976). FDA activities involve both more and less than guidelines development as defined in Chapter 1. The agency engages in both technology assessment and formal regulation, and it has a defined but limited process for securing information on the use of drugs after they have been approved for marketing (postmarketing surveillance). FDA typically does not formally approve off-label uses of drugs (uses for indications other than those for which initial marketing approval was granted),9 and the agency's assessments of safety and efficacy do not include comparisons with alternative therapies. National Institutes of Health A major example of the first government role cited above—direct government involvement in guidelines development or similar activities—is the Office of Medical Applications of Research (OMAR) at NIH, which has traditionally viewed its mission as knowledge building and dissemination. 9 In a departure from its usual practices, FDA reviewed an ACP guideline on methotrexate for rheumatoid arthritis (then an off-label use of that pharmaceutical agent) that had been published in the Annals of Internal Medicine, and approved this use about six months later.
OCR for page 54
Guidelines for Clinical Practice: From Development to Use The NIH Consensus Development Conference Program, administered by OMAR in the Office of the NIH Director, was established in 1977 (Mullan and Jacoby, 1985; Perry, 1987, 1988; IOM, 1990d,g). Through a fairly stylized approach involving small expert panels and invited conference participants, the program develops what it calls consensus statements that it hopes will be useful to health care providers and the public alike. These statements are not meant to be a primary source of data or detailed technical information; rather, the aim is to produce a document that will reflect the commonly held views of an expert panel that grasps the issues and examines relevant scientific information. Thus, consensus statements are expected to "help resolve the issue at hand, advance medical practice, and provide a clear, concise message for clinicians and the public" (OMAR, 1988, p. 2). A notable facet of the NIH consensus development effort has been its wide-ranging dissemination efforts. Although some observers have asserted that the consensus statements are not guidelines (Jacoby, 1985), at least some OMAR-sponsored statements meet the IOM definition of guidelines ("systematically developed statements to guide practitioners and patients"). For example, the statement on the use of intravenous immunoglobulin (IVIG) discusses the safety, risks, effectiveness, and recommended regimens of IVIG for various clinical conditions and immunodeficiencies (NIH Consensus Conference, 1990). Other parts of NIH also contribute to the guidelines scene. For example, the National Heart, Lung, and Blood Institute (NHLBI) has promulgated treatment guidelines in the cardiovascular field through its National Cholesterol Education Program for Adults and its National High Blood Pressure Education Program. Centers for Disease Control Another federal effort in direct development of guidelines is that of the Centers for Disease Control. Its approach tends to be decentralized; individual divisions develop guidelines using different procedures. Some use national committees, appointed by the Secretary of Health and Human Services; others appoint panels directly. By and large, however, guideline development involves a literature search followed by a consensus recommendation by the panel or working group (Steven Teutsch, Centers for Disease Control, personal communication, 1991). As examples of guidelines developed by CDC, those of the Immunizations Practices Advisory Committee are used by many state health organizations; they are published in the Mortality and Morbidity Weekly Report and as stand-alone documents. Another CDC publication, The Prevention and Treatment of Complications of Diabetes Mellitus, comes in a version for primary care practitioners and a version for patients.
OCR for page 55
Guidelines for Clinical Practice: From Development to Use Yet another example of the range of CDC publications is the annual report Health Information for International Travel from the CDC Division of Quarantine, which is updated biweekly with the "Summary of Health Information for International Travel" or "Blue Sheet." The scope of the publication is broad, covering such disparate topics as specific recommendations for vaccination and prophylaxis, geographical analysis of potential health hazards, motion sickness, cruise ship sanitation, and the possibility of anthrax contamination of goatskin products. (Appendix A contains an excerpt on vaccinations for pregnant women.) This publication is used by health departments, private practitioners, travel agencies, international airlines, and shipping companies. Agency for Health Care Policy and Research A hybrid public role—one that involves more than financial sponsorship but less than direct governmental promulgation of guidelines or review criteria—is exemplified by the role that the U.S. Congress mandated for the AHCPR Forum for Quality and Effectiveness in Health Care. The 1990 IOM report (IOM, 1990c) provides details on the early tasks assigned to AHCPR in the Omnibus Budget Reconciliation Act of 1989 (OBRA 89). The agency is to "arrange for"10 the development and periodic review and updating of "clinically relevant guidelines that may be used by physicians, educators, and health care practitioners to assist in determining how diseases, disorders, and other health conditions can most effectively and appropriately be prevented, diagnosed, treated, and managed clinically." (Appendix A includes illustrative excerpts from the first guideline developed through this program.) In addition, the Forum is also to arrange for the development of "standards of quality, performance measures, and medical review criteria through which health care providers and other appropriate entities may assess or review the provision of health care and assure the quality of such care." AHCPR has recently awarded a contract to the American Medical Review Research Center to translate three sets of guideline—on urinary incontinence, postsurgical pain management, and benign prostatic hypertrophy—into medical review criteria. Those criteria are to be applied by Medicare peer review organizations (PROs) to cases that have already been reviewed by PROs and are to be used in educational outreach programs. This project 10 As explained by one individual intimately involved in the development of this legislation, the phrase "arrange for" is a key indicator of the "extent to which the legislation was structured to create a public-private enterprise with respect to guideline development. The Forum develops no guidelines; guidelines are not to be federal creations" (Peter Budetti, George Washington University School of Medicine, personal communication, July 13, 1990).
OCR for page 56
Guidelines for Clinical Practice: From Development to Use is discussed further in Chapter 5.11 Thus, the Forum's broad objectives are to promote the development of instruments to assist clinical decision making and to evaluate the quality of that decision making and the resulting care. OBRA 89 provided an option by which AHCPR could contract with others for guidelines development; exercising this option would perhaps signal greater emphasis on funding (and less on internal development). In fact, AHCPR moved in this direction in September 1991, awarding three contracts for guidelines for otitis media in children, congestive heart failure, and poststroke rehabilitation. The contractors are, respectively, a consortium headed by the American Academy of Pediatrics with subcontracts to the American Academy of Family Practice, the American Academy of Otolaryngology, and the Children's Hospital of Pittsburgh; the RAND Corporation; and the Center for Health Economics Research with a subcontract to the Harvard School of Public Health for the literature review and analysis. The work for these contracts will be done in two phases. By the end of the first, 12-month phase the contractors are expected to produce a science-based, pilot-tested, peer-reviewed guidelines document; at that point AHCPR will determine whether the original contractor should continue with the second phase and translate the guidelines into medical review criteria and standards of quality. For both phases (phase 1 in parentheses), the funding ranges from just over $480,000 ($340,000) to just over $800,000 ($675,000). This approach will leave the agency out of direct involvement in certain guideline development tasks, such as the review of the literature and of the scientific evidence. The Forum does expect, however, to exercise considerable oversight of the work (in view of the fact that the funding is awarded as a contract rather than a grant), especially in such matters as formation of the panels, details of the work plans, and the literature reports. The principal investigators for the three projects, who will not be the panel chairs, have been asked to follow the general precepts for guidelines panels laid out in the agency's interim manual (Woolf, 1990a), although they need not adhere to every detail. AHCPR, through the Medical Treatment Effectiveness Program (MEDTEP) rather than the Forum, also sponsors a dozen or more Patient Outcomes Research Teams (PORTs), as authorized by Congress in 1989 (AHCPR, 1990).12 A considerable fraction of AHCPR's annual budget (about $1 11 In the early 1970s, AHCPR's predecessor, the National Center for Health Services Research and Development, sponsored the Experimental Medical Care Review Organization (EMCRO) program. Some EMCROs promulgated and acted on clear criteria for appropriate and inappropriate health care services. The efforts of the New Mexico EMCRO are probably the best known (Brook and Williams, 1976; Brook et al., 1978). 12 A small unit in AHCPR-the Office of Health Technology Assessment (OHTA)-is responsible for responding to requests from HCFA for "technology assessments," chiefly of new
OCR for page 57
Guidelines for Clinical Practice: From Development to Use million per project per year) is devoted to these multiyear, multisite, and multidisciplinary research projects, which have several effectiveness and patient outcome objectives; guideline development is not an explicit or primary task. Their basic aim is to study the effectiveness of all (reasonable) approaches to care for patients with a specific clinical condition. Among the conditions being studied are those with large variations in clinical practices and outcomes, a criterion also used to set priorities for guidelines development. Each project includes a literature review and synthesis, an analysis of variations in medical practice and patient outcomes, a planned method of targeted dissemination of its findings concerning optimal approaches to patient management, and an evaluation of the impact of this dissemination effort. The latter two activities may involve the development of practice guidelines, although AHCPR and the projects do not necessarily label them as such (AHCPR, 1991). As with the guidelines effort itself, however, expectations that the PORTs (even at the collective levels of funding they now enjoy) will contribute to clear cost savings or significantly improved health care practices—at least in the short run—must be kept realistically cautious. Other Public Agencies Health Care Financing Administration In the 1970s, HCFA-funded researchers associated with Boston University began to perform conceptual and methodological work on ways to measure the appropriateness of hospital use. This effort eventually led to the Appropriateness Evaluation Protocol and other tools that are still widely used in public and private quality assurance and utilization review programs (Gertman and Restuccia, 1981; Payne, 1987). In the 1980s, HCFA also helped fund the RAND Corporation's development of appropriateness criteria. These types of indicators, which have become one of the benchmark categories of medical review criteria, are discussed more fully in later sections of this chapter. More recently, HCFA has also supported guideline development as part of its contracts with Medicare PROs. PROs are private physician-directed organizations funded by HCFA to perform a very specific scope of work as or emerging technologies for which the other agencies must make insurance coverage (e.g., Medicare program) decisions. One aim of the OHTA is to amass information about the safety, efficacy, effectiveness, and other characteristics of certain technologies. However, to the degree that the OHTA literature review and conclusions constitute evidence or statements about appropriate (or inappropriate) use of a technology, they might be regarded as at least kin to guidelines being developed elsewhere. (The OHTA memo regarding reimbursement recommendations [yes or no] is not made public.)
OCR for page 58
Guidelines for Clinical Practice: From Development to Use defined nationally by the agency. One PRO responsibility has been to develop or adapt various kinds of guidelines or criteria for prospective (preadmission and preprocedure) utilization review efforts.13 In keeping with the legislative emphasis on "local" (regional or state) peer review, and in the face of a dearth of accepted national guidelines, the PROs generally have had and have exercised considerable discretion to create, adopt, or adapt review criteria. This has led to substantial state-to-state variation in review criteria as well as to criteria that are not based on much, if any, systematic analysis of the literature and that may be quite minimalist and liberal. All of these factors have played a role in the criticism leveled at past PRO utilization review criteria (Project Hope, 1987; IOM, 1990i [see especially vol. 2]; Kellie and Kelly, 1991; see also Chapter 5 of this report). One way in which AHCPR and the PROs will be working together was noted earlier—the recently initiated project to develop medical review criteria. Another area in which the PRO program and AHCPR might link efforts is in the evolution of HCFA's so-called Uniform Clinical Data Set (UCDS; Krakauer, 1990; Krakauer and Bailey, 1991). The UCDS, which collects 1,600 data elements (typically about 250 to 400 per case), has between 3,000 and 4,000 algorithms governing how those clinical data from hospital inpatient records should be abstracted for various PRO review and data base purposes. It also includes 300 or more algorithms to identify potential cases of substandard care that may require review by a PRO physician advisor. Guidelines from AHCPR panels or other sources might be used to update and upgrade the UCDS algorithms, which date to the late 1980s. Whether such collaboration comes to pass, however, is a question for the future. U.S. Preventive Services Task Force Another example of a public-sector initiative involving substantial private leadership and participation is the U.S. Preventive Services Task Force (USPSTF). The initial objective of this four-year undertaking was "developing recommendations for clinicians on the appropriate use of preventive interventions, based on a systematic review of the evidence on clinical effectiveness" (USPSTF, 1989, p. xxi). The 20-person Task Force, commissioned in 1984 by the Department of Health and Human Services (DHHS), developed guidelines on 169 interventions and involved dozens of advisers, authors of background papers, and reviewers. (The guideline on screening 13 The fourth scope of work for the Medicare PROs was released as this report was being prepared. Over objections from the PRO community, it proposed to abolish the prior-authorization tasks that the PROs have had for the past several years.
OCR for page 59
Guidelines for Clinical Practice: From Development to Use for low visual acuity in children appears in Appendix A.) The Task Force report states that "[r]ecommendations for or against performing these maneuvers should not be interpreted as standards of care but rather as statements regarding the quality of the supporting scientific evidence" (p. vii).14 Congressional Office of Technology Assessment This discussion has focused on the federal executive branch. In the legislative branch, the Office of Technology Assessment (OTA) of the U.S. Congress conducts many studies related to health care technologies; their assessments are largely in the form of reviews of published literature and publicly available data. For instance, in 1990, OTA released Preventive Health Services for Medicare Beneficiaries: Policy and Research Issues, which compiles recommendations published by other groups concerning such topics as periodic health examinations. It also cites assessments of preventive services (e.g., screening for breast cancer, for other types of diseases such as cervical cancer or glaucoma, and for abnormally high levels of cholesterol; vaccines against pneumococcal pneumonia and certain types of influenza) that OTA has itself conducted. The findings and conclusions of these assessments are not guidelines in the typical sense of the term, but they may certainly be regarded as a form of guidance for appropriate use of services. PRIVATE RESEARCH, PAYER, PROVIDER, AND OTHER GROUPS Numerous other kinds of organizations have had some involvement in guidelines development. They include private research organizations, academic medical centers, staff- and group-model HMOs, hospitals and hospital systems, health associations, and payment-related organizations. The following discussion focuses on the relatively few such organizations that (1) have devoted considerable resources to systematic development of guidelines and (2) have reached out to a national audience with these guidelines. For the most part, the guidelines development activities of most hospitals, HMOs, 14 The USPSTF used methods similar to those of the Canadian Task Force on the Periodic Health Examination, which was established in 1976 and issued its first monograph on 78 target conditions in 1979 (Canadian Task Force on the Periodic Health Examination, 1979). The Canadian group reconvened following that work and issued periodic updates in 1984, 1986, and 1988. The Canadian Task Force, following earlier work by Frame and Carlson (1975), analyzes and grades research on different services using a clinical-epidemiological approach. It has made, and continues to update, recommendations to the government that are used by the Canadian provinces to make decisions about the services to be included in the periodic health examination.
OCR for page 60
Guidelines for Clinical Practice: From Development to Use payers, and similar organizations are internally focused and not publicly available. The research organization best known for its involvement in work on appropriate medical care is the RAND Corporation. Its Health Services Utilization Study began in the early 1980s with funding from a variety of sources: the Commonwealth Fund, the John A. Hartford Foundation, the Pew Memorial Trust, the Robert Wood Johnson Foundation, and HCFA (Chassin et al., 1986b). The project devised a formal (modified Delphi) consensus approach to the development of detailed ''appropriateness criteria" (ratings of the appropriateness of up to several thousand separate indications for a given diagnostic or therapeutic procedure). The researchers created such indications for six medical and surgical procedures and conducted an extensive investigation of how variations in the per-capita rates of use of these procedures related to variations in their appropriateness for specific clinical problems; among the procedures studied were coronary artery bypass surgery and carotid endarterectomy (Winslow et al., 1988a, 1988b). (Appendix A contains an excerpt from RAND's publication on the latter procedure.) Appropriateness criteria can be distinguished from guidelines (as understood in this report) because the criteria were not primarily designed to assist physician and patient decision making; rather, they were to help evaluate the appropriateness of clinical decisions. RAND researchers have continued the appropriateness criteria endeavor in various ways (including projects overseas).15 The approach has been adopted by at least one for-profit utilization management group (Value Health Sciences) that has contracts with a variety of private insurers and other health plans. In 1990, the AMA, the RAND Corporation, and the Academic Medical Center Consortium (AMCC) signed a memorandum of agreement to cooperate in an initiative to develop appropriateness criteria and to convert them into practice guidelines (parameters) for everyday use by physicians. The original notion was that the AMCC and RAND would be responsible for conducting research to develop the criteria and for using them to evaluate cases at the individual medical centers. The AMA, through its Specialty Society Partnership and Practice Parameters Forum, would use the research results to facilitate the development and dissemination of practice guidelines. The four procedures that have been under study (and the sites of the principal investigators and administrators) are coronary artery bypass surgery (RAND); cataract surgery (University of Iowa); aortic aneurysm resection (Mayo Clinic); and carotid endarterectomy (Duke University). To date, the Blue Cross and Blue Shield Association (BCBSA) is the 15 However, as a contractor to AHCPR to develop guidelines on congestive heart failure, RAND will be employing different methods that are more in line with those followed by current AHCPR panels (David Hadorn, RAND Corporation, personal communication, October 9, 1991).
OCR for page 61
Guidelines for Clinical Practice: From Development to Use primary, and perhaps only, example of a private insurer that has invested considerable resources in both public guidelines aimed at a national physician audience and proprietary review criteria and other tools for use by local plans (Morris, 1987). Like government efforts, the BCBSA effort has proceeded on two fronts: primarily through funding of other organizations (what it calls a "catalyst role") and secondarily through direct BCBSA guidelines development. The former effort is based on several premises: subscribers (and all patients) are best served by affecting clinical practice positively through provider education rather than through retrospective review and possible claims denial; national medical organizations are a leading and appropriate source of guidance; and physicians will more readily accept such guidance when it is developed and provided by their representative medical organizations than when it is provided by an insurer (Morris, 1987).16 BCBSA has also created the Technology Evaluation and Coverage program to provide proprietary information to Blue Cross and Blue Shield plans to use in making benefit coverage determinations. Another interesting initiative is that being developed in Minnesota by more than 50 health care institutions and other organizations (Borbas et al., 1990; Catherine Borbas, Health care Education and Research Foundation, personal communication, December 31, 1991). Known as the Minnesota Clinical Comparison and Assessment Project (MCCAP), the project has developed guidelines on five conditions and procedures of interest to specific specialties. MCCAP "consensus panels" have convened to draft guidelines, which were then reviewed, revised, and disseminated to affected physicians. Data on physician performance are being collected, analyzed, and compared to the initial guidelines. MCCAP plans to rely as much as possible in future efforts on guidelines developed by national organizations, and it has already revised its locally developed guidelines to reflect and be consistent with new work by national specialty societies. It has gone beyond these national guidelines by developing outcome measures consistent with its data collection and analysis objectives. Among the efforts of individual staff- and group-model HMOs in developing guidelines for their health care professionals, those of the Harvard Community Health Plan (HCHP) have received considerable attention (Gottlieb et al., 1990; Burda, 1991b). The HCHP Clinical Guidelines Program has invested substantial resources in the development of scientifically based clinical algorithms. Although a central objective of the program has been to develop algorithms to guide HCHP clinicians, it also has a major re 16 BCBSA independently considers the payment implications of the guidelines, but the transformation of guidelines into medical review criteria, if recommended, is undertaken primarily at the level of local plans, not at the central association level. Blue Shield of California has been particularly active in developing medical policies and has involved the public as well as professionals.
OCR for page 62
Guidelines for Clinical Practice: From Development to Use search and educational component aimed at teaching a national audience, through publications and training activities, how to develop and implement clinical algorithms. In particular, the program has attended to practical issues regarding guideline use in ambulatory care and in "continuous improvement" strategies for clinical care management. In addition, other HMO systems, such as U.S. Health Care, United Health Care, and Kaiser are involved in initiatives that include some role for guidelines. Some may begin to make such guidelines publicly available. Finally, other types of private initiatives include various kinds of commercially sponsored publications. For example, Merck Sharp & Dohme Research Laboratories produces The Merck Manual, an extensive compendium of information about diagnosis and treatment (Berkow, 1982). Begun nearly a century ago, it has been designed to meet the information needs of medical students, practitioners, and other health professionals on a wide range of medical disorders and patient complaints and concerns. Although the information (contained in volumes that exceed 2,500 pages) is not presented as guidelines per se, the manual supplies commentary on the use, interpretation, and limitations of many common procedures and tests used in diagnosis and patient care management. Other such compendia are the Physicians' Desk Reference (which compiles, indexes, and cross-references information on prescription and nonprescription pharmaceuticals) and Scientific American Medicine (loose-leaf medical and surgical reference volumes that are periodically updated). COSTS OF GUIDELINES DEVELOPMENT Because evaluation of guidelines development eventually must consider the results of the process in relation to development and implementation costs, the committee tried to estimate the costs incurred in developing guidelines. This exercise turned out to be quite difficult, for three major reasons. First, many organizations had no cost estimates; the work was buried in the budgets of one or more organizational units. Second, some who had tried to estimate costs found a full description of costs so time-consuming that they abandoned the effort. Third, those cost estimates that were available were generally not comparable or comprehensive. In some cases, guideline developers can cite direct costs for such items as travel, printing, meeting expenses, consultants, and other line items.17 Few organizations, however, can report costs for staff support, general overhead, 17 For example, the cost of developing appropriateness criteria, evidently counting staff time as well as other direct costs (including nominal honoraria to physician participants) but excluding related research activities, has been estimated at between $250,000 and $500,000 for each set of procedure-specific indicators.
OCR for page 63
Guidelines for Clinical Practice: From Development to Use and the value of volunteer efforts. For example, neither the USPSTF nor OMAR/NIH have good estimates of the total cost of their work to develop specific guidelines (Robinson, 1991; Steven Woolf, U.S. Public Health Service, personal communication, March 27, 1991). John Ferguson, the director of OMAR, estimates that each consensus conference costs about $82,000 in direct costs, but that estimate does not include staff or volunteer participant time (John Ferguson, National Institutes of Health, personal communication, 1991). AHCPR expected to allocate $2 million in 1990 and $3 million in 1991 to guideline development. By the end of calendar year 1991, the agency probably will have three finished guidelines and half a dozen more at various stages of development. Exclusive of staff time, early costs for AHCPR panels appeared to run about $200,000 to $250,000 per panel; more recent estimates put the range between $350,000 and $800,000, depending on the complexity of the topic. The main variable in costs appears to be the number of questions the panel eventually elects to tackle and for which it must review the literature. AHCPR has been tracking some specific costs, such as those entailed in literature reviews, and has found considerable variation across its panels. Some of the variation in these direct costs is attributable to simple differences in the volume of clinical research from topic to topic; some of the variation also appears to be ascribable to differences in panel strategy, frugality, or ingenuity. For example, the depression guideline panel opted not to focus on a single type of depression but to look at a range of diagnostic categories within that broad rubric; the panel thus will produce a "family" of guidelines based on what may eventually total more than 90,000 abstracts from the clinical and research literatures. The GAO survey of medical societies (1991b) found that the cost estimates provided by different groups varied substantially. The estimates, which excluded volunteer time, ranged from $5,000 to $130,000 per guideline or set of guidelines. One society estimated the value of volunteer time over a two-year period at more than $500,000. A recent directory of guidelines initiatives reported estimated costs per guideline from a few thousand dollars to more than $1 million (Robinson, 1991). SUMMARY Planning for successful implementation of guidelines begins with development. This chapter has described selected guidelines development efforts in both the public and private sectors as a means of highlighting how the enterprise is evolving. The pluralistic nature of guidelines becomes quite clear, thereby underscoring the utility of a firm understanding on the part of developers of the attributes of their processes and their eventual products. The diversity of developers, of the topics that are addressed, and
OCR for page 64
Guidelines for Clinical Practice: From Development to Use of the guidelines that are produced are reflected especially in the methods used and in the costs; generally, costs for producing authoritative guidelines are higher than many experts originally anticipate, owing in large measure to the attention directed to a definitive review and analysis of the literature. This report now moves on to issues of implementation. It returns, however, in Chapter 7 to reconsider development in the light of what has been learned about implementation and surrounding policy issues.
Representative terms from entire chapter: