3
Implementing Guidelines: Overview and Illustrative Cases

It is getting to be harder to "run" a constitution than to frame one.

Woodrow Wilson, 1887

Difficult as it is to formulate guidelines, implementation is an even greater challenge. Viewed generally, implementation refers to the concrete activities and interventions undertaken to turn policies into desired results. In the context of guidelines, two overlapping but distinct implementation tasks can be distinguished. One is implementing a public or private program to develop and promote practice guidelines.1 The other is implementing the guidelines themselves.

Implementation in this second sense involves the programs and activities that take guidelines out of the rather abstract phase of development and into the actual world of health care decision making and action. This chapter provides an overview of implementation issues. Chapters 4 and 5 describe some specific programs and activities, focusing on those related to information systems, educational programs, quality assurance, health benefits management, and risk management and medical liability.

Guidelines implementation is a much more diffuse process than guidelines development. The time horizon extends from the near to the indefinite future; the number of involved parties multiplies; responsibilities blur; var-

1  

The Omnibus Budget Reconciliation Act of 1989 gave primary responsibility for establishing a public program to develop and promote practice guidelines to the Department of Health and Human Services, through the Agency for Health Care Policy and Research and its Forum for Quality and Effectiveness in Health Care. Necessary steps for implementing this program include hiring staff, developing a program agenda, letting contracts, convening and assisting expert panels, establishing an advisory council for AHCPR, and generally establishing and administering a broad, ongoing federal program.



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



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

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

OCR for page 65
Guidelines for Clinical Practice: From Development to Use 3 Implementing Guidelines: Overview and Illustrative Cases It is getting to be harder to "run" a constitution than to frame one. Woodrow Wilson, 1887 Difficult as it is to formulate guidelines, implementation is an even greater challenge. Viewed generally, implementation refers to the concrete activities and interventions undertaken to turn policies into desired results. In the context of guidelines, two overlapping but distinct implementation tasks can be distinguished. One is implementing a public or private program to develop and promote practice guidelines.1 The other is implementing the guidelines themselves. Implementation in this second sense involves the programs and activities that take guidelines out of the rather abstract phase of development and into the actual world of health care decision making and action. This chapter provides an overview of implementation issues. Chapters 4 and 5 describe some specific programs and activities, focusing on those related to information systems, educational programs, quality assurance, health benefits management, and risk management and medical liability. Guidelines implementation is a much more diffuse process than guidelines development. The time horizon extends from the near to the indefinite future; the number of involved parties multiplies; responsibilities blur; var- 1   The Omnibus Budget Reconciliation Act of 1989 gave primary responsibility for establishing a public program to develop and promote practice guidelines to the Department of Health and Human Services, through the Agency for Health Care Policy and Research and its Forum for Quality and Effectiveness in Health Care. Necessary steps for implementing this program include hiring staff, developing a program agenda, letting contracts, convening and assisting expert panels, establishing an advisory council for AHCPR, and generally establishing and administering a broad, ongoing federal program.

OCR for page 65
Guidelines for Clinical Practice: From Development to Use ied local circumstances and priorities complicate decision making and generate conflicting incentives; and actions become more difficult to track. These conditions make it hard to specify attributes of good implementation processes in the way that the first IOM report specified the attributes of good guidelines. They also help explain why efforts to implement guidelines and related kinds of recommendations have met with limited success to date (Eisenberg, 1986; Schroeder, 1987; Lomas et al., 1989, 1991). GUIDELINES AND THE REAL WORLD One challenging current reality surrounding the implementation of guidelines is that many potential users are either unaware of guidelines or view them as being of marginal utility in their day-to-day work. Hardly unique is the urban community hospital staff who responded to the committee's inquiry about a site visit to discuss the use of guidelines by saying that there was really nothing related to guidelines going on at their institution. For practitioners, guidelines are just one element in a range of practical and interpersonal challenges of patient care and practice management. Further, the perceived salience of formal guidelines may be lessened by the likely tendency of clinicians to consider accepted, internalized guidelines as something other than guidelines; for example, schedules for well-baby care may be so deeply ingrained that they are simply no longer regarded as guidelines. Some practitioners may well resist guidelines as threats to their autonomy even when the source is a professional organization. For senior health care executives, too, other issues come first: patients, staff, payers, suppliers, competitors, institutional survival-although not necessarily in that order. When these executives frame a vision of their institutions for the future and a management strategy to achieve that vision, guidelines are not likely to appear in mission statements and five-year plans. For patients and their families, guidelines are even more remote. Few laypersons will know about formal efforts to develop clinical practice guidelines; even fewer will know of their initial products or be able to use them directly. The focus of guidelines for patients, therefore, is likely to be educational—for example, handbooks or brochures about proper care for a given ailment, appropriate preventive regimens, or when to seek professional health care and when to manage one's own care. Keeping an overview of implementation reasonably compact but illuminating is difficult, given the scope and variety of implementation efforts, on the one hand, and the lack of systematic literature about the topic, on the other. Compared with the development of guidelines, implementation is not only harder to do but more difficult to describe and analyze. The complexity of the implementation task just in terms of potential users may be illustrated (perhaps overdramatically) by estimated numbers of those who may

OCR for page 65
Guidelines for Clinical Practice: From Development to Use be involved in some aspect of implementation for some category of guidelines. In the United States alone, there are roughly 250 million patients and potential patients, differing in myriad ways; 500,000 physicians, 1.5 million nurses, and 160,000 dentists distributed across a large number of specialties and subspecialties; 6,000 hospitals and thousands of nursing homes, rehabilitation centers, and other health care institutions; faculties and students in 120 medical schools, 1,400 nursing education programs, 50 dental schools, and dozens of other training programs; 600 health maintenance organizations and independent practice associations, and hundreds of preferred provider networks, utilization management organizations, and similar entities; 54 Medicare peer review organizations with hundreds of involved peer reviewers; 6,000 to 10,000 attorneys specializing in health-related issues—even more if personal injury lawyers are included; 1,800 medical libraries; and untold numbers of state, federal, and private health care regulators or administrators as well as technical, lay, and clinical publications with direct or indirect educational purposes or intentions.2 To convey something of the realities of implementation in the absence of much documented description and analysis of actual experience, the committee has devised what it calls "synthetic case studies." They draw on the study's site visits, the diverse experience of the committee members and staff, the limited research and descriptive literature, and conversations with many individuals and groups. Each of the six case studies presented in the next section is a mix of these sources, and none depicts any single organization or individual. The subjects of the case studies were developed to illustrate the perspectives and environment of individuals and organizations, practitioners and patients, and primary, secondary, and tertiary care settings. The cases are intentionally simplified portraits designed to convey some, but by no means all, of the real and practical issues in the effective use of guidelines. They are not intended to portray uniformly flawless application of impeccable guidelines to achieve specifically desired results, nor can they provide the depth of description and analysis possible with true case studies. (The latter would undoubtedly be helpful and might be considered for funding by the government and other organizations interested in guide- 2   Obviously, these are duplicated counts. Clinicians, for instance, can also be patients, "preferred providers." and faculty. No single set of guidelines is likely to involve all these parties in significant ways.

OCR for page 65
Guidelines for Clinical Practice: From Development to Use lines.) These composite cases attempt to show that guidelines can be more (or less) than adequate in meeting the needs of practitioners and patients, that intended users of guidelines vary in their willingness and ability to conform to guidelines, and that systems and incentives differ in the degree to which they support the application of guidelines. Nonetheless, the perceived relevance of guidelines is probably higher than "average" for the individuals and organizations represented in these case studies. Two of these hypothetical case studies focus on ambulatory (office-based) care from the points of view of a physician and an administrator; two focus on inpatient hospital care (one large academic center, one small community hospital); one deals with nursing home and hospice care; and the final case study takes the point of view of a patient. Each case is preceded by several key words; these are intended as quick references to particular implementation issues that are raised by the case. CASE STUDIES Case Study 1: Small Internal Medicine Practice KEY WORDS: patient needs, characteristics, and preferences; conflicts between perceived patient needs and guidelines; specificity and format of guidelines; utility of computer-based information and decision support systems; time constraints; hassle factor Dr. Marcus practices in a typical setting: a small (in this case, five-person) fee-for-service internal medicine group in a middle-class suburb. Although she does not see guidelines as a major issue in her practice, Dr. Marcus can cite a number of them that she uses (for example, those related to preventive services, infectious disease, and pharmaceuticals). For more complex clinical problems, she thinks guidelines that include algorithms and flowcharts are the forms most likely to be precise enough to guide decisions. When talking about guidelines, Dr. Marcus is adamant that the patient comes first—not the guideline. For example, one of her patients—a typical older patient—has high blood pressure, diabetes, asthma, a family history of stroke, and possible macular degeneration. Although guidelines for managing some of these conditions are precise, comprehensive, and relevant, most do not deal with the particular combination of clinical problems and preferences presented by this patient. It is clear to Dr. Marcus that judgment and experience, not arbitrary compliance with guidelines, are what this patient needs. Dr. Marcus points out that the patient comes first not only in making judgments about what to do but in implementing those judgments. This is particularly true for office-based care because the patient, not the physician, has to carry out many steps in a specific course of treatment. Dr. Marcus and her partners audited selected preventive services for patients in their practice and learned that only 50 to 70 percent of their patients had, in fact, received the services that were recommended. In some

OCR for page 65
Guidelines for Clinical Practice: From Development to Use cases, the patients simply did not accept the recommendations; in others, they failed to make return visits or to see referral physicians, despite follow-up calls and notes. For example, Dr. Marcus has for several years urged one of her elderly patients to get a screening mammogram. When the patient was asked why she had not done so, she said she thought the examination would be painful but was finally planning to have one. The reason for her change of heart was that her neighbor had recently had a mammogram that had revealed a small cancerous tumor. Dr. Marcus's patient left the office with a mammography order form, specific advice about where she could obtain the mammogram in the building or elsewhere, and a friendly, concerned parting comment from Dr. Marcus that she would be looking for the radiologist's report. Dr. Marcus estimates the probability that she will eventually get such a report (that is, that this patient will have the mammogram) as not much better than 50/50. Although Dr. Marcus sees both the value and the limitations of existing guidelines for physicians and patients, she is highly critical of the medical review criteria applied by third-party payers. She faces multiple, detailed, and often conflicting review criteria from different organizations. In fact, she often does not even know what criteria are being used. In addition, the language associated with Medicare and private insurer policies, which tends to be aimed at identifying "bad apples," is demoralizing, as is the "hassle factor" that arises in complying with the policies. The burden of paperwork and telephone calls is quite heavy, and Dr. Marcus's group has had to hire an extra person to help handle this workload in addition to the four clerical/data entry/ support staff needed for patient records, receptionist duties, and so on. Dr. Marcus appreciates that some review organizations have made an effort to minimize the burdens on physician offices and to employ clinically knowledgeable reviewers and clinically respectable review criteria; unfortunately, other organizations are less well managed. Dr. Marcus and her partners are atypical in that their group uses computers not only for administrative purposes but also for keeping patient clinical records and for alerting physicians to the need for certain follow-up tests and other activities. The system this group uses relies on software that was first developed in the mid-1970s; it was installed in this practice in the late 1970s. More guidelines for preventive services, diagnostic tests, and other topics could be programmed into this system's alerts and reminders, but Dr. Marcus says that priorities must first be set, because the number of services that could be provided with some marginal probability of benefit was virtually uncountable. Although the practice has been committed to the creative use of computers, it has not yet invested in any on-line clinical information services. The partners feel that available systems do not allow quick enough reference to practice guidelines (while the patient is waiting) in an algorithmic or similarly accessible format. Currently, computers are most useful to the practice in organizing and retrieving information about individual patients. Dr. Marcus still prefers well-known, hard-copy publications (such as the Physicians' Desk Reference) to most available on-line information systems. One of her partners does make limited, off-line use of some new software that helps him calculate objective risk assessments using models and formulas that he could not begin to keep in his head. At this time, however, these assessments are being applied only to a few patients.

OCR for page 65
Guidelines for Clinical Practice: From Development to Use Case Study 2: Managed Care Organization KEY WORDS: local development and adaptation of guidelines; cost management; coverage policy; selective contracting; utilization and quality review; information feedback; patient education and incentives; sanctions Columbia Care, or CC for short, is a 150,000-member health plan. It selectively contracts with physicians and hospitals and pays them on a negotiated basis that involves elements of capitation, per-case payment, discounts, and other payment methods. Its members must accept higher cost sharing if they use physicians or hospitals outside CC's panel of providers. Dr. Potter, the plan's medical director, describes Columbia Care as an eager ''market" for guidelines and review criteria. The plan uses medical review criteria and standards of quality in a variety of ways to influence clinical practice and to make decisions about what services to cover. When professional societies and other "suppliers" of guidelines and review criteria do not meet the plan's needs, Dr. Potter organizes expert panels and consultants to develop guidelines and review criteria for specific purposes. For example, no completely acceptable guidelines were available to advise physicians on how best to diagnose and treat adolescent depression and related mental and emotional conditions. In particular, little consistent advice was available about when to refer marginally symptomatic patients to psychiatrists or clinical psychologists, or about when to manage patients in ambulatory versus inpatient settings. The issue was significant for three reasons: the membership comprises mainly families, so the plan covers a considerable number of adolescents; adolescent mental disorders are rising in prevalence; and the employers with which CC contracts were becoming alarmed at the proportion of expenditures for these conditions. Consequently, Dr. Potter, with the assistance of experts at a nearby research firm, empaneled a group of clinicians (psychiatrists, psychiatric social workers, clinical psychologists, internists, and pediatricians) to examine the existing literature and to reach some consensus on appropriate indications for (1) referral from primary care to specialty care and (2) inpatient treatment. The guidelines will be implemented in several ways. First, CC will disseminate the guidelines (including a description of the development process and participants) through its monthly newsletter to practitioners and hospitals. Second, the inpatient indicators will be enforced (except for emergencies) through a preadmission review program. Third, the plan may require prospective member physicians or hospitals to agree to abide by the guidelines as a requirement for selection. Fourth, Dr. Potter hopes to get internal funds to evaluate whether referral and admission patterns change and whether expenditures are reduced. He has no plans at the moment, however, to monitor patient outcomes, although this is part of the organization's longer term planning. Dr. Potter sees advantages in local practitioner involvement in guideline development. Even if local conditions may not require any changes from national guidelines, local development or adaptation work—participating physicians getting together and "chasing some rabbits," as Kentucky-raised Dr. Potter describes it—can be useful. If nothing more, it makes practitioners more comfortable with guidelines. Dr. Potter also says that the plan is willing to "shoot a few pigeons" (that is, sanction deviant practitio-

OCR for page 65
Guidelines for Clinical Practice: From Development to Use ners) occasionally to show that guidelines are to be taken seriously. In general, however, the plan prefers not to take a punitive approach but to use guidelines as part of an information feedback process and as screens for selecting participating physicians. Because appropriate use of established procedures and services, such as tonsillectomy and hysterectomy, often depends on a variety of patient-specific circumstances, Dr. Potter oversees an array of programs to review the appropriateness of care on a case-by-case basis. This process is contractually permitted under a provision that limits payment to care that is medically necessary. It includes prospective, concurrent, and retrospective review of care. The programs use medical review criteria based on both practice guidelines and statistical norms to screen, for example, hospital length of stay. For preprocedure review, the plan uses a sophisticated system developed by a private company. That system is, in turn, based on detailed appropriateness criteria developed by a well-known research organization. Preprocedure review is viewed primarily as a cost-management tool that can also serve quality assurance objectives by deterring potentially harmful overuse or misuse of care. Dr. Potter insists that the actual review of individual care is based purely on clinical judgments, not cost. For the most part, Columbia Care relies for hospital quality assurance on the processes required by the Joint Commission on Accreditation of Health care Organizations (JCAHO). It has also instituted a quality assurance program for its primary care physicians that, among other features, employs selected practice guidelines. Each year, the plan reviews a sample of medical records for each of its primary care groups to check conformance with two to five guidelines. Last year it considered the percentage of female patients over the age of 49 who were recommended for and who received mammography screening. In addition, the quality assurance program has adapted process-of-care criteria for the outpatient management of adult hypertension, both acute and chronic otitis media among children, and evaluation of abnormal uterine bleeding. These criteria have been distributed to participating physicians, who agree to periodic audits of their charts. The physicians receive reports on how their performance compares with that of their peers—for example, what percentage of the time they use an office-based procedure (endometrial sampling) versus an inpatient procedure (dilatation and curettage) to evaluate abnormal uterine bleeding. Whenever Dr. Potter sends these reports, he asks for suggestions about how the reports and criteria might be improved. As part of its marketing and patient service strategy, Columbia Care provides health education and health promotion services (e.g., hotlines, brochures) that employ various clinical practice guidelines. Some employer-customers want to set up financial rewards (or penalties) for employees who pass (or fail) blood, urine, and other tests related to cholesterol levels, blood sugar levels, blood pressure, weight, and smoking. Dr. Potter has argued against this approach, believing that such programs are too intrusive a way of encouraging patient conformity to guidelines. Instead, he is working with employers to improve employer-based health education programs and to strengthen CC's own efforts, including patient-specific counseling by CC practitioners.

OCR for page 65
Guidelines for Clinical Practice: From Development to Use Case Study 3: Academic Medical Center Hospital KEY WORDS: quality of care and continuous quality improvement; computer-based information and decision support systems; local adaptation of guidelines; practice variation; behavior change; economic incentives University Medical Center Hospital (UMCH) is one of the relatively small number of sophisticated academic medical centers that have complex tertiary care cases, major clinical training responsibilities, and extensive clinical and health services research agendas. UMCH is also atypical in that it has officially adopted continuous quality improvement (CQI) as a management strategy. Dr. Pierce, vice president for quality improvement, has primary responsibility for overseeing implementation of the CQI initiative and has active support and reinforcement from the hospital's governing board and chief executive officer. Consistent with CQI precepts, UMCH seeks to reduce unwarranted variations in key diagnostic and therapeutic processes and to instill a sense of practitioner and staff ownership of the processes and results. Outcomes management is a key strategic aim, and Dr. Pierce cites the argument that one cannot properly understand, measure, and manage patient outcomes until processes of care have been stabilized. UMCH devotes considerable effort to identifying variations in care, determining their sources, and devising remedies where appropriate. Some of the variations can be traced to administrative problems. For instance, reports of radiological examinations performed on Saturdays and Sundays reach the appropriate physicians more slowly than examinations done on weekdays (and thus delay possible Sunday or Monday morning discharges). The reason is that, for months, the hospital has had fewer radiology technicians than it needs on weekends, in part because of indecision about salary increases, which has slowed advertising and recruitment efforts. On the clinical front, the hospital has been concerned about variations in physician practices. A particularly troublesome area has been blood transfusions: transfusion rates have varied from 0 to 70 percent of patients in some heart surgery categories. Efforts were made to determine outcomes associated with different practices and to identify clinical and administrative problems (e.g., inadequate blood retrieval procedures during surgery, long operating times, lack of explicit guidelines for blood use) that created or permitted unwarranted variations in care. Following the feedback of information on variations in blood retrieval rates and operating times and the establishment of guidelines for appropriate use of transfusions (including informed patient decision making), the higher transfusion rates have begun to drop with no measured adverse effects on patients. Although UMCH has invested substantial levels of resources in identifying administrative problems that lead to errors and practice variations, the institution has also devoted considerable effort to feeding back statistical information to physicians in ways that encourage more consistent practice without relying on punitive measures or embarrassment. As a case in point, when information showing the distribution of specific physician practices (e.g., days following major surgery before a patient is ambulated, preferences for certain kinds of antimicrobial agents, lengths of stay for particular conditions) is presented at medical staff meetings, the results for each physician are known to that physician but are blinded for the remainder of the group.

OCR for page 65
Guidelines for Clinical Practice: From Development to Use Sometimes the feedback is linked to practice guidelines ("benchmarking"), but at other times only the statistical information is provided. Dr. Pierce expects to operate this feedback system indefinitely because he knows that practitioners often revert to old patterns of behavior once feedback is stopped. He also believes that clear, quickly available feedback on actual practice variations will change behavior more predictably, and sooner, than will simple dissemination of published guidelines. In the feedback process for physicians, Dr. Pierce relies primarily on the structure offered by the medical staff meeting rather than on that of the hospital's quality assurance (QA) system. This decision is based on his observation that physicians look to their peers for guidance rather than to the administrative units of the hospital. The QA department is oriented mainly toward utilization review, risk management, and quality issues related to accreditation by JCAHO and external regulation by the state's Medicare peer review organization. UMCH management views the future as an era of additional cost pressures on physicians; it believes that these pressures will take the form of resource-based relative value fee scales, selective contracting, and shared financial risk in the context of managed care systems. In recent years, Dr. Pierce has seen physicians become much more interested in information that will help them build a record of practice quality and efficiency that will attract invitations to participate in managed care plans and similar networks. One of the keys to UMCH's quality improvement program is its very sophisticated computer-based information and decision support system, which not only provides a great deal of institutional data but also integrates a variety of practice guidelines in different formats. In describing UMCH's emphasis on this system, Dr. Pierce says that, as a practical matter, no one could work at the institution "without using the keyboard," although long-range plans are to introduce less cumbersome data-entry technologies, such as voice-recognition systems. The current system's clinical applications are related to several features: the computer-based patient record; an integrated data base that permits timely aggregation of and access to essential patient data; pharmacy alerts for possible drug interactions; alerts about questionable or "red-flag" laboratory test results (such as out-of-range electrolyte levels); on-line quality assurance including automated screening, problem alerts, and reports; and clinical decision support involving various kinds of protocols and other tools. For example, ordering of parenteral hyperalimentation is controlled by a protocol that includes both default recommendations and restricted opportunity to depart from the protocol. For blood transfusion orders, the institution has—for the time being—made a deliberate choice against on-line constraints; specifically, the system does not at present reject physician orders when indications for possible transfusion are equivocal, but it does alert the physician to that information. A few years ago, UMCH purchased protocols for certain kinds of hospital care from another institution; consistent with CQI philosophy, they were intended to provide benchmarks for performance. The medical staff did not accept these imported protocols, however, and they were subsequently abandoned. Dr. Pierce says the effort was probably unsuccessful because the protocols were externally created and were not accompanied by any explanation or rationale that could be evaluated by the UMCH staff. Now, protocols and guidelines from other organizations are consulted,

OCR for page 65
Guidelines for Clinical Practice: From Development to Use but they are not adopted without the medical staff's explicit evaluation or, in some cases, modification and adaptation. Case Study 4: Community Hospital KEY WORDS: risk management and medical liability; management decision making and follow-through; provider payment incentives; management of patient care; guideline content Memorial Hospital is a 250-bed hospital serving a community of 58,000 and a large rural area around it. The hospital was facing substantial increases in its malpractice premiums, especially those related to anesthesiology. In response to this problem, Dr. Houlihan, a surgeon and chief of the medical staff, undertook with his anesthesiologist colleagues to learn more about the guidelines of the American Society for Anesthesiology (ASA) for different aspects of anesthesia care. Having determined that these guidelines were being widely accepted throughout the state's hospital sector and were viewed favorably by malpractice insurers, Dr. Houlihan lobbied for their adoption to improve quality of care and to qualify the hospital and its anesthesiologists for lower malpractice premiums. Although in the past many of the medical staff had expressed a distinct lack of enthusiasm for practice guidelines, in this case they unanimously agreed to adopt the anesthesiology guidelines. They also appointed a subcommittee to work with hospital management on practical matters related to the selection of equipment and the establishment of institutional procedures and training programs. Simultaneously, the chairman of the board of the hospital appointed a subcommittee to undertake a special fund-raising effort to secure the funds necessary to quickly purchase the equipment needed to apply the anesthesiology guidelines. (The hospital administrator, however, would rather have used the money to acquire a magnetic resonance imaging machine because a competitor hospital in a nearby community had just ordered one.) Memorial Hospital has a traditional utilization review and quality assurance department that bases most of its activities on retrospective reviews of samples of patient charts, which are judged against common "generic screens." However, Dr. Houlihan prevailed on the QA coordinator to institute a special, concurrent study of all adverse events related to the surgery and anesthesiology departments. Using before-and-after comparisons, they hope to be able to demonstrate a drop in the number and rate of adverse anesthesiology-related events. One of the hospital's QA analysts doubts that they will find enough events over a reasonable period of time to permit valid comparisons-but agrees it is worth a try. As the anesthesiology efforts were gearing up, Dr. Houlihan and Ms. Johns, vice president for nursing, began to discuss how the hospital might develop or use practice guidelines of other sorts. One incentive for doing so is that Memorial Hospital is located in an "all-payer" state, meaning that hospital payments are regulated by a state agency that uses per-case reimbursements to hospitals, a method similar to Medicare's diagnosis-related groups. Dr. Houlihan and Ms. Johns hypothesized that "national" guidelines might be helpful in streamlining the care the hospital provides so that they do better financially under the state's payment scheme. Through the nursing literature that she follows carefully, Ms. Johns began to

OCR for page 65
Guidelines for Clinical Practice: From Development to Use learn more about various kinds of guidelines, including "critical pathways"—documents, often in the form of a matrix, that identify what major elements of care (e.g., ambulation, discontinuation of intravenous fluids or medications) should occur on which day of hospitalization to prepare patients for timely discharge. Development of these pathways does not appear to involve an explicit assessment of scientific evidence about appropriate care; rather, existing practice patterns (in this case, for both medical and nursing care) seemed to be the predominant source of information. Further, as Ms. Johns discovered, both consulting firms and other, similarly sized hospitals are also developing such pathways on their own. To pursue this activity at Memorial Hospital, Dr. Houlihan and Ms. Johns have gone to the hospital's board of trustees. They saw this step as necessary because the pathways will be a new endeavor for medical and nursing staff at the hospital and because they believed that they needed the political backing of the board. A further reason for going to the board was that the acquisition of literature and similar materials would involve costs beyond those normally budgeted. Dr. Houlihan and Ms. Johns hope to start the clinical pathways effort coincidentally with the start of the hospital's next fiscal year. Case Study 5: Nursing Home and Hospice KEY WORDS: limited resources; regulation and interpretation of guidelines; local adaptation of guidelines The Mapletown Home is a long-term care facility in a large metropolitan area. It also serves as an inpatient care unit for a local hospice program. It is a private, nonprofit, nonsectarian organization affiliated with University Hospital Medical Center (case study 3). About two-thirds of the home's patients are on Medicaid. The home has a long history and a considerable reputation in the area; it thus enjoys more-than average support from philanthropic sources. Despite the home's modest endowment and other private resources, Dr. Blake, the medical director, emphasizes how government reimbursement levels have directly limited patient care and patient choices. Furthermore, she notes, those levels have a somewhat pernicious indirect effect: once a "minimum" standard of care is set for a particular clinical problem, it quickly becomes a "ceiling" as well, and the institution's board of directors tends to question additional care or staffing levels based on patient need. For example, to prevent pressure sores, the apparent options indicated by current research are a specially designed mattress, a specially designed bed, or frequent turning and nursing care. Regardless of what would be best for the patient, only the cheapest alternative—the mattress—is really possible. Dr. Blake stressed that exceptions to this and other similar policies that are made for the benefit of a particular individual bring "palpable and present trade-offs" in the care of the other residents. Dr. Blake sees government as virtually incapable of treating guidelines as anything but rigidly applied regulations, noting that the nuances or variations that might be acceptable to guideline developers simply get lost in the insistence on conforming to rules. She cites, for example, regulations that severely circumscribe the use of physical and chemical restraints, which were promulgated following documented evi-

OCR for page 65
Guidelines for Clinical Practice: From Development to Use dence of their widespread overuse. Although in complete agreement with the thrust of the new guidelines, Dr. Blake also points out that physical restraints may be required for some patients (for instance, those who are severely disturbed or demented) to permit use of a feeding tube or to prevent injury to themselves or to other patients. Yet the regulations, as presently interpreted and enforced by government surveyors, limit the use of restraints to such an extreme that for some patients the home has to choose between running the very great risk that they will hurt themselves (or others) and discharging them. Dr. Blake is distressed that more flexibility is not possible in taking account of the needs of all resident patients. Dr. Blake described efforts at the Mapletown Home to develop guidelines—in this case, physician orders for pain control—adapted to the very special circumstances of hospice inpatients. The guidelines were developed empirically after observing the effects of different doses and combinations of medications on many different hospice patients. Compared with what the Physicians' Desk Reference indicates is appropriate for most clinical situations, these guidelines allow a much larger than usual range of pain-killing medications and dosages to prevent recurrent manageable pain. Once orders are initialed by a physician, they can be implemented within the specified ranges as needed by the hospice nursing staff. Physicians are not required to accept or use the guidelines, but if they do not, hospice nursing staff must telephone them to obtain authorization for each change in medication, dosage, timing, or route. Physicians thus have an incentive to accept the guideline, although most appear to have fully supported it from the beginning. Case Study 6: Patient KEY WORDS: human errors, incentives and disincentives, conflicting guidelines Joan Chapman is 41 years of age and has no family history of breast cancer or other risk factors for this disease. She calls her gynecologist to make her yearly appointment for a breast and pelvic examination. When the receptionist learns that Ms. Chapman has not had a mammogram in the past two years (not since a baseline mammogram when she was 38), the receptionist asks that she schedule a mammogram so that the results can be sent before the physical examination. Ms. Chapman arrives for her appointment to find that her gynecologist is ill but that his partner, Dr. Frank, can see her if she wishes. She agrees. After the history and physical examination, Dr. Frank is ready to leave the examining room and end the appointment when Ms. Chapman asks about the results of the mammogram. It becomes clear that Dr. Frank had not checked the file for the radiologist's report. The report describes a suspicious spot in the left breast and recommends that a second mammogram be done in six months—even though the radiologist thinks that the spot is probably just a lymph node. Ms. Chapman asks if this is really necessary. Dr. Frank responds that the practice is purchasing a unit during the next year and suggests that she simply obtain the second mammogram during her next annual visit to their offices. At this point, Ms. Chapman decides to seek advice from the internist she is scheduled to see for another problem. He strongly recommends that the six-month follow-up films be obtained and urges Ms. Chapman to ask the radiologist to make

OCR for page 65
Guidelines for Clinical Practice: From Development to Use sure the suspicious spot is not a large mole that he noted during his physical examination. Ms. Chapman has the second mammogram, remembering to mention the mole. No one, however, remembers to obtain the baseline mammogram for purposes of comparison. While still at the radiology facility, Ms. Chapman is told that the findings are normal. Three days later she gets a call from her internist saying that the results indicate the need for follow-up biopsy. The internist, after hearing the conflicting information that has been given directly to his patient, checks with the radiologist's office and discovers that he has been sent an incorrect summary report. He calls Ms. Chapman back, straightens out the final results, and recommends that she get another routine mammogram in two years. Motivated by this series of events, Ms. Chapman begins to read about mammography screening. She learns that different medical organizations have different guidelines for women in her age group and that many groups no longer recommend a baseline mammogram. She goes on to read some of the background literature and literature reviews on which the guidelines were based. She understands that the evidence of benefit for screening in her age group is weak and contradictory; she further grasps that some evidence suggests that cancers grow faster in younger women, so that a two-year (or even a one-year) screening interval may not be as useful for younger women as it is for older women in whom cancers grow more slowly. Ms. Chapman does not know what to think about when to get another screening mammogram. Her indecision is a function of the conflicting guidelines, the apparently careless and possibly acquisitive behavior of Dr. Frank, the confusion surrounding the results of the follow-up mammogram, and the time and money involved. She also is somewhat annoyed that neither her gynecologist nor her internist mentioned any conflicting recommendations about screening for women her age, and she wonders about other kinds of screening tests she has always taken for granted. She is not prepared to complain to any of the involved physicians about what happened, but she now plans to see a gynecologist who a friend says is willing to explain options and discuss concerns. GENERAL ISSUES IN IMPLEMENTATION Factors Influencing the Effective Use of Guidelines As these synthetic case studies illustrate, the attributes and context of specific guidelines, the characteristics of practitioners and patients, and other factors interact to influence whether and how guidelines are used. Those who develop and implement guidelines need to anticipate how such factors may influence the willingness and ability of individuals and organizations to make effective use of guidelines. Lomas and Haynes (1988) identify five crucial classes of factors: (1) patients and families, (2) practitioners, (3) provider institutions, (4) economics (as it affects practitioners and institutions), and (5) the environment. Interacting with these factors are characteristics of the guidelines themselves, as described in Chapter 2 and elsewhere.

OCR for page 65
Guidelines for Clinical Practice: From Development to Use Patient and family factors include most obviously the patient's health status and particular clinical problems, as well as the expectations, preferences, and knowledge about health care of the patient and his or her family. Age, gender, race or ethnicity, reading skills, income, residence, and similar demographic, social, and economic factors may also influence whether and how guidelines affect health care decisions, behaviors, and outcomes. In the case study of the internal medicine practice, Dr. Marcus's elderly patient clearly feared mammography as painful and just as clearly was more impressed by her neighbor's personal experience than by her doctor's advice. In the nursing home case study, Dr. Blake sees the federal regulations barring physical and chemical constraints as rigid and insensitive to the trade-off (for a family, even if the patient is not aware of it) between a disturbed patient being physically restrained and that patient having to be discharged because of the danger he or she may pose to others. Ms. Chapman, the patient in the last case study, may be atypical in having the educational background and willingness to invest in her own examination of clinical research and practice guidelines related to mammography; she is probably not atypical in her reluctance to complain directly to her physician about some aspect of her care. Practitioner factors also encompass such demographic and social characteristics as age, gender, and residence. Furthermore, the attitudes of practitioners toward the value of guidelines may be affected by the site and type of their professional training, their specialty affiliation, their association with academic medical centers, and the kinds of continuing education to which they are exposed. Other relevant variables include the type, size, and setting of their practices. The physicians in the first and third case studies are representative of a cohort of practitioners who are comfortable with computers. In the community hospital case, Dr. Houlihan's experience as a surgeon and chief of the medical staff has most likely suggested to him the need for careful groundwork with the medical staff and the hospital board as part of his implementation strategy. At the academic center, Dr. Pierce had a lesson in medical staff sensitivity during his unsuccessful effort to import protocols from another institution. Drs. Marcus, Potter, and Pierce all reacted negatively to guidelines that were inadequate in scope, precision, or rationale. Institutional factors include both cultural characteristics, such as management philosophy, objectives, and style of individual institutions, and their operational capacities, including the basic physical plant, equipment, personnel, information and monitoring systems and technologies, and quality-of-care and peer review structures and processes. Relationships with public or private multi-institutional systems may also affect how guidelines are identified and used. In the case studies, the clinicians at University Hospital Medical Center, Memorial Hospital, and Mapletown Home work in worlds that offer

OCR for page 65
Guidelines for Clinical Practice: From Development to Use quite different opportunities for the use of many kinds of practice guidelines. These opportunities are affected by the respective institutional missions and philosophies, governing structures, medical staff and other personnel, information systems, and other variables. At Columbia Care, Dr. Potter tries, for the most part, to emphasize cooperation and positive incentives and to make only limited explicit use of the organization's more regulatory or negative tools. Economic factors that may affect potential users of guidelines include extent of insurance coverage, methods and levels of institutional or practitioner payment (e.g., fee for service, per case, capitated), other financial incentives or disincentives (e.g., for referral to specialist consultants), and provider ownership of related services (e.g., clinical laboratories, diagnostic testing facilities). In the patient-centered case study, Ms. Chapman suspected, rightly or wrongly, that the gynecologist she saw was unduly motivated by economic considerations in recommending that she wait to have a follow-up mammogram until his practice had purchased the necessary equipment. On the other hand, Dr. Pierce at the community hospital and Dr. Potter at the managed care organization clearly view the financial incentives associated with managed care as supporting their efforts to implement guidelines. Dr. Blake just as clearly perceives the constrained economic environment in which Mapletown Home must operate as promoting a narrow application of guidelines and insensitivity to patient needs. Although generally receptive to clinical practice guidelines, Dr. Marcus has few kind words for the way they are used, misused, or not used by third-party payers. Environmental factors include the prevalence and incidence of disease and illness, the composition and capability of the overall system of health care delivery, government regulations, the medical liability system, and the nature and extent of social consensus about matters affecting health care decisions and behaviors. In the nursing home setting, exposure to detailed federal regulation and inspection in a resource-constrained and sometimes bleak environment has evidently left Dr. Blake suspicious that guidelines are vulnerable to public hysteria and regulatory intemperance. These kinds of public and private oversight—that is, regulation and inspection—may increase conformity to some kinds of guidelines but at the (perhaps considerable) cost of irritation and hostility. Although only touched on in the Memorial Hospital case, the medical liability system is clearly a noteworthy environmental variable. Strategies to Encourage Effective Use of Guidelines These six hypothetical cases barely begin to illustrate the diversity of strategies that might be used to encourage the effective use of guidelines and the considerations that might influence the choices made from among

OCR for page 65
Guidelines for Clinical Practice: From Development to Use these strategies. The following discussion covers some of these considerations. First, the particular short- and long-term objectives to be served by the implementation process will make some approaches more relevant than others. For example, an implementation strategy that is appropriate for quickly informing clinicians in private practice (like Dr. Marcus) about newly revised guidelines (as might be prompted by strong new research findings and expert consensus) is unlikely to be appropriate if the objective is to provide comprehensive, continuously accessible information to office-based physicians over the longer term. Further, implementation schemes that are' suitable for either of the preceding situations are not likely to be suitable (at least not alone) for implementing guidelines that are intended to persuade patients (like those Dr. Marcus sees) to ask for and accept recommendations about important screening services or changes in lifestyle or personal habits. Second, decision makers must assess the expected effectiveness of alternative strategies in achieving the objectives in question. As a case in point, if Columbia Care's objective is to reduce the rate of unnecessary repeat cesarean sections consistent with authoritative guidelines, then Dr. Potter is likely to consider the potential impact of several alternative ways of encouraging (and in some cases, enforcing) practitioner conformity with these guidelines. These methods might include provision of written information, education sessions using professional opinion leaders, feedback of comparative information on individual practice patterns, application of some form of utilization review, or perhaps reduction in the payment differential for vaginal delivery versus cesarean section. Lomas and colleagues (1991) tested the first three of these approaches and found the use of opinion leaders to be the most effective. Third, the expected benefits of a particular strategy to promote the use of practice guidelines have to be weighed against predicted costs and available resources. Consider again an organization that is developing sets of guidelines for practitioners in private, office-based practice. To bring its guidelines to their attention, the organization might consider press conferences, direct mailings of announcements of guidelines or of the guidelines themselves, development of hard-copy, desk-top compendia of the guidelines (for example, something like the American Academy of Pediatrics' widely known ''Red Book" [1991] on infectious diseases), and developmental support for computer-based interactive software. The costs of these and other kinds of implementation activities, singly or in combination, must be laid against both the likely results and the available resources. Fourth, in assessing alternative means to encourage the use of practice guidelines, implementers must consider the demands made on target users by different strategies and user receptivity and capacity to change. For example, an organization considering implementation of a computer-based

OCR for page 65
Guidelines for Clinical Practice: From Development to Use decision support system should take into account the benefits of user-friendly software and voice-recognition systems, which demand less of practitioners and patients than hard-to-use programs and keyboard entry systems. Recalling the first case study, Dr. Marcus found hard-copy guidelines more attractive than available on-line information systems. Similarly, the willingness to move on anesthesiology guidelines at Memorial Hospital despite general skepticism about guidelines illustrates the mixed picture of receptivity and resistance that doubtless prevails in the majority of settings. A fifth issue is the manageability of the tasks for administrators or others responsible for implementing a strategy. Administrators face considerable challenges in putting into place viable programs for introducing and using guidelines and for evaluating their impact. Setting up and maintaining a system of financial incentives, for instance, involves both similar and different management tasks than installing and maintaining an information feedback system. The perceptions of Dr. Houlihan and Ms. Johns that they might need the backing of the hospital board, and the support of Dr. Pierce by the UMCH board and chief executive officer, highlight the financial and administrative exigencies of guidelines implementation. Selecting the particular elements of an implementation plan plainly requires that these and other variables be carefully assessed. Inevitably, tradeoffs will be required among some factors such as expected effectiveness and cost or manageability. In theory, these trade-offs may appear to be straightforward and easy to analyze; in the real world, they are unlikely to be so amenable to investigation or understanding. Little immediately relevant empirical research is available to guide decisions. The next chapters look further at supportive conditions for guideline implementation and at some organizational, legal, and policy aspects of implementation efforts. This context includes educational activities and information systems as well as structures and processes to assess and assure quality of care, to manage health care costs, and to reduce medical liability. SUMMARY This chapter has provided an overview of some factors that affect the implementation of clinical practice guidelines. Hypothetical case studies, based on situations encountered in study site visits, literature reviews, and other sources, illustrate the many ways guidelines may be employed and some real-world factors that may encourage or hamper guidelines implementation. Among these factors are variables that relate mainly to patients—for instance, factors that motivate patients to follow or ignore professional advice (or guidelines) and activities relating to patient education and incentives. Also important are variables relating to practitioner behavior and

OCR for page 65
Guidelines for Clinical Practice: From Development to Use decision making in everyday practice: conflicts between perceived patient needs and guidelines, the hassle factor that confronts physicians in implementing guidelines, financial incentives and disincentives facing physicians in private practice, and the exigencies of day-by-day management of patient care. Several important elements cut across settings of care: administrative decision making and follow-through; collaboration (or lack of it) across the main departments of an institutional provider; the role of top management; quality of care, quality assurance, and continuous quality improvement; risk management and liability; computer-based information and decision support systems; a myriad of elements relating to the local development, adaptation, and implementation of guidelines; and simple human error. External factors include the existence of conflicting guidelines, insurance benefit plans and coverage policies, requirements concerning preprocedure review, limited institutional or community resources, and local, state, and federal regulation. Working models of the successful use of guidelines are not now abundant. Thus, opportunities to learn from one's peers, so common in other areas of health care management, seem to be rather scarce. This means that implementation is, in some respects, a challenge that must be met de novo by each health care organization. Until more practical experience with guidelines is available, discussions of implementation will necessarily be somewhat theoretical. The next two chapters consider how educational activities, information systems, and efforts to manage quality, costs, and liability may—in principle—support and be supported by guidelines for clinical practice.