Interdisciplinary Approaches and a Broader View of Health

Interdisciplinary education occurs when “faculty learn, work and teach together” (Gelmon, 1996, p. 218) to prepare students to work as a team driven by the health needs of patients and the goal of providing the services necessary to improve health to the extent possible (Bulger, 2000; Gelmon, 1996). Interdisciplinary education involves more than simply defining the roles of various clinicians (Osterweis, 2001). Health professionals that are well prepared for practice in the 21st century will collaborate across departments and disciplines, and even settings of care, to meet patients' needs.

The term “interdisciplinary” as used here refers to the involvement of different disciplines, such as medicine, nursing, and pharmacy; the term is not used to denote different specialties within a single discipline, such as internal medicine, cardiology, and endocrinology. The notion of interdisciplinary education will assume increasing significance in the future. For example, the needs of people with chronic conditions (who, as noted in Chapter 2, represent a growing proportion of the population) cannot be met by any single health professional. Similarly, applying the latest biomedical advances will increasingly require the expertise of specialized health professionals, such as genetic counselors. Additionally, if patients are expected to be more accountable for maintaining their health and to assume responsibility for self-care in managing chronic conditions, they also need to be recognized as a key member of the health care team. Yet team interactions in practice often fall short of expectations, in part as a result of current approaches in clinical education that emphasize hierarchy, individual decision making, and the organization of work around professional roles rather than patient needs (Institute of Medicine, 2001b). Indeed, the implementation of more interdisciplinary educational approaches will require a level of cooperation that has rarely been demonstrated. As one observer notes, interdisciplinary training is a “goal often espoused but rarely pursued” (LeRoy, 1994, p.337).

As suggested above, clinical education in the 21st century will also need to take a broader view of medicine and health, with greater emphasis on understanding the social, behavioral, cultural, and environmental factors that influence health and disease in addition to understanding the biological basis of disease (LeRoy, 1994; Josiah Macy, Jr. Foundation, 1999; Young and Coffman, 1998). Developing this understanding will in turn require that biomedical science be better integrated with a patient- and population-based approach that addresses the determinants of disease and health, and places greater emphasis on prevention (LeRoy, 1994) and the identification of risk factors and how to mitigate them.

The focus on the biomedical basis of disease that characterizes the current model for clinical education assumes that ill health is fully explained by disease, so that the core of medical science is the diagnosis and treatment of disease (Cassell, 1999). American medicine, however, is being asked to move beyond this model to address issues related to population health, resource allocation, new means for caring for chronic disease, and the management of health information, all areas in which physicians have traditionally not been trained (Schneider and Eisenberg, 1998). Medical schools in particular are believed to produce physicians well equipped to deal with specific organ systems or pathologies, but ill equipped to deal with the behavioral causes of chronic diseases or the social context of illness (Cantor et al., 1993). According to one survey of young physicians, fewer than half reported receiving excellent or good preparation in coordinating patient care with community services, providing cost-effective care, or managing the needs of the frail elderly (Cantor et al., 1993). Nursing tends to be more oriented toward health promotion and disease prevention. Advanced-practice nurses in particular are focused on establishing knowledge partnerships with their patients, educating them about their conditions, and engaging them in illness prevention and health promotion (Mundinger, 2002).

There are a number of barriers to conducting interdisciplinary education, including turf battles, academic credit, recognition of faculty, and scheduling (Gelmon, 1996; Osterweis, 2001). Each college, even each department, guards its own curriculum, and bringing different students together can be viewed as virtually impossible (Kaufman, 1999). The differing academic schedules of schools can also create a significant obstacle (Osterweis, 2001). Although 60 AHCs have identified an individual with responsibility for interdisciplinary education, only about a dozen have established significant activities in this area; most of the latter are public and community-based, have multiple health professional schools, and fall under the broad jurisdiction of an AHC leader (Osterweis, 2001).

Another potential barrier is that faculty may have neither the skills nor the incentives to pursue interdisciplinary approaches to education. Faculty who themselves have not been trained through interdisciplinary approaches may find it difficult to teach that way and be unable to undertake the educational innovations required to implement such approaches. Moreover, interdisciplinary education is not as strongly rewarded as the efforts of independent scientists working in their laboratories. A concern is that students are not being taught explicitly to work in interdisciplinary teams, but implicitly through the work environment (Conway-Welch, 2002; Larson, 2001), which often has not fostered the types of positive, constructive interactions desired across the disciplines. Strengthening efforts to improve the health of patients and populations will necessitate the development of new educational models.

Conducting rounds with students in multiple disciplines is one approach used for encouraging interdisciplinary interactions, but this approach becomes more difficult to implement as hospital stays shorten. It may be relatively easy to design interdisciplinary education for the classroom, but doing so becomes more difficult in a clinical setting, especially as training diversifies into nonhospital clinical sites. Interdisciplinary approaches also become more difficult to implement when attempted across settings of care. For example, there may be opportunities to foster interdisciplinary training between doctors and nurses in a hospital, but it is less clear how to bring public health into the training model. Some have recommended that public health training be incorporated into medical and nursing schools and that schools of medicine and nursing partner with schools of public health to develop interdisciplinary and joint programs (Institute of Medicine, 2003c). Examples of improved public health training for medical students can be found at Duke University, the University of California at San Francisco, and the University of Southern California (Institute of Medicine, 2003c).

Information Management

Health professionals will need to be prepared to manage information so they can deal with a constantly growing evidence base, serve as an information resource, support decision making by patients, and measure care so they can manage it effectively.

Technological and biomedical advances are expanding the evidence base for health and medical care exponentially. The number of clinical trials published in the literature grew from approximately 1,000 in 1966 to more than 10,000 in 1996, with half that growth experienced in more recent years (Chassin, 1998). This growth in information, which will only intensify in the future, will challenge traditional approaches to educating health professionals. Some have even suggested that the traditional emphasis on a core of knowledge is questionable in light of the expansiveness and dynamic nature of the science base (Weed and Weed, 1999). Rather than the traditional approach based on teaching facts, students should be prepared for the types of problem solving they will face in practice (Weed, 1981).

Health professionals will have to know how to obtain and manage new knowledge as it continually emerges. The concept of evidence-based practice is that a clinical problem is defined, and published evidence is obtained, appraised, synthesized, and applied to the problem (Welch and Lurie, 2000). However, there are virtually an unlimited number of clinical strategies, and resources for evaluation are limited. Educators need to teach the evidence where it is certain, and students need to learn to how to obtain and apply evidence as it develops, as well as how to make clinical decisions when the evidence is absent or weak (Welch and Lurie, 2000).

The increasing complexity of disease and expanding treatment options will require that health professionals be able to serve as an information resource for their patients. Health professionals will need to bridge the gap between the evidence base and patient knowledge, evaluating the evidence and turning it into information that can be explained to patients so their preferences can be expressed. They will need to synthesize, explain, and interpret information to support patient decisions and self-management. In some cases, the health professional's primary role may be serving as an information consultant and resource to guide and support decision making by more informed patients, rather than performing a clinical intervention. There is some evidence that patients whose informational needs are not adequately met are likely to make more visits and use more resources in their care (Mundinger, 2002). Indeed, some have suggested that this information role is one of the most important therapies provided to patients, with health professionals serving as coach and adviser to support patients' increased direction over their care (Schneider, 2002). This role should be incorporated into the education of all health professionals, but also reinforced through interdisciplinary training that recognizes the varying contributions different team members can make to a patient's care.

The increasing costs of care and concerns about the quality of care will result in growing demands to measure and manage care. The management of information must include a focus on measuring care so it can be continuously improved. Research, patient care, and therefore health professions education will become increasingly reliant on evaluative disciplines, such as clinical epidemiology, informatics, health services research, outcomes analysis, and value management (Detmer, 1997; Wennberg, 2002).

Managing information to the extent that will also be required in the future cannot be done without more-advanced information systems to acquire and manage the level of information that will be needed for practice. Part of delivering state-of-the-art care in the future will be the use of clinical and other information systems. Students will need to be prepared to use information technology as a more central component of health care. Clinical education programs that fail to incorporate state-of-the-art information systems into their training will be unable to prepare students for practice today, let alone tomorrow.

Nonhospital Training Experiences

To prepare health professionals to deliver care in the 21st century, education should correspond to care delivery. The majority of care is delivered to patients in noninpatient and nonhospital settings. Nearly a billion ambulatory visits were made in 1999, compared with 32 million hospitalizations (Eberhardt, 2001). Ambulatory care as discussed here refers not only to hospital outpatient departments, but also to offices, community health centers, managed care organizations, public health departments, long-term care facilities, and even patients' homes. Any location where care is delivered should be considered a potential training site.

The predominant model of education today, especially for physicians, consists of training in the inpatient setting, delivering tertiary care. The advantage of hospital-based training is that students can learn from the most challenging and difficult cases. Hospitals that see a larger volume of similar patients (e.g., cardiology or cancer patients) are also more likely to demonstrate higher-quality care in that field, which is desirable to teach (Institute of Medicine, 2001b). In addition, seeing patients who are admitted for ambulatory-sensitive conditions or for certain chronic conditions should give students an opportunity to learn what factors contributed to the condition so they can not only treat the symptoms but also consider how patients might be able to avoid such hospitalizations in the future. It is easier to conduct education in the inpatient setting because the acute problems seen are more readily specified, and therefore, the educational content is easier to define (Showstack, 1999). Finally, inpatient settings offer a cluster of faculty, other students, and an infrastructure to oversee the educational process.

The inpatient model for clinical education will be increasingly ineffective in the coming decades, however. The rate of hospital admissions has been declining; lengths of stay are becoming shorter; many diagnostic problems are being handled outside the hospital; patients in hospitals have the most complex conditions and therefore present a relatively narrow spectrum of diseases; and the sicker patients admitted require increasingly technical care (Kassirer, 1996; Goroll et al., 2001). These trends give the learner less time to establish a relationship with the patient and to understand the multiple medical, social, psychological, and other factors that affect not only the course of disease, but also the individual's health and well-being. A short hospital stay provides a poor learning opportunity to understand the influence of behavioral and social factors on health or to foster shared decision making (Ewan, 1985). Furthermore, most patients admitted electively to the hospital have been worked up prior to admission, so they arrive not only with a chief complaint, but also with the results of diagnostic and laboratory tests, and sometimes, a diagnosis. The intellectual challenge to the learner is incomplete, and the learning opportunity is affected.

Training in the inpatient setting, therefore, does not sufficiently prepare health professionals for practice or provide adequate exposure to alternative settings of care. A survey of young physicians revealed that more than half believed there was too little training in physician offices, organized care settings (e.g., health maintenance organizations), or long-term care facilities (Cantor et al., 1993). People with chronic illness that is managed effectively may often avoid hospitalization for the condition altogether. Even more care can be expected to move out of the inpatient setting as biomedical advances affect when an illness is identified and how it is treated. Finally, in the marketplace, there is a trend toward the provision of nonspecialized care in community hospitals and other settings; specialty care is becoming more concentrated in AHCs (The Commonwealth Fund Task Force on Academic Health Centers, 2000). As AHCs become relatively more focused on specialty care and caring for patients with specialized needs, they become less able to prepare health professionals for everyday practice. It has been estimated that, on average each month, less than 1 person in 1,000 is admitted to an AHC (Green et al., 2001).

There has been some progress in increasing the amount of training provided in ambulatory settings; however, the majority of ambulatory training remains within hospitals, and only a small proportion takes place in nonhospital settings. Primary care physicians can be expected to practice predominantly in nonhospital settings, and they undergo about two-thirds of their training in ambulatory settings; however, only about one-quarter of their training is provided in community settings and about one-tenth is in managed care settings (Brotherton et al., 2000). Among non–primary care residents, just over one-third of training is in ambulatory settings, but only about 6 percent is in community settings and about 6 percent in managed care settings (Brotherton et al., 2000). Furthermore, the proportion of training time in nonhospital ambulatory settings (community and managed care settings) showed a decline between 1997 and 1999—a trend in the wrong direction.

Among undergraduate medical education programs, teaching in outpatient settings in required clinical clerkships occupied one-third or more of the time in primary care program areas compared with one-quarter or less in non–primary care program areas (Barzansky and Etzel, 2001). On the other hand, between 1984 and 1994, the percentage of all medical students who participated in one or more clerkships increased from just under half to almost three-quarters, and the average number of weeks in ambulatory settings increased as well (The Commonwealth Fund Task Force on Academic Health Centers, 2002).

Baccalaureate nursing programs are also offering more opportunities for clinical training in noninstitutional community settings, including visiting nurse agencies, home care, schools, and hospices (National Advisory Council on Nurse Education and Practice, 1996). To a lesser extent, training is also provided in such settings as nursing centers, senior citizen centers, and homeless shelters.

Shifting training to ambulatory settings involves more than simply moving or adding training slots. Ambulatory settings will not provide a good learning environment without additional preparation. As a learning environment, they can be unpredictable in terms of the types of patients seen, limited in terms of continuity of care, and variable across sites (Irby, 1995). Short patient visits can make it difficult to provide the observation and feedback needed for teaching (Bowen and Irby, 2002). There is also concern that students in ambulatory settings may lose the conferences, faculty, and general educational surroundings offered by the institutional environment (Kassirer, 1996). Indeed, students rate the quality of their instruction in ambulatory settings lower than that in inpatient settings (The Commonwealth Fund Task Force on Academic Health Centers, 2002).

Oversight of Education Programs

It is estimated that more than 50 groups are involved in the oversight of undergraduate and graduate training programs in the health professions (Gelmon et al., 1999). Some of these groups are identified in Box 3-1. The list intentionally includes the accrediting group for programs in health administration. Although the focus of this chapter is on clinical education, the challenges described also face administrators in terms of both their own education and their support for reform efforts in clinical education. Furthermore, clinicians ought to have knowledge of administrative issues, so it is important to consider programs in health administration when looking at coordination across disciplines.

BOX 3-1

A Sample of Accrediting Organizations for Health Professions Education. Oversight of health professions training occurs through a combination of public and private regulatory activities. A variety of private agencies accredit the education programs for (more...)

Continuing education requirements are overseen by yet other groups. Accreditation of continuing medical education programs is offered by the Accreditation Council for Continuing Medical Education. The American Osteopathic Association has a separate council for continuing education. The American Nurses Credentialing Center of the American Nurses Association administers the association's credentialing programs, providing both accreditation of continuing education programs and certification for specialty nursing practice. Unlike undergraduate and graduate training, which have clearly defined requirements, requirements for continuing education vary within disciplines. Some physician specialty boards require continuing medical education hours to maintain specialty certification, whereas other have no such requirement (Federation of State Medical Boards, 2002).

The proliferation of oversight groups has serious implications for reforming the education of health professionals. There has been a tendency toward expansion in recent years as more specialties have been recognized, a phenomenon that tends to increase subspecialties (as they seek recognition) and extends the length and cost of training (LeRoy, 1994). However, the approaches proposed in this report are not discipline specific, but apply to everyone. As a result, it may be necessary to ask 50-plus groups to amend their standards. For example, achieving a goal such as interdisciplinary education would require not only that each group make changes, but also that the groups work together in making those changes. That is likely to be a time-consuming process, and it is not clear that there is a mechanism for the purpose.

Coordination across the continuum of education is also poor. Coordination of oversight of education has been called fragmented and duplicative (Gelmon, 1996). Responsibilities for undergraduate, graduate, and continuing education reside for the most part in separate organizations (Enarson and Burg, 1992). As a result, accreditation divorces residency training programs from professional schools (Hanft, 1988). Feedback loops between the levels of education could improve all. For example, if one of the purposes of continuing education is to supplement areas in which undergraduate or graduate training has been deficient (Waxman and Kimball, 1999), such information should be provided systematically to those education programs so the deficiencies can be addressed. The current fragmentation inhibits these interactions. A recent Institute of Medicine report (2003a) calls on all education oversight organizations (accrediting, licensing, and certifying bodies) to work together to revise their standards.

Faculty Development and Organization

The school and its faculty are the strongest influences on the design of curriculum and students' educational experiences. Accrediting groups define standards for the structure, performance, and/or functions of the schools, but do not prescribe specific courses or educational experiences. The latter is the responsibility of each school as it designs its own curriculum within the guidelines of the pertinent oversight bodies. Even with a bounty of standards, it is known that schools vary in terms of emphasis, resources, costs, size, centralized or decentralized curriculum, frequency of curricular change, and other factors.

Faculty are being asked to assume new duties in areas in which they may not be adequately prepared to teach the next generation. Faculty teaching today may themselves not have been trained in nonhospital settings, computer-based systems, or interdisciplinary approaches to care (Wilkerson and Irby, 1998). They may not have learned how to develop curricula, evaluate students, or manage educational programs (Gelmon, 1996). Most medical teaching occurs through one-on-one encounters between physician and patient, reflecting the comfort level and expertise of many faculty (Kaufman, 1999). As a result, faculty may be unsure about their own skills for implementing aspects of a new curriculum (Sachdeva, 2000). Being a knowledgeable clinician (or basic scientist) does not necessarily translate to being an effective teacher.

There are also concerns about the availability of faculty in terms of both supply and time. As noted earlier, teaching faculty are under pressure to see patients and conduct research, leaving little time for teaching (The Commonwealth Fund Task Force on Academic Health Centers, 2002; Ludmerer, 1999). Although this concern is often voiced about medical faculty, it has been suggested that as nursing practice plans develop, a similar pattern will ensue. Nursing faculty will also face constraints on time for teaching as the pressure to see patients and raise revenue increases (Conway-Welch, 2002). Furthermore, particularly in nursing, there are concerns about the adequacy of the supply of faculty (American Association of Colleges of Nursing, 2002b; Association of Academic Health Centers, 2002).

The pressures on faculty preparation and time are likely to intensify as training is expanded to encompass a range of sites. More faculty with more-varied backgrounds could enhance the educational experience for students but could also result in even greater variability in student training. Faculty development will be needed to ensure that the faculty available at all training sites are prepared to teach students effectively (Weed, 1981; Griner and Danoff, 2000). Some have suggested using a smaller, full-time faculty (Hanft, 1988), perhaps moving with the students rather than the clinicians in each site taking on faculty duties.

In medical schools, the decentralized structure of faculty with powerful department chairs is viewed as a force that can inhibit educational innovation (Cantor et al., 1991; Regan-Smith, 1998; Petersdorf and Turner, 1995). Faculty identify predominantly with their own department and focus on training in their own discipline, hindering a broad, integrated view of clinical education. The strong departmental structure can also make it difficult to incorporate broad-based education courses that are not departmentally defined; for example, population health or “evaluative” sciences, such as biostatistics or epidemiology. Some schools have moved toward a more centralized curriculum to overcome the problems of a departmentally organized model, and although improvements are seen in terms of curricular reform, they also tend to raise costs because of the increased time needed for faculty coordination (Reynolds et al., 1995).

Weak Evidence Base

The evidence base for clinical education is not as strong as it should be to support the reforms described in this chapter. Better information is needed on the effectiveness of various teaching approaches for clinicians, on how principles of adult education can be applied appropriately to clinical education, on what types of teaching technologies are most effective and under what circumstances, on the characteristics associated with high-quality clinical education, and on the cost of training various health professionals. Good quality measures in clinical education do not currently exist (Blumenthal and Bass, 2001).

The Cochrane Collaboration has been working for many years to develop the evidence base for clinical care, but there is no comparable resource for the evidence base in clinical education. When the Cochrane Collaboration attempted to conduct a systematic review of educational interventions for teaching evidence-based medicine, only one article was found that met their criteria for inclusion (Hatala and Guyatt, 2002).

Two relatively new groups are making such an effort at developing an evidence base. The Campbell Collaboration (formally established in 2000) prepares and maintains systematic reviews of the effects of social and education policies and practices (see www.campbellcollaboration.org). For example, a systematic review is being prepared for problem-based learning for health professionals (Davies and Boruch, 2001). The second group, Best Evidence Medical Education, is a collaboration of individuals and organizations committed to the dissemination of information to people involved in medical education; the production of systematic reviews of medical education; and the creation of a culture of best-evidence medical education among teachers, institutions, and national bodies (see www.bemecollaboration.org). The group has been meeting since 1999. Both groups are international, with a strong European representation.

There is also a lack of information on the actual cost of education programs and its relationship to the quality of education (Henderson, 2000). Spending patterns for public funds are known, but how much training costs is not understood. Medicare payment per resident is known to vary, but it is believed to reflect historical accounting practices rather than true differences in the cost or quality of programs (Young and Coffman, 1998). Many schools have not budgeted systematically for clinical education (The Commonwealth Fund Task Force on Academic Health Centers, 2002). Because current information is so poor, it is difficult to estimate the costs for educational reform or identify areas in which savings might occur (The Commonwealth Fund Task Force on Academic Health Centers, 2002). For example, costs might be incurred to implement computer-based instruction, but could reduce faculty time in some areas.

Financing

As discussed in more detail in Chapter 6, current financing methods for clinical education are not viewed as being supportive of the types of changes advocated in this report (LeRoy, 1994). The current methods have encouraged increases in the number, size, and duration of residency programs (Henderson, 1999, 2000) and programs for the training of specialists in tertiary settings (Young and Coffman, 1998). These methods have also hindered training in nonhospital settings (Henderson, 2000). Moreover, funding is not linked to any workforce goals, whether they be the types of changes described here or other goals related to the supply and mix of the output of the programs.

Interdisciplinary training is also discouraged by variation in how the education of different professions is supported. When Medicare began, educational costs for nursing and allied health professionals were allowable expenses for hospitals. Since 1965, however, many hospital-based training programs have been eliminated. For example, in 1965, 80 percent of training programs for registered nurses were in hospital-operated programs; today the figure is only 7 percent (Medicare Payment Advisory Commission, 2001). Medicare currently supports diploma nursing programs, programs for nurse anesthetists, and training for allied health professionals that are hospital-based programs. About one-half of hospitals with residency training programs also receive money for nursing and allied health training (Medicare Payment Advisory Commission, 2001). However, AHCs offer few hospital-based training programs, so the training support provided for medicine and that for nursing and allied health are going to different organizations, discouraging an interdisciplinary perspective. Another difference is that services provided jointly by a medical resident and supervising physician may be reimbursed, whereas the same is not true for other students. Therefore, nonmedical students do not offer the same advantages in cost recovery to the hospital sponsoring a training program; the result, again, is an emphasis on medicine.