The potential for federal budget cuts and economic upheaval in the health care market to dramatically reduce the financing available for subspecialty training is of grave concern [1]. The benefits of high-quality graduate medical education are realized by all segments of society. Its costs at all levels should be broadly borne, and health insurers, including managed care plans, should not be allowed to opt out of financing this education. But the primary interest of the American Board of Internal Medicine is the quality of training of internists; ways to finance graduate medical education are beyond the Board's purview.

The Board supports the view that too many subspecialists are being trained [2]. The Board also anticipates that funds for excellent training of fewer subspecialists will continue to be available. Although training fewer subspecialists will reduce educational costs, it will also create stress and tension in a system that, until now, has depended on subspecialist fellows to provide much-needed service. These are realities that must be dealt with, but they must not undermine public commitment to highquality training for physicians. In return, the medical profession must take seriously its obligation to use educational resources as efficiently and effectively as possible. To maximize the contribution of fewer subspecialists, rigorous training that is appropriate, relevant, and realistically matched with career opportunities is essential.

Subspecialists have made major and unique contributions to health care by advancing medical knowledge and technology and promoting excellence in the management of complex illnesses. Although discussions about the physician workforce have focused primarily on the clinical contribution of specialists, planning for the future must include continued investment in knowledge that results from subspecialty basic and clinical research. Subspecialists are needed as both investigators and clinicians. All subspecialists must have a comprehensive knowledge base and broad experience in the clinical practice of their subspecialties. However, diversity of the future roles for subspecialists, coupled with new demands and diminishing resources for training, led the Board to recommend different but overlapping training pathways for the subspecialist investigator and the subspecialist clinician (Table 1).

Because there is an oversupply of subspecialist clinicians [3-5], the total number of subspecialty fellowship positions should be reduced, and the proportion of investigator trainees should be increased. Nevertheless, it remains important to train subspecialist clinicians to prevent future gaps in the supply stream, to provide for the needs of patients with complex medical illnesses, and to maintain the number of subspecialist clinical faculty to serve as teachers and role models.

Each subspecialty training program should have the necessary resources to provide high-quality training in both the clinical and investigative paths. However, occasionally concentrating resources in fewer programs may be the most effective way to provide excellent training for a cadre of subspecialist basic scientists and clinical investigators. Consortia comprising several institutions may be useful models for aiding future subspecialty training, allowing geographic pooling of training resources to provide a coordinated approach to training participants for various subspecialty careers.

While recommendations for specific subspecialty training pathways were being formulated, it became clear that each pathway would be enhanced by additional didactic learning and formal coursework that might lead to a separate degree in some cases. Didactic topics for each training pathway are described in more detail later in this article, but examples include the subspecialist basic scientist with a graduate degree in molecular biology and the clinical investigator with a degree in clinical epidemiology. The curricula for the didactic components of training, many of which are not as extensive as that for a separate degree program, are independent of the clinical content of training. In other words, a common body of knowledge is required of investigators and clinicians, regardless of their medical discipline. Therefore, cross crossdivisional crossdivisional divisional modular units of instruction should be considered as an efficient way to extend training opportunities throughout departments of medicine. Many of these educational components would be equally appropriate for internal medicine residents and subspecialty fellows, furthering the potential efficiency of these components.

The Role and Training of the Subspecialist Investigator

Scientific inquiry and innovation are hallmarks of the subspecialist research investigator. The range of research done by subspecialists is wide and varied. The Board anticipates that the career pathways of subspecialist investigators will evolve in two directions, toward basic patient-oriented research and toward applied patient-oriented research [6].

Although the training for each of these two investigative pathways includes several unique components, both require a long and rigorous curriculum. Subspecialties are part of the continuum of internal medicine. Subspecialty training must begin, as it does now, with a solid foundation in internal medicine. Thus, the Board recommends that the training pathway for all subspecialist investigators consist of a coordinated curriculum that includes at least 2 years of internal medicine training, at least 1 or 2 years of clinical subspecialty training (depending on the subspecialty), and at least 3 years of research training under the supervision of the subspecialty program director. The Task Force on the Curriculum of Internal Medicine of the Federated Council of Internal Medicine is expected to issue a final report in 1996. The Board encourages the identification of a core internal medicine curriculum that includes aspects of internal medicine training that are essential to subspecialty investigators. For some trainees, individual programs may require additional protected time for research activities. Training should also include documented continuity experience in the internal medicine ambulatory setting and formal coursework in research methods. The Board will consider subspecialist basic scientists and clinical investigators with 7 years of training to be eligible to apply for certification in internal medicine and cardiology, gastroenterology, or two subspecialties (such as hematology and oncology). Certification of investigators in other single subspecialties will require at least 6 years of training.

The Subspecialist Basic Scientist

Continuous improvement in health care is founded on basic research. The contribution of the physician-scientist who expands the frontiers of biomedical knowledge through hypothesis-testing in the laboratory remains essential to future improvement in health care. Health care benefits can be obtained in the short term by leveraging the existing knowledge base, but long-term advances would be jeopardized if knowledge ceased to expand. It is also crucial that subspecialist physician-scientists maintain an intellectual exchange with their investigator-colleagues in preclinical departments. Yet, according to the Association of Subspecialty Professors, fewer than 2% of all physicians are biomedical scientists, and they recently have been steadily "disappearing." According to Neilson and colleagues [7], "the common denominator that marks a young physician-scientist successful is having had the best training possible..."

Training of the subspecialist basic scientist requires a rigorous and varied curriculum. In addition to training in internal medicine and the clinical subspecialty, the basic scientist needs an in-depth understanding of basic science and biostatistics, study design, and legal and ethical principles, as well as substantial laboratory experience. Formal course work in basic laboratory techniques, computer technology, evaluation of experimental data, biostatistics, and grant and manuscript writing should be included. Duration of research experience during training is well accepted as a predictor of the success of an investigator. Although some argue that 2 years should be the minimum [8], the Board endorses the recommendations of the National Research Council and others who recommend at least 3 years [9]. This stringent requirement makes the total training period required for certification in the subspecialty 1 year longer for an investigator than for a clinician. Although it can be argued that longer training discourages young physicians from considering careers as investigators, the alternative is worse. Insufficiently trained investigators will not be successful in competing for research funds.

The Subspecialist Clinical Investigator

A recent report from the Institute of Medicine predicted that the shortage of clinical investigators apparent in the last decade, when clinical investigation lagged behind basic science research in scientific advances, would continue [9]. If funding for clinical investigation is available, opportunities for translating research from the laboratory bench to the clinic will increase, allowing new knowledge of molecular and cellular biology to affect health care delivery more quickly. New approaches to the prevention and management of disease will evolve from biological and gene therapies. Moreover, fueled in part by changes in health care financing, clinical research will increasingly emphasize outcome studies, health services research, population-based medicine, and cost-effectiveness analyses.

The clinical investigator must understand the fundamentals of the design of clinical studies and their implementation, as well as sampling methods, concepts of sample size, and analytic methods. Formal coursework in experimental design, statistical methods, epidemiology, pharmacology, behavioral sciences, economics, and grant and manuscript writing should be included. The demand for well-trained clinical investigators able to design and conduct clinical studies and to analyze and report their results will increase substantially, especially with regard to studies that link scientific investigation with clinical application. Developing the skills necessary to conduct investigation in all of these areas requires research experience and a training curriculum that is just as lengthy and rigorous as that required of the subspecialist basic scientist. Until now, the training of clinical investigators has been fragmented and diverse, ranging from simple mentoring with no formal training to the adaptation or transference of laboratory techniques learned through experience in bench research [9].

The Role and Training of the Subspecialist Clinician

The growth of managed care will profoundly affect the practice of subspecialty medicine. To be more efficient, generalist physicians and nonphysicians may do tasks previously done by subspecialists. As the number of subspecialists decreases, their role in providing primary care will diminish, while their role in providing technical expertise and highly specialized knowledge and judgment will expand.

Preserving high-quality care will require appropriate patient access to subspecialty care [10]. The coordination of care will require the creation of teams of health care professionals. Common health problems and diseases with known therapies will be managed by generalists, who most often will be the team leaders responsible for directing care, and by subspecialists, who will function as consultants to the primary care providers. But for some patients, the most effective care will require continuing access to the subspecialist who provides principal care. Originally described by Aiken and colleagues in the late 1970s [11], principal care is a special category of primary care in which the specialist provides most of the patient's care. The Board advocates continued use of the principal care model, appropriately modified for managed care environments, for patients with complex diseases that are difficult to manage, such as patients receiving chemotherapy or transplants or patients with end-stage renal disease.

Some patients with serious and complicated illnesses have special needs that differ from those of most patients. Because the health care of such patients is often dominated by their illnesses, most of their needs require the expertise of the subspecialist clinician. In the principal care model, the subspecialist would direct the care of such patients in collaboration with the generalist and other members of the health care team. As the principal care provider, the subspecialist would be expected to assume responsibility not only for providing accessible subspecialty care but also for coordinating other services and communicating directly with patients and their families [12].

Discoveries in immunology, molecular biology, genetics, and the information sciences have added substantially to the knowledge base required of the subspecialist clinician and the subspecialist investigator. As the expert in clinical subspecialty medicine, the subspecialist will serve as a resource for and educator to the entire health care team. The role of the clinician-teacher (who was formerly an academic clinician) will also be borne by physicians in community practice. The subspecialist clinician will be responsible formulating practice protocols and clinical guidelines to facilitate implementation of aspects of subspecialty care by generalists. Coursework in computer science, teaching techniques, quality improvement techniques, and clinical epidemiology should be required. During training, subspecialist clinicians must have firsthand knowledge of and experience with research. Although the clinician in training would not be expected to be the principal investigator, participation in clinical studies would encourage development of a thorough understanding of the scientific principles of research necessary to interpret and critically evaluate the literature. Increasing the number of subspecialists who are training to become investigators while decreasing the number training to become clinicians will increase the opportunity for interaction between subspecialist clinicians and subspecialist investigators, which is valuable to the training of each.

Many subspecialties of internal medicine are distinguishable by the specialized knowledge and understanding of sophisticated technologic procedures required for each. Indeed, competency in doing procedures; understanding the indications, contraindications, and complications of diagnostic and therapeutic procedures; and interpreting the results of procedures are among the essential skills of the subspecialist. But subspecialists do not have a monopoly on subspecialty procedures. As technology evolves, other health care providers will probably be required to become proficient in some of these technologic procedures. In these circumstances, subspecialists will have an important role in teaching procedural skills.

For certification of the subspecialist clinician, the Board will continue to require completion of a coordinated curriculum of 3 years of internal medicine training followed by subspecialty training. More clinical training experience will be required to become a subspecialist clinician than to become a subspecialist investigator. The common background of education and experience shared by general internists and medical subspecialists encourages numerous models of effective clinical teamwork. The fact that subspecialists will perform as principal care providers further justifies comprehensive training in internal medicine before broad, intense training in the subspecialty. A minimum training period of 3 years is required in several of the subspecialty disciplines because of the growth of knowledge and technologic advances and the need for subspecialty clinicians to serve as consultants, teachers, and principal care providers. The Board anticipates that there will be an evolution to 3-year training periods for most, if not all, of the subspecialty disciplines [13]. Although this conclusion is based on the fact that increasing knowledge and experience are required of subspecialists, it is further supported by a study of cardiology training programs, in which the duration of training was found to be associated with better performance on the certification examination, independent of other factors [14].

The Board recommends that each subspecialty community of educators carefully assess the curriculum and length of its training programs to ensure that the education and experience necessary for future subspecialists to develop the appropriate skills are provided. The subspecialty communities should develop detailed descriptions of the skills that should be acquired during subspecialty training, including teaching subspecialty medicine to colleagues, acting as a consultant and as a principal care provider, and performing the technical procedures of the subspecialty.

Certification of Subspecialists

Although passing the certification examination is a dramatic event, it is important to remember that certification is a larger process that includes the satisfactory completion of a prescribed period of training in an accredited program and verification by the training program director that a trainee has mastered the knowledge and skill necessary to provide excellent patient care.

In this article, we describe the American Board of Internal Medicine's new recommendations for subspecialty training, delineating overlapping but different pathways for subspecialist clinicians and investigators. We predict an increase in the required length of clinical training for several subspecialties and a new path for the training of investigators. It should be noted that the Board has offered a special but restricted pathway for clinical investigators during the past decade. Internal medicine and subspecialty program directors were required to formally propose each candidate, and the Board formally approved each one. With implementation of these new recommendations, the subspecialist investigator pathway will be more available to qualified subspecialists seeking investigative careers.

Over the next year, the Board will consider ways to document and monitor the research component of subspecialty training. The creative and diverse nature of research activity makes assessment at the national level difficult, but research experience needs to be recognized as an essential and integral component of subspecialty fellowships. The Residency Review Committee in Internal Medicine, which includes research experience in its minimum requirements for completion of fellowship programs, should strengthen basic and clinical research requirements for the accreditation of subspecialty programs to improve overall quality.

Subspecialty boards will consider expanding the content of certifying examinations to include more of the cognitive aspects of scholarly training, such as the critical evaluation of the literature, principles of epidemiology, and research design. Although two pathways are recommended for subspecialty training, only one certifying examination will continue to be offered in each subspecialty discipline. Sufficient overlap exists in the requisite knowledge of the subspecialist investigator and the subspecialist clinician to justify a common cognitive examination at the end of the training period. Recertification, which will be required to maintain certification, will provide an opportunity to reassess physicians during their careers. One of the goals of the recertification program is to be responsive to the differentiation among careers that is common in a field as broad and diverse as internal medicine.

The Board will continue to require documentation of proficiency in certain essential procedures for certification in the subspecialties [15]. In doing so, the Board prefers to use a uniform set of quantitative standards that is widely endorsed and accepted by the subspecialty and internal medicine communities. Therefore, the development of uniform standards by the Residency Review Committee for Internal Medicine, in collaboration with the Federated Council of Internal Medicine, the Association of Subspecialty Professors, and the subspecialty societies, is encouraged. Similar standards of technical competency should be applied to the same procedures regardless of specialty discipline.

Conclusion

Top Conclusion References

Internal medicine demands both the intellectual pursuit of scientific knowledge and the compassionate provision of high-quality patient care. The training of subspecialist basic scientists and clinical investigators must be protected and nurtured to ensure continued discovery and scholarly application of new knowledge to future patient care. The future health care system will continue to need subspecialists, albeit in reduced numbers, to perform as basic scientists, clinical investigators, and clinicians. The skills needed to perform in these roles will increase. Subspecialist clinicians must be appropriately trained to meet the new challenges that will be faced in managed care environments.

The American Board of Internal Medicine supports the following principles and has unanimously approved (June 1995) the following recommendations to set the course for subspecialty training over the next decade.

1. As the total number of subspecialty fellowship training positions is reduced to conform to societal need, the proportion of subspecialist clinicians should be reduced and the proportion of investigators increased.

2. Subspecialty practice and, therefore, training must evolve from the continuum of internal medicine.

3. For some patients with complex illnesses, the subspecialist clinician should be the principal care provider.

4. Specifically planned training paths for the subspecialist investigator and the subspecialist clinician should be organized.

5. Each subspecialty should assess itself to determine the optimal duration of training and the appropriate balance of clinical and research training required to produce enough subspecialists who are competent as clinicians and investigators.

6. As fellowship training programs are reduced in size, opportunities to improve their quality should be realized; standards required for accreditation should be increased accordingly.

7. Departments of medicine should provide cross-divisional didactic learning experiences to enhance the scholarly aspect of training.

8. Responsibility for the cost of graduate medical education and research should be shared by all health care payers. Subspecialty training, an integral part of internal medicine, deserves continuing support.

9. The American Board of Internal Medicine should develop questions to be included in certification examinations to evaluate the nonclinical didactic components of subspecialty training.

10. Common standards for establishing and maintaining the technical proficiency of subspecialists in doing procedures should be created for all specialty disciplines.

Dr. Toskes: Division of Gastroenterology, Hepatology, and Nutrition, University of Florida, College of Medicine, 1600 SW Archer Road, Box 100214, Gainesville, FL 32610.