Editor’s Note: This is a commentary on Huddle TS, Heudebert GR. Internal medicine training in the 21st century. Acad Med. 2008;83:910–915.
Medical specialties and subspecialties rapidly evolve to meet patient and market needs. Internal medicine has been debating the criteria for recognizing new forms of its practice and the training needed for them. In this issue of Academic Medicine, Huddle and Heudebert1 caution against practice differentiation too early in training and urge us to keep our core science-based diagnostic skills idealized by the hospital-based Oslerian internist. This commentary explores how specialty differentiation rests both on formal training and on practice-based learning and improvement. Training for specialty practice should become a continuum of learning fundamental knowledge and technique and demonstrating progression from competence to expertise through objective measures of competence in actual practice, rather than through months of training or a particular sequence of rotations in residency.
The Evolving State of Internal Medicine
American internal medicine is not a static specialty. The discipline has been called “bidirectionally pleuripotential.” Its practice uses common diagnostic methods as well as subspecialized or focused practice to meet patient needs and market forces. The specialty continues to evolve through research in basic and applied science. Individual practices evolve through practice-based learning and development of focused expertise. A brief history of the specialty’s 92 years traces this evolution.
The age of the consultant diagnostician.
American internal medicine emerged as a specialty with the founding of the American College of Physicians in 1916 and the American Board of Internal Medicine (ABIM) in 1936. Our forefathers linked British “naturalist” observation and taxonomy methods and German laboratory technology for examining tissue, microbes, and body fluids with French skills in bedside examination to solve clinical problems. German “innere medizin” became the British and American science-based internal medicine.2 The internist–scientists helped general practitioners and surgeons diagnose and manage patients in hospitals and, at times, in offices. They rarely practiced primary or principal care. This was the only variety of internal medicine for approximately 50 years.
The age of the internal medicine subspecialist.
Medical science and technology exploded from teaching hospital laboratories. After their initial hospital training, some internists became fellows, learning research techniques and expanding scientific knowledge about organ-specific diseases. These fellows became subspecialty practitioners, applying the methods of consultation, diagnosis, and treatment using scientific knowledge learned in the laboratories. By the mid 1970s, certification in nine internal medicine subspecialties had been defined by robust research agendas, training requirements, and technology for diagnostic and therapeutic procedures. Internists not pursuing laboratory fellowships practiced the specialty called general internal medicine (GIM). They continued consultant diagnostician practices based on the new knowledge and techniques learned in residency programs, incorporating subspecialty fellowships. By the late 1980s, subspecialty practice attracted 70% of graduates from internal medicine residencies.
The primary care general internist.
The specialty of primary care was born when the American Board of Family Practice was formed in the early 1970s. Not to be excluded, other generalist specialties declared proficiency in first-contact, continuing, and coordinated care for adults (in internal medicine), for children (pediatrics), and for women (obstetrics–gynecology). Insurance plans created a huge demand for primary care gatekeepers to rationalize referrals to other specialty and subspecialty services. This market force and the growth of family practice specialists, who bypassed internists seeking consultation directly from medicine subspecialists, shifted GIM practice from being hospital consultants to office-based primary care “doctors for adults.” By the end of the 20th century, GIM practice provided continuing care over time for a panel of patients in the office or the hospital. Internal medicine had evolved to a single specialty, with many varieties of practice rising to meet market and patient needs.
Residency programs struggled to balance the demands to redesign GIM training for office-based primary care with the insatiable demands to cover hospital services and provide hospital care for uninsured patients. Hospital training worked well for learning diagnostic skills, but training in hospital follow-up clinics proved inadequate for learning diagnosis of undifferentiated symptoms, managing minor problems never seen in hospitals, and keeping chronic care patients from needing readmission. Furthermore, research advancing scientific knowledge and technology in ambulatory and community practice got the academic short shrift. The growing cadre of GIM faculty pushed their way from low-status practitioners to higher-status university professors by incorporating the new “basic sciences” of epidemiology, economics, sociology, and psychology into theories of medicine. With lagging science and education in primary care practice, internal medicine graduates entered practice unprepared to be primary care doctors for adults. They remained true to the principles of internal medicine and engaged in self-education in how to apply the exploding medical science and technology to shape internal medicine office practices. The first few years in GIM practice became an on-the-job “fellowship” in office-based internal medicine. Throughout the 1990s the percentage of residents entering subspecialties reversed again, with 70% of graduates entering primary care GIM.
Focused practice in GIM (hospital medicine).
During the past decade, market forces have radically changed GIM practice again. Higher-paying hospital positions attracted general internists to become “hospitalists.”3 Resident work hours limitations, combined with caps on federal support for positions, limited further expansion. Hospitals, therefore, replaced resident services with hospitalists. This variety of GIM practice improved resident education by providing experts in hospital medicine and a stable generalist physician staff, as well as improved hospital quality, safety, and efficiency. The applied science of hospital medicine expanded, not through research fellowships, but through practice-based learning and improvement. At the turn of the millennium, the hospital patient census for many general internists in nonteaching hospitals dropped so low and the complexity of problems rose so high that many GIMs turned their patients over to the care of hospitalists. Subsequently, in 2007, most GIM graduates became hospitalists.
Focused practice in GIM (comprehensive care).
Freed from hospital obligations, general internists began focusing on how to apply techniques of practice-based learning and improvement in office practice. The health care demands of aging patients, and the increased burden of chronic illness, call for better information technology and office staff organization. An increasing volume of GIM office practice is no longer comprised of patients with straightforward clinical problems, but of patients with complex and unique concatenations of multiple chronic diseases, multiple medications, and myriad subspecialty services. The general internist’s knowledge, applied science, attitudes, and skills are ideally trained for framing these unique problems and creating customized, yet scientifically based, solutions involving coordination of care by multiple other specialists. The core principles of Oslerian applied science and discovery permit quality care when scientific knowledge from controlled clinical trials fails to guide decisions for this patient with this unique pattern of confounding problems.
During the past three years, the ABIM has explored future shapes of GIM practice with thought leaders in medical education, public policy, and practice, as well as with purchasers, payers, consumers, other specialists, and patients.4 What became clear is that coordination of longitudinal care for patients with complex problems is the most pressing need for future GIM practice. Moreover, the current information technology, organization of office staff and specialist network, and applied knowledge in community-based office practice are underdeveloped in many GIM office practices today.
Training for Many Practices in Internal Medicine
How, then, should we structure GME to prepare internists for satisfying and relevant careers that nimbly adapt to new knowledge, technology, and institutional practice contexts and to changing patient needs? We must first abandon the notion that valid learning for medical practice occurs only during residency or fellowship. Next, trainees and practitioners must become accountable for their continuous professional development by routinely engaging in guided reflection on measures of performance in actual practice and by providing assurance that they incorporate new knowledge, technology, skills, and attitudes to align practice with patient needs.
The current 36-month internal medicine training after medical school does a good job in teaching knowledge about the discipline and basic sciences relevant to internal medicine, the applied science and technology for diagnosis and treatment, and the skills and attitudes for providing service to patients.5 We can, however, do better in teaching lifelong practice-based learning and improvement.
The first 24 months of residency should continue to be a foundational practicum using reflection-on-action mentored by faculty, and reflection-in-action practiced by second- or third-year residents working with PGY1s.6 The “action” is solving actual patient problems in the many contexts of internists’ workshops. Hospitals, offices, nursing homes, and intensive care units should be selected for their importance in defining internal medicine practice. Experience in multiple contexts teaches tacit knowledge and attitudes about systems-based practice. Such rotational training embodies the knowledge needed to provide medical care through multiple specialists working across multiple contexts of care, and know-how in coordinating patient transitions among them.
The curriculum for internal medicine residency should not be an expanding list of topics. Instead, it should dynamically arise through residents learning the scientific knowledge and technique needed to solve their patients’ problems. Working in pairs, with or without medical students, senior and junior residents learn by thinking out loud and finding answers to their questions as they practice medicine. This approach to learning is called “reflection-in-action.” Residents learn how to solve clinical problems by performing standardized interviews and examinations and reviewing existing data to frame clinical problems explained by scientific theories about the causes and solution of potential disease. They order studies according to the pretest probability of making a diagnosis and estimating its trajectory without intervention, and they execute a coordinated plan for modulating the disease with pharmaceuticals and myriad specialist treatments. These steps shape the curriculum for study needed to answer questions arising from reflecting on the care. Assignment to care for various types of patients in different contexts ensures experience with solving the full spectrum of internal medicine problems.
Faculty mentors for practice-based learning teach by asking residents to think out loud and answer questions about surprises in care. They provide feedback about their direct observation of resident performance, and they stimulate anticipatory problem solving by asking “what if … ” questions. This technique, called “reflection-on-action,” teaches practice-based learning and improvement and epitomizes the Oslerian tradition promoted by Huddle and Heudebert.1 It can be applied to any patient-care situation in any context, not only for hospitalized patients.
That said, the hospital is an attractive context for teaching applied internal medicine because it can provide ample opportunity for faculty observation, feedback, and guided reflection-on-action during bedside teaching rounds, morning reports, and hand-off rounds. Obviously, the hospital is the only context for learning system-based hospital care. Additionally, compared with many teaching clinics, hospital services have more sophisticated infrastructures, teams of professional personnel, and carefully integrated processes for care. The outcomes of actions in hospital care are more immediate, and therapeutic actions do not depend on patients to follow through.
Reflection-on-action in real settings, solving real problems, is the only way residents learn the skills and attitudes needed to actually provide internal medicine services. Feedback and counsel from senior residents and faculty teaches junior residents to shape attitudes of unselfish service, quest for safety, respect for patients’ unique personhood, and the promise to do no harm while transforming suffering and disease into healing and health. Working together, members of the clinical team refine their professional attitudes in the crucible of conflicting priorities, enormous demands, emotional tension, conflicting values, and rapidly evolving knowledge and technology. Explicit dialogue about how physicians cope with these tensions turns the “hidden curriculum” into explicit lessons.
Many teaching hospital patients may be so complex that Dreyfus model “advanced beginners,” who use clinical rules of thumb and evidence-based guidelines for diagnosis and therapy, may find their care to be frustratingly impossible.7 These complex patients are unique, and although evidence-based guidelines fall short in their care, proficient and expert specialists who focus their practice on these cases manage their problems by applying general principles and making wise judgments intuitively. These hospitalized patients are ideal for reflection-on-action mentoring for internists who have mastered the basics and are ready to tackle learning to become virtuoso practitioners. Expert hospital-based internists who have mastered the hospital system of care are valuable mentors in guiding competent internists through the maze of “no-evidence” decision-making necessary for handling unique and complex clinical problems.
Once residents demonstrate competence in applying the internal medicine method of diagnosis and treatment with professional attitudes and skills, they are ready to progress to focusing these skills on the care of particular types of patients within particular contexts of care. They are ready to begin subspecialty training or to develop proficiency in focused areas through practice-based learning.
Competence cannot be determined by an arbitrary number of months of training; however, it seems unreasonable that competence could be achieved in fewer than 24 months of foundational training, and many residents may require 36 months or more. The model of foundational training in method and knowledge, followed by focus through guided practice or fellowships, makes sense. I agree that the continuing care clinics are difficult to incorporate in many residencies, and learning the knowledge and context of ambulatory care might be better achieved through long block assignments. The transition from foundational training to subspecialty or focused GIM practice should be based on objective evaluation of competence to practice in actual contexts of care, to meet real patients’ needs. Structured direct observation of clinical practice skill and attitude during foundational training, and passing an examination of knowledge and judgment, are robust evaluations of competence that provide the evidence for progression. Measurement of practice quality, patient experience of care, volume of experience in practice, and 360-degree evaluations (including continuing medical education mentors) can be aggregated to demonstrate increasing levels of proficiency and expertise in practice. Fellowships provide a mixture of foundational training in subspecialty techniques, and practice-based learning in a particular context of care, for patients with specific problems.
Practice-based learning relies on discipline in self-directed reflection-in-action, stimulated by surprise with unexpected observations or outcomes. But self-directed reflection is not enough.8 Self-regulating professionals value accountability through regularly measuring their performance in practice against standards for best practices and reporting these results to those to whom we are accountable. Self-improving professionals use the performance measurement and reflection-on-action to change practice and the practice system to improve performance. Board certification and maintenance of certification (MOC) by an American Board of Medical Specialties member board provide robust and valid methods for professional accountability to serve the health needs of our patients and society.9
The ABIM has proposed that MOC might be used to demonstrate proficiency and expertise in a particular focus of internal medicine practice.10 Focusing MOC in a particular area of internal medicine practice paves the way for demonstrating and publicly acknowledging proficiency. This approach converts certification and recertification into a valued growth process for professional development, rather than a hoop to jump through to show that one is still “good enough.”
So, asking “what is the best duration of training for internal medicine” misses the point that training for professional practice should be a seamless process from basic education, professional practice formation, and subspecialization or developing expertise in focused areas of practice. The concept of competency-based education and certification frees us from the mental prison of time-based formal training and lifelong one-time certification. Using reflection-in-action with mentors guiding reflection-on-action over sufficient contexts of care, with a sufficient number and variety of patients, should be the criteria for training in internal medicine. Robust evaluation that demonstrates competence permits passing objective specialty board evaluations to demonstrate achievement of the next stage of professional development. The framework for this model is available now, and many measures of competence are already in use. It’s time to do it.
The opinions expressed in this commentary are those of the author and do not represent the official position of the American Board of Internal Medicine.