Share this article on:

Beyond Flexner: A New Model for Continuous Learning in the Health Professions

Miller, Bonnie M. MD; Moore, Donald E. Jr PhD; Stead, William W. MD; Balser, Jeffrey R. MD, PhD

doi: 10.1097/ACM.0b013e3181c859fb
Flexner Centenary: Article

One hundred years after Flexner wrote his report for the Carnegie Foundation, calls are heard for another “Flexnerian revolution,” a reform movement that would overhaul an approach to medical education that is criticized for its expense and inefficiency, its failure to respond to the health needs of our communities, and the high cost and inefficiency of the health care system it supports. To address these concerns, a group of Vanderbilt educators, national experts, administrators, residents, and students attended a retreat in November 2008. The goal of this meeting was to craft a new vision of physician learning based on the continuous development and assessment of competencies needed for effective and compassionate care under challenging circumstances. The vision that emerged from this gathering was that of a health care workforce comprised of physicians and other professionals, all capable of assessing practice outcomes, identifying learning needs, and engaging in continuous learning to achieve the best care for their patients. Several principles form the foundation for this vision. Learning should be competency based and embedded in the workplace. It should be linked to patient needs and undertaken by individual providers, by teams, and by institutions. Health professionals should be trained in this new model from the start of the educational experience, leading to true interprofessional education, with shared facilities and the same basic coursework. Multiple entry and exit points would provide flexibility and would allow health professionals to redirect their careers as their goals evolved. This article provides a detailed account of the model developed at the retreat and the obstacles that might be encountered in attempting to implement it.

Dr. Miller is senior associate dean for health professions education, Vanderbilt University School of Medicine, Nashville, Tennessee.

Dr. Moore is director, Division of Continuing Medical Education, and director, Education and Evaluation in Graduate Medical Education, Vanderbilt University School of Medicine, Nashville, Tennessee.

Dr. Stead is associate vice chancellor for strategy and transformation, Vanderbilt University School of Medicine, Nashville, Tennessee.

Dr. Balser is vice chancellor for health affairs and dean, Vanderbilt University School of Medicine, Nashville, Tennessee.

Correspondence should be addressed to Dr. Miller, 201 Light Hall, Vanderbilt University School of Medicine, Nashville, TN 37232-0685; telephone: (615) 343-7536; e-mail:

When Abraham Flexner visited each of the 155 existing North American medical schools in the years preceding the publication of his report in 1910, medical education in the United States was slowly emerging from its unregulated, proprietary origins. Relatively few schools offered a strong foundation in the newly developing disciplines of human biology, and he cited those that did as exemplars. His resulting recommendations to link medical education to the university, embed both didactic scientific teaching and laboratory experience within the curriculum, provide ample clinical opportunities, and demand strong premedical preparation were quickly adopted and changed the face of medical education for decades.1 In a dialectical swing, the study of medicine became the study of science and physicians became applied human biologists who at times seemed to disregard medicine's ancient roots as a healing art. By midcentury, educators began to perceive weaknesses in the model, which they attempted to remedy through such strategies as the biopsychosocial model of disease, case-based pedagogies, and the early integration of clinical experience with basic science instruction.

If we focus only on Flexner's specific recommendations, however, we may miss the real meaning of his message. What Flexner argued against was an educational system that operated with little accountability and no regulation and that was far more concerned with providing a livelihood for the teachers than effective medical care for the public. It had failed to keep up with the growth in knowledge that was applicable to the maintenance of health and the understanding of disease. One hundred years after Flexner, despite numerous calls for reform, medical education seems to be clinging once more to models that no longer respond to the health care needs of the community. And although our educational system is now highly regulated, it is not yet accountable for the most important of outcomes—the health of the public.

Back to Top | Article Outline

A New Vision for Learning

The current approach to medical education poorly prepares our graduates to confront the challenges of an increasingly complex health care system.2,3 Because of this, we suggest that a totally different educational system is required to address the learning needs of health professionals along the educational continuum. Incremental modifications to existing models will no longer suffice for the following reasons. First, the body of biomedical knowledge is larger than any one person can master, and it continues to grow at an unmanageable pace. Although we can no longer expect health professionals to recall all the biomedical information they may need during patient encounters, our educational models are still based on that expectation. Second, despite advances in knowledge about the multidisciplinary determinants of health, the dominant focus of didactic study remains on the biomedical sciences. Third, the current approach does not develop in learners the capacity to deal with the uncertain environment in which they will work. Some observers have suggested that successful performance will depend on effective responses to unpredictable factors that emerge from the dynamics of the workplace. This environment demands a new set of skills, including the ability to work in interprofessional teams. Finally, the current system of medical education does not adequately nurture the skills needed for lifelong learning, nor does it develop in learners the ability to analyze practice performance and make changes that improve patient outcomes.

To address these concerns, a group of Vanderbilt educators, administrators, residents, and students gathered with a panel of U.S. and Canadian continuing education experts at a three-day retreat in November 2008. The invited students and residents had previously demonstrated their interest in medical education through participation in curriculum committees; the participating educators and administrators hold responsibility for Vanderbilt's education enterprise. The group reported to the dean of the medical school, who also attended the conference. The goals of this meeting were to examine physician learning comprehensively, craft a new vision of physician learning based on the continuous development and assessment of the competencies necessary to practice as effective and compassionate caregivers under challenging circumstances, and outline the initial steps needed to implement this vision not only for physicians but for all members of the health care team.

The vision that emerged from this gathering was that of a health care workforce comprised of physicians and other health professionals, all capable of assessing their performance and making the changes needed to provide the very best care for their patients. Several principles are central to this vision (List 1). Learning should be competency based and embedded in the workplace. It should be linked to patient needs and undertaken by individual providers, care teams, and health care institutions. All providers and patients should be members of in-person and/or virtual learning communities, which would be based on clinic placements, specialty needs and interests, and academic affiliations. Medical homes, which provide and coordinate health care needs for patients, would also be learning homes that provide and coordinate educational needs for learners along the continuum. An informatics infrastructure would support this learning environment, and continuing education professionals would provide the needed coaching for the development and implementation of appropriate learning plans. With learning situated in the workplace, learning outcomes and health care outcomes would be directly linked.



The group also determined that health professionals must be trained in this environment from the start of their educational experience. Interprofessional learning/working teams would share facilities and the same foundational coursework. As learners progressed through this educational system, they would perform health care work of increasing responsibility and complexity, which might offset the cost of attendance. To realize this vision, the group determined that three interrelated domains would be required: a rich electronic learning system that enables the creation of in-person and virtual learning communities, a program for constant and iterative workplace learning, and a new model for health workforce development.

Back to Top | Article Outline

Learning Communities

Learning communities would be both in-person and virtual, supported by systems that provide access to evidence-based information resources, learning activities, and communication services. In-person learning communities would include the patients, providers, and learners who share a workplace and a medical learning home. Virtual communities would be defined not by physical location or role within the community but by common interests, concerns, or affiliations. For example, a virtual learning community might be created around the diagnosis, management, and understanding of schizophrenia. Members of this community might include patients, families, community mental health workers, social workers, pharmacists, nurse and physician providers, postgraduate trainees, basic scientists, and health professions students, all of whom might work, learn, or receive care at distant and diverse geographical locations. While in-person learning communities would be based within institutions, virtual communities would be inclusive and expansive. The electronic platforms connecting these communities would provide both public and secure role-based access.

The information resources provided by this portal-based system would facilitate just-in-time answers to questions regarding new symptoms, unique presentations, and best practices and could include access to the formal curriculum of a medical school, links to future forms of commercial sources such as WebMD, UpToDate, or MD Consult, and links to presaved literature searches conducted by members of the community. The system could also send alerts and updates regarding fundamental discoveries, new management strategies, and medication safety and could link the community to important stories in the news regarding health and policy. Learning activities provided to the communities would be blended and would include case conferences, workshops, and workplace seminars. The communities would also provide access to digital libraries and interactive learning modules. The latter could include not only internally created programs but also those provided by professional organizations, such as the Association of American Medical Colleges' MedEdPORTAL ( Members of the learning community would be encouraged to contribute to the collective resources through the creation of wikis and refereed compendiums. Social networking technologies would connect members of the learning community with each other, and messaging systems would allow patients to communicate with providers and providers to communicate with each other as well as with their students. Finally, members of the community could post formal queries that would be answered by either specialty consultants or health librarians. Individual learning communities would no doubt overlap to create a complex, interwoven, fluid network that would respond to change and rapidly transmit knowledge and information.

The informatics capabilities and information technology required for the learning communities already exist, and Vanderbilt's Department of Bioinformatics has developed many programs that could serve as building blocks. Star Panel is a comprehensive electronic medical record (EMR), and Star Tracker sorts patient panels by diagnosis so that quality indicators can be applied and measured. KnowledgeMap serves as a searchable data repository for the curriculum, and KM Portfolio interfaces with the EMR and automatically captures all student entries in individual learning portfolios, thereby creating patient panels for each student. Academic health centers should take the lead in further developing and disseminating such tools and in applying them to education. While federal funding from the National Institutes of Health and the National Library of Medicine might support development efforts, it is possible that user fees will be needed to support the extensive infrastructure required to maintain the system.

Back to Top | Article Outline

Workplace Learning

Workplace learning is defined as “the way in which individuals or groups in a workplace acquire, interpret, reorganize, change, or assimilate a related cluster of information, skills or feelings in response to challenges in the workplace.”4 The curriculum for workplace learning derives naturally from the workplace itself when individuals, teams, and organizations strive to improve performance, even if performance by usual standards might be considered adequate. Workplace learning deserves critical attention because it addresses many of the concerns driving reform in medical education. It relies on the skillful use of information resources, an understanding of process and contingency, and the input of all professionals on a team, thus decreasing reliance on the knowledge memorized by a sole provider, which can no longer be considered adequate to achieve the highest quality of patient care. Ideally, learning in the workplace should foster the sustained development of both the individual and the organization, within the context of present and future organizational goals and individual career aspirations. It can be formal, informal, incidental, or experiential,5 and in health care, it is beginning to emerge as “practice-based learning”6,7 or experiences-based learning in practice.8

Recently, educators have developed strategies to address workplace issues in formal curricular programs. One strategy brings a simulated workplace into the learning experience. Using a variety of techniques, educators can situate learning in simulated work environments. These techniques include case presentations, role-plays, computer simulations, and standardized patient encounters. The authenticity of the simulation increases with the use of actual practice data. A second tactic brings “formal learning experiences into the workplace.” A variety of educational activities can be situated in an actual practice setting, such as academic detailing and video or Web conferencing. While both of these general strategies have demonstrated effectiveness, their primary value has been in supporting a predetermined curriculum that does not necessarily address the dynamic and unpredictable learning needs of health professionals in the real world.

To support the more informal, responsive, and dynamic model of workplace learning that we are advocating, two elements must be present: an infrastructure that supports learning opportunities in the workplace, and providers who have the skills to recognize learning needs and follow through by engaging in learning activities. The electronic learning resources described above could supply the infrastructure. In addition to the functions already outlined, the electronic learning platform would link to patient health records, facilitate outcomes analysis, and guide providers to learning projects based on identified performance deficiencies. Learning opportunities could be provided synchronously, when an individual patient presents whose management falls out of recommended guidelines, or asynchronously, when provider or team performance for a panel of patients is considered suboptimal.

Teaching health professionals to identify learning needs and to develop effective learning strategies will pose considerable challenges. Two skill sets seem to be critical to this process: self-assessment and reflective practice. Studies have demonstrated the inaccuracies of self-assessment in the absence of supporting data,9–11 and recent attempts to encourage the development of reflective skills have met with varying degrees of success.12 Nonetheless, for a culture of constant workplace improvement to take hold, learners must begin to practice these skills from the start of the educational process. Curricula should provide at least a basic understanding of improvement science, and workplaces should include safe settings for teams to discuss and process both successful and disappointing outcomes. As a move in this direction, many morbidity and mortality conferences are moving away from the classic “shame and blame” format to one that is systems based and focused on improvement.13 In response to the Accreditation Council for Graduate Medical Education core competencies, some residency programs are requiring that residents participate in practice improvement projects. Success of the workplace learning enterprise will also require fundamental changes in the functions of offices of continuing education.14 Practice improvement continuing education15 should become a required component of a provider's continuing education portfolio, and continuing education professionals should assist learners in the creation of both improvement projects and individual learning plans (ILPs). With guidelines and recommendations from specialty societies, ILPs could be used for maintenance of certification.

Back to Top | Article Outline

Health Workforce Development

In the new model that we envision, the education of all health care professionals would begin in the same setting, a blended work and school environment. Conceptually, the model is based on the principles that even novices are capable of making significant contributions to health care delivery and that health care education gains greater meaning and effectiveness when carried out in the context of authentic work and responsibility. On matriculation, all students would be assigned to a clinic-based learning community consisting of providers and postgraduate trainees who oversee the care of a population of patients and serve as teachers, advisors, and mentors. Students would immediately learn basic patient care skills, such as taking vital signs and performing intake interviews, and would learn basic principles of microsystems by studying how the clinic works. With these skills acquired, they would become members of the team that provides care for patients and administers the clinic. They would join the clinic's electronic learning community and would maintain individual portfolios that interface with the community platform and document their achievements. Within this portfolio, learners would maintain ILPs that outline learning goals as well as a road map for achieving them. Learners would formulate their ILPs with the guidance of continuing education coaches who would review them on a regular basis, as learners either attained their goals or changed them. Whereas most medical schools now include clinical experiences in the preclinical years, this model eliminates the “preclinical” and “clinical” distinction by giving students real responsibility for clinic operations and patient care from the start of their professional education and by making the learning–working team the focus of their educational experience.

Learning activities would be organized as modules that could be taken in a variety of sequences, based on assessment of learner capabilities, interests, and goals. Certain formal learning modules would be required of all students, regardless of the professions they intend to enter. These would cover topics such as health care macrosystems, the determinants of health and illness, public health and prevention, basic communication skills, and quality, safety, and improvement. Students would select other modules required for their particular career paths. For example, those students who plan to enter the biomedical science-oriented fields, such as medicine, nursing, and pharmacy, would take modules that introduce foundational concepts in human biology. Those who plan to enter health-related professions in the social sciences, such as social work and health systems administration, would take modules related to those fields. As learners' paths became more differentiated, the modules would become more specific for one profession or another. Learners who intend to become physicians would still be required to take hospital-based modules, and all students could elect to take specialty experiences in a variety of clinical settings that would facilitate career decisions. Designed to convey the essential knowledge needed at specific developmental stages, this sequenced system of modules would create a more efficient “just-in-time” quality that eludes our current curricula.

The model could accommodate multiple entry and exit points, based on the learner's prior experiences, competencies already acquired, and professional goals. Learners could even enter this school before completing baccalaureate degrees; parent universities could award “preprofessional” bachelor degrees by assigning double credit to the early modules of the health professions curriculum described above. Initially, these learners would transition into existing individual professional schools as they achieve prerequisite competencies. Eventually, we envision the coalescence of individual professional schools into a single entity that provides learning activities for the development of the competencies required for the various degrees. This demands that the competencies for all careers must be clearly defined and measurable. As aspirations evolved, the learner could reengage the system and begin a new pathway, leveraging existing competencies and minimizing the need for repeat coursework. The infrastructure of postgraduate training programs and teaching hospitals and clinics would remain intact.

The health care workforce school should maintain lifelong connections with learners, even after they complete programs and leave the physical learning community. In addition to learning and information resources, virtual communities established by the school could provide access to practice guidelines and information technologies that facilitate the measurement of outcomes. If needed, providers would also be able to return to their learning homes for on-site immersion experiences, which might be simulation based or classroom based. This capability would be especially helpful if new procedures replaced those previously learned or if new bodies of knowledge developed that could best be mastered with an in-depth, focused course of study.

Some examples of how learners might progress through the new health professions system that we suggest follow.

Back to Top | Article Outline

Example 1

Jane is a 20-year-old woman who enters the health professions school after two years of undergraduate school; she is uncertain of the exact path she will take. She becomes embedded in a clinic team, takes core learning modules, and after a year determines that she wishes to become a physician. As she progresses through the system, her learning modules become more specialized both to her chosen profession and to her individual interests. As her competencies develop, she is given more responsibilities in clinic, and she serves as a mentor to new students. Over the next three years, she becomes intrigued by the overwhelming problem of obesity that she sees in her clinic patients. She takes a year to perform research with an endocrinologist, then returns to school with plans to become a bariatric surgeon. She achieves the competencies required for the MD degree, then begins postgraduate training in surgery, focusing on minimally invasive and bariatric procedures.

After Jane completes her surgical training, she takes a position at a large multidisciplinary clinic. She remains connected to several learning communities, including her original clinic community, a minimally invasive surgery community, and an obesity prevention and treatment community. She uses tracking and improvement tools that are made available through the learning communities, and because of her outstanding results, she is asked to lead the clinic's bariatrics program. After many years, she decides that she would like to focus her efforts more on prevention. She leaves her practice and returns to the health professions school as a full-time learner. Her competencies are assessed and a special yearlong learning program is devised that will allow her to retrain as a primary care provider. She completes this program, becomes credentialed, and returns to her original clinic as a primary care attending.

Back to Top | Article Outline

Example 2

Daniel is a 21-year-old student who enters the new health professions school after three years of undergraduate school. He grew up in a small rural community and had always wanted to return there as a pharmacist. He completes all of the required core modules and is able to differentiate quickly toward a pharmacy-specific curriculum. As he progresses through the school, he is eager to oversee his clinic's formulary, perfect its computerized order entry system, and initiate other improvement projects that reduce the risk of adverse drug events. These projects are so successful that they are adopted by other clinics in the system. After attaining the competencies required for his doctor of pharmacy degree, he returns to his home town and practices with a physician, a nurse practitioner, and a dentist in a rural health clinic. He adds tremendously to the clinic with his insights into process management and his ability to counsel patients about their medications. After several years, however, the group decides that clinic flow could be improved if Daniel also had the ability to prescribe. He returns to the health professions school as a part-time student, taking a series of modules distributed across 18 months that allows him to continue working while attaining additional competencies required for credentialing as a family nurse practitioner.

Back to Top | Article Outline

Example 3

Roger is a physician alumnus of the health professions school who completed his postgraduate training in medicine and cardiology 10 years earlier. He practices in a medium-sized midwestern town, where he sees a large volume of patients with heart failure. Using tools that are available through his cardiology learning community, he measures quality indicators and finds that his patients have a relatively high rate of hospital admission. He analyzes potential reasons and determines that his patients are not doing as well as expected after transcatheter implantation of stem cells, a treatment that was proven to be effective after he completed his fellowship. He sends his patient data along with recordings of selected procedures to a health professions school faculty member for review. The faculty member suggests advanced training in the technique, so Roger returns to the health professions school for a four-week course that includes an overview of the pertinent basic science, a review of selection criteria, intense simulation experience, and faculty-assisted cases on real patients.

Back to Top | Article Outline

Obstacles to Implementing the Vision

We can imagine many potential obstacles to the implementation of this model, originating from within institutions as well as from the external organizations that regulate the professions and the services they deliver. The first local hurdle to be cleared would be political. The creation of the competency-based, modular, “just-in-time” curriculum outlined above and the technologic innovations required to deliver it will demand an enormous amount of faculty and staff effort, which in turn will require the rechanneling of finite institutional resources, including money and space as well as people. To garner support, the case must be made that change in our education system is critical to solving the health care crisis and, thus, that educational innovation is just as worthy of investment and risk as innovation in research and technology. In addition, even minor curricular revisions at medical schools tend to meet with strong faculty resistance. Disruptions of the magnitude we envision would require skilled facilitation by a team of dedicated champions who understand the science of organizational change. Even with political hurdles cleared, financial obstacles would remain. Although grants and foundations might support the initial phases of planning and implementation, could the model be sustainable? Would the work that learners contribute to health care delivery have enough value to offset tuition, and with a teaching mission that might slow the pace of clinical encounters, can the learning homes generate enough income to be self-supporting? Finally, current faculty providers may resist changes in continuing education systems that would require them to attain new competencies related to practice-based learning and improvement and the use of information technology.

The external obstacles facing this model are those posed by the organizations that regulate health professions education and the delivery of health services. Can the educational programs we advocate be designed so that they still satisfy the standards set by the various accreditation organizations? While we envision the eventual evolution of a system in which transitions between the current phases of learning and practice become blurred, in the meantime will graduates of these programs be competitive for postgraduate training at more traditional institutions? Will students be able to sit for licensing examinations if they have taken a nonlinear path through the educational process, and will licensing boards allow such students extended time to licensure? Will the use of interprofessional supervisors in the provision of clinical care have any implications for reimbursement from either public or private third-party payers? Although certification in some specialties is moving toward a competency-based approach that encourages workplace learning and practice improvement, a system which aligns requirements for continuing education, certification, and credentialing would provide a much-needed impetus in this direction. The determined support of talented and visionary leaders will be required to negotiate these obstacles on local, regional, and national fronts.

Back to Top | Article Outline

The Model as a Disruptive Technology

In articles related to business, health care, education, and continuing medical education, Clayton Christensen and colleagues16,17 describe “disruptive technologies” as new approaches that emerge in rudimentary forms in unexpected areas of an enterprise and then improve rapidly, eventually displacing the technology that formerly dominated the enterprise. The key finding of their research is that the technology of these enterprises met basic customer needs in simple ways at early points in time, thus making it accessible to new markets. This disruptive phase was followed by waves of “sustained innovation,” in which incremental modifications responded to the demands of high-end users until the product overshot the needs of the majority of users.

From Christensen's perspective, Flexner's model of medical education might have been considered a disruptive technology in 1910, if one uses a broad definition of technology that encompasses systems and organizations as well as tools. The important but minimally disruptive curricular reforms introduced over the past several decades might be considered sustaining innovations that improved delivery but did so within the framework of the deeply embedded preexisting model. As it has evolved, this medical education system has delivered an increasingly sophisticated product that focuses on high-end, high-tech, specialized knowledge while becoming increasingly out of touch with the most basic needs of the population it serves.

We believe that our proposed model represents a disruptive technology. As we have outlined, it faces formidable obstacles, and we do not have answers to all of these obstacles. In their research, Christensen and colleagues found that a disruptive technology will develop and prosper only if it is provided an independent existence, separate from the parent organization with its own governance and guaranteed budget, as the traditional enterprise is not structured to create so novel a product. The success of student-run free clinics throughout the country, including our own, may represent disruptive technologies in our midst, operating as they do outside of the required curriculum but providing what many students consider their most meaningful educational experiences. For academic health centers to lead in the creation of completely new systems, we must be willing to create innovative educational subsidiaries that offer our learners alternative pathways while demonstrating proof of concept. Whereas the initial development phases will most likely take the form of small pilot programs, we believe that as these models mature they should detach from the parent schools and operate with separate budgets and administrative structures. Eventually, we believe these alternate schools would be so effective and attractive that they would replace existing models.

In the coming academic year, Vanderbilt plans to launch a pilot program that will include first-year nursing, medical, and pharmacy students. They will matriculate a month early to participate in an immersion course that will build the culture of their interprofessional teams while introducing them to the fundamental principles of professionalism and health care systems. Students will practice basic clinical skills in simulated settings until they demonstrate the competence needed to perform these tasks in real patient settings. Supervised by faculty representing all three professions and by postgraduate trainees, they will then begin to work in assigned community clinics one half-day a week and spend another half-day reviewing patients' outcomes, developing improvement projects, and discussing the social, cultural, and behavioral determinants of health. During the remainder of the week, they will participate in their regular curricula. We hope this pilot will test the principles that learning can be accomplished in interprofessional teams and that novice learners can add significantly to the care of patients and populations. In addition, we hope that this pilot will inform us about the cost of innovative programming and help us determine whether the model is scalable. Our assumptions are that faculty and residents will be able to meet the productivity targets needed to ensure sustainability and that fewer support staff will be needed as students take on more responsibilities for patient care and clinic management. It may take several years to validate these assumptions and determine whether the savings generated by leaner staffing models would be enough to reduce tuition costs.

Back to Top | Article Outline


The model for health professions education described above represents a radical departure from Flexner's model, but the complexity of contemporary health care delivery and the rapidly increasing rate of change demand a radically different approach. Framed on competency-based, interprofessional, continuous workplace learning, we believe the model has the capacity to sustain constant improvement in our systems of care and constant growth in our learners. We have little doubt that as we implement this model, it will itself become a product of workplace learning, shaped by factors that we cannot predict or even imagine, just as Flexner could not have possibly imagined the factors that influence us now.

Funding/Support: None.

Other disclosure: None.

Ethical approval: Not applicable.

Back to Top | Article Outline


1 Flexner A. Medical Education in the United States and Canada: A Report to the Carnegie Foundation for the Advancement of Teaching. Bulletin No. 4. Boston, Mass: Updyke; 1910.
2 Plsek PE, Greenhalgh T. The challenge of complexity in healthcare. BMJ. 2001;323:625–628.
3 Fraser SW, Greenhalgh T. Coping with complexity: Educating for capability. BMJ. 2001;323:799–803.
4 Marsick VJ. Learning in the workplace: The case for reflectivity and critical reflectivity. Adult Educ Q. 1988;38:187–198.
5 Matthews P. Workplace learning: Developing an holistic model. The Learning Organization. 1999;6:18–29.
6 Moore DE Jr, Pennington FC. Practice-based learning and improvement. J Contin Educ Health Prof. 2003;23(suppl 1):S73–S80.
7 Margolis PA, Lannon CM, Stuart JM, Fried BJ, Keyes-Elstein L, Moore DE Jr. Practice based education to improve delivery systems for prevention in primary care: Randomised trial. BMJ. 2004;328:388.
8 Dornan T, Boshuizen H, King N, Scherpbier A. Experience-based learning: A model linking the processes and outcomes of medical students' workplace learning. Med Educ. 2007;41:84–91.
9 Davis DA, Mazmanian PE, Fordis M, Van Harrison R, Thorpe KE, Perrier L. Accuracy of physician self-assessment compared with observed measures of competence: A systematic review. JAMA. 2006;296:1094–1102.
10 Eva KW. Self-assessment in the health professions: A reformulation and research agenda. Acad Med. 2005;80(10 suppl):S46–S54.
11 Eva KW, Cunnington JP, Reiter HI, Keane DR, Norman GR. How can I know what I don't know? Poor self assessment in a well-defined domain. Adv Health Sci Educ Theory Pract. 2004;9:211–224.
12 Mann K, Gordon J, Macleod A. Reflection and reflective practice in health professions education: A systematic review. Adv Health Sci Educ Theory Pract. 2009;14:595–621.
13 Deis JN, Smith KM, Warren MD, et al. Transforming the Morbidity and Mortality Conference into an instrument for systemwide improvement. In: Henriksen K, Battles JB, Keyes MA, Grady ML, eds. Advances in Patient Safety: New Directions and Alternative Approaches. Vol 2: Culture and Redesign. Rockville, Md: Agency for Healthcare Quality and Research; 2008.
14 Oyler J, Vinci L, Arora V, Johnson J. Teaching internal medicine residents quality improvement techniques using the ABIM's practice improvement modules. J Gen Intern Med. 2008;23:927–930.
15 American Medical Association. The Physician's Recognition Award and Credit System. Chicago, Ill: American Medical Association; 2006.
16 Christensen CM, Bohmer R, Kenagy J. Will disruptive innovations cure health care?. Harv Bus Rev. September–October 2000;78:102–112, 199.
17 Hwang J, Christensen CM. Disruptive innovation in health care delivery: A framework for business-model innovation. Health Aff (Millwood). September–October 2008;27:1329–1335.
© 2010 Association of American Medical Colleges