How can medical education be better designed to address the current complaints that new doctors are less prepared for residency and practice than society expects? Critics of medical education systems have alleged that these shortcomings are to blame for mismatches between society's medical needs and the specialties new doctors choose,1 and for graduating people who are not adequately prepared for residencies and practice.2–4 Langdale and colleagues2 go further in predicting that as funding pressures increase, residencies will expect an entry level of skill and competence greater than that which schools are currently providing. The frequency with which reforms have been introduced and the similarity of the recommendations on which they are based suggests that the medical community has agreed that there are problems in the current system, but that it has yet to arrive at a satisfactory solution for present and future needs.5–7
Recent reform proposals have emphasized changing curriculum content (e.g., adding courses in medical ethics, communications, and patient interviewing)8 and pedagogical approach (e.g., switch from lecture format to case studies).8–11 Others have called for the lengthening of medical school because of tremendous advances in medical science and the greater corpus of knowledge that is considered essential for physicians to possess.12–14
We did not write this article to join others in assessing the performance of the medical education system. Rather, our goal was to gain a better understanding of the nature of the problems already reported in the literature. By describing and analyzing a case study we conducted, we attempt to demonstrate that the absence of a systems approach in a curriculum results in a cascade of negative effects. Curriculum-design decisions made with regard to the individual program components only (e.g., the clerkships, and each activity within the clerkships) are not necessarily attuned to the objectives of the school.
Importance of a Systems Approach
Medical education reforms usually focus on the addition (but seldom the subtraction) of particular topics or focus on how specific topics are taught, without addressing medical education as a longitudinal process—a system of interdependent elements in which relationships among the elements must be actively managed over time. In contrast, there is ample evidence that the most outstanding organizations in other demanding, high-tech, knowledge-intensive industries manage complex processes such as design, production, engineering, logistics, and training both in terms of the pieces of the system (e.g., how individual people do work) and in terms of the relationships among the pieces. Complex systems, once the layers of complexity have been peeled away, operate according to a simple set of rules. The rules that govern complex learning systems within the manufacturing industry provide a compelling set of guiding principles for medical education reform. This approach to managing systems involves (1) defining objectives for the organization in terms of the characteristics (e.g., volume, variety, features) of the products or services that they provide; (2) designating who should do what specific work in what sequence for what other person to achieve organizational objectives; (3) designating characteristics for products or services one person provides, given the capabilities and needs of those processes that precede and follow; and, finally, (4) determining how the individual work assignments should be done given their relationship to the system overall, with feedback possible at every stage. The system's performance at each of these four stages is continuously monitored by way of embedded tests, and, if the system's performance is found wanting, managers then have latitude to redesignate each of these aspects of the system rather than tinkering with the components alone.15,16
By employing such an approach, which looks at how the interactions among elements of the system should be managed, outstanding organizations avoid the potential disconnects between managing parts of the system and managing the system as a whole. It is critical that processes be designed and managed not in isolation but as elements of an interdependent system. If high-level objectives have not been defined, then the production tasks to meet them cannot be accurately defined, the roles and responsibilities associated with those tasks cannot be defined and properly assigned, and so forth throughout the system of processes. Rather, the managers are forced into a reactive mode, coping with whatever demands are placed on them and contending with whatever comes their way in terms of instructions and supplies. In this scenario, the value of changing the system at the local level is short-lived due to the uncertainty that managers must contend with.
In medical education, applying an approach that looks at how the interactions among elements of the system should be managed could serve to overcome disconnects between varying components of the learning experience, and better clarify the relationships between objectives and pedagogy. Opportunities for improvement in medical education ought not to be constrained by questions such as whether or not a medical student should have a particular clerkship, or how that clerkship should be designed. Rather, opportunities for improvement should also be shaped by (1) a definition of the skills and knowledge that should be acquired in medical school generally and in clerkships specifically, (2) the order in which skills and knowledge should be conveyed, and (3) how students should be evaluated within each stage to ensure a level of preparation that respects the learners’ needs without compromising the needs of patients whose care provides educational experiences. Finally, the school must determine (4) how students should be trained within each educational increment, given the objectives assigned to that component in relation to its overall role within the educational process. If clear standards are understood at every stage, the faculty and students can work to close the gap between the students’ incoming level of competency and the desired learning encounters.
A Case Study
We now describe a case study to provide a context for identifying the disconnects that hinder learning and stifle improvements within the current system. We first outline the method we used to obtain our data. We then present the case study itself, which shows how a typical student experienced the third and fourth years of medical school with respect to the content and sequence of her clerkships. Finally, we draw on the experience of high-performing organizations in other industries to provide principles for generating learning-based improvement. We conclude with suggestions for how these principles might be applied to medical education.
Obtaining the Data
The case study of Emily Wilson, a fourth-year medical student at Worthington—a well-regarded, private medical college*—is based on three interviews, totaling seven hours. While drawing conclusions from a single case presents limitations, Wilson's experience is likely representative of that of other medical students. Wilson never suggested that she perceived her experience as unique relative to that of her Worthington classmates. Moreover, we interviewed ten other medical students and housestaff from seven different medical schools around the country, who shared similar narratives. In addition, our interviews with five medical educators about their perceptions of medical education's strengths and weaknesses painted a picture similar to that imparted by Wilson. Finally, we vetted this case with 68 medical educators in a session of the Harvard-Macy Physician Educator course and, when polled, 64 agreed that Wilson's experience was similar to those of their own students. (The 68 physician–educators were from 31 institutions, and 16 were from countries outside North America.)
The Wilson case study focuses on third- and fourth-year clerkships because educators whom we interviewed perceived these as less homogenous and predictable than the first and second years and more impervious to reform despite documented need for change. However, problems evident in Wilson's clerkship experience may have analogues in other stages of medical training.
Emily Wilson majored in philosophy and minored in chemistry in college, and fulfilled ten premedicine requirements. In her first two years of medical school, Wilson completed her basic-sciences courses and passed Step 1 of the United States Medical Licensing Examination. Prior to clinical clerkships, she attended a half-day orientation at Worthington that discussed professional manner, health concerns, and whom to contact if she had questions (usually the clerkship director). There were also 15 minutes of overview on the first day of each new rotation by instructors.
Wilson and her classmates each had idiosyncratic experiences in their third year. Wilson completed five clinical clerkships, starting with internal medicine and continuing onto obstetrics, pediatrics, psychiatry, and surgery, delaying family medicine to the fourth year because of a broken arm. A classmate began with psychiatry and family medicine before taking surgery, pediatrics, and obstetrics, with internal medicine at the end. A third was assigned to obstetrics and psychiatry before internal medicine; then she took a dermatology elective while delaying family medicine to the fourth year, and concluded with pediatrics and surgery. That clerkships were assigned based on openings at the 11 teaching hospitals affiliated with the school added extra variation.
Additional sources of variation among the students’ experiences arose because of the way in which teams—which included attending physicians responsible for particular patients, housestaff (residents and interns), and three to five medical students—were managed. First, due to idiosyncratic clerkship assignments, Wilson was never teamed with classmates with whom she had previously trained. Second, responsibility for Wilson's training varied across rotations. In some, medical students did morning rounds with an assigned attending physician. In others, no attending had overall responsibility for the students, so learning depended on the particular relationships the students had with housestaff at various stages of their own training.
Little continuity existed even within a rotation among those responsible for teaching and providing evaluations and those being trained, since attendings, housestaff, and students were scheduled according to different calendars. For instance, Wilson was on a six-person team during internal medicine, during which the chief resident and the fourth-year student changed. During obstetrics and gynecology, Wilson was on an eight-person team; after two weeks, one of two residents and both interns changed. One week into pediatrics, the chief resident changed. Halfway through surgery, the third-year resident and three of four interns changed. Only during psychiatry did Wilson's team remain intact.
Wilson's fourth year was also structured around eight clerkships, four of which were electives. These were for preparing for residency—in Wilson's case, anesthesiology—or for studying subjects outside intended specialties, and electives were drawn from approximately 200 offered through Worthington and thousands more through other institutions.
Wilson and her classmates placed heavy emphasis on fourth-year clerkships and evaluations, believing that they would affect dramatically what residency they matched with. However, inconsistent interactions with senior physicians contributed to Wilson's perception that grading and evaluations were arbitrary. For instance, during Wilson's first anesthesiology rotation, she was assigned to shadow different physicians each day. Wilson and her classmates speculated that because of the limited contact students had with any doctor, assessments from the unit's 45 attendings and 24 residents depended more on subjective factors like personality, timeliness, enthusiasm, and curiosity than on objective factors such as clinical ability and longitudinal improvement.
To have time for residency interviews, Wilson chose a radiology clerkship with two one-hour lectures per day, five days per week, for four weeks. A different instructor gave each lecture, except for those on Tuesday morning, which the clerkship director taught. Because the instructors were different and attendance was not taken, Wilson and her peers postulated that students who knew the director and attended his sessions received an honors grade and that students whom the director saw in lecture but did not know well received a pass grade.
Not all clerkships seemed as impersonal, with evaluations as arbitrary. Wilson's pain management rotation team included the clerkship director. For a surgery ICU rotation, Wilson was on a team that included the ICU's attending, who supervised the clerkship, a second-year resident, an intern, and another fourth-year student. Staffing was stable throughout the rotation, and it was better run than other clerkships, in Wilson's view. The director took the unusual step, from Wilson's perspective, of issuing manuals to students and residents that included his expectations for them and had articles on the concepts the attending considered to be important.
In reading about Wilson's clerkships, most readers will recognize that the haphazard nature of her training does not do justice either to her or her preceptors. Some may argue that the answer lies in changing the clerkships; while well intentioned, this approach hardly addresses the systemic problem or its cascading effects. These effects originate with the failure to adequately define what skills and knowledge a student should have mastered at the end of each stage in the curriculum. This leads those responsible for teaching into difficult territory, left to make decisions in isolation about how and what to teach, to offer experiences as they arise, and to hope against hope that enough variety of experience will compensate for the lack of deliberate planning.
What is noteworthy in the case is the dominant role that chance played in determining what Wilson learned, where, how, with, and from whom. Certainly, we might expect that the fourth year, which was dedicated in large part to electives and residency auditions, would be idiosyncratic to each student and that fourth-year clerkship directors and educators would have to be prepared to teach students who were different both by preparation and by objectives. However, the third-year experience, for students and educators alike, was also shaped by the chance assignment of students to clerkships. The third year presented 720 possible sequences for Wilson's classmates to complete six required clerkships. In a class of approximately 150, then, no one was likely to encounter material in the same order as anyone else. That clerkships were at one of 11 hospitals with a large pool of housestaff and faculty at each, many operating according to calendars independent of those of clerks, further increased the chance that each student's education was unique.
What is problematic about this approach becomes clear when we consider the oft-heard comments of attendings and housestaff, who report that when new students arrive for a clerkship, some educators assume that students know nothing and that they must start from scratch. As cynical as this might sound, it is a rational response that points to the major concern of this article: that the lack of a system to guide the medical school experience makes it difficult to structure and improve the education process so that the students achieve clearly defined core competencies in a professional manner.
With little ability to anticipate levels of preparation or to foresee expected accomplishment, educators are disadvantaged in designing clerkship experiences that are optimal in terms of the students’ overall development. Even though what a student is prepared to learn in a pediatrics rotation, for example, would seem to depend on whether that student has had family medicine previously, educators and administrators have little choice but to treat each clerkship as if it were in isolation, since the alternative of preparing a course of study appropriate for each student would be impossible given the large number of possible combinations of experiences both before and after. That educators must also account for the possibility that a student's experience may have been at one of 11 hospitals, each with its own idiosyncratic case load, staff, and teaching emphasis, multiplies the number of contingencies educators must prepare for. Therefore, assuming that students know nothing does not necessarily reflect a condescending attitude. Rather, because dialogue among instructors about what happened in previous stages and what is supposed to happen in future stages is precluded by system complexity and unpredictability, it may be a risk-averse position to take when planning.
Treating each student as a tabula rasa, however, carries its own risks. Students might learn the same few things over and over. While repetition helps to reinforce previously learned knowledge and skills, redundancy in which students relearn what they have mastered on a prior occasion may be viewed as wasteful in a costly system, particularly if it comes at the expense of learning something else important. In contrast, faculty who mistakenly assume mastery of prerequisite skills that are not yet part of the students’ repertoire create expectations that cannot be met by the students and which may be destructive for faculty and patients as well. Our case study focuses on how this scenario affects students; it is not difficult to imagine as well the frustration felt by the educators attempting to work within this scenario.
Principles for Managing Complex Systems
The industrial domain may seem an unusual place to look for organizational or management models appropriate to education generally and medical education specifically. Manufacturers typically transform materials whereas educators transfer knowledge and imbue students with important attitudes—they transform people. In addition, particularly in high volume manufacturing, the variety of output/demand may be far less than the variety of patient needs in a clinical setting. Despite these distinctions, however, similarities between the two domains make relevant to medical education the application of approaches used by outstanding manufacturers. Notably, success in both—be it delivering outstanding products, services, or students—depends on learning.
Most outstanding manufacturers make systemic adjustments quite effectively, and their success depends on the presence of learning throughout the system, and the ability to examine gaps and disconnects immediately to enable the highest performance possible. Within a wide variety of industries, companies compete to meet the needs of the same market by making similar products using nearly identical production technology. Yet some excel by a significant margin along multiple performance metrics (quality, productivity, flexibility, profitability, and safety) that is sustained over years if not decades. Often, these differences result from the superior rates at which the leaders are able to learn about and improve the complex processes they use to create products and services. These differential rates in learning can express themselves as differences as considerable as two or three to one in productivity, ten to one in safety, and 1,000 to one in quality.17–23 The manufacturers that achieve the shortest lead time, highest quality and productivity, and best safety record are those most capable of ongoing learning—that is, those dedicated to rapidly modifying and improving their processes based on their experiences and outcomes. They learn because clarity and specificity are built into the system, allowing them to identify areas requiring remediation and make the appropriate changes, with the certainty that the adjustments will resolve the problem. This ability to correct can only occur in a carefully designed pathway.
Therefore, despite the differences between education and manufacturing, the fact that the best manufacturers lead because of how they learn invites us to look to manufacturing for transferable lessons on how to organize complex systems that use learning to achieve high performance.
Toyota is an example of a company that has exhibited leading rates of improvement and learning and superior performance. The company's approach is to specify (1) what products or services the system (e.g., work cell, unit, department, section, factory, or plant) as a whole is to provide to whom; (2) how responsibilities will be assigned within the system in terms of who will provide what product, service, information, training, and the like to whom; (3) how interactions and interdependencies will be managed in terms of handoffs of products, services, and information between linked work activities; and (4) how individual work will be done. These designs are repeatedly tested as they are used, and when difficulties are encountered, problem solving is triggered immediately.15,16
The clerkship learning process that Emily Wilson encountered ran counter to this approach. The clerkship sequence was defined only ex post. Therefore, those who managed the clerkship process as a whole relegated it to those responsible for the individual system-elements (i.e., clerkship directors, residents, or attendings) to adjust accordingly across the wide range of inputs that they received. Housestaff and attendings had to be able to simultaneously train students who had had a wide range of prior experiences with a wide range of subsequent learning events. Educators and administrators, denied the opportunity to establish reasonable expectations of what students should know coming into a clerkship and what they should master by a rotation's completion, instead must engage in teaching that is less well connected to students’ needs. The result may then become an ad hoc approach to teaching, based on opportunistic responses to stochastic events, leading to evaluations and reviews that are perceived to be arbitrary. While relying on the regular flow of patients to provide adequate learning opportunities as compensation for the lack of an adequate system design may have been appropriate when the current system was anchored in mentor–protégé, master–apprentice relationships, it is problematic now.24
The simple rules underlying complex manufacturing systems are instructive as they relate to instituting clarity of both purpose and outcome in the elements of the medical education curriculum, while minimizing the strains placed on patients, educators, and students. While viewing the medical school experience systematically in this manner is but one approach to the current problem, we believe that introducing specification into medical education to achieve clarity is ultimately liberating rather than constraining for the schools because, as in the manufacturing environment, it creates a flexibility that enables improvements to be made more efficiently.
The rules that guide manufacturing systems can be applied to medical education reform efforts as outlined below.
System Rule: Define the skills and knowledge a student should be able to competently and professionally demonstrate upon completion of his or her education.
Applying this rule clearly establishes what it means for the school to be successful in terms of educating a student. Medical educators, in large part, are already moving in this direction, in concert with the work of numerous organizations—including the Association of American Medical Colleges,25 the Accreditation Council for Graduate Medical Education,26 the Institute for International Medical Education,27 and the World Federation for Medical Education28—that have identified and recommended to medical schools core objectives and competencies to meet minimal essential standards of practice, while still allowing the freedom to diversify beyond the minimum. What is missing at this juncture is a strong connection between what students should master and how that mastery should be acquired. Determining how is the realm of the next rules.
Pathway Rule: Develop a transparent curriculum and assign responsibility for delivering subsets of specific skills and knowledge to particular clinical educators working in a well-defined and understood sequence.
Applying this rule involves designing a specified sequence of learning modules—making clear who is responsible for transferring what knowledge to students—in which students encounter certain types of clinical situations prior to and as preparation for more complex or specialized situations.
Schools that have developed clinical courses that represent a planned developmental sequence of clinical skills teaching in multiple settings, insuring mastery of basic skills before requiring performance in more complex or critical skills, are working in ways consistent with this rule. Another alternative guided by a systems approach might involve forming groups of students who would progress through the same sequence at the same time—rather than each student having an idiosyncratic experience—with faculty who understood the sequence and its predetermined activities and outcomes.29 If a planned and carefully specified sequence were developed, students could play a more significant role in monitoring their own professional growth, and develop habits of lifelong learning.
Schools in which objectives have been specified for each clerkship have taken a partial step toward specifying a sequence in which skills and knowledge are to be acquired. However, creating a transparent pathway for students and faculty will require a level of planning that goes well beyond the definition of objectives for isolated blocks of time. Learners should experience a developmental educational process that is monitored with success in mind and mentored with clarity of purpose so that teachers and learners will know what is expected of them at each stage of progress through the sequence.
Connection Rule: Define criteria for determining whether a student is prepared to advance from one learning stage to the next.
This rule is perhaps the one most violated in the case of Emily Wilson, wherein transitions between clerkships and mentors are haphazard. To correct this problem, medical schools might specify a sequence of modules that allows the directors of each to establish expectations as to what each ought to deliver, given the knowledge and skills that incoming students have and what, therefore, students must learn. Orientations that define expected roles and skills for new work along with assessments that ensure successful completion of prerequisite skills would be basic to meeting the educational purpose of this rule, and the systematic assessment exercises that curriculum reformers have attempted to embed throughout the educational continuum would make timely remediation, when required, more likely.
Work Activity Rule: Develop each specialty's pedagogy in light of the educational responsibilities assigned to it (by application of the Pathway Rule) and the level of preparation students are expected to have upon entering and exiting each stage (by application of the Connection Rule).
Were this rule applied, the current focus on work within the individual elements of medical education would continue to enrich each learning activity in ways consistent with the philosophy of the school. For example, in order to overcome the variation in clinical experiences, some clerkship directors are developing teaching sections to compensate for clinical teaching that may not occur during a given clerkship because certain common medical problems are not encountered.
This approach creates flexibility that enables improvements to be made more efficiently, both for individual students and the process as a whole. For instance, the clarity such an approach instills proves crucial when a student struggles to meet the criteria for advancement from one stage to the next; this is evident immediately rather than at some distant point, and educators can provide additional learning opportunities in a very responsive, targeted way—thereby avoiding the associated negative consequences to students, educators, and patients alike. However, the case may arise where the system is at fault. If a sizable portion of students are found to be having difficulty, this would provide a clear signal to those responsible for managing the system that there need be a reassessment of what gets taught, when, by whom, and what the expectations might be for handing off students from one stage to the next. † Such remediation could occur quickly enough to have an impact on each student's learning experience while simultaneously informing future program implementation.
We believe that the complex systems of quality-driven industries offer pertinent lessons to medical educators engaged in curriculum reform. When work and learning systems are managed so that what is expected to occur is clear, with mechanisms to identify where expectations are not met, and with protocols and resources to make improvements when problems develop, then customers, shareholders, and employees all benefit from increased quality, productivity, reliability, and safety. Or, translated into medical education, patients, physician educators, administrators, and students all benefit. In contrast, the lack of specification at all levels in the case study of Emily Wilson makes it hard to establish expectations, and difficult for everyone who is part of that system to do their work and influence the systems within which they work. Any system that handicaps those who are part of it from doing their best work is inherently disrespectful to all involved. This need not be.
Our challenge in curriculum reform is to address all levels of the medical education system. If we focus on individual courses or activities without examining the relationship between these elements, we deny ourselves the opportunity to create the substantive reform we all seek. While many medical schools have begun to specify order to the clerkship sequence, the Emily Wilson case demonstrates that specificity of the learning modules within each clerkship, student assessments, and the hand-offs between clerkships remain problematic. As students progress through a series of clinical learning modules, specifying what students should have learned in a preceding module will make it easier to define incoming students’ baseline of knowledge and skills and what that module should convey to the students, and to identify those students or those topics that are not getting sufficient attention. Then, the design of the individual modules, the handoffs between them, and the sequencing of them can be modified so that students achieve higher competency, possibly at less cost and in less time. While the optimal sequence in which clinical experiences might be completed will unlikely be established on the first try, each experiment with the order will create the opportunity for improvement. In fact, we have assumed that the basic clerkship process may be immutable in the short-term. It may be that by managing the medical education process so that educating students is tightly integrated with bettering the process of educating students, some schools will discover alternative approaches that are less bound to the legacy structures with which we are currently working. This is a far cry from approaches that are based, unrealistically, on “starting with a clean slate,” as if the current system would suddenly disappear. Rather, this approach relies on recognizing that the current system is the de facto starting point, and that a far better system can be achieved through incremental, but frequent, experimentation and respecification.
Ultimately, utilizing this systems approach will provide patients, students, and educators a safer, more respectful environment where high-quality health care delivery and education can coexist.
The authors acknowledge the scholars of the Harvard-Macy Institute for their willingness to experiment with the ideas expressed in this article during teaching sessions of the Institute's two annual programs, and for their insightful feedback. The authors also acknowledge Dr. Clayton Christensen of Harvard Business School for inspiring the design of our case study.
1. Adams D. Student debt helps drive students into specialty matches. Am Med News. 2003;46:10–2.
2. Langdale LA, Schaad D, Wipf J, Marshall S, Vontver L, Scott CS. Preparing graduates for the first year of residency: are medical schools meeting the need? Acad Med. 2003;78:39–44.
3. Finocchio LJ, Bailiff MA, Grant RW, O'Neil EH. Professional competencies in the changing health care system: physicians’ views on the importance and adequacy of formal training in medical school. Acad Med. 1995;70:1023–8.
4. Graber DR, Bellack JP, Musham C, O'Neil EH. Academic deans’ views on curriculum content in medical schools. Acad Med. 1997;72:901–7.
5. Enarson C, Burg FD. An overview of reform initiatives in medical education, 1906 through 1992. JAMA. 1992;268:1141–3.
6. Ludmerer K. Time to Heal: American Medical Education from the Turn of the Century to the Era of Managed Care. Oxford, New York: Oxford University Press, 1999.
7. Windish DM, Paulman PM, Goroll AH, Bass EB. Do clerkship directors think medical students are prepared for the clerkship years? Acad Med. 2004;79:46–61.
8. Bloom SW. Structure and ideology in medical education: an analysis of resistance to change. J Health Soc Behav. 1988;29:294–306.
9. Mann K. Thinking about learning: implications for principle-based professional education. J Contin Educ Health Prof. 2002;22:69–75.
10. Pratt DD, Collins JB. Reconsidering “good teaching” across the continuum of medical education. J Contin Educ Health Prof. 2001;21:70–81.
11. Tosteson DC, Aldenstein SJ, Carver ST. New Pathways to Medical Education at Harvard Medical School. Cambridge, Massachusetts, London: Harvard University Press, 1994.
12. Whitcomb ME. The general professional education of the physician: is four years enough time? Acad Med. 2002;77:846.
13. Nutter D, Whitcomb M. The AAMC Project on the Clinical Education of Medical Students. Washington, DC: Association of American Medical Colleges, 2002.
14. Weinberger S, Whitcomb M. The clinical education of medical students: report on millennium competencies I & II. Carl J Shapiro Institute for Education and Research, Harvard Medical School and Beth Israel Deaconess Medical Center, AAMC, 2001.
15. Spear SJ, Bowen HK. Decoding the DNA of the Toyota Production System. Harv Bus Rev. 1999;77:96.
16. Spear SJ. The essence of just-in-time: imbedding diagnostic tests in work-systems to achieve operational excellence. Product Plann Cont. 2002;13:754–67.
17. Cusumano M. Manufacturing innovation: lessons from the Japanese auto industry. Sloan Manage Rev. 1988;30:29–40.
18. Garvin DA. Quality on the line. Harv Bus Rev. 1983;61:65, reprint 83505.
19. Hayes RH, Clark KB. Explaining observed productivity differentials between plants: implications for operations research. Interfaces. 1985;15:3–14.
20. Hayes R, Wheelwright SC, Clark KB. Dynamic Manufacturing. New York: Free Press; London: Collier Macmillan, 1988.
21. Leonard-Barton DA. The factory as a learning laboratory. Sloan Manage Rev. 1992;34:23–8.
22. Lieberman MB, Lau LJ, Williams MD. Firm-level productivity and management influence: a comparison of U. S. and Japanese automobile producers. Manage Sci. 1990;36:1193–1215.
23. Womack JP, Jones DT, Roos D. The Machine That Changed The World. New York: Rawson Associates; Toronto: Collier Macmillan; New York: Maxwell Macmillan, 1990.
24. Jolly B. Charting the course: designing a medical curriculum. In: Jolly B, Rees L. eds. Medical Education in the Millennium. New York: Oxford University Press, 1998:22
25. Medical School Objectives Writing Group. Learning Objectives for Medical Student Education—Guidelines for Medical Schools: Report I of the Medical School Objectives Project. Acad Med. 1999;74:13–8.
26. Accreditation Council for Graduate Medical Education (ACGME). Outcome Project: enhancing residency education through outcomes assessment 〈http://www.acgme.org/Outcome/
〉. Accessed 20 January 2004.
27. Core Committee, Institute for International Medical Education. Global minimum essential requirements in medical education. Med Teach. 2002;24:130–5.
29. Green, Gordon J, dean of basic science and medical education, St. Matthew's University School of Medicine (Grand Cayman, British West Indies). Personal communication, June 16, 2002.
*Names are disguised to protect confidentiality.
†Even with existing constraints that might make a specified linear approach difficult, the current system could be simplified with a specified “ring” approach. At the start of the year, students could be assigned across the specialties such as internal medicine, surgery, pediatrics, and so on, thereby respecting the current approach of everyone starting in July and all clerkships being evenly loaded with students throughout the year. However, in contrast to current practice, every four weeks, students would advance in the same order; for instance, internal medicine always clicking forward to family medicine and then to pediatrics, and so on. This would mean there would be far less variety for which educators would have to adapt than the 720 possible sequences currently available at Worthington.