Over the last 30 years, many physicians likely received some part of their clerkship or residency training on internal medicine medical teaching units (MTUs) within university hospitals.1,2 Clinical education units similar to MTUs exist for many specialties, and comparable teaching models are used around the world.3–5 The ward attending physician, or preceptor, plays a central role in the MTU. She or he cares for ill patients, teaches learners of various skill levels, and leads multidisciplinary teams of health care providers. Accountable to many stakeholders and highly visible within the health care and medical education systems, preceptors instruct and act as role models for the future physician workforce.6
Despite the paramount importance of this role, there is scant empirical evidence that attending physicians can use to guide their performance as preceptors. Medical education frameworks describe ideal physician attributes but were not developed to guide the multidimensional preceptor role specifically.7 The main focus of research and faculty development regarding the preceptor role has been the clinician–educator component of the work.8,9 Direct observation of preceptors in practice has been limited to studies of their teaching abilities.10–12 As health care and medical education systems evolve,13,14 the attending physician’s role likely will transform to adapt to the new environment. The context in which teaching, learning, and clinical practice occur is now recognized as a powerful influence over these important activities.15 Academics advocate for an updated and more holistic perspective on the preceptor role that includes the consideration of the preceptor’s complex work tasks and environments, demographics, and training, as well as the influence of humanism and physician wellness on performance.16–22 Recent worldwide emphasis on competency-based medical education training, which includes entrustable professional activities, described as units of professional practice, further demands evidence of physicians’ real-world work.23
A current, comprehensive, evidence-based view of the attending physician preceptor role is lacking. Observing physicians during their workday provides a unique opportunity to mirror their behaviors back to the physicians, the medical education system, and the health care system. Given that physicians rely on evidence-based practice, the provision of empirical knowledge about their behaviors in important roles may help to drive change where needed. The purpose of this study, then, was to generate an empiric, detailed, and updated view of the attending physician preceptor role and its interface with the complex work environment.
We used a constructivist approach, whereby knowledge is constructed through the interaction of the researcher and research participants, with an inductive reasoning methodology (i.e., hypothesis construction research process).24 We gathered data using observations of preceptors in real-world MTUs from January through November 2013.25 We undertook a collective ethnography26 with two sets of observers: (1) nine insider observers who were MTU preceptors from university hospitals across Canada and (2) two outsider observers who were social scientists without in-depth knowledge of the preceptor role or environment. With this study design, we expected overlapping perspectives as well as unique contributions from each set of observers, generating a more comprehensive view of the preceptor role.25 Our research team included both MTU stakeholders and social scientists. The principal investigator (J.B.L.) was familiar with the study subjects and setting because of her work as an MTU preceptor, medical educator, and researcher in the field of physician wellness.
Context and sampling strategy
We conducted the study at two Canadian university hospitals that are part of medical teaching centers serving patients of high acuity and complexity (> 1,000 beds at Site A and > 600 beds at Site B). At both sites, there are dedicated MTU wards with patients placed on other hospital wards because of overcapacity or patient care needs. Three MTU teams per site care for approximately 20 patients each. All MTU preceptors (23 at Site A, 12 at Site B, all trained as general internal medicine consultants) were eligible to participate. Each observation consisted of one observer following one preceptor in whatever aspect of work and in whatever work setting she or he was engaged during the observation period (e.g., interacting with the MTU team or colleagues, formally teaching, working alone). A typical MTU team included a rotating core of learners (i.e., senior medical resident, three to four junior residents, and one to three medical students who spent one month with the team) complemented by other team members (e.g., nurses, allied health care professionals) and those from other hospital units (e.g., critical care, emergency room).
A disproportionate sampling strategy stratified by site and preceptor rotation was used to select participants. Observations were arranged at approximately two-week intervals with deliberate variation in factors such as time of day, day of week, month, and timing within rotations. At both sites, we randomly selected one of the three preceptors who was scheduled to work during the projected observation times throughout the study period and approached her or him for consent. If that preceptor refused, we randomly selected one of the remaining two preceptors, and so on. Preceptors could participate once in each of the insider and outsider observations. We felt that the final number of 32 observations (16 per site, 16 per observation set) was sufficient to allow for variation in contextual factors, such as team composition, patient load and acuity, and preceptor experience.
Preceptors were contacted by e-mail and telephone. Written informed consent was obtained. All other stakeholders who were indirectly involved (e.g., patients, health care providers) were informed of the planned observations and could decline to participate. Ethics approval was obtained from the Conjoint Health Research Ethics Board of the University of Calgary.
Data collection and processing
During 2013, observers conducted 32 observations of 26 preceptors (6 participated in both insider and outsider observations), lasting approximately 3 hours each (total 99.5 hours). To ensure consistency in data collection, a social scientist (also an observer) prepared a detailed set of written guidelines, which was distributed in advance to the observers and reviewed in a face-to-face training session. Observers were coached to watch the MTU preceptors as they carried out their work and to provide thorough and accurate descriptions of the activities, events, and interactions they observed. Key aspects of this standardized training included instructing observers how to (1) be unobtrusive; (2) observe but not judge; (3) write brief notes documenting descriptions of physical settings as well as what they see and hear; (4) write comprehensive, detailed field notes after the observations; (5) document initial impressions about the possible meaning of what they observed; and (6) complete a preliminary analysis of their field notes. One nonobserver from the research team (J.B.L.) was charged with reviewing the observers’ work immediately after completion as a further check for consistency in approach to data collection.
Nine insider observers observed 16 preceptors (8 per site) between January and June (total 50.5 hours). Two observations (1 per site, 2 preceptors) were excluded because of incomplete data collection, leaving 44.5 hours. Two outsider observers observed 16 preceptors (8 per site) between May and November (total 49.0 hours). In total, there were 30 observations (93.5 hours) of 24 preceptors included in our analysis. With this sample, data saturation was achieved, whereby similarities of ideas and themes emerged from the descriptions of what was seen and documented.25 The observers’ documentation included extensive handwritten notes of events, analytic memos, and comments. After the observations, usually the same day, the observers wrote detailed field notes based on this documentation. Data collected from each observation were assigned a randomly generated number to mask the identity of the preceptor, observer, and site, and securely stored.
The insider observers observed 6 female (37.5%) and 10 male (62.5%) preceptors with an average age of 40 years (range 31–50) and 8 years of experience (range < 1–18). Thirteen (81.3%) were cohabiting or married, and 13 (81.3%) were parents. Their yearly MTU rotations ranged from 6 to 26 weeks. The outsider observers observed 6 female (37.5%) and 10 male (62.5%) preceptors with an average age of 45 years (range 33–59) and 12 years of experience (range 1–24). All were cohabiting or married, and 14 were parents (87.5%). Their yearly MTU rotations ranged from 4 to 20 weeks.
We analyzed the data collected by each group of observers separately, then we performed a further abstraction that included both data sets to produce a comprehensive description of the preceptor role. Observers performed a preliminary thematic analysis of their field notes generally within 24 to 48 hours of the observation. For the data set collected by the insider observers, the principal investigator (J.B.L.) and a social scientist analyzed the observer-generated preliminary thematic analyses, referring back to the field notes to clarify themes and concepts as needed. This approach was necessary as it was not feasible to have the nine insiders from across the country complete this additional analysis. For the data set collected by the outsider observers, they performed the preliminary thematic analysis of their field notes soon after each observation; then they performed a full analysis of the preliminary data after all of the observations had been completed.
We used an inductive thematic approach, where analyses moved from empirical data to more abstracted ideas and theory building from the bottom up.24 Concrete observations were analyzed without a priori categorization to label, define, and develop themes that emerged from the data. These initial themes were further grouped and abstracted to identify larger groups of phenomena and achieve theoretical understanding.24 Data were stored and managed using NVivo 10 (QSR International, Doncaster, Australia). Differences in coding and categorizing of the data were addressed through discussion to achieve consensus.
A further abstraction was performed on the two data sets through discussion and consensus of all four data analysts (one of whom was J.B.L.). This final empiric, detailed, and updated view of the attending physician preceptor role and its interface with the complex work environment was shared with all observers. A symposium was held in September 2014 to which the observers were invited to review and discuss the study results. These interactions allowed for a form of member checking and triangulation of study results. Following this symposium, we performed axial coding where we searched for connections between categories,25 allowing us to understand the relationships among the phenomena and to identify connections between elements.
Our analysis produced 123 components of the attending physician preceptor role (i.e., subthemes) that abstracted to 14 themes, and further abstracted to 3 overarching elements (i.e., major themes) of the role—competence, context, and conduct. Highlights are presented here, and a detailed analysis is available from the authors. The competence element is defined as the execution of traditional physician competencies (i.e., professional knowledge, skills, and attitudes). The context element is defined as the extended MTU environment. The conduct element is defined as the manner of acting or behaviors and attitudes in the role. We identified multiple connections between these elements, suggesting that all three contribute to preceptor performance. In particular, it appeared that preceptors, while trying to do their work, exhibited behaviors and attitudes that were affected by their adaptability, humanism, interpersonal skills, and wellness, to negotiate their complex work environment.
Chart 1 describes the eight themes derived from the 70 role-related components of the competence element. These themes include (1) clinician, (2) teacher/role model, (3) leader/manager, (4) communicator, (5) collaborator, (6) professional, (7) information manager, and (8) safety architect. The observers saw all of the preceptors executing many or all of these competencies at any given time. This practice often resulted in role fusions or overlaps, with competing or even conflicting aspects at times (e.g., morning review of newly admitted patients timed to coincide with ward patient discharges). Although many of these competencies have been described previously in medical education frameworks,27,28 more detail and nuance emerged from our data. For example, the observers saw the preceptors working not only as managers but also as team leaders, responsible for building, empowering, and supporting the team. They displayed evidence of big-picture thinking and provided a positive influence.
Novel competencies were also identified. The observers saw preceptors as information managers who received, verified, and processed vast amounts of written and verbal information (see Table 1 Section A). In addition, the safety architect theme emerged. The observers saw preceptors safeguarding patients from suboptimal care caused by learner inexperience, protecting health care information from inappropriate dissemination, and guarding the extended health care team from physical harm (e.g., enforcing infection prevention and control techniques) (see Table 1 Section B). The safety architect theme was also featured prominently in the connections between the role elements. The observers’ documentation portrayed preceptors as effectively carrying out these competencies, with sparse examples of concrete displays of ineffectiveness (e.g., telling learners what to do rather than probing knowledge, excluding bedside nurses from relevant patient care discussions).
Chart 1 describes the three themes derived from the 33 role-related components of the context element. These themes include (1) the work, (2) the team, and (3) the institutional setting. The observers described the MTU environment as busy and hectic, with autonomous factors that could be dynamic or static. These factors could either help or hinder the preceptor’s performance.
The work theme included the nature of the work, workflow, and work demands. The observers saw the preceptors’ work as intense, with heavy workloads, diverse tasks, time constraints, and incessant and competing demands. There were difficult patient encounters and frequent transitions of patients. The workflow was often disorderly and disrupted with waits and delays, interruptions, unpredictability, and urgency (see Table 1 Section C). However, routines around patient care and teaching duties also provided structure to the workflow. The work demands on the preceptors were physical, emotional, and cognitive and included unrestricted access to the preceptor by patients, learners, and other health care workers (see Table 1 Section D).
The team theme included membership and structure, lexicon, and teamwork. The observers saw the preceptors’ teams as varied in composition, membership, skill set, and experience, with frequent transitions between members (see Table 1 Section E). Team routines and rituals created a sense of order. Team members used metaphors and medical language to expedite communication around patient care (see Table 1 Section F). Teamwork resulted in cohesion, with easy, cordial interactions and respectful, effective working and personal communications.
The institutional setting theme included the physical space, technology, and the health care system. The observers saw the physical work space as one that required frequent movement between ward locations in a large hospital. There was difficulty locating equipment and a paucity of quiet, private work spaces (see Table 1 Section G). Technology was prevalent, with preceptors using electronic patient care information systems, pagers for communication, and mobile devices for calculations and accessing information. The observers saw deficient system resources, such as poor computer operations and process gaps (see Table 1 Section H).
Chart 1 describes the three themes derived from the 20 role-related components of the conduct element. These themes include (1) negotiating the context, (2) interpersonal skills, and (3) wellness behaviors. The observers saw the preceptors using behaviors, attitudes, and relationships in their workday.
The negotiating the context theme included being adaptable and flexible; juggling roles and responsibilities; balancing efficiency and professionalism; dealing with distractions and interruptions; and using the art of medicine, instinct, and experience. The observers saw the preceptors being highly efficient, by monitoring and managing their time. They prioritized, juggled, and balanced work demands and stakeholder expectations, often integrating responsibilities. They recognized the time required for particular tasks while maintaining quality work. The observers saw the preceptors being creative in their decision making, recognizing when things were not right, finding discrepancies in information, identifying oversights in care, and showing foresight in mitigating potential problems (see Table 1 Section I).
The interpersonal skills theme included the preceptors’ relationality, or deep and invested relationships, with the extended MTU team and learners, with patients, and with oneself, and maintaining positivity. The observers saw the preceptors constantly engaging with the people around them and building relationships, documenting empathetic, compassionate, and respectful interactions. They shared reflections on past experiences or mistakes with learners, offered reassuring words to patients regarding difficult diagnoses, and celebrated learners’ achievements (see Table 1 Section J). The observers saw the preceptors displaying satisfaction and enjoying their work.
The wellness behaviors theme included effecting wellness, promoting wellness, and assuming responsibility for wellness. The observers saw preceptors practicing health-conscious behaviors like taking the stairs for exercise. They used humor and irony to diffuse stress and provided and received collegial support (see Table 1 Section K). Preceptors assigned breaks to learners and ensured that they left on time. They recognized learners’ moral distress, offering support and coping strategies. The observers also saw the preceptors permitting barriers to impede their own wellness and allowing professionalism to trump wellness at times as they grappled to fulfill the responsibilities of the preceptor role (see Table 1 Section L).
Connections between the three elements
Multiple connections emerged between the three elements of the attending physician preceptor role; specifically, the conduct element appeared to function as an interface between the competence and context elements. Preceptors seemed to carry out their work within a complex, dynamic environment using behaviors, attitudes, and relationships as navigational tools. We present examples of these connections in Table 2.
Preceptors navigated the tension between patient care and medical education regarding safety (see Table 2 Section A). They showed the ability to sense distress in learners, recognizing when to intervene and ensure both patient and learner safety. This observation demonstrates the connections between the competence (safety architect, clinician, and teacher) and conduct (interpersonal skills and wellness behaviors) elements. Preceptors exercised tenacity and flexibility appropriate to the challenging and variable work environment around them (see Table 2 Section B). They attempted to stick to scheduled tasks despite competing responsibilities but knew when to yield and diverge from the schedule. This observation demonstrates the connections between the competence (clinician, teacher, manager, and professional), context (work and team), and conduct (negotiating the context) elements. Preceptors used the security of routines and rituals to counterbalance an environment fraught with distractions and interruptions (see Table 2 Section C). They maintained focus and relied on frameworks around patient care and team schedules to keep them on track. This observation demonstrates the connections between the competence (clinician, professional, and manager), context (work and team), and conduct (negotiating the context) elements. Preceptors managed the variable impacts of technology and physical space, sometimes at a cost to their physical wellness (Table 2 Section D). They showed focus despite noisy, cramped quarters, and were creative in improvising office and classroom space. They integrated technology to improve efficiency. This observation demonstrates the connections between the context (physical space and technology), competence (teacher, communicator, and clinician), and conduct (wellness behaviors and negotiating the context) elements.
Our study provides an updated, more detailed but also more expansive, and humanized view of the attending physician preceptor role where three overarching elements (competence, context, and conduct) appear to contribute to preceptors’ performance. This complex role is thus not only dependent on preceptors’ mastery of professional skills but also vulnerable to contextual factors that are independent of these skills. We propose that some of this vulnerability is tempered by preceptors’ manner of acting, where their adaptive behaviors, attitudes, and relationships help them to reconcile the work to be done and the influence of the context of that work.
The attending physician preceptor role is functionally nestled within health care and medical education organizations. Process and procedure are often emphasized within these hierarchical and task-oriented systems.29 As a result, preceptors could be viewed as conforming to a blueprint of role competencies, with the clinician–educator skills at the forefront. But preceptors do not simply perform work tasks unhindered. A study of preceptors’ hospital ward teaching rounds identified several work context tensions, including patient census, time sensitivity, and multiple and conflicting commitments of the participants.12 Attending physicians also have related the seemingly unsafe number of patients they have to see, a variable over which they often have no control, to the quality of the patient care they provide.30 Goldszmidt and colleagues31 reported on preceptors’ views of the clinician–educator aspect of their role and focused on how they balanced teaching and patient care. The authors advanced a model where relative prioritization of patient care, trainee supervision, and teaching drove supervisory practice. They acknowledged that supervisory style may be vulnerable to, but also represent an approach to, contextual triggers.31 Historically, according to Ludmerer,32 preceptors have brought research scholarship into their teaching and supervision work; this role appears to be missing from our study findings. It is possible that some important aspects of the preceptor role are less prevalent now or have been redirected to other teaching settings because of the evolution of the health care system and current intense time demands on preceptors.
Our results fit within the theoretical framework of sociomateriality,17 where it is acknowledged that we work with people, places, and things, and complex adaptive systems.18,19,33 The latter is defined as “a way of thinking about and analyzing things by recognizing complexity, patterns, and interrelationships rather than focusing on cause and effect.”19 This framework and its applications in health care systems were the topics of a 2010 evidence scan by the Health Foundation.19 The authors described how representatives or “agents” within complex adaptive systems may be forced to interact and associate in ways that are unpredictable and unplanned. They also suggested that this variety and “edge of chaos” generate creativity. Complexity science supports the need to optimize people and relationships within the living systems in which they work, allowing them to be creative.19,33 In our study, the observers documented a hectic and unpredictable environment in which preceptors’ manner of acting appeared to reflect creativity born from the need to continually adapt. Other scholars have demonstrated that high-performing academic clinicians have both “reflective clinical practice and scholarship” and “people skills, engagement and adaptability.”34 In our study, the observers documented that the preceptors optimized “the people” within the system, including themselves, using humanism, interpersonal skills, positivity, and wellness behaviors. Health care and medical education systems should recognize that attending physicians operate within these living systems as well as support them in their optimization of both the traditional competencies and their behaviors and attitudes in the role.
We recommend keeping in mind the strengths and limitations of the study design when interpreting our results. Qualitative research serves to produce and explain models that help to improve our understanding of complex settings, rather than to produce generalizable results.24 Thus, a strength of our study is the observational qualitative approach we used which provided in-depth findings that could not be produced using either quantitative methods such as surveys or other qualitative methods such as focus groups or interviews. Several limitations deserve consideration as well. First, our findings may not translate to nonuniversity hospitals or to noninternist preceptors. Next, the preceptor’s voice about the motivation behind her or his actions is absent. Also, observers focused on the preceptors rather than on the other stakeholders whose perspectives may not be well represented in our findings. In addition, preceptors’ behaviors may have been affected because they were aware that they were being observed (i.e., the Hawthorne effect). Finally, the observers, in particular the insiders, may have experienced empathy for the preceptors, which could have affected their interpretation of the observed events.
The implications of our findings for medical education programs are many. Overall, we see a need to recognize the value of humanizing the attending physician preceptor role and applying this construct to how we prepare preceptors for optimal role performance. First, training objectives should reflect the impact of the work context on preceptor performance and highlight the critical importance of being able to negotiate the interface between performance and work context. Second, education around newly identified aspects of the competence (e.g., safety architect) and conduct elements of the preceptor role should be incorporated into current training objectives and included in orientation and mentoring tools. For example, adaptability, creativity, leadership, information management, and the safety architect competency could become learnable and measurable assets. Third, current preceptor performance assessment frameworks should move beyond role descriptions and evaluations to recognize that some aspects of performance are independent of the preceptor her- or himself or require additional skills to achieve. Lastly and perhaps most importantly, we should acknowledge that health care and medical education organizations hold valuable assets in their human resources.13,19,34 Armed with leadership, interpersonal, and creativity skills, physicians are well positioned to advocate for change within health care systems where needed. By promoting innovation, wellness, and healthy work environments that enable physicians to perform optimally,22 physicians themselves, the medical profession, and health care institutions could work collectively to ensure sustainable and resilient systems.
The authors acknowledge the invaluable contributions of the visiting attending physician preceptors from across the country who served as insider observers. The authors wish to thank Alicia Polachek for expertly implementing the study as well as for her role in data collection and analysis. The authors also thank Kristen Desjarlais-deKlerk and Jaya Dixit for their expertise in data collection and analysis; Jill de Grood, Jennifer Cowles, and Mary Widas for their support of this study; Holly Wong for her assistance with the preparation of the manuscript; and Dr. Christopher Brown and the many others for their helpful feedback on the manuscript and study results.
6. Cruess SR, Cruess RL, Steinert Y. Role modelling—Making the most of a powerful teaching strategy. BMJ. 2008;336:718721.
7. Frank JR, Danoff D. The CanMEDS initiative: Implementing an outcomes-based framework of physician competencies. Med Teach. 2007;29:642647.
8. Sutkin G, Wagner E, Harris I, Schiffer R. What makes a good clinical teacher in medicine? A review of the literature. Acad Med. 2008;83:452466.
9. Steinert Y, Mann K, Centeno A, et al. A systematic review of faculty development initiatives designed to improve teaching effectiveness in medical education: BEME guide no. 8. Med Teach. 2006;28:497526.
10. Koeijers JJ, Busari JO, Duits AJ. A case study of the implementation of a competency-based curriculum in a Caribbean teaching hospital. West Indian Med J. 2012;61:726732.
11. Nilsson MS, Pennbrant S, Pilhammar E, Wenestam CG. Pedagogical strategies used in clinical medical education: An observational study. BMC Med Educ. 2010;10:9.
12. Hoffman KG, Donaldson JF. Contextual tensions of the clinical environment and their influence on teaching and learning. Med Educ. 2004;38:448454.
13. Borden WB, Mushlin AI, Gordon JE, Leiman JM, Pardes H. A new conceptual framework for academic health centers. Acad Med. 2015;90:569573.
14. West DC, Robins L, Gruppen LD. Workforce, learners, competencies, and the learning environment: Research in Medical Education 2014 and the way forward. Acad Med. 2014;89:14321435.
15. Bates J, Ellaway RH. Mapping the dark matter of context: A conceptual scoping review. Med Educ. 2016;50:807816.
16. Rosenbaum ME, Ferguson KJ, Lobas JG. Teaching medical students and residents skills for delivering bad news: A review of strategies. Acad Med. 2004;79:107117.
17. Fenwick T. Sociomateriality in medical practice and learning: Attuning to what matters. Med Educ. 2014;48:4452.
18. Holden LM. Complex adaptive systems: Concept analysis. J Adv Nurs. 2005;52:651657.
19. Complex Adaptive Systems: Evidence Scan. 2010.London, England: The Health Foundation.
20. Wachter RM, Verghese A. The attending physician on the wards: Finding a new homeostasis. JAMA. 2012;308:977978.
21. Stone S, Ellers B, Holmes D, Orgren R, Qualters D, Thompson J. Identifying oneself as a teacher: The perceptions of preceptors. Med Educ. 2002;36:180185.
22. Wallace JE, Lemaire JB, Ghali WA. Physician wellness: A missing quality indicator. Lancet. 2009;374:17141721.
23. Ten Cate O, Chen HC, Hoff RG, Peters H, Bok H, van der Schaaf M. Curriculum development for the workplace using entrustable professional activities (EPAs): AMEE guide no. 99. Med Teach. 2015;37:9831002.
24. Tavakol M, Sandars J. Quantitative and qualitative methods in medical education research: AMEE guide no 90: Part II. Med Teach. 2014;36:838848.
25. Berg BL, Lune H. Qualitative Research Methods for the Social Sciences. 2012.8th ed. Boston, MA: Pearson.
26. Buford May RA, Pattillo-McCoy M. Do you see what I see? Examining a collaborative ethnography. Qual Inq. 2000;6:6587.
29. Brookes K, Davidson PM, Daly J, Halcomb EJ. Role theory: A framework to investigate the community nurse role in contemporary health care systems. Contemp Nurse. 2007;25:146155.
30. Michtalik HJ, Yeh HC, Pronovost PJ, Brotman DJ. Impact of attending physician workload on patient care: A survey of hospitalists. JAMA Intern Med. 2013;173:375377.
31. Goldszmidt M, Faden L, Dornan T, van Merriënboer J, Bordage G, Lingard L. Attending physician variability: A model of four supervisory styles. Acad Med. 2015;90:15411546.
32. Ludmerer KM. Let Me Heal: The Opportunity to Preserve Excellence in American Medicine. 2015.New York, NY: Oxford University Press.
33. Plsek PE, Wilson T. Complexity, leadership, and management in healthcare organisations. BMJ. 2001;323:746749.
34. Mahant S, Jovcevska V, Wadhwa A. The nature of excellent clinicians at an academic health science center: A qualitative study. Acad Med. 2012;87:17151721.