A key goal for interns is to become competent in clinical settings. This requires a capacity to move beyond university-based competence into performance within hospital and other health care settings. This transition is a reciprocal process dependent on the interns’ abilities and the support that they encounter in the workplace setting. We are aware that interns can easily identify “good” clinical teams from whom they will learn effectively. Their more experienced colleagues also know about “good” interns. One of their characteristics is the ability to get effectively involved with the team.1 However, getting involved is a two-way process. There are behaviors interns need to display, but there are also factors within the clinical team itself that determine involvement. Understanding these factors could make life easier for both parties in organizing effective clinical experiences. In this study, we aimed to establish characteristics of desirable clinical teams and to identify factors that make it easier for interns to become involved with these teams.
The literature on effective learning within the workplace suggests that participation in the social context2 and activities of the workplace are key ingredients.3 Being accepted by and participating fully in a work team has consequences for individuals beyond positive working relationships and effective work performance. A key outcome of individuals working and communicating together effectively is the development of intersubjectivity or shared understanding.4 Within a clinical setting, intersubjectivity occurs when all members of a health care team understand each other's preferences and idiosyncrasies and where working together can occur without the need for constant negotiation, which can be reserved for dealing with new or novel tasks or problems. People may not always consciously recognize that intersubjectivity is a learned outcome and is not something that just happens. A lack of intersubjectivity can contribute to a clinical environment being perceived as alienating to some novices.
The knowledge and skills required by medical interns have their origins in practices that have evolved over time and have been learned by more expert practitioners. This knowledge is developed through extensive and diverse experiences until the skills become almost unconscious responses.5,6 There is therefore a need to engage with those expert practitioners in order to learn that knowledge. Textbooks provide one means of securing some forms of knowledge. However, the workplace offers learning outcomes that move beyond tacit knowledge.7,8
Acknowledgment of this is evident in the long-held model of medical apprenticeship where clinical competence is gained in the reality of supervised practice,9 with the experienced clinician guiding reflection and exploration of learning from the real cases and the problems those cases present.10,11 Working, or vocational, knowledge can be seen as a repertoire of exemplars and experiences that the clinician draws on to solve clinical problems. Reflection is necessary to promote learning from action and experience.7,11
Learning is thus both a collaborative and social process. It is difficult for a new practitioner to develop an accurate and full understanding of the workplace unless there are opportunities for people to share their own understanding and then clarify their views by comparing them with the views of others. To understand how best to maximize the opportunities to learn through interaction in hospitals, it is important to understand how individuals participate in medical settings and how that participation might be improved.
This study was instigated by two converging sets of concerns. Firstly, we had noted that final-year medical students regarded as “borderline” are often seen to be uninvolved in clinical team activities,1 that is, they do not engage in interpersonal interactions as frequently as those whose performance is judged to be more effective. Secondly, we noted that active involvement in the workplace is important to effective learning3, and that certain characteristics of workplace environments are more invitational to individuals’ learning than are others.3 Thus, individuals may struggle in a workplace because they are not involved or because workplace factors actually discourage their involvement. In this study we sought to identify those factors in the clinical workplace setting in hospitals that might enhance effective learning, especially in terms of encouraging or inviting novice doctors to engage in interactions with peers and more experienced practitioners.
This study had three stages during which we identified themes and developed and refined a model. In stage 1, we interviewed five interns about their experiences during their recent clinical rotations. In stage 2, we took the initial findings to a focus group of 12 additional interns to validate, reinforce, and extend the data gathered in the first stage and to refine the tentative model. This occurred in Christchurch hospital, New Zealand, during 2002 and 2003. The third stage comprised two phases of validation of the model by clinical educators and health educators.
Stage 1: Interviews and initial identification of themes
Five interns were selected and interviewed individually about experiences in their two most recent clinical rotations. We purposefully selected these participants to ensure a representation of experiences across specialties, settings, and favorable and less favorable rotations. Although there was some commonality of rotations, this sampling provided data across seven distinct clinical environments, including two provincial hospitals. One of us (DS) conducted the interviews in privacy, in the clinical settings. DS helped coordinate the training program of interns, but had no involvement in the assessment of performance or local employment or disciplinary issues. The interview used open-ended questions to elicit information about experiences during the clinical rotation, the level of interaction and involvement with the clinical team, the identification of useful learning experiences, and aspects of the attachment that facilitated or inhibited learning. In addition, the participants were asked about advice they might give a subsequent intern on how to make the most of the attachment. DS followed up on participants’ responses, pursued themes, and sought clarification or elaboration where required. Each participant was also asked to recall a critical incident where significant learning occurred. He or she was asked to describe the event, the context, and the factors that helped his or her learning and/or contributed to this becoming an effective learning event.
The ethics committee of the Christchurch College of Education in New Zealand approved the project, and all participants gave written informed consent. Each interview lasted 45–90 minutes, was taped, and then transcribed verbatim. We coded the data to ensure anonymity of individuals and rotations. We analyzed the transcripts and identified key themes using inductive analysis and principles of grounded theory and the constant comparative method.12,13 We then used the synthesis of the data and the emerging themes to propose a tentative model.
Stage 2: Focus-group interview
The themes and tentative model were further elaborated and validated in the second stage through an interview with a focus group comprising 12 interns from a cohort group that was different from the original sample. These participants were asked to review the model and to add comments about their own attachment experiences with a particular focus on factors that encouraged or inhibited their participation and learning in the clinical settings. Overall, this group's comments reinforced the themes synthesized from the interview data from the initial interview group and added some contributions to the initial data. One of us (TW) facilitated the meeting while another (DS) recorded the comments and suggestions. We also analyzed these comments and additional data for themes and then combined them with the initial data. We then developed the more elaborated version of the model.
Stage 3: Validation of themes and model
In the third stage, we presented tentative findings in the form of key themes to two groups of informed practitioners for their perspectives and judgments. The first group comprised 11 clinical educators, from a broad range of institutions within New Zealand, who were involved with a Graduate Certificate in Clinical Teaching. We invited them to comment on whether the themes were consistent with their own experiences of effective learning within a clinical environment (a form of face validity) and, in small groups, to derive their own model from the themes. These educators gave us oral permission to use their models in order to challenge and inform our ideas and model development. From these diagrams and comments, we modified and refined the initial model. We presented the themes and tentative model to a second group of health professional educators and interns at an annual health professional education conference. The feedback from the participants supported the themes and the model.
We categorized the findings from the initial analysis of the data under three related themes: factors encouraging participation and learning, barriers to participation and learning, and behaviors the interns could adopt to enhance their involvement and learning. These themes emerged initially from the analysis of the individual interview data and remained largely unchanged following the focus-group triangulation. Illustrative quotations for each theme that emerged from each set of data collection are shown in Table 1.
Theme 1: Factors that encouraged participation
This first theme had a number of subthemes.
Taking responsibility often emerged through absences of more senior team members, although the focus group also indicated that supervisor behavior was important. In the absence of more experienced staff, the capacity of interns to engage in work, make decisions, and undertake the whole task was seen to enhance both participation and learning. However, this opportunity brings the obligation of being clinically competent. Taking responsibility and engaging with the task of doctoring provides rich bases for participation and learning, in so far as it is within each intern's range of existing competence and development.
Team attributes and guidelines.
Both groups identified the attributes of teams and the guidelines under which they operate as important factors that encourage participation. These attributes and guidelines were seen often to result from the idiosyncrasies of senior team members, that is, a learner needed to know not so much the right way to do something but the way preferred by the supervisor. Here, the importance of understanding an individual's status, preferences, and the means of operating were likely to be a key factor in participating as a member of the team.
Relationship with supervisor.
Both groups identified the need to have some valued connection between supervisor and learner. The learners needed to feel valued and the supervisors needed to be approachable, so that the learners would feel invited to participate in the group and could believe that supervisors would value their contributions.
Interactions with more experienced staff.
Having opportunities to interact with more experienced staff was expanded by highlighting the value of serendipitous moments, such as sharing coffee times. Most interviewees strongly identified the value of perceived informal opportunities to interact at a personal level. Placing all participants on an equal footing, such as when all are puzzling over a diagnostic dilemma, was also deemed to facilitate learning interactions.
Teaching and learning behaviors.
The capacity of more experienced practitioners in the clinical settings to effectively use questioning and feedback was valued in terms of the quality of learning outcomes. The opportunities provided by patients to engage and elaborate the interns’ knowledge also was also perceived to have strong pedagogic properties in that it provided a kind of learning in action, guided by more expert partners (e.g., Schön11).
Theme 2: Barriers to participation and learning
This second theme centered mainly on insufficient time and/or access to the supervisor. This theme is significant because a key concern raised in an earlier study was that interns can easily feel marginalized if they lack confidence to persist and engage.1 Having too many people on the team limits the amount of patient and supervisor contact possible, and a kind of hierarchy of learners develops that may be hard to overcome unless the trainees are assertive.
Theme 3: Learner behaviors
This third theme relates to the second theme in that overcoming any hierarchy of learners requires a degree of proactivity. This extends to making offers to do work and undertake tasks, rather than waiting to be told what to do. However, there is an expectation that interns will have a level of proficiency that will allow them to make useful contributions to the team. Interns’ taking on a task that is well beyond their capabilities was not welcomed by fellow workers, who may have to retrieve the situation.
This theme also relates to the first theme in that it is expected that interns will learn to understand the preferences of senior staff members. These preferences are not always explicit and can only be understood through interactions with more experienced members of the team. It also seems that the capacity to listen and think carefully is an important quality in terms of making progress with other staff, particularly senior clinicians. Appearing confident seems to be the key factor here. Learners who initiate activities, offer to do things, ask questions, and get involved with the patients and the team appear to be in a reinforcing cycle. These attributes also encourage intersubjectivity with supervisors and other clinicians, so making tasks easier and communication more efficient as time goes on.
Validation of the themes and development of the model
When we presented these themes to the focus group of clinical educators for their comment, there was consensus that by operationalizing the themes into actions, supervisors could do more to induct interns into the team, and that these actions may then be used to provide a list of key points for supervisors to cover at orientation. The discussion around these issues alerted us to the possibility that some of the drivers for participation were important at the beginning of an attachment (i.e., initiation) and that others were more important later (i.e., maintenance). For instance, for each trainee, knowing where he or she fits in the team is an initiation activity. Finding this out early helps early communications and facilitates team membership. Once confidence builds and communication is easier, the informal conversations with colleagues reinforce and maintain the shared understanding and build the team. The learners’ focus-group data also supported the notion that there were some things that were essential early in an attachment, which we classified as initiation activities, and there were other drivers to maintain participation. This precipitated a consideration of factors relating to patient care and factors relating to interactions with team members.
Finally, we presented the three initial themes to a group of supervisors, educators, and trainees at a health education conference. They were asked to identify factors de novo, and we found that all the factors contributed by this group fitted within the developing model. When we then shared our model, there was consensus that the model described the features of effective learning experiences within clinical workplaces. The trainees particularly highlighted the theme of “interactions with more experienced staff.”
A model for participation in clinical settings
The refined model arising from these considerations is shown in Figure 1. In this model, two critical components are identified as those involved in the task of providing patient care and those in the task of engaging with the clinical team. These two components are further divided into two aspects: initiation and maintenance.
The interview data suggested that, on many clinical placements, some of the activities described in the model are not always relevant. Some rotations can be seen as very task focused, and so there is a very clear articulation and understanding of the patient care duties and their associated roles and expectations. However, other aspects, especially teamwork and communication aspects (e.g., “understand the politics of the people”) seem to receive less attention. The reverse can also occur, in which it is possible to feel welcome, involved, and part of the team, but very unclear about roles in relation to patient care. One of the barriers identified by the interviewees was the queue of learners needing experience but not getting it because they are simply one of too many people on the team. This situation suggests that beyond those personal relationships and workplace activities that assist the learning of interns are other and perhaps more strategic concerns. These concerns include the number of people that might be attached to a particular clinical setting, the range of patient care opportunities likely to arise in a particular setting, and the scope of the work team that the individuals will be participating with. Thus, as with any curriculum, consideration must be given to issues concerning the organization, timing, and flow of the learning experiences that are most appropriate to learners at particular points in their development.
Initiation and maintenance tasks
Initiation tasks can form the basis of orientation to an attachment, as they orientate the learner as a member of a clinical team. This practice not only invites participation, but also gives permission for open, safe participation and, when done well, enhances the newcomer's confidence. Initial tasks involve the behaviors expected of both the supervisor and supervisee within an understanding of the team's organization. If these factors converge, then the trainee feels invited to participate and to enter the maintenance phase. Key initiation tasks include getting to know the team, the preferences of its members, the norms and practices of the clinical settings, and the interns’ scope of expectations.
Maintenance tasks are more day-to-day and are mutually reinforcing for both learner and supervisor. In other words, the more ongoing feedback and guidance learners receive regarding their patient care activities, the more their confidence increases, as do their opportunities for engaging in shared activities that can lead to intersubjectivity or shared understanding between them and their supervisors. This scenario means that less routine direction may be needed and that more time is available to engage fully in discussions, to ask questions, and to participate actively in professional conversations and problem-solving. Ongoing participation accordingly can be encouraged and maintained through a cycle where there are formal and informal conversations about patient care, a sense of partnership in problem solving, and opportunities to provide feedback on decisions. These factors then build on each other to increase confidence, participation, and learning.
The model we propose here identifies necessary components to make clinical settings more positive learning environments. It indicates that attention must be paid not only to patient care but also to team tasks. Furthermore, attention must be paid to initiation factors and maintenance factors. The importance of the two dimensions for participation and learning and the concept of changes in supervisor behavior across time are also represented in Hersey and Blanchard's14 model of workplace supervision, but there is one key difference. Hersey and Blanchard suggest the balance of task-directed supervision and relationship tasks changes with time and experience. Our suggestion is that these two dimensions remain critical over time because practice requires engaging in novel activities and because practice contexts (clinical rotations) vary. Moreover, novelty is in some ways person dependent.
What, then, might an invitational clinical environment be like? Newly arrived trainees would have an initiation that provided practical information on timetables, specific aspects of patient care related to the setting, clarification of the roles of the trainees, and a clear description of the expectations that others have in relation to their performance on the attachment. The idiosyncrasies of other team members would be made explicit. In turn, trainees would be expected to be enquiring but also to show initiative by reading around problems. Once the initiation phase had been satisfactorily completed, a maintenance phase could occur, where ongoing review and feedback relating to patient care could take place alongside activities to promote and encourage interactions within the team. Attention would be paid not only to the performance of the trainees in relation to patient care but also to making time for interactions between team members, thus highlighting the significance of co-participation in workplace learning.3,15 The place of conversations around patient care over morning coffee and between all members of the team also should not be underestimated as a base for their engagement in the reciprocal process of workplace participation.
Clinical areas are busy workplaces, and patient care and service delivery are priorities. It is therefore encouraging that the activities recommended in the model, particularly in the maintenance phase, are part of everyday workplace activities and of the normal flow of the workplace and the supervisor's workday. The model does not ask clinical supervisors to acquire a large new set of skills or teaching techniques but instead to be systematic in using these skills and to consciously engage learners in these activities. It also reminds learners that engagement and participation is a two-way process and that there is an equal responsibility to participate. Each trainee must take responsibility as an active learner and a team member. The social nature of learning2–4 is reinforced and reframed in the model, and no supervisor, no matter how good, can create an environment of shared understanding without input from the learner. Furthermore, when learners, no matter how keen and willing, are not allowed access to key information about the tasks of the clinical area or are excluded from the team in subtle ways, the more likely it is that they will find the required learning difficult. Although not explicitly addressed in this study, relationships within a team are also likely to be important. Future research aimed at identifying patterns of personal communication or team behavior that include or exclude individuals would be useful in this regard.
Others have noted the importance of interactions between team members in promoting learning. Richards,16 when discussing learning acquired from physician colleagues, observes that self-directed, workplace learning involves more interactions with colleagues than does formal learning. Jennett et al.17 tested the effectiveness of a range of teaching methods in continuing medical education and identified a significant enhancement in the learning for those involved in small group discussions with peers. A number of studies provide support for the positive effect of modeling by a respected peer.17–19 Studies in occupational therapy indicate that debriefing with peers in occupational therapy fieldwork practice helps learners appraise themselves and discuss confusing issues.20 Studies in nursing reveal that similar practices boost confidence to apply knowledge and skills.21,22 These studies all demonstrate that interactions with peers or more experienced counterparts in conjunction with everyday learning are well established.
The hierarchy of communication that can occur if there are too many people in the team has been noted in other settings and may exclude learners. For example, full-time workers have been found to inhibit the activities of part-time workers in order to preserve their standing,23 non-English-speaking workers have been marginalized and scapegoated in workplaces,24 and production workers in manufacturing have been denied the standing and involvement that an objective analysis of their work warrants.25 Nevertheless, respondents in our study identified observing conversations as being one component that encouraged professional thinking; this suggests that some important learning can occur when learners are quite passive—consistent with Lave and Wenger's “legitimate peripheral participation.”2
Our model also incorporates notions of effective supervision. Health professional learners asked about the attributes they desire in a clinical supervisor identified modeling competent practice, demonstrating the role, planning learning experiences, explaining own expectations of the supervisee, giving feedback, allowing the supervisee a measure of independence, and encouraging self-evaluation through questioning.26–28 Those familiar with traditional approaches to apprenticeship will recognize practices that have been tested and refined in a number of areas of work, particularly the trades. Drawing upon apprenticeship learning, Collins and Newman29 identified a model of learning for reading, writing, and mathematics that they termed cognitive apprenticeships. In their model, the expert (a teacher) modeled, coached, and scaffolded the learner before withdrawing support incrementally until the learner could achieve independent performance. This approach is clearly analogous to practices proven to be successful in medicine.
Feedback is central to a good learning experience. Ende stresses formative feedback as a tool to keep learners on course and involved, and suggests that without it “the sense of being adrift in a strange environment is amplified.”30, p.778 Irby emphasizes that such feedback requires interactive thinking and improvisation by supervisors because they must simultaneously diagnose a patient's problems and their learners’ levels of understanding. He notes that excellent teachers “incorporate the whole team in discussion.”10, p. 636 Able clinicians and teachers can also include the patient as part of this whole team.31
Formative feedback paired with reflection is even more powerful as this allows the assimilation and reordering of concepts and a consideration of meaning. The exploration of the meaning and implications of experiences and action in the hands of the skilled facilitator keeps discussions on higher levels, avoiding the repetition of mundane facts and more importantly it grows teams and team understanding.32 The skills involved in these tasks should not be underestimated and have important implications for faculty development. A good doctor is not automatically a good teacher or provider of feedback. Likewise, reflection may not occur automatically for new practitioners. While reflection can be a self-directed, private activity, it is also a cognitive process that can be modeled and promoted by supervisors and mentors.7,11
In presenting this model, we invite further discussion and suggestions for refinement. We see the main limitations being uncertainty about its generalizability to other contexts and its utility in a variety of work practices. We have been encouraged by feedback from other health educators, but interns working in other contexts, other centers, or other countries may identify different factors. These areas are worthy of further exploration.
Our model has important implications for faculty development. Since its development, we have used the model for faculty development in three ways. It has been introduced within a five-week “train-the-trainer” seminar series for multidisciplinary health professionals. In an introductory session participants discuss the role of the supervisor and the needs of the learner using the model as a framework. Feedback has been positive and it appears that the model allows clinicians to create a clearer link with later sessions that focus on setting objectives, giving feedback, learning from experience, and facilitating reflection. Adapting teaching and learning strategies for different stages in learning can be discussed, particularly distinguishing between orientation and then ongoing learning within clinical teams. Secondly the model has been used as a foundation for the design of a new clinical rotation in an emergency department. The model was used to set the teaching and supervision standards for the rotation. Feedback from the first two interns on the run has been overwhelmingly appreciative. This highlights the third use of the model: educating students and interns about the nature of clinical learning and their part in the process. The model has thus not only been used as a blueprint for supervisors but also for learners, and stresses the reciprocal roles of teacher and learner.
Within our proposed model, the outcome of all four factors working well is potentially a reinforcing cycle of activities that promote and encourage effective participation and learning in clinical settings. As stated, the activities recommended in the model are seen and reported elsewhere in medical education and are common features of apprenticeship style programs. Our model brings these together into a coherent and simple whole that illuminates the relationship between the components of good initiation, supervision, and workplace guidance. It is a pattern that, once established, allows both supervisors and learners to gain confidence and satisfaction. The model therefore could provide a blueprint or best-practice guide for clinical supervisors, learners, and quality monitors of clinical rotations and vocational training programs.
1 Wilkinson TJ, Harris P. The transition out of medical school - a qualitative study of descriptions of borderline trainee interns. Med Educ. 2002;36:466–71.
2 Lave J, Wenger E. Situated Learning: Legitimate Peripheral Participation. New York: Cambridge University Press, 1991.
3 Billett S. Learning in the Workplace: Strategies for Effective Practice. Sydney: Allen and Unwin, 2001.
4 Rogoff B. Apprenticeship in Thinking: Cognitive Development in Social Context. New York: Oxford University Press, 1991.
5 Anderson JR. Acquisition of cognitive skill. Psychol Rev. 1982;89:369–406.
6 Mann KV. Educating medical students: lessons from research in continuing education. Acad Med. 1994;69:41–7.
7 Boud D, Garrick J (eds). Understanding Learning at Work. London: Routledge, 1999.
8 Billett S. Situated learning: a workplace experience. Aust J Adult Community Educ. 1994;34:112–30.
9 Cox K. What is included in clinical competence? Med J Aust. 1988;148(1):25–7.
10 Irby DM. How attending physicians make instructional decisions when conducting teaching rounds. Acad Med. 1992;67:630–8.
11 Schön DA. The Reflective Practitioner: How Professionals Think in Action. New York: Basic Books, 1983.
12 Strauss A, Corbin J Grounded theory methodology. An overview. In: Denzin NK, Lincoln YS (eds). Handbook of Qualitative Research. Thousand Oaks, CA: Sage Publications, 1994: 273–85.
13 Green J. Grounded theory and the constant comparative method. BMJ. 1998;316(7137):1064–5.
14 Hersey P, Blanchard KH, Johnson DE. Situational Leadership: Management of Organizational Behavior. 7th ed. London: Prentice Hall, 1996:188–228.
15 Billett S. Workplace pedagogic practices: co-participation and learning. Br J Educ Stud. 2002;50:457–81.
16 Richards RK. Physicians’ self-directed learning: a new perspective for continuing medical education III. The physician and self-directed learning projects. Mobius. 1986;6(4):1–14.
17 Jennett PA, Laxdal OE, Hayton RC, et al. The effects of continuing medical education on family doctor performance in office practice: a randomized control study. Med Educ. 1988;22:139–45.
18 Stross JK, Hiss RG, Watts CM, Davis WK, Macdonald R. Continuing education in pulmonary disease for primary-care physicians. Am Rev Respir Dis. 1983;127:739–46.
19 Stross JK, Bole GG. Continuing education in rheumatoid arthritis for the primary care physician. Arthritis Rheum. 1979;22:787–91.
20 Mackenzie L. Briefing and debriefing of student fieldwork experiences: exploring concerns and reflecting on practice. Aust Occup Ther J. 2002;49(2):82–92.
21 Davies E. Reflective practice: a focus for caring. J Nurs Educ. 1995;34:167–74.
22 Shields E. Reflection and learning in student nurses. Nurs Educ Today. 1995;15:452–8.
23 Bernhardt A. The Future of Low-wage Jobs: Case Studies in the Retail Industry. Institute on Education and the Economy Working Paper Number 10, 1999.
24 Hull G. Preface and introduction. In: Hull G (ed). Changing Work, Changing Workers: Critical Perspectives on Language, Literacy and Skills. New York: State University of New York Press, 1997.
25 Darrah CN. Learning and Work: An Exploration in Industrial Ethnography. New York: Garland Publishing, 1996.
26 Onuoha ARA. Effective clinical teaching behaviours from the perspective of students, supervisors and teachers. Physiotherapy. 1994;80:208–14.
27 Neville S, French S. Clinical education: students’ and clinical tutors’ views. Physiotherapy. 1991;77:351–4.
28 Harel BH. Supervision. In: Jacobs K, Logigian MK (eds). Functions of a Manager in Occupational Therapy. Thorofare, NJ: Slack, 1994.
29 Collins A, Brown JS, Newman SE. Cognitive apprenticeship: teaching the crafts of reading, writing and mathematics. In: Resnick LB, ed. Knowledge, Learning and Instruction: Essays in Honor of Robert Glaser. Hillsdale, NJ: Erlbaum & Associates, 1989: 453–94.
30 Ende J. Feedback in clinical medical education. JAMA. 1983;250:777–81.
31 Stewart M, Brown JB, Weston WW, McWhinney IR, McWilliam CL, Freeman TR. Patient-Centred Medicine: Transforming the Clinical Method. 2nd ed. Abingdon: Radcliffe Medical Press, 2003.
32 Branch WT Jr, Paranjape A. Feedback and reflection: teaching methods for clinical settings. Acad Med. 2002;77(12 Pt 1):1185–8.