The hospital inpatient service has been the predominant setting of internal medicine education for over a century.1 However, training in the inpatient learning environment is changing rapidly; the number and acuity of hospitalized patients have risen substantially, many patients arrive with a diagnosis already made, and patients spend increasingly shorter periods of time in the hospital.2 Moreover, as the locus of care for many clinical problems has shifted to the outpatient setting, the diversity of inpatient patient care experiences has diminished. Residents generally do not follow patients to the outpatient setting, and thus they see only the fraction of an illness that is treated at the hospital.3 Finally, the traditional “physician-centric” approach to care for hospitalized patients, in which patient care decisions lay with an individual physician, has evolved to a predominantly interdisciplinary team approach, leading to greater role specialization that impacts physicians, nurses, case workers, pharmacists, therapists, and others.4
The literature is replete with discussions regarding the need for substantial redesign and comprehensive reform of internal medicine training,5 even calling for a new paradigm for medical education.6 Recommendations include restructuring training programs with more attention to ambulatory settings,7–9 improvements in supervision, and further limits to duty hours.10 Despite these significant concerns and developments in internal medicine training, the overall structure and process of medical education in the United States have changed relatively little.11,12
The traditional “rotating” model of inpatient training1 continues to be the gold standard of residency, requiring residents to move through different systems every two to four weeks. This rapid transitioning is considered necessary to expose trainees to a diversity of experiences that guide specialty choice and promote clinical expertise, yet little is known about whether or how this experience relates directly to practice. Although rotations have been studied and enhanced at the undergraduate level in innovative, integrated models of training,13,14 there is limited consensus among the medical community regarding the optimal processes and structures of clinical training, and resident experiences remain largely unexamined. If redesigning residency in internal medicine is a national priority, then an effort to gain a greater understanding of the pedagogical impact of the experience of frequent transitions is warranted.
In this study, we examined the “rotating” model on which residency training is constructed. Specifically, we explored the experience and impact of frequent transitions from the resident perspective. Our goals were to understand more deeply how residents transition between systems, which aspects of transitions facilitated or hindered the experience, what strategies residents relied on to transition successfully, and what impact transitions had on patients.
In March 2007, the American Board of Internal Medicine convened an international panel of sociologists, educators, and physicians to talk through current challenges to professional development in internal medicine. One goal was to understand more fully the extent to which resident professional behaviors were embedded within their larger institutional culture, or the hidden curriculum. Because the training environment changes so frequently for trainees, the panel concluded that research exploring transitions was essential to understanding the hidden curriculum and resident professional development.
We used purposive sampling to select three sites representing a range of geographic locations and structural characteristics (Table 1). At each site, internal medicine residency program administration recruited, via e-mail, interested residents, faculty, nurses, social workers, and ancillary (e.g., clerks, patient care associates) staff to participate in focus groups. Care was taken to recruit a sample representative of each program with regard to gender, year of residency, specialty, and years of experience (Table 1). When the goal of 5 to 12 participants in each focus group15 was reached (total N = 97), recruitment ceased. To protect the identities of our participants, we refrain from using the names of the participating institutions in this report; instead, we refer to them as Sites A, B, and C.
Focus-group method and data collection
Focus groups followed recommended methodology.15 At each site, we held four focus groups with (1) residents only, (2) faculty only, (3) nurses, social workers, and ancillary staff only, and (4) a mixed group. Focus groups took place over a period of eight weeks, from February to March 2009, and lasted 60 to 90 minutes. All were conducted by one researcher (E.B.), with support from other researchers (M.H. and E.H.). A note taker observed and documented nonverbal communication. Prior to the focus groups, participants provided consent and completed a brief survey of sociodemographic questions. After the focus groups, participants were also e-mailed a Web-based survey to anonymously share information they may not have felt comfortable sharing in the larger group, and/or to reflect on additional information generated in the group discussion. Data generated from the Web survey were included in the present analysis. All focus groups were audiotaped and transcribed. Each site received local IRB approval prior to the data collection, and participants received a $100 to $250 stipend for their time.
Focus-group discussions were loosely structured around the experience of transitions. Open-ended questions were used to guide discussion on key topics such as preparation, orientation, and support for transitions; learning of unwritten rules or norms; relationships; and the perceived impact of transitions on professional development and patient care (Supplemental Digital Appendix 1, http://links.lww.com/ACADMED/A46). The topic guide allowed for flexibility on emergent themes16 and developed iteratively as the study progressed, with no questions being deleted and several being added as emergent topics and themes were identified.
We analyzed the data using constant comparative methods consistent with grounded theory, allowing themes to emerge from the data.17 Two researchers (E.B. and M.H.) individually coded the transcripts line by line to identify themes and subthemes. Higher-order themes were further developed and iteratively tested against the full dataset. Three additional researchers (S.G., J.R., E.H.) then reviewed all the codes and transcripts in detail, challenging the developing framework and further modifying and clarifying the themes. Coding was terminated when no new themes emerged, and all members agreed on final interpretation. Qualitative software (NVivo; QSR International, Doncaster, Australia), was used to organize the data.
As seen in Table 2, our final coding schema was organized around three main themes: relationships, the impact of transitions on resident professional development, and the perceived impact of transitions on patient care. Additional subthemes arose under the main theme of relationships, including dwell time, faculty involvement, and geographic decentralization.
Our data suggest that, for residents, establishing and maintaining relationships are critical and expected skills to display on each new transition. Residents consistently described the feeling of being thrust into a “chaotic” and “insane” new environment, forced to master the names, roles, trust, and confidence of a multitude of new trainees, nurses, patients, and faculty. Though residents' levels of interest and success in developing relationships were variable, the majority felt that positive relationships mitigated the stress and anxiety experienced during transitions and fostered effective communication and collaboration.
Among those relationships required for success, residents prioritized relationships with peers:
The biggest thing is the relationship between the intern and the upper-year, more so than the interaction with the attending or the nursing staff. Because the attending is around for a couple of hours and that is it.... And if you have a good relationship [with an upper-year] then things go smoothly. (Resident, Site C)
Many residents also emphasized relationships with nurses, often striving to “get the nurses' respect,” asserting that nurses possessed the “deepest” knowledge of patients and the system. As one resident reflected,
... you are so unsure of yourself. And then to be able to make that jump to make any kind of decision and put the first order in, you're like, what is going to happen to this patient when I prescribe that? And the nurse just says, this is what you need to write, this is what needs to happen. I feel like I gained [confidence] more from the nurses than from anybody else in the first week. I felt more comfortable going to them and being like, you know what? Can you help me out? Maybe it is because of the hierarchy. I don't want to sound like I don't know anything. And nurses know a lot. (Resident, Site C)
Effective communication, personality, trust, prior exposure (both to the staff and to the environment), and possessing clear expectations were all described as facilitators to establishing relationships. Many residents also felt that resolution of conflict was very important to optimize the working relationship:
... if I am feeling ruffled, I either apologize or compensate. Then I try to restore some sort of relationship so we can work better together the next day. (Resident, Site A)
However, there was a great degree of variability in the frequency and intensity of these facilitating factors, and our findings further exposed several aspects of the current rotation schedule that undermine relationships: a lack of dwell time, minimal faculty involvement, and geographic decentralization.
Lack of dwell time.
Frequent transitions seemed to be a significant threat to relationships. Residents reported that they do not have sufficient time, energy, or opportunity to get to know new team members. In a four-week schedule overloaded with service and educational activities, most found it difficult to continuously invest in new relationships, despite the reported value of such relationships in transitions:
One problem is when we have interns switch, we'll have one attending for one day ... so you never really get a rapport, and it feels like a day where you don't get to establish any sort of relationship. And there is not much education that can really take place without building a rapport. (Resident, Site A)
I remember my first day, the attending told me, your job is to be efficient.... I think you are bombarded by work and you can't really think about establishing rapport, even with your attending. You don't have time to pay attention to that. (Resident, Site C)
For many residents, challenges in forming relationships led to a feeling of isolation. For example,
I feel like I was getting to know people, but after three weeks they disappeared. So then I don't have time to hang out with them outside to continue relationships. I don't think it was until I saw people on a [subsequent] basis that I felt like my friendships were growing. I felt pretty isolated in terms of not being able to know people very well. (Resident, Site A)
This concept of isolation was prominent in our data, and linked to working with colleagues in parallel, with little knowledge of what other team members were doing or were capable of, particularly within the first few days of transitions. One resident, for example, declared that nurses and residents “operate on two different levels and rarely interact, except for sometimes the occasional phone call when you need to put in antibiotics” (Resident, Site A). Another resident confessed it was “hard to figure out what roles some of the staff” play and had “no idea what the difference was between a social worker and a care coordinator.” When asked how he worked with these individuals, he reported he would “consult both” and, “whatever the issues are,” let them “figure it out” (Resident, Site C). For other residents, relying on each other was reportedly “easier” than collaborating with others, many of whom were unfamiliar, around issues of patient care:
A lot of us may not know the best person to go to at times, so in terms of communication regarding patients, I think the residents mainly handle that among themselves. (Resident, Site C)
Minimal involvement of faculty.
Our participants described an extremely low level of faculty involvement in supporting residents through transitions. Many faculty disclosed that they didn't explicitly teach goals and expectations, instead stating that “it is much more time-efficient to tell [residents] what you want them to do than it is to ask them what they think is going on” (Faculty, Site C) and acknowledging that they often hold residents to “standards that may not exist” (Faculty, Site B). Further, faculty were generally not aware of the details of how residents experienced transitions, admitting that “two-thirds of resident learning, [faculty] don't see” (Faculty, Site B). The majority believed that, ultimately, residents determined “the rhythm of the floor” (Faculty, Site A), and most relied on senior residents to instruct new residents:
There is a fair amount of peer-to-peer orientation that happens. And I think that is very important. I think that the resident leadership role is very important for the intern as well as for the attending. (Faculty, Site A)
In this context, self-directed learning was highly valued, and most faculty and residents agreed that even under periods of duress, residents should adapt without complaining. Residents and faculty agreed that transitions are “just the way it is” and that the experience of frequent transitions, though frustrating, builds self-confidence. One faculty participant offered an interesting analogy for the experience: “The stress of transitions builds confidence, and helps residents help others in the future, like a wilderness experience” (Faculty, Site A).
Residents expressed a strong reliance on each other for support or guidance in transitions, with the senior resident playing a prominent role. Senior residents established junior residents' expectations for transitions. As one resident described, “75% of how [residents] go from rotation to rotation is talking to other residents” (Resident, Site B). Residents further conveyed a strong message of only knowing what is needed at any given moment, admitting: “If something happens, you look it up. There is no set plan. It's just learning as you go” (Resident, Site C).
Only a few found hospital-wide orientation helpful, and most agreed that having one's “hand held” through transitions would be unnecessary and redundant.
Very few opportunities for formal introductions were described, so most interactions, whether with faculty, nurses, or staff, occurred opportunistically and on a case-by-case basis depending on patients being seen. In units where patients and charts were centrally located and/or geographically localized, a culture of physical nearness facilitated conversation and rapport. All residents agreed on the importance of “face time,” believing that both the experiences of transitions and patient care were optimized when the team had sufficient time to get to know and work with each other. One resident reflected on the benefits of geographic centralization, noting,
We're floor-based now. So that makes the transition a little easier because you're not all over the place. You're in one specific area, you know the nurses and patients already that were there before you got there, and know [various] dispositions and all that. (Resident, Site C)
Resident professional development
Residents seem to value transitions, and they believe that transitions facilitate several skills and habits necessary to becoming a practicing physician, including the ability to multitask, work with difficult and varying personalities, be adaptable and flexible, and build confidence in decision making, especially under uncertain conditions. For one resident, the importance of these experiences for professional development was apparent:
One thing that's good [during transitions] is developing a sense of confidence that I can do this. To not let intimidation take over. Even though we are tired, and uncertain, we have to learn that we can serve ... we can do excellent things and be helpful to patients, be of service to a peer who might also have the same uncertainties. I think there is value in that because it's a maturing process. And as we gain experience, it's like a scout living outdoors in the woods. We adapt. (Resident, Site A)
The majority of residents found that transitions were best managed through trial and error, choosing to embrace concordant attributes of the many individual faculty styles they were exposed to, while discarding the discordant ones: “I take from [faculty] what I want, and then adjust myself—it's not difficult” (Resident, Site C).
A variety of coping strategies were used to endure frequent transitions. Most residents asserted that transitions are so much a part of training that it is best to just “put your head down” and “get through them.” There was an explicit acceptance that a lot would go wrong during transitions, and many residents rationalized the extreme stress and anxiety they felt by “just showing up and doing it,” using each other for support. For many, the experience of transitions got easier with time and experience: “When you first start doing [transitions], it is difficult. And then you just get used to it. Then you can walk into a new service and it is not really a big deal” (Resident, Site B).
To cope with gaps and dysfunctions in the system, a number of residents relied on innovative processes or workarounds,18 some of which were beneficial to residents and to patient care. One resident, for example, rounded on patients the night before each transition to mitigate stress. Nurses also came up with innovative workarounds, including creative ways to help each other learn the names of new residents during each new transition. In these, and several other accounts, residents and nurses developed positive coping strategies to manage the inherent limitations of the system and rotation schedule. Unfortunately, workarounds were often designed in isolation and rarely shared with the rest of the team or unit, so the good ones, even if successful, were lost as residents moved on to new systems.
Several residents described negative coping mechanisms or workarounds used to deal with the stress associated with transitions, particularly in the first few days of each rotation. One resident explained, “It is impossible to [provide] care thoroughly, and you learn to cope with it” by learning how to “play the game” (Resident, Site B). Many residents admitted to cutting corners; for example,
I think there is a huge tension between what you learn in medical school and residency. Where in medical school you have to be thorough in terms of everything, I mean do the whole exam and take the medical history starting from A to Z ... [in residency] it is impossible. You're going to see one patient and then you have to round. You have to streamline. And that process is never taught; you gain it very quickly because you have to. You don't have any choice. If you only saw one patient and you started rounds, the next day you would not make the same mistake. But the reality is that you have to hit the ground running and you only have a finite amount of time to do it in. So like it or not, you're going to cut some corners. (Resident, Site C)
Residents admitted to many other workarounds, including writing fewer progress notes or deliberately not answering telephones or pages, lying about the status of a test or lab, hiding information, and walking away from situations in which they were needed. Consistent with prior research,19,20 many residents avoided asking questions, even when accurate information was critical to patient care.
Faculty were frequently aware of residents' counterproductive habits during transitions, citing accounts of residents who coped by lying, mismanaging emotions, or, in some cases, abusing substances. They further reported that the current rotation schedule contributed to pretense, a lack of patient ownership, and superficial relationships, comparing residents to a “plant with a shallow root” (Faculty, Site C). However, faculty varied in the level of responsibility they felt to address these issues, with many feeling already burdened by increasing responsibilities and admitting that they didn't have the time or skill to properly remediate disruptive behavior.
Finally, it is worth noting the depth of negative affect expressed in our data. Residents and nurses, in particular, frequently experienced heightened levels of stress, anxiety, frustration, intimidation, and fear during the first few days of each transition. The majority of our participants acknowledged the deleterious effects of these emotions on communication, personal well-being, decision making, and the management of day-to-day tasks, yet, consistent with attitudes about transitions more generally, they accepted that negative emotions “increased resident confidence” and were just “the way it is.”
Perceived impact of frequent transitions on patient care
Despite a few reports of enhanced patient care through “fresh eyes” and new perspectives on patient care, our data highlight numerous ways in which frequent transitions may negatively affect quality of care. Residents described how frequent transitions detracted from their ability to relate to patients, leading to “putting patients last,” particularly during handoffs, when they felt their job was “task and information transfer” rather than “assessing whether or not the patient is getting better” (Resident, Site A). One resident likened the experience of transitions on patients as “little boxes going through UPS [postal service]” (Resident, Site B). Many residents also found it difficult to manage patients they just met without knowing a complete history, admitting they were “not as present” because no one “owned” the patient:
... in the days where it was one team per month in transition, you had a senior resident who knew he was leaving, interns who knew they were leaving and this concept of the off-service note and the passing of the guard to the next team. I think this was much more formal. What you've lost is that kind of formal passing of the guard because the team is not a team for a month. (Resident, Site B)
These perspectives were shared by nurses, who felt awkward and guilty that patients often “didn't know who their doctor was.”
Both residents and nurses told several additional stories regarding inefficient, redundant, and burdensome care. Multiple residents described faulty communication and technological processes, largely attributable to the complicated rotation schedule; for example, one resident reported being paged for patients who were no longer his: “They [nurses] may be calling somebody who's not even in the hospital, and meanwhile the patient is sitting there and nothing is being done” (Resident, Site C). In these cases, finding the right person was difficult, and it often delayed or, in worst cases, prevented critical treatments or interventions.
Discussion and Conclusions
Given transformations in health care delivery over the last 30 years, it may be time to test the theoretical underpinnings of the current process and structure of graduate medical education. The rotating model remains the standard for medical education, despite little evidence that frequent transitions support training.21 Importantly, our findings highlight several ways in which transitions may undermine fundamental goals of residency, and vividly illustrate their potentially unfavorable effects on both resident development and patient care.
In their seminal work on the hidden curriculum in medicine, Hafferty and Franks22 urged educators to look beyond the formal, stated goals of medical training to the implicit messages embedded within the day-to-day experiences of trainees that may undermine their professional and ethical development. Hafferty23 later suggested that good places to look for elements of the hidden curriculum include (1) institutional “slang,” (2) evaluation, (3) policy development, and (4) resource allocation. Our findings touch most directly on the latter two of these elements.
A powerful implicit message embedded within the current, rotation-based training model is that residents' autonomy, efficiency, and adaptability are valued over relationship building, teamwork, and the development of deep system knowledge. These values are embodied within the language residents used to describe transitions (“chaotic,” “insane”) and their role relative to transitions (“surviving” and “getting through” transitions; being “efficient,” striving to “figure it out,” “learning as you go,” not needing to have one's “hand held”). They are also evidenced in the apparent lack of resources devoted to faculty management and supervision of residents' transitions. The absence of faculty mentors in the transition process may have negative implications for residents' professional development and patient safety, and this absence dilutes the possibilities for effective role modeling,24 teamwork, and “situated learning” as it deprives residents of the opportunity to become part of a “community of practice.”25,26
The extent to which these values were internalized by residents is remarkable. During the group discussions, a number of advanced residents articulated that they scarcely recognized the disorientation and stress that junior residents and interns described in association with transitions. With just two years of experience, senior residents professed not to spend much energy preparing for transitions, describing them as “no big deal.” In this view, our findings support further examination of the current model's likely dampening effect on reflection, help seeking, support, and effective supervision in residency training.19,26,27
The amount of time and energy residents expend on each transition, mostly through stress and inefficiency, takes away much needed strength and effort for resident learning. Thus, a likely, yet unanticipated, consequence of the implicit imperative for maximal autonomy, flexibility, and efficiency is that frequent transitions may make it difficult for residents to become proficient, let alone expert, in systems or human interactions.28 Consistent with other work implicating elements of teaching and assessment in fostering incompetence,29 the current structure and process of medical education may simply maintain minimal acceptable standards, limiting physicians from aspiring to or achieving expertise.30
Equally problematic is the potential for residents to learn negative coping strategies at a critical period in residency training, when they are acquiring their own styles. Many residents reported negative coping mechanisms and workarounds that may undermine patient care and professionalism. Residency reform initiatives must therefore be undertaken with a focus on what students learn rather than on merely what they are taught.23
Our findings challenge the utility and value of the traditional and frequent “rotating” model in graduate medical education. Although we acknowledge that frequent transitions may provide a well-balanced exposure to both patient acuity and the quality and diversity of learning environments, we have to wonder, given the changes in the inpatient setting, do we still have the right model for training? Are transitions every four weeks between systems necessary? More important, are they the best way for residents to develop and learn? Do frequent transitions contribute to or undermine quality patient care? Without critical analysis of these questions, assumptions of training may remain hidden, and we may mistakenly promulgate ideas that are detrimental to residents, patients, and health care more generally.
To maximize the validity of our findings, we triangulated data from different points of view, including those of residents, nurses, and faculty. We also included mixed groups to mitigate the “groupthink” that sometimes occurs during group sessions. However, our research approach is qualitative, and emergent themes, though transferable across professional groups and three sites, should be tested in additional settings to confirm conceptual generalizability. A second consideration is that some participants were selected by institutional administration, and thus our sample may be biased toward the opinions of those who had existing relationships or familiarity with leadership. However, investigators and interviewers did not personally know the participants, and the critical and often-negative nature of all of the discussions suggests that participants were not inhibited or constrained in their participation, adding credibility to the results.
This study raises awareness of some of the problems currently associated with frequent transitions in graduate medical education not only for residents but also for nurses, other health care staff, and faculty. These are timely issues, given the recent interest in this topic within the medical community.12–14 We hope to stimulate others to think carefully about how the process and structure of residency training cultivate potentially unconstructive attitudes and behaviors, and we encourage additional research to explore alternative models for training in internal medicine.
The authors would like to thank all of our participants for speaking so openly and candidly about their experiences. Without them, this project would not be completed. They also gratefully acknowledge the contributions of Kate Ross, MBE, who assisted in organizing and facilitating the focus groups. The authors further express their sincere gratitude to Fred Hafferty, PhD, for such thoughtful comments on the project, and to Benjamin Chesluk, PhD, Lorna Lynn, MD, and Siddharta Reddy, MPH, for manuscript review.
This work was supported by an American Board of Internal Medicine Foundation (ABIMF) grant.
This study was approved by the institutional review boards at the three institutions where the study was conducted. All researchers involved in the study additionally completed the online Collaborative Institutional Training Initiative ethical training course (https://www.citiprogram.org).
The opinions expressed in this article are those of the authors alone and do not reflect the views of the ABIMF.
The abstract of an earlier version of this article was presented at the Society of General Internal Medicine meeting in Minneapolis, Minnesota (April 2010) and at a workshop conducted on relevant data at the Accreditation Council for Graduate Medical Education Annual Educational Meeting in Nashville, Tennessee (March 2010).