Hoellein, Andrew R.; Feddock, Christopher A.; Wilson, John F.; Griffith, Charles H. III; Rudy, David W.; Caudill, T Shawn
Background: Inpatient internal medicine education occurs in a fragile learning environment. The authors hypothesized that when medical students are involved in teaching rounds, residents may perceive a decrease in value of attending teaching.
Method: During two summer periods, trained research assistants shadowed teaching rounds, tracking patient census and team call status, recording basic content of rounds, and delivering a survey instrument to the learners, asking them to rate the quality of the attending’s teaching that day.
Results: One hundred sixty-six rounds were analyzed. Attending teaching ratings peaked when students were highly involved. In fact, high student involvement was an independent predictor of higher resident evaluation of teaching rounds (P < .0001).
Conclusions: The best teaching occurred when involvement of medical students was greatest and their involvement was not necessarily a zero-sum game. The authors conclude that attending investment in medical student education during teaching rounds benefits all members of the inpatient team.
Attending rounds are attributed to Franciscus de le Boe Sylvius (1614–1672) and were popularized by William Osler.1 Given this time-honored history, ward rounds conducted by the teaching attending are traditionally considered the cornerstone of clinical education.1–3 Further, some espouse the notion that all hospitals should be teaching hospitals2 and that all rounds should be teaching rounds.3 However, the inpatient setting is likely one of the most fragile environments for formal instruction. Attending rounds are subject to frequent interruptions,4,5 fatigue and sleep deprivation,5 documentation requirements,5,6 patient turnover,6 and patient availability.4 This highly complex setting is further complicated by the presence of multiple learners at different levels (resident–physicians and medical students). In the academic medical hospital, attending physicians are expected to both supervise and take ultimate responsibility for patient care as well as improvise and adapt their teaching to different leaner needs and clinical conditions. Moreover, attending physicians must respect the residents’ compliance with Accreditation Council for Graduate Medical Education work hours limitations (WHL).7 Little is known of how teaching rounds occur in the hectic hospital setting and how the balance of resident and student learning is maintained.
It has long been known that when students are mere onlookers, the educational value of attending rounds suffers.8 In addition, it is suggested that an ideal learning climate is one in which students are actively involved or even personally invested.4 In a recent nationwide survey of fourth-year medical students, Guarino and colleagues9 confirmed this concept when they found certain attending characteristics associated with student satisfaction with rounds, especially engaging students in substantive discussion. In fact, in their study of 1,530 respondents, they found active engagement significantly predictive of medical student satisfaction, with an odds ratio of 3.0.
What is happening on today’s attending rounds? How does an attending efficiently teach medical students or residents in the increasingly hectic milieu of academic medical centers? Are students being taught, and where is it happening? Furthermore, can one engage and teach medical students during rounds without compromising resident focus and WHLs? We hypothesize that if the balance of teaching rounds tips toward the students—that is, longer patient presentations by students and the teaching of basic concepts to students—residents may feel pressured by their time commitments and their own learning needs, leading to a perceived decrease in value of attending rounds. On the other hand, if the students are less engaged in teaching rounds, their perceptions of teaching quality will be adversely affected. Therefore, the purpose of this study is to determine the impact of student involvement on perceptions of the quality of teaching rounds for all learners.
In the summer periods of 2004 and 2005, five medical student research assistants who had completed human subject protection training shadowed a convenience sample of teaching rounds on general medical inpatient services. After extensive pilot testing of the data collection and survey instruments and training of the research assistants, reliable standardized raters were further assured with a rotating “float rater” system so that two raters might follow a team on any given day.
The research assistants tracked the daily census, admissions, and discharges (including deaths), and they coded the content of the first two hours of rounds in 2.5-minute intervals into patient care, other teaching, or administration. These contextual variables were felt to be important to our conceptual model of how external factors can affect teaching and the resident–student balance. After rounds, they delivered an institutional review board-approved survey instrument to the residents and third-year medical students on the teams. The survey asked residents and students to rate their level of agreement with “my attending spent enough time teaching today” on a five-point, Likert-type scale (1 = strongly disagree, 5 = strongly agree). Finally, the research assistants independently recorded their own perceptions of the quality of teaching rounds on a 10-point, Likert-type scale (1 = poor, 10 = excellent) with regards to the involvement of students and residents, also on a 10-point, Likert-type scale (1 = little involvement, 10 = very involved). Student and resident involvement was rated as a function of active participation in the rounds, such as case presentations, questions, or patient interactions. For purposes of analysis, high involvement was defined as 8 or higher on the Likert scale. Data were analyzed with simple descriptive statistics, correlation coefficients, and multiple regression approaches, with ratings of attending teaching as dependent variables and involvement of learners and the contextual factors listed above as independent variables.
At our institution, there are four internal medicine teams, each consisting of an attending physician, senior resident physician, two interns, and two third-year medical students. Other occasional members of the team include an acting intern (fourth-year medical student) and a pharmacist or pharmacy student. The team takes overnight call every fourth day. Typically, the interns and students “preround” on their patients before the 8:00 am morning report, and attending teaching rounds occur at 9:00 am but end before the 12:00 pm resident conference. Our usual model is to combine teaching rounds with management rounds as approved by the Residency Review Committee program requirements.10 In general, the entire team sees every patient every day, with the student or intern presenting to the attending and resident at room side.
One hundred sixty-six daily rounds sessions were evaluated, canvassing 17 unique attending physicians. Individual resident and student data were not collected, because of the anonymous nature of the survey instrument. Patient care was the focus of the majority of the first two hours of attending rounds (median 66%, range 55%–83%), and the median time (range) spent at the bedside was 28% (1%–40%). Separate time dedicated to teaching comprised a median of 6% (3%–11%) of rounds, whereas a median of 3% (0%–9%) of time was spent on administrative work. The mean patient census was 7.7 ± 3.0. Teaching quality as rated by the residents, medical students, and research assistants was 4.2 ± 0.8, 4.0 ± 1.0, and 7.1 ± 2.1, respectively. All evaluators’ assessments of teaching quality were strongly positively correlated with each other (P < .0001). Mean level of involvement for residents was 8.2 ± 1.5 and for students was 5.3 ± 2.2.
High student involvement occurred in 20% and high resident involvement in 79% of attending rounds; both were rated as high in 10%. Resident involvement was negatively correlated with student involvement (r = −0.26; P = .008). Adequacy of teaching time ratings peaked when students were highly involved: both students (F = 5.11; P = .002) and residents (F = 4.74; P = .003) rated teaching the highest when students were highly involved (students 4.5 and residents 4.6) compared with when the students were not involved (students 3.9 and residents 4.1). In fact, high student involvement was an independent significant predictor of higher resident evaluation of teaching rounds (model F = 8.21; P < .001; r2 = 0.13; variable F = 24.4; P < .0001). Research assistant evaluations of attending teaching were correlated positively with both resident involvement (0.42; P < .001) and student involvement (0.21; P = .01), but they correlated negatively with census (−0.22; P = .006). On further analysis of contextual factors (census, call status, admissions, discharges, etc.) associated with rounds on which both high resident and student involvement occurred, only the occurrence of a patient’s death was significantly associated with both high resident and student involvement (26% versus 5.6% base rate; P < .05).
In our study, the most important independent variable predictive of quality attending teaching was not patient census, call structure, or predominant content of teaching rounds but, rather, the level of learner involvement. Moreover, the very best teaching occurred when involvement of medical students was greatest. We did find a significant negative correlation between resident and student involvement, and although the involvement of residents may interfere with student perceptions of teaching quality, residents seem to interpret the involvement of students in teaching rounds as satisfying basic teaching expectations. In the end, heavy student involvement is not necessarily a zero-sum game, because everyone benefits.
Factors producing maximal involvement of all learners are unclear. In an observational case study, Irby11 followed six distinguished clinical teachers, and although widely disparate teaching styles and diverse teaching scripts were evident, some common themes emerged: enthusiasm, planning, improvisation, and simultaneously diagnosing the patient’s problem and the learner’s level of understanding, all in a climate of mutual support. Perhaps when students are more involved, attending physicians exhibit more of the characteristics described by Irby.11 Anecdotally, we have observed that when students are on rounds, teaching is often more deliberate and planned, covering a deeper range of material. When students are absent, attendings may not discuss basic pathophysiology or disease manifestations in as much detail, because they inaccurately assume that residents have already mastered this material. Residents may value these student-centric rounds more, because they may still have gaps in basic knowledge or appreciate the planned teaching sessions. Future research could further explore other contextual factors of rounds (e.g., different call systems or team structures) that maximize both learner involvement and educational value. Faculty development interventions emphasizing enthusiastic bedside teaching may naturally promote student engagement and maximize the educational value for all.12,13 We were pleasantly surprised that 28% of rounding time in our sample was spent at the bedside, but we did note the wide variation from 1% to 40% of rounding time.
Inpatient rounds are inherently chaotic, and therefore actual teaching may be very difficult to separate from patient care and may explain our 6% other teaching. Although trainees may consider mini-lectures on topics that arise during rounds to be the sine qua non of teaching, lessons from attending physicians are often integrated into patient care discussions. Perhaps in this state of “clinical entropy,” teaching trainees in the traditional passive pedagogy conceptualized as a series of mini-lectures on topics that happen to arise may be nearly impossible.1 In addition, attending physicians may consider modeling the doctor–patient interaction as one of the most important teaching moments. Thus, students and educators may differ on the situations which demonstrate teaching. In a review of inpatient internal medicine interventions, Di Francesco and colleagues14 concluded that there are few data about effective inpatient training as a result of the complexity of the learning environment, different patient experiences, and balance of autonomy and supervision. Nevertheless, it takes a skilled physician–educator to balance the clinical needs with the needs of various learners for a supportive learning environment for all.
In addition to the difficulties of studying inpatient teaching listed above, our data have certain limitations to consider in the interpretation of the results. The sample was collected over two summer periods at a single institution and may not be generalizable to other training programs and medical schools. The timing of data collection, however, did allow us to capture both seasoned and novice housestaff and students. Further, we analyzed only those attending rounds in which both students and residents were present. The absence of students may create rounds which are even less structured and less centered on quality teaching. By contrast, these rounds may be more educational if residents are more willing to ask basic medical questions when students are not present. Also, for purposes of standardized analysis, we analyzed the content of only the first two hours of rounds, potentially missing valuable teaching moments. Further, the validity of our attending teaching evaluation and using the unit of time might be perceived as subjective, and time teaching may not equate quality teaching. Finally, an important confounder may be the attending him- or herself—better teaching attendings may be more likely to involve students. Future research could investigate resident teaching ratings of attending rounds with and without students and compare satisfaction. Despite these limitations, we conclude that attending investment in medical student education during teaching rounds benefits all members of the inpatient team and that both residents and students believe the best teaching rounds were when students were most involved.
1Linfors EW, Neelon FA. The case for bedside rounds. N Engl J Med. 1980;303:1230–1233.
2Hurst JW. The overlecturing and underteaching of clinical medicine. Arch Intern Med. 2004;164:1605–1608.
3Shankel SW, Mazzaferri EL. Teaching the resident in internal medicine: present practices and suggestions for the future. JAMA. 1986;256:725–729.
4Ende J. What if Osler were one of us? J Gen Intern Med. 1997;12:S41–S48.
5Lurie N, Rank B, Parenti C, Wolley T, Snoke W. How do house officers spend their call nights? A timed study of internal medicine house staff on call. N Engl J Med. 1989;320:1673–1677.
6Ahmed MEK. What is happening to bedside clinical teaching? Med Educ. 2002;36:1185–1188.
8Reichsman F, Browning FE, Hinshaw JR. Observations of undergraduate clinical teachers in action. J Med Educ. 1964;39: 147–163.
9Guarino CM, Ko CY, Baker LC, Klein DJ, Quiter ES, Escarce JJ. Impact of instructional practices on student satisfaction with attendings’ teaching in the inpatient component of internal medicine clerkships. J Gen Intern Med. 2006;21:7–12.
11Irby DM. How attending physicians make instructional decisions when conducting teaching rounds. Acad Med. 1992;67:630–638.
12LaCombe MA. On bedside teaching. Ann Intern Med. 1997;126:217–220.
13Lehman LS, Brancati FL, Chen M-C, Roter D, Dobs AS. The effect of bedside case presentations on patients’ perceptions of their medical care. N Engl J Med. 1997;336:1150–1155.
14Di Francesco L, Pistoria MJ, Auerbach AD, Nardino RJ, Holmboe ES. Internal medicine training in the inpatient setting: a review of published educational interventions. J Gen Intern Med. 2005;20:1173–1180.