education in family practice Analysis of Resident and Attending Physician Interactions in Family Medicine Mark P. Knudson, MD, Frank H. Lawler, MD, Steven C. Zweig, MD, Carlos A. Moreno, MD, Michael C. Hosokawa, EdD, and Robert L. Blake, Jr., MD Winston-Salem, North Carolina; Greenville, North Carolina; Columbia, Missouri; and San Antonio, Texas Clinical teaching does not fit neatly into traditional teaching-learning models. The in­ teraction between a resident and an attending physician is of particular interest be­ cause it has several functions including education, supervision, socialization, and quality control. The purpose of this study was to observe, classify, and record verbal teaching and learning behaviors in the resident-attending physician interaction. During a 12- month period, 125 observations of resident-attending physician interactions were re­ corded; the average length of the interactions was 4.27 minutes. The six most frequent resident verbal behaviors compared by postgraduate year level did not vary significantly. Only one of the six most frequent attending physician verbal behaviors varied significantly. In the average interaction of about 4 minutes, three fourths of the interaction was on patient care issues, leaving little time for teaching. There are many unanswered questions about the resident-attending physi­ cian interaction and its contribution to the training of a physician. here has been increasing interest in defining, describ­ answer, inquiry-problem solving, modeling, consultation, T ing, and analyzing clinical teaching, particularly in demonstration, clarification, or seeking information from the ambulatory setting. Most clinical teaching does not fit such outside resources as texts or consultants. When a neatly into traditional teaching or learning models; how­ diagnosis and a management plan have been developed ever, the interaction between a resident and an attending that are acceptable to the resident and to the attending physician is of particular interest because of the broad physician, the resident returns to the examination room to range of behaviors that come into play. This one-to-one discuss the findings and the plan with the patient. interaction has several functions, including education, su­ There are many variations to this basic scenario. In some pervision, socialization, and quality control. training programs, all residents present all patients to the Traditionally, in the ambulatory setting the resident- attending physicians, and in others, the interns present all attending physician interaction follows the completion of a patients, but second- and third-year residents present a history and physical examination by the resident. The resi­ proportion of their patients or present only at the residents’ dent presents the findings to the attending physician, and discretion. The resident and the attending physician may the ensuing dialogue becomes a teaching-learning experi­ see the patient together, the resident may be viewed from ence focused on the care of the patient. Teaching and an observation room or by video, or the entire interaction learning may be categorized as instruction, question-and- may be centered on the resident’s presentation of a patient whom the attending physician never sees. Few observational studies of one-to-one clinical teaching Submitted, revised, March 7, 1989. have been published. Foley and others1 videotaped 17 ran­ domly selected clinical teaching sessions in a core medical From the Department of Family and Community Medicine, Bowman-Gray School of Medicine, Winston-Salem, and the Department of Family Medicine, East Carolina school clerkship and analyzed the verbal behavior and the University School of Medicine, Greenville, North Carolina; the Department of Family level of verbal interaction. Foley defined low-level informa­ and Community Medicine, University of Missouri-Columbia School of Medicine, Columbia, Missouri; and the Department of Family Practice, University of Texas tion and low-level questions as reporting, reading, sum­ Health Science Center at San Antonio, San Antonio, Texas. Requests for reprints marizing, giving or asking directions, giving information, should be addressed to Michael Hosokawa, Department of Family and Community or asking about procedures or facts. High-level information Medicine, University of Missouri, MA 303 Health Sciences Center, Columbia, MO 65212. and questions included comparing, contrasting, evaluating, © 1989 Appleton & Lange THE JOURNAL OF FAMILY PRACTICE, VOL. 28, NO. 6: 705-709, 1989 705 RESIDENT AND ATTENDING INTERACTION TABLE 1. PROPORTION OF TIME THE RESIDENT TALKED Resident COMPARED WITH THE TIME THE ATTENDING PHYSICIAN Exploration; Open discussion related to TALKED patient-com pare, evaluate, hypothesis Inquiry: Open-ended question, probing, Attending asking for evaluation, hypothesis Question: Request specific factual or Resident Talked* Physician Talked procedural information Case Presentation: Details of patient's Residency Percent of Percent of history and physical Year Minutes Interaction Minutes Interaction Verification: Seeking agreement with diagnosis or treatm ent plan First 2.62 56.4 2.03 43.6 Response: Answer to request for specific information Recommendation: Proposal of hypothesis Second 2.61 59.4 1.78 40.6 or management plan Acceptance: Agreement with finding, hypothesis or management plan Third 2.10 59.8 1.41 40.2 Seeking: Consulting outside resource (text, chart, consultant) * The differences between the 3 postgraduate years were not statisti­ Silence or confusion: Lack of progress cally significant. Random conversation: Discussion unrelated to the patient Attendinq Exploration: Open discussion related to patient-com pare, evaluate, hypothesize Inquiry: Open-ended questioning, probing, and 26% for second- and third-year residents, respectively asking for evaluation, hypothesis (P C.005). The mean duration of the interactions de­ Question: Request for specific factual or procedural information creased from 7.7 minutes for first-year residents to 6.9 and Response: Answer to request for specific information 6.1 minutes for second- and third-year residents, respec­ Lecturing: Informing, giving information tively (P C.05). More senior residents engaged in more on specific subject Recommendation: Proposal of hypothesis focused interactions and demonstrated clinical indepen­ or management plan Acceptance: Agreement with finding, dence and educational assertiveness. hypothesis or management plan The authors of this study were interested in developing Disagreement: Rejection of finding, hypothesis or management plan an observational model to better describe the resident- Seeking: Consulting outside resource (text, chart, consultant) attending physician interaction. The purpose of this study Silence or confusion: Lack of progress was to observe, classify, and record verbal teaching-learn­ Random conversation: Discussion unrelated ing behaviors in the resident-attending physician interac­ to the patient tion. Figure 1. Interaction analysis grid METHODS synthesizing, predicting, and hypothesizing. The 17 ob­ Family medicine residents at the University of Missouri- servations included teaching rounds, working rounds, Columbia present patients to the attending physician in a morning reports, lectures, patient management confer­ conference room adjacent to the examining rooms. Interns ences, grand rounds, and journal clubs. Low-level informa­ are expected to present all patients to the attending physi­ tion and questions accounted for 78% of instructor talk and cian. Second- and third-year residents present their pa­ ranged from 69% in morning reports to 86% in lectures. tients or cases at their discretion; all encounters are re­ Glenn and others2 observed 949 interactions between viewed by the attending physician, but not necessarily at residents and attending physicians in a family medicine the time the patient is in the clinic. Following the case ambulatory care center. Teaching behaviors were placed in presentation and discussion of a patient, the resident re­ ten categories on an interaction analysis recording form. turns to the patient in the examining room; in some cases Clarifying information and concluding statements were ob­ the attending physician accompanies the resident to see the served in 90% of the interactions; statements that recalled patient. Most teaching or learning occurs during the case didactic information or involved analysis of information presentation and discussion of the patient. Attending phy­ and options were observed in two of every three interac­ sicians use this opportunity differently depending on their tions. Multiple recurrence of clarifying, recalling, analyti­ attending style, the number of patients to be seen by the cal, and concluding behaviors occurred in the same teach­ resident, the chief complaint, and the skills of the resident. ing interaction, and Glenn et al concluded that resident- Using videotapes of resident-attending physician inter­ attending physician teaching was best described as team actions, common verbal behaviors were noted and classi­ problem solving. fied. A grid was developed for recording the verbal Williamson and others3 observed resident-attending behaviors at 15-second intervals. This grid was tested and physician interactions to determine clinical independence revised, and a final version was prepared for use in this and assertiveness. First-year residents consulted the at­ study (Figure 1). tending physician in 48% of the visits