Communication is not only an exchange of ideas but also a form of social behavior that negotiates relationships. How two parties talk with each other reveals their relative status, level of rapport, and value for each other. Not surprisingly, the power that a speaker derives from his or her status may jeopardize a conversation. The teacher's role, particularly as evaluator, often leads the teacher to dominate conversation with the learner. In one-on-one teaching and other dyadic interactions, the less powerful party expects to adapt to the dominant party's speech and initiations.1 When adaptation is extreme, communication is authoritarian; the more powerful party rejects the less powerful party's speech by interrupting, taking over, or monopolizing the conversation. When adaptation is minimal, communication is autonomous; the more powerful party encourages the other to dominate or lead the conversation by verbal and nonverbal behaviors.
Although several studies have demonstrated that residents perceive autonomy as important to their learning,2,3 research about the effects of interactions between residents and attending physicians on the development of clinical independence has produced contradictory findings.4,5 No study has described how autonomy emerges from communication between residents and their teachers. Analysis of conversations between teachers and learners has generally been limited to determining the amount and duration of contact. Most content analysis has focused on the topics discussed and categorization of utterances.6 However, analysis of another dyadic interaction, physician-patient communication, has identified several distinct patterns based on communication control and verbal dominance.7,8
An in-depth examination of communication patterns between the physician-teacher and the physician-in-training may increase our understanding of the types of interactions that help residents learn. This study analyzed how preceptors and residents interact during teaching encounters in ambulatory pediatrics primary care settings. This study focused primarily on autonomous communication when residents dominate the conversation.
The Institutional Review Board approved this study, which was conducted in the continuity care clinics of the general pediatrics residency program at Baylor College of Medicine, Houston, Texas. The study involved both academic and community (private practice) sites. Preceptors were selected based on diversity of teaching reputation, teaching and pediatrics experience, interpersonal skills, and practice setting (solo to large group). The final sample was made up of six academic and seven community preceptors. Four to nine clinical teaching encounters were observed and audiotaped for each preceptor. Each encounter was a unique opportunity to capture a communication pattern. In all, 76 preceptor-resident interactions were analyzed using the grounded-theory method.9 An experienced educator re-examined and independently coded a portion of the transcribed encounters. Intercoder agreement was about 95%. Participating preceptors were also asked to confirm the analyses.
The encounters included acute care, follow-up, and well-child visits and involved first-, second-, and third-year residents. Four distinct patterns of communication were identified based on conversational input and verbal dominance. Of 76 interactions, 54 (71%) showed a conversational balance between speakers: 47 mutual (high preceptor and high resident input) and seven default (low preceptor and low resident input). The remaining 22 interactions showed imbalances between speakers: 15 autonomous (high resident and low preceptor input) and seven authoritarian (high preceptor and low resident input).
Almost 20% of the interactions were classified as autonomous. Of these, 12 (80%) occurred in community settings. Two academic and four community preceptors engaged in autonomous interactions. No preceptor relied on autonomous interactions exclusively, although one academic and one community preceptor used only authoritarian communication. Thus, 11 of the 13 preceptors used more than one communication pattern.
Further analysis of autonomous interactions revealed specific preceptor behaviors. In every autonomous interaction observed, the preceptor recognized the resident's “expertise” and allowed the resident to dominate communication during the interaction or, at least, the conversation about the patient. Generally, the preceptor's approval resulted from the preceptor's identifying the resident's level of understanding as appropriate for a case. The examples reported in the following sections represent behaviors observed across the series of autonomous interactions.
Probing Questions. Preceptors used probing questions to assess residents' understanding. In one community encounter, a first-year resident presented an 8-year-old child complaining of nighttime coughing and congestion related to physical exertion. After listening to a concise but detailed exposition of subjective and objective findings, the preceptor asked, “What do you think of his sequelae to his respiratory infection, exercise cough, and that kind of thing?” The residents' response confirmed a level of understanding appropriate for diagnosing the patient's condition:
R: Well, it's pretty likely that he has some kind of twitchy airway. He's had a recent infection and recent irritation to his lung and he just had another little cold. So anything that he might get on top of it might cause him to have a little bit tighter air flow. So maybe at night, [and] that might be one aspect of reactive airway disease, especially when he's active.
In another encounter, an academic preceptor used variations of the sample probe throughout a third-year resident's presentation of a 5½-year-old girl with Angelman's syndrome and a febrile seizure disorder. The preceptor began probing after the resident had finished presenting the subjective findings:
R: Also [she was] seen by Neurology for a history of febrile seizures seen with infections. She's been on Depakene. She has been on several medicines. First Dilantin liquid, then Dilantin tablets that were crushable, but she had a lot of drooling and would drool out most of it. Depakene, first the sprinkles and now the elixir.
P: They really thought this was a seizure disorder and not just febrile seizures?
R: Thought so. [Looking through chart for Neurology entry] Impression is seizure disorder, febrile. Recommended an EEG [electroencephalogram] which Mom said was done but was not a good study, and hasn't been seen in the [Neurology] Clinic in about two years. Mom wants to take her off the Depakene and I told her that I would want her to be seen by the physicians [neurologists] here.
In this encounter, the resident responded to the question by giving an “expert” answer: citing the chart entry made by Neurology. Then, the resident elaborated by expressing the need and rationale for obtaining a blood test to determine whether (1) the current dosage of anticonvulsant was therapeutic and (2) stopping the medication would do no harm. The preceptor's remarks to the resident's concise assessment and plan for this complicated patient with multiple medical problems signaled support of the resident's autonomy:
R: So, impression is history of Angelman's syndrome, developmental delay, history of complex febrile seizures, left exotropia, and [patient] would probably benefit from visits back to her multiple subspecialists.
P: So, we're going to check her Depakote level today and then maybe decide [about taking her off Depakote]. And you want her to go to Neurology as well.
The preceptor signified concurrence with the resident's plan by using “we” to show agreement (“we're going to check her Depakote level today and then maybe decide”). Similarly, the preceptor confirmed the resident's autonomy and dominance by using the pronoun “you” (“you want her to go to Neurology as well”).
Inference. In other encounters, the preceptor inferred the resident's level of knowledge and understanding from the organization, thoroughness, and conciseness of the case presentation, a finding also demonstrated by Irby.10 The case presentations of the only third-year resident observed in the community setting (four interactions) were so well organized and articulated that the preceptor rarely commented other than to agree with the resident's findings. The resident almost monopolized the conversation, with a smooth, confident, and complete case presentation in SOAP format (subjective-objective-assessment-plan). The preceptor spoke only when the resident paused and showed agreement by using minimal reinforcers, such as “all right,” “okay,” and “that sounds right.” Conversationally, minimal reinforcers cue the speaker that the listener is involved and following the speaker's thoughts.11 In a clinical interaction, these utterances also cued the resident that the preceptor was willing to allow the resident to dominate the talk and the encounter.
Besides smoothness, proper terminology, and adherence to SOAP format, other characteristics of the presentation permitted a different community preceptor to assess a first-year resident's level of knowledge. In this encounter, the resident was confident but more relaxed in style and language than the previously described third-year resident. Satisfied with the resident's presentation of the subjective and objective findings, the preceptor asked for the resident's assessment and plan. The resident replied, “Her right TM looks like really white. I just, I guess that there's pus behind it…. It looks way different than the left side…. So right otitis. And since she's never had any problems before, just do amoxicillin.” Although not eloquent, the resident's response brought agreement from the preceptor and closed the encounter.
Admittedly, without probing the extent of a resident's knowledge, a preceptor might wrongly infer a resident's understanding was appropriate. By engaging in the behavior of “showing,” that is, confidently presenting findings and knowledge of certain entities of a case, a resident might hide actual deficiencies of other entities within the same case. The preceptor's own knowledge of a resident's past performance, particularly for a disease or family of diseases, may prevent some mistakes. For example, in one encounter, an academic preceptor asked a second-year resident to limit the case presentation and “give the big points.” The resident condensed the subjective and objective findings into two sentences: “These kids are here for well-child checks. The bottom line is that the older one has lice and ringworm, and the younger one has only lice.” The resident's response probably tapped into two important pieces of information available to the preceptor. First, the preceptor knew what a second-year resident should know about lice and ringworm, both common pediatric problems. Second, the preceptor knew this resident specifically from interactions over the preceding two years. Despite this knowledge, the preceptor listened attentively through-out the resident's speech and offered minimal reinforcers like “oh no” and “okay.”
Nonverbal Behaviors. In the 15 autonomous interactions, all preceptors listened attentively and used verbal and nonverbal behaviors to indicate that they followed the residents' reasoning and talk. Nonverbal behaviors, such as eye contact, facial expressions, head nods, and alert body posture, more accurately disclose how well a party is listening to a conversation than do verbal behaviors.12 Another important nonverbal behavior observed was the control of the patient's chart. In most autonomous interactions, the resident controlled the patient's chart. Controlling the chart prevented the preceptor from reading the chart during case presentation and diverting the conversation to an unrelated chart entry, behaviors observed in the seven authoritarian interactions characterized by preceptor dominance. In both autonomous and authoritarian interactions, the dominant speaker controlled the patient's chart.
Absence of Teaching Scripts. Preceptors did not use a clinical teaching script10 in any autonomous interaction. Fatigue or time of an encounter, such as the last encounter of a late-running clinic, might have affected a preceptor's decision not to use a clinical teaching script, but instead to permit the resident's autonomy. Arguably, a preceptor who failed to assess a resident's understanding might have missed an opportunity to use a teaching script.
At least one autonomous interaction exemplified how a clinical teaching script might have been less effective than the teaching created by follow-up of the case itself. This encounter was the follow-up visit of a “fussy” 2-week-old child, who had been consoled only by feeding when initially seen by the same academic preceptor and first-year resident. The initial teaching encounter, which corresponded to the first visit, contained a clinical teaching script regarding diagnosis and management of suboptimal weight gain in a baby. At that time, the preceptor and the resident negotiated a treatment plan to increase the number of feedings and rule out gastroesophageal reflux as an organic cause for fussiness and poor weight gain. Between the initial and follow-up visits, the patient had been x-rayed and started on appropriate medication for reflux by another preceptor and resident. By the follow-up visit (and second teaching encounter), the resident was able to learn from the case itself and to see the results of the patient's treatment. The resident's speech reflected understanding of the case and delight with the patient's improvement:
R: She did not cry once, the whole time I was in there.
P: You're kidding. Was this the same baby we saw one week ago [laughing]?
R: No. I was thinking, this was a completely different baby. You know, when I examined her, I did everything. You know? She's alert, looking around. I mean she wasn't lethargic. She was fine, screaming at the top of her lungs.
R: Mom says that she's not fussy at home the way she had been. This is the way she is like at home as well. [She] is breastfeeding about every 2 to 2½ hours, 15 minutes on each breast. Seems satisfied after each feed. The feedings are going better since she started the Zantac and Cisapride.
Observing the post-treatment changes in the patient reinforced the preceptor's earlier teaching script. This pair of encounters also demonstrated the benefits of the continuity care experience in which a resident develops and maintains a continuing physician—patient relationship with a panel of patients. The preceptor's experience with a resident and that resident's panel of patients may increase the likelihood that the preceptor will allow the resident greater autonomy in the teaching interaction.
Autonomous preceptor-resident communication is characterized by high resident and low preceptor input and preceptors' behaviors that confirm and recognize the residents' speech. The preceptor assesses the resident's level of understanding in a particular case through questioning or inference from the organization, thoroughness, conciseness, and confidence of the resident's speech. Twelve of the 15 autonomous interactions occurred in community settings. The reasons for this difference in the rate of autonomous interactions between academic and community settings were not identified. Conversation is a response not only to a person but also to environmental conditions, such as the available time and space for teaching. Economic factors may encourage community preceptors to permit autonomy rather than provide more directed teaching.
Academic and community preceptors were alike in other respects, particularly the use of multiple communication patterns. This finding suggests a spectrum of the relationships between teacher and learner that encourage different conversational behaviors. The use of multiple communication patterns may indicate “scaffolding,” an overarching process within the preceptor-resident relationship in the continuity care experience.13
Scaffolding refers to techniques that support learners in their efforts to solve difficult problems or perform difficult tasks. For a novice resident, a preceptor may behave authoritatively to provide maximum support by modeling desired behaviors, such as how to perform an examination or give anticipatory guidance. As the resident's experience and skill in relating to the preceptor and patients increase, the need for support decreases. Thus, the preceptor behaves less authoritatively, and more collaborative and autonomous interactions occur.
Scaffolding requires the preceptor to know what support a resident truly requires. The space and time available for teaching may increase the preceptor's reliance on autonomous interactions, even when there is a recognizable teaching moment. In this study, two extremes were found: some first-year residents were involved in autonomous interactions and some third-year residents in authoritarian interactions. Possibly, preceptors select the amount of support to give a resident based on the resident's specific experience with a problem. Whitman and Schwenk14 seem to advocate “selective” scaffolding by suggesting that medical teachers alternate between assuming active and passive roles, depending on learners' needs. Training preceptors in scaffolding techniques is not likely to eliminate default communication patterns, particularly when fatigue undermines conversation.
This study is limited because it was performed in a pediatrics setting. Despite its setting, the study has potential implications for all clinical teaching that involves one-on-one interactions between learner and teacher. It is probable that the communication patterns identified may be observed in other practice areas because this study did not limit participation to exemplary teachers and sampled for diversity. The effect of observation on participants' behaviors cannot be entirely discounted. However, over a clinic session, both preceptors and residents seemed to forget that they were being observed.
Future studies should delineate how different communication patterns affect teaching and learning. The relationship of the teaching interaction to future physician—patient communication also deserves investigation. For example, do the ways that preceptors talk with residents predict the ways that residents talk with patients?
Teaching preceptors active listening and supportive verbal and nonverbal behaviors may benefit both patients and residents. When preceptors carefully attend to their residents' presentations, by listening and acknowledging the residents' concerns and opinions, they model how physicians should attend to their patients' conversations, by listening and acknowledging patients' concerns. Studying the conversation of clinical teaching should illuminate how physicians-in-training learn the conversation of healing.