Secondary Logo

Journal Logo

Special Theme: End-of-life Care: Education and Practice: ARTICLES

How Do Physicians Teach Empathy in the Primary Care Setting?

Shapiro, Johanna PhD

Author Information
  • Free


A key professional attribute that medical students and residents must cultivate is the ability to empathize with patients. Despite some concerns about whether empathy is truly a component of professionalism,1 in general it is recognized that to effectively render help to the patient, the physician needs empathy for the patient's situation and lived experience.2,3 Clinical training experiences are particularly important in this regard, as research suggests that it is in these settings that “process understanding” of the practice of medicine is acquired.4,5 Yet, paradoxically, clinical exposure often has the opposite effect. Both anecdotal reports and research studies point to significant negative shifts in student attitudes toward patients between the preclinical and clinical years.6,7 Similarly, several articles note a disturbing trend among residents toward cynicism and self-protective strategies as their training progresses.8–10

Most formal teaching approaches have focused on teaching empathy as a behavioral skill.11 In these models, empathy is defined as a set of discrete behaviors that can be analyzed and learned.12 Typically, such approaches are presented in a concentrated workshop format, and as supplemental curriculum with short-term successful outcomes.13,14 We know very little about how clinical teachers actually teach empathy to learners on an ongoing basis, and whether they endorse a skill-based method or another approach. In 1999-2000 I undertook this study to address the following question: How do physicians teach empathy to learners in the primary care setting?


Since little information exists specifically addressing the above question, for this study I adopted a qualitative method based on in-depth, open-ended interviewing. Twelve primary care physician—teachers (four each from the specialties of family medicine, pediatrics, and general internal medicine) were selected through a snowball nomination process. That is, they were identified for inclusion in the study because they were recipients of one or more student-initiated teaching awards and/or because colleagues identified them as “an outstanding clinical teacher.” All participating pediatricians were women, and all participating internists, men. Family physicians were split equally between women and men. With one exception, all were members of the same academic institution, the University of California Irvine College of Medicine. All were between 35 and 50 years old, and had between five and 20 years of teaching experience. In addition to their academic duties, all (except one hospitalist) maintained outpatient clinical practices.

All those who were invited formally consented to participate in the study. Each physician—teacher was interviewed for approximately one hour using an open-ended question route and prompts. Topics addressed included how the participants (1) defined empathy, (2) conveyed empathy to their patients, (3) taught empathy to students and residents, (4) recognized that a learner was successfully acquiring the attributes and skills of empathy, (5) worked with learners who were not successful in demonstrating empathy, and (6) regarded the risks and benefits of empathy. All interviews were audiotaped. Brief (30-minute) follow-up interviews were conducted with nine of the participants to clarify earlier statements. All respondents received a summary of the results, with a request to clarify, modify, or express disagreement.

To triangulate the data, I observed eight of the 12 faculty between one and three times each in actual teaching situations with either students or residents. While these sessions did not focus exclusively on empathy, they did provide an opportunity to evaluate the extent to which respondents' perceptions of their teaching methods accorded with their actual behaviors. There were frequent examples of respondents' doing in the clinical setting at least some of what they described when reflecting on their teaching practices, although the limited nature of the observations made it impossible to draw firm conclusions about the extent of concordance between faculty members' verbal descriptions and their actual behaviors. In these contexts, whenever possible, students also were asked about the teaching skills of these faculty, and how empathic they perceived them to be as faculty preceptors.

Grounded theory was used as the basis for analyzing the participants' responses to the interviews. The purpose of this method is to discover categories of a phenomenon (in this case, teaching empathy) evident in respondents' reports that will begin to generate a theory to explain the phenomenon. All tape recordings were transcribed. Line-by-line coding was used in analysis to identify emergent categories and compare them with other categories that addressed how respondents tried to teach empathy. Major themes were noted and used to generate “theoretical notes,” which in turn were reviewed to interpret or attribute meaning.15

This study used the constant comparative method, in which data collection and data analysis occur simultaneously. In this process, the data collection influences the results, and the emerging results influence the way in which data are collected. Thus, the actual question route was modified periodically as new issues or refinements of old issues arose based on previous answers from respondents.


What Is Empathy?

All respondents described empathy with phrases such as “putting myself in the patient's shoes” or “climbing into the same boat as the patient.” All distinguished empathy from sympathy (empathy being “feeling what the patient is feeling rather than what you would be feeling in the same circumstances”), and agreed that empathy is more than intellectual understanding or cognitive analysis. Rather, the respondents recognized that empathy involves a personal relatedness. As to the precise balance between cognitive and affective components, some disagreement did emerge. Some respondents cautioned that emotion, unfiltered by cognition, could be un-helpful and misleading to the clinician. Others stressed that the key component of empathy was the emotional connection with the patient, and without this affective bond, mere behavioral attempts at empathy would produce its antithesis.

When asked specifically whether empathy is a skill or an attitude, most respondents agreed it was both, although pediatricians and women physicians generally tended to emphasize its innate qualities (“empathy has always been a part of who I am”), while others favored a more reductionistic, behavioral definition. Some noted that they could not imagine not being empathic, while others stressed the importance of analyzing, specifying, and “working on” empathic skill development.

One respondent, who described himself as “not big on touchy—feely language” emphasized the importance of “empathy in action.” In this view, the purpose of empathy is to render more meaningful assistance to the patient by truly understanding the patient. Other respondents agreed with this concept, pointing out that empathy must include an implementation component—the willingness to help the patient in concrete, specific ways (e.g., streamlining the hospital stay, or prescribing a less expensive medication), not simply listening to the patient's problems.

Conveying Empathy to Patients

Although the respondents differed as to whether empathy was an attitude or a skill, in practice most advocated a combination of global attitudes and specific skills in communicating empathy to patients (List 1). One physician described the process as “What you allow to happen in the session— not avoiding tough, complex issues … whether your priority is getting the patient turfed, or deep listening.” Another spoke about assuming the stance of servant, and said she always tries to “stay closer to [the patient's] heart than to their face.”

List 1. Ways in Which 12 Physician-Teachers Believed They Conveyed Empathy to Patients, 1999–2000

In addition to cultivating general attitudes, most faculty also mentioned using more behavioral, observable communication skills to convey empathy, including both verbal and nonverbal behaviors. Several respondents believed that, generally, body language conveys empathy much more effectively than spoken dialogue. Finally, faculty mentioned specific, concrete actions that they felt conveyed empathy to their patients by improving their conditions and reducing suffering.

Teaching Empathy

All respondents agreed empathy should be taught as part of medical education, but several felt it was difficult to structure or formalize such teaching, and believed it was conveyed to learners “as part of the informal curriculum.” For some, teaching empathy was a somewhat “haphazard,” intuitive process. By contrast, other respondents employed an explicit, well-developed teaching schema.

Skill versus attitude

Whether respondents understood empathy primarily as a skill-based technique or an attitude also influenced their approaches to teaching. A subset of the faculty had strong feelings on this point. On the one hand, adherents of “attitudinal” teaching argued that focusing on techniques makes empathy artificial, and can actually impede the expression of sincere empathy. Respondents of this persuasion felt pat phrases that learners could memorize were not useful, and stressed that learners must develop their own empathic style, “find their own words.” But “behavioral” (i.e., skill-based) faculty felt it was extremely difficult to teach an attitude of empathy, whereas specific language and techniques could be easily transmitted and evaluated. The majority of teachers recognized that some sort of formal training in empathy could lay a foundation, but that the most important aspects of empathy “can't be conveyed theoretically.”

Role modeling: the core of teaching empathy

Almost all faculty endorsed role modeling as the most effective way of teaching empathy, although attitudinal teachers stressed the importance of whole-person, global modeling, while behavioral or skill-based teachers paid more attention to the more reductive modeling of specific language patterns and behaviors. For example, attitudinal respondents talked about showing learners how to “step into the patient's world,” while skill-based respondents focused on “eye contact” and “reflective listening skills.” However, closer questioning suggested that behavioral skills and global concepts were closely related. For example, when attitudinal teachers were asked how learners could “step into the patient's world,” they tended to talk about concrete behaviors such as eye contact and reflective listening. Conversely, when skill-based teachers were asked about the purpose of eye contact and reflective listening, they often used phrases similar to the global language of the attitudinal teachers.

Attitudinal teachers often preferred to simply demonstrate empathy, then have the learner question them afterwards about what they had done, rather than initiating explanation or instruction. One described it as “just being who you are in front of the learner.” These faculty seemed to have a somewhat fatalistic attitude toward the outcomes of such teaching, stating that “role modeling will be successful if the learner wants to learn … they'll pick it up if they want to; otherwise they won't.” One respondent elaborated on this approach to role modeling by stating that, from the learner's standpoint, direct observation of the physician should be “like watching a movie or reading a poem,” in that it should have a similar emotional impact. In her perception, learner observation should be an affective experience, as well as one involving knowledge and skill acquisition. “The student should feel something during the encounter … the student should be moved.”

Faculty with a more skill-based approach liked to make the point that “role modeling without explanation is an unreliable way of teaching,” and employed a combination of pre- and/or post-teaching encounters to highlight particular aspects of the modeled interview. For example, these faculty were more likely to mention doing preparatory work with learners, by priming learners to pay attention to empathic behaviors or verbalizations. One individual mentioned the importance of developing an empathic “strategy” prior to the encounter, and a couple of faculty suggested setting a specific empathy goal (e.g., having the resident decide to practice a specific empathic skill during an interview, such as paraphrasing).

Debriefing: the second key component of teaching empathy

Both kinds of teachers also used debriefing, although not as consistently as role modeling. A major constraint cited was time limitations. Attitudinal teachers who used debriefing tended to ask open-ended questions such as What did you notice? What was going on? What was I trying to accomplish? Skill-based teachers often concentrated their questions more narrowly: What specifically did you do to show empathy toward this patient? One skill-based respondent pointed out that debriefing was important to “break down” the “complete experience” of modeling into digestible chunks that the learner could absorb and learn from. However, faculty from the two schools of thought endorsed both kinds of questioning.

Learner-centered approaches

Faculty appeared to describe teaching techniques that emphasized a synergy between learner- and patient-centered approaches to help learners gain knowledge of empathy. Some of their efforts were directed toward getting the learner to alter, modify, or improve his or her own behavior and attitudes. Faculty who favored a skill-based approach emphasized breaking down empathy into a series of verbal and behavioral steps that learners could first visualize themselves doing and then implement. For example, these faculty commonly encouraged learners to “sit down,” “maintain eye contact,” and “avoid writing when you're listening to the patient” as ways of signaling empathy to the patient. In a skill-based mode, teachers also paid significant attention to learners' verbal communication skills and frequently used verbal correction as a teaching style, helping learners formulate more empathic responses to patients. Respondents were generally less explicit about how to get learners to focus on global attitudes judged to be conducive to empathy. However, one faculty member tried to create an “attitude of empathy” in learners by exhorting students to “take on” the patient, “adopt” the patient, “really care,” and “get involved.”

Patient-centered approaches

From an attitudinal perspective, the respondents wanted, in the words of one, to “place the person of the patient front and center” in the learner's awareness, and frequently globally reminded learners of the importance of understanding the patient for the practice of good medicine. Faculty in this mode urged students to “get to know each patient as a person,” stressed the importance of personal knowledge about patients, and often provided additional personal/historical/contextual details about patients in an effort to have learners assume a more empathic stance. In a similar vein, the respondents cautioned learners not to make assumptions about patients based on race, ethnicity, gender, culture, or socioeconomic status, or even past encounters. One faculty member stressed acceptance of the patient's perceived weaknesses and limitations, and pointed out that “you can't choose your patients any more than you can choose your family.” Faculty frequently reported encouraging learners “to look for the reasons behind a patient's problematic behavior,” “to make the assumption there is always an explanation for the way a patient responds,” and that it is the doctor's job to figure this out. A more behavioral patient-centered approach that all faculty used to get at this issue was to encourage learners to pay close attention to the patient's nonverbal cues, then use these to help the patient articulate frustrations, worries, or fears.

Another approach mentioned by several respondents that involved both attitudinal and skill-based components was encouraging the learner to pay close attention to patients' feelings. One faculty member stressed that “the patient's feelings are the key to understanding the patient,” while another tried to overcome learners' inclination to avoid patient emotion as a “private issue” by pointing out that “medicine is all about invasion of privacy.” However, the majority of respondents tended to ask learners directly about patient feelings or points of view “only with failures,” or “only when negativity is present in the encounter.” In these cases, they might get students “to imagine what's going on with the patient.” One attitudinally-oriented faculty member “used patient point of view questions all the time,” routinely making inquiries such as “How do you think the patient feels about this?” and “How do you think the patient will take this news?”

Parallel process

Several respondents, especially those with a more attitudinal orientation, commented on the importance of expressing empathy for the learner in the teaching process. Many recognized that training programs are “brutal,” and that residents in particular “are suffering and in pain.”15 Others noted modestly that it is much easier to notice gaps in empathy as the teacher than on the front line as the learner, and were sensitive to learners' feelings of being overwhelmed by patient care while being negatively judged by faculty. The respondents were all acutely aware that if there was a “disconnect” between the teaching style they demonstrated in interacting with the learner and what they required the learner to do with the patient, they would not be successful in their efforts to teach empathy. In one memorable quote, a faculty member noted, “The path to empathy must lead through the learner,” meaning that the learner will not be able to learn how to express empathy for the patient unless he or she has experienced empathy from the faculty member.

Knowing the learner is “getting it.”

Whenever possible, respondents with an attitudinal orientation tended to look directly at the patient's response to evaluate whether the learner was mastering the requisite skills and attitudes of empathy: How did the patient respond to the learner? Did the patient identify the learner as his or her doctor? Did the patient return to the resident, or ask to see the student again? It appeared to matter less to them what the learner was doing than how the patient reacted to how the learner was being. Skill-based respondents preferred to pay attention to learners' behaviors in the patient encounter such as tone of voice, manner, the use of open-ended questions, and reflective listening, as well as actual empathic statements.

Primarily, faculty of either orientation judged learners' skill and attitude acquisition by the nature of their case presentations, since, for practical reasons, they did not always have access to direct observation of the learner and the patient. Their criteria for success were remarkably consistent. Specifically, they expected that presentations of successful learners would begin to show evidence of close attention to the patient's actual language, more empathic statements about the patient, more psychosocial information, and more mention and recognition of the patient's feelings. Learners would begin to assess and prioritize the patient's problems differently, in a way that reflected the patient's perceptions as well as their own. Similarly, their treatment plans would contain more ideas and suggestions about how to help the patient. Finally, the learner would be more willing to take action to benefit the patient, such as coordinating care or making referrals. Attitudinal teachers also commented on global changes they observed in learners, describing them as “more compassionate,” “less dismissive,” “more caring,” and seeming to get more personal satisfaction out of what they were doing.

What to do when the learner isn't “getting it.”

Confronted with this situation, faculty responses were varied. Some maintained they rarely encountered this problem. Others, particularly those with more attitudinal approaches, admitted to not having any good remedies. Still others suggested “just keep pushing” the learner—continuing to ask questions about the patient as a person, the patient's life context, and point-of-view questions. Not surprisingly, behavioral teachers emphasized skill repetition and practice. Several respondents in both camps were fatalistic. “Not much can be done—some people aren't in medicine to care for patients.”

Risks and benefits of empathy

Faculty seemed aware of the risks of empathy for themselves and for their learners. Several faculty holding both attitudinal and behavioral orientations mentioned that empathy can be an overwhelming and emotionally burdensome experience. One faculty member commented that empathy makes the physician more vulnerable to being taken advantage of by a demanding or exploitative patient. This same respondent thought there was a fine line between empathy and co-dependence. Another respondent worried that an “excess of empathy” might make the physician unable to be objective and provide reliable care. But all felt that in their own practices, they knew how to cope with the downside of empathy. Several mentioned compartmentalization, or the ability to set aside feelings for one patient as they moved on to the next (as one faculty member memorably put it, “At some point you have to climb out of the boat”). Another described the experience of “touch-and-go” empathy, “diving into” an empathic position, only to “resurface quickly” to a more superficial level of interaction.

All respondents agreed that empathy made them better and happier clinicians. They stated it not only improved their relationships with patients, but also made the practice of medicine more rewarding, more interesting, less frustrating, and more pleasurable, “a way of making medicine feel more human.” Because of empathy, these faculty agreed that they could take better care of their patients by devising better treatment plans, and were more likely to get patient buy-in and compliance with their recommendations. Several attitudinal teachers stated bluntly that having an empathic connection with the patient was “why they went into medicine.”


Based on their interview data, it was fairly easy to place the 12 respondents on a continuum from “attitudinal” to “behavioral” and, to some extent, clinic observation confirmed this perception. Behavioral, skill-based teachers tended to be more explicit, analytic, and reductive, while attitudinal teachers favored providing a more global experience for learners. Nevertheless, in general, both kinds of faculty tended to move comfortably between both kinds of teaching. Behavioral teachers might exhort learners periodically to “think about the patient as a person,” a decidedly global, non-demonstrable injunction; whereas attitudinal faculty were likely to model paraphrasing or body language, techniques drawn from a skill-based armamentarium. Similarly, informal student feedback did not distinguish between the quality of teaching of the two kinds of faculty. While students seemed to feel closer to attitudinal faculty, and to have more personal relationships with them, they often waxed eloquent about the uncanny ability of skill-based faculty to “really understand what was going on with the patient, not just on the surface, but down deep, things I hadn't seen at all.”

Although faculty differed philosophically as to whether empathy is primarily a measurable, behavioral skill or a global attitude, in terms of both their actual clinical practices and their teaching, all seemed to move comfortably between attitudes and skills, sometimes disclosing global dispositions, at other times focusing on demonstrable behaviors. Indeed, the key phenomenon to emerge from this study seemed to be the balance of attitudinal and behavioral components of empathy that all respondents achieved. In the “attitudinal” group, the skill-based dimension was often more implicit in their teaching, but was definitely present. Similarly, in the “behavioral” group, while these faculty were less likely to verbalize sentiments about “caring for the whole person” or “knowing the patient's heart,” they did use this rhetoric on occasion; and were acknowledged by learners to be deeply caring physicians, not mere technicians. The respondents consistently revealed a willingness to make the process of patient care more transparent and accessible to learners, and a desire to coach learners in the thoughts, feelings, and attitudes that promote empathic responses to patients. The acknowledged excellence of these teachers suggests the importance of teaching empathy in its totality, acknowledging both behavioral and attitudinal tools for enhancing empathic awareness in learners.


1. Landau RL. And the least of these is empathy. In: Spiro H, Curnen MGM, Peschel E, St. James D (eds). Empathy and the Practice of Medicine. New Haven, CT: Yale University Press, 1993:103–9.
2. Gianakos D. Empathy revisited. Arch Intern Med. 1996;156:135–6.
3. Markakis KM, Beckman HB, Suchman AL, Frankel RM. The path to professionalism: cultivating humanistic values and attitudes in residency training. Acad Med. 2000;75:141–50.
4. Prislin MD, Feighny K, Stearns JA, Hood J, Arnold L, Johnson L. What students say about learning and teaching in longitudinal ambulatory primary care clerkships: a multi-institutional study. Acad Med. 1998;73:680–7.
5. Prislin MD, Fitzpatrick CF, Munzing T, Oravetz P, Radecki S. Ambulatory primary care medical education in managed care and non-managed care settings. Fam Med. 1996;28:478–83.
6. Lu MC. Why it was hard for me to learn compassion as a third-year medical student. Cambridge Quarterly of HealthCare Ethics. 1995;4:454–8.
7. Prislin MD, Giglio M, Lewis EM, Ahearn S, Radecki S. Assessing the acquisition of core clinical skills through the use of serial standardized patient assessments. Acad Med. 2000;75:480–3.
8. Davis M. Intern discussion group: a supportive educational experience for junior doctors. Hosp Med. 1999;60:435–9.
9. Wipf JE, Pinsky LE, Burke W. Turning interns into senior residents: preparing residents for their teaching and leadership roles. Acad Med. 1995;70:591–6.
10. McKegney CP. Medical education: a neglectful and abusive family system. Fam Med. 1987;21:452–7.
11. Platt FW, Keller VF. Empathic communication: a teachable and learnable skill. J Gen Intern Med. 1994;9:222–6.
12. Burack JH, Irby DM, Carline JD, Root RK, Larson EB. Teaching compassion and respect: attending physicians' responses to problematic behaviors. J Gen Intern Med. 1999;14:49–55.
13. Smith RC, Lyles JS, Mettler J, et al. The effectiveness of an intensive teaching experience for residents in interviewing: a randomized controlled study. Ann Intern Med. 1998;128:118–26.
14. Winefield HR, Chur-Hansen A. Evaluating the outcome of communication skill teaching for entry-level medical students: does knowledge of empathy increase? Med Educ. 2000;34:90–4.
15. Strauss A, Corbin J. Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Newbury Park, CA: Sage, 1991.
© 2002 Association of American Medical Colleges