Prior studies have reported that, despite the importance of empathy in doctor/patient interactions, medical education leads to deterioration in empathy among medical students and residents.1,2 Clinical and therapeutic advantages of empathetic patient care include improved doctor–patient communication,3 increased patient satisfaction,4,5 greater patient compliance,4,6 decreased litigation,7 increased physician job satisfaction,8 and decreased physician burnout.9
In health care, empathy is defined as “a cognitive attribute that involves an understanding of the inner experiences and perspectives of the patient as a separate individual, combined with a capability to communicate this understanding to the patient”10 and “act on that shared understanding in a helpful and therapeutic way.”11 A possible explanation for a decline in empathy during medical education may be found in seminal papers by Hafferty12 and by Hafferty and Franks.13 These describe a “hidden” or “informal curriculum” in medical school and a learning environment in which students adopt behaviors in an ad hoc and unstructured manner. In this environment, negative role models may exert a powerful formative influence.14 Thus, Kenny and colleagues15 describe the imperative for medical students to obtain positive “professional character formation” and to “develop safe spaces where negative role modeling can be reflected on and translated into an effective learning experience.” More recently, medical educators have described moral distress, burnout, and depression in third- and fourth-year medical students.16 Burnout has been associated in third-year students with “self-reported unprofessional conduct and less altruistic professional values.”17 Spiegel and Siegel18 write of the need for medical students to understand the “chaotic and challenging circumstances of medical school” so that it does not “threaten their identity or progress in the medical profession,” thus “ensur[ing] that they become the compassionate, effective physicians they envision.”
Researchers have recently conducted many studies on student empathy and have generally found that empathy in medical students decreases toward the end of their undergraduate medical education experience.1,2,19,20 These studies of empathy in medical students are of varying design (cross-sectional versus longitudinal), and they have employed varying instruments, including the Jefferson Scale of Physician Empathy Medical Student Version (JSPE-MS),1 the Balanced Emotional Empathy Scale (BEES),2 and Hogan's Empathy Scale (HES).21 Diseker and Michielutte,21 for example, observed decreased emotional empathy, measured by the HES, among medical students before and after clinical experiences. Newton and colleagues2 conducted a longitudinal study of empathy, pooling data from four contiguous medical school classes, and observed a statistically significant decrease in empathy during the third year of medical school as measured by the BEES. Hojat and colleagues20 conducted a longitudinal study of third-year medical students at Jefferson Medical College using the JSPE-MS and noted a statistically significant decline in empathy. In a subsequent report on 456 students across four years of medical school, they found no significant change in empathy during the first two years, but a significant decline in empathy at the end of the third year through graduation; however, 27% of the students in the study did not show this decline in empathy.1 Finally, Chen and colleagues22 conducted a cross-sectional study at Boston University School of Medicine, measuring empathy using the JSPE-MS for all four classes in 2006. They found significant differences in mean empathy scores when comparing data from the end of the second year to the end of the third year. However, a recent study questioned whether “reports of the decline of empathy during medical education are greatly exaggerated”; in particular, the alleged magnitude of the decline and the subjective nature of the assessment received criticism.23
Several authors have suggested that educational efforts to enhance humanism and incorporate it into the curriculum may help overcome the perceived decline in empathy during medical school.24–26 Shapiro and colleagues,27 measuring empathy with the Empathy Construct Rating Scale and the BEES, reported a significant increase in empathy in 22 first-year students exposed to an eight-hour (total) literature and medicine elective. Shapiro et al28 reported the preservation of empathy in 26 third- and fourth-year students after a multidimensional humanities elective. Qualitative research by DasGupta and Charon29 suggests a positive change in medical student empathy through reflective writing and the “personal illness narrative.” However, limitations of these studies include heterogeneous measurement of empathy, small sample sizes, and variable timing of curricular interventions.
Innovative programs designed to enhance the importance of compassionate patient care, such as those developed by the Arnold P. Gold Foundation for Humanism in Medicine, may also work to preserve empathy.30 Recently, the Gold Foundation created the Gold Humanism Honor Society (GHHS) to recognize students, residents, and physicians who are exemplars of humanism (defined as empathy, compassion, altruism, responsibility, and respect) in doctor/patient interactions. Third-year students select their peers for the GHHS through a well-validated process,31 and membership entails a service requirement. Membership in the GHHS is distinct from membership in Alpha Omega Alpha (AOA), the national medical honor society for medical students, residents, physicians, and scientists in the United States and Canada.32 Nomination for AOA requires class rank in the top quartile and leadership characteristics, whereas membership in the GHHS is based on peer nomination using a validated questionnaire.
In 2007, 10 Robert Wood Johnson Medical School (RWJMS) student, resident, and faculty GHHS members met with their RWJMS–GHHS faculty advisor (who was, at the time, also a dean of student affairs) to design an educational intervention to maintain empathy in third-year students. We believed, on the basis of student reports, that erosion of empathy was an issue in medical education, especially during the clinical years. Further, none of us had had formal opportunities to discuss the challenges of the clerkship years during our own years in undergraduate medical education. We hypothesized that a curricular intervention during third-year clerkships might be a way to attenuate the loss of empathy.
The purpose of our study was to evaluate JSPE-MS scores of two consecutive medical school classes in order to assess the impact of an empathy-preserving curricular innovation. In addition, we evaluated the relationship between the maintenance of empathy during the clerkship year and student demographic characteristics, including membership in the GHHS.
We evaluated, using the JSPE-MS, 107 students in the RWJMS class of 2009 and 102 students in the RWJMS class of 2010 before they began their clinical clerkships. The JSPE-MS consists of 20 questions measured on a Likert-like scale of 1 to 7 (a higher score implies more empathy, and the maximum score is 140). Prior research has shown the instrument to be valid for use with physicians, medical students, and other health professionals.1,33,34
We administered the JSPE-MS to RWJMS medical students in a large-group setting and coded answer sheets with a unique identifier to maintain anonymity. At the conclusion of the third year, we readministered the JPSE-MS, using the identifier. Also at the end of the clerkship year, we administered two supplementary questionnaires to gather demographic information on each student and to assess student satisfaction with the intervention. Demographic information collected included gender, age, race, history of hospitalization, having close family members with illness, and membership in GHHS. Student satisfaction questions queried students about the effect of the intervention curriculum on their awareness of positive and negative role models, the preservation of their innate empathy, their awareness of the importance of empathy in patient care, their ability to cope with everyday stressors in their clerkships, and their ability to recognize burnout in themselves and others. We offered no incentives for completing the questionnaires.
During the six required third-year rotations (medicine, surgery, pediatrics, obstetrics–gynecology, family medicine, and psychiatry), students participated in interactive sessions as part of an intervention entitled “Humanism and Professionalism” (H&P). The clerkship directors agreed that we could dedicate one hour of didactic time per rotation for each of these H&P sessions. Two of us (S.R. and B.G.) divided the sessions between us and communicated regularly to ensure that we independently delivered the same curriculum. All students on each rotation (approximately 15) attended their clerkship-specific session (see Table 1 for a comparison of curricular details in the 2009 and 2010 cohorts). During each session, we allotted time for students to debrief about the emotionally intense events they experienced and to share observations about positive and negative role models. During the first clerkship, students suggested that they would like to post reflective comments via anonymous blogs on our password-protected WebCT system (Blackboard, Inc., Washington, DC). WebCT is an online educational software system that is sold to colleges and other institutions for electronic learning.35 Blogging soon became an H&P requirement; we asked each student to post one entry per clerkship, and students were on their honor to comply. During H&P sessions, we used the blog posts as triggers for discussions on students' reactions to their clerkship experiences. In addition, one of us (S.R.) culled reflective journal articles from the New England Journal of Medicine's Perspectives column,36,37 from the Annals of Internal Medicine's “On Being a Doctor” column,38 from Academic Medicine,39 from the American Medical Association's Virtual Mentor,40,41 and from The New York Times 42,43 to discuss at the sessions. The facilitators explicitly told students that the main purpose of the H&P sessions was to maintain their innate humanism and professionalism. We provided consistent reinforcement of this objective through discussions of positive and negative role models, patient care experiences, morally distressing events, and students' reactions to all of these.
At midyear, many of the RWJMS class of 2009 students reported burnout; thus, we involved all students in an “appreciative inquiry” exercise to discuss positive aspects of their clerkship experience. Appreciative inquiry is a process whereby individuals in an organization ask questions that strengthen and highlight the positive potential of a system.44,45 We held two required large-group (about 50 students) evening exercises for the class of 2009 to discuss “health care as a human right.” During these events, students viewed films and heard panel discussions concerning health care in the United States and abroad. We asked for feedback from students in January and at the end of the third year (RWJMS class of 2009), or only at year's end (RWJMS class of 2010), about their satisfaction with the H&P curriculum and their perceptions of its effects on their professional development (Table 1).
The RWJMS curriculum was evolving, and students in the class of 2010 were the first to be enrolled in a new, four-year, “Patient-Centered Medicine” (PCM) course (Table 1). This course involves a significant time commitment, including early clinical experience, seminars with mentors, large-group discussions, and a longitudinal clinical experience. During their first and second years of medical school, the class of 2010 experienced 80 contact hours in PCM I and PCM II. The H&P curriculum was integrated into their PCM III course. Although the H&P clerkship sessions remained the same for the class of 2010, a requirement of 24 hours (12 seminar hours and 12 longitudinal clinical hours) replaced the two evening sessions experienced by the class of 2009 (Table 1).
Because of the introduction of the PCM course for the class of 2010, we analyzed JSPE-MS scores separately for the two classes. We compared the change in JSPE-MS scores for each class after completion of all six required rotations using paired t tests (MS Excel, Data Analysis Module, 2007, Redmond, Wash). We set the α level for significance at P < .05. We excluded one extreme outlier in the class of 2010 whose scores were greater than two standard deviations below the mean. The RWJMS institutional review board approved this study.
We detected no significant decreases in empathy scores at the end of the third-year clerkships for either the RWJMS class of 2009 (mean JSPE = 115.4 versus 113.9, P = .14) or the RWJMS class of 2010 (mean JSPE = 112.4 versus 110.5, P = .07; Table 2). When we compared the pretest empathy scores of the class of 2009 with the pretest empathy scores of the class of 2010, we detected no statistically significant differences (class of 2009 = 115.1, class of 2010 = 112.4, P < .10; Table 2). Figures 1 through 3 and Tables 3 and 4 present the analyses of and relationships between (1) the change in empathy scores and (2) demographics, both before and after the intervention for each class.
Overall, feedback from students' self-reported perceptions of the H&P initiative showed that they were satisfied with the curriculum. Students in both classes commented that it helped them identify positive and negative role models and prevent burnout.
RWJMS class of 2009
Change in empathy scores.
We received both pretest and posttest data from a total of 89 (83% of 107) students (43 female, 46 male). Mean change in pretest/posttest empathy score was not significant for the group as a whole (pre = 115.4, post = 113.9, P = .135; Table 2). These scores approximated pretest scores reported in previous studies, but our sample did not decline as sharply, or significantly.1,10
We found no statistically significant change between pretest and posttest empathy scores when analyzing data by gender, intended specialty choice, age, career prior to entering medical school, experience of illness/death of a close friend and/or family member, or prior hospitalization (Table 3). When analyzing pretest scores by demographic characteristics, female students, students entering core specialties (internal medicine, pediatrics, family medicine, obstetrics–gynecology, and psychiatry), and those with prior hospitalization had significantly higher empathy scores compared with male students, students entering noncore (all other) specialties, and those not previously hospitalized (Figure 1). Female students, those entering core specialties, and those previously hospitalized continued to have higher posttest empathy scores when compared, respectively, with male students, students entering noncore specialties, and students who had not been hospitalized previously (Figure 2 [top]).
Subgroup analysis for GHHS.
Subgroup analysis showed that JSPE-MS scores of students selected for GHHS (n = 15) showed significant differences from their classmates' scores. Although their pretest scores did not differ significantly from the class as a whole before the intervention (Figure 1), GHHS students had posttest empathy scores significantly higher than the other students in the class (n = 67) who were not elected to the GHHS (120.6 versus 112.0, P < .00022; Table 3). In addition, those students who were not GHHS members had significant declines in empathy between the pretest and posttest (114.5 versus 112.0, P < .02), whereas GHHS members' scores did not change, even increasing, albeit not significantly (118.5 versus 120.6, P < .32; Table 3). GHHS students in the class of 2009 were aware that they had been selected for GHHS when we administered the posttest JSPE, unlike the subsequent cohort.
RWJMS class of 2010
Change in empathy scores.
We received both pretest and posttest data from a total of 73 (71% of 102) students (39 female and 34 male). Mean change in pretest/posttest empathy score was not significant for the class as a whole (pre = 112.4, post = 110.5, P < .07; Table 2).
Demographic analysis of the class of 2010 showed lower pretest and posttest empathy scores in males than in females (pretest score: males 110.6, females 114.2, P < .1; posttest score: males 108.1, females 112.6, P < .05; Table 4). Unlike students in the class of 2009, JSPE scores for students in the class of 2010 showed significant decline if they were older than 24 years of age when entering medical school, had another career prior to medicine, or had experienced an illness in a loved one (Table 4). Similar to the female students in the class of 2009, the female students in the class of 2010 had significantly higher posttest empathy scores (compared with their male classmates); however, unlike in 2009, students in core specialties, GHHS members (prenotification), and those who had been previously hospitalized did not have significantly higher posttest scores compared with their classmates who were in the noncore specialties, who were not GHHS members, and who had not been hospitalized (Figure 2 [bottom]).
Subgroup analysis for GHHS.
Because of the striking posttest difference in JSPE score in GHHS versus non-GHHS students in the class of 2009, we hypothesized that the GHHS “label” may have had an effect on the GHHS students' posttest JSPE scores; therefore, we did not inform the students in the class of 2010 who were selected for GHHS of their selection until after we administered the posttest JSPE. For these students, empathy scores initially showed a significant decrease on the JSPE posttest (116.1 versus 109.7, P < .03; Table 4). Two months later, we notified the class of 2010 of the results of the GHHS selection; those students who had been selected for GHHS completed the posttest once again. The only information they received was that the investigators wished them to take the JSPE posttest a second time. Scores on the JSPE-MS postdisclosure rose significantly, returning to the level of pretest scores (109.7 versus 115.3, P < .016; Figure 3).
Discussion and Conclusions
To our knowledge, this is the first primary data study that demonstrates preservation of empathy in two consecutive third-year medical school classes. Although we cannot be certain that the H&P intervention was responsible for the lack of decline in empathy, student feedback indicated that the sessions helped them “prevent burnout” and recognize positive and negative role models. Throughout the sessions, students expressed excitement and pride in helping to make a diagnosis, altruism and concern in their attitudes toward patients, and admiration and respect for positive role models. We used blogs and trigger articles to initiate discussion of fear of failure, dismay at the behavior of negative role models, and guilt at being privy to very private moments in patients' lives—as well as appreciation for that same privilege. Students agreed that both admitting their own mistakes and watching others disclose medical errors to patients was difficult. They admitted to feelings of insecurity in their own knowledge and skills, and they expressed relief both at the opportunity to discuss these feelings and at the realization that their classmates shared these feelings. Feedback from students indicated that small-group discussion and blogging were the most useful components of the H&P sessions. Anonymous blogging may be more efficacious than traditional forms of narrative writing for Generation Y medical students46; it is comfortable, anonymous, interactive, and shared.
We are encouraged by the fact that, by the end of the year, our students' self-reported agreement with the statement, “Viewing things from a patient's perspective is not difficult,” increased (as measured by the JSPE). To illustrate, one student wrote in his final blog:
We've all seen examples this year of sarcastic and uncompassionate behavior. Just remember that as the only physician in the room of 10-plus caregivers, you're the top dog and you may very well control the tone of the meeting. If you lack compassion and empathy, it may make it more difficult for others around you, or worse, your behavior may be contagious…. Showing compassion and approaching each patient with empathy is never naïve; it's called being a good doctor. If bad behavior can be contagious, then maybe empathy and compassion can be too.
The difference in JSPE empathy scores between students whose peers selected them for the GHHS and their non-GHHS classmates was a serendipitous finding. The GHHS was instituted at RWJMS as a means of reinforcing humanistic values in our medical school community. GHHS students' behavior, assessed by peers, identifies them as a distinct group.31 Results for GHHS students in both classes show that these students scored differently than classmates on both pretests and posttests. The RWJMS class of 2009 pretest and posttest scores were higher for GHHS students whose scores rose significantly at the conclusion of their clerkship year. We hypothesized that because GHHS students in this class knew of their selection prior to their posttest, being chosen for GHHS may have reinforced their self-identification as empathetic physicians and resulted in a rise in JSPE scores. This explanation is corroborated by data from GHHS members in the class of 2010 whose posttest JSPE scores initially declined but later rose significantly when we notified them of their GHHS status. In contrast, students elected into AOA showed no significant differences in JSPE scores before and after the third year (data not shown). Peer validation of GHHS students as empathic caregivers may have restored their perceived identity, underscoring the interpersonal nature of empathy assessment.
Two factors that limited our study are the inclusion of only two classes of medical students at a single institution (which limits generalizability) and the before-and-after design (which limits inferences about intervention effects). Further, we were unable to account for the differences in JSPE-MS posttest scores between the classes of 2009 and 2010 for two subgroups of students: those who had previously been hospitalized and those who were in the GHHS (before learning of their GHHS status). Students in these two subgroups in the class of 2010 showed a decrease in empathy on the JSPE posttest, as opposed to students in the class of 2009. Possibly, the differences in numbers of students who were available for the posttest in each year (because of scheduling conflicts and absenteeism) reduced our power to detect significant differences in these factors across the two classes. In both years, we recorded higher empathy scores at the pretest for older students, those who had a close friend or family member who had experienced illness, and those who were in the GHHS.
The JSPE-MS, although a validated tool, is a self-reported and, thus, subjective measure that may not objectively capture empathic behavior. Three-hundred-sixty-degree assessment may be preferable but is difficult to obtain in a clerkship setting. Others have noted the need for patient assessment of physicians to validate empathy.23
Members of the RWJMS class of 2010 differed from those of the RWJMS class of 2009 in prior experience, age at entering medical school, and exposure to a PCM course. Future studies with larger samples and a more homogeneous curricular experience may minimize sample variability and reveal stronger relationships in trends we observed.
Nonetheless, our findings suggest that empathy may be preserved in medical school despite prior evidence that a decline in empathy is pervasive; we believe that the H&P intervention may have attenuated this decline. Future studies that employ a large controlled trial in multiple institutions are needed to confirm these findings.
On the basis of our experience with two classes of students at RWJMS, we found that empathy may be preserved in third-year medical students. Furthermore, a curriculum that includes time for third-year students to share feelings in a protected and familiar venue during their rotations may attenuate a decline in empathy. In addition, programs like the GHHS, which validate humanistic behavior, may contribute to preservation of positive professional identity.
The authors thank Kaye Maxwell of the Thomas Jefferson Medical School for scanning and preparing the data for analysis. The authors also wish to thank Robert Wood Johnson Medical School for providing support services for this study.
Dr. Rosenthal reports having received grant support (but no consulting fees) for this study from the Arnold P. Gold Foundation for Humanism in Medicine. The Gold Foundation did not consult with the authors on the study design, did not monitor either the conduct of the study or the collection of data, did not assist the authors with either the analysis or the interpretation of data, and was not involved in either the preparation or review of the manuscript.
Thomas Jefferson Medical School received compensation for scanning and preparing the data for analysis.
The Robert Wood Johnson Medical School (New Brunswick, New Jersey) institutional review board approved this study.
The authors presented portions of this work at the Arnold P. Gold Foundation for Humanism in Medicine Third Biennial Conference in Chicago, Illinois, September 2008.
The opinions in this article are those of the authors alone and do not necessarily reflect those of the Arnold P. Gold Foundation for Humanism in Medicine or Robert Wood Johnson Medical School.
3Beckman HB, Frankel RM. The effect of physician behavior on the collection of data. Ann Intern Med. 1984;101:692–696.
4Kim SS, Kaplowitz S, Johnston MV. The effects of physician empathy on patient satisfaction and compliance. Eval Health Prof. 2004;27:237–251.
5Roter DL, Stewart M, Putnam SM, Lipkin M Jr, Stiles W, Inui TS. Communication patterns of primary care physicians. JAMA. 1997;277:350–356.
6Squier RW. A model of empathic understanding and adherence to treatment regimens in practitioner–patient relationships. Soc Sci Med. 1990;30:325–339.
7Levinson W, Roter DL, Mullooly JP, Dull VT, Frankel RM. Physician–patient communication. The relationship with malpractice claims among primary care physicians and surgeons. JAMA. 1997;277:553–559.
8Suchman AL, Roter D, Green M, Lipkin M Jr. Physician satisfaction with primary care office visits. Collaborative Study Group of the American Academy on Physician and Patient. Med Care. 1993;31:1083–1092.
9Thomas NK. Resident burnout. JAMA. 2004;292:2880–2889.
10Hojat M, Gonnella JS, Nasca TJ, Mangione S, Vergare M, Magee M. Physician empathy: Definition, components, measurement and relationship togender and specialty. Am J Psychiatry. 2002;159:1563–1569.
11Mercer SW, Reynolds WJ. Empathy and quality of care. Br J Gen Pract. 2002;52(suppl):S9–S12.
18Spiegel J, Siegel BS. The Mindful Medical Student: A Psychiatrist's Guide to Staying Who You Are While Becoming Who You Want to Be. Lebanon, NH: Dartmouth College Press; 2009.
19Hojat M, Mangione S, Nasca TJ, et al. The Jefferson Scale of Physician Empathy: Development and preliminary psychometric data. Educ Psychol Meas. 2001;61:349–365.
22Chen D, Lew R, Hershman W, Orlander J. A cross-sectional study of medical student empathy. J Gen Intern Med. 2007;22:1434–1438.
25Stepien KA, Baernstein A. Educating for empathy. A review. J Gen Intern Med. 2006;21:524–530.
27Shapiro J, Morrison E, Boker J. Teaching empathy to first year medical students: Evaluation of an elective literature and medicine course. Educ Health (Abingdon). 2004;17:73–84.
28Shapiro J, Rucker L, Robitshek D. Teaching the art of doctoring: An innovative medical student elective. Med Teach. 2006;28:30–35.
33Sherman JJ, Cramer A. Measurement of changes in empathy during dental school. J Dent Educ. 2005;69:338–345.
34Glaser KM, Markham FW, Adler HM, McManus PR, Hojat M. Relationships between scores on the Jefferson Scale of physician empathy, patient perceptions of physician empathy, and humanistic approaches to patient care: A validity study. Med Sci Monit. 2007;13:CR291–CR294.
36Treadwell K. The code. N Engl J Med. 2007;357:1273–1275.
37Klass P. The moral of the story. N Engl J Med. 2008;358:2313–2316.
38Shapiro AE. On being a doctor: An uneasy understanding. Ann Intern Med. 2007;147:811–812.
44Inui TS. A Flag in the Wind: Educating for Professionalism in Medicine. Washington, DC: Association of American Medical Colleges; 2003.
45Cooperrider DL, Whitney DK. A Positive Revolution in Change: Appreciative Inquiry. San Francisco, Calif: Berrett-Koehler; 2005.