2012; 34: 305–311

Characterizing changes in student empathy

throughout medical school
DANIEL C. R. CHEN1, DANIEL S. KIRSHENBAUM1, JUN YAN2, ELAINE KIRSHENBAUM3 & ROBERT
H. ASELTINE4
1
Boston University School of Medicine, USA, 2University of Connecticut, USA, 3Massachusetts Medical Society, USA,
4
University of Connecticut Health Center, USA

Abstract
Background: Empathy is important in the physician–patient relationship. Prior studies suggest that medical student empathy
declines with clinical training.
Aims: We examined the trend of empathy longitudinally; determined differences in empathy according to gender and medical
specialty preferences; and determined empathy and career preference differences among students admitted through different
medical school admission pathways.
Method: The data for this study were collected using a longitudinal cohort design and included 2652 observations nested within
1162 individuals. Participants were medical students at a university-based medical school surveyed yearly from 2007 through 2010.
Empathy was measured by the Jefferson Scale of Physician Empathy-Student Version (JSPE-S), a validated, 20-item self-
administered questionnaire. Predictors of JSPE-S scores included gender, age, anticipated financial debt upon graduation and
future career interest.
Results: Empathy scores of students in preclinical years were higher than in clinical years. Gender was a significant predictor of
empathy, with women having higher empathy scores than men. Students preferring technology-oriented specialties had lower
empathy scores. When career preference was controlled, higher levels of debt were significantly associated with greater empathy.
Students with high baseline empathy decreased less than students with low baseline empathy during medical school. Students in
traditional four-year medical school programs had higher baseline empathy than those in early pathway programs.
Conclusions: Self-reported empathy for patients, a possibly critical factor in high-quality patient-centered care, wanes as
students advance in clinical training, particularly among those entering technology-oriented specialties. In the era of new health
care policy and primary care shortages, our research may have implications for the medical education system and admission
policy.

Introduction
Practice points
Empathy in the physician–patient relationship is the physi-
cian’s ability to recognize and understand a patient’s perspec- . Medical student empathy declines during medical
tives and experiences, and convey such an understanding school.
back to the patient (Coulehan et al. 2001; Hojat et al. 2001). . Students with higher baseline empathy, at the start of
This understanding allows patients to feel respected and medical school, have a slower rate of decline than those
validated (Beckman et al. 1994), promotes patient and with lower baseline empathy.
physician satisfaction, and may improve patient outcomes . Students in traditional four-year medical school pro-
(Suchman et al. 1993; Coulehan et al. 2001; Vermeire et al. grams have higher baseline empathy than students in
2001; MacPherson et al. 2003; Bikker et al. 2005). Empathy is early admissions pathway programs.
one of the Association of American Medical Colleges’ (AAMC
1998) goals for the development and education of altruistic the clinical years when compared to the preclinical years
and compassionate physicians. (Hojat et al. 2004; Chen et al. 2007).
There is concern among educators that clinical training may The aging of the US population and the expansion of health
adversely affect medical resident and student empathy. In insurance coverage due to health reform may result in
internal medicine residents, empathy was measured to be shortages of primary care physicians over the next decade
highest at the beginning, but decreasing by the end of (Gordon 2010). Prior research shows that practicing physicians
internship, and remained low throughout residency (Bellini with higher measured empathy are found in primary care
et al. 2002; Bellini & Shea 2005). Among medical students, specialties – internal medicine, family medicine (Hojat et al.
studies have demonstrated that empathy tends to be lower in 2002) – and that students with higher measured empathy are

Correspondence: D.C.R. Chen, Boston University School of Medicine, 801 Massachusetts Avenue, 2nd Floor, Boston, MA 02118, USA. Tel: (617)
638-8058; fax: (617) 414-4676; email: daniel.chen@bmc.org
ISSN 0142–159X print/ISSN 1466–187X online/12/040305–7 ß 2012 Informa UK Ltd. 305
DOI: 10.3109/0142159X.2012.644600
D. C. R. Chen et al.

often more interested in primary care related fields (Chen et al. historically black institutions, as well as four institutions with
2007). As such, empathy may be predictive of those students under-represented minority populations. During the summer
who enter into primary care specialties. Educational interven- after their sophomore year in college, students accepted into
tions and policies could target empathy to address the demand EMSSP take Boston University undergraduate courses for
for primary care physicians. credit toward their college degree. Their senior year of college
Many medical schools have sought to ‘‘answer the call’’ is spent at Boston University, and students take BUSM courses
concerning physician shortages by increasing medical school for advanced standing.
enrollment and by providing alternative admission pathways,
such as early medical school admission and pathways
specifically for underrepresented minorities (Boston Study design and measures
University 2011). The AAMC (2006) has called for a 30% This study utilized a longitudinal cohort design involving 1162
increase in medical school class size from 2002 to 2012. medical students from the 2007 to 2013 graduating classes at
However, how alternative educational pathways and increased BUSM. For all participating classes, brief self-administered
class sizes will impact the potential for schools to cultivate surveys were administered on up to five occasions. Incoming
empathetic physicians committed to patient-centered care is first-year medical students were emailed yearly to voluntarily
not clear. complete an online survey measuring ‘‘student attitudes
This study seeks to extend prior empathy research by toward medicine’’ during Orientation Week of medical
examining long-term trajectories of empathy among medical school (last week of August to first week of September),
students at the Boston University School of Medicine (BUSM). prior to the beginning of first-year medical school classes.
Additionally, this study seeks to determine how measured Subsequently, first- through third-year medical students were
empathy differs by gender; expected debt at the end of asked to complete the survey yearly from March to May. The
medical school; different medical student career-specialty survey was administered to each class during their end-of-year
preferences; and medical school admission pathways. Objective Structured Clinical Examinations (OSCEs). Fourth-
year medical students were asked to complete the survey in
February, one month prior to Match Day. Students in each
Methods class were reminded to complete the survey in bi-weekly
emails for a month following the initial invitation. Survey
Study participants
response rates by study wave and cohort ranged between 54%
All medical students (incoming medical students and those and 99%, with an average of 81% across all waves and cohorts.
completing first- through fourth-year of medical school) at The number of participants in each study wave and the
BUSM from August 2006 through June 2010 were eligible to response rates in each wave are presented in Table A1.
participate in the study. The BUSM curriculum is a traditional The Jefferson Scale of Physician Empathy-Student Version
four-year medical school with two years of preclinical study, (JSPE-S) is a self-administered 20-item instrument measuring
with limited patient contact in the form of weekly physician components of empathy among health professionals in
shadowing for 10 weeks and a weekly patient interviewing patient-care situations (Hojat et al. 2001, 2002, 2004).
and examination course for six weeks, followed by two years Respondents indicated their level of agreement on a seven-
of clinical clerkships and electives. point Likert scale. Scores ranged from 20 to 140, with higher
BUSM is known for the various medical school admission values indicating a higher degree of empathy (Hojat et al.
pathways that are available to incoming students. These are 2001). In past studies, total scores among medical students
the Traditional, Seven-Year Liberal Arts/Medical Education ranged from 115 to 123.1 and standard deviations (SD) ranged
Program (SMED), Modular Medical Integrated Curriculum from 9.9 to 14.1 (Hojat et al. 2001, 2002, 2004; Sherman &
(MMEDIC), Engineering Medical Integrated Curriculum Cramer 2005; Chen et al. 2007).
(ENGMEDIC), and Early Medical School Selection Program Participants were asked to indicate their gender, age,
(EMSSP). The alternative pathways make up approximately anticipated financial debt, and likelihood of choosing various
30% of each entering class. The SMED program, established in specialties. Career-specialty intentions were categorized into
1961, admits highly qualified high school graduates into an two groups, ‘‘people-oriented’’ specialties and ‘‘technology-
accelerated pathway, where they receive their BA and MD in oriented’’ specialties, which were based on specialty groupings
only seven years. This compression is achieved using summer determined in prior studies (Table 1) (Hojat et al. 2002; Chen
coursework to complete the undergraduate coursework in et al. 2007). Students indicated their career specialty intentions,
three years. The MMEDIC program, established in 1977, is in terms of likelihood of entering each of the specialties listed
available to students of all colleges at Boston University. in Table 1, on a five-point Likert scale (very unlikely ¼
Students enter the program at the beginning of their junior year 1, . . . , neutral ¼ 3, . . . , very likely ¼ 5). Two dummy variables
of college, and then fulfill some preclinical requirements at were created from these selections. Each student was assigned
BUSM during their junior and senior years of college. The a value of 1 for ‘‘people-oriented’’ if they indicated they were
ENGMEDIC program, established in 1990 by the College of ‘‘very likely’’ to choose a career in a people-oriented specialty,
Engineering, is similar to the MMEDIC program; however, to 0 otherwise. Similarly, each student was assigned a value of 1
be eligible students must complete two years of undergraduate for ‘‘technology-oriented’’ if they indicated they were ‘‘very
biomedical engineering curriculum. The EMSSP pathway, likely’’ to choose a career in a technologically focused
established in 1983, is a partnership between BUSM and 10 specialty, 0 otherwise. There was no missing data on any of
306
 Changes in student empathy

Table 1. Career preference categories. Table 2. Descriptive characteristics.

‘‘People-Oriented’’ specialties ‘‘Technology-Oriented’’ specialties Gender Percent N

Internal Medicine Pathology
Family Medicine Surgery and Surgical Subspecialties Men 46.5 540
Pediatrics Radiology Women 53.5 622
Neurology Radiation Oncology 100.0 1162
Rehabilitation Medicine Anesthesiology
Psychiatry Medical school pathways Percent N
Emergency Medicine
EMSSP 7.3 51
Obstetrics and gynecology
4 years 53.9 379
Ophthalmology
7 years 15.6 110
Dermatology
Dual degree 23.2 163
100.0 703
Source: Chen et al. (2007) and Hojat et al. (2002).
Empathy Mean SD

Time 0 112.97 12.53

Time 1 114.58 11.70
Time 2 116.02 12.42
these questions, as students were not able to submit their
Time 3 114.24 13.74
survey with incomplete fields. Time 4 113.29 12.59
Procedures used to collect and analyze these data were Anticipated debt Mean SD
approved by the Boston University Medical Center Institutional Time 0 5.13 2.27
Review Board. Time 1 5.08 2.30
Time 2 5.08 2.32
Time 3 5.22 2.25
Time 4 5.15 2.31
Data analysis Specialty affinity Neither Tech People Both
(%) oriented (%) oriented (%) (%)
To examine changes in students’ levels of empathy over time,
a series of linear mixed models were estimated using package Time 0 29.4 17.3 36.4 16.9
lme4 for the R statistical software package (Bates & Maechler Time 1 28.7 14.9 44.2 12.1
Time 2 31.9 13.3 41.2 13.6
2011; R Development Core Team 2011). The basic model Time 3 9.3 26.2 56.3 8.3
examining changes in empathy over time was as follows:
Notes: 4 years, Traditional Admissions Pathway in the United States; MMEDIC
Yij ¼ 0 þ 1 x1 þ 2 x2 þ 3 x3 þ 4 x4 þ 5 x5 þ i þ "ij , and ENGMEDIC are included with 7-year programs (SMED); Dual Degree,
include MD-PhD, MD-MBA, MD-MPH candidates and students who have
where Yij is the level of empathy of student i in year j; x14 a already obtained Advanced Degrees (PhD, MBA, MPH) prior to entering
series of dummy variables for time (time 0 is the reference medical school.
level); x5 a dummy variable for gender (female ¼ 1); i a
random effect at the student level; and "ij a random noise
independent of the student-level random effect. Following
estimation of the average student trajectory of empathy over
‘‘Neither’’ or ‘‘Both’’ as possible career choices by the end of
time, we examined the role of students’ affinity for technology-
their third year of medical school (Table 2).
vs. people-oriented specialty areas and level of medical
Results from multivariate models capturing changes in
student debt in predicting empathy. We subsequently
students’ levels of empathy over time are presented in Table 3
expanded these models to examine the impact of students’
and Figure 1. This analysis is based on 2652 observations
educational pathway (e.g., traditional four-year medical school
nested within 1162 individuals. As would be anticipated given
program; seven-year program; dual degree programs) and
the trends presented in Table 2, the coefficient estimates
baseline levels of empathy on empathy trajectories over time.
presented in Table 3 indicate that by the end of their second
year of medical school (Time 2), students’ empathy levels had
significantly increased (B ¼ 2.48, p 5 0.05). This increase in the
Results first two years was followed by declines in levels of empathy in
Table 2 presents basic descriptive information about the study years three and four. Although the difference in levels of
sample. The sample was 54% female, with over half the empathy between Time 0 and Time 4 was not statistically
sample in a traditional four-year program. Empathy for patients significant (B ¼ 0.76, p 4 0.05), the decline in empathy
tended to increase over time through the second year of scores from year two to year four was statistically significant
medical school, after which point it tended to decline. at the 0.05 level (B ¼ 3.25, SE ¼ 0.591, t ¼ 5.49; data not
Anticipated debt was fairly consistent over all data collection shown). Gender was also a significant predictor of empathy
periods, with a mean of just over five corresponding to the levels, with women much more likely to express empathy for
$100,000–$150,000 category. Finally, preferences for techno- patients than their male colleagues (B ¼ 6.53, p 5 0.05).
logically oriented vs. people-oriented specialty areas varied Further analysis examined the degree to which decreasing
over time, solidifying as students reached the end of their third levels of empathy might be attributed to cohort differences as
year of medical school. This is seen as fewer students selected opposed to actual within student changes over time. If later
307
D. C. R. Chen et al.

Table 3. Effects of time, medical student debt, and specialty affiliation on students’ empathy.

Model 1a Model 2b

Fixed effects Estimate SE t-value Estimate SE t-value

Intercept 109.99 0.63 174.52* 109.75 0.74 149.28*

Female 6.53 0.63 10.44* 5.86 0.68 8.64*
Time 1 1.24 0.61 2.03* 1.11 0.61 1.82
Time 2 2.48 0.60 4.16* 2.35 0.59 3.94*
Time 3 0.51 0.63 0.80 0.27 0.63 0.43
Time 4 0.76 0.70 1.09 – – –
People – – – 2.15 0.50 4.26*
Tech – – – 1.42 0.57 2.52*
Debt – – – 0.40 0.14 2.91*

Random effects Variance SD Variance SD

Intercept 75.24 8.67 73.56 8.58

Residual 74.01 8.60 72.83 8.53

Notes: aAnalysis based on 2545 observations nested within 1162 participants.

b
Analysis based on 2168 observations nested within 1010 participants.
*Effect statistically significant at 0.05 level.

Differences in empathy trajectories by gender,

medical school pathway, and baseline levels of
empathy
Further analyses sought to examine differences in empathy
trajectories over time among men and women, those in
different medical school pathways, and by levels of empathy
at the start of medical school. First, to assess gender
differences in empathy trajectories, product terms for the
interaction of gender and time were added to the baseline
model presented in Model 1 above. None of the product
terms capturing the gender  time interactions achieved
Figure 1. Students’ empathy toward patients over time.
statistical significance, indicating that the higher levels of
empathy observed among female students were consistent
over time (data not shown). Second, product terms for the
graduating classes were higher in average levels of empathy, interaction of student pathways and time revealed one
the inclusion of their data from Time 0 to Time 2 could be significant effect, with those enrolling in a 7-year program
erroneously interpreted as change over time. To examine this significantly lower in empathy at baseline than those enrolled
possibility, dummy variables for cohort (e.g., class year) were in a traditional 4-year program (data not shown). Note that
added to Model 1 above. An F-test for differences between pathway information was only available for 703 of the 1162
these two models did not achieve statistical significance at the individuals included in the study.
0.05 level (F ¼ 10.97, df ¼ 6, p ¼ 0.089), indicating that cohort Our final set of analyses revealed that empathy trajectories
differences could not account for the differences over time significantly differed based on empathy level at the beginning
observed in Model 1. of the first year of medical school. In this analysis, we divided
Model 2 of Table 3 presents results from an analysis students into tertiles corresponding to high, medium, and low
including measures of students’ affinity for more technology- levels of baseline empathy, which were then used to calculate
oriented vs. people-oriented specialties and anticipated debt product terms capturing their interaction with time. Results
levels as time-varying covariates. (Note that data from Time 4 from this analysis are presented in Figure 2 and reveal that the
was omitted from this analysis due to the fact that specialty curves for these three groups are not parallel, indicating that
affinity was not measured at the end of year four, reducing the where one starts on the empathy scale affects changes in
number of observations to 2168 nested within 1010 individ- empathy over the course of their studies. In contrast to the
uals.) Greater attraction to technology-oriented specialties was subtle decline over time for those high in empathy at baseline,
strongly associated with lower levels of empathy (B ¼ 1.42, the lower two groups increase in empathy through year two,
p 5 0.05), and affinity for people-oriented specialties was only to drop sharply between years two and three. Note that
significantly related to higher levels of empathy. Finally, larger there are no observations at Time 4 here due to the absence of
levels of anticipated medical school debt were associated with any Time 4 data from the cohorts who were assessed at time 0
higher levels of empathy (B ¼ 0.40, p 5 0.05). (i.e., they have not reached the end of their careers yet).
308
 Changes in student empathy

A number of educational interventions have been devel-

oped to address this decline in empathy, including: the
incorporation of clinical narrative or critical incident writing
into the clinical curriculum; classes on medically themed
creative writing, literature and art; journal writing; and using
standardized patient scenarios in the medical education
curriculum (Lichstein & Young 1996; Shapiro & Lie 2000;
Charon 2001; Hatem & Ferrara 2001; Anderson &
Schiedermayer 2003; Charon 2004; DasGupta & Charon
2004). A recent review suggests that empathy may be
amenable to a range of intervention strategies. Qualitative
Figure 2. Empathy trajectories over time, by baseline levels
of empathy (tertiles). data from independent observations and unvalidated surveys
note that these interventions improve student communication
skills and empathy (Stepien & Baernstein 2006). Student
course evaluations and feedback suggest that students respond
Discussion positively to these educational interventions and perceive
themselves to be more sensitive and empathic toward their
Empathy is a critical factor in fostering patient-centered care. patients from such activities. Quantitative results associated
This study confirms and extends prior research on physician with these types of interventions have been mixed, however.
empathy by examining empathy trajectories over the course of One group preliminarily measured an increase in empathy in
the medical student career in a large student population. students who participated in role-playing and simulated
This study confirmed previous research that observed signif- patient scenarios. In contrast, an entire medical school class
icant changes in empathy over time, as empathy levels taking a four-month patient-interviewing course designed to
increased from the beginning of medical school until the end teach communication and emphasize empathy did not show
of the preclinical years, followed by a fairly steep decline an improvement in self-reported empathy (Diseker &
during the third year of medical school (first full clinical year) Michielutte 1981).
that persisted through graduation. Second, although women One noteworthy finding is that students admitted earlier to
expressed significantly higher levels of empathy than men, medical school through the seven-year SMED program have
there were no gender differences in the trajectory of empathy lower measured empathy than their counterparts in the
over time. In other words, women were just as likely to
traditional four-year program during their medical school
express declining levels of empathy during the clinical years as
years. This finding is particularly interesting because few
were men. Third, stratifying students by their baseline levels of
schools have established admission pathways similar to BUSM,
empathy revealed substantial differences in empathy trajecto-
and to date, there has been no research into the differences in
ries over the medical school career. Although empathy peaked
empathy between different medical school admission path-
at the end of the second year for all three groups, students who
ways. Students in the SMED program decide on a career in
began medical school with high levels of empathy were less
medicine in high school, often much earlier than normal
likely to experience steep declines over the course of medical
pathway students. The program’s accelerated coursework and
school. Fourth, empathy trajectories may be related to degree
emphasis on biological sciences may leave students with fewer
pathways, with students in some non-traditional pathways
opportunities for course electives and self-reflection through-
showing more severe decreases in empathy during the clinical
out their undergraduate years, thus impacting their develop-
years than their counterparts in traditional four-year pathways.
ment of empathy. Demographic characteristics of these early
Regarding declining levels of empathy during the third- and
admitted students, such as younger age and relative lack of life
fourth-years of medical school, various educators have
experiences, may also contribute to the initial lower levels of
commented that the current process of medical education
empathy of the SMED students. A better understanding of
and clinical training environment may foster the perceived
students who enter into medical school early or who make
decline in empathy (Kay 1990). The third year is typically
medical career choices earlier is needed as our findings may
when students begin to rotate through various clinical clerk-
impact admission policies.
ships or ward rotations, which can lead to what authors have
Students entering medical school with high levels of
coined the ‘‘traumatic de-idealization’’ of medicine. Students
empathy have been found to better maintain their levels of
lose empathy as they are subjected to challenges of practicing
empathy throughout the medical school careers. We deter-
modern day medicine. Various authors have commented on
the increased volume of patients seen, acuity of illnesses mined that empathy increases in all students in the preclinical
treated, lack of proper student role models during their clinical years, but those having the lowest empathy at the beginning of
clerkships and the increasing dependence on technology to medical school have a steeper decline in empathy than those
make diagnoses which result in less direct patient/bedside initially in the highest tier during the clinical years.
contact (Kay 1990). There is also concern that, because Characteristic differences in these two groups may be initially
students often deal with emotionally challenging and difficult due to a selection effect through the different admission
situations, their empathy declines as a protective defense pathways, but further research may be needed to determine if
mechanism (Chen et al. 2007). the training programs differently impact these students.

309
D. C. R. Chen et al.

Our findings have the potential to impact admission JUN YAN, PhD, is an Associate Professor and Statistician in the Department
of Sociology, University of Connecticut Storrs, CT, USA.
policies, particularly in light of the critical need to recruit
students into primary care. Empathy is the cornerstone of the ELAINE KIRSHENBAUM, MPH, Med, is the Vice President for Policy,
Planning and Member Services at the Massachusetts Medical Society,
physician–patient relationship and as such should be valued in
Waltham, MA, USA.
prospective medical school students and actively cultivated
ROBERT H. ASELTINE Jr, PhD, is a Professor in the Division of Behavioral
throughout their training. Recognizing that there are specialties Sciences and Community Health and Director at the Institute for Public
whose provision of care is based on the relationship with the Health Research at the University of Connecticut, Farmington, CT, USA.
patient, such as the ‘‘people-oriented’’ specialties, suggests the
possibility that if empathy can be enhanced by educational or
experiential interventions, student career preference can also Acknowledgments
be modified. Research is needed to determine whether
increasing empathy increases the likelihood of medical Permission to use the JSPE-S was obtained from the Jefferson
students choosing primary care professions or choosing Medical College Center for Research in Medical Education and
primary care professions fosters the development of higher Health Care. The Massachusetts Medical Society provided
empathy. Following students longitudinally will help deter- statistical funding support. Research was conducted at Boston
mine which relationship holds true. University School of Medicine, Boston, MA, USA.
If further research determines that it is possible to predict
the specialty that an incoming student will choose based on Declaration of interest: The authors report no conflicts of
their initial empathy levels during first year of medical school, interest.
policy decisions can be made affecting resource allocation and
medical curricula which would be driven by societal or
regional needs, such as increasing the number of primary care References
clinicians. Anderson R, Schiedermayer D. 2003. The art of medicine through the
Our study has several limitations. First, our research humanities: An overview of a one-month humanities elective for fourth
focused on self-reported empathy and not behaviors. Various year students. Med Educ 37:560–562.
studies have conflicting results regarding the association Association of American Medical Colleges. 1998. Learning objectives for
medical student education: Guidelines for medical schools. [Published
between self-report and actual observed empathy (Hojat 1998 January]. Available from: https://www.aamc.org/initiatives/msop.
et al. 2005; Chen et al. 2010). We did not survey students in Association of American Medical Colleges. 2006. AAMC statement on the
alternative pathways before they entered the traditional four- physician workforce. [Published 2006 June]. Available from: https://
year curriculum, yet those students had already had exposure www.aamc.org/download/55458/data/workforceposition.pdf
to medical school courses. Although our results regarding Bates D, Maechler M. 2011. lme4: Linear mixed-effects models using S4
classes. [Published 2011 October]. Available from: http://CRAN.
empathy trends over the medical school years and among
R-project.org/package=lme4
different genders were consistent with prior research, our Beckman HB, Markakis KM, Suchman AL, Frankel RM. 1994. The doctor-
results regarding the early pathways may not be fully patient relationship and malpractice: Lessons from plaintiff depositions.
generalizable to other programs that do not have well- Arch Intern Med 154:1365–1370.
established early admissions pathways. Bellini LM, Baime M, Shea JA. 2002. Variation of mood and empathy during
internship. JAMA 287:3143–3146.
Bellini LM, Shea JA. 2005. Mood change and empathy decline persist during
three years of internal medicine training. Acad Med 80:164–167.
Conclusion Bikker AP, Mercer SW, Reilly D. 2005. A pilot prospective study on the
consultation and relational empathy, patient enablement, and health
Given its role in bolstering the physician–patient relationship
changes over 12 months in patients going to the Glasgow
and promoting patient-centered care, empathy should be Homoeopathic Hospital. J Altern Complement Med 11:591–600.
cultivated among students and all practicing physicians. To Boston University. 2011. Boston University Early Medical School Selection
mitigate the declining levels of empathy in the later years of Program. [Published 2011 January]. Available from: http://
medical school, universities should consider incorporating www.bu.edu/academics/busm/programs/early-medical-school-selection-
program
evidence-based interventions into the curriculum and adopting
Charon R. 2001. Narrative medicine: Form, function, and ethics. Ann Intern
admissions policies that favor prospective students who Med 134:83–87.
identify and empathize with patients. Such policies may Charon R. 2004. Narrative and medicine. N Engl J Med 350:862–864.
produce medical school graduates who are more likely to Chen D, Lew R, Hershman W, Orlander J. 2007. A cross-sectional
gravitate toward ‘‘people-oriented’’ specialties and specifically measurement of medical student empathy. J Gen Intern Med
22:1434–1438.
toward primary care, potentially addressing the current short-
Chen DC, Pahilan ME, Orlander JD. 2010. Comparing a self-administered
ages in primary care physicians in the US. measure of empathy with observed behavior among medical students. J
Gen Intern Med 25:200–202.
Coulehan JL, Platt FW, Egener B, Frankel R, Lin CT, Lown B, Salazar WH.
Notes on contributors 2001. ‘‘Let me see if I have this right. . .’’: Words that help build
empathy. Ann Intern Med 135:221–227.
DANIEL C.R. CHEN, MD, MSc, is an Assistant Professor of Medicine, DasGupta S, Charon R. 2004. Personal illness narratives: Using reflective
Associate Clerkship Director for Internal Medicine, and Assistant Dean for writing to teach empathy. Acad Med. 79:351–356.
Student Affairs at Boston University School of Medicine, Boston, MA, USA. Diseker RA, Michielutte R. 1981. An analysis of empathy in medical
DANIEL S. KIRSHENBAUM, MD, is an Internal Medicine Resident at Boston students before and following clinical experience. J Med Educ
Medical Center, Boston, MA, USA. 56:1004–1010.
310
 Changes in student empathy

Gordon EE. 2010. Medical staffing: Critical shortages on the horizon. Surg Lichstein PR, Young G. 1996. ‘‘My most meaningful patient’’: Reflective
Neurol Int 1:7. learning on a general medicine service. J Gen Intern Med 11:406–409.
Hatem D, Ferrara E. 2001. Becoming a doctor: Fostering humane caregivers MacPherson H, Mercer SW, Scullion T, Thomas KJ. 2003. Empathy,
through creative writing. Patient Educ Couns 45:13–22. enablement, and outcome: An exploratory study on acupuncture
Hojat M, Gonnella JS, Nasca TJ, Mangione S, Veloksi JJ, Magee M. 2002. patients’ perceptions. J Altern Complement Med 9:869–876.
The Jefferson Scale of Physician Empathy: Further psychometric data R Development Core Team. 2011. R: A language and environment for
and differences by gender and specialty at item level. Acad Med statistical computing. [Published 2011 September]. Available from:
77:S58–S60. http://www.R-project.org
Hojat M, Mangione S, Nasca TJ, Cohen MJM, Gonnella JS, Erdmann JB, Shapiro J, Lie D. 2000. Using literature to help physician-learners
Veloski J. 2001. The Jefferson Scale of Physician Empathy: understand and manage ‘‘difficult’’ patients. Acad Med 75:765–768.
Development and preliminary psychometric data. Educ Psychol Meas Sherman JJ, Cramer A. 2005. Measurement of changes in empathy during
61:349–365. dental school. J Dent Educ 69:338–345.
Hojat M, Mangione S, Nasca TJ, Gonnella JS, Magee M. 2005. Empathy Stepien KA, Baernstein A. 2006. Educating for empathy: A review. J Gen
scores in medical school and ratings of empathic behavior in residency Intern Med 21:524–530.
training 3 years later. J Soc Psychol 145:663–672. Suchman AL, Roter D, Green M, Lipkin M. 1993. Physician satisfaction with
Hojat M, Mangione S, Nasca TJ, Rattner S, Erdmann JB, Gonnella JS, Magee primary care office visits. Collaborative study group of the American
M. 2004. An empirical study of decline in empathy in medical school. Academy on Physician and Patient. Med Care 31:1083–1092.
Med Educ 38:934–941. Vermeire E, Hearnshaw H, van Royen P, Denekens J. 2001. Patient
Kay J. 1990. Traumatic deidealization and the future of medicine. JAMA adherence to treatment: Three decades of research. A comprehensive
263:572–573. review. J Clin Pharm Ther 26:331–342.

Appendix

Table A1. Total number of responses and response rates by study wave and graduating cohort.

Survey totals by wave and cohort

Wave Number of
0 1 2 3 4 eligible students
2008 126 120 152
2009 128 149 111 154
2010 104 154 149 122 155
2011 109 90 144 161 8 168
2012 152 170 164 2 196
2013 178 131 56 17 196
439 495 646 604 361

Response rates by wave and cohort

2008 0.00 0.00 0.00 0.83 0.79
2009 0.00 0.00 0.83 0.97 0.72
2010 0.00 0.67 0.99 0.96 0.79
2011 0.65 0.54 0.86 0.96 0.05
2012 0.78 0.87 0.84 0.01 0.00
2013 0.91 0.67 0.29 0.09 0.00

Notes: Gray shaded cells contain students who were retained from prior years or had an alternative (split second-year) curriculum and were therefore not included in
response rates.
Average response rate ¼ 0.81; total number of observations ¼ 2545; and total number of students ¼ 1162.

311
Copyright of Medical Teacher is the property of Taylor & Francis Ltd and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.