Cultural competency in medical education: demographic differences associated with medical student communication styles and clinical clerkship feedback.

Abstract

We tested the significance of associations among students' demographics, communication styles, and feedback received during clerkships.
US medical students who completed at least one required clinical clerkship were invited between April and July 2006 to complete an anonymous, online survey inquiring about demographics, communication styles (assertiveness and reticence), feedback (positive and negative), and clerkship grades. The effects of self-identified race/ethnicity, gender, and generation (immigrant, first- or second-generation American) and their 2-way interactions on assertiveness, reticence, total positive and total negative feedback comments were tested using factorial analysis of covariance, controlling for age, clerkship grades, and mother's and father's education; pairwise comparisons used simple contrasts. Two-sided P values < .05 were considered significant.
Medical students from 105 schools responded (N = 2395: 55% women; 57% white). Men reported more assertiveness than women (P = .001). Reticence (P < .001) and total positive comments (P = .006) differed by race/ethnicity; in pairwise contrasts, black, East Asian, and Native American/ Alaskan students reported greater reticence than white students (P < .001), and white students reported receiving more positive comments than black, and South and East Asian students. Race/ethnicity-by-generation (P = .022) and gender-by-generation (P = .025) interaction effects were observed for total negative comments; white first-generation Americans reported receiving the fewest and male immigrants reported receiving the most negative comments.
Demographic differences in students' communication styles and feedback they received highlight a need for cultural competency training to improve medical student-teacher interactions, analogous to training currently advocated to improve physician-patient interactions.

Full text

Page 1

116 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
VOL. 101, NO. 2, FEBRUARY 2009
original

communication
INTRODUCTION
M
this goal is the need for medical students to acquire the
skills to effectively communicate with patients across
culture, gender, and socioeconomic status.1 In addition
to being important to physician-patient interactions with
regard to patient satisfaction and patient participation
during the physician encounter,2,3 communication skills
also are associated with clinical competency,4 and the
ability to elicit, analyze, and communicate relevant clin-
ical information to patients.5
Communication skills also are important during the
third and fourth years of medical school when medical
students enter the clinical clerkships. During these years,
medical students’ clinical performance is largely evalu-
ated by physician’s subjective ratings of medical stu-
dents’ clinical knowledge, patient interaction, and com-
munication and interpersonal skills. An earlier study
reported that certain communication styles—greater
assertiveness and less reticence—were independently
associated with receiving higher grades in clinical clerk-
ships.6 Briefly, assertiveness reflects students’ comfort
with on-the-spot questioning and confidence in their pub-
lic speaking abilities; reticence reflects discomfort with
speaking out, responding to questions, being modest or
reserved. Questionnaire items used to measure assertive-
ness and reticence are shown in Table 1. Other studies
also have indicated a relationship between demographic
characteristics and communication skills among medical
students. A recent study by Fernandez and colleagues
found that Asian and black US medical students scored
edical schools strive to train physicians who
are capable of treating patients from diverse
social and cultural backgrounds. Crucial to
Author Affiliations: Washington University School of Medicine in St. Louis, St
Louis, Missouri (Drs Lee, Andriole, and Jeffe; Mr Vaishnavi); University of Cali-
fornia, San Diego School of Medicine, San Diego, California (Mr Lau).
Corresponding Author: Katherine B. Lee, MD, MA, Washington University
School of Medicine, c/o Dr Donna B. Jeffe, 4444 Forest Park, Ste 6700, St
Louis, MO 63108 (kathyleemd@gmail.com).
Purpose: We tested the significance of associations among
students’ demographics, communication styles, and feed-
back received during clerkships.
Methods: US medical students who completed at least one
required clinical clerkship were invited between April and
July 2006 to complete an anonymous, online survey inquir-
ing about demographics, communication styles (assertive-
ness and reticence), feedback (positive and negative), and
clerkship grades. The effects of self-identified race/ethnicity,
gender, and generation (immigrant, first- or second-genera-
tion American) and their 2-way interactions on assertiveness,
reticence, total positive and total negative feedback com-
ments were tested using factorial analysis of covariance,
controlling for age, clerkship grades, and mother’s and
father’s education; pairwise comparisons used simple con-
trasts. Two-sided P values < .05 were considered significant.
Results: Medical students from 105 schools responded (N =
2395; 55% women; 57% white). Men reported more assertive-
ness than women (P = .001). Reticence (P < .001) and total
positive comments (P = .006) differed by race/ethnicity; in
pairwise contrasts, black, East Asian, and Native American/
Alaskan students reported greater reticence than white stu-
dents (P < .001), and white students reported receiving more
positive comments than black, and South and East Asian stu-
dents. Race/ethnicity-by-generation (P = .022) and gender-
by-generation (P = .025) interaction effects were observed
for total negative comments; white first-generation Ameri-
cans reported receiving the fewest and male immigrants
reported receiving the most negative comments.
Conclusions: Demographic differences in students’ com-
munication styles and feedback they received highlight a
need for cultural competency training to improve medical
student-teacher interactions, analogous to training currently
advocated to improve physician-patient interactions.
Keywords: cultural competency n education
J Natl Med Assoc. 2009;101:116-126
Cultural Competency in Medical Education:
Demographic Differences Associated With
Medical Student Communication Styles and
Clinical Clerkship Feedback
Katherine B. Lee, MD, MA; Sanjeev N. Vaishnavi, BS; Steven K.M. Lau, BA; Dorothy A. Andriole, MD, FACS;
Donna B. Jeffe, PhD

Page 2

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATIONVOL. 101, NO. 2, FEBRUARY 2009 117
CULTURAL COMPETENCY IN MEDICAL EDUCATION
significantly lower than white students on the communi-
cation portion of clinical performance examinations.7
In addition, racial/ethnic disparities in clinical clerk-
ship grades have been reported; underrepresented minor-
ity (URM; black, Hispanic, Native American and Native
Alaskan) and Asian students reported receiving lower
clinical clerkship grades compared with white students.6
Because race/ethnicity, gender, and communication
styles were found to be associated with clinical clerk-
ship grades,6 in the present study, we sought to identify
associations among students’ demographics, their com-
munication styles, and the feedback students receive
regarding their communication skills during clinical
clerkships. An understanding of the extent to which
race- or ethnicity-based differences in students’ commu-
nication styles may contribute to disparities in clinical-
clerkship grades could inform strategies to help medical
students acquire the communication skills that are val-
ued in clinical medicine. A better understanding of race-
or ethnicity-based differences in students’ communica-
tion styles also could inform the efforts to educate an
increasingly diverse medical school population, which
will be critical for the creation of a culturally competent
physician workforce, well prepared to meet the health-
care needs of an increasingly demographically diverse
American society.8-10
METHODS
Sample, Survey Design, and
Administration
Between April 6 and July 1, 2006, medical students
were contacted through e-mail using the membership
database of a national medical student association and
from the medical school deans of all 125 US allopathic
medical schools, who were asked to forward an e-mail
survey invitation and link to their third- and fourth-year
students. US medical students who had completed any
of the following clerkships—internal medicine, neurol-
ogy, psychiatry, obstetrics/gynecology, surgery, and
pediatrics—were eligible to participate in this online
anonymous survey. Completion of the questionnaire
implied consent. The protocol received institutional
review board approval at Washington University School
of Medicine in St Louis, Missouri.
Measures
The questionnaire included items that measured com-
munication-related variables (assertiveness, reticence,
and positive and negative comments received during
clerkships) as well as items about students’ and their par-
ents’ demographics and students’ clerkship grades.
Communication-Related Variables
Assertiveness and reticence. Development of the
assertiveness and reticence measures have been described
previously.6 Students indicated their degree of agree-
ment using a 5-point scale ranging from 1, strongly dis-
agree to 5, strongly agree with 12 statements (Table 1),
measuring assertiveness regarding on-the-spot question-
ing, confidence in public speaking, networking, taking
risks, making sure important people know who they are,
and “working the room.” Students also indicated their
degree of agreement using the same 5-point scale with 6
statements (Table 1) measuring reticence regarding dis-
comfort with speaking out, fear of appearing foolish,
Table 1. Assertiveness and Reticence Items
Assertiveness
• If I don’t know an answer, I still act as though I do.
• I feel comfortable with direct questioning that puts me on the spot.
• I feel comfortable with my current communication skills.
• I see a competitive side to medicine and it drives me to do well.
• I feel confident in my potential to be a good doctor.
• I have made sure that important people know who I am.
• I am a logical, goal-directed speaker when I need to be.
• I know how to “work a room” or use a meeting to my advantage.
• Powerful is a word I would use to describe myself.
• I enjoy taking risks, especially those that might advance my career.
• When I know an answer to a question, I am eager to respond.
• I am confident when interacting with attendings and residents.
Reticence
• I do not readily respond to questions from residents and attendings because I am uncomfortable
speaking out.
• I am uncomfortable responding to questions because I’m not always sure I’m right.
• I do not always ask questions, because I’m afraid people will think my questions are foolish.
• I speak and participate as much as other students but am still labeled as being quiet/reserved.
• I am modest or reserved when interacting with attendings and residents.
• Out of respect for my superiors, I prefer not to challenge or openly question their decisions.

Page 3

118 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
VOL. 101, NO. 2, FEBRUARY 2009
CULTURAL COMPETENCY IN MEDICAL EDUCATION
being modest or reserved, being labeled as quiet, and
being respectful of superiors. Higher mean scores indi-
cated greater assertiveness (Cronbach a = 0.82) and
greater reticence (Cronbach a = 0.74).
Positive and negative feedback comments. A list of
feedback comments was developed from informal dis-
cussions with the first author with a convenience sample
of 40 medical students from 20 medical institutions
about their clerkship experiences. The questions were
reviewed with 5 medical school deans for content and
were then piloted on a convenience sample of 10 resi-
dent physicians from different medical institutions to
refine item wording and clarity. Based on the residents’
input, we then eliminated some items and revised the
wording of other items prior to surveying the medical
students for our study. Students completing the survey
were asked to indicate if they ever received (yes or no)
specific positive and negative feedback regarding their
communication skills during clinical clerkships (Table
2). The total number of yes responses to 4 “positive”
comments was the positive feedback score, and the total
number of yes responses to 7 negative comments was
the negative feedback score. Medical students who par-
ticipated in the original discussions did not see any of
the questions included in the survey prior to the survey
being sent out.
Demographics
Gender, race/ethnicity, and generation. Students
indicated their gender as male or female. Students indi-
cated their race and ethnicity by selecting all that applied
from the following race/ethnicity category selections:
white, not of Hispanic origin; black, not of Hispanic ori-
gin; Chinese; Taiwanese; Korean; Japanese; Vietnam-
ese; other southeast Asian; Asian Indian; Pakistani;
Native Hawaiian; Filipino; other Pacific Islander; other
Asian; American Indian or Alaska native; Mexican
American/Chicano; Puerto Rican; other Hispanic,
unknown; and other, please specify. The white racial/
ethnic group included all students who identified them-
selves as white, European, or Canadian. East Asian race/
ethnicity included all students who identified themselves
as Chinese, Taiwanese, Korean, Japanese, Filipino, Viet-
namese, Thai, Burmese, or Malaysian. Students who
identified themselves as Asian Indian, Pakistani, or
Afghani were categorized as South Asian. Black stu-
dents included respondents who indicated they were
black, non-Hispanic. Hispanic students were all partici-
pants who indicated they were of Hispanic origin. Stu-
dents reporting they were Native American or Native
Alaskan were grouped as Native American/Native Alas-
kan. Students who identified themselves as Middle East-
ern; multiracial (eg, self-identifying as both white and
Table 2. Associations between Communication Styles and Each Positive and Negative Feedback
Comment Receiveda
Positive Comments
Student Takes
InitiativeItem
Received comment
Yes
No
Assertiveness, OR (95% CI) 3.625 (2.786-4.716) 3.310 (2.417-4.534) 2.753 (1.623-4.671) 2.962 (1.279-6.859)
Reticence, OR (95% CI) 0.589 (0.487-0.711) 0.589 (0.487-0.711) 1.036 (0.683-1.572) 1.549 (0.796-3.012)
Student Is Very
Assertive
Student Is Very
Hardworking
Student Interacts
With Patients Well
553
1232
1515
281
1724
70
1774
25
Negative Comments
Item
Received
comment
Yes
No
Assertiveness
OR (95% CI)
Student
Is Too
Quiet
Student Is
Very
Timid/
Reserved
Student Is
Not
Independent
Minded
Student Lacks
Adequate
Communication
Skills
Student
Does Not
Ask Enough
Questions
Student
Lacks
Leadership
Skills
Student’s
Interaction With
Patients Needs
Improvement
464
1321
0.658
(0.507-
0.853)
3.970
(3.170-
4.971)
381
1405
0.684
(0.519-
0.901)
4.456
(3.490-
5.690)
61
1723
0.536
(0.311-0.924)
48
1730
0.389
(0.199-0.759)
196
1588
0.610
(0.435-
0.855)
2.439
(1.850-
3.216)
22
1756
0.325
(0.130-
0.809)
1.293
(0.618-
2.705)
57
1717
0.815
(0.451-1.473)
Reticence
OR (95% CI)
1.336
(0.867-2.059)
1.521
(0.886-2.610)
1.132
(0.713-1.798)
Abbreviations: OR, adjusted odds ratio; CI, confidence interval.
a We included both assertiveness and reticence in each model, comparing receipt of the comment with not having received this
comment as reference. Students who did not respond to a particular feedback comment were excluded.

Page 4

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATIONVOL. 101, NO. 2, FEBRUARY 2009 119
CULTURAL COMPETENCY IN MEDICAL EDUCATION
Asian or as both East and South Asian); unknown; pre-
ferred not to answer; or other, unspecified were excluded
from analysis.
In addition, it has been reported that college students’
degree of acculturation may affect their academic per-
formance and communication skills.11 For this reason,
students also were asked to indicate their country of
birth and the country of birth of their parents and grand-
parents. Responses to these questions were used to cate-
gorize students into 3 groups: immigrants (students born
outside of the United States), first-generation Americans
(students born in the United States but whose parents
were both immigrants), and second-generation Ameri-
cans (one or both parents born in the United States).
Parents’ education and students’ age. Because
medical students are likely to have parents who are
highly educated and who might serve as role models for
their children in medical school,12-14 we asked students to
indicate the highest level of education completed by
their mother and father using the following response
options: less than a high school education, high school
graduate or some postsecondary education, college
graduate or some postgraduate education, or postgradu-
ate education. Respondents also were asked to indicate
their current age.
Table 3. Sample Demographics

Race/Ethnicity
White
East Asian
South Asian
Black
Hispanic
Native American/Alaskan
Otherb
No responseb
Gender
Male
Female
No responseb
Generation
Immigrant
First generation
Second generation
No responseb
Mother’s highest education
Less than high school
High school graduate or some post secondary
College grad or some post graduate
Postgraduate degrees
No response or unknownb
Father’s highest education
Less than high school
High school graduate or some post secondary
College grad or some post graduate
Postgraduate degrees
No response or unknownb
Grading systems used in clinical clerkships
Honors/high pass/pass/fail
Honors/pass/fail
A/B/C/D/F
Pass/failb
Otherb
No responseb
N = 2369a
Percent
1357
353
171
115
116
57.3
14.9
7.2
4.9
4.9
1.4
4.0
5.5
32
95
130
937
1314
118
39.6
55.5
5.0
298
406
1361
304
12.6
17.1
57.5
12.8
622.6
24.8
29.5
29.9
13.2
588
698
709
312
59 2.5
16.1
20.3
47.4
13.8
382
480
1122
326
1301
217
336
54.9
9.2
14.2
2.0
3.7
16.1
47
87
381
a Excludes 22 osteopathic and four foreign students. N = number of respondents.
b Excluded from further analysis.

Page 5

120 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
VOL. 101, NO. 2, FEBRUARY 2009
CULTURAL COMPETENCY IN MEDICAL EDUCATION
Percentage of Clerkships With
Highest Grade
Students were asked to indicate the system used to
grade clinical clerkships at their schools and their grade
for each of the clerkships they had completed. For stu-
dents who reported either honors/high pass/pass/fail,
honors/pass/fail, or A/B/C/D/F grading systems, grades
were recoded into a dichotomous variable with 1, high-
est grades (honors or A) and 0, lower grades (high pass/
pass/fail or B/C/D/F). We then computed the percentage
of clerkships they completed in which they received the
highest grade (ie, honors or A). This variable was not
computed for students whose schools used a grading
system that had no “highest” passing grade with which
to compare other passing grades (eg, pass/fail) or did not
grade clerkships, or if students did not report receiving
any grades.
Analysis
Pearson product-moment correlations were used to
measure the associations between the percentage of
clerkships with highest grade and each of the positive
and negative comments received as well as the relation-
ships among continuous variables (age, percentage of
clerkships with highest grade, assertiveness, reticence,
total number of positive and negative comments
received). The c2 test was used to measure associations
among categorical variables (gender, race, mother’s and
father’s level of education). Logistic regression models
were used to determine associations between each of
assertiveness and reticence and receipt of each positive
and negative feedback comment. We report adjusted
odds ratios (ORs) and 95% confidence intervals (CI)s.
Differences in each outcome (assertiveness, reticence,
total positive and total negative feedback comments) by
race, gender, and generation were tested using factorial
analyses of covariance (ANCOVAs), after controlling for
age, percentage of clerkships with highest grade, and
mother’s and father’s level of education. Pairwise com-
parisons were made using simple contrasts. Because of
small or empty cells in 3-way interactions, only 2-way
(gender-by-race/ethnicity, gender-by-generation, and
race/ethnicity-by-generation) interactions were tested in
the ANCOVAs. Two-sided p values are reported.
RESULTS
Respondent Characteristics
We received 2395 responses from students in 105
medical schools (94 of 125 US allopathic, 7 U.S. osteo-
pathic, and 4 foreign medical schools). We excluded 22
students from osteopathic and 4 students from foreign
medical schools due to small numbers, leaving 2369 US
allopathic medical students in the sample. Characteris-
tics of these 2369 respondents are shown in Table 3. This
sample did not differ significantly from the overall pop-
ulation of third- and fourth-year US allopathic medical
students in terms of age at matriculation (mean age at
matriculation = 24 in our sample and in US allopathic
students overall), gender, and race/ethnicity.15-17 The
mean percentage of clerkships with highest grade
reported by the students in our sample was 37.1% (SD
34.4%; range, 0-100%). Students with missing data and
students who reported “other” race or either pass/fail or
“other” clerkship grading systems were excluded from
further analysis.
Table 4. Pearson Product-Moment Correlations

Assertiveness
% of
Clerkships
With Highest
Grade
0.216C
(n = 1505)
-0.206C
(n = 1492)
Assertiveness Reticence
1.000
Negative
Comments
Positive
Comments Age
Mother’s
Highest
Education
Reticence-0.574C
(n = 1492)
-0.352C
(n = 1498)
0.414C
(n = 1505)
0.017
(n = 1505)
0.062a
(n = 1505)
0.026
(n = 1505)
1.000
Negative comments -0.186C
(n = 1593)
0.445C
(n = 1485)
-0.345C
(n = 1492)
-0.050
(n = 1492)
-0.085b
(n = 1492)
-0.038
(n = 1492)
1.000
Positive comments 0.134C
(n = 1601)
-0.049
(n = 1615)
0.110C
(n = 1615)
0.071b
(n = 1615)
-0.295C
(n = 1593)
-0.016
(n = 1593)
-0.071b
(n = 1593)
-0.046
(n = 1593)
1.000
Age 0.044
(n = 1601)
0.014
(n = 1601)
-0.019
(n = 1601)
1.000
Mother’s highest
education
Father’s highest
education
-0.094C
(n = 1615)
-0.068b
(n = 1615)
1.000
0.510C
(n = 1615)
a P < .05.
b P < .01.
c P < .001.