Culture, Language, and The Doctor-Patient Relationship: Escholarship@Umms
Culture, Language, and The Doctor-Patient Relationship: Escholarship@Umms
Culture, Language, and The Doctor-Patient Relationship: Escholarship@Umms
eScholarship@UMMS
Family Medicine and Community Health
Family Medicine and Community Health
Publications and Presentations
5-2002
Lucy M. Candib
University of Massachusetts Medical School
Repository Citation
Ferguson, Warren J. and Candib, Lucy M., "Culture, language, and the doctor-patient relationship" (2002). Family Medicine and
Community Health Publications and Presentations. 61.
https://escholarship.umassmed.edu/fmch_articles/61
This material is brought to you by eScholarship@UMMS. It has been accepted for inclusion in Family Medicine and Community Health Publications
and Presentations by an authorized administrator of eScholarship@UMMS. For more information, please contact [email protected].
Vol. 34, No. 5 353
Background: This review’s goal was to determine how differences between physicians and patients in
race, ethnicity, and language influence the quality of the physician-patient relationship. Methods: We
performed a literature review to assess existing evidence for ethnic and racial disparities in the quality of
doctor-patient communication and the doctor-patient relationship. Results: We found consistent evidence
that race, ethnicity, and language have substantial influence on the quality of the doctor-patient relation-
ship. Minority patients, especially those not proficient in English, are less likely to engender empathic
response from physicians, establish rapport with physicians, receive sufficient information, and be en-
couraged to participate in medical decision making. Conclusions: The literature calls for a more diverse
physician workforce since minority patients are more likely to choose minority physicians, to be more
satisfied by language-concordant relationships, and to feel more connected and involved in decision mak-
ing with racially concordant physicians. The literature upholds the recommendation for professional in-
terpreters to bridge the gaps in access experienced by non-English speaking physicians. Further evidence
supports the admonition that “majority” physicians need to be more effective in developing relationships
and in their communication with ethnic and racial minority patients.
Over the course of the last decade, disparities in health to socioeconomic status, the literature on the doctor-
outcomes among ethnic minority and racial groups have patient relationship has not addressed the influence of
become increasingly clear.1,2 Differences include access cultural difference between physicians and patients on
to care, screening, diagnostic and treatment interven- communication effectiveness. Additionally, appropri-
tions, and morbidity and mortality.Accordingly, the US ate care of ethnically and racially diverse populations
government set goals for narrowing these differences, requires the ability to communicate with individuals
first for the year 2000 and now for the year 2010. 3 While who have limited English proficiency. Only 25% of the
the reasons for these disparities remain poorly under- important investigations on doctor-patient relationships
stood,1 calls for cultural competency training in medi- have considered non-English-speaking patients.11
cal school and residency,4-6 as well as efforts to recruit We examined the literature on doctor-patient com-
a more diverse medical workforce,7,8 suggest that as- munication and culture, looking for recommended strat-
pects of the doctor-patient relationship may be impor- egies for improving the doctor-patient relationship. In
tant causative factors. this literature, several themes stand out: approaches to
Research on doctor-patient communication has gen- language barriers; the recognition of physician bias, in-
erated considerable evidence that effective communi- cluding racism; and relationship building. The latter
cation can improve outcome measures such as patient includes the ability to use empathy and foster trust. It
satisfaction, adherence to treatment, and disease out- also includes effective communication skills to facili-
comes.9,10 Provision of adequate information, elicita- tate participatory decision making with patients and the
tion of patient worries, and a participatory decision- provision of culturally competent care.12
making style have all correlated with improved effec- The goal of this review is answer three questions:
tiveness. However, apart from an occasional reference (1) Is there evidence that differences in language,
ethnicity, and race between physicians and patients af-
fect the quality of their relationship and communica-
From the Department of Family Medicine and Community Health, Uni- tion, and if so, are there outcome measures to substan-
versity of Massachusetts, and the Family Health Center of Worcester. tiate such an effect? (2) Is there evidence that improv-
354 May 2002 Family Medicine
ing such communication makes a difference in health Five studies were identified that showed a correla-
care outcomes? (3) Is there evidence that acting on rec- tion between LES ability of patients and perceived qual-
ommendations to diversify the physician workforce and ity of outcomes in comparison to English-speaking
train the existing workforce to be culturally and lin- patients. The results are summarized in Table 1. Four
guistically effective will make a difference in outcomes of these studies measured quality by surveying patients.
of care? For example, one survey of Hispanics in Arizona re-
garding health status, access barriers, and care satis-
Methods faction showed that language of interview was a more
Using MEDLINE, we performed a literature review significant variable than ethnicity.15 Three other stud-
using the key words “culture,” “racism,” “minority popula- ies surveyed patients following medical encounters,
tions,” “ethnic groups,” “language,” “interpreters,” “phy- reporting on differences in satisfaction, provision of
sician-patient relationship,” “physician-patient communi- information, and compliance.16-18
cation,” “patient satisfaction,” “compliance,” “negotiation,” The study by Baker17 was unique in that he surveyed
and “empathy.” Key words were “exploded” to cover a 467 patients from one of three groups: those interviewed
large number of MeSH headings. Searches were lim- in English, those interviewed with an ad hoc interpreter,
ited to articles published in English from 1966 to 2000. and a third group interviewed with no interpreter de-
Additionally, we searched a database developed at the spite a patient’s report that one was needed. Those who
University of Massachusetts that includes published ar- used ad hoc interpreters or who went without a needed
ticles on access and health outcome barriers. interpreter indicated that providers were less friendly,
We included articles reporting investigator-initiated less respectful, and less concerned. For those needing
research and secondary data analyses with quantitative an interpreter but not using one, these findings were
methods that controlled for covariates such as age, edu- magnified. These patients were also less satisfied with
cation, socioeconomic status, and measures of wellness. time spent by provider and with interpersonal aspects
Articles and recommendations based on expert opin- of care. Another emergency room study with pediatric
ion or anecdotal experience were excluded, as were case patients demonstrated that children with LES parents
studies. No articles using rigorous qualitative research had longer, more-costly visits with more testing due to
methodology were found. the inability to communicate with parents.19
Table 2 summarizes those studies examining lan-
Results guage concordance between a physician and patient and
More than 400 articles were initially identified, but methods to bridge language difference. Three studies
only 21 met the inclusion criteria. Most articles reported of Spanish-speaking Latino patients observed a corre-
on outcome problems with limited-English speaking lation between doctor-patient language concordance
patients (five studies) or on strategies to overcome this and quantifiable outcomes. For example, Latino patients
language barrier by using bilingual physicians or pro- with a chronic condition (asthma) and cared for by a
fessional interpreters (seven studies). All of these are language-concordant physician asked more questions,
categorized below as language studies and summarized had greater recall of recommendations, had lower use
in Tables 1 and 2. A second group of articles reported of the emergency room, and had more compliance with
on the evidence for potential physician bias in racially follow-up care.20 In another study of a homogeneous
and ethnically discordant physician-patient relation- population, poor, LES Spanish-speaking Hispanics with
ships (four studies). These are presented in Table 3 and a language-concordant physician had more information
categorized below as bias studies. The last group of recall and asked more questions of their physicians than
articles (Table 4) examined relationship-building be- those cared for by an English-speaking physician.21 A
haviors of physicians, focusing on differences in rap- stratified random analysis of Latino and Caucasian pa-
port building, demonstration of empathy, and quality tients with diabetes and hypertension from the Medi-
of communication skills beyond language of interview cal Outcomes Study found a correlation between physi-
(five studies). cal function and better well-being when the primary
care physician spoke the same language.22
Language Four studies have focused on outcome measures with
In 1990, more than 14 million Americans were not interpretation methods. In the first study (not included
proficient in English.13 Existing strategies for improv- in Table 2), physician and patient satisfaction with in-
ing access for limited-English-speaking (LES) patients terpretation methods were surveyed using a Likert scale
include care by bilingual/bicultural providers, use of developed by the authors. Validity and reliability test-
bilingual/bicultural employees as community health ing of the instrument were not reported. Both physi-
workers and culture brokers, use of bilingual employ- cians and patients were most satisfied with professional
ees who interpret in addition to their regular work, the interpreters. Patients, but not physicians, were satis-
use of professional interpreters, and the use of transla- fied with use of a family member or with use of a bilin-
tion via written or other technologies.14
Modern Culture and Physician-Patient Communication Vol. 34, No. 5 355
Table 1
David and 139 Cases=57 ES (controls) Outpatient Cross-sectional No control for MD explanation LES < ES
Rhee,16 1998 Controls= versus LES internal medi- patient survey measures of educa- Satisfaction LES < ES
47 (cases) cine clinic tion or socioeco- MD perception LES < ES
Hispanics nomic status Time with MD NS
Nonprofessional No report on survey Mammogram NS
Interpreters used validity
for LES
Baker,17 1998 467 36 Hispanics ED Cross-sectional Age, gender, literacy, MD friendliness LES < ES
ES versus LES ± patient survey health status, antici- MD respect LES < ES
nonprofessional pated satisfaction MD show of LES < ES
interpreter concern
Sufficient time NS
Carrasquillo 2,333 ES=47 ES (controls) Five EDs Cross-sectional Age, gender, race/ Satisfaction LES < ES
et al,18 1999 LES=41 versus LES patient survey ethnicity, education, Discharge LES < ES
(cases) income, insurance instruction
Latino, African status, chief com- Overall care LES < ES
American, Asian, plaint, urgency, Courtesy LES < ES
and Eastern having primary MD Respect LES < ES
European
Hampers 2,467 LB=30 209 children with ED Prospective cohort Race, ethnicity, Length of time LB > nLB
et al,19 1999 months LB for family/ study insurance, MD level, in ED
No LB=36 MD triage category Total ED test LB > nLB
months 2,258 children charges
with no LB for
family/MD
ED—emergency department, ES—English speaking, LB—language barrier, LES—limited English speaking, MD—physician, nLB—no language barrier,
NR—not reported, NS—nonsignificant
* The following results were all significant with P<.05 or lower. LES < ES=scores or measures lower with limited English-speaking patients compared to
English-speaking patients; LB > nLB=amount larger for families with language barrier than families with no language barrier.
gual physician colleague. Physicians, but not patients, In a third study, use of health services and preven-
were satisfied with interpretation by telephone.23 tive screening exams was studied in four health main-
The second study was a randomized study of 49 post- tenance organization practice sites prior to and follow-
partum visits that compared two types of professional ing the addition of professional interpreters to on-site
interpretation: proximate-consecutive (typically per- staff. Retrospective record review, after the interven-
formed in the triadic interview with a patient, physi- tion, revealed that patients with limited English profi-
cian, and interpreter) and remote-simultaneous (the ciency had more office visits and increased use of pre-
form of simultaneous translation used in the United scription drugs, as well as increased numbers of rectal
Nations with special technology). The remote type of exams, flu vaccines, and fecal occult blood testing.
interpretation was judged to be superior in many ways. However, there was no statistical change in use of mam-
Patient and physician utterances were both increased mography, PAP testing, or physical breast exam.25
using the remote method. There were 12% fewer inac- One other study demonstrated that even professional
curacies of words spoken by physicians and 13% fewer interpretation might have its limitations. A cross-
inaccuracies of words spoken by patients. Both patients sectional sample of patients was videotaped during vis-
and physicians preferred the remote method, although its with physicians in a multi-ethnic university clinic.
interpreters preferred the proximate method.24 English-speaking patients made almost three times as
356 May 2002 Family Medicine
Table 2
Perez-Stable 236 61 37 ES Latinos University Patient survey Age, gender Physical LC > LD
et al,22 1997 73 LES Latinos IM: A/R/NP Education function
126 Caucasians Illness burden Psychological LC > LD
63% female Ethnicity well-being
± bilingual MD Health LC > LD
perceptions
Pain LC < LD
Interpreter Strategies
Hornberger 49 NR 49 SS mothers Well baby Randomized inter- Homogenous sample MD utterances RS > PC
et al,24 1996 with infants clinic vention using two Mother RS > PC
interpretation utterances
methods Interpreter RS > PC
errors
MD satisfaction RS > PC
Patient RS > PC
satisfaction
Jacobs,25 4,380 ISG=46 327 LES Spanish HMO practice Retrospective Gender Office visits LES > CG
2001 CG=43 and Portugese sites chart review Age Rx written LES > CG
4,053 CG Income Rx filled LES > CG
Years enrolled Phone calls NS
Urgent care NS
visits
Mammograms NS
Breast exams NS
Pap tests NS
FOB testing LES > CG
Rectal exams LES > CG
Flu vaccines LES > CG
A—attending, C—Caucasian, CA—Central American, CG—control group, CH—Chicano, ED—emergency department, ES—English s peaking,
FOB—fecal occult blood, IM—internal medicine, ISG—interventional study group, MD=physician, LC—language concordant, LD—language discordant,
LES—limited English speaking, M—Mexican, NP—nurse practitioner, NR—not reported, NS—nonsignificant, PC—proximate consecutive interpretation,
R—resident, RN—registered nurse, RS—remote spontaneous interpretation, SS—Spanish speaking
* The following results were all significant with P<.05 or lower. LC > LD=language concordant higher than language discordant; LC < LD=language
concordant lower than language discordant; RS > PC or RS < PC= remote spontaneous interpretation higher or lower than proximate consecutive
interpretation; SS < ES=scores with Spanish speakers less than English speakers; LES > CG=intervention with limited English-speaking rates higher
than control group.
Modern Culture and Physician-Patient Communication Vol. 34, No. 5 357
Table 3
Studies of Physician Bias, Rapport Building, and Patient Preference, With Race/Ethnicity As Variables
STUDY POPULATION
Mean Age Outcome
Source # (Years) Demographics Practice Study Type Controls Measure Results*
Van Ryn,27 618 Patients=65 57% C, 43% AA, Eight New Cross-sectional Education Feelings of C > AA
2000 MD=45 53% male York hospitals physician and SES affiliation
Years of education patient survey Age Risk of C < AA
=12 Gender substance abuse
SES: 33% high, Intelligence C > AA
33% moderate, Interest in C > AA
33% poor active lifestyle/
cardiac
rehabilitation
MD participa- C > AA
tory style
AA—African American, C—Caucasian, ED—emergency department, H—Hispanic, MD=physician, NS—nonsignificant, SES—socioeconomic status
* The following results are statistically significant at P<.05 or lower: C > AA/C < AA=measure higher/lower for Caucasians than African American
patients. C > H/C < H=measure higher/lower for Caucasians than Hispanic patients.
many offers of information. Spanish-speaking patients somewhat less likely to have a positive perception of
were less likely to receive facilitation from physicians African Americans on a number of issues. Physicians
and were more likely to have their comments ignored rated African Americans as less likely to be the kind of
despite the presence of a professional interpreter.26 We person they could be friends with, as being less likely
could not find any studies that measured outcomes of to be free of substance abuse problems, and less likely
training providers in the use of interpreters. than Caucasians to be interested in an active lifestyle
and cardiac rehabilitation. Finally, physicians rated
Evidence of Physician Bias African Americans as less intelligent and less educated
Table 3 summarizes those studies that either directly than Caucasians. All of these relationships were stron-
or indirectly examine stereotyping and bias in physi- ger if the patient was from a lower socioeconomic
cian-patient interactions. The most direct evidence of class.27
such physician bias comes from a study of 618 post- Additional indirect evidence of racial bias emerged
angiogram visits performed by mostly Caucasian phy- from studies comparing pain treatment for long-bone
sicians with Caucasian and African American patients. fractures in emergency departments for Caucasians
Eight New York hospitals participated in the study. The versus Hispanics and Caucasians versus African Ameri-
authors surveyed physicians’ perceptions of and atti- cans. Studies in Los Angeles revealed that Hispanic
tudes toward patients, focusing on patients’ personal males were half as likely to receive analgesia despite
and psychosocial characteristics, behavior, and likely equivalent estimates of pain intensity by both physi-
role demands. They studied whether these perceptions cians and patients.28,29 The same authors performed a
or attitudes were affected by patient race or socioeco- retrospective review of African Americans and Cauca-
nomic status as independent variables. Physicians were sians with long-bone fractures in Atlanta. Blacks had a
358 May 2002 Family Medicine
Table 4
Hooper 150 < 45=28 67 C, 74 H Teaching Cross-sectional Age, gender, patient Measure of C> H
et al,38 1982 15 45–64=68 All English hospital observation appearance physician
MDs 65–74=40 speaking outpatient empathy
≥ 75=12 All MDs C clinic Rating of C> H
interview skill
Sleath,39 407 18–37=103 62% ES-H Internal Cross-sectional Age, gender, health MD demonstra- NS
2000 38–50=108 38% C medicine and observation perception, tion of empathy
51–57=90 family practice education, MD MD demonstra- C > H
58–88=106 resident clinics familiarity tion of
positiveness
Cooper- 1,816 All > 18 784 C patients IPA: 32 internal Telephone survey Age, gender, education, Participatory
Patrick,40 814 AA patients medicine and marital status, self- decision-making
1999 36 C MDs family reported health, style
16 AA MDs practices length of MD-patient C and AA AA < C
10 Asian MDs relationship patients with
2 Latino MDs C MD
Race
Concordant RD < RC
versus discor-
dant relation-
ships
Kaplan,41 8,316 46.5 78.3% non- NR Secondary analysis MD age, gender, Participatory
1995 minority of MOS specialty, ethnicity, decision-making
21.7% minority practice type, style
61.3% female geography Physician style Minority
19.6% report fair Patient age, gender, with minority score <
or poor health education, ethnicity, versus non- non-
193 general health status minority patients minority
internist MD Minority patients Minority
92 family physicians with minority MD <
versus non- non-
minority MD minority
MD
AA—African American, C—Caucasian, ES—English speaking, ES-H—English-speaking Hispanic, H—Hispanic, IPA—independent practice association,
LES—limited English speaking, MD—physician, MOS—Medical Outcomes Survey, NR—not reported, NS—nonsignificant, RC—race concordant, RD—
race discordant.
* The following results are statistically significant at P<.05 or lower. C > AA/C < AA=measure higher/lower for Caucasians than African American patients.
C > H/C < H=measure higher/lower for Caucasians than Hispanic patients. M > C=effect higher for minority than for Caucasian patients. RD < RC=race
discordant scores lower than race concordant scores.
Modern Culture and Physician-Patient Communication Vol. 34, No. 5 359
66% greater risk of receiving no analgesia, when com- by physicians also correlated strongly with patient sat-
pared to whites, after controlling for several covariates.30 isfaction. This trend was enhanced when there was gen-
der concordance as well.40 A secondary analysis of the
Relationship Building Between Patient and Physician Medical Outcomes Study, a 4-year longitudinal obser-
Rapport building via the use of empathy and effec- vational exploration, assessed participatory decision-
tive communication skills is critical to forming effec- making styles of physicians with both minority and non-
tive and trusting relationships with patients. Empathy minority patients. Minority patients on average rated
has demonstrated importance in the positive building physicians lower than non-minority patients. Interest-
of relationships.31 Positive nonverbal32 and verbal33 ex- ingly, however, minority patients scored non-minority
pressions to patients that demonstrate active listening physicians somewhat higher than minority physicians. 41
and respect for the patient’s positive attributes also im- While the results were statistically significant (P<.05),
prove the physician-patient relationship. Additionally, we point out that score differences were small.
communication skills that assist in patient assessment, We could not identify any studies that examined im-
particularly elicitation skills to understand the patient’s provements in physician communication as a result of
perspective of symptoms and explanatory health belief training. In contrast, there is growing evidence that train-
models, increase patient satisfaction, trust, and com- ing of patients to be more assertive is an effective strat-
pliance. Negotiation skills are also crucial to elucidate egy to improve doctor-patient communication.42
the patient’s perspective and encourage patient empow-
erment.34-36 Studies that compared physician commu- Discussion
nication and rapport-building skills in racially and eth- This review provides evidence that race, ethnicity,
nically concordant and discordant relationships with and language all affect the quality of the doctor-patient
patients are summarized in Table 4. relationship. Minority patients, especially those not
Three observational studies measured the ability of proficient in English, are less likely to engender em-
medical students and residents to build relationships pathic responses from physicians, less likely to estab-
with Hispanic patients in comparison to Caucasian pa- lish rapport with physicians, less likely to receive suf-
tients. Two of the studies showed significantly decreased ficient information, and less likely to be encouraged to
rapport building and empathy with Hispanic compared participate in medical decision making. These charac-
to Caucasian patients, even when interviews were con- teristics have all been linked to patient satisfaction,
ducted in English.37,38 In the third study, internal medi- patient compliance, and care outcomes in the general
cine and family practice residents spoke significantly literature on the doctor-patient relationship. Some of
fewer positive expressions to Hispanics than to non- the literature also validates calls for a more diverse
Hispanic patients. More positive expressions occurred physician workforce, since minority patients are more
if patients knew the physician and if they were more likely to choose minority physicians, be more satisfied
educated and rated as healthier. However, in this study, by language-concordant relationships, and feel more
expressions of empathy, while few, were equivalent in connected and involved in decision making with ra-
interviews with Hispanics and non-Hispanic white pa- cially concordant physicians. Studies support the con-
tients.39 clusion that professional interpreters are more likely to
Two of the above studies comparing care for Cauca- bridge the gaps in access experienced by non-English-
sians and Hispanics also demonstrated poorer interview- speaking patients, although at least one study demon-
ing skills by English-speaking resident physicians with strated persistently poor communication skills on the
English-speaking Hispanics. Interviewing skills of resi- part of the physicians using such interpreters.
dents with Anglos were rated significantly higher than While the evidence is convincing that “majority”
with Hispanics.38 In the second study, trainees provided physicians need to be more effective in developing re-
better explanations and more feedback of higher qual- lationships and in their communication with ethnic and
ity to Anglo patients.37 racial minority patients, we found no studies that dem-
Two additional studies measured participatory deci- onstrate improvement through training. This finding is
sion making by physicians with racially concordant and likely to be due to the paucity of formal training pro-
discordant patients. Negotiation and encouragement of grams in medical schools and residencies.43,44
patient participation in problem management by phy-
sicians were rated worse by African American patients Limitations
compared with Caucasians in a telephone survey of Our review of the literature has limitations. While
privately insured patients from 32 group practices. we define culture broadly in practice, we limited our
Ratings of white and minority physicians were not dif- review only to ethnicity, race, and language. Addition-
ferent overall, but patients in race-concordant relation- ally, we could not find studies using rigorous qualita-
ships believed visits were more participatory than in tive methods from peer-reviewed journals. Moreover,
discordant ones. A participatory decision-making style the broad scope of work published in books, a rich
360 May 2002 Family Medicine
source of medical anthropology, was not reviewed. tient trust and disclosure of concerns? Without address-
Additional limitations include the significant number ing issues such as these, the goals of Healthy People
of studies conducted in emergency medicine settings 2010 may still be our goals in 2050.
and involving trainees that may not generalize to a larger
population of patients in continuity-based relationships Acknowledgments: We thank Andrew Dzaugis, Rosemary Leary, and Nancy
Harger, all staff of the Homer Gage Library of UMass Memorial Health
with experienced physicians. Care, for their assistance with the literature search and retrieval of articles.
22. Perez-Stable EJ, Napoles-Springer A, Miramontes JM. The effects of 38. Hooper EM, Comstock LM, Goodwin JM, Goodwin JS. Patient char-
ethnicity and language on medical outcomes of patient care with hy- acteristics that influence physician behavior. Med Care 1982;20(6):
pertension and diabetes. Med Care 1997;35(12):1212-9. 630-8.
23. Kuo D, Fagan MJ. Satisfaction with methods of Spanish interpretation 39. Sleath B, Rubin RH, Arrey-Wastavino A. Physician expression of em-
in an ambulatory care clinic. J Gen Intern Med 1999;14:547-50. pathy and positiveness to Hispanic and non-Hispanic white patients
24. Hornberger JC, Gibson CD, Wood W, et al. Eliminating language bar- during medical encounters. Fam Med 2000;32(2):91-6.
riers for non-English-speaking patients. Med Care 1996;34(8):845-56. 40. Cooper-Patrick L, Gallo JJ, Gonzales JJ, et al. Race, gender, and part-
25. Jacobs EA, Lauderdale DS, Meltzer D, Shorey JM, Levinson W, Thisted nership in the patient-physician relationship. JAMA 1999;282:583-9.
RA. Impact of interpreter services on delivery of health care to limited- 41. Kaplan SH, Bandek B, Greenfield S, Rogers W, Ware JE. Patient and
English-proficient patients. J Gen Intern Med 2001;16:468-74. vis it characteristics related to phys icians’ p articip atory decision-
26. Rivadeneyra R, Elderkin-Thompson V, Silver RC, Waitzkin H. Patient making style. Results from the Medical Outcomes Study. Med Care
centeredness in medical encounters requiring an interpreter. Am J Med 1995;33:1176-87.
2000;108(6):470-4. 42. Post DM, Cegala DJ, Miser WF. The other half of the whole: teaching
27. Van Ryn JB. The effect of patient race and socioeconomic status on patients to communicate with physicians. Fam Med 2002;34(5):344-
physicians’ perceptions of patients. Soc Sci Med 2000;50:813-28. 52.
28. Todd KH, Samaroo N, Hoffman JR. Ethnicity as a risk factor for inad- 43. Flores G, Gee D, Kastner B. The teaching of cultural issues in US and
equate emergency room department analgesia. JAMA 1993;269(12): Canadian medical schools. Acad Med 2000;75:451-5.
1537-9. 44. Culhane-Pera KA, Like RC, Leben soh n-Chialvo P, L oewe R.
29. Todd KH, Samaroo N, Hoffman JR. The effect of ethnicity on physi- Multicultural curricula in family practice residencies. Fam Med 2000;
cians’ estimates of pain severity in patients with isolated extremity 32(3):167-73.
trauma. JAMA 1994;271(12):925-8. 45. Gray B, Stoddard JJ. Patient-physician pairings: does racial and ethnic
30. Todd KH, Deaton C, D’Adamo AP, Goe L. Ethnicity and analgesic congruity influence selection of a regular physician? J Comm Health
practices. Ann Emerg Med 2000;35(1):11-6. 1997;22(4):247-59.
31. Brock CD, Salinsky JV. Empathy: an essential skill for understanding 46. Moy E, Bartman BA. Phys ician race and care of minority and medi-
the physician-patient relationship in clinical practice. Fam Med 1993; cally indigent patients. JAMA 1995;273(19):1515-20.
25(4):245-8. 47. Corkery E, Palmer C, Foley ME, Schechter CB, Frisher L, Roman SH.
32. Koss T, Rosenthal R. Interactional synchrony, positivity, and patient Effect of a bicultural community health worker on completion of dia-
satisfaction in the physician-patient relationship. Med Care 1997;35(11): betes education in a Hispanic population. Diabetes Care 1997;20(3):
1158-63. 254-7.
33. Hall J, Rotor DL, Katz NR. Meta-analysis of correlates of provider 48. Cook-Gotay C, Wilson ME. Social support and cancer screening in
behavior in medical encounters. Med Care 1988;26:657-75. African American, Hispanic, and Native American women. Cancer Prac-
34. Kleinman A, Eisenberg L, Good B. Culture, illness, and care: clinical tice 1998;6(1):31-7.
lessons from anthropologic and cross-cultural research. Ann Intern Med 49. Palos G. Cultural heritage: cancer screening and early detection. Semin
1978;88:251-8. Oncol Nurs 1994;10(2):104-13.
35. Berlin EA, Fowkes WC. A teaching framework for cross-cultural health 50. Hayes RP, Baker DW. Methodological problems in comparing English-
care. Application in family practice. West J Med 1983;139:934-8. speaking and Spanish-speaking patients’ satisfaction with interpersonal
36. Zweifler J, Gonzalez AM. Teaching residents to care for culturally di- aspects of care. Med Care 1998;36(2):230-6.
verse populations. Acad Med 1998;73:1056-61.
37. Shapiro J, Saltzer E. Cross-cultural aspects of physician-patient com-
munication patterns. Urban Health 1981;Dec:10-5.