You are on page 1of 6

Issue Brief

Findings from HSC N O . 1 3 0 • F EBRUARY 2010

MODEST AND UNEVEN: While nearly half of U. S. physicians identify language or cultural communication
PHYSICIAN EFFORTS TO barriers as obstacles to providing high-quality care, physician adoption of practices
to overcome such barriers is modest and uneven, according to a new national study
by the Center for Studying Health System Change (HSC). Despite broad consen-
sus among the medical community about how physicians can help to address and,
By James D. Reschovsky and Ellyn R. Boukus ultimately, reduce racial and ethnic disparities, physician adoption of several recom-
mended practices to improve care for minority patients ranges from 7 percent report-
ing they have the capability to track patients’ preferred language to 40 percent report-
ing they have received training in minority health issues to slightly more than half
reporting their practices provide some interpreter services. The challenges physicians
face in providing quality health care to all of their patients will keep mounting as the
U.S. population continues to diversify and the minority population increases.

Physicians' Role in Addressing Racial and Ethnic Disparities

B eginning in 2003, the federal govern-

ment through the National Healthcare
Disparities Report has annually documented
the quality of care provided across racial
and ethnic groups.2
Drawing on these recommendations, the
• whether their practice has information
technology (IT) to identify patients’ pre-
ferred language; and
wide disparities in the quality of health nationally representative 2008 HSC Health
• whether they receive reports about the
care received by different racial and ethnic Tracking Physician Survey asked physi-
quality of care delivered to minority
groups in the United States. And, little prog- cians about steps they and their practices
ress has been made in closing these quality have taken to reduce language and cultural
gaps, according to the latest report issued in barriers and generate information at the
May 2009. practice level to improve care for minor- Reducing Language and
Although differences in insurance ity patients (see Data Source). The survey Cultural Barriers
coverage and other patient, community asked physicians the following:
and health system factors contribute to A prerequisite for quality medical care is
• whether their practice provides inter- effective communication between patient
disparities, studies indicate disparities also
preter services; and caregiver. Nearly half (48.6%) of all
can arise during the patient-physician
encounter.1 The Institute of Medicine and • whether their practice provides patient- U.S. physicians in 2008 reported that diffi-
the National Quality Forum, among oth- education materials in languages other culty communicating with patients because
ers, have outlined measures physicians and than English; of language or cultural barriers was at least
physician practices can take: 1) to reduce a minor problem affecting their ability to
• whether they have received training in provide quality care, though less than 5
language and cultural barriers that hinder
minority health issues; percent viewed it as a major problem. The
communication between some minority
patients and their physicians; and 2) to • whether they receive reports containing failure of physicians to address communi-
establish practice-level information systems patient demographic information, such cation barriers, coupled with the potential
to facilitate the elimination of language and as race or ethnicity; ineffectiveness of measures taken, may con-
cultural barriers and provide feedback on tribute to disparities in the quality of care

Providing Insights that Contribute to Better Health Policy

Center for Studying Health System Change Issue Brief No. 130 • February 2010

Table 1 by providing patients with written informa-

U.S. Physicians Implementing Select Tools Aimed at Reducing Racial/Ethnic tion about their conditions and self-care
Disparities, 2008 instructions.8 Offering patient-education
materials in appropriate languages for
Practice Provides Interpreter Services1 55.8%
patients with limited-English proficiency also
Practice Provides Patient-Education Materials in Languages can promote health literacy. Among physi-
other than English2 cians in practices treating patients with any of
Physician Received Training in Minority Health3 40.3 four prevalent chronic conditions—asthma,
Physician Receives Reports on Own Patients' Demographic diabetes, congestive heart failure and depres-
23.2 sion—72 percent in 2008 reported their prac-
Information Technology to Access Patients' Preferred tice provides patient-education materials for
7.3 at least one of the four conditions.9 Yet, only
Language is Available and Used Routinely1
Physician Receives Reports on Quality of Care for Own 40 percent of physicians in these practices
11.8 reported providing patient-education materi-
Minority Patients3
Excludes physicians who reported having no non-English speaking patients. als in languages other than English for at least
Population consists of physicians whose practices treat at least one of the following chronic conditions: diabetes, asthma, depres- one of the conditions.
sion, congestive heart failure. Population excludes physicians who report having no non-English speaking patients. Physician training in minority health.
Excludes physicians who report having no minority patients. Culturally competent care emphasizes com-
Source: HSC 2008 Health Tracking Physician Survey prehension of cultural differences and the
interaction with individuals’ health expec-
tations and behaviors, disease incidence
across racial and ethnic groups. may reflect greater demand for interpreter and prevalence, and treatment outcomes.
Interpreter services. An ever-increasing services among practices experiencing The goal of minority health education is to
number of people in the United States language barrier problems, it also may develop practitioner skills to tailor care to
speak a language other than English at indicate that interpreter services are not patients’ culturally unique needs.10 Roughly
home (56 million people aged 5 and older always readily available or are inadequate.5 four in 10 physicians in 2008 reported they
in 2008, 44% of whom reported they speak Nearly one in five physicians (18.8%) have received some training in minority
English less than “very well”).3 These facts reported being unable to obtain inter- health, such as cultural competency training,
highlight the need for interpreter services preter services in the past 12 months that through professional meetings, workshops
to improve the quality of care for patients they believed were medically necessary. or continuing medical education courses.
with limited-English proficiency.4 Moreover, of physicians reporting that However, the survey question did not assess
Moreover, providers have legal obliga- their practices provided interpreter servic- the nature of the training, its comprehen-
tions to provide needed interpreter ser- es, it is unclear how many provide profes- siveness or how recently it was received.
vices, at least for patients with public insur- sional interpretation services, as opposed Although more likely to have received train-
ance. While nearly 97 percent of physicians to using staff members who may be less ing in minority health than other physicians,
have at least some non-English speaking than fluent in the language, unfamiliar only half of physicians in high-minority
patients, only slightly more than half of with medical terminology or unaware of practices (defined as 50% or more minority
physicians (56%) were in practices that cultural nuances. patients) had received such training.
provided interpreter services in 2008 (see Non-English, written patient-education
Table 1). Of physicians in practices that materials. Low health literacy—defined Information Feedback to
provided interpreter services, 44 percent as limited capacity to obtain, process and Physicians
were in practices that offered interpretation understand health information and services
in only one language, 16 percent were in needed to make appropriate decisions6—is Other prerequisites for improving care for
practices offering two languages, with the associated with less use of preventive ser- minority patients include identifying them,
rest in practices providing interpretation vices, more frequent hospital and emergen- knowing what language they prefer and
in three or more languages, including tel- cy department visits, and poorer health.7 monitoring what quality of care they receive.
ephonic translation services. Minorities, particularly those not proficient Although such information could exist in
Interestingly, among physicians with in English, are disproportionately repre- purely paper recordkeeping systems, recent
patients who speak different languages, sented among individuals with low health national efforts have focused on electronic
those in practices providing interpreter literacy. medical records (EMRs). Movement in this
services were more likely to report com- Although other care management prac- direction is likely to accelerate in the near
munication difficulties than those without tices have been found to be more effective future because of incentives in the American
access to interpreter services, even after than written materials in educating and Recovery and Reinvestment Act of 2009 and
adjusting for the percentage of minority engaging patients in their own self-care, emerging guidelines from the Office of the
patients treated (see Table 2). While this physicians can help promote health literacy National Coordinator for Health Information

Center for Studying Health System Change Issue Brief No. 130 • February 2010

Table 2
Patient Communication Difficulties among U.S. Physicians, by Level of Interpreter Support, 2008

All Unable Reported Difficulties Proportion of

Physicians to Obtain Communicating with Patients Physician
Interpreter Patients Because of has Difficulties
Services in Past Language or Cultural Communicating with
Year Barriers a major Because they Speak a
Problem Different Language
No Interpreter Services (R) 42.8% 15.8% 3.8% 4.3%
Practice Provides One Language Only 23.5 24.0** 4.4 7.2**
Practice Provides Two Languages 8.7 20.9* 4.2 7.8**
Practice Provides Three or More
13.2 20.1* 5.5 8.5**
Practice Provides Translator
8.3 18.0 4.8 8.2**
Notes: “Translator services” refers to any on-demand language resource, such as a telephone interpreter bank. Physicians who indicated that “most,” “any” or “all” languages were available in their practice
were assumed to have access to such a service. This table presents regression-adjusted means that control for the minority composition of physicians’ patients. The All Physicians column does not sum to
100% because 3.5% of physicians reported having no non-English speaking patients.
* Difference from reference group, as indicated by (R), is statistically significant at p<.05. ** at p<.01.
Source: HSC 2008 Health Tracking Physician Survey

Technology and the Centers for Medicare Addressing Disparities Varies by Apart from the minority composition
and Medicaid Services that include these Practice Characteristics of physicians’ patient panels, practice type
capabilities in definitions of “meaning- also demonstrates a strong association with
Adoption rates vary not only across the dif- efforts to address disparities.12 Physicians
ful use” of EMRs qualifying for subsi-
ferent tools, but also according to practice in solo and group practices were less likely
characteristics (see Table 3).11 Physicians to adopt measures to address disparities
Reporting of patient demographic
in practices with a greater share of minor- than those in institutional practices, such as
information and access to patient-pre-
ity patients were more likely to adopt each hospitals, health maintenance organizations
ferred language. Less than one in four
of the measures to address disparities. For (HMOs) and medical schools. For example,
physicians (23%) indicated they receive
example, almost twice as many physicians nearly 90 percent of physicians in group-
reports on patient demographics. Among
reported providing interpreter services or staff-model HMOs reported providing
physicians in high-minority practices, only
in practices with a majority of minority interpreter services, compared with 34 per-
a slightly higher percentage (29%) received
patients, relative to those in low-minority cent of physicians in solo or two-physician
demographic reports.
practices—less than 10 percent minor- practices. Physicians in HMOs also were
Knowing what language a patient ity patients (72.3% vs. 39.2%). Similarly,
prefers is important for arranging and more than 10 times as likely as those in
there were large differences in provision of solo or two-physician practices to routinely
coordinating interpreter services for patient-education materials in foreign lan-
patients with limited-English proficiency use IT to access information on patients’
guages (60% vs. 24%), routine use of IT to preferred language.
or matching appropriate physicians to access patients’ preferred languages (10.5%
patients. Twenty-two percent of physicians vs. 4.3%) and quality reporting by racial or
indicated their practice has IT capable Policy Implications
ethnic group (16.8% vs. 8.2%).
of reporting patients’ preferred language Weighting physicians by the propor- Despite broad consensus among the medi-
(findings not shown), but only a third of tion of minority patients they treat allows cal community about how physicians can
these physicians (7%) routinely used this a rough estimate of the distribution of help to address and, ultimately, reduce
capability. minority patients treated by physicians disparities, physician adoption of several
Reporting of quality of care delivered using various tools to reduce disparities. recommended practices to improve care
to minority patients. Nearly nine out of 10 For example, while 56 percent of physicians for minority patients remains low. Cost and
physicians lacked a formal means to assess provided interpreter services, 64 percent lack of reimbursement for these activities
the quality of care provided to patients of minority patients were treated by physi- are likely among the largest obstacles to
across racial and ethnic groups. Only 11.8 cians in practices with interpreter services. implementation in physician practices.
percent of physicians reported access to And, 14 percent of minority patients were The tools most commonly adopted tend
reports on the quality of care they provide treated by physicians who received reports to be the least expensive to implement:
stratified by patient race or ethnicity. on the quality of care for minority patients provision of patient-education materials
(findings not shown). in foreign languages, which can often be

Center for Studying Health System Change Issue Brief No. 130 • February 2010

Table 3
U.S. Physicians and Disparity Reduction Efforts, by Minority Patient Composition and Practice Type, 2008
Disparity Reductiton Tools
All Average Providing Providing Trained in Receiving Routinely Receiving
Physicians Minority Interpreters2 Any Minority Patient Use IT to Quality
Patients1 Patient- Health Demographic Access Reports4
Education Issues4 Reports4 Patients'
Materials Preferred
in Non- Language2
All Physicians 100% 32.7% 55.8% 40.1% 40.3% 23.2% 7.3% 11.8%
Percent Minority Patients
Low (<10%) (R) 17.8 3.2 39.2 24.1 35.7 17.9 4.3 8.2
Medium (10-50%) 61.4 27.3 54.5** 37.3** 37.3 22.3* 7.0** 10.8
High (>50%) 20.8 73.8 72.3** 59.9** 51.5** 28.8** 10.5** 16.8
Type of Practice
31.2 30.9 34.4 29.8 36.1 19.7 3.1 13.9
Physicians (R)
Group (3-5
15.4 28.1 42.4** 32.5 31.7 18.4 4.2 9.2
Group (6-50
19.2 29.5 51.5** 33.3 32.5 21.0 5.0* 9.3
Group (51+
6.1 25.9 72.7** 46.9** 38.4 25.5 10.4** 8.8
3.5 35.7 90.6** 75.1** 71.3** 48.7** 33.2** 23.4
23.6 41.6 85.7** 53.0** 52.5** 28.2** 11.9** 12.2
* Difference from reference group, as indicated by (R), is statistically significant at p<.05. ** at p<.01.
This is the percent of patients treated who are black or Hispanic, as reported by physicians.
Excludes physicians who report having no non-English speaking patients.
Population consists of physicians whose practices treat at least one of the following chronic conditions: diabetes, asthma, depression, congestive heart failure.. Population excludes physicians who report
having no non-English speaking patients.
Excludes physicians who report having no minority patients.
Institutional practice includes community health centers, hospitals, and medical school/university.
Source: HSC 2008 Health Tracking Physician Survey

downloaded from the Internet for free, and vices. This may help to explain their higher affected and the financial burdens imposed
training in minority health. On the other adoption rates relative to solo and small- on providers. Low payment rates, especially
hand, IT systems that can support report- group practices. in Medicaid and SCHIP, mean that aggres-
ing on patient care by race, ethnicity or Competent interpreter services are a key sive enforcement could hinder physicians’
language, as well as interpreter services, are step in improving physician-patient com- willingness to treat these patients.
expensive and less common. For instance, an munication when language barriers are a Therefore, it is not surprising that physi-
encounter with an interpreter involves the problem. Under federal law—Title VI of cians working in hospital and other insti-
cost of paying the interpreter and requires the Civil Rights Act—health care providers tutional settings were more likely to have
the physician to spend more time with the who treat patients with public insurance— interpreter services available than physicians
patient—time that could have been devoted Medicare, Medicaid and the State Children’s in solo or group practices. As of 2008, all 50
to seeing more patients. Larger physician Health Insurance Program (SCHIP)—have states had laws governing language services
organizations and those with higher concen- an obligation to provide interpreter services in health care settings.13 However, these
trations of patients needing interpretation to those patients. However, enforcement is laws vary by state, languages and/or condi-
services can take advantage of scale econo- subject to judgments about the number of tions covered, provider setting, and level of
mies to more efficiently provide these ser- patients with limited-English proficiency enforcement. Moreover, only some states

Center for Studying Health System Change Issue Brief No. 130 • February 2010

provide reimbursement for interpreters and Ethnic Disparities in Health Care,

under Medicaid and SCHIP, and often these which identified physician bias, ste-
states rank among those with the fewest resi- reotyping and uncertainty on the part
dents with limited-English proficiency. of providers as factors influencing dis-
Even when interpreters are available, parities. Similar recommendations are
the benefit to the patient can be uncertain. contained in National Quality Forum
Providers often rely on patients’ relatives, guidelines. See http://www.qualityforum.
untrained bilingual staff or other ad-hoc org/Topics/Disparities.aspx.
interpreters. The fact that the provision of
3. These estimates come from the 2008
interpretation services shows little relation-
American Community Survey 1-Year
ship to physician reports of difficulties with
Estimates published by the U.S. Census
language and cultural barriers raises ques-
Bureau. The full table can be found on
tions of effectiveness. Little policy atten-
the Census Bureau Web site at Table Despite broad
tion has been given to clarifying the legal
S1601, Language Spoken at Home.
framework governing interpreter services
and establishing uniform standards for inter- 4. Use of interpreters in clinical encoun- consensus among the
preter services. However, in September 2009, ters, especially those who are profession-
a new group, the Certification Commission ally trained, has been associated with
for Healthcare Interpreters, was formed to higher provider and patient satisfaction medical community
establish national certification for health (Carrasquillo, Olveen, et al., “Impact of
care interpreters, and the National Council Language Barriers on Patient Satisfaction about how physicians
on Interpreting in Health Care is developing in and Emergency Department,” Journal
national training standards for interpreters. of Internal Medicine, Vol. 14, No. 2
Many private insurers’ physician direc- (February 1999)) and improved health can help to address
tories list languages spoken by participat- outcomes via reductions in errors and
ing physicians, and the public sector might enhanced patient comprehension and and, ultimately, reduce
consider similar steps to direct patients to utilization (Jacobs, Elizabeth A., et
physicians who speak their language or offer al., “Overcoming Language Barriers
appropriate interpreter services. Moreover, in Health Care: Costs and Benefits of disparities, physician
policy makers likely will need to consider Interpreter Services,” American Journal
additional subsidies to support interpreta- of Public Health, Vol. 95, No. 5 (May adoption of several
tion services. 2004)). Conversely, language barri-
The challenges physicians face in provid- ers are commonly associated with lack
ing quality health care to all of their patients of a usual source of care, lower health recommended prac-
will keep mounting as the U.S. population service utilization, poor adherence to
continues to diversify and the minority and misunderstanding of treatment tices to improve care
population grows. Although disparities can and follow-up regimens, lower satisfac-
stem from factors beyond the physician- tion, and higher incidence of medical
patient encounter, the ability of physicians complications (Karliner, Leah S., et al., for minority patients
to communicate effectively with patients and “Do Professional Interpreters Improve
understand their cultural and social context Clinical Care for Patients with Limited remains low.
is important in caring for a diverse patient English Proficiency? A Systematic
population. Review of the Literature,” Health Services
Research, Vol. 42, No. 2 (April 2007)).
Notes 5. The relationship between the percentage
1. Smedley, Brian D., Adrienne Y. Stith of patients with whom physicians have
and Alan R. Nelson (eds.) Unequal difficulty communicating and the num-
Treatment: Confronting Racial and Ethnic ber of languages for which their main
Disparities in Health Care, Institute of practice provides translation services is
Medicine (2003). only slightly attenuated after adjusting
for the percentage of Latinos, Asians and
2. These measures were included in the Native Americans who are treated.
2003 Institute of Medicine (IOM) report,
Unequal Treatment: Confronting Racial 6. U.S. Department of Health and

Center for Studying Health System Change Issue Brief No. 130 • February 2010

Data Source Human Services, Healthy People 2010: ity health training, have IT available to
Understanding and Improving Health obtain patients’ preferred languages, and
This Issue Brief presents findings from (November 2000). receive quality reports according to race
the HSC 2008 Health Tracking Physician or ethnicity, while surgeons were more
7. Baker, David W., et al., “Health Literacy
Survey, a nationally representative mail likely to provide interpreters.
and Mortality Among Elderly Persons,”
survey of U.S. physicians. The sample of Archives of Internal Medicine, Vol. 167, 13. Au, Melanie, Erin Fries Taylor and
physicians was drawn from the American No. 14 (July 2007). Marsha Gold, Improving Access to
Medical Association master file and Language Services in Health Care: A
8. Roter, Debra L., et al., “Effectiveness
included active, nonfederal, office- and Look at National and State Efforts, Policy
of Interventions to Improve Patient
hospital-based physicians providing at Brief, Mathematica Policy Research,
Compliance: A Meta-Analysis,” Medical
Washington, D.C. (April 2009).
least 20 hours per week of direct patient Care, Vol. 36, No. 8 (August 1998).
care. Residents and fellows were excluded. 9. Roughly three-quarters of physicians are
The survey includes responses from more in practices that treat patients with at
than 4,700 physicians and had a 62 per- least one of the four focal chronic condi-
cent response rate. Estimates from this tions. Ninety percent of primary care
survey should not be compared to esti- physicians are in practices that treat all
of the four conditions.
mates from HSC’s previous Community
Tracking Study Physician Surveys because 10. Betancourt, Joseph R., et al., “Defining
of changes in survey administration mode Cultural Competence: A Practical
Framework for Addressing Racial/Ethnic
from telephone to mail, question wording,
Disparities in Health and Health Care,”
skip patterns, sample structure and popu- Public Health Reports, Vol. 118 (July/
lation represented. More detailed infor- August 2003).
mation on survey content and methodol-
11. All six measures are not relevant to all
ogy can be found at physicians. For example, interpreter
services are not relevant to the 3.5 per-
cent of physicians with no non-English
speaking patients. Moreover, we are not
able to assess whether the quarter of
physicians in practices that do not treat
any of the four major chronic conditions
provide patient information materials in
languages other than English for condi-
Funding Acknowledgement tions they do treat. See Table 2 notes for
This research was supported by the more information.
Robert Wood Johnson Foundation. 12. In addition to practice type and percent-
age of minority patients, other physi-
cian and practice characteristics are
ISSUE BRIEFS are published by the
associated with adoption of practices
Center for Studying Health System Change. to reduce racial and ethnic disparities.
Minority—black and Hispanic—physi-
600 Maryland Avenue, SW, Suite 550
Washington, DC 20024-2512
cians are more likely to implement each
Tel: (202) 484-5261 of the tools than white, non-Hispanic
Fax: (202) 484-9258 physicians. With the exception of receiv-
ing patient demographic reports, the
President: Paul B. Ginsburg same is true for physicians practicing
in large urban areas relative to rural
areas. Across specialties, primary care
physicians and medical specialists were
more likely than surgeons to provide
patient-education materials in languages
other than English, have received minor-

HSC, funded in part by the Robert Wood Johnson Foundation, is affiliated with Mathematica Policy Research