ARTICLE |
a Division of General Pediatrics and Adolescent Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland
b Division of Health Services Research, American Academy of Pediatrics, Elk Grove Village, Illinois
c Center for the Advancement of Underserved Children, Departments of Pediatrics and Epidemiology, Health Policy Institute, Medical College of Wisconsin and Children's Research Institute, Milwaukee, Wisconsin
d Women's and Children's Health Policy Center, Department of Population, Family and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
| ABSTRACT |
|---|
|
|
|---|
METHODS. Data were obtained from the Periodic Survey of Fellows No. 60, a nationally representative survey of members of the American Academy of Pediatrics. A total of 1829 surveys were mailed, and responses were obtained from 58%. Use of 6 language services was assessed. Factors associated with language services use were examined after adjusting for physician, practice, and state characteristics.
RESULTS. Bilingual family members (70%) and bilingual staff (58%) were the most frequently reported language services; 40% of respondents report the use of professional interpreters, 28% use telephone interpreters, and 35% of practices report provision of translated written materials. Pediatricians in smaller and rural practices and in states with higher proportions of limited English proficiency persons report less use of professional interpreters. Pediatricians in states with third-party reimbursement for language services are more likely to report use of professional interpreters.
CONCLUSIONS. Most pediatricians report using untrained interpreters to communicate with limited English proficiency patients and their families. Pediatricians in regions with high proportions of limited English proficiency persons may be less likely to provide appropriate language services. Third-party reimbursement for professional language services may increase the use of trained interpreters and quality of care.
Key Words: child health services multilingualism translating
Abbreviations: LEPlimited English proficiency AAPAmerican Academy of Pediatrics SCHIPState Children's Health Insurance Program ORodds ratio CIconfidence interval
Rapid growth is occurring in the number of Americans with limited English proficiency (LEP), defined as those with a self-reported ability to speak English less than "very well."1 Twenty-one million Americans had LEP in 2000, a substantial increase from 14 million who had LEP in 1990.1,2 Patients with language barriers are at risk for impaired health status,35 lack of health insurance,6 increased test charges and lengths of stay in emergency departments,7 increased adverse events during hospitalization,8 decreased adherence to medications and follow-up appointments,9 and a lower likelihood of receiving follow-up appointments after emergency department visits.10
The array of language services used in health care settings includes bilingual physicians, bilingual staff, bilingual family members and friends, professional interpreters, telephone interpreters, and written materials in the primary language. Different methods, however, have been shown to have varying levels of effectiveness. The use of bilingual physicians and professional interpreters results in optimal communication and improved medical outcomes5 and has been linked with higher patient adherence, increased use of screening tests, an increased number of office visits, higher rates of prescriptions being filled, fewer laboratory tests ordered, increased adherence to guidelines, and increased patient satisfaction.1115 In contrast, untrained ad hoc interpreters, such as untrained staff or family members, friends, or strangers from the waiting room, are associated with poorer self-reported understanding of diagnoses, increased numbers of interpreter errors of clinical consequence,1618 and higher rates of testing and admission from emergency departments.12
Title VI of the Civil Rights Act of 1964 mandates that health care providers receiving federal funds provide "meaningful access to their programs and activities by LEP persons" without cost to the patient.19 The use of trained interpreters among internists and family physicians is low, with cited barriers being cost, inconvenience, limited availability of trained interpreters, and an ongoing perception that ad hoc interpreters are sufficient.20,21 Little is known about pediatricians' use of language services for families with LEP. The objectives of this study were to examine how pediatricians communicate with families with LEP and to examine the pediatrician, practice, and state characteristics associated with use of various language services.
| METHODS |
|---|
|
|
|---|
Respondents estimated the percentage of their patients with LEP, defined in the question as those for whom English was not the primary language and was spoken less than "very well." Questions were asked about the primary languages spoken by patients with LEP. For each primary language encountered, respondents were asked whether they provided each of 6 different methods of communication: bilingual physicians (self or other), bilingual staff, bilingual family members, professional interpreters, written materials in the primary language, and telephone interpreters.
Additional information was collected regarding physician and practice characteristics. Physician characteristics included age, gender, race, and Hispanic/Latino ethnicity (henceforth reported as Latino). Practices were categorized by setting (solo/2 physician, group with 310 pediatricians, group with >10 pediatricians/multispecialty group/staff model health maintenance organization, or hospital/clinic/medical school), location (inner-city, other urban, suburban, or rural), and region (Northeast, Midwest, South, or West). Practices were dichotomized at the mean by the percentages of patients with public insurance as low (
36%) or high (>36%) and also as having <20% or
20% Latino patients.
States in which pediatricians practice were categorized by US Census-reported prevalence of LEP persons, growth in the Latino population, and the availability of public third-party (Medicaid or State Children's Health Insurance Program, SCHIP) reimbursement for language services. Ten states were identified as having high LEP, with
9% of the population speaking English less than "very well" relative to the national mean LEP proportion of 8%.23 Ten states were categorized as having high LEP among Spanish speakers, defined as having
5% of the population speaking Spanish at home and speaking English less than very well.23 Eight states were identified as having high LEP among Asian speakers, defined as having
2% of the population speaking an Asian or Pacific Island language and speaking English less than very well.23 Twelve states were categorized as having a high interval population increase of Latinos, defined as an increase of >200000 Latinos between the 1990 and 2000 US Census.24 Ten states were categorized as providing Medicaid/SCHIP reimbursement for interpreter services, ranging from direct provider or interpreter reimbursement to contracts with interpreter organizations for language services.25
Analyses were conducted using SPSS 12.0 (SPSS Inc, Chicago, IL). Tests for significance of means were conducted by t test and medians by Mann-Whitney and Kruskal-Wallis tests. Bivariate analyses were performed using
2 for categorical and analysis of variance for continuous variables. Multivariate analysis was performed by logistic regression using 3 domains of independent variables selected with health policy significance in mind: physician (age, gender, and Latino ethnicity), practice (setting and location), and state (LEP prevalence and Medicaid/SCHIP reimbursement) characteristics. The 6 methods of communication were selected as dependent variables, and all of the variables were forced in.
| RESULTS |
|---|
|
|
|---|
Analyses were limited to the 835 respondents who finished their residency training, did not have a specialty fellow (certified by a board other than a pediatric board) designation, reported having patients with LEP, and responded to the subset of questions focusing on language services provided. A total of 51.8% of respondents included for analysis were women, with a mean age of 45.2 years and 7.6% reporting Latino ethnicity. Compared with 2006 national AAP member data, respondents included for analysis were more likely to report a younger age (mean age: 45.2 vs 46.3 years; P < .05); no significant differences were found in gender or geographic distribution. Because the national AAP membership database does not contain sufficient information on ethnicity, aggregate data from AAP Periodic Surveys 59 to 61 were used for national comparison; survey respondents included for analysis were more likely to report Latino ethnicity (7.6% versus 5.2%; P < .001).
Respondents self-reported a median percentage of 5.0% (mean: 13.4%) of patients with LEP. No patients with LEP were seen at 13% of practices, and >20% of patients had LEP in 19.3% of practices. Pediatricians reported 62 different languages spoken at home among their patients with LEP, most commonly Spanish (reported by 94.0%), Chinese (11.8%), and Vietnamese (10.5%). Pediatricians estimated that 53.8% of their patients were non-Latino white, 19.6% were Latino, 18.2% were black, and 5.8% were Asian/Pacific Islander. A mean of 53.2% of the patients had private insurance coverage, 36.7% had public insurance, 3.3% had TRICARE (military) insurance, and 7.0% were uninsured.
Pediatricians who were <45 years old, women, or Latino reported caring for significantly higher proportions of LEP patients (P < .05; Table 1). In addition, higher median proportions of patients with LEP were reported being seen by pediatricians practicing in hospitals, clinics or medical schools, the inner city, the Western United States, and those caring for high percentages of Latino and publicly insured patients.
|
|
|
|
In multivariate analysis (Table 5), female physicians had increased odds of using a bilingual family member (odds ratio, OR: 1.49; 95% confidence interval, CI: 1.052.11) and written materials in the primary language (OR: 1.60; 95% CI: 1.152.23). Non-Latino physicians had decreased odds of using a bilingual physician and increased odds of using bilingual family members, bilingual staff, and professional interpreters. Suburban, rural, and noninner city urban practices were substantially less likely to use professional interpreters. Smaller practices had 8 to 9 times lower odds of using professional interpreters. Physicians in states with higher proportions of LEP persons were more likely to use bilingual physicians (OR: 2.17; 95% CI: 1.553.05) and bilingual staff (OR: 4.63; 95% CI: 3.226.67) but less likely to use professional interpreters (OR: 0.46; 95% CI: 0.310.69). Physicians in states with public third-party reimbursement for language services were more likely to use professional interpreters (OR: 2.05; 95% CI: 1.103.83).
|
| DISCUSSION |
|---|
|
|
|---|
Several findings have key policy implications. Smaller and rural practices are less likely to use professional interpreters, even after adjustment for state LEP prevalence and third-party reimbursement for interpreter services. Further research is needed to determine whether this finding relates to insufficient availability of professional interpreters, inadequate reimbursement for language services, or the need for greater education on the importance of professional interpreters in providing high-quality care and optimal communication. Non-Latino physicians report greater use of bilingual staff and family members, possibly reflecting the need for improved training on cultural competency and on the hazards of using family members as interpreters. Physicians in states with higher proportions of persons with LEP report lower use of a professional interpreter; similar reported use of professional interpreters among respondents who report use of bilingual physicians or staff members compared with respondents reporting no use suggests that this finding cannot be explained by higher use of bilingual personnel in such states. Our results may suggest an inadequate supply of professional interpreters to meet the increasing demand in states with a high LEP population.
It is encouraging that third-party reimbursement for language services is associated with greater use of professional interpreters, suggesting an important intervention that could increase the use of appropriate language services. States with public third-party reimbursement for language services may have a paucity of bilingual providers and staff; according to the 2000 US Census, 8 of 10 states providing Medicaid and SCHIP reimbursement for interpreter services (at the time of the survey) have an LEP population proportion that is less than or equal to the median LEP population proportion for the United States, with only Hawaii having both a high proportion of LEP patients and third-party language services reimbursement. Although there was a significant association between third-party reimbursement and professional interpreter use in adjusted analysis, reported use of professional interpreters was <60% even in states with third-party reimbursement. Levels of and mechanisms for payment vary by state,25 and we do not know how often practices bill for these services, making it difficult to fully assess the effect of third-party reimbursement. We also have no information regarding private third-party reimbursement, but patients with LEP disproportionately have public insurance.30
The distribution of LEP patients is uneven across pediatricians, practices, and states and offers starting points to improve the delivery of language services. The higher proportion of LEP patients reported by female and younger pediatricians highlights the importance of education on proper language services use during the early years of physician training. Despite an emphasis on providing culturally effective care in residency training,31 a recent study found that 22% of pediatric residents report being very or somewhat unprepared to treat patients with LEP.32 Misperceptions exist among practicing physicians that family members provide sufficient interpretation services,20 highlighting the need for provider education and practice policies that ensure uniform delivery of effective language services. Successful strategies already being used by small practice providers include determining language needs at initial contact, use of trained bilingual staff, extensive use of written translations available either through community resources or the Internet, and use of telephone language lines.19,33 Finally, an increase in interpreter use has been reported among physicians with previous training in interpreter use.34
The overall cost of providing language services may be relatively modest. A federal report by the Office of Management and Budget estimated that the cost of interpreter services for LEP persons, when averaged over all inpatient, outpatient, and dental visits, would be an average of $4.04 more per visit, equivalent to 0.5% of the average cost per health care visit.35 Latinos and Asians incur annual medical costs 20% to 60% less than the mean of non-Latino whites, and the overall cost of providing interpreter services is less than existing cost disparities.30 However, health care providers in many states currently assume the burden of the cost of language services, potentially creating disincentives for providing language services. Outpatient providers, in particular, bear a disproportionate share of the cost. Third-party reimbursement would alleviate the burden30 and may increase preventative medical services, which could further lower the overall cost of interpreter services.13
Certain study limitations should be noted. First, the response rate was 58%; however, this is similar to that of other large national physician postal surveys,36 and analysis of response rates in previous AAP surveys reveals minimal nonresponse bias.29 Respondents included for analysis tended to be younger relative to national AAP membership and to have a higher proportion reporting Latino ethnicity compared with overall Periodic Survey respondents, which could overestimate LEP patient prevalence and frequency of language service provision. Second, survey data were self-reported and subject to recall bias. Providers may vary in how they estimate the percentage of their patients with LEP, and reported use of a language service may not reflect actual frequency of use; for example, it is unlikely that the 52% of respondents reporting use of a bilingual physician provide a bilingual physician for each encounter with a patient with LEP. Third, no additional definitions for methods of communication were provided for survey respondents, possibly resulting in respondents having differing definitions of what may compose a "professional" interpreter or "written materials in the primary language." Finally, our study does not address the quality or effectiveness of language services that are provided, and the quality of the training of individual interpreters is not known. Health care interpreter standards were issued recently by the National Council on Interpreting in Health Care37; how interpreter standards will impact the supply and costs of interpreter services is unknown.
| CONCLUSIONS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Address correspondence to Dennis Z. Kuo, MD, MHS, Division of General Pediatrics and Adolescent Medicine, Johns Hopkins University School of Medicine, 200 N Wolfe St, Baltimore, MD 21287. E-mail: dkuo5{at}jhmi.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
The views expressed in this article are those of the authors and do not represent policies of the American Academy of Pediatrics.
This work was presented at the annual meeting of the Pediatric Academic Societies; April 29, 2006; San Francisco, CA; and at the AcademyHealth Annual Research Meeting; June 27, 2006; Seattle, WA.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
Y. Schenker, B. Lo, K. M. Ettinger, and A. Fernandez Navigating Language Barriers under Difficult Circumstances Ann Intern Med, August 19, 2008; 149(4): 264 - 269. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. X. Vo and M. J. Park Racial/Ethnic Disparities and Culturally Competent Health Care Among Youth and Young Men American Journal of Men's Health, June 1, 2008; 2(2): 192 - 205. [Abstract] [PDF] |
||||
Read all P3Rs
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||