Health care service delivery and clinical research need to accommodate the specialized needs of diverse populations. Unfortunately, the degree to which the health care system and research institutions respond to this challenge is driven by funding incentives and legal mandates, rather than by consumers' needs. Recently, the increased focus on ethnic and socioeconomic health disparities has resulted in policy changes that direct more attention to minority health issues. For example, an Executive Order on “Improving Access to Services for Persons with Limited English Proficiency” issued by the White House on August 11, 2000, resulted in written guidelines being provided, by the U.S. Department of Health and Human Services, to clinicians to ensure language assistance for persons with limited English skills.1 In fact, even the term “minority health” has been challenged since we are quickly becoming a minority-majority country.

The two articles appearing in this month's issue, by Jacobs et al.,2 and by Carrete et al.,3 highlight two issues that have become critically important due to recent health policy initiatives. Jacobs et al. emphasize the need to address the service delivery requirements of limited English proficient (LEP) patients in many areas of the U.S. Carrete et al. offer an excellent example of the need to validate self-reported measures of health for ethnically diverse populations.

The Jacobs study compared utilization of preventive and clinical services between a group of patients receiving professional interpreter services and a group that did not (comparison group). Compared to the comparison group, utilization of clinical services increased more in the group of patients with interpreter services, for office visits, number of prescriptions written, prescriptions filled, and rectal examinations. Although we have insufficient information to determine the medical appropriateness of the increases in utilization, or to determine whether more services of this type lead to better health outcomes, provision of professional interpreter services appeared to improve access to clinical services for non-English speaking patients. However, as stated by the authors, improvement in access to care does not imply improved quality or outcomes of care. Therefore more studies are needed.

Although tremendous progress has been made in the field of medicine and clinical practice, it remains incumbent upon a clinician to convey the relevance of such advances to individual patients. To do so requires the clinician to convey information and emotional support in ways that are understood and acceptable to the patient. Language is a fundamental vehicle for doing so. Out of necessity, especially when having a translator present would require long waiting times, clinicians will rely on their limited foreign language skills to “get by” during medical encounters.4 These conditions occur often in the realm of daily practice. Little is known about the processes and outcomes that occur during these visits. One recent study showed that among monolingual Spanish speaking patients, patient-centered care was enhanced by language-concordant interactions between patients and doctor, when compared to interpreter-mediated visits.5 In fact, our own work found that Spanish-speaking patients with a language-concordant physician had significantly better scores on medical outcomes study measures than their discordant counterparts.6 We need additional research focusing on the quality of information that LEP patients receive about their conditions and treatments, the level of satisfaction the patients have with their care, and the rates of adherence with recommendations for screening tests, behavior change and medications. Clinicians who speak the language are most likely to provide optimal care compared to interpreters, but it is unrealistic to expect these skills to be widely available in the physician workforce. However, it is intuitive that interpreter services are helpful in facilitating access to care as demonstrated in the Jacobs study.

Until recently, the language needs of immigrant populations were largely ignored by the U.S. health care system. The piecemeal nature of its response is evidenced by the lack of standardized language assessment tools, training methods, interpretation standards, quality assurance programs and continuing education for professional translation service providers.7 Although the U.S. census collects data on linguistically isolated individuals, data on the cultural and linguistic background of patient populations is often lacking. To fully understand the health needs of LEP patients, a concerted effort needs to be made to collect data systematically and consistently on linguistic and cultural factors, and to link such data to data on health care processes and outcomes. The first generation of studies of the impact of language proficiency and interpreters on health care has focused on access issues, as did the Jacobs study. Future studies need to expand our efforts to assess the quality and nature of culturally competent care on the process and outcomes of health care. Linguistic appropriateness of health care services is only one component of a comprehensive program to serve the health care needs of ethnically and linguistically diverse populations.

Carrete et al.3 examine the psychometric properties of a telephone administered Spanish version of the Geriatric Depression Scale (GDS-T) among older Latinos. The authors attest to the clinical importance of such an instrument to help detect and plan timely interventions in a group of patients that are at high risk of depression who might easily remain undetected. In this study, the GDS-T demonstrated acceptable specificity, sensitivity, internal consistency reliability, and preliminary construct validity using the DSM IV criteria for depression as the gold standard. The authors used rigorous translation protocols to develop the translation of the instrument, and traditional psychometric methods to evaluate its measurement properties in a random sample of 282 older Latino members of a university affiliated health maintenance organization (HMO) in Buenos Aires, Argentina.

The Carrete study involves translation of an existing instrument for use in another language and culture. Rather than assume that the “shelf instrument” was valid and reliable in their ethically diverse sample, the authors assessed the conceptual, linguistic, and psychometric suitability of the instrument for use in a diverse group for which there is no evidence that such an assumption can be safely made. In international studies of health-related quality-of-life, instruments often undergo rigorous scrutiny with the assumption that some adaptation of the instrument will be necessary. In the United States, a more widely used approach is to take an existing instrument and apply it in a new population with the assumption that it will measure the same constructs in a similar way as it did in the population in which it was developed. Thus, the methodological importance of assessing the conceptual and psychometric adequacy (within a group) and equivalence across groups is most often overlooked.8

Recent policy and research initiatives have emphasized the elimination of health disparities indicating that research studies in diverse populations will increase.8 To address potential causes of, and interventions for reducing, health disparities involves comparing groups on mean levels of health. Most health disparities have been linked to differences among ethnic or socioeconomic groups. Comparing ethnic, socioeconomic or other groups rests on the assumption that self-report measures of health are understood and responded to in a similar manner across groups. Given that most self-report instruments of health are products of U.S. or British cultures, these instruments are characterized by a “cultural hegemony” that assumes that all groups interpret the items as do middle class Whites.9 The increasing percentage of immigrant groups, admixture, and language diversity of our society challenge the defensibility of this approach. Study findings that rely on comparisons between groups on self-rated health measures may be invalid, meaning that observed differences in scores may actually reflect bias or measurement error and not actual differences in levels of health.

As we operate in this world, we carry with us socially and individually derived representations of ourselves, of the way we view others, and of our preferred modes for interactions. In many instances, these cognitive representations are barely acknowledged within our consciousness and much less in society, yet they have a tremendous amount of influence on our behavior and the ways in which we interpret our environment. When we increase the diversity of our environments, we create challenges to our predetermined ways of thinking that can result in conflict, negotiation, self-reflection and potentially growth. These established notions of the ways in which our world functions affect our socially constructed reality of research and clinical practice. Research that is imbedded in mainstream American culture will reflect and be dominated by that culture. When we use self-report instruments that have been developed for largely White, middle class respondents, we must reflect on their usefulness, appropriateness and salience for ethically diverse groups. If we assume that an instrument performs equally well when applied to other socioeconomic or cultural groups without evidence that this is the case, we run the risk of conducting research based on invalid assumptions.

It should be readily accepted by most people that health care services and measures of health that are not linguistically and culturally appropriate are inherently inferior to those that are appropriate linguistically and culturally, yet we have little evidence to support this tenet. Measures of health and health care services that cannot be understood by, or are not relevant for, particular groups represent substandard quality and increase the risk of substandard outcomes of care, or substandard research methods. For both the provision of clinical services and the conduct of research in diverse populations, addressing issues of cultural and linguistic suitability can only improve our ability to address health disparities and develop service delivery models that constitute best practices for these populations. —Anna Nápoles-Springer, PhD and Eliseo J. Pérez-Stable, MD,Division of General Internal Medicine, Department of Medicine, Medical Effectiveness Research Center for Diverse Populations, Center for Aging in Diverse Communities, University of California, San Francisco, Calif.


  1. Top of page
  • 1
    U.S. Department of Health and Human Services Office for Civil Rights. Washingon, DC: Federal Register; 2000.
  • 2
    Jacobs EA, Lauderdale DS, Meltzer D, Shorey JM, Levinson W, Thisted RA. Impact of interpreter services on delivery of health care to limited–English-proficient patients. J Gen Intern Med. 2001;16: 46874.
  • 3
    Carrete P, Augustovski F, Gimpel N, Fernandez S, et al. Validation of a telephone-administered geriatric depression scale in a Hispanic elderly population. J Gen Intern Med. 2001;16: 44650.
  • 4
    Hornberger J, Itakura H, Wilson SR. Bridging language and cultural barriers between physicians and patients. Public Health Rep. 1997;112: 4107.
  • 5
    Rivadeneyra R, Elderkin-Thompson V, Silver RC, Waitzkin H. Patient centeredness in medical encounters requiring an interpreter. Am J Med. 2000;108: 4704.DOI: 10.1016/s0002-9343(99)00445-3
  • 6
    Pérez-Stable EJ, Nápoles-Springer A, Miramontes JM. The effects of ethnicity and language on medical outcomes of patients with hypertension or diabetes. Med Care. 1997;35: 12129.
  • 7
    Riddick S. Improving access for limited English-speaking consumers: A review of strategies in health care settings. J Health Care Poor Underserved. 1998;9(suppl:)4061.
  • 8
    Stewart AL & Nápoles-Springer A. Health-related quality-of-life assessments in diverse population groups in the United States. Med Care. 2000;38(9 suppl II):10224.
  • 9
    Guyatt GH. The philosophy of health-related quality of life translation. Qual Life Res. 1993;2: 4615.