When nurses double as interpreters: a study of Spanish-speaking patients in a US primary care setting

https://doi.org/10.1016/S0277-9536(00)00234-3Get rights and content

Abstract

The United States is experiencing one of its largest migratory waves, so health providers are caring for many patients who do not speak English. Bilingual nurses who have not been trained as medical interpreters frequently translate for these patients. To examine the accuracy of medical interpretations provided by nurses untrained in medical interpreting, we conducted a qualitative, cross-sectional study at a multi-ethnic, university-affiliated primary care clinic in southern California. Medical encounters of 21 Spanish-speaking patients who required a nurse-interpreter to communicate with their physicians were videorecorded. Encounters were transcribed by blinded research assistants. Transcriptions were translated and analyzed for types of interpretive errors and processes that promoted the occurrence of errors. In successful interpretations where misunderstandings did not develop, nurse-interpreters translated the patient's comments as completely as could be remembered and allowed the physician to extract the clinically-relevant information. In such cases, the physician periodically summarized his/her perception of the problem for back-translation and verification or correction by the patient. On the other hand, approximately one-half of the encounters had serious miscommunication problems that affected either the physician's understanding of the symptoms or the credibility of the patient's concerns. Interpretations that contained errors that led to misunderstandings occurred in the presence of one or more of the following processes: (1) physicians resisted reconceptualizing the problem when contradictory information was mentioned; (2) nurses provided information congruent with clinical expectations but not congruent with patients’ comments; (3) nurses slanted the interpretations, reflecting unfavorably on patients and undermining patients’ credibility; and (4) patients explained the symptoms using a cultural metaphor that was not compatible with Western clinical nosology. We conclude that errors occur frequently in interpretations provided by untrained nurse-interpreters during cross-language encounters, so complaints of many non-English-speaking patients may be misunderstood by their physicians.

Introduction

Immigration now accounts for 37 percent of the national population growth in the United States (Horner, 1995). Many of these immigrants have suffered malnutrition, poverty, torture and disease during their migration with resulting physical and mental consequences (Castillo, Waitzkin, Ramirez & Escobar, 1995). Additionally, immigrants come from every continent as opposed to the primarily European emigration at the turn of the century, so they represent diverse languages and cultures. According to the last census, more than 150 languages are spoken in the United States within almost 300 racial and ethnic groups (Horner, 1995). Because of the size of the surge in immigration, many of the newcomers are able to move into established ethnic communities, which delays their acculturation and acquisition of English (Szapocznik, 1994). The largest group of immigrants comes from Mexico, and this group accounts for 10 percent of the total US population. Frequently, many Mexican immigrants maintain homes and family ties in Mexico while they pursue economic advantages by working in the US (Haffner, 1987). Consequently, they may avoid acculturation, and family members, anticipating a temporary stay in the US, may choose not to learn English.

When immigrants need medical services, they usually must visit caregivers of different cultural backgrounds. The caregivers may not understand the implications that particular symptoms hold for patients because the perception and interpretation of somatic sensations are frequently defined by cultural idioms (Eisenberg, 1977; Kleinman, Eisenberg & Good, 1978; Kleinman, 1980; Pachter, 1994). Furthermore, the burden of a lack of a common language hinders patients’ ability to explain their concerns or perceptions of illness to the physicians because the emphasis moves toward trying to convey the clinical symptoms. Almost all communication between physicians and non-English-speaking patients is concerned with symptomatology to the exclusion of feelings, causes, or patient questions (Rivadeneyra, Elderkin-Thompson, Silver & Waitzkin, 2000). The clinical focus probably explains, in part, why non-English-speaking patients who use an interpreter, or feel that they need one, rate their providers as less friendly and less respectful than do patients without a language or cultural barrier (Baker, Hayes & Fortier, 1998; Cooper-Patrick et al., 1999). As a result of the impersonal interaction, non-English-speaking patients are at risk for not developing the trust in their physicians that facilitates cooperation with suggested treatment regimens (Marcos, Urcoyo, Kesselman & Alpert, 1973; Quesada, 1976; Marcos, 1979; Triandis, Marin, Lisansky & Betancourt, 1984; Perez-Stable, 1987; Bertha, 1992).

Immigrants rate language and cultural differences as their biggest barriers to receiving health care. Parents with little or no English skill, who tried to secure care for their children at a Latino clinic, cited the difference in language as the cause of misdiagnoses, poor medical care, and inappropriate medications and/or hospitalizations of their children (Flores, Abreu, Olivar & Kastner, 1998). The expected problems of poverty, lack of insurance, transportation, and long waiting times were rated after language. The health care system increasingly depends on patients to make critical informed decisions regarding their health care, e.g., the use of life support systems or the choice of therapeutic approaches for cancer or diabetes (Kent, 1996). However, distinct differences in therapeutic choices have been reported even across racial and ethnic groups that use the same language (Blackhall et al., 1999). When language and cultural barriers are present, the probability is further increased that patients might misunderstand or miscommunicate their perceptions of risks and benefits of treatment options.

Other factors affect patients’ ability to implement therapeutic recommendations from cross-language encounters. Patients receive less information about the therapeutic regimen, understand less of the medication instructions (Shapiro & Saltzer, 1981), are less likely to keep subsequent appointments, and are more likely to make emergency room visits than are patients in same-language encounters (Manson, 1988). Non-English speakers are also less likely to receive preventive services (Woloshin, Schwartz, Katz & Welch, 1997). Among Spanish-speaking patients, those seen by physicians with even limited Spanish ability feel that they understand the disease and treatment better, have better recall, and ask more questions than Latinos seen by non-Spanish-speaking physicians (Erzinger, 1987; Seijo, Gomez & Freidenberg, 1991; Baker, Parker, Williams, Coates & Pitkin, 1996).

In addition to the problems associated with conveying information about diagnoses or medications, a language barrier impedes the effective use of information for encouraging compliance. Some patients prefer to be well-informed so that they can monitor their own progress, while others prefer to avoid medical details. Giving too much information to a person who prefers to avoid the medical detail, or not giving enough information to a person who wants to be informed, can increase the anxiety level of the patient enough to delay recovery (Suls & Fletcher, 1985). In addition, reducing the patient's anxiety level by use of the appropriate level of information encourages calm decision-making if choices between alternatives must be made (Kerrigan et al., 1993; Kent, 1996). Similarly, some patients prefer to feel in control of their medical care, while others prefer a more dependent role and trust the physician to make decisions (Thompson, 1981). However, even if a physician discerns the appropriate amount of information and control desired by an immigrant patient regarding treatment options, the information might not be understandable to the patient if the symptoms are perceived within a cultural idiom.

A commonly held idiom among Latinos as well as many other immigrants is the humoral concept of illness (Bastein, 1987; Messer, 1987; Nichter, 1987; Tedlock, 1987; Weiss et al., 1988). Illness is perceived as a hot–cold imbalance within a fluid of the body, such as blood, phlegm, or semen (Bastein, 1987). When illness is associated with an imbalance the imbalance perceived as a reaction to an unfavorable situation or an unusual action by another person (Weiss et al., 1988). Illnesses, medicines, and people themselves may be located anywhere on the hot–cold continuum, and recovery from illness requires the appropriate balancing of all elements to achieve a “temperate” outcome (Messer, 1987). Solving the problem, then, requires addressing both the symptoms and the context in which the symptoms developed. Western medicine, which relies heavily on technology, often treats symptoms independent of their context. Although many immigrants have a complex perception of caregiving that incorporates popular, folk and professional medical options (Tedlock, 1987), immigrants may still have difficulty understanding the physician's logic and be hesitant about following a regimen perceived as addressing only a portion of the problem.

Physicians who are not proficient in their patients’ languages must use interpreters. The problems when using ad hoc interpreters are well known. Ebden and colleagues found that between 23 and 52 percent of physicians’ questions were either misinterpreted or not interpreted at all by non-trained, ad hoc, staff interpreters (Ebden, Carey, Bhatt & Harrison, 1988). Other researchers have corroborated the frequency of editing errors and omissions by non-trained interpreters (Marcos, 1979; Diaz-Duque, 1982; Ebden et al., 1988; Serrano, 1989). Misdiagnoses among psychiatrists have also been attributed to cross-language difficulties (Alpert, Kesselman, Marcos & Urcuyo, 1973; Marcos et al., 1973; Marcos, 1979; Seijo et al., 1991).

It has been recommended that physicians utilize medical interpreters who can avoid the problems inherent in the use of a family member or ad hoc interpreters (Erzinger, 1987; Hardt, 1991; Seijo et al., 1991). Professional interpreters allow patients to speak freely, and they provide patients the opportunity to explain problems without modifications by family members who can bring their own objectives and perceptions to the encounter. However, for economic reasons, many physicians and health care facilities employ bilingual nurses rather than hire trained medical interpreters. Nurses understand the need for a flexible blending of questioning, listening, eliminating irrelevant information, educating, and collecting of facts for hypothesis formulation — the necessary steps in medical inquiry. Patients often mention important material casually or indirectly, and only a medically trained person would recognize its significance (Putsch, 1985). Additionally, nurses are likely to understand the physician's medical assumptions and rationale for repeated questions, so they can elicit the type of information needed for clinical decision-making. Patients may not view nurses’ presence as invasive in the intimate context of a medical encounter as they might view the presence of other interpreters.

However, the extent to which nurses’ professional standing mitigates their willingness to identify and explain points of confusion or medical conflicts to the practitioners, to whom they are subordinate, is unknown. The relationship between the patient and nurse-interpreter is also vulnerable to the class, gender, age and educational biases of their shared ethnic culture. Nurses may have social positions that they perceive as superior to the immigrant patients, and this perception may influence their interpretations of patients’ narratives. Interpretations are framed by the beliefs and assumptions of the interpreter, so how an interpreter views a patient personally may influence how he or she understands and interprets the patient's comments (Poma, 1983; Seijo et al., 1991).

Yet the use of bilingual nurses as interpreters for non-English-speaking patients in lieu of trained medical interpreters has received little critical attention, and there is almost no empirical research literature of which we are aware that examines the accuracy of these interpretations. To examine the effectiveness of nurses who had excellent bilingual skills but no prior translation training, we qualitatively analyzed interpretations provided by nurses during cross-language medical encounters. The research focus was on the accuracy of the interpretations. If inaccuracies were found, we were interested in the nature of the inaccuracies and the behavioral processes that promoted the occurrence of the errors.

Section snippets

Participants

Patients between ages 18 and 66, seeking first-time episodic care at a walk-in primary care clinic sponsored by a Southern California university, were approached by bilingual/bicultural research assistants to participate in a research project on mental health issues in primary care funded by the National Institute of Mental Health (NIMH). The clinic serves a low socioeconomic area with a large number of immigrants from Mexico and Central America. Over a 6-month period, patients who consented to

Results

The 21 encounters divided evenly between complicated and uncomplicated cases. Ten contained minor interpretive errors — usually editing changes — that did not become clinically significant. In these successful interpretations, the symptoms were generally conveyed accurately and were understood by the physicians. Medical chart summaries of patients’ symptoms were accurate and subsequent notations within the following six-month period indicated that follow-up procedures were completed and

Discussion

Physicians who do not have trained medical interpreters available for their non-English-speaking patients frequently turn to bilingual nurses to provide interpreting services. This study examined the accuracy of the interpretations provided by nurses who were untrained in medical interpreting but who possessed excellent bilingual skills. Twenty-one encounters were examined, and approximately two-thirds of the uncomplicated cases did not contain errors that jeopardized the diagnosis and

Conclusions

The use of nurses as medical interpreters has become common as the number of non-English-speaking patients increases. During the interpretations, physicians can improve the quality by taking time to articulate points on which they remain unsure, proceeding slowly and systematically with their queries, reflecting information back to the interpreter for back-translation and patient verification, and remaining flexible about the possibility of reformulating their working hypotheses when

Acknowledgements

The authors would like to thank Dr. Peter Kaneshige for reviewing medical charts, providing assistance in videotaping encounters, and making suggestions regarding communication errors. We appreciate the efforts of Fanay Loiselle, our senior translator who reviewed all videotapes and transcripts for accuracy, as well as the efforts of our bilingual research assistants who provided the interpretations of the transcripts. We could not have done the study without the cooperation of the medical and

References (60)

  • D.W Baker et al.

    Interpreter use and satisfaction with interpersonal aspects of care for Spanish-speaking patients

    Medical Care

    (1998)
  • D.W Baker et al.

    Use and effectiveness of interpreters in an emergency department

    Journal of the American Medical Association

    (1996)
  • J Bastein

    Cross-cultural communication between doctors and peasants in Bolivia

    Social Science and Medicine

    (1987)
  • H.B Beckman et al.

    The effect of physician behavior on the collection of data

    Annals of Internal Medicine

    (1984)
  • M.O Bertha

    Modesty, sexuality, and breast health in Chinese-American women

    Western Journal of Medicine

    (1992)
  • D Buchwald et al.

    Caring for patients in a multicultural society

    Patient Care

    (1994)
  • R Castillo et al.

    Somatization and primary care, with a focus on immigrants and refugees

    Archives of Family Medicine

    (1995)
  • L Cooper-Patrick et al.

    Race, gender, and partnership in the patient–physician relationship

    Journal of the American Medical Association

    (1999)
  • T.W Day

    Cross-cultural medicine at home

    Minnesota Medicine

    (1992)
  • O.F Diaz-Duque

    Overcoming the language barrierAdvice from an interpreter

    American Journal of Nursing

    (1982)
  • D Edwards

    Script formulationsA study of event descriptions in conversation

    Journal of Language and Social Psychology

    (1994)
  • Eisenberg, L. (1977). The search for care. Daedalus Winter,...
  • S.L Erzinger

    Communication between Spanish-speaking patients and their doctors in medical encounters

    Culture, Medicine & Psychiatry

    (1987)
  • G Flores et al.

    Access barriers to health care for Latino children

    Archives of Pediatrics and Adolescent Medicine

    (1998)
  • R.M Frankel

    From sentence to sequenceUnderstanding the medical encounter through micro-interactional analysis

    Discourse Processes

    (1984)
  • L Haffner

    Translation is not enoughInterpreting in a medical setting

    Western Journal of Medicine

    (1987)
  • E.J Hardt

    The bilingual medical interview I

    (1991)
  • J.C Hornberger et al.

    Eliminating language barriers for non-English-speaking patients

    Medical Care

    (1996)
  • C Horner

    Reclaiming the vision

    The Brookings Review

    (1995)
  • D Kerrigan et al.

    Who's afraid of informed consent?

    British Medical Journal

    (1993)
  • Cited by (241)

    View all citing articles on Scopus
    View full text