Communicating across language barriers is a challenge for clinicians all across the United States. The American Community Survey estimated that about 20 % of the adult US population speaks a language other than English at home, with almost half of these speaking English less than very well.1 Spanish and Chinese are the most common preferred languages of limited English proficient (LEP) adults, but hundreds of additional languages are in use throughout the U.S.

The intersection of race and ethnicity, social class, and language skills in determining effective communication between patients and clinicians remains difficult to disentangle. However, we do know that patients with LEP are less likely than English-speaking patients to achieve rapport, or to receive empathy, information, and encouragement to participate in decision-making.2,3 LEP status has also been associated with fewer physician visits, receipt of fewer preventive services, less satisfaction with clinical care and worse control of diabetes.4,5 Such shortcomings may be on a critical path toward increasing health disparities. In turn, improving these relational aspects of the patient-clinician interaction may decrease health disparities for LEP patients, given that good patient–physician relationships have been linked to positive patient outcomes, including medication adherence, comprehension and satisfaction, lower blood sugar among patients with diabetes, and better self-rated health.2,5,6

Clinicians rely on the use of interpreters—sometimes professional, often not—to help them and their LEP patients navigate the medical visit and subsequent care. Federal policies are clear and, based on an August 2000 Executive Order by President Clinton under Title VI policy guidance to avoid discrimination against LEP persons, dictate that “health and social service providers must take adequate steps to ensure that such persons receive the language assistance necessary to afford them meaningful access…free of charge.7 Furthermore, an Institute of Medicine Report in 2010 recommended standardized collection of data on LEP status among patients, in order to provide adequate language services to their patients.8 However, health systems and clinicians are often unable to meet the demand for professional interpretation due to cost, availability of trained interpreters or lack of information about why this is important; there are no accepted standards on defining LEP status or need for interpreter services. Patients often rely on their own limited English language abilities or on family members or friends acting as ad hoc interpreters when professional interpreters are not available. On the other hand, English-speaking clinicians may use their own limited foreign language skills to bridge the gap if possible, and as a result, the quality of clinical communication may be compromised.9

Empirical scientific research on LEP patients’ utilization of health services and their interactions with clinicians has emerged in the past 20 years. Most studies have been conducted with LEP Spanish speaking Latinos, some Asian national origin groups, and recent immigrants or refugees with unique social and political circumstances. In this issue of JGIM, Flynn and colleagues10 evaluated how the use of interpreter services, as coded in the electronic medical record, affected primary care health utilization and diagnoses over 1 year at the Mayo Clinic in Rochester, Minnesota. In this retrospective cohort study, patients requiring an interpreter were more likely to use primary care and emergency care, were disproportionately represented among the high utilizers, and were less likely to have mental health-related diagnoses, despite similarly high rates of somatic symptom diagnoses. The 1,566 patients requiring interpreter services were only 2.5 % of all patients in the sample, and by far the most common language was Somali. (Spanish was second, and an additional 32 languages were used.) Such linguistic and ethnic heterogeneity may not be unusual in areas with limited immigration; however, the findings from the Mayo Clinic experience may not be generalizable to the majority of LEP patients needing interpreters, as Somali immigrants may have specific social and cultural factors that drive utilization, and somatic symptoms leading to communication barriers that would be difficult to bridge even with the best professional interpreters at every encounter. Regardless, one possible implication of this study is that poor communication resulting from language barriers leads to fewer mental health diagnoses and potentially more primary care and emergency department visits.

In defining the research and policy questions to be addressed in the care of LEP patients, it is critical to properly identify constructs. The laboratory to address empirical research in this context is the patient–clinician encounter, whether it is in the ambulatory setting or hospital. Secondary data analyses from electronic medical records are valuable for identifying patterns of care and outcomes and to help identify questions, but cannot provide the level of detail needed to understand what really goes on in the interpreter-mediated clinical encounter. While one would assume that language concordance (i.e. clinician speaks the patient’s primary language fluently) is the “gold-standard” in communicating with LEP patients, empirical data need to be generated to support this hypothesis in order to consider policy changes in workforce development and compensation. Research on use of interpreters should focus on the value of professional versus ad hoc interpreters, a distinction that was not clearly made in the Mayo Clinic study, and the mode of providing interpretation as in person or remote via telephone or videoconferencing. Finally, research examining the effects of language concordance and type of interpretation on care of LEP patients should systematically measure outcomes that matter to patients and clinicians.

This study does highlight three important policy issues and research questions that need additional work in order to make progress in this area. First, health systems must standardize data collection on LEP status, collect this routinely and incorporate it into the electronic record. The US Census question (How well do you speak English? Very well, well, not well, not at all), plus a second item asking preferred language for health care, was associated with improved outcomes of effective communication by Spanish and Chinese speaking patients in both ambulatory and inpatient settings.11 The Institute of Medicine (IOM) Report endorsed these two items for use in health care, and we should all be lobbying our systems to incorporate these as standard questions on registering a patient.8

Second, if an interpreter is needed, a trained professional interpreter should be routinely used. Use of an ad hoc interpreter may be better than no interpreter at all, but in comparison to using professional interpreters, clinicians who use ad hoc or no interpreters have to accept delivering lower quality of care to these patients and recognize that this may lead to more errors in communication and worse clinical outcomes.12 Recent data have shown that professional interpretation through remote telephone or video conferencing modes is as effective in most instances as in-person interpretation in facilitating quality communication between patient and clinician.13 Health systems need to be persuaded that the improved quality of care and patient satisfaction would justify the increased costs of making professional interpretation available. Examples of medical centers that have invested successfully in language services include San Francisco General Hospital, The Children’s Hospital of Philadelphia and Bellevue Hospital Center in New York City. Furthermore, it is common sense that in a clinical encounter with an interpreter where the conversation is repeated, in order to have an equivalent amount of information exchange and rapport building as in an English language encounter, it will take twice as much time for the visit or alternatively, half as much communication will actually happen. Thus, in order to avoid the scenario of lower quality, modifications in work relative value unit (RVU) credit or compensation may also help improve care.

Third, the optimal way for clinicians to manage LEP patients is to have language concordant care. Among monolingual Spanish speaking patients, language concordance has been found to improve patient measures of health outlook, pain, and anxiety,6 lower the rate of poorly controlled diabetes,5 and lead to less confusion, frustration and perceived bias in health care encounters.14 In a broader survey of LEP patients speaking 11 different languages, those with language discordant clinicians were more likely to report problems understanding a medical situation, trouble understanding labels, and bad medication reactions.15 Language concordance can attenuate the communication gaps created by limited health literacy, and overall will lead to more patient-centered care.16 Thus, whenever possible, clinicians who are fully fluent in a non-English language should be in a position to care for these LEP patients. There is a need for nationally accepted standards to evaluate and certify non-English language fluency of clinicians to care for LEP patients.9

We are realistic and understand that these three goals of standardized data collection, universal access to professional interpretation, and moving towards language concordant care will require structural change, resources and time. General training of all clinicians in the social and political history of immigrants and their home country, and in the proper use of professional interpreters, can help address these inequities in the meantime. LEP patients will be in our clinics and hospitals for many years and more empirical research is needed to understand mechanisms of optimal care of this disadvantaged population.