Introduction
Language barriers (LBs) hinder health care delivery in settings with culturally diverse populations. LBs decrease access to primary and preventive care [
1], patient comprehension, and patient satisfaction [
2]. The implications of LBs may be greatest in the emergency department (ED). Often considered the “safety net” for the health care system, EDs are used disproportionately by non-English speakers facing barriers to conventional forms of primary care within the health care system [
3‐
8]. ED care is characterized by brief encounters between health care staff and patients [
9] and is increasingly pressured by ED overcrowding [
10], increased patient volumes [
11], and relative understaffing, all of which hamper optimal physician-patient communication.
Non-English-speaking patients have cited language as the single greatest barrier to health care in the ED [
12]. LBs have been reported to increase ED visit times [
13], impede diagnoses [
14], and possibly lead to inappropriate medication and hospitalization [
12]. They may decrease patient compliance upon discharge from the ED and lower comprehension levels of discharge instructions [
15,
16]. Interventions to accommodate non-English-speaking patients have been shown to shorten length of stay and reduced non-urgent visits [
17].
The handful of ED interventions employed to bridge LBs all have limitations. Cost constraints often limit the availability of in-house language interpreters to weekday business hours. Patients are often less satisfied with such intermediaries than when speaking directly to the physician, and training physicians in communication skills has been advised [
18]. Telephone translation services are effective, but introduce a third party, and can be costly. Websites and computer programs that translate common medical vocabulary are frequently cumbersome to use [
19,
20]. Hospital employees and family members serving as unskilled translators compromise patient confidentiality and raise potential for medical misunderstanding from using an unskilled translator, who may be reluctant to relay bad news or embarrassing questions to the patient [
21].
We designed and pilot tested a brief language education and translation aid for physicians in the ED, as a practical and economical adjunct to currently available strategies. The primary objective of this pilot study was to assess whether the ability of the ED clinician to speak limited Portuguese could enhance emergency physician (EP) ability to obtain a directed medical history among non-English-speaking Portuguese patients. A secondary objective was to investigate clinician and patient perceptions of this intervention. This study represents the first of its kind within Canada, and the first of its kind in Portuguese in North America.
Methodology and methods
The study protocol was approved by the hospital Research Ethics Board. All consent forms for patients were available in Portuguese, and a Portuguese-speaking research assistant was available to assist if the patient was illiterate.
The protocol consisted of (1) development of a paper-based, problem-oriented language translation aid and companion audio compact disk (CD) demonstrating proper pronunciation of the text in Portuguese, (2) distribution of the instrument and audio CD to ED clinicians for home practice, (3) piloting the instrument at bedside, with post-application survey of patient and clinician, and (4) post-pilot evaluation survey to assess EP impressions of the instrument.
(B) Audio guide: The purpose of the audio guide was to reinforce proper Portuguese pronunciation. Audio clips were created for each Portuguese phrase by a native Portuguese-speaking ED nurse (HL). The CD audio guide graphically reproduced the pocket brochure, with “icons” alongside the text that would play the spoken Portuguese text upon a mouse click.
The CD was placed in the ED mailbox of each EP (n = 30) and over a 3-week period, EPs were sent regular weekly e-mail reminders, along with an electronic copy of the paper instrument, describing the upcoming pilot and encouraging review of the instrument and audio guide. Though clinicians were encouraged to orient themselves to the CD contents, failure to do so would not preclude study participation.
During data collection shifts, a trained Portuguese-speaking researcher (AH) recruited eligible patients and their clinicians for informed consent and enrollment in the study. Participating clinicians were instructed to use the instrument for obtaining a history from the patient. They could select any questions they chose from the printed instrument, but were to ask each question first in English and then in Portuguese. (Each patient thus served as his or her own control.) The observer scored the response to each question (first in English, then in Portuguese) as to whether it yielded a response that was appropriate for the question asked and comprehensible to the non-Portuguese-speaking clinician. (For example, for the question “Are you having chest pain?” an indication in the positive or affirmative would be deemed appropriate; a response that did not speak directly to the question, i.e., “I’m short of breath” or “It’s okay” would not be regarded as appropriate). Physicians were permitted to terminate use of the instrument at their discretion. After completion of history taking, both patient and physician were interviewed in a standardized manner regarding the intervention’s effectiveness in assisting with communication.
Results
None of the clinicians had reviewed the audio CD prior to using the language translation tool at the bedside. However, half of the clinicians (4/8) agreed or strongly agreed that the tool helped them to communicate, though two were neutral and two disagreed that the instrument would help them in the future. Most physicians (7/8) agreed or strongly agreed that they would want to use the tool in the future. Physicians expressed difficulties with Portuguese language pronunciation (Table
1).
The majority of emergency physicians (20/25) had read the pre-study e-mails introducing the study, though only two thirds (16/25) had reviewed the attached document copy of the language translation instrument, and few (2/25) tried the sound clips on the audio CD.
Free text clinical comments included apprehension regarding potential “false understanding” created by translation aids, a suggestion for incorporation of phonetic texts in the instrument, and an expression of dissatisfaction with the suggestion of acquiring language skills to accommodate non-English-speaking patients without their own interpreter.
Discussion
This study piloted an innovative intervention to improve communication with a subset of non-English-speaking patients in the ED. The intervention was well received by patients and physicians and resulted in improved communication despite suboptimal physician preparation. Our study supports previous literature indicating the beneficial impact of attempting cultural competency [
17]. Patients were extremely appreciative of EP attempts at language competency, despite the fact that the clinician spoke only a limited number of phrases in Portuguese. This model has the potential to be utilized with other languages and is easily adaptable for use in other EDs.
Other interventions designed to bridge LBs in the ED have been described. Bischoff et al. utilized brief training aimed at communicating with foreign-language patients and working with interpreters and demonstrated improved quality of communication as perceived by patients [
23]. Koff et al. developed a medical Spanish website for use in the ED, utilizing phrases and sound clips to facilitate history taking [
19]. Mazor et al. demonstrated that a 10-week Spanish course for pediatric emergency physicians was associated with decreased interpreter use and increased family satisfaction [
24].
Brief language instruction is associated with hazards as well as benefits. An intervention in the USA assessed eight PGY1 emergency medicine residents who underwent a 45-h medical Spanish course in their first residency month. Upon completion, major errors, such as misunderstanding of duration of symptoms and misunderstanding of vocabulary, were found in 14% of 34 subsequent physician-patient interactions by these residents over 6 months. The authors caution that although medical language courses may be a useful adjunct to interpreters, they are not designed to replace them, and that significant errors may occur when participants in such courses assume their knowledge is sufficient to obtain a good history, give patient release instructions, and provide medical care in general without an interpreter present [
25].
Pronunciation of even short Portuguese phrases is difficult to master and requires practice by clinicians to be successful. Shortening and simplifying several of the phrases may be beneficial. Although the tool employed close-ended questions constructed to elicit meaningful responses to the non-Portuguese speaker, patients would occasionally respond with long narratives that were incomprehensible to the clinician. A possible solution might be to design questions even more strictly close-ended than they presently are and to make available to the clinician the translated phrase “Please answer with only yes or no, or a number.”
Limitations of the study include small sample size and poor staff recruitment. Small sample size in this pilot study related chiefly to an unusually low number of ED patients who met the inclusion criteria, and possibly incomplete recruitment of eligible patients by the triage nurse and ED clinician given time constraints. (A repeat study with a revised instrument and larger sample size will allow for more meaningful statistic analysis.) Poor staff compliance with practicing with the audio CD prior to bedside pilot of the instrument limited usefulness of the communication tool, and inclusion of phonetic pronunciation on the printed instrument might have mitigated this weakness. Clinician indication of support for using the tool with future patients may promise greater compliance with practicing with the audio guide and greater success with the instrument after successful use. Having patients translate back the question to a bilingual observer would have further confirmed their comprehension and appropriateness of response.
Conclusion
Language barriers, patient fear, and apprehension impede optimal care. Attempts at cultural competency may go far in enhancing the patient’s positive impression of the encounter with the clinician, reducing patient stress, and helping the clinician to more easily elicit patient concerns. Improving language concordance even by learning a few key phrases is likely to have a positive impact on clinical care and demonstrates concern and respect for the patient.
Open Access This is an open access article distributed under the terms of the Creative Commons Attribution Noncommercial License (
https://creativecommons.org/licenses/by-nc/2.0
), which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.