Patient safety/original research
Errors of Medical Interpretation and Their Potential Clinical Consequences: A Comparison of Professional Versus Ad Hoc Versus No Interpreters

Presented at the Pediatric Academic Societies annual meeting, May 2003, Seattle, WA; and the AcademyHealth annual meeting, June 2003, Nashville, TN.
https://doi.org/10.1016/j.annemergmed.2012.01.025Get rights and content

Study objective

To compare interpreter errors and their potential consequences in encounters with professional versus ad hoc versus no interpreters.

Methods

This was a cross-sectional error analysis of audiotaped emergency department (ED) visits during 30 months in the 2 largest pediatric EDs in Massachusetts. Participants were Spanish-speaking limited-English-proficient patients, caregivers, and their interpreters. Outcome measures included interpreter error numbers, types, and potential consequences.

Results

The 57 encounters included 20 with professional interpreters, 27 with ad hoc interpreters, and 10 with no interpreters; 1,884 interpreter errors were noted, and 18% had potential clinical consequences. The proportion of errors of potential consequence was significantly lower for professional (12%) versus ad hoc (22%) versus no interpreters (20%). Among professional interpreters, previous hours of interpreter training, but not years of experience, were significantly associated with error numbers, types, and potential consequences. The median errors by professional interpreters with greater than or equal to 100 hours of training was significantly lower, at 12, versus 33 for those with fewer than 100 hours of training. Those with greater than or equal to 100 hours of training committed significantly lower proportions of errors of potential consequence overall (2% versus 12%) and in every error category.

Conclusion

Professional interpreters result in a significantly lower likelihood of errors of potential consequence than ad hoc and no interpreters. Among professional interpreters, hours of previous training, but not years of experience, are associated with error numbers, types, and consequences. These findings suggest that requiring at least 100 hours of training for interpreters might have a major impact on reducing interpreter errors and their consequences in health care while improving quality and patient safety.

Introduction

More than 6,900 known living languages are spoken in the world.1 A sizable, growing body of literature documents that language barriers can have a major adverse effect on health and health care, including suboptimal health status; a lower likelihood of having a regular health care provider; lower rates of mammograms, pap smears, and other preventive services; greater likelihood of a diagnosis of more severe psychopathology and leaving the hospital against medical advice among psychiatric patients; increased risks of drug complications; and higher resource use for diagnostic testing.2

More than 59 million Americans speak a language other than English at home, and 25.2 million have limited English proficiency (ie, a self-rated ability to speak English less than “very well”).3 Despite the large number of Americans with limited English proficiency and federal policy4 requiring providing adequate language assistance to patients with limited English proficiency, many such patients do not receive professional medical interpretation, but rather must resort to using ad hoc interpreters, such as family members, friends, or strangers from the waiting room, or having no interpreter. One study of 530 Latino adult ED patients revealed no interpreter use for 46% of patients with limited English proficiency, and 39% of interpreters used had no training.5

To our knowledge, no study has compared errors in medical interpretation and their potential clinical consequences among professional hospital interpreters, ad hoc interpreters, and encounters with no interpreters. The aim, therefore, was to compare interpreter errors and their potential clinical consequences in pediatric encounters with hospital versus ad hoc versus no interpreters. A secondary aim was to determine whether, among professional interpreters, hours of previous interpretation training or years of interpreter experience are associated with error numbers, types, and potential clinical consequences, which might provide useful data for interpreter training programs, hospitals, and policymakers.

Section snippets

Materials and Methods

Encounters were audiotaped in the 2 largest pediatric emergency departments (EDs) in Massachusetts. To be eligible for the study, the child's principal caretaker had to identify Spanish as the primary language spoken at home and had to have limited English proficiency, according to the US Census definition (self-rated ability to speak English of less than “very well”).6 During 30 months (May 2000 to November 2002), pediatric ED visits were purposively sampled between 8 am and 11 pm on weekdays

Results

A total of 57 interpreted encounters were audiotaped, yielding 2,367 pages of transcripts and 61,478 dialogue lines. Interpreters present during these encounters included professional interpreters, 20 (35%); ad hoc interpreters, 27 (47%); and no interpreter, 10 (18%). There were no significant differences in the median number of lines of dialogue per encounter for professional versus ad hoc versus no interpreters (501 versus 630 versus 420, respectively; P=.20), consistent with previous work

Limitations

This research was conducted in 2 urban EDs in Massachusetts, so the findings may not necessarily generalize to other nonurban settings, other states, and other countries. This study was restricted to children; similar work on interpreter errors in adult encounters is needed. Only families with limited English proficiency caretakers who spoke Spanish as their primary language were enrolled in the study; additional studies are needed to determine whether these findings pertain to limited English

Discussion

The study findings document that the proportion of errors of potential clinical consequence is significantly lower for professional hospital interpreters, compared with ad hoc interpreters and having no interpreter, both of which were found to have approximately double the odds of professional interpreters of committing errors of potential clinical consequence. These findings are consistent with a substantial body of research showing that optimal communication, patient satisfaction, and

References (35)

  • G. Flores et al.

    Errors in medical interpretation and their potential clinical consequences in pediatric encounters

    Pediatrics

    (2003)
  • H. Barik

    Simultaneous Interpretation: Temporal and Quantitative Data [unpublished PhD dissertation]

    (1972)
  • J.C. Hornberger et al.

    Eliminating language barriers for non-English-speaking patients

    Med Care

    (1996)
  • G. Flores et al.

    Access to hospital interpreter services for limited English proficient patients in New Jersey: a statewide evaluation

    J Health Care Poor Underserved

    (2008)
  • P. Hudelson et al.

    Overcoming language barriers with foreign-language speaking patients: a survey to investigate intra-hospital variation in attitudes and practices

    BMC Health Serv Res

    (2009)
  • F.M. Gany et al.

    Reducing clinical errors in cancer education: interpreter training

    J Cancer Educ

    (2010)
  • P.N. Butow et al.

    Interpretation in consultations with immigrant patients with cancer: how accurate is it?

    J Clin Oncol

    (2011)
  • Cited by (325)

    • Association between no-show rates and interpreter use in a pediatric otolaryngology clinic

      2023, International Journal of Pediatric Otorhinolaryngology
    View all citing articles on Scopus

    Supervising editor: Robert L. Wears, MD, PhD

    Author contributions: GF had full access to all study data, takes responsibility for the data integrity and accuracy of the data analysis, was responsible for study concept and design, obtained funding, and provided administrative, technical, and material support. GF, MA, CPB, and RB were responsible for acquisition of data. GF, MA, and HL were responsible for analysis and interpretation of data. All authors drafted the article and were responsible for critical revision of the article for important intellectual content and final approval of the version to be published. HL was responsible for statistical analysis. GF and MA supervised the study. GF takes responsibility for the paper as a whole.

    Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). The authors have stated that no such relationships exist. Funded in part by a grant from the Office of Minority Health, US Department of Health and Human Services.

    Please see page 546 for the Editor's Capsule Summary of this article.

    A feedback survey is available with each research article published on the Web at www.annemergmed.com.

    A podcast for this article is available at www.annemergmed.com.

    Publication date: Available online March 15, 2012.

    View full text