Patterns of communication through interpreters: a detailed sociolinguistic analysis

Abstract

Background: Numerous articles have detailed how the presence of an interpreter leads to less satisfactory communication with physicians; few have studied how actual communication takes place through an interpreter in a clinical setting.

Objective: Record and analyze physician-interpreter-patient interactions.

Design: Primary care physicians with high-volume Hispanic practices were recruited for a communication study. Dyslipidemic Hispanic patients, either monolingual Spanish or bilingual Spanish-English, were recruited on the day of a normally scheduled appointment and, once consented, recorded without a researcher present in the room. Separate postvisit interviews were conducted with the patient and the physician. All interactions were fully transcribed and analyzed.

Participants: Sixteen patients were recorded interacting with 9 physicians. Thirteen patients used an interpreter with 8 physicians, and 3 patients spoke Spanish with the 1 bilingual physician.

Approach: Transcript analysis based on sociolinguistic and discourse analytic techniques, including but not limited to time speaking, analysis of questions asked and answered, and the loss of semantic information.

Results: Speech was significantly reduced and revised by the interpreter, resulting in an alteration of linguistic features such as content, meaning, reinforcement/validation, repetition, and affect. In addition, visits that included an interpreter had virtually no rapport-building "small talk," which typically enables the physician to gain comprehensive patient history, learn clinically relevant information, and increase emotional engagement in treatment.

Conclusions: The presence of an interpreter increases the difficulty of achieving good physician-patient communication. Physicians and interpreters should be trained in the process of communication and interpretation, to minimize conversational loss and maximize the information and relational exchange with interpreted patients.

FIGURE 1

Expected time spent talking as percentage of the visit, by participant.

FIGURE 2

Actual time spent talk as percentage of the visit, by speaker type.