Evaluating a conceptual framework for quality assessment of medical interpretation

We aimed to evaluate a proposed conceptual framework for quality assessment of medical interpretation using actual data from clinical settings. A mixed methods approach was used. Routine outpatient encounters involving Portuguese-speaking Brazilian patients, Japanese physicians, and hospital-provided and ad hoc interpreters were digitally recorded and transcribed. A questionnaire survey examining participants’ satisfaction was conducted; 111 and 13 encounters by hospital-provided and ad hoc interpreters, respectively, were recorded. Segments of consecutively interpreted utterances were coded as “altered,” whereby the interpreter changed the meaning of the source utterance, or “unaltered (accurate).” Frequency and type of alteration were analyzed. The effect of positive interpretation alterations on physician-patient interactions was qualitatively described. Interpretation accuracy was significantly higher for hospital-provided interpreters, but was not associated with overall patient satisfaction. Overall physician satisfaction was associated with accurate interpretation, clinically negative altered interpretations, and positive voluntary interventions (p