The Impact of Medical Interpretation Method on Time and Errors

Abstract

Background: Twenty-two million Americans have limited English proficiency. Interpreting for limited English proficient patients is intended to enhance communication and delivery of quality medical care. Objective: Little is known about the impact of various interpreting methods on interpreting speed and errors. This investigation addresses this important gap. Design: Four scripted clinical encounters were used to enable the comparison of equivalent clinical content. These scripts were run across four interpreting methods, including remote simultaneous, remote consecutive, proximate consecutive, and proximate ad hoc interpreting. The first 3 methods utilized professional, trained interpreters, whereas the ad hoc method utilized untrained staff. Measurements: Audiotaped transcripts of the encounters were coded, using a prespecified algorithm to determine medical error and linguistic error, by coders blinded to the interpreting method. Encounters were also timed. Results: Remote simultaneous medical interpreting (RSMI) encounters averaged 12.72 vs 18.24 minutes for the next fastest mode (proximate ad hoc) (p = 0.002). There were 12 times more medical errors of moderate or greater clinical significance among utterances in non-RSMI encounters compared to RSMI encounters (p = 0.0002). Conclusions: Whereas limited by the small number of interpreters involved, our study found that RSMI resulted in fewer medical errors and was faster than non-RSMI methods of interpreting.