Comparing in-person, video, and telephonic medical interpretation.

BACKGROUND

Using trained interpreters to provide medical interpretation services is superior to services provided on an ad hoc basis, but little is known about the effectiveness of providing their services remotely, especially using video.

OBJECTIVE

To compare remote medical interpretation services by trained interpreters via telephone and videoconference to those provided in-person.

DESIGN

Quasi-randomized control study.

PARTICIPANTS

Two hundred and forty-one Spanish speaking patient volunteers, twenty-four health providers, and seven interpreters.

APPROACH

Patients, providers and interpreters each independently completed scales evaluating the quality of clinical encounters and, optionally, made free text comments. Interviews were conducted with 23 of the providers, the seven interpreters, and a subset of 30 patients. Time data were collected.

RESULTS

Encounters with in-person interpretation were rated significantly higher by providers and interpreters, while patients rated all methods the same. There were no significant differences in provider and interpreter ratings of remote methods. Provider and interpreter comments on scales and interview data support the higher in-person ratings, but they also showed a distinct preference for video over the phone. Phone interviews were significantly shorter than in-person.

DISCUSSION

Patients rated interpretation services highly no matter how they were provided but experienced only the method employed at the time of the encounter. Providers and interpreters were exposed to all three methods, were more critical of remote methods, and preferred videoconferencing to the telephone as a remote method. The significantly shorter phone interviews raise questions about the prospects of miscommunication in telephonic interpretation, given the absence of a visual channel, but other factors might have affected time results. Since the patient population studied was Hispanic and predominantly female care must be taken in generalizing these results to other populations.