Google Translate

Purpose #

A free, general-purpose machine translation app increasingly used as an ad-hoc communication tool in healthcare settings to bridge language barriers with LEP patients, including via voice-to-voice translation.

"One such translation app is Google Translate, which enables translation between over 100 languages and is being increasingly investigated as a translation tool in healthcare settings." (Hwang, 2022, p. 579)

Mechanism #

Free-text (free-translation) machine translation over 100+ languages, including a voice-to-voice feature, allowing bidirectional translation of arbitrary phrases (unlike fixed-phrase phrasebook apps). Because translation is unconstrained and context-free, accuracy varies by language, accent, and audio conditions, and it is generally suited only to short or simple phrases.

"Different to Google Translate's feature of "free-­text" phrases, other language translation tools prefer "fixed-­phrase translation"" (Hwang, 2022, p. 579)

Example(s) of usage #

Trialed on aged-care hospital wards (6 of 21 analyzed observations); its voice-to-voice feature could translate patient responses but was often slow, e.g. "Needed to ask, 'Have you had a shower yesterday' 7 tries before translation." (Observer, site 1, Google Translate).

Other Notes #

Included in the trial specifically because ward staff were already using it and similar apps on an ad-hoc, unofficial basis. Prior studies cited indicate it is only suitable for short or simple healthcare phrases, with accuracy varying across languages.

Single canonical artifact for Google Translate across the corpus (merged 2026-07-30 from four duplicate ART nodes created by parallel extraction agents). One real-world product, used in two distinct modalities — the specific modality is recorded in each EVD's How?:

• Text / document translation (statistical MT at the time of the older studies): translating written patient-education material (@Khanna_2011_Performance_online) and public-health documents in an MT-plus-postediting workflow (@Turner_2014_comparison_human, @Turner_2015_Machine_Translation); text quality of safety statements across 20 languages (@Das_2019_Dangers_Machine).
• Conversation / voice mode (neural MT): real-time spoken bedside symptom assessment in the PACU (@Kapoor_2022_Use_Neural); ad-hoc voice-to-voice ward use (@Hwang_2022_Testing_use, @Panayiotou_2020_perceptions_translation).