NeFut Logo NeFut
中 Admin Login

[CS.AI] I Just Assumed It Would Translate: Examining Healthcare Staff’s MT Risk Awareness Using Abbreviations as a Use Case

Published at: 2026-10-05 22:00 Last updated: 2026-10-06 12:11
#AI #Machine Learning #GPT

In the UK, public healthcare workers frequently turn to machine translation—predominantly Google Translate—to communicate with patients who do not share a common language, aiming to fulfil their duty of care. However, uninformed reliance on MT in high‑stakes contexts such as patient records can jeopardise safety. Existing literature mainly examines MT in interpersonal exchanges or patient‑facing documents, leaving a gap in understanding staff awareness of MT risks in more critical scenarios.\ \ Medical abbreviations are already known to increase the likelihood of misinterpretation and medication errors even in monolingual settings, with consequences ranging from minor harm to death. Consequently, this study uses abbreviations as a use case to identify the specific risks introduced by MT. Researchers extracted authentic clinical sentences in French and Spanish from authoritative corpora, translated them with Google Translate, and presented the outputs during semi‑structured interviews with 21 healthcare staff members across diverse roles and specialties.\ \ Qualitative coding and cross‑analysis of the interview data revealed several key points:\

Original Source: https://arxiv.org/abs/2610.02496

[h] Back to Home