Sign in Subscribe
Machine learning

Machine Translation in Patient-Centered Care: A Research Breakthrough

Strategic Insights

— 1 min read

Research from the University of Texas MD Anderson Cancer Center has shed light on the potential of machine translation in patient-centered care. The study focuses on the ability to rapidly and inexpensively create accurate translations of English language patient-reported outcome measures (PROMs) to facilitate global uptake. The researchers explored the effectiveness of machine translation tools, including Generative Pretrained Transformer (GPT)-4, GPT-3.5, and Google Translate, in translating English versions of selected PROMs scales into various languages.

Key Takeaways:

  • The study compared the translation quality of machine translation tools, including GPT-4, GPT-3.5, and Google Translate, in translating PROMs scales into six languages: Arabic, Vietnamese, Italian, Hungarian, Malay, and Dutch.
  • The research found that machine translation tools can produce high-quality PROM translations, but substituting human translation with machine translation is not advisable at the current stage.
  • The study used the Metrics for Evaluation of Translation with Explicit Ordering (METEOR) and Kruskal-Wallis test or analysis of variance to evaluate the translation quality.
  • The findings suggest that large language models, such as GPT-4 and GPT-3.5, provide high-quality PROM translations to support human translations and reduce costs.
  • The study emphasized the importance of human expertise in machine translation, especially in cases where high accuracy is crucial.

Statistics:

  • The study translated PROMs scales into six languages: Arabic, Vietnamese, Italian, Hungarian, Malay, and Dutch.
  • The research used Generative Pretrained Transformer (GPT)-4, GPT-3.5, and Google Translate as machine translation tools.
  • The METEOR scores significantly varied depending on target languages for all MT tools (p-value < 0.01).
  • The study compared the METEOR scores between different translation versions using the Kruskal-Wallis test or analysis of variance, as appropriate.

Sources:

  • Can machine translation match human expertise? Quantifying the performance of large language models in the translation of patient-reported outcome measures (PROMs). Journal of Patient-Reported Outcomes, 2025, 9(1):1-11.
  • University of Texas MD Anderson Cancer Center. Research by Sheng-Chieh Lu, Department of Symptom Research, University of Texas MD Anderson Cancer Center, and additional authors Cai Xu, Manraj Kaur, Maria Orlando Edelen, Andrea Pusic, Chris Gibbons.

Read more