ArticlePDF Available

Use of Google Translate in medical communication: Evaluation of accuracy

Authors:
Sumant Patil at Deenanath Mangeshkar Hospital
Sumant Patil
Patrick Davies at Nottingham University Hospitals NHS Trust
Patrick Davies
Download full-text PDFDownload full-text PDF
Read full-text
Download citation

Abstract

Communication is the cornerstone of medicine, without which we cannot interact with our patients.1 The General Medical Council’s Good Medical Practice states that “Doctors must listen to patients, take account of their views, and respond honestly to their questions.”2 However, we still often interact with patients who do not speak the local language. In the United Kingdom most hospitals have access to translation services, but they are expensive and often cumbersome. A complex and nuanced medical, ethical, and treatment discussion with patients whose knowledge of the local language is inadequate remains challenging. Indeed, even in a native language there is an element of translation from medical to lay terminology. We recently treated a very sick child in our paediatric intensive care unit. The parents did not speak English, and there were no human translators available. Reluctantly we resorted to a web based translation tool. We were uncertain whether Google Translate was accurately translating our complex medical phrases.3 4 Fortunately our patient recovered, and a …
bmj.g73
92.pdf
Content uploaded by Patrick Davies
Author content
All content in this area was uploaded by Patrick Davies on Jan 30, 2015
Content may be subject to copyright.
Use of Google Translate in medical communication: Evaluation of accura
cy.pdf
Available via license: CC BY-NC 4.0
Content may be subject to copyright.
CHRISTMAS 2014: FOUND IN TRANSLATION
Use of Google Translate in medical communication:
evaluation of accuracy
OPEN ACCESS
Sumant Patil senior clinical fellow, Patrick Davies consultant
Paediatric Intensive Care Unit, Nottingham Children’s Hospital, Nottingham, UK
Communication is the cornerstone of medicine, without which
we cannot interact with our patients.1The General Medical
Council’s Good Medical Practice states that “Doctors must
listen to patients, take account of their views, and respond
honestly to their questions.”2However, we still often interact
with patients who do not speak the local language.
In the United Kingdom most hospitals have access to translation
services, but they are expensive and often cumbersome. A
complex and nuanced medical, ethical, and treatment discussion
with patients whose knowledge of the local language is
inadequate remains challenging. Indeed, even in a native
language there is an element of translation from medical to lay
terminology.
We recently treated a very sick child in our paediatric intensive
care unit. The parents did not speak English, and there were no
human translators available. Reluctantly we resorted to a web
based translation tool. We were uncertain whether Google
Translate was accurately translating our complex medical
phrases.3 4 Fortunately our patient recovered, and a human
translator later reassured us that we had conveyed information
accurately.
We aimed to evaluate the accuracy and usefulness of Google
Translate in translating common English medical statements.
Methods
Ten commonly used medical statements were chosen by author
consensus. These were translated via Google Translate to 26
languages. Translations only were sent to native speakers of
each of these languages and translated back to English by them.
The returned English phrases were compared with the originals
and assessed for meaning. If translations did not make sense or
were factually incorrect they were considered as wrong. Minor
grammatical errors were allowed.
Results
Ten medical phrases were evaluated in 26 languages (8 Western
European, 5 Eastern European, 11 Asian, and 2 African), giving
260 translated phrases. Of the total translations, 150 (57.7%)
were correct while 110 (42.3%) were wrong. African languages
scored lowest (45% correct), followed by Asian languages
(46%), Eastern European next with 62%, and Western European
languages were most accurate at 74%. The medical phrase that
was best translated across all languages was “Your husband has
the opportunity to donate his organs” (88.5%), while “Your
child has been fitting” was translated accurately in only 7.7%
(table⇓). Swahili scored lowest with only 10% correct, while
Portuguese scored highest at 90%.
There were some serious errors. For instance, “Your child is
fitting” translated in Swahili to “Your child is dead.” In Polish
“Your husband has the opportunity to donate his organs”
translated to “Your husband can donate his tools.” In Marathi
“Your husband had a cardiac arrest” translated to “Your husband
had an imprisonment of heart.” “Your wife needs to be
ventilated” in Bengali translated to “Your wife wind movement
needed.”
Discussion
Google Translate is an easily available free online machine
translation tool for 80 languages worldwide.5However, we have
found limited usefulness for medical phrases used in
communications between patients and doctor.367
We found many translations that were completely wrong.
Google Translate uses statistical matching to translate rather
than a dictionary/grammar rules approach, which leaves it open
to nonsensical results.4 8
In today’s world “just Google it” is considered to be the answer
to everything, but for health related questions this should be
treated with caution.9Google Translate should not be used for
Correspondence to: P Davies patrick.davies@nuh.nhs.uk
No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe
BMJ 2014;349:g7392 doi: 10.1136/bmj.g7392 (Published 15 December 2014) Page 1 of 3
Research
RESEARCH
taking consent for surgery, procedures, or research from patients
or relatives unless all avenues to find a human translator have
been exhausted, and the procedure is clinically urgent. We have,
however, not assessed the accuracy of human translators, who
cannot be assumed to be perfect and may be subject to
confidentiality breaches.
We looked at translations from and to English only. Western
European languages were the most accurately translated,
implying a bias in translating algorithms towards those
languages more commonly used in computing. Previous research
has used one phrase, using the same algorithm to translate and
retranslate, which is likely to increase the stated accuracy.10 11
Conclusion
Google Translate has only 57.7% accuracy when used for
medical phrase translations and should not be trusted for
important medical communications. However, it still remains
the most easily available and free initial mode of communication
between a doctor and patient when language is a barrier.
Although caution is needed when life saving or legal
communications are necessary, it can be a useful adjunct to
human translation services when these are not available.
Contributors: SP and PD shared the conception of the idea, sourcing
suitable translators, and drafting the paper. PD is guarantor.
Funding: None.
Competing interests: We have read and understood the BMJ Group
policy on declaration of interests and have no relevant interests to
declare.
1 Brindley PG, Smith KE, Cardinal P, LeBlanc F. Improving medical communication: skills
for a complex (and multilingual) clinical world. Can Respir J 2014;21:89-91.
2 General Medical Council. Good medical practice . GMC, 2013: paragraph 31. www.gmc-
uk.org/guidance/good_medical_practice.asp.
3 Börner N, Sponholz S, König K, Brodkorb S, Bührer C, Roehr CC. [Google Translate is
not sufficient to overcome language barriers in neonatal medicine]. Klin Padiatr
2013;225:413-7.
4 Balk EM, Chung M, Chen ML, Trikalinos TA, Kong Win Chang L. Assessing the accuracy
of Google Translate to allow data extraction from trials published in non-English languages.
AHRQ Methods for Effective Health Care 2013 Report No 12(13)-EHC145-EF.
5 Google Translate. http://translate.google.com/about/intl/en_ALL/.
6 Palluzi J. Results of using Google Translate for medical communication on the Android
OS. iMedicalApps.com, 2010. www.imedicalapps.com/2010/07/results-of-using-google-
translate-for-medical-communication-on-the-android-os/.
7 Wade R. Try Google Translate to overcome language barriers. BMJ 2011;343:d7217.
8 Gomes L. Google Translate tangles with computer learning. Forbes 2010 Jul 22. www.
forbes.com/forbes/2010/0809/technology-computer-learning-google-translate.html.
9 Scullard P, Peacock C, Davies P. Googling children’s health: reliability of medical advice
on the internet. Arch Dis Child 2010;95:580-2.
10 Rubin M. Lost in translation. What happens when patients don’t speak English? EMS
World 2013;42:122.
11 Kaliyadan F, Gopinath Pillai S. The use of google language tools as an interpretation aid
in cross-cultural doctor-patient interaction: a pilot study. Inform Prim Care 2010;18:141-3.
Accepted: 14 November 2014
Cite this as: BMJ 2014;349:g7392
This is an Open Access article distributed in accordance with the Creative Commons
Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute,
remix, adapt, build upon this work non-commercially, and license their derivative works
on different terms, provided the original work is properly cited and the use is
non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe
BMJ 2014;349:g7392 doi: 10.1136/bmj.g7392 (Published 15 December 2014) Page 2 of 3
RESEARCH
Table
Table 1| List of medical phrases translated via Google Translate
Percentage correctSample or most common errorPhrase translated
53.8Your wife cannot fall overYour wife is stable
53.8Your husband’s heart was imprisonedYour husband had a cardiac arrest
73.1Your husband’s heart was attackedYour husband had a heart attack
26.9Your wife needs to be airedYour wife needs to be ventilated
69.2Your child’s state is not life stoppingYour child’s condition is life threatening
7.7Your child has been constructingYour child has been fitting
76.9Your child is sleeping earlyYour child will be born premature
88.5Your husband is now ready to donateYour husband has the opportunity to donate his organs
61.5We need your consent for operating (such as machinery)We will need your consent for operation
65.4Your home temperature was highDid he have high fever at home?
No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe
BMJ 2014;349:g7392 doi: 10.1136/bmj.g7392 (Published 15 December 2014) Page 3 of 3
RESEARCH
... Google Translate [96] marked a significant advancement in machine translation using Google Neural Machine Translation [97], leveraging statistical methods and later incorporating deep learning models [98]. Google Translate supports translation between over 100 languages, ranging from widely spoken languages like English, Spanish, and Chinese to less common ones such as Icelandic and Maori [99].Some studies have reviewed the effectiveness of Google Translate, such as translation in medical communication [100] and Sanskrit to English translation [74]. Tsai et al. [101] demonstrated that Google Translate English texts presented a number of components of significantly higher writing quality with fewer mistakes in spelling and grammar and fewer errors per word than its counterparts [101]. ...
An evaluation of LLMs and Google Translate for translation of selected Indian languages via sentiment and semantic analyses
Preprint
Full-text available
  • Mar 2025
Large Language models (LLMs) have been prominent for language translation, including low-resource languages. There has been limited study about the assessment of the quality of translations generated by LLMs, including Gemini, GPT and Google Translate. In this study, we address this limitation by using semantic and sentiment analysis of selected LLMs for Indian languages, including Sanskrit, Telugu and Hindi. We select prominent texts that have been well translated by experts and use LLMs to generate their translations to English, and then we provide a comparison with selected expert (human) translations. Our findings suggest that while LLMs have made significant progress in translation accuracy, challenges remain in preserving sentiment and semantic integrity, especially in figurative and philosophical contexts. The sentiment analysis revealed that GPT-4o and GPT-3.5 are better at preserving the sentiments for the Bhagavad Gita (Sanskrit-English) translations when compared to Google Translate. We observed a similar trend for the case of Tamas (Hindi-English) and Maha P (Telugu-English) translations. GPT-4o performs similarly to GPT-3.5 in the translation in terms of sentiments for the three languages. We found that LLMs are generally better at translation for capturing sentiments when compared to Google Translate.
View
... Compared to other communication tools, Google Translate was reported to be the most common translation app used in healthcare [37]. Satisfaction and habit of using google translation among midwives could lead to resistance to use other dialog support tool, despite the fact of the inaccuracies in automatic translations [50]. Moreover, resistance to change in healthcare could be a reason for challenging implementation of dialog support. ...
Midwives’ perceptions of communication at antenatal care using a bilingual digital dialog support tool– a qualitative study
Article
Full-text available
  • Mar 2025
Background Sweden has a large population of migrant women, which contributes to communication challenges and, consequently, suboptimal maternity care. Compared with native-born women, migrant women have an increased prevalence of adverse pregnancy outcomes. Miscommunication and language barriers are among the reasons for these results. Thus, language barriers can also lead to providing less information to migrant women. A digital Swedish-Arabic dialog support tool was developed and tested at antenatal care, to facilitate communication between midwives and Arabic-speaking women. This study aimed to describe midwives’ perceptions of communication via Swedish-Arabic dialog support (Sadima) in antenatal care. Methods A qualitative study was conducted with 14 midwives in antenatal care with experience communicating using a Swedish-Arabic dialog support tool. The data were collected via semi structured individual interviews and were analyzed via phenomenographic analysis. Results The analysis resulted in three categories: (1) Dialog support - the skill of constructing bridges, comprised the main finding that dialog support facilitated communication by providing a multimodal way of communication including intercultural evidence-based content;(2) Dialog support - challengingly implementing adaptive efficiency, represented the implementation of dialog support to be time-consuming and, eventually, time-efficient when midwives gained digital skills; and (3) Women and their partners - the ability to be empowered, included the main finding of increased women’s empowerment and control over their lives by being less dependent on interpreters. Conclusions The findings contribute to the understanding of communication via dialog support based on midwives’ experiences. This study highlights that communication via dialog support facilitates communication between midwives and Arabic-speaking women and enhances midwives’ working conditions. Within our increasingly heterogeneous societies, health care could provide support for communication via digital dialog support that is women-centered and culturally sensitive to avoid misunderstandings and delayed or incorrect treatment of migrant women.
View
... Even though Google Translate does provide some extremely wrong results (as pointed out by Patil and Davies, 2014). So, to gauge how well Google Translate (also called Machine Translation or MT) performs, an error analysis should be run. ...
Inaccuracy of An Examination of Grammatical Error
Article
Full-text available
  • Mar 2025
When it comes to the construction of a communication process, grammar plays a significant part. It is utilized to make a clear declaration about the time at which a verb happens. The purpose of this study is to identify grammatical faults in the translation of abstracts from the Indonesian language to English that were produced by Google Translate. These errors pertain to tense, active, and passive voice abstracts. In this particular research project, the research method that was utilized was the descriptive qualitative method. The twenty abstracts that were retrieved from the repository at the University served as the population for this investigation. Documentation and checklists were the instruments that were utilized in the process of doing the data collection. When analyzing the data, Keshavarz's (1999) hypothesis was utilized. Following an examination of the data, it was discovered that the most common error that Google Translate generated was related to the use of tense. In the meanwhile, errors involving active and passive voices were only sometimes discovered. It is possible to conclude that Google Translate, operating as a tool, is unable to determine the tense of the sentences that are being translated. This is a possibility since tenses are strongly connected to the context in which the moment a verb occurs in a sentence, which is something that can only be broken down and comprehended by humans.
View
... However, when these systems fail, the consequences can be severe (Canfora and Ottmann, 2020). The results of a study conducted by Patil and Davies (2014) show Google Translate incorrectly translating the phrase "Your child is fitting" (which denotes a child having seizures) into the Swahili equivalent of "Your child is dead". While translation errors in medical contexts can lead to life-threatening misunderstandings, it is not the only domain where MT systems can fail. ...
Evaluating Gender Bias in German Machine Translation
Preprint
Full-text available
  • Feb 2025
We present WinoMTDE, a new gender bias evaluation test set designed to assess occupational stereotyping and underrepresentation in German machine translation (MT) systems. Building on the automatic evaluation method introduced by arXiv:1906.00591v1 [cs.CL], we extend the approach to German, a language with grammatical gender. The WinoMTDE dataset comprises 288 German sentences that are balanced in regard to gender, as well as stereotype, which was annotated using German labor statistics. We conduct a large-scale evaluation of five widely used MT systems and a large language model. Our results reveal persistent bias in most models, with the LLM outperforming traditional systems. The dataset and evaluation code are publicly available under https://github.com/michellekappl/mt_gender_german.
View
... Multilingual demands (Moorkens and Arenas, 2024), paired with the accessibility of online systems, has put MT at the forefront of user-facing language technologies. Once confined to professional settings, MT is now used by millions (Pitman, 2021), bringing into its fold an array of lay users in contexts ranging from casual interactions (Gao et al., 2015) to critical domains such as healthcare and employment (Patil and Davies, 2014;Dew et al., 2018;Liebling et al., 2022;Valdez et al., 2023). Despite MT's broad reach and potential for social impact in sensitive scenarios (Vieira et al., 2021), still little is known about its evolving relationship with the general public, how non-expert users interact with it, or how it caters to their needs. ...
Translation in the Hands of Many:Centering Lay Users in Machine Translation Interactions
Preprint
Full-text available
  • Feb 2025
Converging societal and technical factors have transformed language technologies into user-facing applications employed across languages. Machine Translation (MT) has become a global tool, with cross-lingual services now also supported by dialogue systems powered by multilingual Large Language Models (LLMs). This accessibility has expanded MT's reach to a vast base of lay users, often with little to no expertise in the languages or the technology itself. Despite this, the understanding of MT consumed by this diverse group of users -- their needs, experiences, and interactions with these systems -- remains limited. This paper traces the shift in MT user profiles, focusing on non-expert users and how their engagement with these systems may change with LLMs. We identify three key factors -- usability, trust, and literacy -- that shape these interactions and must be addressed to align MT with user needs. By exploring these dimensions, we offer insights to guide future MT with a user-centered approach.
View
Extreme Weather, Vulnerable Populations, and Mental Health: The Timely Role of AI Interventions
Article
Environmental disasters are becoming increasingly frequent and severe, disproportionately impacting vulnerable populations who face compounded risks due to intersectional factors such as gender, socioeconomic status, rural residence, and cultural identity. These events exacerbate mental health challenges, including post-traumatic stress disorder (PTSD), anxiety, and depression, particularly in low- and middle-income countries (LMICs) and underserved areas of high-income countries (HICs). Addressing these disparities necessitates inclusive, culturally competent, intersectional, and cost-effective strategies. Artificial intelligence (AI) presents transformative potential for delivering scalable and culturally tailored mental health interventions that account for these vulnerabilities. This perspective highlights the importance of co-designing AI tools with at-risk populations, integrating these solutions into disaster management frameworks, and ensuring their sustainability through research, training, and policy support. By embedding mental health resilience into climate adaptation strategies, stakeholders can foster equitable recovery and reduce the long-term mental health burden of environmental disasters.
View
AI Translation Tools in Language Learning: Bridging Gaps and Enhancing Skills
Chapter
  • Mar 2025
The chapter highlights the impact of AI for teaching translation skills focusing on the usage of AI translation tools by investigates the impact of AI translation tools on English language learning. It highlights how these technologies bridge gaps in language acquisition and enhance various language skills by examining the studies in the contexts of AI and teaching and learning translation. It covers various aspects of the impact of AI translation tools on the teaching and learning of translation in various contexts by unveiling the ways AI translation tools facilitate vocabulary expansion, grammar understanding, and pronunciation improvement while providing real-time translation and contextual support. It also addresses the challenges of using AI tools, such as over-reliance, contextual inaccuracies, and the need for critical engagement, offering strategies for educators and learners to integrate these tools effectively taking into consideration ethical issues. The chapter covers a wide range of studies in the context of translation and AI tools to present a comprehensive picture for the readers' insight into various aspects of the impact of artificial intelligence on the teaching and learning of translation in various contexts. It concludes that AI has transformed the field of translation studies completely and most of the influence is positive but the ethical issues and fake translations are the challenges need to be addressed more efficiently.
View
From Bytes to Bites: Application of Large Language Models to Enhance Nutritional Recommendations in CKD
Article
Full-text available
  • Mar 2025
Background Large language models (LLMs) such as ChatGPT are increasingly positioned to be integrated into various aspects of daily life, with promising applications in healthcare, including personalized nutritional guidance for patients with chronic kidney disease (CKD). However, for LLM-powered nutrition support tools to reach their full potential, active collaboration of healthcare professionals, patients, caregivers, and LLM experts is crucial. Methods We conducted a comprehensive review of the literature on the use of LLMs as tools to enhance nutrition recommendations for patients with CKD, curated by our expertise in the field. Additionally, we considered relevant findings from adjacent fields, including diabetes and obesity management. Results Currently, the application of LLMs for CKD-specific nutrition support remains limited and has room for improvement. Although LLMs can generate recipe ideas, their nutritional analyses often underestimate critical food components such as electrolytes and calories. Anticipated advancements in LLMs and other generative AI technologies are expected to enhance these capabilities, potentially enabling accurate nutritional analysis, the generation of visual aids for cooking, and identification of kidney-healthy options in restaurants. Conclusion While LLM-based nutritional support for patients with CKD is still in its early stages, rapid advancements are expected in the near future. Engagement from the CKD community, including healthcare professionals, patients, and caregivers, will be essential to harness AI-driven improvements in nutritional care with a balanced perspective that is both critical and optimistic.
View
View
Bridging language barriers in healthcare: a patient-centric mobile app for multilingual health record access and sharing
Article
  • Feb 2025
Introduction Access to health data for patients is hindered by a fragmented healthcare system and the absence of unified, patient-centric solutions. Additionally, there are no mechanics for easy sharing of medical records with healthcare providers, risking incomplete diagnoses. To further intensify the problem, when patients seek care abroad, language barriers may prevent foreign doctors from understanding their health data, further complicating treatment. Methods Our study presents the development and evaluation of a mobile application designed to enable users to access and share their health records directly from their device, in multiple languages, ensuring ease of use and convenience. The solution utilizes OpenNCP for translating patient summaries into multiple languages and the FHIR Smart Health Links Protocol for secure sharing. We conducted a user acceptance study with 45 participants to evaluate our mobile app's interface and functionality. Results The feedback was positive, highlighting the app's user-friendliness and usefulness. The participants felt it would enhance communication between physicians and patients and the features of sharing and translating are going to give more control of their medical data to the patients. Discussion Based on the results and participants feedback, our mobile solution significantly enhances healthcare accessibility and efficiency by enabling easy access and sharing of health records in multiple languages, using relevant protocols and standards, reducing medical errors and ensuring personalized care.
View
Improving Medical Communication: Skills for a Complex (And Multilingual) Clinical World
Article
Full-text available
It has been reported that suboptimal communication represents the largest source of preventable error during acute medical care. Because a significant proportion of ongoing care relies heavily on verbal communication, it is incumbent on clinicians to develop, hone and maintain these skills in the interests of their patients and, at the same time, contribute to a more reliable and patient-focused health care system. This review briefly discusses why communication matters, practical strategies from both inside and outside clinical medicine, the implications of poor translation and the state of medical communication in Canada.
View
Try Google Translate to overcome language barriers
Article
Full-text available
  • Nov 2011
  • Br Med J
In 2007 Jones debated whether the NHS should curb spending on translation services. 1 In 2011 the NHS might consider Google Translate—a free application for iPhones and Android devices and an effective alternative way of communicating with non-English speakers. The user speaks or types the text into the phone and selects the desired languages for translation; conversion takes a matter of seconds. The application currently understands and translates between 66 languages. If a speaker's nationality is uncertain, Google Translate detects the language automatically. Translated text can then either be read on-screen or spoken aloud by the phone. Informal testing with multilingual staff and non-English speaking patients has yielded satisfying results, with a high degree of accuracy. No studies of its use in healthcare environments have apparently been published, but a recent randomised trial in a US paediatric emergency department concluded that remote telephonic interpretation was liked as much as having a dedicated medical interpreter in person and more than having a bilingual physician in person. 2 Given that Google Translate is free and easy to use and allows questions to be customised, doctors should try it when other methods of translation are unavailable or inadequate.
View
View
Googling children's health: Reliability of medical advice on the internet
Article
Full-text available
  • Apr 2010
  • Arch Dis Child
To assess the reliability and accuracy of medical advice, over a range of types of websites, found using the Google search engine, thus simulating a patient's experience. Advice was sought for five common paediatric questions using the Google search engine. The first 100 results of each question were classified as either being consistent or inconsistent with current recommendations or as 'no answer given'. Record of the type of site and its visibility was noted. 39% of the 500 sites searched gave correct information; 11% were incorrect and 49% failed to answer the question. Where an answer was available, 78% of sites gave the correct information. The accuracy of information varied depending on the topic and ranged from 51% (mumps, measles and rubella and autism) to 100% (breast feeding with mastitis/the sleeping position of a baby). Governmental sites gave uniformly accurate advice. News sites gave correct advice in 55% of cases. No sponsored sites were encountered that gave the correct advice. The authors have shown that the advice on the internet is very variable. Patients are known to use the internet for their own research and as such the authors encourage healthcare workers to recommend government or NHS websites.
View
View
[Google Translate is not Sufficient to Overcome Language Barriers in Neonatal Medicine.]
Article
  • Aug 2013
  • Klin Pädiatr
Language barriers hinder the interaction with patients and relatives. The use of language services increases knowledge, satisfaction and the use of medical care and thus improves patient's clinical outcome. The recommended use of professional interpreters (PI) is not always feasible. We tested an online translation tool as an alternative for PI for the transla-tion of standardized sentences from a neonatal doctor-/nurse-relative-interview.Translation of 20 sentences from a German neonatal intensive care unit parent information brochure to English, Portuguese and Arabic, using Google Translate (GT). Assessment of accuracy concerning grammar and content, in a second step simplification of all incorrect sentences, translation by GT and critical re-assessment and evaluation.An average of 42% of the sentences was correctly translated concerning grammar and content. The proportion of incorrectly translated sentences varied between 45-70%. By simpli-fication another 23% were translated correctly.Translations by GT were often incorrect in content and grammar. We suppose that the design of GT, which is a statistical translation engine, might be an explanation for this phenomenon. Presently, GT cannot guarantee unambiguous translations and cannot substitute PIs, only in particular circumstances, the use of GT or similar engines may be justified. For future use of electronic translation services, we suggest to compile a catalogue of sentences containing central information, which can be translated into defined foreign languages without misinterpretation or loss of information. © Georg Thieme Verlag KG Stuttgart · New York.
View
General Medical Council Good medical practice
  • Jan 2013
General Medical Council. Good medical practice. GMC, 2013: paragraph 31. www.gmcuk.org/guidance/good_medical_practice.asp.
a complex (and multilingual) clinical world
  • Jan 2014
  • 89-91
a complex (and multilingual) clinical world. Can Respir J 2014;21:89-91. 2
Assessing the accuracy of Google Translate to allow data extraction from trials published in non-English languages. AHRQ Methods for Effective Health Care
  • Jan 2013
  • E M Balk
  • M Chung
  • M L Chen
  • T A Trikalinos
  • Kong Win Chang
Balk EM, Chung M, Chen ML, Trikalinos TA, Kong Win Chang L. Assessing the accuracy of Google Translate to allow data extraction from trials published in non-English languages. AHRQ Methods for Effective Health Care 2013 Report No 12(13)-EHC145-EF.
Google Translate tangles with computer learning
  • Jul 2010
  • L Gomes
Gomes L. Google Translate tangles with computer learning. Forbes 2010 Jul 22. www. forbes.com/forbes/2010/0809/technology-computer-learning-google-translate.html.