Japan’s Hospitals Need More Professional Medical Interpreters

76 views
Skip to first unread message

John Stroman

unread,
Oct 27, 2023, 9:34:30 AM10/27/23
to hon...@googlegroups.com

Fred Uleman

unread,
Oct 27, 2023, 9:58:01 AM10/27/23
to hon...@googlegroups.com
I wonder if their insurance covers medical interpretation.

----
fred uleman

Bill Lise

unread,
Oct 27, 2023, 10:20:12 AM10/27/23
to hon...@googlegroups.com
I noticed that the Minato Red Cross hospital here in Yokohama was recently advertising for interpreters (CN<>JA) at 1500 yen/hour, so maybe the insurance might cover that cost, quality concerns aside. 

--
You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CAA%2BTmpG7zRJU5F_ewOmDhwQjbtWDiv9jTJej7XSVicddVmvytg%40mail.gmail.com.

Tom Gally

unread,
Oct 27, 2023, 10:23:02 AM10/27/23
to hon...@googlegroups.com
It’s too bad that the article didn’t mention the languages needed for interpretation. The language most in demand is almost certainly Chinese, followed by two or three other Asian languages before English.

Two personal anecdotes:

One hospital I go to for treatment is in Omotesando, Tokyo. It has a sign in the stairwell in Chinese, Korean, and Russian announcing the availability of medical interpreters. According to the hospital’s website, they also offer interpretation in English, but apparently there wasn’t enough demand for English for it to be included on the sign. (Russian is included because the hospital specializes in thyroid conditions and it gets—or used to get—Russian-speaking patients who had been affected by the Chernobyl accident.)

Last week, I spent three nights in a hospital in Yokohama for some dental surgery. (Now fully recovered, thank you.) The hospital has a full-time Chinese interpreter, whom I have seen several times helping Chinese outpatients. The hospital’s website says that it also provides interpreting in English and Korean. One of the nurses told me, though, that I was the first English-speaking inpatient she has dealt with at the hospital. She also told me that, at the senmongakkō affiliated with the hospital, the most popular foreign language chosen by future nurses is Chinese, followed by English and then Spanish. She said that she had chosen English but has never had any use for it. (We were speaking Japanese, of course.)

Tom Gally

Matthew Schlecht

unread,
Oct 27, 2023, 11:54:04 AM10/27/23
to hon...@googlegroups.com
I have no experience with hospitals in Japan, but I did have the unique experience of spending 12 hours in an emergency room in Flushing, Queens, NY, USA a few months ago (I am fine now).
The ER scene is much different from what admitted patients will experience, but I noted that the medical staff made extensive use of smartphone translation apps to communicate with patients who had limited/no English capability; mainly Chinese speakers but also some Spanish speakers. There was usually a Spanish speaking staff member (not an interpreter) available, but far less often a Chinese speaker.
I didn't witness a working interpreter during the entire stay.

Matthew Schlecht, PhD
Word Alchemy Translation, Inc.
Newark, DE, USA
wordalchemytranslation.com

Jon Johanning

unread,
Oct 27, 2023, 1:32:36 PM10/27/23
to Honyaku E<>J translation list
These phone apps are getting better and better for lots of conversations, I guess. But how are they for medical terminology? Perhaps they have enough detail for communicating with patients, who don't need very precise terms. 

They have certainly left the old printed phrasebooks that I remember far behind, which didn't go much beyond "I'm in pain" and various parts of the body you could add to that to explain where it hurt.

Jon Johanning

Tom Gally

unread,
Oct 27, 2023, 9:45:13 PM10/27/23
to hon...@googlegroups.com
Six or seven years ago, when the Google Translate smartphone app was still relatively new, some graduate students and I did experiments to find out how well spoken conversations could be conducted with various language pairs. We had native speakers of Japanese, English, Chinese, and Polish, and we gave each pair a simple situation to role play—getting to know each other, making an appointment to meet later, etc. We recorded all of the conversations and analyzed them later.

The conversations started okay, but they all eventually broke down, with one person unable to understand what the other wanted to say and with attempts at clarification also failing. The Japanese-English conversations were able to continue for a minute or two, while the Chinese-Polish ones broke down more quickly. The Polish student, who also knew Chinese, said that the Polish sounded as though it had been translated through English rather than directly from Chinese. (I read reports at the time that Google used English as an intermediate language for less common language pairs. That might no longer be done with LLM-based translation.)

The translation performance of the apps has presumably improved since then, but they still have additional points of failure compared with text-based machine translation: speakers are more likely to elide key information than writers are, and voice recognition mistakes can make accurate translation impossible. At that time, I concluded that the apps could not be used for extended real-life conversations unless both users had been trained on how to use them.

Thinking back on my recent hospital stay and my multiple visits to the hospital in the weeks beforehand, I suspect that translation apps would be of only limited use in that hospital. They might allow patients to explain their symptoms, medications, medical histories, etc. to doctors and nurses, but I imagine that the medical personnel would hesitate to make decisions based only on the app translations, as even first-language self-reporting may not be reliable. Medical personnel could also use the apps to give instructions to patients—“Take off your shirt,” “Turn your head to the left,” “Go to room E123,” etc.—but many of those instructions can be done just as well with gestures. On my outpatient visits, more than once I saw Japanese hospital staff using gestures to communicate effectively with Chinese patients who didn’t understand much or any Japanese. (The hospital is between Yokohama Chinatown and the Kotobukichō skid row area, making for a diverse clientele.)

Perhaps the biggest need for translation in a hospital is to enable patients and their families to feel confident that they have been understood and are being treated appropriately. For that purpose, a human interpreter would definitely be better than an app.

Tom Gally

P.S. On a lighter note, I have been playing with DALLE-E 3 lately. Here is an attempt to express the spirit of this mailing list:


--
You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/f5a28f27-d50a-49bf-838b-b1e4ee82a809n%40googlegroups.com.

Tom Gally

unread,
Oct 27, 2023, 11:43:45 PM10/27/23
to hon...@googlegroups.com
I like this version better:


My original prompt was “A photo of three translators sitting around a table having a lively discussion about language. The translators are three different animals. On the wall behind them is the word ‘Honyaku’.”

After several iterations, in most of which the animals were fine but the word “Honyaku” was garbled, DALL-E 3 came up with the following prompt, which yielded the above image:

“Photo of a fluffy rabbit, a bulky walrus, and a color-shifting chameleon gathered around a dark wooden table in a modern office illuminated by ceiling lights. They are animatedly discussing linguistic topics, using their paws, tusks, and tails for emphasis. The table surface is cluttered with dictionaries and reference books. On the pristine white wall in the background, 'HONYAKU' is prominently written in black letters.”

The whole process took about ten minutes. I’m glad I’m not trying to make a living as an illustrator.

Tom Gally

On Sat, Oct 28, 2023 at 10:44 AM Tom Gally <tomg...@gmail.com> wrote:
...

P.S. On a lighter note, I have been playing with DALLE-E 3 [sic] lately. Here is an attempt to express the spirit of this mailing list:


cpta...@ozemail.com.au

unread,
Oct 28, 2023, 2:00:43 AM10/28/23
to hon...@googlegroups.com

Dear Tom and everyone else who is interested in AI or machine translation.

At the risk of making a contribution again that may seem to be off-topic, I would challenge this comment:

“Perhaps the biggest need for translation in a hospital is to enable patients and their families to feel confident that they have been understood and are being treated appropriately. For that purpose, a human interpreter would definitely be better than an app.”

I think this is both wrong, and another indication of why many translators should legitimately fear being replaced by AI, while others should have no fear at all.

The ones who should worry are too close to the process of drafting translations, and don't spend enough time standing back and looking at how translation enables larger processes. The latter is the only thing people pay for. This is the value we add.

I have argued this point before and found little in response on Honyaku. But I offer it again in the hope that it provides people with a more positive view of their work and of their future.

The biggest need for translation in a hospital is to protect the interests of the hospital and the patient. Patients rarely bring their own interpreters, and so the interests of the hospital generally drive the interpreting budget. Hospitals for example have an interest in minimising their exposure to potential litigation.

So one of the most conspicuous examples of why a human interpreter is the only possible solution is in the translation of the information required for a patient to give informed consent to a surgical procedure. (in Australia if informed consent is not obtained then the procedure constitutes assault). This is true no matter how low the quality and speed of a human interpreter is compared to what Google translate produces.

Hospitals will pay for professional interpreters not to obtain necessarily accurate and faithful translation, but so they don’t have to say under oath in court in a malpractice litigation “Well your honour the patient seemed to understand this gesture..”

They are buying the right to say “Nothing to do with me Your Honour, please examine the translator, they took full responsibility for the translation.”

It offends me as a language geek, but to fail to understand that that is our product, is to be a slave on a sinking ship of language geeks.

AI will never replace human interpreters. The reasons are not linguistic. They are legal. AI cannot be sued. Whereas a person, subject to the terms of a commercial agreement between them and an incorporated entity such as the hospital or whoever their client is, can be held to account for the quality of their work.

That is how large organisations like hospitals outsource risk. They outsource it to the people with a professional skill set that will enable them to most competently manage that risk.

How competently they do or should is a secondary matter. Discussions about grammar and pronunciation and intentionality and so on, are surely the nuts and bolts of translation I agree, and we can sit around discussing those things and we do so on Honyaku. They are all interesting discussions.

But that is not the definition of our job. If you think it is, you will be replaced. If you can make yourself a manager of all the risks associated with the language barrier, you won’t.

This applies to all our work, not just medical interpreting.

Chris

Tom Gally

unread,
Oct 28, 2023, 5:07:01 AM10/28/23
to hon...@googlegroups.com
Chris and all,

Those are all interesting points.

After reading them, I wondered how much the emphasis on protection from legal exposure in Australia also applies in Japan, so I did a little Googling.

The following site for a company selling medical interpretation services lists 医療通訳を利用するメリット without mentioning legal protection for the hospital:


The following paper refers to lawsuits in the case of bad interpretation only in non-Japanese contexts:

家族や知人などが行う通訳は「にわか通訳」「ad hoc 通訳」 などと呼ばれ、誤訳が多く、医療専門用語(以下、専門用語)を知らず通訳者自身が医師の説明を理解できなかったり、 医師に病状を充分伝えられなかったりしたことが報告されており、海外では適切な医療通訳者を確保できずに、訴訟や死亡事故につながった事例も発生している。(p. 306)


On the other hand, the following, more detailed site does name lawsuit avoidance as one reason for hiring medical interpreters:

外国人患者の診察においては，医療通訳を介して，患者が理解できる言語で説明・同意を取ることがトラブル・訴訟予防に有効である.


It’s more than ten years old, but the following document goes into a lot of detail on legal issues with foreign patients:


Here’s one more interesting item:

「医療通訳制度」の必要性が全国的に言われるようになったひとつのきっかけに、ある公立病院での医療訴訟がありました。東南アジアから来日しコンピュータ関係の仕事をしているある患者さんが、手の手術をしなければならなくなりました。公立病院の医師は英語ができ、患者さんも英語ができるため、お互い手術の内容とリスクについて、「説明と同意」（インフォームドコンセント）を英会話で行いました（行ったつもりでした）。ところが、手術後に手の神経に痛みが残ったことについて、その患者さんは「充分な説明を受けていなかった」と訴え、医療訴訟に発展してしまったのです。結果論ですが、医師側の話す英語と、患者側の話す英語で、手術後のリスクについての充分なコミュニケーションは取れていなかったことが明らかになったのです。


Tom Gally


John Stroman

unread,
Oct 28, 2023, 6:58:42 AM10/28/23
to hon...@googlegroups.com
In response to Chris and Tom, one of the main takeaways from my initial post on the thread  "Say No to term full post-editing,'" which provided a link to a 42 min video interview with Slator, was that no one has ever successfully sued a machine for a translation error. The interviewee is an agency manager based in Slovakia, and his main beef is against the term "full post-editing" that appears in ISO certification regulations, which he considers meaningless.  He contends that an MT/AI leveraged translation that has been reviewed and corrected by an experienced translator with field-specific expertise is equivalent to "human translation" and should be compensated as such. 
Link: https://www.youtube.com/watch?v=DjQiw-e_PEw
(if you choose to watch it, start about 15 min into the interview to avoid the initial unrelated chit-chat)

John Stroman
----------------
To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CACnS3Ne2d5YgOQTTNJ--MV%3DYtSuFzHzkp0kXwr4qU-0g3Ni9ug%40mail.gmail.com.

cpta...@ozemail.com.au

unread,
Oct 28, 2023, 7:20:51 AM10/28/23
to hon...@googlegroups.com

Hi John,

Wow sorry I didn’t read that thread. I will have a look at the vid.

Chris

 

From: hon...@googlegroups.com <hon...@googlegroups.com> On Behalf Of John Stroman
Sent: Saturday, October 28, 2023 9:58 PM
To: hon...@googlegroups.com
Subject: Re: Japan’s Hospitals Need More Professional Medical Interpreters

 

In response to Chris and Tom, one of the main takeaways from my initial post on the thread  "Say No to term full post-editing,'" which provided a link to a 42 min video interview with Slator, was that no one has ever successfully sued a machine for a translation error. The interviewee is an agency manager based in Slovakia, and his main beef is against the term "full post-editing" that appears in ISO certification regulations, which he considers meaningless.  He contends that an MT/AI leveraged translation that has been reviewed and corrected by an experienced translator with field-specific expertise is equivalent to "human translation" and should be compensated as such. 

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

cpta...@ozemail.com.au

unread,
Oct 28, 2023, 7:33:40 AM10/28/23
to hon...@googlegroups.com

Very interesting Tom and thanks for those links. I am indebted to you for the Japanese term for “ad hoc interpreting” (Please see my English definition here) https://pooletranslation.com.au/file/7636/776

To get hardcore about this though, you need to completely separate the legal liability question from the 「誤訳が多く、医療専門用語」question. I know it will grind the gears of anyone who has spent the last thirty years getting their words right, but the latter in not the issue. It is that any “bilingual subject matter expert” i.e. the natural enemy of all translators and interpreters, if they get something wrong, can legitimately say “Well, it’s not my job”. That is where the function of professional translation has completely fallen through the cracks of all kinds of governance, and the bulk of our efforts should be concentrated on this pain point if we are ever to establish T&I as a respected and remunerated profession.

Reply all
Reply to author
Forward
0 new messages