Karen Boyd, ASQ CQA Great topic, John! Does accuracy factor in with that #1 of consistency concerns?
I look forward to viewing all the responses as well.

Karen Boyd, ASQ CQA As a side note and potential solution, I think the more usage / integration of harmonized international symbols would help ease and streamline content.

John Herzig Hi Karen - in response to the first comment: Accuracy absolutely factors into my point 1. Frequently, in larger medical device companies, we see situations where the web marketing team may be working with one vendor (or many) and the tech pubs team with a different one (or ones!), and the accuracy of translations may not be the same.

The challenge is the same with consistency. For example, if the medical device manufacturer has a style guide and maybe technical glossary for core terminology, brand voice, currency, etc. it can be a challenge ensuring that it is applied consistently across outputs when multiple translation firms are involved.

And I completely agree with the use of harmonized symbols to decrease the amount of translatable content, and thereby mitigating the risk associated with translation.

Thanks for the comments!

Stuart Briefer I don't see anywhere here discussions on documentation of procedures, patient response, vital signs taken at certain intervals, and a way to save this documentation without future modification to insure reliable data gathering from the onset

Marcus Gould Hi John

Thanks for the discussion, and for everyone's comments. Very interesting.

One important, but overlooked element of translation, is the possible change in the usability risk when an IFU, or System Operating Manual, has been translated. When these large bodies of text are compiled they, of course, should go through usability studies/human factor studies to ensure the text is readily understood by the intended user. These studies are usually conducted on text written in the English language (after all a copy of all labels and labelling need to be submitted as part of the regulatory submission, and this is often in English). When the text has been reviewed and ‘approved’ by the regulatory body (be it the FDA, a notified body, Health Canada etc.) it is then, dependent upon the market requirements, translated into the relevant language using the English text as the source language. How often is the translated text then taken back through the preceding usability studies/human factor studies to determine if the text, throughout its translation, shall lead to the device being used in a different manner?

I have heard of a number of devices, which when placed on the Spanish market, and supplied with Spanish Instructions for Use (IFU), were used in a different manner to those placed on markets supplied with English Instructions for Use. It was originally thought regional, cultural and clinical differences were the root cause for the same device being used in a different manner, however an investigation determined that, during the translation of the text, some elements of the safety and warnings information inadvertently changed enough that it was used in a manner different to that it was designed for. If the Spanish IFU has gone back through the usability studies/human factor studies used to compile the English language IFU, this would have probably been noticed and rectified.

The process of the translation of text, especially for medical devices, may, without knowing it, actually increase the risk of a device. This needs to be factored in when determining translation needs.

Marcus

John Herzig Hi Marcus - GREAT observation. This is indeed a challenge in localization where client budgets and timelines may not allow for usability/human factor studies. Though as you rightly point out, when those are not conducted in local language, the risk factors may indeed increase. Thanks for the insight!

John Mitchell Do you use blinded circular translation? That is, have one person translate something from original into local language, and then have a different person translate from the local translation back to the original. Next, compare the two originals side-by-side to identify losses of fidelity. Finally, follow-up with the translators to determine why the translated certain words the way they did.

This has been our standard practice when localizing market research studies--especially in medical devices where the nuances of clinical terminology can really make translation challenging--and we find it works quite well.

John Herzig John - thanks for the question and observation. Yes, we do perform a similar process of Forward Translation, Back Translation, and Reconciliation depending on the content type, level of risk associated with the translation, etc. I agree that it works well. Thanks for weighing in!

John Mitchell You're welcome. Good luck.

Robert Barber Apart from the 3 items listed above, one of the main issues is that companies often rely on local distributors to perform translations of Instructions For Use - and local regulatory requirements can vary from those of the originating country - so as well as ensuring the translation is accurate, regard also has to be given to regional text differences to comply with each regional regulatory requirement. For example, changes in warning statements require review by the regulatory authorities in Japan, but not in most other countries, so co-ordinating the timing of a change can get complicated if one country is demanding a quick change (for good safety reasons) while another enforces a review for the change to be officially sanctioned (which takes time). Another issue is that accurate translation is a critical safety factor - how do you qualify the competence of translators within your own organisation? Are your external translation companies subject to supplier selection and evaluation (they probably should be as a critical supplier)? How do you really determine their competence if you are not bi-lingual (including technical jargon) yourself? There are still debates about the equivalence of English, German and French versions of even the EU Directives, so what hope is there for those of us who only speak one language?! I feel especially for non-US companies that are required to present all their documentation to FDA in English - and I don't believe FDA yet provides non-English translations of its regulations. (If anyone knows different, please share). Having said that, there is an urban myth that the country that has the biggest misunderstanding of the FDA regulations is the UK - American English isn't necessarily the same as English English! No wonder we all get confused!

John Herzig Robert - those are keen observations! You are 100% correct that not only do the translations need to be perfect, but we also need to consider local regulatory requirements for labeling. Thanks for sharing your thoughts!

Marie Suetsugu With regard to Karen's emphasis on 'harmonized international symbols', I'd be interested to hear what people are doing with software devices, where there is so much non-symbolisable information in GUI...

One issue I have observed over the years is ensuring the meaning of the translation at the local level resonates. Words and symbols are valuable and different amongst languages. Ensuring the meaning is communicated (along with regulatory guidelines of course) adds more value.

John Herzig Hi Flora - Great insight! I completely agree that good, local translators (with relevant medical experience) should be used at all times!

As a translator and localiser, a good one, I can tell you that it is key to find a local translator, to invest in the relationship and not to rely on other companies to arrange for translations. Trying to cut corners by bargaining online with unknown translators at rock bottom rates leads to disaster. I can hardly believe that good translators are hard to find. For English to Dutch, just google translatedutch.nl