Dear Reza

Salam

I routinely include non-English articles in all of my systematic reviews. The process is very simple conceptually: do not restrict your search in the databases you are searching in, namely PUBMED, SCOPUS, ISI, etc.

Usually by reviewing title and abstract (which is always available in English) you can decide which article is relevant and which is not. Be sure to consider duplicate publications as many groups have published duplicate results in English and non-English articles and you do not need their non-English articles to be included in the systematic review.

The next step is to find the full texts. It is a little bit hard however usually the publishers will provide full texts and many of the non-English articles are open access in their website.

The last step is data extraction. You have two different pathways:

1- either the article is in pdf and in fonts and you can copy and paste its content. The best way is google translate site: translate.google.com. It works for all languages (thus far I have included and extracted data in Spanish, German, Czech, Greek, Russian, Turkish, Thai, Japanese, Chinese, Korean, French, Hungarian, and Polish.) Be advised that the translation is not perferct but gives you the data you need to perform the systematic review and meta-analysis

2- or the article is in pdf in pictures (the scaned pictures of the articles have been provided for you). This is a major problem as you can not copy and paste the data in translate google. What you need to to is to OCR the picture to text. I use "finereader" software. After conversion of picture to text then you can follow the next step above. This is important for Chinese, Japanese, Korean, Russian, Greek, and Hebrew articles as you can not identifiy the script.

Best wishes,

Ramin Sadeghi

Let me recommend you the document on systematic reviews from the Centre for Reviews and Dissemination in the University of York (UK) http://www.york.ac.uk/inst/crd/pdf/Systematic_Reviews.pdf

Dear Reza

Salam

I routinely include non-English articles in all of my systematic reviews. The process is very simple conceptually: do not restrict your search in the databases you are searching in, namely PUBMED, SCOPUS, ISI, etc.

Usually by reviewing title and abstract (which is always available in English) you can decide which article is relevant and which is not. Be sure to consider duplicate publications as many groups have published duplicate results in English and non-English articles and you do not need their non-English articles to be included in the systematic review.

The next step is to find the full texts. It is a little bit hard however usually the publishers will provide full texts and many of the non-English articles are open access in their website.

The last step is data extraction. You have two different pathways:

1- either the article is in pdf and in fonts and you can copy and paste its content. The best way is google translate site: translate.google.com. It works for all languages (thus far I have included and extracted data in Spanish, German, Czech, Greek, Russian, Turkish, Thai, Japanese, Chinese, Korean, French, Hungarian, and Polish.) Be advised that the translation is not perferct but gives you the data you need to perform the systematic review and meta-analysis

2- or the article is in pdf in pictures (the scaned pictures of the articles have been provided for you). This is a major problem as you can not copy and paste the data in translate google. What you need to to is to OCR the picture to text. I use "finereader" software. After conversion of picture to text then you can follow the next step above. This is important for Chinese, Japanese, Korean, Russian, Greek, and Hebrew articles as you can not identifiy the script.

Best wishes,

Ramin Sadeghi

Thank you very much everyone, your responses have been very helpful.

Ramin, Wa'Alaikassalam.. Thank you very much for a very detailed step-by-step description.. I will certainly work along these lines...

Just wanted to know if you have any experience with databases other than the regularly used ones like medline and scopus etc. i have found articles in chinese through a database of chinese language articles that i cannot access the full text for using google translate feature but will certainly try to go through the publisher website..

regards, raza

In addition to Ramin´s response - indeed ,sometimes there are duplicates of articles, but I want also to put your attention of so called - additional publications. You may come across on a publication, which uses a subgroup of the same sample of another study. In other case you may find an article which describes the extension phase of previous trial. In both type I recommend you to use those additional publications as a valuable source of additional data and often a source of info about the methodology, as the number of words in journals is limited.

@Raza:

It is really good to see somebody acknowledging that we rarely include foreign language reviews in our searches - this goes both ways for English & non-English speaking countries.

My interest was sparked for 2 reasons: firstly, I came across a PhD thesis on health & safety in nursing, from a small European country. They claimed in their literature review that virtually no literature existed for the topic & therefore their approach was innovative, new, much needed, ground breaking etc ..... (this is often the case when we search solely in our own languages and do not acknowledge anything that is outside them).

I have also conducted & published 'comprehensive' & 'state of the art' reviews in my own field of poster presentation & retrieved very little in the way of formal results. I therefore asked myself, that if poster presentation is a worldwide media of knowledge transfer at conferences, then surely so many people must be voicing their opinions outside the restrictions of formal systematic reviews.

I am currently completing a search & compare returns in the world top 20 languages from formal databases, Google, Google Scholar & other search tools. I am advocating a new type of review that looks to include (rather than exclude) & uses innovations such as auto translation. Most of the searching is done & it just needs categorizing & writing up. Given the diverse range of 'data' we now have access to, it is perhaps time to reconsider if 'robust' means 'exclusive', i.e. not every worthy study is an RCT and just because Stephen Hawking made some interesting comment in a common-old-garden documentary meant for the masses or a blog, it does not make his professional opinion any less noteworthy or valued.

It is interesting the York document (http://www.york.ac.uk/inst/crd/pdf/Systematic_Reviews.pdf ) says that:"An investigation of the inclusion of non-English language reports of RCTs

in systematic reviews concluded that language restrictions do not appear to bias

the estimates in reviews of conventional interventions, but may bias the results of

complementary or alternative medicines". Although contextualized in medicine, this can also be read that the 'mainstream' topics of cancer, genetics, nan-tech etc, are more likely to be sufficiently represented in English to make reviews valid, but understudied areas (like poster presentation ;-) ) are more likely to be biased or non representative, if we do not adopt a wider approach to literature retrieval. Of course, it is preferable to follow the York papers advice & include non-English work, but it may also be tempting to then simply put a red pen through them & write 'language' as an excluder if we are not inclined to lower ourselves to using rudimentary and imprecise technology ...

Send me a message if you are interested in pursuing this topic further.

Quid-pro-quo ... do you know anyone who can translate 'Poster Presentation' into Lahnda ??? 78.3 millions speakers, yet it is really difficult to find auto translation tools that include it :-)

Thank you Nicholas for highlighting this interesting aspect of research... Your idea of compiling information from various databases is quite impressive..

I tend to disagree with the views of the York document since looking at a recent topic of interest, There has been meta-analysis describing no benefit of the intervention looking at english only studies. However, a recent meta-analysis looking at english and non-enlish literature found a significant benefit from the intervention. This alone is sufficient for me to pursue non-english literature for a systematic review.

I must admit my ignorance of Lahnda language... but looking it up on wikipedia, i figured the two common spoken versions are quite familiar and i can certainly get in touch with people who can translate things you need. Will get in touch with you as soon as I hear from them.

regards, raza

I have a Russian paper to translate. Problem is it is written in cyrillic. Can copy and paste into translaters such as Google but the text is just a jumble of different letters and not cyrillic so the translation does not work. Any ideas?

@Anthony Frais, if you are looking for professional translation feel free to check out Academic Language Experts (www.aclang.com).