Why, asked the medlib naif? - Steele Lawman

My main reason is reproducibility - PubMed is freely available to anyone to use. Ovid MEDLINE is part of the Ovid subscription. While Ovid (and prob some other subscription versions) has some additional options like adjacency searching, I find using a public database preferable, and personally find PubMed searches to be more readily understood when reported, especially as a "whole" search rather than line by line pieces like the one I'm looking at now. - Rachel Walden

[This is also why this library teaches PubMed to students and residents instead of Ovid MEDLINE - because they will continue to have access to PubMed no matter where they end up] - Rachel Walden

I don't like the Ovid interface. - Hedgehog

I don't like it either. - Rachel Walden

So, the output, the product is search results? That's not an interim step that ends on the cutting-room floor? Because in the disciplines I'm familiar with, no one would say "I searched MLA International Bibliography using the EBSCO platform to find Nabokov criticism." - Steele Lawman

Well, the product would be a systematic evaluation of and reporting on the existing research. The whole point would be to say "here's what currently exists as published evidence on these key medical questions, how good it is, and what it found." For medical systematic reviews, the search methods are a piece of the overall research methods, because if you searched the wrong places, searched poorly, omitted relevant terms, etc., then everything that follows is suspect. - Rachel Walden

So knowing exactly where you searched matters, because future researchers might do an update of your review, or want to reproduce your methods, and the different handling of a MEDLINE search on different platforms makes the resulting set of citations to deal with different. - Rachel Walden

Here's an example report - http://www.ncbi.nlm.nih.gov/books... - with search methods at http://www.ncbi.nlm.nih.gov/books... - and some more background http://en.wikipedia.org/wiki... - Rachel Walden

Excellent, thanks for the lesson. - Steele Lawman

No problem! I do a lot of work with systematic reviews as part of my job here, and am facilitating a roundtable on the topic at MLA next week. Happy to answer any questions about it. - Rachel Walden

We had a similar debate over ERIC - Ebsco, which students find easier, or the free one, which students can use after graduation? I lost. So did public ERIC, which didn't try very hard. - barbara fister from iPhone

"public ERIC, which didn't try very hard" - lol, too true. :) - Rachel Walden

In 2009-2010, I was looking in to "evidence based librarianship" and wound up co-authoring a systematic review. We had a hard time explaining to authors of the Journal of Academic Librarianship and College and Research Libraries that it was not simply a literature review, but an actual research methodology. Even still the editor of the journal that did publish it asked us to take out our search strategies because they said, "everyone will have their own way of structuring the search" and therefore would be judging our strategy. - LibrarianOnTheLoose

that's hilarious - Meg V. Meg

uh huh - LibrarianOnTheLoose from BuddyFeed

Maybe the title should have been "Using the Methodology of Systematic Reviews to Examine Evidence-Based Librarianship" with a paragraph explaining what this means and why EVERYONE should do it. A lot of lib lit seems to be arguing for a methodology borrowed from another discipline as much as for its findings. But evidently it needs to be labeled "hot new methodology" or it will be seen as just a weird way of doing it. - barbara fister

Rachel if one wants to start learning how to do medical systematic reviews do you have any recommendations on what to read? How difficult is it to do a decent one? - aarontay

Hey, Aaron, catching up from MLA, but will share some stuff soon. Remind me if I don't get this done in a few days. Barbara - in my situation, we are doing systematic reviews on medicine, for publication in medical journals - so it's not applying another discipline's methodology, it's working collaboratively in that discipline, just to clarify. - Rachel Walden

Jumping in quite late here, but really, a good systematic review could probably use both Ovid MEDLINE and PubMed. And Embase. And WoS. And Scopus. And whatever other database is relevant. - mlrethlefsen

Feel a bit shy asking but... Rachel anything to share....? I do notice we have quite a few books on that.. Maybe I should just crack open a book and read. Hmm there's a systematic review for social sciences? http://linc.nus.edu.sg/record=... not just medical... - aarontay

Aaron - for starters, the PRISMA statement is a good thing to be aware of - it is a checklist of what "should" be in a systematic review: http://www.plosmedicine.org/article... Also, Cochrane puts their review methods online - http://www.cochrane-handbook.org/ - although of course other groups like AHRQ may have slightly diff methods. For examples, AHRQ is a good one. - Rachel Walden

I don't want to overload you, but am happy to follow up with more. - Rachel Walden

Thanks Rachel you have being very kind. We don't provide support for medical users who want to learn how to do systematic studies might want to ask someday, so good to learn. I hope you don't mind me asking Rachel presumably your library teaches users how to do systematic reviews. How did you learn how to do that? Was it covered in library school or did you learn on your own by doing or you were mentored by someone? - aarontay

I have learned a lot by being mentored and through our work - my institution has one of the AHRQ evidence-based practice centers, and we heavily involved with that. Internally we have a lot of mentoring on study design, biostats, evaluating published research, etc., but it has been through our EPC work that I've really been able to see the process start to finish. Unfortunately I'm not sure AHRQ's methods are online the way Cochrane's are. - Rachel Walden