Entries in Systematic Reviews (24)


Systematic Review Filter

I'll be running an SR search for SRs later this year.  Will revisit this.


Objectively Developing Searches

The authors suggest using text mining software to objectively develop searches.  I agree with the approach, but don't think it fundamentally differs from what should already be standard practice; that is, using a test set of relevant articles to a) harvest text words and controlled terms and b) objectively validate the search.  Both can be accomplished manually, though a program would likely be more efficient and comprehensive (& probably more accurate, reproducible, and systematic now that I think about it).  Worth a read.  


How To Review A Search

An interesting conversation via Evidence-Based Health.  I've wondered how closely (and effectively) systematic review and clinical guideline searches are scrutinized during peer review. My guess is not very.  But given their significant role in such projects, they certainly warrant careful review.  Fortunately, as someone mentioned in the exchange, exceptional research has been done in this area:


It's a great paper. The authors offer a number of recommendations that I think provide a nice framework for critically evaluating searches.  Definitely worth a read.  


Edit: Also take a look at the checklist the authors put together based on their research.  Another fine contribution.  


EPID 757 - Introduction to Systematic Reviews and Meta-Analyses

Earlier this year, I was invited to provide a lecture on the systematic review search process for a course in the University of Michigan School of Public Health.  The course materials, including my lecture, are now available as part of the Open.Michigan initiative, which facilitates the opening up and sharing of locally created educational resources.  Check out the site.


Literature Search Reporting in Systematic Reviews

Nice study on search reporting.  The meat of the discussion is basic for librarians, but it's nice to see the topic get acknowledged in an medical journal.
The International Committee of Medical Journal Editors (ICJME)’s Uniform Requirements for Manuscripts Submitted to Biomedical Journals instruct authors to “[i]dentify the methods, apparatus (give the manufacturer’s name and address in parentheses), and procedures in sufficient detail to allow others to reproduce the results."  Although all three of the journals included in our study recommend that submitting authors follow ICJME guidelines, none of  the reviews we examined provided descriptions of their search strategies in enough detail as to be—based on our criteria—reproducible. Most often, the absence of the Boolean operators used to combine search terms and the absence of a specific search date in reporting precluded the ability to reproduce a search. These absences are not unique to medical education: A study of randomly selected meta-analyses in clinical medicine reported that only 6.7% reported a search strategy in enough detail to be reproduced.