2
$\begingroup$

In keeping with the guidelines of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) Statement, I am trying to fill in their flowchart (Fig. 1) with the details of the literature search performed for my meta-analysis.

That template asks for the "# of records identified through database searching", and another one for "records identified through other sources" - however, most meta-analyses I've seen report having used multiple databases and/or search engines for the literature search, e.g. Google Scholar, PubMed, Web of Science, etc.

So how is one meant to arrive at the two numbers? Are researchers who wish to follow the PRISMA Statement meant to only decide upon one primary database, and report the number of records returned by that one database in the first box, then sum up the number of records returned by all other databases (or non-database sources such as manual perusal of reference lists) for the second box?

$\endgroup$
3
$\begingroup$

'Other' refers to grey literature, hand-searching conference abstracts, trial registries, etc. Things that aren't indexed in bibliographic DBs. Some people report the total number of citations across all bibliographic databases searched and then below it in the same box will break down the numbers per database. For example:

3000 records identified through database searching
 Medline (n = 1000)
 Embase (n = 1500)
 Cochrane (n = 500)
$\endgroup$
3
$\begingroup$

The first figure is for all the bibliographic data-bases. It is usual to report somewhere how many remained after stripping duplicates. The second figure is for all other sources and it is best to report for each source like reference lists, internet searching, expert advice, and so on.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.