What are the recommended databases? Any consensus available?
It depends upon the topic and the field of study but for a systematic review on a medical topic the minimum number of databases should include the following
MedLine: https://pubmed.ncbi.nlm.nih.gov/
Scopus: https://www.scopus.com/
Cochrane Library: https://www.cochranelibrary.com/search?cookiesEnabled
The Cumulative Index to Nursing and Allied Health Literature: https://health.ebsco.com/products/the-cinahl-database
Google Scholar: http://scholar.google.com
Thanks all for your recommendations - I have found this interesting paper, I guess it can give us all a better idea.
https://pubmed.ncbi.nlm.nih.gov/29208034/
It depends on the question and the field, but some journals, e.g. BMJ Open doesn't accept articles with under four databases.
A comprehensive search should include searches on the relevant databases within the field, e.g. in biomedical research at least PubMed/Medline, Embase, Scopus, Web of Science, and Cochrane. Furthermore, a search in the grey literature e.g. ProQuest and OpenGrey should be conducted. And finally, you should do a backward snowballing where you look at the references in the included studies.
An article about the number of databases has been published: Article Optimal database combinations for literature searches in sys...
AMSTAR-2, a validated tool to asses systematic reviews, values the use of at least 2 databases for the relevat research question. I would say that is a fairly realistic benchmark. All the best for your research!
A minimum of 2 databases.
We always search Medline, Embase and Cochrane library (as a minimum).
you could search many database .Systematic review allows you to choose from as many types of database published unpublished papers.The only terms are that rather than quantity you choose quality.All papers should have the statistical value you want for meta analysis after systematic review.I have come across many A grade journals in my domain mainly social science aggregating 700-1000 papers and then rejecting on the basis of abstract research question language etc and then doing a research of mainly 30-50 papers only for meta analysis.
I think that depend on the field of your topic. in my opinion , OSA, IEEE, Springer, and Elsevier are a good database.
Andrew Abue This is true, there really isn't a consensus. That's why reference searching and snowballing is so useful at tracking down results that may not have come up in your search. I don't think there's such a thing as a perfect search either, getting the balance between sensitive and specific, deciding which filters or limits to use or not use etc.
In my opinion, you should access at least 2 databases. Your systematic review is likely to be considered rigorous and accepted by a good journal if it covers 2 or more databases. Of course, like people have mentioned here, some journals have very strict rules of 4 databases and above, but as a rule of thumb, 2 databases should be a bare minimum. You could include PubMed and google scholar, if not anything else. Apart from these Cochrane library is also freely accessible and has high quality articles. The ultimate purpose of a systematic review is to collate as much relevant literature as possible. It is only fair to try to search through 2 databases or more. :)
There is no perfect search though and you would never find a review that would have covered all possible articles on a subject ever available online. It is also important to not end up with an unmanageable number of hits. A lot of factors come into play. Hope this helps!
When i did Cochrane course, they recomended 5 databases. So I peformed my systematic review folowing this, and worth it because i found different articles in inch database, of course elevate the numbers of duplicates but it was manageable with endnote software.