Before you undertake your search there are three important pieces of information researchers need to remember.
The primary databases are the key databases used for undertaking systematic searching. Which database you use is dependent on the research being undertaken.
Primary databases contain records for a wide range of original (primary) research. There are key databases that are regarded as having relatively comprehensive coverage in specific disciplinary areas such as Ovid Medline and EbscoHost CINAHL.
The narrower focus of these primary databases compared to other database types allows for greater focus on subject headings and research methodology used in the included research.
Some databases are multidisciplinary and cover multiple subject areas, such as Proquest Central.
Multidisciplinary databases can also be useful to search, as these typically index material from diverse subject areas. Searching one or more of these is typically done to increase the sensitivity of the search.
Other databases are discipline specific and only contain resources from a specific subject area. HeinOnline, for example, is a legal database with content from Australian and international law resources.
Full-text databases have the record and the full text for the resources. Not all databases are full text.
Databases that only contain records for resources are called indexing or citation-only databases. SCOPUS and Web of Science are examples of indexing databases.
Indexing databases while useful for an early review of literature on a topic are not suitable for use in a high level review. They lack the elements, such as controlled vocabularies of medical subject headings, that are needed to approach a search in a highly structured manner.
Hand searching is the manual method of searching through selected journals or books, page by page, from cover to cover looking for information relevant to the high level review question being investigated, e.g. checking reference lists of journal articles or book chapters. This technique is also known as snowballing, reference harvesting or pearl growing.
Hand searching can assist researchers in:
Researchers should always consider grey literature when undertaking a high level review. The information available via this avenue is important in showing that researchers have taken a rigorous approach to searching all the evidence available in answering the research question.
Some of the information that can be found in searching grey literature are results from studies where the trial or study indicated a negative impact or no effect in the results. Papers or reports which have a less than positive outcome are less inclined to be published in scholarly journals, this known as publication bias.
As research takes a long time from initial planning to publication, study results are often presented at conferences giving researchers the opportunity to disseminate their findings before appearing in a peer-reviewed publication.
Note: Researchers who use grey literature when undertaking a high level review need to ensure the information is credible and does not show bias. Grey literature by its nature will need to be found as part of hand search and not the structures search strategies.
Jess Tyndal from Flinders University developed the AACODS checklist (Authority, Accuracy, Coverage, Objectivity, Date, Significance) to assist in evaluating grey literature.
Cochrane Handbook, part 2, chapter 4, section 4.3: sources to search
This chapter outlines the gold standard guidelines for SR searching and contains a section recommending sources to search.