Depending upon your topic, it may be appropriate to search the grey literature and/or snowball your search.
Snowballing, also known as citation chaining, is the process of searching the references and/or the citations of your articles to identify other relevant material. Indexes such as Scopus and Web of Science enable you to carry out this procedure.
Grey literature refers to material not usually found in the commercial publisher databases, e.g., government reports, conference proceedings, theses, etc.
British Standards Online (BSI Online)
DTIC Online (Defense Technical Information Center)
espacenet (European Patent Office)
EThOS - Electronic Thesis Online Service
Open Access Theses and Dissertations (OATD)
SAE Aerospace Standards and Materials Specifications
United States Patent and Trademark Office (USPTO)
As with your search-string searches, it is important to document the process you went through to locate additional studies through snowballing and/or ad-hoc searching of the literature.
Any studies identified through snowballing and/or ad-hoc searching, that are to be included in your review, must be evaluated against your inclusion and exclusion criteria at the full text screening stage and against the quality appraisal checklist in the quality appraisal stage.
The number of articles located through these processes should be separately documented at each of these two stages, along with the number of articles from your search string searches.
Here is an example of this process:
At the full text-based screening stage:
articles from search strings (n = 246), articles from snowballed searches (n = 12), articles from ad-hoc searches (n = 8).
Once the full text screening has been completed:
rejected articles (n = 134).
(Note: you also need to document the basic details of the studies excluded, as well as the reasons for exclusion)
At completion of the quality appraisal screening stage:
rejected articles (n = 35).
(Note: you also need to document the basic details of the studies excluded, as well as the reasons for exclusion)
Final number of articles left:
articles from search strings (n = 86), articles from snowballed searches (n = 7), articles from ad-hoc searches (n = 4).
Total articles selected for the review (n = 97)