Systematic reviews involve one major trade-off:

1. You basically have to find everything, or look like you would
2. You don't want to screen so much rubbish that you never finish the review

To navigate the trade-off, there broadly two approaches.

The 'conceptual approach' to finding literature (don't do this!)

1. Brainstorm a list of search terms
2. Pull terms from existing reviews
3. Peer review your search strategy
4. Run your search and complete your screening
5. Complete forward and backward searching to completion
6. Ask authors for any other studies

This approach is fine for experts who have published in a field for a long time. They know the literature already, what the jargon is, what they're trying to find, etc. Hence, it's recommended by Cochrane. But, even experts do this terribly.

In four systematic reviews, the conceptual approach failed to identify 28 of the relevant MEDLINE citations

Translation: experts creating searches in their own field can't find the research they know exists.

Hausner, E., Guddat, C., Hermanns, T., Lampert, U., & Waffenschmidt, S. (2016). Prospective comparison of search strategies for systematic reviews: an objective approach yielded higher sensitivity than a conceptual one. Journal of Clinical Epidemiology, 77, 118–124. https://doi.org/10.1016/j.jclinepi.2016.05.002

This method fails because:

• You don't need to know what you're looking for before you run the database search. I've had students run a whole review and find nothing.
• You tend to add keywords that do not find what you're looking for, but do find lots of other rubbish
• You can't tell if the keywords you're using actually find the papers you're looking for without running the search.

In short, don't do the conceptual approach on your first review. Be wary of it, even on your 10th.

The 'objective approach' to finding literature (do this)