For each database you search, the structure of the search strategy remains the same. However, all databases have a slightly different search interface and offer slightly different search options. It is important that you explore the search options for each database you will search for the systematic review, and adapt the search strategy accordingly.
Most databases use unique controlled vocabulary to index their articles - for each database adapt the controlled vocabulary terms. For more details on the controlled vocabulary terms for different databases, see Step 4.
The fields you can search free-text terms (or the combination of them) vary between the databases. For more details on the search fields available for different databases, see Step 5.
Different databases offer different wildcard characters and commands. For more details on differences between databases for phrase searching, truncation, wildcards, Boolean and proximity operators, see Step 6 and Step 7.
The types of pre-set limits/filters vary across the databases. For more details on limits, see Step 8.
The Polyglot Search Translator is a tool for translating search strings from PubMed or Medline via Ovid across multiple databases. It has been developed by the Institute for Evidence-Based Healthcare at Bond University.
Clark, JM, Sanders, S, Carter, M, Honeyman, D, Cleo, G, Auld, Y, Booth, D, Condron, P, Dalais, C, Bateup, S, Linthwaite, B, May, N, Munn, J, Ramsay, L, Rickett, K, Rutter, C, Smith, A, Sondergeld, P, Wallin, M, Jones, M & Beller, E 2020, 'Improving the translation of search strategies using the Polyglot Search Translator: a randomized controlled trial', Journal of the Medical Library Association, vol. 108, no. 2, pp. 195-207.
As each database has its own indexing system, you need to identify the controlled vocabulary terms for each database separately.
These are the controlled vocabulary terms for Concept 1 for some of our UTAS health databases (note the slightly different spelling):
|MeSH Terms||PsycInfo Thesaurus||CINAHL Headings||Emtree Terms|
Also, note how the syntax between different databases varies, even if they use the same controlled vocabulary; e.g. demonstrated for the three databases that use MeSH as the controlled vocabulary:
MeSH descriptor: [Dementia] explode all trees
MeSH descriptor: [Alzheimer Disease] explode all trees
Not all databases offer users to customise the search fields as freely as Ovid. Simply try to keep the search fields as consistent as possible. This table lists similar options in UTAS Health databases:
|Medline via PubMed||dementia [tw]|
|Medline via Ovid||dementia.ab,kf,ti.|
|CINAHL||TI dementia OR AB dementia|
|Web of Science||TOPIC: (dementia)|
|Scopus||TITLE-ABS-KEY ( dementia )|
Some databases offer a wider range of proximity operators and wildcards than others; plus the function of the symbols used varies between the databases. It is best to
Here are a few examples for the example search to illustrate the difference of symbols and meaning.
? stands for zero or one character within a word
* is used to match multiple characters within a word
|CINAHL||behavio#r.ab,kw,ti.||# finds all citations of the word that appear with or without the extra character
finds behavior and behaviour
|Web of Science||TOPIC: (behavio$r)||
$ useful for finding British and American spellings of same word
|Scopus||Scopus finds common American/British English variant spellings automatically|
|Ovid databases||animal adj2 therapy.ab,kf,ti.||animal and therapy need to be within 2 words of each other: one or no word between them|
|Cochrane||animal near/2 therapy:ti,ab,kw||Finds the terms when they are maximum number of words between them is two|
|CINAHL||(TI animal N2 therapy) OR (AB animal N2 therapy)||N2 finds the words if they are within two words of one another regardless of the order in which they appear|
|Web of Science||TOPIC: (animal NEAR/2 therapy)||NEAR/2 maximum 2 words separate the terms|