Skip to Main Content

Systematic Reviews

Developing a search strategy for your Systematic Review can seem overwhelming. Carefully planning it out and following the steps should make this process a little less daunting.

  1. Start by getting some background information (aka Term Harvesting)
    Identify "gold standard" articles (3-10 minimum) to use to identify some potential search terms and also to double check that your search strategy is locating the articles you know should be found in your results. If these articles are missing you should ensure they meet your inclusion/exclusion criteria and then edit your search strategy to ensure that it is comprehensive enough.

  2. Formulate the research question
    Some of your inclusion and exclusion criteria can be built into your literature search and some are more appropriate to apply during the screening phase. Use limits/filters carefully - make sure have a documented rationale for why you are limited by date, language, etc.
    Remember to select your databases, search for grey literature and include some handsearching as well. See "Databases' and 'Grey Literature and Hand Searching' tabs above for more information.
    • Identify the key concepts
      Start with one database that is appropriate and that you are most comfortable using.
  • Develop search terms by thinking of key words.
  • Develop search terms by checking for controlled vocabulary (e.g. MeSH in PubMed, CINAHL Headings etc).
  • Include phrases, wildcards and truncation.
    • Phrase = "pediatric migraine"
    • wildcard = wom#n includes woman or women
    • truncation=child* includes all forms child, children, childhood

      Pro Tip!
      Yale MESH Analyzer tool extracts MeSH and other citation data from PubMed articles to all you to review terms in a grid for relevant articles to help you develop a search strategy for your review. You can analyze a maximum of 20 articles at a time.

  1. Think about the search fields you want to target for your key words (e.g. in PubMed you can target specifically the title and abstract [Title/Abstract] but consider text word [Text Word] as this includes author assigned key word that title and abstract might not include.
  2. Combine all of key words and controlled vocabulary using Boolean Operators AND (both terms must be included in the results), OR (one or the other or both can be included in the results), NOT (exclude a specific word or concept e.g. common cold NOT cold temperature). This process will take a while. Make sure to look for certain key articles that you have identified being in the search results. 
  3. Consider search limits such as language and dates (use these last after your search strategy is finished).
    Be careful regarding using limits in databases for age group and human (e.g. PubMed)! It is better to search these as a concept. 
  4. Run the search several times and tweak terms as needed.
    Pro tip!
    Run one concept at a time e.g. all key words and controlled vocabulary with OR for concept one, all key words and controlled vocabulary with OR for concept two THEN combine each set of concepts with AND.
  5. Take your final search strategy and run it in the first selected database.
    Make sure to keep track of everything!   
    1. All the terms in the search used.
      Copy the search string or down load a spreadsheet (PubMed offers this)
    2. Any filters used.
    3. Date it was run.
    4. Number of results.
      Pro tip! 
      Carefully documenting your search strategy is a critical part of reporting your systematic review methodology. It must be clearly stated, replicable, and dated.  Best practice is to document the strategy for each database you search even if only one representative example is published in your final manuscript (typically as a supplement).
      Try the Cochrane's Template for a Search Report Form to document your search!
      Pro tip!

      Sign up for a free account in each database and save the search. See our guide for more information. This allows you to run the search again without having to put in all of the search terms again AND you can also sign up for alerts for any new articles.
  6. Run the search in several other databases.
    • The original search will need to be adapted to each databases keeping in mind e.g. PubMed MeSH terms may not be used in a database. Several databases have their own. 
      Pro tip!
      Document, document, document! Keep records of search strategies, dates run, databases used and search strategies. This has been covered earlier in the text but it is so important!

What is Grey Literature?
Cochrane Handbook defines
Grey Literature as ..."understood to be literature not formally published in books or journals. This can include theses or dissertations, conference proceedings, clinical trials registries, white papers, government reports, and more."

What is Handsearching?
Searching for trial reports in databases may not retrieve all relevant available studies. Reports may either be missing from the database or may not have been indexed adequately due to lack of detail in titles and abstracts. Some reports are only ever published as abstracts in conference proceedings books.

Handsearching is the task of searching through medical journals or conference abstract books for reports of controlled trials which are not indexed in the major electronic databases like MEDLINE and Embase. For complete identification of reports of trials, electronic searching may need to be supplemented by conducting page by page searches of a variety of sources including journals and supplements, conference proceedings and abstracts, and correspondence.

Further Searching Advice
Remember to check the reference lists of good articles as well.
Do not forget to check for unpublished studies such as clinical trials. Find clinical trials at


Make sure to include all relevant databases when developing your search strategy. It is important to search several databases and grey literature. Below is a list of databases through Pratt. Feel free to email for advice or fill in the form below.