Skip to main content
It looks like you're using Internet Explorer 11 or older. This website works best with modern browsers such as the latest versions of Chrome, Firefox, Safari, and Edge. If you continue with this browser, you may see unexpected results.

Systematic Review

Resources for conducting a systematic review research.

Developing a Search Strategy

A path covered with well aligned green, yellow and red apples flowing out of a basket

Image: Pinterest

Techniques for search term harvesting

Begin brainstorming search terms by using the following techniques:

  • Explore gold standard articles supplied by principal investigator or found through preliminary searches.
  • Look at search strategies from published systematic reviews.
  • Scan records, articles and searches for usable controlled vocabulary and natural language. Use database tools (e.g. thesauri, index, subject headings) to find controlled vocabulary terms; Oxford English Dictionary to locate word variants or synonyms, Text Mining tools to find Medical Subject Headings (MeSH) terms or "implicit" keywords.

​Look for relevant and/or frequently occurring terms. List all terms in an Excel harvesting form.

Example:

  • cancer[Title/Abstract]
  • neoplasm[MeSH]
  • malignant[Text word]
  • tumor[All fields]

Learn how to build the search with the PubMed Tutorial

_____________________________________________________________

Take control of your search and turn off Pubmed's Automatic Term Mapping (ATM)! It will not include all variant terminology, and automatically explodes MeSH terms. Not using ATM allows for clearer documentation of search method.

For more information on Automatic Term Mapping, watch the video below.

A worldle displaying Text mining as the major word and three workers digging up words and carry them out.

Image by Prakhyath Rai, Assistant Professor, SCEM, Mangaluru

Text Mining Tools

These tools can help you with building your search strategy.

  • PubMed Pub ReMiner - Text mining for PubMed
  • Go PubMed - Text mining tool for PubMed or MeSH terms. This article explains the features of this text mining tool.
  • Yale MeSH Analyzer - Watch this tutorial (7 min.). This tool allows users to enter up to 20 PubMed ID numbers, which it uses to aggregate the metadata from the associated articles into a spreadsheet. For systematic reviews, it is useful in search strategy development to quickly aggregate the Medical Subject Heading (MeSH) terms associated with relevant articles. While it only works for PubMed, it can be useful for developing searches in medical-adjacent fields, such as psychology, nutrition, and animal health. 
  • NCBI MeSH on Demand identifies MeSH® terms in your submitted text (abstract or manuscript). MeSH on Demand also lists PubMed similar articles relevant to your submitted text.
  • SWIFT-Review - Desktop text mining tool specific to systematic reviews. To obtain your free license for SWIFT Review, simply browse to the Sciome Software web page to login and/or create your SWIFT-Review account.  
  • Voyant - General text mining (this is the download). For the web version go to http://voyant-tools.org
  • TerMine - General text mining
  • JSTOR Text Analyzer - Recommends journal articles in JSTOR relevant to text.
  • CREBP-SRA Word Frequency Analyser (WFA) - This tool helps determine which words you should use to construct and refine a search strategy
  • HelioBLAST is a free internet service that can be used to find PubMed records that are similar to a highly relevant record you have already identified. As well as providing a list of 50 best PubMed records and displaying a relevance score, it also provides a list of ‘implicit’ keywords. These keywords can help to identify additional terms and concepts to add to your search strategy. Type in the search box an abstract or a paragraph.
  • Medline Ranker requires a set of known relevant records with PubMed identifiers and a test set of records (e.g. search results from a highly sensitive search). Medline Ranker sorts the records in the test set and presents those that were most similar to the relevant records first. Medline Ranker also provides a list of discriminating terms which discriminate relevant records from non-relevant records.

_________________________________________________________________________

For more information on text mining tools - review and comparison, read the following article:

Paynter, R., Bañez, L. L., Berliner, E., Erinoff, E., Lege-Matsuura, J., Potter, S., & Uhl, S. (2016). EPC methods: an exploration of the use of text-mining software in systematic reviews.

Flying books through a filter and a small pile of books on the other side of the filter

Image modified from www.buzzfeed.com, image #24

Search Filters / Hedges

Search hedges are vetted strategies created by expert searchers to find specific literature.

Keep in mind!

  1. Date of creation of filter
  2. Pace of change in terminology (controlled  and free text)

If you edit a filter, note this in the manuscript.

Example:

“We used a prognosis filter based on that developed by Smith (2015).”

You might limit to a particular publication type in Pubmed. See a full list of Pubmed publication types.

Hedges by Topic (in alphabetical order)

A hand checking a box next to Excellent. Other bullets under the performance heading are good and poor.

Image by Steven Hunt

Evaluating Your Search Strategy

Check your search strategy for any errors using the check list below.

Source:

McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer Review of Electronic Search Strategies: 2015 guideline statementJournal of Clinical  Epidemiology, 75, 40-46. 

Graphics with three overlapping paper sheets

Image by http://www.duluthtownship.org

Documenting Your Search Strategy

Keep track of all your search terms and search strategy that you have used for each database.

You will need this information as a supplemental material to accompany your manuscript.

Systematic Literature Review Worksheet

Use the Database Search Log to record your search terms, search strategy and databases searched.

Guidance on Reporting Systematic Reviews

Cochrane strongly encourages that review authors include a study flow diagram as recommended by the PRISMA statement.

Other checklists include:

  • Equator Network list of guidelines -  Browse for reporting guidelines by study type, clinical area, and section of report. Some examples include:
    • ARRIVE and DSPC for animal studies
    • MOOSE - meta-analysis of observational studies in epidemiology
    • STARLITE - general health policy and clinical practice
    • TIDier-PHP - population health and policy interventions

Examples of documented search methodologies:

  • Full search strategies for all database searches provided in the Appendices:

Bath, P. & Krishnan, K. (2014). Interventions for deliberately altering blood pressure in acute strokeCochrane Database of Systematic Reviews, 10.

  • A summary of sources searched and keywords used in the Sources section:

McIntyre, S, Taitz, D, Keogh, J, Goldsmith, S, Badawi, N & Blair, E. (2013). A systematic review of risk factors for cerebral palsy in children born at term in developed countries. Developmental Medicine & Child Neurology, 55(6), 499-508.

Callout with the word Blog inside over the heads of people

Image by Eternal allegiance

Blogs & Discussion lists

Learn from other experienced searchers and get professional advice from the library community.