Skip to Main Content

Responsible Literature Searching Guide

Refine and Save Search

Objectives

After reviewing this section content, the reader will:

  • Explain the concepts of precision and recall
  • Apply techniques to increase precision or recall of a completed search
  • Describe basic output options
  • Understand the benefits of setting up a personal account in a database
  • Be able to set up alerts and save searches

Balancing Recall and Precision

Searching is an iterative process. The researcher or clinician determines whether the searched primary, secondary, and tertiary information resources have responsibly addressed the research topic or clinical dilemma. When completing a database search, it is important to achieve a balance between precision (or specificity) and recall (or sensitivity).

  • Precision (or specificity):is "the proportion of citations in a given search that are relevant to the search question”. Precision is important if the intended purpose of the literature search is to get background information on the topic or an overview of the subject area. Retrieving all primary literature is not necessary.
  • Recall (or sensitivity) is "the proportion of relevant citations in a given search among the total number of relevant citations available in the bibliographic database". Recall is important if the intended purpose of the literature search is to locate primary literature to support a grant application, publication, or systematic review or to protect human research subject(s) from adverse events or harm. In this case, identifying all primary literature is necessary.

The following table outlines different search techniques that can increase precision or recall of the literature search.

Too Many Results (increase precision or specificity) Too Few Results (increase recall or sensitivity)
Use major subject terms Use the broad subject terms
Use the "focus" option, if available Use the "explode" option, if available
Select more specific, narrower subject headings Check similar/related articles for additional terms
Use appropriate subheadings for each subject heading Do not use subject heading’s subheadings
Decrease the number of keywords, synonyms Use the database's search fields to identify keywords
Precision can be enhanced by decreasing the use of the Boolean operator "OR". This eliminates synonyms or like concepts. Include alternate spellings, variant endings
Precision can also be increased by adding concepts with the Boolean operator "AND". Increasing recall requires greater use of synonyms and the Boolean operator "OR", as well as a reduction in the use of the Boolean operator "AND".
Searching for precision should result in retrieving many relevant citations and very few irrelevant citations. Be aware that recall will have an inverse effect on the performance of the literature search by also retrieving many irrelevant citations.

(Allison, Kiefe, Weissman, Carter, & Centor, 1999; Jankowski, 2008; McKibbon & Walker-Dilks, 1995; Wessel, 2019)

Search History

The database’s search history feature is essential to keep track of the search strategy. This feature allows the searcher to use Boolean operators to combine – and recombine – search lines and apply limits or filters to refine search results. Additionally, the searcher can use the search history to translate the search strategy to other databases, i.e., from PubMed/Ovid MEDLINE to Ovid Embase.

As expected, since PubMed and Ovid MEDLINE search the same records, there is minimal difference in the number of citations retrieved.

PubMed Search Strategy

Ovid MEDLINE Search Strategy

When translating the search to Embase, notice the different subject headings as well as the number of citations retrieved. It is not unusual for the Embase search to retrieve more results than PubMed/Ovid MEDLINE as the Emtree subject headings are more granular than MeSH.

Embase Search Strategy

(Gray, 2012; U.S. National Library of Medicine, 2019; Wessel, 2019; Wolters Kluwer, 2019)

More Information

See Appendix 3 - Database Comparison Chart to view a chart that summarizes search tips by commonly searched databases.  This tool is helpful when translating a search from one database to another database.

(Elsevier, 2019; Jankowski, 2008; Patrick & Munro, 2004; U.S. National Library of Medicine, 2019; Welch Medical Library, 2019; Wolters Kluwer, 2019)

Output

Most databases have output choices, which can include but not be limited to:

  • Export: The database may allow the searcher to export results in different formats:
    • DOC or DOCX (Microsoft Word)
    • TXT (text file)
    • RIS (standardized format to export results to citation management software, i.e., EndNote)
    • Delimited or XML for database import (Microsoft Excel)
    • PDF
  • Email/Print: The database may allow the searcher to specify:
    • Fields to display
    • Citation format style (i.e., JAMA, APA, MLA)
    • Citation layout (i.e., citation only; citation plus abstract; citation plus abstract and subject headings; customizable, etc.)

Citation management software, such as EndNote (UT Southwestern campus standard), RefWorks, Zotero, Mendeley, etc. allow the user to:

  • Organize citations (results can be saved in folders and shared with others)
  • Attach full-text articles
  • Save time (e.g., deduplicate results, retrieve full-text)
  • Interface with Microsoft Word (build a reference list in the Cite While You Write function)

Refer to the specific database handout for information related to output, saved searches, and alerts.

(Gray, 2012; U.S. National Library of Medicine, 2019; Wessel, 2019; Wolters Kluwer, 2019)

Additional Resources

Saved Searches and Alerts

Many databases permit the searcher, researcher, or clinician to create an account to save the search strategy and/or create an alert to receive scheduled updates. Retrieving and re-running a saved search saves time and allows the searcher, researcher, or clinician to edit and document changes to the search strategy. Depending on the database, the user can schedule the frequency of the alert, report format, number of citations to send, etc.

Refer to the specific database handout for information related to output, saved searches, and alerts.

(Gray, 2012; U.S. National Library of Medicine, 2019; Wessel, 2019; Wolters Kluwer, 2019)

References

Allison, J. J., Kiefe, C. I., Weissman, N. W., Carter, J., & Centor, R. M. (1999). The art and science of searching MEDLINE to answer clinical questions. Finding the right number of articles. Int J Technol Assess Health Care, 15(2), 281-296.

Gray, J. R., Grove, S.K., and Burns, N. (2012). The practice of nursing research: Appraisal, synthesis, and generation of evidence. London: Elsevier Health Sciences.

Jankowski, T. A. (2008). The Medical Library Association Essential Guide to Becoming an Expert Searcher: Proven Techniques, Strategies, and Tips for Finding Health Information. New York: Neal-Schuman Publishers.

McKibbon, K. A., & Walker-Dilks, C. J. (1995). The quality and impact of MEDLINE searches performed by end users. Health Libr Rev, 12(3), 191-200. doi:10.1046/j.1365-2532.1995.1230191.x

U.S. National Library of Medicine. (2019, November 15, 2019). PubMed User Guide. Retrieved from https://pubmed.ncbi.nlm.nih.gov/help/

Wessel, C. B. (2019). Responsible Literature Searching for Research: A Self-Paced Interactive Program. Retrieved from https://cme.hs.pitt.edu/ISER/app/learner/loadModule?moduleId=8381&dev=false

Wolters Kluwer. (2019). Ovid Help. Retrieved from http://site.ovid.com/site/help/documentation/osp/en/index.htm#CSHID=advanced.htm