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.

PubMed: Tips & Tricks: Home

Background

PubMed is a freely available public interface to more than 32 million citations and abstracts that cover life science and medical journals.  While PubMed does not include full-text articles, Penn's PubMed Plus, does include links to Penn subscribed as well as freely accessible journals.  The great majority of citations found using PubMed come from the Medline database produced by the National Library of Medicine.

In addition to bibliographic information (author/title/journal), citations from Medline include Medical Subject Headings, a controlled vocabulary used to index concepts discussed in the article.

https://pubmed.ncbi.nlm.nih.gov/about/

Navigation and Simple Searching

The PubMed homepage offers a single search box.  Enter words or phrases in the box.  Pubmed will search for the words you type in and additional terms 'mapped' to your topic. Use the Advanced search option to review the search strategy created.  Select Advanced below the search box, find your search in the History area, and then select Details.

  • Results, by default, appear in 'summary' format and in 'best match' order. You can change these settings by selecting 'Display Options'.
  • Click on an article title to display the 'full' record.  The PennText button will appear to the right.  PennText will guide you to full-text holdings as well as request options if the Library does not have full-text access to the article.
  • Use the Filter options that appear on the Results page to reduce your results. Many important filters are hidden so click the "Additional filter" button to review options.

"Basic Searching" Video
"PennText Service" Video

Using Medical Subject Headings (MeSH)

MeSH terms are added to Medline citations by indexers at the National Library of Medicine.  Using these terms can improve search results. Because MeSH terms are hierarchical in nature, PubMed allows you to 'explode' a term which will retrieve articles indexed to more specific MeSH terms.  For example, searching for the MeSH term "Neoplasms" will automatically retrieve articles indexed to any type of neoplasm.

Finding MeSH terms in the MeSH database

  • Do a quick search and then review the MeSH terms attached to relevant citations.  They will appear in the 'full' display.  Left-click on a MeSH term and then select "Search in MeSH" to be taken to the subject heading in the MeSH database.
  • Go directly to the MeSH database using the link found on the PubMed homepage.

In addition to selecting a MeSH term to search. You can attach "subheadings" and also specify that a MeSH term must be one of the "major topics" discussed in the article.

Use the "PubMed Search Builder" to construct a search strategy and then send it back to PubMed for searching.

MeSH Database
MeSH Tutorial

Creating a Personal Account

Creating a personal account in PubMed offers a number of benefits.  As long as you are logged into your account, you can:

  • permanently save a search strategy.
  • set-up a search that will automatically be executed monthly/weekly and then send you the latest results via email.
  • save collections of citations that can be shared with others.
  • set preferences to highlight the terms you searched.
  • PubMed NO LONGER automatically saves searches you have executed in the past six months.  If you wish to save a search strategy, you must use the "create alert" feature.

Create a Personal Account Video

Truncation, Phrase, & Field Searching

While truncation can increase your results, Phrase and Field searching can help limit your results.  Rather than finding a word anywhere in a record, you can specify the field where the word should appear. 
Warning: the use of field, phrase, or truncation commands will stop PubMed from 'mapping' your search term to other appropriate terms.

  • migraine[ti] - term must appear in the article title field
  • migraine[tiab] - term must appear in either the article title or abstract field
  • asthm*[tw] - will retrieve citations that include the word asthma or asthmatic in the 'text word' fields
  • "health policy" - forces PubMed to search for the two words next to each other.  (By default, PubMed searches for two words anywhere in a record.  They do not need to be next to each other.)
  • health policy[ti] - adding a field tag will also search for two words as a phrase

Additional Field Tags

Clinical Queries Searching

If attempting to answer a clinical question, consider using this PubMed feature to retrieve better evidence articles. By adding your search terms and then selecting the appropriate clinical study category, PubMed will use built-in search filters to restrict retrieval.

Clinical Queries Access
Clinical Queries Video

What do I do?

If too many citations retrieved

  • Use the filters to limit time period, type of article, language, age group, etc.
  • Add another concept to your search.
  • Consider using MeSH terms in your search.  Select subheadings if appropriate. Specify that the MeSH term must be the main topic of the article.
  • Is it a clinical question?  Consider using the Clinical Queries feature.
  • Use the field label to limit some terms to appear in the title of the article.
  • Have you used the correct syntax for the PubMed interface?  Are your parentheses in the right place? Check with a librarian.

If none, or too few citations retrieved

  • Remove a concept to simplify your search strategy.
  • Consider using a broader search term.
  • Try adding synonyms to your search.
  • If you have limited a term to a single field, allow the term to be searched in any field.
  • If you find a relevant citation, use the Related Citations feature found in the 'full' display.
  • Have you used the correct syntax for the PubMed interface?  Are your parentheses in the right place? Check with a librarian.
  • Consider searching a different database (Embase or Scopus).