Why search in PubMed?
The biggest subset of PubMed is the MEDLINE database of the U.S. National Library of Medicine (NLM). PubMed is also supplemented with references from articles that are not yet indexed for MEDLINE or that fall outside the scope of MEDLINE. In addition, PubMed contains references to e-books published on the NCBI Bookshelf.
- Scopus and Web of Science are ideal for citation search, Embase and PubMed are less focused on this. In Embase the number of citing articles in Scopus is mentioned per article and you can click through to an overview of these articles in Scopus.
- In Embase and PubMed you can conduct a more advanced search by using index terms: synonyms and terms higher or lower in the hierarchy are automatically included in your search. Scopus does contain all the Embase index terms (field code INDEXTERMS) but does not integrate the synonyms and hierarchically higher and lower terms in your search. As a consequence you get considerably fewer search results in Scopus and Web of Science than in Embase and PubMed. For example: if you search on “heart attack” in Scopus, you will miss all the publications that mention “myocardial infarction” or that are indexed with the Emtree term “heart infarction”.
- Subheadings, where the focus is on a certain aspect of a search term, are not included in Scopus or Web of Science. So your search will be less accurate.
How to search in PubMed?
You can search in PubMed with free text (see Automatic Term Mapping and Searching via Tags). You can also search using MeSH terms, the thesaurus keywords (see Searching via MeSH terms). However, a good search action requires a combination of the two (see Searching via free text and MeSH terms).
In PubMed the following Boolean operators can be used to set up your search query:
- AND: all terms should appear. In a standard search query AND is automatically placed between search terms.
- OR: one of the terms should appear.
- NOT: this term should not appear.
Use double quotation marks in PubMed to search for an exact word or word combination, e.g. “cardiovascular disease”. The use of round brackets changes the order in which the AND, OR and NOT combinations are carried out. The latter is also called ‘nesting’.
When entering a search term – without tags and quotation marks – both free text and MeSH terms, keywords from the controlled thesaurus, are automatically searched on.
Via Details – which you can view from the Advanced Search – you can see how PubMed translates your search term(s) via Automatic Term Mapping using operators and tags.
With tags, you yourself indicate the fields in which your search term should appear. The tags or field codes are placed between square brackets and follow the search term, e.g. hypotension[All fields]. The tag [All fields] is automatically added (e.g. via Automatic Term Mapping). Be aware that in relation to this tag, the search term may also appear in the author name or affiliation (but never in the full text). However, if you use the tag [TIAB], you limit the search results to those in which the search terms only appear in Title, Abstract and/or Author Keywords.
An overview of the tags or field codes in PubMed is available here.
In MEDLINE keywords, the so-called MeSH or Medical Subject Headings, are assigned per article.
MeSH terms are searchable via the MeSH-database.
You will find a number of building blocks for your search query on the detail page of a MeSH term:
- Subheading: index term that explains an aspect of a MeSH term.
- Major Topic: MeSH term indexed as main subject.
- Location of the term in the tree structure, with indication of MeSH terms higher and lower in the hierarchy. In a standard search query all underlying terms are included, but here you have the choice to switch off this option if it would be relevant.
TIP: Key articles can provide MeSH terms that are useful for further search queries.
The advantage of searching on MeSH terms has the advantage that the article deals with the subject (regardless of how the author described the subject). Moreover, it also leads you to articles without an abstract. It is recommended to combine this search action with free text, so that articles that have not (yet) been assigned keywords appear in your search query. It may be useful to add the tag [TIAB] to your search term(s) for that second search action, so that the search terms only appear in fields relevant to the content i.e. Title, Abstract and/or Author Keywords.
You can develop a search strategy using the Advanced Search Builder.
Via the Add terms to the query box option, you can add tags of field codes to your search term(s).
In the Query box you can build your search strategy based on different search terms and Boolean operators.
The previously entered search terms are saved under History. You can combine terms from History with (a) new search term(s).
Search results can be refined by filtering in the left column on: Text Availability, Article Attribute, Article Type, Publication Date and Journal Category.
Additional filters or options of a standard filter can be added via ‘Additional Filters’. After adding, you still need to activate the filter via the check box in front of the filter.
Tip: after logging in via MyNCBI, you can see and adjust your personal filters at the top of the left column. For example, you can add a frequently used search block (that applies to your research domain), so you can always activate it while searching in PubMed.
- Use a Truncation operator i.e. asterisk (*) to allow variability for several characters in your search query. Note: this can only be used in PubMed at the end of a word or phrase. For example: the search query creat* will give results for creature, creation, create, creating, creator, etc. Asterisks can in the New PubMed be used in combination with quotation marks!
- In PubMed you can easily consult related articles via Similar Article, on the right side of a detail page of a record.
- And it is also definitely worth searching in Embase, the European equivalent of PubMed. Both PubMed and Embase have a number of unique references on most subjects, since you search the MEDLINE database through both of them. For more information, click here.
- When you search via an author name, enter the surname first and then one or more initials.
- If you are searching for an article and you know the title, copy and paste it into the Single Citation Matcher on the PubMed homepage.
With PubCrawler and MyNCBI you can stay up to date with new articles on your subject. You will need to create an account for both services, but its use is free. It works best if you formulate a specific search query, so that you are not overloaded. The results can be seen on the websites, but can also be sent to your email address. It is easy to create Alerts via MyNCBI while searching in PubMed. In PubCrawler you can also receive updates of articles related to one specific article.
Link with EndNote
If you want to import one or more references to your personal EndNote library, select them in the list of search results.
Then select Send to > Citation manager.
An nbib file is created which you can open when your EndNote is opened. Articles are then automatically imported into your EndNote library.
Tip: when using EndNote via Athena, you also need to have PubMed opened via a browser on Athena to enable direct import.
The maximum number of references you can import to you EndNote library in one step is 10000.
Subsequently click on the Import File button in EndNote. Select the txt file and choose PubMed (NLM) as Import Option.