Do you find too much or too little material? Search techniques using AND OR NOT can help you to develop your search further.
Many databases have their own lists of subject headings that you can use. Subject headings are a form of controlled vocabulary. Searching with controlled vocabulary often provides increased relevance, compared to a free text search. The articles in the databases are tagged with subject headings describing the main focus, which makes it easier for you to find relevant articles.
A free text search, where you search with keywords without using subject headings, becomes a less specific search. The free text search becomes broader and also provides material where your words are mentioned, for example in the abstract, but are not necessarily the essence of the article.
If you are going to conduct a more comprehensive search, for example in a literature study, you may need to combine searching with subject headings with a free text search.
Each database has their own list of controlled vocabulary, that is valid within that database. Examples of that are MeSH (Medical Subject Headings) in PubMed, Emtree in Embase and CINAHL Subject Headings in CINAHL. You can select headings and create your search directly in the database using the subject heading lists.
There may be different words for breast cancer in different articles, for example breast tumor or mammary cancer, but there is a subject heading that every article is tagged with. In this case, breast neoplasms is the MeSH term for breast cancer. If you search with it, you will find all the articles in the database that are relevant in relation to breast cancer no matter what synonym the authors used. At the same time, articles where breast cancer is mentioned but is no significant part of the article are cleared from the result list.
All databases contain filters that can help you get even more relevant results. There are filters for age groups, article or study designs, publication date, language etc.
Video (5:17) How to search with Subject Headings combined with text words in the CINAHL database.
Library and ICT, Faculty of Medicine, Lund University.
With the building block search strategy you do a search in several steps. You start by dividing your question into blocks according to which search terms belong together, and search separately for each set of words. Which patient group or disease? Which therapy or intervention? Which effect or outcome do you want to know more about? Then you combine your searches. The video shows an example on a building block search.
NB! If you are doing a search for a systematic review, you should also combine the subject headings with free text words (keywords/natural language) since you have to make a search with a broad scope. The newest publications might not have subject headings and only searching with subject headings can also be a to narrow search, since it only searches the subjects field, not for example the title and abstract.