Skip to Main Content

AAAAReusable content

info for reuse on our pages

Why use a database?

Databases allow you to do different kinds of searches.

Basic keyword search: similar to a library catalogue or Google search where you use whatever keywords you think describe your topic.

Subject heading search:  use the thesaurus terms (also known as controlled vocabulary) of the database to find all articles tagged with specifically assigned subject headings.

Combined keyword & subject heading search:  for advanced research, use combined searching  to ensure retrieval of current and relevant  resources to support your research.  Start with your own basic keywords as concepts but also find out what the thesaurus term in the database is for that concept and combine them. 

search mesh on medline complete

To search within MeSH:

  1. To browse a list of subject headings available in the database, click the MeSH button on the blue sub-toolbar.

        Users can search MeSH in MEDLINE in three different ways:

  • Term Begins WitEnter a term in the Browse for field and choose Term Begins With. An alphabetical list is  displayed.
  • Term Contains - Enter a term in the Browse for field and choose Term Contains. The term searched appears first, followed by terms that contain the term in alphabetical order.
  • Relevancy Ranked - Enter a term in the Browse for field and choose Relevancy Ranked. The exact match for the term appears first in the list, followed by the relevancy ranked list.

   2. Mark the check box tho the left of a term.

   3. Click Search Database. Your search query finds articles with the term as a MeSH Heading.

Specific or Sensitive Searching?

Specific or Sensitive Search. Which one do i need?

The number, type & combination of terms you use in your search will depend on both your question & the objective of your research. A search cannot be both specific & sensitive. Your job as a researcher is to find the balance.

A specific search (also called precise or narrow search)

  • requires a specific answer, is most likely clear & easily described with keywords &/or thesaurus terms.
    • e.g. What is the recommended daily dose of paracetamol for a child with fever?
  •  may be one where you only need a few articles on a specific topic
    • e.g. use of flipped classroom techniques in medical education to stimulate student participation in lectures or you want a recent RCT on the effectiveness of diclofenac for osteoarthritis
A specific search will retrieve very relevant results, but not all relevant papers will be found.

 A sensitive search (also called a broad search)

  • requires researcher to gather all relevant literature & includes the most recent evidence which may not yet have thesaurus term assigned;  
    • e.g. find all studies on the influence of diet on Rheumatoid arthritis from the last 10 years for a Systematic Review you wish to publish
  • it may include concepts that do not have precise thesaurus terms & may require combinations of synonyms to capture all aspects of your research question.
    • e.g. you are doing PhD research & you are worried about missing any relevent studies
A sensitive search is less precise & will retrieve a combination of relavant & less relevant results which you will need to include or exclude according to your question.

Recording your search strategies

Literature searches often produce a large number of citations & require use of several databases.

Good search practice includes keeping a search diary or a document including details of your search strategy to allow others to reproduce your steps & get the same results.

Record your search strategies by logging details on:

  • all database & any other sources you search - eg., Medline, Cochrane Library, Google Scholar, etc.
  • note down journal titles that appear often during your searches to familiarize yourself with journals in your area of interest.
  • which search terms you used (keywords, subject headings, phrases)
  • any search techniques you applied (truncation, proximity, etc.)
  • how you combined your search terms (AND/OR/NOT)
  • the number of search results from each source & each strategy used.

Download the citations you have retrieved to a reference manager &/or excel spreadsheet to deduplicate citations if necessary.

PubMed Basic search

pubmed basic search screen

Advpmsrcbx

Tips to increase Sensitivity

  • Refine  your question or objective
  • Remove a concept from your search    
               Instead of CANCER AND CHEMOTHERAPY AND NAUSEA
         > search CHEMOTHERAPY AND NAUSEA 
  • search more synonyms (keywords) for each concept in your question (find the outliers!)
                Instead of CHEMOTHERAPY AND NAUSEA 
> search (chemotherapy OR alemtuzumab OR cisplatin) AND (nausea OR vomiting OR emesis)
  • search more than 1 database

 

Tips to increase Precision

  • Refine question or objective
  • Add a concept to your search, or combine two concepts into one
             Instead of TYLENOL AND FEVER > search TYLENOL AND FEVER AND RANDOMIZED
             Instead of CHEMOTHERAPY AND NAUSEA
> search ''CHEMOTHERAPY INDUCED NAUSEA AND VOMITING"
  •  Restrict search words to title, title/abstract, or author-provided keywords
  •  Use tools of the database to restrict search results to only those where your concept is the MAJOR SUBJECT of the article
  • Use study filters, date limits, or clinical queries to restrict results of the search
  • Use different sort options to force more relevant or higher cited articles to the top of your results (in some databases)
  • Search a distilled information resource, such as BMJ Best Practice instead of a database

Search Filters for Systematic Reviews

PubMed Search Filter used to find Randomized Controlled Studies: These search strings can be added to your subject & keyword searches to limit your results to sensitive or sensitive and precise results depending on the type of research you are doing.

1. Sensitivity-Maximizing Version (2008 revision) - Maximizes recall. Preferred search for a Systematic Review on trials (exhaustive)

(randomized controlled trial[pt] OR controlled clinical trial[pt] OR randomized[tiab] OR placebo[tiab] OR drug therapy[sh] OR randomly[tiab] OR trial[tiab] OR groups[tiab] NOT (animals [mh] NOT humans [mh]))

 

2. Sensitivity and Precision-Maximizing Version (2008 revision) - Provides a balance of recall and precision. Preferred for a Review or RCTs (not exhaustive)

(randomized controlled trial[pt] OR controlled clinical trial[pt] OR randomized[tiab] OR placebo[tiab] OR clinical trials as topic[mesh:noexp] OR randomly[tiab] OR trial[ti] NOT (animals[mh] NOT humans [mh]))

 

 

Tools for evaluation of studies

Randomized control studies (RCTs)

Study Evaluation Tools