Evidence Based Medicine

MEDLINE is produced by the National Library of Medicine and is the premiere biomedical database. 

• Coverage: generally 1947 to the present, with some older material.
• Indexes 5,400+ international biomedical journals.
• Includes primarily journal articles, reviews, letters, editorials, comments.
• Does include some books from NCBI's Bookshelf.
• Abstracts are available for most entries.
• Different platforms are available from the National Library of Medicine, commercial vendors, and universities.
• Free MEDLINE access is available through NLM's PubMed.

MeSH is the National Library of Medicine's controlled vocabulary thesaurus. It consists of sets of terms naming descriptors in a hierarchical structure that permits searching at various levels of specificity.  An advantage of controlled vocabularies, like MeSH, is that a single term has been used to represent a particular concept no matter how it was referred to in the original article.

• MEDLINE uses MeSH for indexing database articles.
• 6-15 subject headings are assigned per article, with up to 3 assigned for major emphasis of the article.
• MeSH is updated annually to reflect current terminology usage.
• MeSH allows for more precise subject searching. Articles are indexed to the most specific term available.
• Using MeSH to search MEDLINE allows the searcher to apply subheadings, limits, add major emphasis, and "explode".
• MEDLINE allows users to expand a category of terms from the MeSH TREES (MeSH terms in a hierarchy from general to specific) to retrieve all of the articles on the general term and all of the specific terms listed underneath. Terms are automatically expanded in PubMed unless you change the default.  For Example:
  Cohort Studies would include:  Longitudinal Studies, Follow-up Studies, and Prospective Studies
  Longitudinal Studies would include: Follow-up Studies and Prospective Studies
• MeSH definitions, history, and TREES can be found in the MeSH Database at http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=mesh.
• Subheadings are terms which cover general, frequently discussed aspects of a subject (e.g. diagnosis, etiology, surgery, therapy). Subheadings are combined with MeSH terms to indicate a specific focus. (Example: ASPIRIN/adverse effects)
• Limits are descriptors used for commonly occurring concepts (e.g., human, animal, female, male, infant, child, adult, aged).
• Publication Type is a modifier available for publication types such as Randomized Controlled Trial, Meta-Analysis, and Review.

** Note: Publication Type [PT] is always singular (Practice Guideline) while MeSH terms are always plural (Practice Guidelines). [PT] indicates the report in the article is a Practice Guideline. An article which has been indexed with the MeSH term Practice Guidelines is an article about this methodology or process. Some searchers, in order to be as comprehensive as possible, will use the strategy: Practice Guideline [PT] OR Practice Guidelines [MeSH].

MeSH tutorials are available from the MeSH web page at http://www.ncbi.nlm.nih.gov/mesh.

These sample MEDLINE searches will give you practice using MeSH, applying clinical filters, and putting limits on your search retrieval. Be aware that different MEDLINE search systems have unique functions and commands.

Sample Search on Therapy

What is the evidence for the use of antibiotics for treating otitis media in children?

Sample Search on Diagnosis

Are there any evidence based articles on the differential diagnosis of Alzheimer's Disease?

Sample Search on Prognosis

What is the prognosis for a successful pregnancy in a young woman with Systemic Lupus Erythamatosus?

Sample Search on Clinical Prediction Guides

Are there any evidence based practice guidelines on the use of mammography in screening for breast cancer?

Sample Search on Etiology

Does hormone replacement therapy increase the risk for pulmonary embolism in postmenopausal women?

Medical Subject Headings:

• Clinical Trials
• Confidence Intervals
• Cost-Benefit Analysis
• Decision Making
• Decision Trees
• Diagnosis
• Guidelines
• Odds Ratio
• Outcome Assessment
• Practice Guidelines
• Predictive Value of Tests
• Probability
• Prospective Studies
• Sensitivity and Specificity
• Survival Analysis
• Treatment Outcome

Publication Types:

• Clinical Trial
• Consensus Development Report
• Controlled Clinical Trial
• Guideline
• Meta-Analysis
• Multicenter Study
• Practice Guideline
• Randomized Controlled Trial
• Review

Subheadings:

• adverse effects
• diagnosis
• drug therapy
• etiology
• mortality
• prevention and control
• therapeutic use
• therapy

Try these techniques for limiting your results:

• Use AND with another concept
• Limit to English
• Limit by Age, Sex, Human
• Limit to Publication Type (for example: Randomized Controlled Trial or Review)
• Use Subheadings (for example: Diagnosis or Therapy)
• Use a more specific term (Femoral Neck Fractures instead of Hip Fractures)
• Restrict to Major MeSH Topic
• Choose "Do not include MeSH terms found below this term in the MeSH hierarchy"
• Use Text Word in Title
• Limit to Subsets (e.g., Core Clinical Journals)

Try these techniques for increasing your results:

• Use OR with synonymous/related concepts
• Include all languages (English abstracts are often available)
• Remove all Limits
• Do not restrict to Publication Types
• Choose All Subheadings
• Use a broader term (Eye Diseases instead of Retinal Diseases)
• Do not Restrict to Major MeSH Topic
• Do not choose "Do not include MeSH terms found below this term in the MeSH hierarchy"
• Truncate text words (recommend*)
• Use the "See Related" feature in PubMed

This guide has been created by the University of Illinois at Chicago's Library of the Health Sciences at Peoria. 
Please feel free to contact us with any questions at lib-pref@uic.edu.